Stereotactic Radiosurgery Technology

Central Nervous
System Metastases
Diagnosis and Treatment
Rohan Ramakrishna
Rajiv S. Magge
Ali A. Baaj
Jonathan P.S. Knisely
Editors
123
Central Nervous System Metastases
Rohan Ramakrishna • Rajiv S. Magge
Ali A. Baaj • Jonathan P.S. Knisely
Editors
Central Nervous System

Metastases
Diagnosis and Treatment
Editors
Rohan Ramakrishna Rajiv S. Magge
Department of Neurological Surgery Department of Neurology
NewYork-Presbyterian Hospital/ NewYork-Presbyterian Hospital/
Weill Cornell Medicine Weill Cornell Medicine
Director - Brain Metastases Program New York, NY
New York, NY USA
USA
Ali A. Baaj Department of Radiation Oncology
Department of Neurological Surgery NewYork-Presbyterian Hospital/
NewYork-Presbyterian Hospital/ Weill Cornell Medicine
Weill Cornell Medicine New York, NY
New York, NY USA
USA
ISBN 978-3-030-42957-7 ISBN 978-3-030-42958-4 (eBook)

https://doi.org/10.1007/978-3-030-42958-4
© Springer Nature Switzerland AG 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, expressed or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To Priyanka, Surya, and Samira – you make every day better.
To my parents – thank you for your example.
To my patients – thank you for the daily inspiration.
Rohan Ramakrishna
My heartfelt gratitude goes to my family for their support and

of course to our patients who continually demonstrate how to
approach adversity with kindness, dignity, and determination.
Rajiv S. Magge
To my beautiful daughter, Hannah, who has filled our hearts

with endless joy.
Ali A. Baaj
To my parents, Samuel Emerson and Kristine Sandberg Knisely,

and to my wife, Mary Jean Hu. You always supported me as I
engaged myself with problems that I found particularly
challenging, and for that, I am grateful beyond where language
can take me.
Foreword
The editors and authors are to be congratulated for organizing this compre-
hensive, multidisciplinary text reviewing the contemporary management of
brain and spinal metastasis. They draw upon the breadth of experience from
an international group of authors while still integrating epidemiology and
basic biology with diagnosis, evaluation, and treatment. All aspects of meta-
static disease including local control, leptomeningeal disease, paraneoplastic
syndromes, and neurocognitive implications are reviewed. Beyond standard
care, alternative and complimentary therapies are also reviewed in relation to
their impact on quality-of-life issues, neurocognition, and pain control.
Advances in our biological understanding of radiation, new technologies,
and combinational therapies have transformed our approach toward central
nervous system metastasis. Recent data supporting the synergistic effects of
radiation therapy and immunotherapy has modified the treatment paradigms
used. This text provides a framework for understanding the biology of radia-
tion therapy as it relates to the multiple technological choices available to the
treating physician. The authors also provide a detailed perspective of the spe-
cific advantages and disadvantages of the multiple radiation therapies avail-
able. New technologies are reviewed from the perspective of maximizing
efficacy and minimizing toxicity, independently and as combinatorial ther-
apy. The newest advances in radiation therapy which maximize the tumori-
cidal effect while minimizing neurocognitive decline are clearly reviewed in
concise prose.
Management of metastatic disease requires the integrated efforts of sur-
geons, oncologists, neuro-oncologists, neurologists, radiation oncologists,
pathologists, precision medicine, imaging specialists, neuropsychologists,
neuro-psycho-pharmacologists, and social workers. A complex disease man-
agement team applying the newest medical, surgical, and technological tools
to alleviate this devastating oncologic process is required to work in conjunc-
tion with cognitive experts in order to maximize the quality of life and main-
tain family stability. This book does an excellent job of tying all of these
facets together in order to give the reader a clear and concise treatment para-
digm. The complexity in managing metastatic disease is not only biological
but also psychosocial. The authors emphasize the importance of integrating
all aspects of cancer care for our patients.
vii
viii Foreword
I strongly recommend Central Nervous System Metastases: Diagnosis and

Treatment to all practitioners in this arena. Being the first text of its kind, it is
really a “must have” for the serious student and caregiver. Students, nurses,
physicians, psychologists, physicians, psychologists, social workers, and
psychiatrists will all find value in digesting components of this book.
Management of metastatic tumors requires both scientific and emotional
tools emphasized by the authors.
Philip E. Stieg, PhD, MD
NewYork-Presbyterian Hospital/
Weill Cornell Brain and Spine Center
New York, NY, USA
Preface
We live in an interesting time as we treat patients with central nervous system

metastases. Though systemic therapies have become more successful and
diverse, the number of patients with central nervous system metastases is
growing. Furthermore, patients are living longer with CNS metastatic dis-
ease. As such, it is imperative in the modern era that patients with CNS
metastases are treated in a holistic, multidisciplinary fashion. Our goals as
treating physicians can no longer be singularly focused on the local control of
a CNS lesion. Rather, we must consider not only oncologic control but also
quality of life, the interaction of our treatments with systemic therapies, pain
control, and much more.
With that in mind, we have endeavored to produce a multi-specialty book
on the diagnosis, evaluation, and treatment of CNS metastases of the brain
and spine. Our authors span the globe and are noted experts in their fields. We
have chosen authors who bring unique perspectives to the field of CNS metas-
tases and hope that you find their contributions both educational and useful.
As you will see, we have designed this book to cover what we consider
essential contemporary topics in CNS metastases care. Truly, a book like this
would have not have been particularly interesting even 10 years ago given the
subsequent advancements in systemic and targeted therapy, radiation therapy,
and surgical therapy. In general, we have begun each section with chapters
covering the fundamental biology of disease so that subsequent chapters on
imaging, diagnosis, and treatment can be properly contextualized.
This book represents a herculean effort made possible by many individu-
als. I would like to thank all of our authors who have selflessly contributed
their time and knowledge. Their expertise and perspective have proven
invaluable. A huge thanks is also due to Philip Stieg at Weill Cornell Medicine
who encouraged me to pursue this project and the development of a brain
metastases program with gusto. Other mentors have been similarly influential
including Richard Ellenbogen (University of Washington), Raymond Sawaya
(MD Anderson), and Frederick Lang (MD Anderson), and I owe them all a
debt of gratitude as well. In addition, I would like to thank Weill Cornell
Medicine and New York Presbyterian Hospital – these institutions make
coming to work a pleasure each day. I would like to also thank Springer,
Richard Hruska, and Connie Walsh who have provided invaluable editorial
assistance and author support. Finally, I would like to thank my coeditors
Rajiv S. Magge, MD; Ali A. Baaj, MD; and Jonathan P.S. Knisely, MD, who
have been instrumental partners in the completion of this book.
ix
x Preface
Please contact me directly (ror9068@med.cornell.edu) should you have

feedback or constructive criticism on how to improve this text. We hope this
text helps you take care of your patients with CNS metastases.
New York, NY, USA Rohan Ramakrishna, MD

Contents
Part I Fundamentals
1 Epidemiology and Socioeconomic Impact of
CNS Metastases�� 3
Jessica A. Wilcox and Lisa M. DeAngelis
2 Basic Biology of Brain Metastasis�� 19
Monika Vishnoi, Robert A. Scranton, Samuel K. Asante,
and Robert C. Rostomily
3 Preclinical Models of Brain Metastasis�� 37
Lucía Zhu and Manuel Valiente
4 Pathology of Brain Metastases�� 53
David J. Pisapia
5 Role of Precision Medicine in Patients with
CNS Metastasis�� 69
Albert Eusik Kim and Priscilla K. Brastianos
6 Classification of Brain Metastases�� 83
Paul W. Sperduto
7 The Role of Advanced Imaging in the
Management of Brain Metastases�� 95
Eaton Lin and Gloria C. Chiang
Part II Evaluation of CNS Metastasis

8 Clinical Presentation of Central Nervous
System Metastases�� 117
Laura E. Donovan and Rajiv S. Magge
9 Management of Seizures in Brain Metastases�� 125
Ankush Bhatia and Edward K. Avila
10 Cerebrovascular Complications in Patients with Cancer�� 139
Jaclyn E. Burch and Alan Z. Segal
11 Mood Disorders in Patients with CNS Metastases�� 151
Kaleena Chilcote
xi
xii Contents
12 Leptomeningeal Disease and the Role of

Intrathecal Therapy�� 169
Fadi Saadeh and Adrienne Boire
13 Paraneoplastic Neurological Disorders�� 187
Monica Weaver Buckley and John C. Probasco
14 Systemic Therapy of Brain Metastases: Lung Cancer �� 207
Adam Lauko, Vyshak Alva Venur, and Manmeet S. Ahluwalia
15 Systemic Therapy of Brain Metastases: Breast Cancer �� 219
Leigh Klaus Swartz and Aki Morikawa
16 Systemic Therapy for Brain Metastases: Melanoma�� 235
Sarah Weiss and Harriet Kluger
17 Systemic Therapy for Brain Metastases in
Other Primary Cancers (Genitourinary,
Gastrointestinal, Gynecology, Head/Neck)�� 245
Karishma M. Parikh and Rajiv S. Magge
18 Management of Solid Tumor CNS Metastases in Children �� 259
Whitney E. Parker, Shahiba Q. Ogilvie, Lily McLaughlin,
and Mark M. Souweidane
Part III Radiation Therapy of CNS Metastasis

19 Basic Radiobiology and Radiation Physics Primer�� 271
Emily S. Lebow, Marc R. Bussière, and Helen A. Shih
20 Role of Whole-Brain Radiotherapy�� 281
Connor Lynch, Jeffrey P. Gross, and Vinai Gondi
21 Stereotactic Radiosurgery Technology�� 299
Diana A. R. Julie and Jonathan P.S. Knisely
22 Stereotactic Radiosurgery: Indications and Outcomes
in Central Nervous System and Skull Base Metastases�� 315
Henry Jeison Ruiz-Garcia, Daniel M. Trifiletti,
and Jason P. Sheehan
23 Hypofractionated Stereotactic Radiosurgery
(HF-SRS) in the Treatment of Brain Metastases�� 329
Jordan A. Torok, Scott R. Floyd, Peter E. Fecci,
and John P. Kirkpatrick
24 Challenges and Controversies in Stereotactic
Radiosurgery�� 343
Jugal K. Shah and Douglas Kondziolka
25 Synergy of Immunotherapy and Radiosurgery�� 355
Andrew G. Brandmaier, Rohan Ramakrishna,
and Silvia C. Formenti
Contents xiii
26 Salvage Irradiation for Patients with Recurrent

Brain Metastases�� 371
Christian Iorio-Morin, Laurence Masson-Côté,
and David Mathieu
27 Applications of Stereotactic Radiosurgery for
Akshay V. Save, Dominique M. O. Higgins, Mark D. Mayeda,
and Tony J. C. Wang
28 Radiation Necrosis Following the Radiosurgical
Treatment of Brain Metastases�� 393
Stephanie M. Robert and Veronica L. Chiang
29 Neurocognitive Effects of Brain Metastases and
Their Treatment �� 407
Karine A. Al Feghali, Mariana E. Bradshaw, Caroline Chung,
and Jeffrey S. Wefel
Part IV Surgical Treatment of Brain Metastasis

30 The Role of Surgery in the Management of
Jeffrey I. Traylor, Aditya Srivatsan, Rajan Patel,
Rohan Ramakrishna, and Ganesh Rao
31 Intraoperative Brachytherapy for Resected
A. Gabriella Wernicke, Sean S. Mahase,
and Theodore H. Schwartz
32 Laser Interstitial Thermal Therapy for
Brain Metastases and Radiation Necrosis �� 457
Jeffrey I. Traylor, Ahmed Habib, Vittorio Stumpo,
Dhiego Chaves de Almeida Bastos, and Sujit S. Prabhu
33 Preventing Cranial Wound Complications in
Cancer Patients�� 473
James C. Lee, Jimmy Xia, Rohan Ramakrishna,
and David M. Otterburn
Part V Spinal Metastases: Foundations

34 Introduction to Spinal Metastases�� 487
Ibrahim Hussain, Brenton H. Pennicooke, and Ali A. Baaj
35 Epidemiology of Spinal Metastatic Disease �� 495
John Berry-Candelario, Mark H. Bilsky, Ilya Laufer,
C. Rory Goodwin, and Ori Barzilai
36 Bone Metabolism in Cancer�� 503
Tilman D. Rachner, Lorenz C. Hofbauer, and Andy Göbel
xiv Contents
37 Systemic Therapies for Patients with Metastatic

Spinal Disease �� 513
Panagiotis J. Vlachostergios and Ashish Saxena
38 The Role of Advanced Imaging in Spinal Metastases�� 523
Sasan Karimi, Nicholas S. Cho, Kyung K. Peck,
and Andrei I. Holodny
39 Decision-Making Algorithms for Surgical
Treatment of Spine Metastatic Disease�� 535
Brenton H. Pennicooke, Ibrahim Hussain, and Ali A. Baaj
Part VI Surgical Treatment of Spinal Metastases

40 Biomechanics of Spinal Fixation in Metastatic Disease�� 543
Allen L. Ho and Atman M. Desai
41 Separation Surgery for Spinal Metastases�� 559
Robert J. Rothrock, Ori Barzilai, Ilya Laufer,
and Mark H. Bilsky
42 Vertebrectomy for Spinal Metastases�� 569
Samuel Kalb and Juan S. Uribe
43 Minimally Invasive Surgery for Spinal Metastases�� 575
Robert J. Rothrock, Ori Barzilai, Mark H. Bilsky,
and Ilya Laufer
44 Deformity Secondary to Vertebral Body Metastases�� 583
Zach Pennington, A. Karim Ahmed, and Daniel M. Sciubba
45 Postoperative Complications and Spinal Metastases�� 603
Bushra Yasin and Michael S. Virk
46 Vertebral Augmentation Procedures for
Treatment of Pathologic Vertebral Body Fractures�� 613
Justin Schwarz, Alejandro Santillan, Adham Mushtak, and
Athos Patsalides
47 Spinal Laser Interstitial Thermal Therapy for
Metastatic Tumors �� 623
Linton T. Evans, Rafael A. Vega, and Claudio E. Tatsui
48 Optimizing Wound Healing in Metastatic
Spine Surgery �� 635
Jaime L. Bernstein, Matthew A. Wright, and Jason A. Spector
49 Contemporary Radiation for Spinal Metastasis and
Spinal Cord Compression �� 643
John Roberson, Bernard Newman, and Samuel Ryu
50 Intraoperative Radiation for Spinal Metastatic Disease�� 659
Brandon S. Imber, Michael R. Folkert, and Yoshiya Yamada
Contents xv
Part VII Pain
51 Approach to Pain in Patients with Central Nervous System
Metastases�� 673
Thomas Chai, Jennifer Erian, Mihir Joshi, Larry C. Driver,
and Dhanalakshmi Koyyalagunta
52 Interventions for Refractory Pain in Cancer Patients�� 687
Michael G. Kaplitt
53 Complementary and Integrative Therapies (CIM) in
Patients with CNS Metastasis �� 695
Santhosshi Narayanan, Wenli Liu, and Gabriel Lopez
54 The Palliative Care of Patients with Brain Metastases�� 705
Rebecca A. Harrison and Eduardo Bruera
Index�� 719
Contributors
Manmeet S. Ahluwalia, MD Burkhardt Brain Tumor and Neuro-Oncology

Center, Cleveland Clinic, Cleveland, OH, USA
A. Karim Ahmed, BS Department of Neurosurgery, Johns Hopkins
Hospital, Baltimore, MD, USA
Karine A. Al Feghali, MD, MSc Department of Radiation Oncology, The
University of Texas MD Anderson Cancer Center, Houston, TX, USA
Samuel K. Asante, MD Department of Neurosurgery, Houston Methodist
Hospital, Houston, TX, USA
Edward K. Avila, DO Department of Neurology, Memorial Sloan Kettering
Cancer Center, New York, NY, USA
Ali A. Baaj, MD Department of Neurological Surgery, Weill Cornell
Medical College/New York Presbyterian Hospital, New York, NY, USA
Department of Neurological Surgery, Weill Cornell Medical Center,
New York, NY, USA
Ori Barzilai, MD Department of Neurosurgery, Memorial Sloan Kettering
Jaime L. Bernstein, MD, MS Division of Plastic Surgery, Department of
Surgery, NewYork-Presbyterian/Weill Cornell Medicine, New York, NY,
USA
John Berry-Candelario, MD, MPH Department of Neurosurgery,
Memorial Sloan Kettering Cancer Center, New York, NY, USA
Ankush Bhatia, MD, MS Department of Neurology, Memorial Sloan
Kettering Cancer Center, New York, NY, USA
Mark H. Bilsky, MD Department of Neurosurgery, Memorial Sloan
Adrienne Boire, MD, PhD Human Oncogenesis and Pathology Program,
Department of Neurology, Memorial Sloan Kettering Cancer Center, New
York, NY, USA
xvii
xviii Contributors
Mariana E. Bradshaw, PhD, ABPP Section of Neuropsychology,

Department of Neuro-Oncology, The University of Texas MD Anderson
Cancer Center, Houston, TX, USA
Andrew G. Brandmaier, MD, PhD Department of Radiation Oncology,
NewYork-Presbyterian Hospital, Weill Cornell Medicine, New York, NY,
USA
Priscilla K. Brastianos, MD CNS Metastasis Center, Department of
Medicine, Massachusetts General Hospital, Harvard Medical School, Boston,
MA, USA
Eduardo Bruera, MD Department of Palliative, Rehabilitation, and
Integrative Medicine, The University of Texas MD Anderson Cancer Center,
Houston, TX, USA
Monica Weaver Buckley, MD, PhD Department of Neurology, Johns
Hopkins University School of Medicine, Baltimore, MD, USA
Jaclyn E. Burch, MD Department of Neurology, University of Rochester
Medical Center, Rochester, NY, USA
Marc R. Bussière, MSc Department of Radiation Oncology, Massachusetts
General Hospital, Boston, MA, USA
Thomas Chai, MD Department of Pain Medicine, UT MD Anderson Cancer
Center, Houston, TX, USA
Gloria C. Chiang, MD Department of Radiology, Division of
Neuroradiology, NewYork-Presbyterian/Weill Cornell Medicine, New York,
NY, USA
Veronica L. Chiang, MD Department of Neurosurgery, Yale University
School of Medicine, New Haven, CT, USA
Kaleena Chilcote, MD Psychosocial Oncology, Department of Behavioral
Medicine and Psychiatry, West Virginia University Cancer Institute,
Morgantown, WV, USA
Nicholas S. Cho, BS Department of Radiology, Memorial Sloan Kettering
Medical Scientist Training Program, David Geffen School of Medicine at
UCLA, Los Angeles, CA, USA
Caroline Chung, MD, MSc, FRCPC, CIP Department of Radiation
Oncology, The University of Texas MD Anderson Cancer Center, Houston,
TX, USA
Dhiego Chaves de Almeida Bastos, MD, MSc Department of Neurosurgery,
The University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
Lisa M. DeAngelis, MD Department of Neurology, Memorial Sloan
Contributors xix
Atman M. Desai, MD, FAANS Department of Neurosurgery, Stanford

University School of Medicine, Stanford, CA, USA
Laura E. Donovan, MD Department of Neurology, Columbia University
Irving Medical Center, New York, NY, USA
Larry C. Driver, MD Department of Pain Medicine, UT MD Anderson
Jennifer Erian, MD Department of Pain Medicine, UT MD Anderson
Linton T. Evans, MD Department of Neurosurgery, Dartmouth-Hitchcock
Medical Center, Lebanon, NH, USA
Peter E. Fecci, MD, PhD Center for CNS Metastasis, Department of
Neurosurgery, Duke Cancer Institute, Durham, NC, USA
Scott R. Floyd, MD, PhD Department of Radiation Oncology, Duke Cancer
Institute, Durham, NC, USA
Michael R. Folkert, MD, PhD Department of Radiation Oncology,
University of Texas Southwestern Medical Center, Dallas, TX, USA
Silvia C. Formenti, MD Department of Radiation Oncology, NewYork-
Presbyterian Hospital, Weill Cornell Medicine, New York, NY, USA
Andy Göbel, PhD Department of Medicine III, Division for Endocrinology,
Diabetes and Bone Disorders; University Hospital “Carl Gustav Carus”,
Dresden, Germany
Vinai Gondi, MD Brain and Spine Tumor Center, Department of Radiation
Oncology, Northwestern Medicine Cancer Center Warrenville, Warrenville,
IL, USA
Northwestern Medicine Chicago Proton Center, Warrenville, IL, USA
C. Rory Goodwin, MD, PhD Department of Neurosurgery, Duke University
Medical Center, Durham, NC, USA
Jeffrey P. Gross, MD, MS Department of Radiation Oncology, Northwestern
University, Robert H. Lurie Comprehensive Cancer Center, Chicago, IL,
USA
Ahmed Habib, MD Department of Neurosurgery, The University of Texas
M.D. Anderson Cancer Center, Houston, TX, USA
Rebecca A. Harrison, MD Department of Neuro-Oncology, The University
of Texas MD Anderson Cancer Center, Houston, TX, USA
Dominique M. O. Higgins, MD, PhD Department of Neurological Surgery,
Columbia University Irving Medical Center, New York, NY, USA
Allen L. Ho, MD Department of Neurosurgery, Stanford University School
of Medicine, Stanford, CA, USA
xx Contributors
Lorenz C. Hofbauer, MD Department of Medicine III, Division for

Endocrinology, Diabetes and Bone Disorders, University Hospital “Carl
Gustav Carus”, Dresden, Germany
Andrei I. Holodny, MD Department of Radiology, Memorial Sloan
Ibrahim Hussain, MD Department of Neurological Surgery,
Weill Cornell Medical College/New York Presbyterian Hospital, New York,
NY, USA
New York, NY, USA
Brandon S. Imber, MD Department of Radiation Oncology, Memorial
Sloan Kettering Cancer Center, New York, NY, USA
Christian Iorio-Morin, MD, PhD Division of Neurosurgery, Department
of Surgery, Université de Sherbrooke, Centre de Recherche du Centre
Hospitalier Universitaire de Sherbrooke, Sherbrooke, QC, Canada
Mihir Joshi, MD Department of Physical Medicine & Rehabilitation,
Baylor College of Medicine, Houston, TX, USA
Diana A. R. Julie, MD, MPH Department of Radiation Oncology,
NewYork-Presbyterian/Weill Cornell Medicine, New York, NY, USA
Samuel Kalb, MD Department of Neurological Surgery, Barrow
Neurological Institute, Phoenix, AZ, USA
Michael G. Kaplitt, MD, PhD, FACS Department of Neurological Surgery,
Sasan Karimi, MD Department of Radiology, Memorial Sloan Kettering
Albert Eusik Kim, MD Department of Neurology, Medical Oncology,
Massachusetts General Hospital, Boston, MA, USA
John P. Kirkpatrick, MD, PhD Center for CNS Metastasis, Departments of
Radiation Oncology and Neurosurgery, Duke Cancer Institute, Durham, NC,
USA
Harriet Kluger, MD Medicine (Medical Oncology), Yale University, New
Haven, CT, USA
Jonathan P.S. Knisely, MD, FASTRO Department of Radiation Oncology,
Douglas Kondziolka, MD, MSc, FRCSC Department of Neurosurgery,
New York University Langone Health, New York, NY, USA
Dhanalakshmi Koyyalagunta, MD Department of Pain Medicine, UT MD
Anderson Cancer Center, Houston, TX, USA
Contributors xxi
Ilya Laufer, MD Department of Neurosurgery, Memorial Sloan Kettering

Adam Lauko, MS Burkhardt Brain Tumor and Neuro-Oncology Center,
Cleveland Clinic Foundation, Cleveland, OH, USA
James C. Lee, MD Department of Plastic Surgery, NewYork-Presbyterian/
Weill Cornell Medicine, New York, NY, USA
Eaton Lin, MD, MEd Department of Radiology, Division of Neuroradiology,
Wenli Liu, MD Palliative, Rehabilitation & Integrative Medicine, The
Gabriel Lopez, MD Palliative, Rehabilitation & Integrative Medicine, The
Connor Lynch, BA Northwestern University Feinberg School of Medicine,
Chicago, IL, USA
Rajiv S. Magge, MD Department of Neurology, NewYork-Presbyterian/
Sean S. Mahase, MD Department of Radiation Oncology, NewYork-
Presbyterian/Weill Cornell Medicine, New York, NY, USA
Laurence Masson-Côté, MD, FRCPC Nuclear Medicine and Radiobiology,
Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, QC, Canada
David Mathieu, MD, FRCSC Division of Neurosurgery, Department of
Surgery, Université de Sherbrooke, Centre de Recherche du Centre Hospitalier
Universitaire de Sherbrooke, Sherbrooke, QC, Canada
Mark D. Mayeda, MD Department of Radiation Oncology, Columbia
University Irving Medical Center, New York, NY, USA
Lily McLaughlin, BS Department of Neurosurgery, Memorial Sloan-
Aki Morikawa, MD, PhD Division of Hematology/Oncology, Department
of Internal Medicine, University of Michigan, Ann Arbor, MI, USA
Adham Mushtak, MD Weill Cornell Medical College in Qatar, Ar-Rayyan,
Qatar
Santhosshi Narayanan, MD Palliative, Rehabilitation & Integrative
Medicine, The University of Texas MD Anderson Cancer Center, Houston,
TX, USA
Bernard Newman, MD Department of Radiation Oncology, Stony Brook
University Hospital, Stony Brook, NY, USA
Shahiba Q. Ogilvie, MPH Department of Neurosurgery, Memorial Sloan-
xxii Contributors
David M. Otterburn, MD Department of Plastic Surgery, NewYork-

Karishma M. Parikh, MD Department of Neurology, NewYork-
Whitney E. Parker, MD, PhD Department of Neurosurgery, New York-
Department of Neurosurgery, Memorial Sloan-Kettering Cancer Center,
New York, NY, USA
Rajan Patel, BS Department of Neurosurgery, The University of Texas
Athos Patsalides, MPH, MD Department of Neurological Surgery,
Kyung K. Peck, MSc, PhD Department of Medical Physics and Radiology,
Memorial Sloan Kettering Cancer Center, New York, NY, USA
Brenton H. Pennicooke, MD, MS Department of Neurological Surgery,
Weill Cornell Medical College/New York Presbyterian Hospital, New York,
NY, USA
New York, NY, USA
Zach Pennington, BS Department of Neurosurgery, Johns Hopkins
Hospital, Baltimore, MD, USA
David J. Pisapia, MD Department of Pathology and Laboratory Medicine,
Sujit S. Prabhu, MD Department of Neurosurgery, The University of Texas
John C. Probasco, MD Department of Neurology, Johns Hopkins University
School of Medicine, Baltimore, MD, USA
Tilman D. Rachner, MD Department of Medicine III, Division for
Endocrinology, Diabetes and Bone Disorders, University Hospital “Carl
Gustav Carus”, Dresden, Germany
Rohan Ramakrishna, MD Department of Neurological Surgery, NewYork-
Ganesh Rao, MD Department of Neurosurgery, The University of Texas
John Roberson, MD Department of Radiation Oncology, Stony Brook
Stephanie M. Robert, MD, PhD Department of Neurosurgery, Yale
University School of Medicine, New Haven, CT, USA
Contributors xxiii
Robert C. Rostomily, MD Department of Neurosurgery, Houston Methodist

Research Institute, Houston, TX, USA
Robert J. Rothrock, MD Department of Neurosurgery, Memorial Sloan
Henry Jeison Ruiz-Garcia, MD Department of Radiation Oncology, Mayo
Clinic, Jacksonville, FL, USA
Samuel Ryu, MD Department of Radiation Oncology, Stony Brook
Fadi Saadeh, MD Department of Neurology, Memorial Sloan Kettering
Alejandro Santillan, MD Department of Neurological Surgery, NewYork-
Akshay V. Save, BS Department of Neurological Surgery, Columbia
Ashish Saxena, MD, PhD Department of Medicine, Division of Hematology
& Medical Oncology, NewYork-Presbyterian Hospital/Weill Cornell Medicine,
New York, NY, USA
Emily S. Lebow, MD Department of Medicine, Memorial Sloan Kettering
Theodore H. Schwartz, MD Department of Neurological Surgery,
Justin Schwarz, MD Department of Neurological Surgery, NewYork-
Daniel M. Sciubba, MD Spine Tumor and Spine Deformity Surgery,
Department of Neurosurgery, Johns Hopkins Hospital, Baltimore, MD, USA
Robert A. Scranton, MD Department of Neurosurgery, Goodman Campbell
Brain and Spine, Indiana University, Indianapolis, IN, USA
Alan Z. Segal, MD Department of Clinical Neurology, NewYork-
Jugal K. Shah, MD Department of Neurosurgery, New York University
Langone Health, New York, NY, USA
Jason P. Sheehan, MD, PhD Neurosurgery and Radiation Oncology,
Department of Neurosurgery, University of Virginia, Charlottesville, VA,
USA
Helen A. Shih, MD, MS, MPH Central Nervous System & Eye Services,
Department of Radiation Oncology, Massachusetts General Hospital, Boston,
MA, USA
xxiv Contributors
Mark M. Souweidane, MD, FACS, FAAP Department of Neurosurgery,

New York-Presbyterian Hospital, Weill Cornell Medicine, New York, NY,
USA
Department of Neurosurgery, Memorial Sloan-Kettering Cancer Center, New
York, NY, USA
Jason A. Spector, MD, FACS Division of Plastic Surgery, Department of
USA
Paul W. Sperduto, MD, MPP, FASTRO Minneapolis Radiation
Oncology & University of Minnesota, Minneapolis, MN, USA
Aditya Srivatsan, MSc Department of Neurosurgery, The University of
Texas M.D. Anderson Cancer Center, Houston, TX, USA
Vittorio Stumpo, MD Università Cattolica del Sacro Cuore School of
Medicine and Surgery, Rome, Italy
Leigh Klaus Swartz, MD Division of Hematology/Oncology, Department
of Internal Medicine, University of Michigan, Ann Arbor, MI, USA
Claudio E. Tatsui, MD Department of Neurosurgery, MD Anderson Cancer
Center, Houston, TX, USA
Jordan A. Torok, MD Department of Radiation Oncology, Duke Cancer
Jeffrey I. Traylor, BS Department of Neurosurgery, The University of Texas
Daniel M. Trifiletti, MD Department of Radiation Oncology, Mayo Clinic,
Jacksonville, FL, USA
Juan S. Uribe, MD, FAANS Division of Spinal Disorders, Department of
Neurological Surgery, Barrow Neurological Institute, Phoenix, AZ, USA
Manuel Valiente, DVM, PhD Brain Metastasis Group, Spanish National
Cancer Research Centre (CNIO), Madrid, Spain
Rafael A. Vega, MD, PhD Department of Neurosurgery, Beth Israel
Deaconess Medical Center, Boston, MA, USA
Vyshak Alva Venur, MD Department of Medical Oncology, University of
Washington Seattle Cancer Care Alliance, Fred Hutchinson Cancer Research
Center, Seattle, WA, USA
Michael S. Virk, MD, PhD Department of Neurological Surgery, Weill
Cornell Medicine, New York Presbyterian Hospital, New York, NY, USA
Monika Vishnoi, PhD Department of Neurosurgery, Houston Methodist
Research Institute, Houston, TX, USA
Contributors xxv
Panagiotis J. Vlachostergios, MD, PhD Department of Medicine, Division

of Hematology & Medical Oncology, NewYork-Presbyterian Hospital/Weill
Cornell Medicine, New York, NY, USA
Tony J. C. Wang, MD Department of Radiation Oncology, Columbia
Jeffrey S. Wefel, PhD, ABPP Section of Neuropsychology, Department of
Neuro-Oncology and Department of Radiation Oncology, The University of
Texas MD Anderson Cancer Center, Houston, TX, USA
Sarah Weiss, MD Medicine (Medical Oncology), Yale University, New
Haven, CT, USA
A. Gabriella Wernicke, MD, MSc Department of Radiation Oncology,
Jessica A. Wilcox, MD Department of Neurology, Memorial Sloan Kettering
Matthew A. Wright, BA Division of Plastic Surgery, Department of
USA
Jimmy Xia, BS Department of Plastic Surgery, NewYork-Presbyterian/
Yoshiya Yamada, MD Department of Radiation Oncology, Memorial Sloan
Bushra Yasin, MD, PhD Department of Neurological Surgery, Weill
Cornell Medicine, New York Presbyterian Hospital, New York, NY, USA
Lucía Zhu, MSc Brain Metastasis Group, Spanish National Cancer Research
Centre (CNIO), Madrid, Spain
Part I
Fundamentals
Epidemiology and Socioeconomic
Impact of CNS Metastases
1
Jessica A. Wilcox and Lisa M. DeAngelis
Introduction secondary to many factors, including longer

patient survival, improvements in screening pro-
Metastatic brain tumors are the most common grams, and increasingly sensitive imaging tech-
intracranial neoplasm in adults and affect up to niques allowing for earlier detection.
one-third of adults with cancer [1]. Most patients
present with neurologic symptoms such as head-
ache, focal weakness or numbness, cognitive Epidemiologic Studies
impairment, or seizures. The diagnosis of central
nervous system (CNS) metastases often requires Epidemiologic studies are important for under-
focal therapy, including neurosurgical or radio- standing the burden of disease, impact of
therapeutic options, as most conventional chemo- advances in treatment, and appropriate allocation
therapies have limited ability to penetrate the of resources. The three major means of analyzing
blood-brain barrier. However, pharmacologic incidence of brain metastases are via population,
treatment of brain metastases has grown in the hospital, and autopsy series.
last two decades due to the advent of immuno-
therapies and targeted therapies based on molec-
ular and genomic tumor profiling. Despite such Autopsy Series
advances, brain metastases remain a significant
cause of morbidity and mortality with a poor Autopsy studies often cite a greater incidence
prognosis for many. Furthermore, the presence of than population studies, with intracranial metas-
brain metastases has historically been an exclu- tasis rates as high as one-third of all patients with
sion criterion for many clinical trials, leaving an cancer [2, 3]. In 1978, Posner et al. found an
unmet need for these patients. intracranial metastasis rate of 24% of 2375
The most common cancers to spread to the patients who died of cancer [2]. Fifteen percent
CNS include lung, breast, melanoma, renal, and of patients had parenchymal metastases, 8% had
colorectal malignancies. The incidence of meta- leptomeningeal metastases, and 20% had dural
static brain tumors from most primaries is on the disease. Takakura et al. quoted a similar inci-
rise. This epidemiologic trend is thought to be dence of 26% with intracranial metastases in
3359 autopsied patients in 1982 [3]. A much
J. A. Wilcox · L. M. DeAngelis (*) larger 1983 autopsy series including 10,916
Department of Neurology, Memorial Sloan Kettering patients found an intraparenchymal metastasis
Cancer Center, New York, NY, USA incidence of 8.7% [4].
e-mail: deangell@MSKCC.ORG
© Springer Nature Switzerland AG 2020 3

R. Ramakrishna et al. (eds.), Central Nervous System Metastases,
https://doi.org/10.1007/978-3-030-42958-4_1
4 J. A. Wilcox and L. M. DeAngelis
Autopsy studies, which are the most accurate ability of neuroimaging; computed tomography
assessment of brain metastasis frequency in ter- (CT) scans became available in 1974 and mag-
minal patients, do have some limitations. The netic resonance imaging (MRI) in the 1990s.
incidence of CNS disease for end-stage cancer Additionally, neurologic symptoms in older
patients will always be higher than that of newly patients or those with end-stage metastatic dis-
diagnosed patients. In addition, due to the dra- ease might not have been investigated or
matic reduction in autopsy rates in the last 30 recognized.
years, the existing autopsy series are outdated More recent studies give a more accurate pic-
and do not reflect the current landscape of onco- ture of current epidemiologic trends of brain
logic care and outcomes. For this reason, most metastases (Table 1.1). Barnholtz-Sloan et al.
recent epidemiologic studies are either popula- calculated the incidence proportions (IPs) for the
tion based or hospital based. most common primary malignancies to spread to
the brain by analyzing the Metropolitan Detroit
Cancer Surveillance System (MDCSS) between
Population and Hospital Series the years 1973 and 2001 [8]. The total IP of brain
metastases for the five most common primary
National population-based registries have long sites combined—lung, breast, melanoma, renal,
been used to gauge epidemiologic trends. Primary colorectal—was 9.6%. The IP for each specific
brain tumors are often recorded in large-scale malignancy was 19.9% for lung, 6.9% for mela-
cancer data sets such as the Surveillance noma, 5.1% for breast, 6.5% for renal, and 1.8%
Epidemiology and End Results (SEER) database; for colorectal cases. Risk was further stratified
however, it was only recently that data regarding based on ethnicity, age, gender, and SEER stage.
brain metastases were included. Hospital-based The malignancy with the highest IP of CNS dis-
studies, conversely, are reliant on autopsy results, semination was metastatic melanoma with an IP
imaging data, pathology, and medical records. of 36.8% across all age groups. African Americans
These results may be biased in the selection for had higher rates of brain metastases from lung,
patients from large tertiary referral centers, which melanoma, and breast cancers as compared to
are not commonly reflective of the population at white patients, and significantly lower rates of
large. Regardless of the methodology, examining brain metastases from renal cancer. With the
series over time demonstrate the rising incidence exception of lung cancer, men had a higher IP of
of CNS metastases in most studies. For example, brain metastases than women. Age at initial diag-
in 1970 a study from Iceland estimated an annual nosis was also influential. For example, those
incidence of 2.8 brain metastases per 100,000 diagnosed with lung cancer between ages 60 and
persons compared to an incidence of 7.8 primary 69 years exhibited the highest absolute frequency
brain tumors per 100,000 [5]. A 10-year Finnish of brain metastases; however, the peak IP for
study reviewing hospital and death records from brain metastases was among those diagnosed
1975 to 1985 found brain metastases and primary between 40 and 49 years of age. Melanoma,
brain tumors to occur in 3.4 and 12.3 per 100,000, renal, and colorectal cases all shared a common
respectively [6]. Within the United States, records IP peak for brain metastasis when the primary
from the Mayo Clinic over a 33-year time period site was diagnosed between 50 and 59 years of
revealed slightly more comparable incidences of age, whereas the IP peak for brain metastasis
11.1 and 12.5 per 100,000 for metastatic and pri- from breast cancer occurred when the primary
mary brain tumors, respectively, in 1972 [7]. tumor was diagnosed between ages 20 and 39.
These older studies have several limitations pre- Although younger breast cancer patients had a
cluding a true estimation of brain metastasis inci- higher risk for brain metastases with an IP of
dence. Asymptomatic brain metastases may have 10%, the absolute frequency of brain dissemina-
escaped clinical detection as screening of the tion was relatively lower in this population com-
CNS was impossible prior to widespread avail- pared to older age groups. The authors surmised
Table 1.1 Incidence proportions of brain metastases from different primary cancers in recent epidemiologic studies
Lung NSCLC SCLC Breast Melanoma Renal Colorectal
No. No. No. No. No. No.
No. with with with with with with with
Reference BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP%
Schouten 156 938 16.3 96 742 12.6 60 196 29.7 42 802 5.0 12 150 7.4 12 114 9.8 10 720 1.2
et al.a [9];
1986–1995;
Maastricht
Cancer
Registry,
Netherlands;
n = 2724
Barnholtz- 11,763 59,038 19.9 2635 51,898 5.1 566 8229 6.9 467 7205 6.5 779 42,817 1.8
Sloan et al.
[8];
1973–2001;
Metropolitan
Detroit Cancer
Surveillance
System,
USA;
1 Epidemiology and Socioeconomic Impact of CNS Metastases
n = 169,187
Duell et al. 118 678 17.4c
[13];
2003–2010;
Asklepios
Lung Hospital,
Germany;
n = 678
Goncalves 2712 30,466 8.9 760 4235 17.9
et al. [11];
1973–2011;
Metropolitan
Detroit SEER
Registry,
USA;
n = 34,681b
(continued)
5
6
Table 1.1 (continued)
Lung NSCLC SCLC Breast Melanoma Renal Colorectal
No. No. No. No. No. No.
No. with with with with with with with
Reference BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP% BM Total N IP%
Cagney et al. 21,804 185,199 11.8 18,241 162,689 11.2 3563 22,510 15.8 973 239,102 0.4 508 77,876 0.7 809 54,495 1.5 365 134,813 0.3
[10]; 75,043 24.3c 15,186 23.5c 12,844 7.6c 1804 28.2c 7463 10.8c 26,923 1.4c
2010–2013;
National
SEER Dataset,
USA;
n = 1,302,166
Zhang et al. 1,547 116,119 1.3
[55]; 4369 35.4c
2010–2015;
National
SEER Dataset,
USA;
n = 121,255
Abbreviations: BM brain metastasis, IP% incidence proportion, NSCLC non-small-cell lung cancer, SCLC small-cell lung cancer, SEER Surveillance Epidemiology and End Results
a
IP% listed is 5-year cumulative incidences published by Schouten et al.
b
n limited to nonmetastatic (local and regional) disease at diagnosis
c
IP% of BM at diagnosis of de novo metastatic disease
J. A. Wilcox and L. M. DeAngelis
1 Epidemiology and Socioeconomic Impact of CNS Metastases 7
that this peak IP in younger breast cancer patients noma were only 0.4%, 1.5%, and 0.7%,
may reflect the increasing trend for longer overall respectively. However, the presence of systemic
survival, giving patients more time to develop metastases at initial presentation significantly
brain metastases after initial diagnosis; it may increased the rates of brain metastases for all can-
also reflect the biology of breast cancer in the cer types: 28.2% for metastatic melanoma, 26.8%
young adult. for lung adenocarcinoma, 23.5% for small-cell
A smaller, population-based Netherlands lung cancer, 15.9% for squamous cell lung can-
study also reported on IP of brain metastases cer, 10.8% for renal cell carcinoma (RCC), and
between the years 1986 and 1995 using the 7.6% for breast cancer.
Maastricht Cancer Registry [9]. A total of 2724
patients were included in this registry, of which
232 (8.5%) were ultimately diagnosed with brain pidemiologic Trends Per Primary
E
metastases. At 5 years, the cumulative incidences Malignancy
for brain metastases were 16.3% for lung, 5.0%
for breast, 7.4% for melanoma, 9.8% for renal, Lung Cancer
and 1.2% for colorectal carcinomas. Among the
lung cancers, the 5-year cumulative incidence Despite the general trend toward increasing inci-
was greater for small-cell carcinoma (29.7%) dence of brain metastases for solid tumor malig-
than non-small-cell carcinoma (12.6%). Unlike nancies in general, this pattern has not been
other studies suggesting an increasing trend in observed for primary lung cancers. Data from the
brain metastases for breast and lung carcinomas, Maastricht Cancer Registry between the years
this study found a non-statistically significant 1986 and 1995 included 938 patients with small-
decrease between the early and later years, albeit cell lung cancer (SCLC) and non-small-cell lung
subject numbers were lower than other larger- cancer (NSCLC) [9]. Both groups showed a drop
scale population studies. in cumulative incidence of brain metastases in
The SEER program, sponsored by the National the latter diagnosis years. NSCLC patients had a
Cancer Institute, publishes cancer incidence and 5-year cumulative incidence of brain metastases
survival data from various population-based reg- of 8.4% for stages I and II disease, 4.3% for stage
istries in the United States. In 2010, the SEER III disease, and 10.8% for stage IV disease. The
database began to include data on the presence or SCLC group had a much higher 5-year cumula-
absence of brain metastases at primary cancer tive incidence at 29.7%. A second study using the
diagnosis. With this new information, Cagney Metropolitan Detroit SEER registry was con-
et al. reviewed the SEER database from 2010 to ducted to ascertain the incidence of brain metas-
2013, capturing 1,302,166 patients diagnosed tases for patients who presented initially with
with extracranial solid tumor malignancies and nonmetastatic SCLC and NSCLC [11]. Between
known status of CNS disease [10]. Of this cohort, the years 1973 and 2011, the IPs for CNS dis-
a total of 26,430 patients had brain metastases at semination for NSCLC and SCLC were 9% and
diagnosis, accounting for 2.0% of all patients and 18%, respectively (Table 1.2). The incidence of
12.1% of patients with systemic metastatic dis- brain metastases was also significantly decreased
ease. The authors estimate that this translates to a in the latter years for both lung cancer types.
brain metastasis incidence of 23,598 per annum
for patients with newly diagnosed cancer in the on-Small-Cell Lung Cancer
N
United States. The most common primary malig- NSCLC is the most common primary malignancy
nancies to present with brain metastases at initial to metastasize to the brain, at an incidence of
diagnosis were small-cell lung cancer (SCLC; 17–44% [12, 13]. This risk is higher for those with
15.8%) and lung adenocarcinoma (14.4%). Brain advanced disease, with rates for CNS recurrence of
metastasis IPs at the time of initial primary diag- 30–50% after initial treatment of locally advanced
nosis for breast cancer, renal cancer, and mela- stage III NSCLC [14, 15]. For early-stage disease,
Table 1.2 Incidence and characteristics of brain metasta- Given the high incidence of brain metastases
ses from SCLC and NSCLC among lung cancer patients, debate has arisen as to
No. with BM Total N IP% whether neurologic screening at NSCLC disease
SCLCa [11] 760 4235 17.9 diagnosis is indicated. The presence of asymptom-
Sex atic brain metastases is common, and a retrospec-
Male 385 2251 17.1
tive review of 809 patients who had routine
Female 375 1984 18.9
Age screening brain MRI or CT at initial NSCLC diag-
<60 303 1325 22.9 nosis found that 22% of patients actually harbored
≥60 457 2910 15.7 brain metastases and 34% were asymptomatic
NSCLCa [11] 2712 30,446 8.9 [19]. Adenocarcinoma and large-cell carcinoma
Sex had higher odds of producing brain metastases as
Male 1553 18,719 8.3 opposed to squamous cell carcinoma, particularly
Female 1159 11,727 9.9
in the absence of nodal involvement in the latter. In
Age
<60 1225 8414 14.6
fact, 33% of patients with de novo brain metasta-
≥60 1487 22,032 6.7 ses were N0 by imaging criteria, and 31% had no
Histology evidence of extra-thoracic metastases at NSCLC
Adenocarcinoma 1181 10,543 11.2 diagnosis, indicating that complete resection for
Squamous cell 722 12,432 5.8 what is believed to be local disease does not pre-
Large cell 243 1984 12.2 clude the possibility of asymptomatic brain metas-
NSCLC, not specified 566 5487 10.3 tases. As a result, routine screening for brain
Tumor genotypeb [21]
metastases is recommended as per standard guide-
EGFR-mutated 19 78 24.4
ALK-rearranged 5 21 23.8 lines for stages III and IV NSCLC [20].
Data from Goncalves et al. [11] and Rangachari et al. [21]
The development of targeted therapies with
Abbreviations: IP% incidence proportion, BM brain blood-brain barrier penetration may reduce the
metastasis, SCLC small-cell lung cancer, NSCLC incidence of brain metastases in certain subsets of
non-small-cell lung cancer NSCLC patients. Patients harboring EGFR muta-
a
IP% calculated using the Metropolitan Detroit
Surveillance Epidemiology and End Results registry for
tions and anaplastic lymphoma kinase (ALK)
patients with non-metastatic first primary lung cancer rearrangements carry a high incidence of brain
diagnosed between 1973–2011. metastases, with nearly 50% of those patients
b
IP% calculated at initial evaluation for a cohort of developing CNS dissemination within 3 years of
patients treated at Beth Israel Deaconess Medical Center
between 2012–2014.
initial diagnosis [21]. Gefitinib and erlotinib are
small molecule tyrosine kinase inhibitors (TKIs) of
predictors for the development of brain metastases EGFR that are efficacious in patients with relapsed
include younger age, larger tumor size, lymphovas- NSCLC or as initial therapy in those with advanced
cular invasion, and hilar lymph node involvement NSCLC with specific EGFR mutations. Brain
[14]. Increasing primary tumor size has been dem- metastasis response rates to these two agents have
onstrated in other studies to be a strong predictor of varied from 10% to 70%, with higher response
metastases to the brain, for both early and advanced rates reported for treatment-naïve never-smokers
stages [16, 17]. The median time from primary [15]. However, CNS progression can occur in
diagnosis to the development of brain metastases patients who otherwise respond systemically to
has ranged from 7.5 to 12.5 months [14]. Women EGFR TKIs [22]. CNS treatment failure may be
were found to have a higher incidence of NSCLC attributed to relatively lower drug concentration in
brain metastases in multiple studies, for reasons the CNS, longer patient survival, and acquisition of
that are unclear [8, 11]. This may be partially TKI-resistance mutations within the CNS. Heon
explained by the observation that women are more et al. evaluated the incidence of brain metastases in
likely to harbor activating epidermal growth factor patients with stage IIIB/V or relapsed NSCLC with
receptor (EGFR) mutations, which confers a sur- EGFR mutations treated with gefitinib or erlotinib
vival advantage and candidacy for certain targeted as initial therapy for advanced disease [15, 23], and
therapies [18]. Women are also more likely to be found an incidence of CNS progression of 28% at
nonsmokers and have adenocarcinoma subtypes. a median follow-up of 42 months. This is much
lower than the 40–55% crude incidence reported in [9]. A similar trend was noted in the Metropolitan
the pre-gefitinib era [24, 25]. Nearly 20% of these Detroit SEER database for more recent years
patients had preexisting brain metastases, the vast [11]. Incidence of brain metastases are much
majority of whom received CNS-directed treatment higher in the younger (<60 years) than older (>80
with surgery or radiotherapy prior to administration years) populations [11]. As opposed to NSCLC,
of the TKI [15]. The risk for CNS progression was there does not appear to be a consistent gender
slightly higher in the cohort with preexisting brain bias for brain metastases from SCLC [11]. One
metastases compared to those without known brain small study found that men had a statistically sig-
metastases, with an overall median time to CNS nificant higher rate of brain metastasis relapse
progression of 19 months. Two patients underwent and shorter brain metastasis-free survival after
surgical resection of brain metastases that devel- chemoradiotherapy for limited disease SCLC
oped while on TKI treatment; both had EGFR-TKI [28], but this has yet to be replicated. At least two
resistance mutations within the CNS lesion. As studies suggest that SCLC brain metastases inci-
reported in other studies, young age was associated dence may be slightly higher in African
with a higher likelihood of CNS progression. More Americans compared to Caucasians; however,
recently, osimertinib has also emerged as an attrac- this has not always reached statistical signifi-
tive third generation TKI with CNS penetration cance [8, 11].
and activity against EGFR mutant lung cancer with Given its propensity to spread to the brain,
T790M resistance mutations. isolated relapse in the CNS is common.
Brain metastases are present in approximately Incomplete thoracic surgical resection and higher
20% of patients with ALK-positive NSCLC at pathologic stage are independent predictors of
initial diagnosis [21]. Crizotinib, a first genera- CNS relapse [29]. Therefore, current guidelines
tion ALK inhibitor, has shown significant activity recommend radiographic CNS screening in all
in the treatment of ALK-positive NSCLC, but it patients who present with SCLC and PCI for
has low durable intracranial response rates due to those with limited- or extensive-stage SCLC who
its poor blood-brain barrier penetration. To cir- achieve a complete or partial response to initial
cumvent this issue, second- and third-generation therapy. This practice guideline is endorsed both
ALK inhibitors such as ceritinib, alectinib, briga- by the American College of Chest Physicians and
tinib brigatinib, and lorlatinib were designed to the American Society of Clinical Oncology due
confer improved CNS permeability. A recent to multiple clinical trials demonstrating increase
review of existing data regarding efficacy of ALK in overall survival and decreasing incidence of
inhibitors for the treatment of brain metastases brain metastases for those who receive PCI [30–
found pooled intracranial objective response 33]. A meta-analysis of four phase II/III trials
rates of 59% for alectinib, 57% for ceritinib, and found 1- and 3-year survival rates of 56% and
26% for crizotinib as first-line therapies [26]. Use 18% for those who received PCI, respectively, as
of these agents as part of initial treatment for sys- compared to 32% and 5% for those who did not
temic disease may ultimately reduce subsequent [30]. These findings were irrespective of age,
intracranial disease progression, but this awaits gender, and stage. Total radiation doses greater
future study. than 30 Gy are more toxic than a 25 Gy regimen,
with resultant shortened survival and increased
mall-Cell Lung Cancer
S chronic neurotoxicity, particular for those older
Multiple reports suggest decreasing overall inci- than 60 years of age [30, 34]. For this reason,
dence of SCLC brain metastases, perhaps owing 25 Gy is the standard regimen for those undergo-
to the steady decline in SCLC incidence [27] and ing PCI. However, despite clear benefit of PCI on
prophylactic cranial irradiation (PCI). The overall survival and the development of brain
Netherlands population-based study of the metastases, PCI is not universally utilized due to
Maastricht Cancer Registry indicated a cumula- concern for long-term cognitive damage. In one
tive incidence of brain metastases from SCLC of institution-based study of 283 patients with
32.5% in 1986–1990 and 26.0% in 1991–1995 limited-stage SCLC, only 55% of eligible patients
ultimately received PCI. The most common Table 1.3 Incidence and median OS of patients with
reasons for PCI omission were patient refusal brain metastases from newly diagnosed metastatic breast
cancer stratified
due to concerns for neurotoxicity, followed by
physician assessment of the patient being medi- No. with Total Median
BM N IP%a OSb
cally unfit and advanced age [35]. A hospital-
HR+/HER2+ 136 1704 8.0 21.0
based review of SCLC patients treated in France months
between the years 1997 and 2017 found slightly HR+/HER2− 361 6607 5.5 14.0
increased PCI utilization in more recent years, months
which might provide an explanation for the recent HR−/HER2+ 106 926 11.5 10.0
months
decline in SCLC brain metastasis incidence on a
HR−/HER2− 173 1522 11.4 6.0 months
global scale [36]. However, the same review also
Data from Martin et al. [41]
found no improvement in overall survival or Abbreviations: BM brain metastasis, IP% incidence pro-
response to chemotherapy over the years portion, OS overall survival, HR hormone receptor, HER2
assessed, indicating a significant need for new human epidermal growth factor receptor 2
treatment strategies for SCLC.
a
IP% calculated using the Surveillance Epidemiology and
End Results registry for patients with newly diagnosed
metastatic breast cancer between 2010 and 2013
b
Median OS for patients with brain metastases at the time
Breast Cancer of initial breast cancer diagnosis
Breast cancer is estimated to be the second most respectively, of being hospitalized for complica-
frequent cause of brain metastases, and recent tions related to brain metastases.
studies suggest that the incidence of brain metas- Several factors have been associated with a
tases in this population might be increasing. greater incidence of brain metastases. These
Population studies indicate that brain metastases include: younger age at diagnosis, advanced
are diagnosed in approximately 5% of patients, stage, aggressive histologic features, BRCA1
but the incidence was found to be much higher in mutations, triple-negative subtypes, and human
prior autopsy series at 18–30% of patients [8, 9, epidermal growth factor receptor 2 (HER2)
37]. Although radiographic screening for brain amplification (Table 1.3) [12, 41]. Within the
metastases in breast cancer patients is not stan- HER2-amplified group, the administration of
dard, Miller et al. reviewed screening imaging of trastuzumab, a non-blood-brain barrier penetrat-
155 breast cancer patients and found that 14.8% ing drug, is further associated with later CNS dis-
had occult asymptomatic brain metastases [38]. semination of disease [37, 42].
Taking this discrepancy into account, one can One small study discovered that among 15
theorize that approximately 20–30% of all patients with germline BRCA1 mutations, 67%
women with breast cancer will develop symp- developed parenchymal brain metastases com-
tomatic or asymptomatic brain metastases during pared to none of the germline BRCA2 and 10.3%
the course of their illness, with certain subpopu- of the BRCA noncarrier patients [43]. Median
lations being at greater risk [37]. Median time time from first breast cancer metastasis to the
from breast cancer diagnosis and detection of diagnosis of brain metastases in BRCA1 carriers
brain metastases, regardless of tumor subtype, is was 7.8 months. These findings are also in line
approximately 35 months [39]. Admission rates with the observation that BRCA1 mutations are
for brain metastases have been rising steadily often associated with triple-negative breast can-
since the late 1990s, according to the National cer, another factor associated with increased inci-
Cancer Register in Sweden [40]. Compared with dence of brain metastases [12].
patients diagnosed with breast cancer in 1998– The HER2/neu proto-oncogene is amplified in
2000, those diagnosed between 2001–2003 and 25–30% of primary breast cancers [44]. The inci-
2004–2006 had a 17% and 44% increased risk, dence of brain metastases in this group of patients
has ranged from 30% to 40% [45], significantly zumab and control groups, respectively [49]. Bria
higher than that of the entire breast cancer popu- et al. concluded that more than 160 patients
lation. Amplification of this receptor is a ssociated would need to be treated with trastuzumab in
with cellular proliferation, migration, and neo- order to observe one event [46]. Furthermore,
angiogenesis [44]. The introduction of trastu- although the brain has a higher rate of first recur-
zumab, a humanized monoclonal antibody rence in the trastuzumab-treated groups, this is
directed at HER2, has significantly improved partially explained by early failure in other organs
extracranial disease control and survival in in the control arms. Lapatinib, an oral HER2 and
HER2-amplified breast cancer patients [46]. EGFR inhibitor, has emerged as an efficacious
However, despite its success with systemic dis- treatment for patients who have developed brain
ease, trastuzumab has been associated with an metastases after pretreatment with trastuzumab,
increased incidence of brain metastases affecting due to the ability of lapatinib to cross the blood-
25–48% of such patients [42, 46]. The median brain barrier. A meta-analysis including 799
time from start of treatment with trastuzumab to patients with brain metastases found that lapa-
the development of brain metastases ranges from tinib in combination with capecitabine achieved
4 to 24 months [47]. This is presumed to be sec- response rates in 30% of patients that had been
ondary to “gained deaths,” meaning that the sur- pretreated with trastuzumab [50].
vival benefit from trastuzumab results in patients
living long enough to develop brain metastases at
a later, more advanced stage. Additionally, pos- Melanoma
sible loss of HER2 overexpression in brain
metastases and the failure of trastuzumab to cross Melanoma is the third most common cause of
the blood-brain barrier might also account for brain metastases, with an incidence of 7–75%
this rise in brain metastases. Due to its high according to population and autopsy data [8, 9,
molecular weight of 145 kDa, trastuzumab pen- 51–54]. More recent population-based studies
etration into the CNS is highly impaired, leaving suggest that melanoma brain metastases are pres-
the brain relatively unprotected during treatment ent at primary diagnosis in 1.3% of all-comers;
as compared to extracranial sites. however, that number rises to 28.2–35.4% when
To better delineate this observation, four patients present with de novo metastatic disease
major randomized clinical trials were conducted [10, 55]. Although melanoma only represents 1%
(NSABP B-31, NCCTG N9831, HERA, PACS of all malignancies, the incidence of cutaneous
04) that explore the safety and efficacy of adju- melanoma has been steadily increasing [53].
vant trastuzumab. All trials indicate a trend Melanoma also has the highest propensity of all
toward increase in brain metastases in the primary cancers to disseminate to the brain [2].
trastuzumab- treated groups, which in meta- Brain metastases from melanoma are notoriously
analyses reached statistical significance [46, 48, hemorrhagic, and are more likely to seed the cor-
49]. The overall relative risk for the development tex rather than the gray-white junction as seen in
of CNS metastases as a site of first recurrence other malignancies. Data suggest that melanoma
ranged from 1.35 to 1.57 for patients that received brain metastases are most likely to occur before
adjuvant trastuzumab as compared with patients or during the first line of systemic therapy, lead-
who did not receive trastuzumab [46, 49], and the ing some practitioners including the National
ratio of CNS to non-CNS metastases was dou- Comprehensive Cancer Network to recommend
bled in the trastuzumab-treated arm [49]. brain imaging for initial staging of patients with
However, despite this reproducible trend, the advanced melanoma [56, 57].
overall incidence of brain metastases as a site of Among primary melanoma tumor characteris-
first recurrence remained low in the time inter- tics, primary ulceration and origin on the head
vals studied: 2.56% versus 1.94% in the trastu- and neck are the strongest independent predictors
of development of brain metastases [58]. Other tors conferred a greater reduction in brain
significant associations include thick lesions, metastasis size compared to ipilimumab, an
nodular melanoma, and tumors with higher antibody against cytotoxic T lymphocyte-asso-
mitotic index. Certain molecular phenotypes are ciated antigen-4 (CTLA-4) [64].
also associated with the development of brain Combination of the BRAF and mitogen-
metastases, notably BRAF, NRAS, and PTEN activated protein kinase (MEK) inhibitors dab-
mutations [56]. rafenib and trametinib, respectively, in patients
Overall survival after the development of with BRAF V600-mutant melanoma also halted
brain metastases from melanoma has historically brain metastasis growth, albeit for a shorter dura-
ranged from only 3 to 6 months [56, 58]. Patients tion when compared to the combination’s extra-
with solitary or oligometastatic disease that can cranial disease control [66]. Despite the
be treated with surgical resection or stereotactic demonstration of CNS activity of the newer
radiosurgery (SRS) have a survival advantage of agents, the rate of de novo brain metastases was
an additional 3–7 months [59]. Fortunately, data not significantly lower for those receiving BRAF
suggest that survival in recent years has improved or checkpoint inhibitors when compared to stan-
from the time of brain metastasis diagnosis [56]. dard chemotherapies [56].
One retrospective study found that median over-
all survival was nearly three times as long in
patients diagnosed with brain metastases in 2011 Renal Cell Carcinoma
as compared to the years 2000–2008 (22.7
months versus 7.5 months, respectively), likely Brain metastases are observed in patients with
owing to improvement in radiation techniques, renal cell carcinoma (RCC) at an estimated inci-
targeted agents, and immunotherapies [56]. dence of 2–17% [8, 9, 67]. However, brain metas-
Combination therapy with ipilimumab and tases are only present in approximately 1.5% of
nivolumab in patients with asymptomatic, patients at diagnosis [10]. Factors associated with
untreated brain metastases demonstrated an higher odds of brain metastases at diagnosis
intracranial response rate of 57% in a phase II include larger primary tumors (>10 cm), higher
study, although more than half experienced stage and tumor grade, nodal metastases, clear
grade 3 or 4 treatment-related adverse events cell histology, white race, and lower socioeco-
[60]. A recent review of the National Cancer nomic status [68]. Between 2010 and 2013, the
Database revealed that among melanoma incidence of brain metastases at diagnosis was
patients who present with brain metastases, shown to be lower than the 6.5% reported in
treatment with immune checkpoint inhibitors led 2001, perhaps owing to escalation of body imag-
to significant improvements in both median ing in the last decade leading to earlier local diag-
(12.4 versus 5.2 months) and 4-year overall sur- nosis of small renal masses.
vival (28.1% versus 11.1%) [61]. A retrospective review of the phase III
The use of SRS with either BRAF inhibitors Treatment Approaches in Renal Cancer Global
or immune checkpoint inhibitors has also Evaluation Trial (TARGET) evaluated the devel-
yielded improvement in intracranial control opment of brain metastases for patients with met-
with a trend toward increased survival [53, 56, astatic RCC receiving sorafenib, an oral TKI that
62, 63]. SRS, when delivered within 4 weeks of targets vascular endothelial growth factor recep-
initiation of immunotherapy, has been shown to tor (VEGFR) and platelet derived growth factor
improve response rates with a trend toward lon- receptor (PDGFR) [69]. In this trial, all patients
ger overall survival, theoretically because had negative head imaging at treatment initiation,
radiation increases immunogenicity and sus- and those who received sorafenib achieved a two-
ceptibility of tumors to immune checkpoint fold increase in progression-free survival com-
inhibitors [64, 65]. Concurrent treatment with pared to placebo. Treatment with sorafenib was
programmed cell death protein-1 (PD-1) inhibi- associated with a significant reduction in the
occurrence of brain metastases, with incidence of general. Median age at brain metastasis develop-
3% and 12% in the sorafenib- and placebo-treated ment was the seventh decade for most studies.
groups, respectively, at a median follow-up of 19 Diagnosis of CNS dissemination averaged
months. This retrospective study is limited by between 20 and 40 months from the primary
small patient numbers; however, it does suggest diagnosis, with more advanced systemic disease
efficacy of sorafenib in treating or delaying the at initial diagnosis leading to earlier CNS metas-
development of brain metastases. This finding is tases. The presence of lung metastases seems to
supported by a case report demonstrating reduc- be uniformly associated with an increased risk of
tion of gadolinium-contrast enhancement from brain metastasis development, with an incidence
renal cell leptomeningeal carcinomatosis in a of brain metastases of 6.2–22.6% in those with
patient treated with sorafenib [70]. A similar anti- lung metastases. Interestingly, the presence of
angiogenic TKI, sunitinib, has also shown effi- liver metastases seems to have an inverse rela-
cacy for the treatment of RCC brain metastases in tionship with CNS dissemination [77]. One plau-
case reports [71, 72]. Few studies suggest intra- sible explanation for this discrepancy is that CRC
cranial response when using immune checkpoint has more routes of hematogenous spread to the
inhibitors for the treatment of brain metastases lung and brain than it does to the liver. This
from RCC [73]. However, this response tends to includes direct extension from the vertebral
be less robust than what is observed in other plexus to the brain and indirect extension from
malignancies. the vena cava to the lungs and then the brain.
Molecular markers have also been studied in
their relationship to CNS dissemination; how-
Colorectal Cancer ever, the only mutation routinely associated with
brain and lung metastases involve RAS [78].
Colorectal cancer (CRC) is the fourth most com- Other markers including PIK3CA and BRAF
mon adult malignancy, but only infrequently mutations, EGFR expression, and CEA and
metastasizes to the brain. Recent review of exist- CA19.9 levels have all been investigated, but
ing literature on brain metastases from CRC sug- none with any clear certainty of relationship to
gests an incidence of only 0.6–3.2% [74]. Part of brain metastases [74].
this stable trend may be due to the relative pau-
city of effective agents for metastatic CRC. As
treatment options for advanced CRC have Cancers with Rare CNS Metastases
expanded in recent years to include irinotecan,
oxaliplatin, and bevacizumab to name a few, Prostate cancer incidence and mortality have
patient survival has been prolonged [75, 76]. It both declined in the United States in the last
has been suggested that with this survival advan- decade, although this trend has not been reported
tage, metastatic patterns of CRC might be evolv- in all countries [79, 80]. A retrospective series of
ing to include uncommon sites of disease 16,280 patients treated at M.D. Anderson Cancer
dissemination, including the CNS. However, Center between 1944 and 1998 found a 0.63%
recent hospital- and population-based studies on incidence of parenchymal brain metastases [81].
this question have demonstrated mixed results. Nearly 90% of the patients presented with a soli-
A large review of existing literature in 2016 tary metastasis, with squamous cell carcinoma
found a weighted mean of 1.55% incidence of and cribriform subtypes metastasizing to the
brain metastases from CRC, with no significant CNS more commonly than adenocarcinomas.
difference in reported incidence with the year of Following the incorporation of docetaxel as a
data collection [74]. The weighted incidence of first-line treatment for castration-resistant pros-
brain metastases from rectal cancer was 48.5% tate cancer (CRPC) in 2004, the longer overall
compared to colon cancer, which is striking given survival and lack of docetaxel penetration through
the relatively lower incidence of rectal cancer in the blood-brain barrier may lead to an increased
incidence of brain metastases. Patients with plastic cancer in 23% and medullary cancer in
CRPC treated with docetaxel between 2002 and 9%. Affected patients tended to be older and have
2010 had a 3.3% incidence of brain metastases distant metastases at initial cancer diagnosis.
[82]. While this figure is much higher than what
has been reported historically, direct comparison
is challenging because discrimination by ocioeconomic Impact of Brain
S
castration-resistance status in the pre-docetaxel Metastases
era had not always been reported in prior epide-
miologic studies. With the increasing incidence of brain metasta-
Similarly for urothelial cancer, the incidence ses across the general population, one would
of brain metastases is thought to have risen from expect a rise in socioeconomic burden. While
a historical 1–3% to 16% after the advent of studies comparing the overall economic cost of
MVAC (methotrexate, vinblastine, doxorubicin, treating brain metastases now versus prior
cisplatin) due to improvements in overall survival decades are lacking, the literature does reveal a
but lack of blood-brain barrier penetration [83]. marked increase in health care expenditure after
More recent studies elucidating the incidence of the diagnosis of brain metastases. A claims anal-
brain metastases in urothelial cell carcinoma are ysis of patients with breast cancer between the
lacking. years 2002 and 2004 found that the average total
Gynecologic malignancies generally have a cost at 6 months was $60,045 for those with
low propensity to spread to the brain. Although brain metastases versus $28,193 for those with-
the incidence of brain metastases for ovarian can- out [90]. Similarly, a second claims analysis for
cer varies widely per study, with historical reports individuals with lung cancer found a rise in total
as high as 11.6%, most current studies hypothe- 6-month cost per patient from $70,157 to
size a relatively stable incidence of 1–2.5% [84, $86,027 when comparing the pre- and post-brain
85]. There is a slight tendency toward multiple as metastasis diagnosis intervals [91]. Among a
opposed to solitary brain metastases [85]. Rates cohort of patients with ALK-rearranged NSCLC
of new cervical cancer diagnoses have been treated with crizotinib, monthly health care
steadily decreasing according to SEER statistics, expenditure increased from $5983 to $22,645
likely owing to improved screening practices and per patient following the diagnosis of brain
the human papillomavirus vaccination [80]. metastases, with the main economic contributors
Cervical cancer brain metastasis incidence is also being pharmacy (42.0%), inpatient (29.6%), and
very low, with reports ranging 0.4–2.3%; how- outpatient (26.0%) costs [92]. A similar rise in
ever, it may be increasing due to longer overall monthly health care expenditure from $7277 to
survival in this patient population [86]. $14,489 has been observed for patients with
Endometrial cancer has the lowest propensity to melanoma following the diagnosis of brain
spread to the brain, with a pooled rate of only metastases [93].
0.6% of patients developing brain metastases as Among all studies included, one of the largest
per a recent review [87]. drivers for increased health care expenditure
Brain metastases from thyroid cancer remain arises from inpatient hospitalizations, in terms of
exceedingly rare, and are generally thought to the number of admissions, length of stay, and
occur in only 0.15–1.3% of patients [88]. One total cost. In 6 months, breast cancer patients
retrospective review of 3117 patients with thy- with brain metastases averaged 1.1 hospitaliza-
roid carcinoma at a single institution found an tions of 8.0 days duration compared to 0.5 admis-
incidence of brain metastases of 1.5% between sions of 2.5 days duration in controls [90]. This
clinical and autopsy data [89]. The most common correlated with an increase in 6-month hospital-
thyroid cancer to cause brain metastases was the ization costs from $5362 to $17,462 [90]. Lung
differentiated subtype in 68%, followed by ana- cancer patients had higher rates of admission fol-
lowing brain metastasis diagnosis with 10.7-day References

longer lengths of stay on average [91]. Mean pre-
scription costs, radiology services, physician vis- 1. Gavrilovic IT, Posner JB. Brain metastases: epidemiol-
its, and other outpatient visits are also uniformly ogy and pathophysiology. J Neurooncol. 2005;75:5–14.
2. Posner JB, Chernik NL. Intracranial metastases from
increased for those diagnosed with brain metasta- systemic cancer. Adv Neurol. 1978;19:579–92.
ses [90–93]. 3. Takakura K, Sano K, Hojo S, Hirano A. Metastatic
In addition to the dramatic rise in direct health tumors of the central nervous system. Tokyo/New
care expenses, the cost of productivity loss on the York: Igaku-Shoin; 1982.
4. Pickren J, Lopez G, Tsukada Y, Lane W. Brain
patients, payers, and employers is also signifi- metastases: an autopsy study. Cancer Treat Symp.
cantly affected [91]. Salary losses averaging over 1983;1983(2):295–313.
$8000 per 6 months, largely secondary to unpaid 5. Guomundsson KR. A survey of tumors of the central
sick days, were calculated for lung cancer patients nervous system in Iceland during the 10-year period
1954–1963. Acta Neurol Scand. 1970;46:538–52.
after brain metastasis diagnosis. The absentee 6. Fogelholm R, Uutela T, Murros K. Epidemiology of
rate from work approached 50% in this popula- central nervous system neoplasms: a regional survey in
tion. This is felt to be a gross underestimation of Central Finland. Acta Neurol Scand. 1984;69(3):129–36.
global employment days lost as this number does 7. Percy A, Elveback L, Okazaki H, Kurland
L. Neoplasms of the central nervous system: epidemi-
not include the additional consequences on fami- ologic considerations. Neurology. 1972;22(1):40–8.
lies and caregivers. Furthermore, quality of life 8. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau
scores vary among studies, but generally are sta- FD, Lai P, Sawaya RE. Incidence proportions of brain
ble to worse following treatment with whole metastases in patients diagnosed (1973 to 2001) in the
Metropolitan Detroit Cancer Surveillance System. J
brain radiotherapy, reflecting a high unmet need Clin Oncol. 2004;22(14):2865–72.
for improved treatment options [94]. 9. Schouten LJ, Rutten J, Huveneers HAM, Twijnstra
A. Incidence of brain metastases in a cohort of patients
with carcinoma of the breast, colon, kidney, and lung
and melanoma. Cancer. 2002;94(10):2698–705.
Conclusion 10. Cagney DN, Martin AM, Catalano PJ, Redig AJ,

Lin NU, Lee EQ, et al. Incidence and prognosis of
patients with brain metastases at diagnosis of sys-
Brain metastases are the most frequently encoun- temic malignancy: a population-based study. Neuro
tered tumors of the CNS, and epidemiologic data Oncol. 2017;19(11):1511–21.
suggest that the incidence of brain metastases is 11. Goncalves PH, Peterson S, Vigneau FD, Shore RD,
Quarshie WO, Islam K, et al. Risk of brain metas-
increasing. Lung, breast, renal, melanoma, and tases in patients with non-metastatic lung cancer:
colorectal cancers are the most common malignan- analysis of the Metropolitan Detroit Surveillance,
cies to spread to the brain. The general rise in brain Epidemiology, and End Results (SEER) data. Cancer.
metastasis frequency can be attributed to longer 2016;122(12):1921–7.
12. Nayak L, Lee EQ, Wen PY. Epidemiology of brain
patient overall survival with better therapies and metastases. Curr Oncol Rep. 2012;14:48–54.
earlier diagnosis of brain metastases. In the era of 13. Duell T, Kappler S, Knoferl B, Schuster T, Hochhaus
targeted therapeutics based on genetic and molecu- J, Morresi-Hauf A, et al. Prevalence and risk factors
lar cancer subtypes, improved systemic disease of brain metastases in patients with newly diagnosed
advanced non-small-cell lung cancer. Cancer Treat
control has led to the brain as a common site of late Commun. 2015;4:106–12.
recurrence. In addition to the detrimental impact on 14. Hubbs JL, Boyd JA, Hollis D, Chino JP, Saynak M,
quality of life, this overall trend also has significant Kelsey CR. Factors associated with the development
socioeconomic implications given the dramatic rise of brain metastases. Cancer. 2010;116:5038–46.
15. Heon S, Yeap BY, Britt GJ, Costa DB, Rabin MS,
in health care expenditure following the diagnosis Jackman DM, et al. Development of central nervous
of brain metastases. Further population-based epi- system metastases in patients with advanced non-
demiologic studies are needed to determine if small cell lung cancer and somatic EGFR mutations
newer CNS-penetrating treatments will be effec- treated with gefitinib or erlotinib. Clin Cancer Res.
2010;16(23):5873–82.
tive in preventing CNS relapse.
16. Bajard A, Westeel V, Dubiez A, Jacoulet P, Pernet Epidemiologic, and End Results database. J Clin
D, Dalphin JC, et al. Multivariate analysis of fac- Oncol. 2016;24(28):4539–44.
tors predictive of brain metastases in localized 28.
Roengvoraphoj O, Eze C, Niyazi M, Li M,
non-small cell lung carcinoma. Lung Cancer. Hildebrandt G, Fietkau R, et al. Prognostic role of
2004;45(3):317–23. patient gender in limited-disease small-cell lung
17. Mujoomdar A, Austin JH, Malhotra R, Powell C,
cancer treated with chemoradiotherapy. Strahlenther
Pearson GD, Shiau MC, et al. Clinical predictors of Onkol. 2017;193:150–5.
metastatic disease to the brain from non-small cell 29. Gong L, Wang QI, Zhao L, Yuan Z, Li R, Wang
lung cancer: primary tumor size, cell type, and lymph P. Factors affecting the risk of brain metastasis in
node metastases. Radiology. 2007;242(3):882–8. small cell lung cancer with surgery: is prophylactic
18. Bell DW, Brannigan BW, Matsuo K, Finkelstein DM, cranial irradiation necessary for stage I-III disease?
Sordella R, Settleman J, et al. Increased prevalence Int J Radiat Oncol Biol Phys. 2012;85(1):196–200.
of EGFR-mutant lung cancer in women and in East 30. Schild SE, Foster NR, Meyers JP, Ross HJ, Stella PJ,
Asian populations: analysis of estrogen-related poly- Garces YI, et al. Prophylactic cranial irradiation in
morphisms. Clin Cancer Res. 2008;14(13):4079–84. small-cell lung cancer: findings from a North Central
19. Shi AA, Digumarthy SR, Temel JS, Halpern EF,
Cancer Treatment Group Pooled Analysis. Ann
Kuester LB, Aquino SL. Does initial staging or tumor Oncol. 2012;23(11):2919–24.
histology better identify asymptomatic brain metasta- 31. Rudin CM, Ismaila N, Hann CL, Malhotra N, Movsas
ses with non-small cell lung cancer? J Thorac Oncol. B, Norris K. Treatment of small-cell lung cancer:
2006;1(3):205–10. American Society of Clinical Oncology endorsement
20. Detterbeck FC, Lewis SZ, Diekemper R, Addrizzo- of the American College of Chest Physicians guide-
Harris D, Alberts WM. The stage classification of line. J Clin Oncol. 2015;33(34):4106–11.
lung cancer: diagnosis and management of lung can- 32. Auperin A, Arriagada R, Pignon J-P, Le Pechoux C,
cer, 3rd ed: American College of Chest Physicians Gregor A, Stephens RJ, et al. Prophylactic cranial
evidence-based clinical practice guidelines. Chest. irradiation for patients with small-cell lung cancer in
2013;143(5 suppl):7S–37S. complete remission. Prophylactic Cranial Irradiation
21. Rangachari D, Yamaguchi N, VanderLaan PA, Folch Overview Collaborative Group. N Engl J Med.
E, Mahadevan A, Floyd SR, et al. Brain metas- 1999;341:476–84.
tases in patients with EGFR-mutated or ALK- 33. Ge W, Xu H, Yan Y, Cao D. The effects of prophy-
rearranged non-small-cell lung cancers. Lung Cancer. lactic cranial irradiation versus control on survival of
2015;88(1):108–11. patients with extensive-stage small-cell lung cancer: a
22. Omuro AM, Kris MG, Miller VA, Franceschi E, Shah meta-analysis of 14 trials. Radiat Oncol. 2018;13:155.
N, Milton DT, et al. High incidence of disease recur- 34. Wolfson AH, Bae K, Komaki R, Meyers C, Movsas
rence in the brain and leptomeninges in patients with B, Le Pechoux C, et al. Primary analysis of a phase
nonsmall cell lung carcinoma after response to gefi- II randomized trial Radiation Therapy Oncology
tinib. Cancer. 2005;103(11):2344–8. Group (RTOG) 0212: impact of different total doses
23. Heon S, Yeap BY, Linderman NI, Joshi VA, Butaney and schedules of prophylactic cranial irradiation on
M, Britt GJ, et al. The impact of initial gefitinib or chronic neurotoxicity and quality of life for patients
erlotinib versus chemotherapy on central nervous with limited-disease small-cell lung cancer. Int J
system progression in advanced non-small cell lung Radiat Oncol Biol Phys. 2011;81(1):77–84.
cancer with EGFR mutations. Clin Cancer Res. 35. Lok BH, Ma J, Foster A, Perez CA, Shi W, Zhang
2012;18(16):4406–14. Z, et al. Factors influencing the utilization of pro-
24. Mamon HJ, Yeap BY, Janne PA, Reblando J, Shrager phylactic cranial irradiation in patients with limited-
S, Jaklitsch MT. High risk of brain metastases in sur- stage small cell lung cancer. Adv Radiat Oncol.
gically staged IIIA non-small-cell lung cancer patients 2017;2(4):548–54.
treated with surgery, chemotherapy, and radiation. J 36. Lattuca-Truc M, Levra MG, Ruckly S, Villa J, Dumas
Clin Oncol. 2005;23(7):1530–7. I, Julian P. Trends in response rate and survival in
25. Chen HY, Yu SL, Chen CH, Chang GC, Chen CY, small cell lung cancer patients between 1997 and
Yuan A, et al. A five-gene signature and clinical out- 2017. J Clin Oncol. 2018;36:e20572.
come in non-small-cell lung cancer. N Engl J Med. 37. Stemmler H-J, Heinemann V. Central nervous sys-
2007;356(1):11–20. tem metastases in HER-2-overexpressing metastatic
26. Petrelli F, Lazzari C, Ardito R, Borgonovo K,
breast cancer: a treatment challenge. Oncologist.
Bulotta A, Conti B, et al. Efficacy of ALK inhibi- 2008;13:739–50.
tors on NSCLC brain metastases: a systematic 38. Miller KD, Weathers T, Haney LG, Timmerman R,
review and pooled analysis of 21 studies. PLoS One. Dickler M, Shen J, et al. Occult central nervous sys-
2018;13(7):e0201425. tem involvement in patients with metastatic breast
27. Govindan R, Page N, Morgensztern D, Read W,
cancer: prevalence, predictive factors and impact on
Tierney R, Vlahiotis A. Changing epidemiol- overall survival. Ann Oncol. 2003;14:1072–7.
ogy of small-cell lung cancer in the United States 39. Bachmann C, Schmidt S, Staebler A, Fehm T, Fend F,
over the last 30 years: analysis of the Surveillance, Schittenhelm J, et al. CNS metastases in breast cancer
patients: prognostic implications of tumor subtype. noma brain metastases treated with stereotactic radio-
Med Oncol. 2015;32:400. surgery in the era of immunotherapy and targeted
40. Frisk G, Svensson T, Backlund LM, Lidbrink E,
agents. Adv Radiat Oncol. 2017;2:572–80.
Blomqvist P, Smedby KE. Incidence and time trends 54. Sloan AE, Nock CJ, Einstein DB. Diagnosis and

of brain metastases admissions among breast cancer treatment of melanoma brain metastasis: a literature
patients in Sweden. Br J Cancer. 2012;106:1850–3. review. Cancer Control. 2009;16(3):248–55.
41. Martin AM, Cagney DN, Catalano PJ, Warren LE, 55. Zhang D, Wang Z, Shang D, Yu J, Yuan S. Incidence
Bellon JR, Punglia RS, et al. Brain metastases in and prognosis of brain metastases in cutaneous mela-
newly diagnosed breast cancer: a population-based noma patients: a population-based study. Melanoma
study. JAMA Oncol. 2017;3(8):1069–77. Res. 2019;29(1):77–84.
42. Musolino A, Ciccolallo L, Panebianco M, Fontana E, 56. Sloot S, Chen YA, Zhao X, Weber JL, Benedict JJ,
Zanoni D, Bozzetti C, et al. Multifactorial central ner- Mule JJ, et al. Improved survival of patients with
vous system recurrence susceptibility in patients with melanoma brain metastases in the era of targeted
HER2-positive breast cancer: epidemiological and BRAF and immune checkpoint therapies. Cancer.
clinical data from a population-based cancer registry 2017;124(2):297–305.
study. Cancer. 2011;117(9):1837–46. 57. Coit DG, Thompson JA, Algazi A, Andtbacka R,
43. Albiges L, Andre F, Balleyguier C, Gomez-Abuin Bichakjian CK, Carson WE, et al. Melanoma, version
G, Chompret A, Delaloge S. Spectrum of breast can- 2.2016, NCCN clinical practice guidelines in oncol-
cer metastasis in BRCA1 mutation carriers: highly ogy. J Natl Compr Canc Netw. 2016;14(4):450–73.
increased incidence of brain metastases. Ann Oncol. 58. Zakrzewski J, Geraghty LN, Rose AE, Christos

2005;16(11):1846–7. PJ, Mazumdar M, Polsky D, et al. Clinical vari-
44. Slamon DJ, Godolphin W, Jones LA, Holt JA, Wong ables and primary tumor characteristics predictive
SG, Keith DE, et al. Studies of the HER-2/neu of the development of melanoma brain metasta-
proto-oncogene in human breast and ovarian cancer. ses and post-brain metastases survival. Cancer.
Science. 1989;244(4905):707–12. 2011;117(8):1711–20.
45. Slimane K, Andre F, Delaloge S, Dunant A, Perez A, 59. Ramakrishna N, Margolin KA. Multidisciplinary

Grenier J, et al. Risk factors for brain relapse in patients approach to brain metastasis from melanoma; local
with metastatic breast cancer. Ann Oncol. 2004;15:1640–4. therapies for central nervous system metastases. Am
46. Bria E, Cuppone F, Fornier M, Nistico C, Carlini P, Soc Clin Oncol Educ Book. 2013;33:399–403.
Milella M, et al. Cardiotoxicity and incidence of brain 60. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS,
metastases after adjuvant trastuzumab for early breast Moschos SJ, et al. Combined nivolumab and ipilim-
cancer: the dark side of the moon? A meta-analysis umab in melanoma metastatic to the brain. N Engl J
of the randomized trials. Breast Cancer Res Treat. Med. 2018;379:722–30.
2008;109:231–9. 61. Iorgulescu JB, Harary M, Zogg CK, Ligon KL,

47. Duchnowska R, Szczylik C. Central nervous system Reardon DA, Hodi FS, et al. Improved risk-adjusted
metastases in breast cancer patients administered survival for melanoma brain metastases in the era
trastuzumab. Cancer Treat Rev. 2005;31:312–8. of checkpoint blockade immunotherapies: results
48. Romond EH, Perez EA, Bryant J, Suman V, Geyer from a national cohort. Cancer Immunol Res.
CE, Davidson NE, et al. Trastuzumab plus adjuvant 2018;6(9):1039–45.
chemotherapy for operable HER2-positive breast can- 62. Gaudy-Marqueste C, Dussouil AS, Carron R, Troin
cer. N Engl J Med. 2005;353:1673–84. L, Malissen N, Loundou A, et al. Survival of mela-
49. Olson EM, Abdel-Rasoul M, Maly J, Wu CS, Lin NU, noma patients treated with targeted therapy and
Shapiro CL. Incidence and risk of central nervous sys- immunotherapy after systematic upfront control
tem metastases as site of first recurrence in patients of brain metastases by radiosurgery. Eur J Cancer.
with HER2-positive breast cancer treated with adju- 2017;84:44–54.
vant trastuzumab. Ann Oncol. 2013;24:1526–33. 63.
Gabani P, Fischer-Valuck BW, Johanns TM,
50. Petrelli F, Ghidini M, Lonati V, Tomasello G,
Hernandez-Aya LF, Keller JW, Rich KM, et al.
Borgonovo K, Ghilardi M, et al. The efficacy of lapa- Stereotactic radiosurgery and immunotherapy in mel-
tinib and capecitabine in HER-2 positive breast cancer anoma brain metastases: patterns of care and treat-
with brain metastases: a systemic review and pooled ment outcomes. Radiother Oncol. 2018;128:266–73.
analysis. Eur J Cancer. 2017;84:141–8. 64. Qian JM, Yu JB, Kluger HM, Chiang VLS. Timing
51. Patel JK, Didolkar MS, Pickren JW, Moore and type of immune checkpoint therapy affect the
RH. Metastatic pattern of malignant melanoma: a study early radiographic response of melanoma brain
of 216 autopsy cases. Am J Surg. 1978;135(6):807–10. metastases to stereotactic radiosurgery. Cancer.
52. Davies MA, Liu P, McIntyre S, Kim KB, Papadopoulos 2016;122:3051–8.
N, Hwu W-J, et al. Prognostic factors for survival in 65. Patel KR, Shoukat S, Oliver DE, Chowdhary M,

melanoma patients with brain metastases. Cancer. Rizzo M, Lawson DH, et al. Ipilimumab and stereo-
2011;117(8):1687–96. tactic radiosurgery versus stereotactic radiosurgery
53. Acharya S, Mahmood M, Mullen D, Yang D, Tsien alone for newly diagnosed melanoma brain metasta-
CI, Huang J, et al. Distant intracranial failure in mela- ses. Am J Clin Oncol. 2017;40(5):444–50.
66. Davies MA, Saiad P, Robert C, Grob J-J, Flaherty KT, 80. SEER cancer statistics review, 1975–2015, National
Arance A, et al. Dabrafenib plus trametinib in patients Cancer Institute. Bethesda, MD [Internet]. 2018 [cited
with BRAF V600–mutant melanoma brain metastases December 15, 2018]. Available from: https://seer.can-
(COMBI-MB): a multi-cohort, open-label, phase 2 cer.gov/csr/1975_2015/.
trial. Lancet Oncol. 2017;18(7):863–73. 81. Tremont-Lukats IW, Bobustuc G, Lagos GK, Lolas
67. Schuch B, La Rochelle JC, Klatte T, Riggs SB, Liu W, K, Kyritsis AP, Puduvalli VK. Brain metastasis from
Kabbinavar FF, et al. Brain metastasis from renal cell prostate carcinoma: the M. D. Anderson Cancer
carcinoma. Cancer. 2008;113(7):1641–8. Center experience. Cancer. 2003;98(2):363–8.
68. Sun M, De Velasco G, Brastianos PK, Aizer AA, 82. Caffo O, Gernone A, Ortega C, Sava T, Carteni G,
Martin AM, Moreira R, et al. The development of Facchini G, et al. Central nervous system metasta-
brain metastases in patients with renal cell carcinoma: ses from castration-resistant prostate cancer in the
epidemiologic trends, survival, and clinical risk fac- docetaxel era. J Neurooncol. 2012;107:191–6.
tors using a population-based cohort. Eur Urol Focus. 83. Mahmoud-Ahmed AS, Suh JH, Kupelian PA, Klein
2018;S2405–4569(17):30294–8. EA, Peereboom DM, Dreicer R, et al. Brain metasta-
69. Massard C, Zonierek J, Gross-Goupil M, Fizazi K, ses from bladder carcinoma: presentation, treatment
Szczylik C, Escudier B. Incidence of brain metastases and survival. J Urol. 2002;167:2419–22.
in renal cell carcinoma treated with sorafenib. Ann 84.
Pietzner K, Oskay-Oezcelik G, Khalfaoui K,
Oncol. 2010;21:1027–31. Boehmer D, Lightenegger W, Sehouli J. Brain
70. Ranze O, Hofmann E, Distelrath A, Hoeffkes
metastases from epithelial ovarian cancer: over-
HG. Renal cell cancer presented with leptomenin- view and optimal management. Anticancer Res.
geal carcinomatosis effectively treated with sorafenib. 2009;29(7):2793–8.
Onkologie. 2007;30(8–9):450–1. 85. Pakneshan S, Safarpour D, Tavassoli F, Jabbari

71. Thibault F, Billemont B, Rixe O. Regression of brain B. Brain metastasis from ovarian cancer: a systematic
metastases of renal cell carcinoma with antiangio- review. J Neurooncol. 2014;119:1–6.
genic therapy. J Neurooncol. 2008;86(2):243–4. 86. Fetcko K, Gondim DD, Bonnin JM, Dey M. Cervical
72. Medioni J, Cojocarasu O, Belcaceres JL, Halimi P, cancer metastasis to the brain: a case report and
Oudard S. Complete cerebral response with suni- review of literature. Surg Neurol Int. 2017;8:181.
tinib for metastatic renal cell carcinoma. Ann Oncol. 87. Piura E, Piura B. Brain metastases from endometrial
2007;18(7):1282–3. carcinoma. ISRN Oncol. 2012;2012:581749.
73. Lauko A, Thapa B, Jia X, Ahluwalia MS. Efficacy of 88. Vrachimis A, Schmid KW, Jurgens H, Schober

immune checkpoint inhibitors in patients with brain O, Wekesser M, Riemann B. Cerebral metastases
metastasis from NSCLC, RCC, and melanoma. J Clin from thyroid carcinoma: complete remission fol-
Oncol. 2018;36(5_suppl):214. lowing radioiodine treatment. Dtsch Arztebl Int.
74. Christensen TD, Spindler K-LG, Palshof JA, Nielsen 2013;110(50):861–6.
DL. Systemic review: brain metastases from colorec- 89. Chiu AC, Delpassand ES, Sherman SI. Prognosis and
tal cancer — incidence and patient characteristics. treatment of brain metastases in thyroid carcinoma. J
BMC Cancer. 2016;16:260. Clin Endocrinol Metab. 1997;82(11):3637–42.
75. Grothey A, Sargent D, Goldberg RM, Schmoll
90. Pelletier EM, Shim B, Goodman S, Amonkar

HJ. Survival of patients with advanced colorectal MM. Epidemiology and economic burden of brain
cancer improves with the availability of fluorouracil- metastases among patients with primary breast can-
leucovorin, irinotecan, and oxaliplatin in the course of cer: results from a US claims data analysis. Breast
treatment. J Clin Oncol. 2004;22(7):1209–14. Cancer Res Treat. 2008;108:297–305.
76. Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, 91. Guerin A, Sasane M, Dea K, Zhang J, Culver K,
Hainsworth J, Heim W, et al. Bevacizumab plus irinote- Nitulescu R, et al. The economic burden of brain
can, fluorouracil, and leucovorin for metastatic colorec- metastasis among lung cancer patients in the United
tal cancer. N Engl J Med. 2004;350(23):2335–42. States. J Med Econ. 2016;19(5):526–36.
77. Sundermeyer ML, Meropol NJ, Rogatko A, Wang 92. Guerin A, Sasane M, Zhang J, Culver KW, Dea K,
H, Cohen SJ. Changing patterns of bone and brain Nitulescu R. Brain metastases in patients with ALK+
metastases in patients with colorectal cancer. Clin non-small cell lung cancer: clinical symptoms, treat-
Colorectal Cancer. 2005;5(2):108–13. ment patterns and economic burden. J Med Econ.
78. Yaeger R, Cowell E, Chou JF, Gewirtz AN, Borsu 2015;18(4):312–22.
L, Vakiani E, et al. RAS mutations affect pat- 93. Vekeman F, Cloutier M, Yermakov S, Amonkar MM,
tern of metastatic spread and increase propensity Arondekar B, Duh MS. Economic burden of brain
for brain metastasis in colorectal cancer. Cancer. metastases among patients with metastatic melanoma
2014;121(8):1195–203. in a USA managed care population. Melanoma Res.
79. Wong MCS, Goggins WB, Wang HHX, Fung FDH, 2014;24:602–10.
Leung C, Wong SYS, et al. Global incidence and 94. Peters S, Bexelius C, Munk V, Leighl N. The impact
mortality for prostate cancer: analysis of tempo- of brain metastasis on quality of life, resource utiliza-
ral patterns and trends in 36 countries. Eur Urol. tion and survival in patients with non-small-cell lung
2016;70(5):862–74. cancer. Cancer Treat Rev. 2016;45:139–62.
Basic Biology of Brain Metastasis
2
Monika Vishnoi, Robert A. Scranton,
Samuel K. Asante, and Robert C. Rostomily
Table 2.1 SEER-based incidence of brain metastases

Introduction and Epidemiology determined by primary tumor site
Estimated new Estimated
Brain and other nervous system cancers are Primary cancer site cases, 2018 deaths, 2018
extremely lethal and represent 1.4% of all new Breast cancer 266,120 40,920
cancer cases in the United States (SEER 2018, (female)
Lung and bronchus 234,030 154,050
2008–2014). In 2018, there were an estimated
cancer
23,380 new cases of brain and other nervous sys- Prostate cancer 164,690 29,430
tem cancers and 16,380 estimated cancer deaths. Colorectal cancer 140,250 50,630
Brain cancer (BC) is divided into two different Melanoma of the 91,270 9320
types depending on origin site: (1) primary cancer, skin
confined to the brain; (2) secondary cancer, metas- Bladder cancer 81,190 17,240
Non-Hodgkin 74,680 19,910
tasized to the brain from a different primary site.
lymphoma
Secondary brain tumors are extremely aggressive Kidney and renal 65,340 14,970
and about 30–40% of cancer patients with primary pelvis cancer
tumor (melanoma, breast, lung, etc.) have been Uterine cancer 63,230 11,350
diagnosed with brain metastasis (BM) at some Leukemia 60,300 24,370
stage after initial cancer diagnosis (Table 2.1).
Lung and breast cancer are the most frequent can-
and anaplastic lymphoma kinase (ALK) positivity
cers that metastasize to the brain in men and
in non-small-cell lung cancer (NSCLC) carry a
women respectively [1]. Certain molecular sub-
higher rate of brain metastasis [2, 3]. Other factors
types such as HER2 amplification in breast cancer
associated with incidence of brain metastasis include
age, ethnicity, and geographic location [1].
Among all primary cancer types, the overall
M. Vishnoi · R. C. Rostomily (*)
Department of Neurosurgery, Houston Methodist 2-year and 5-year survival rates for brain meta-
Research Institute, Houston, TX, USA static patients are 8.1% and 2.4%, respectively [1,
e-mail: rrostomily@houstonmethodist.org 4]. The therapeutic management of brain metasta-
R. A. Scranton sis depends on the number and location of meta-
Department of Neurosurgery, Goodman Campbell static tumors, and can include whole brain
Brain and Spine, Indiana University, radiation therapy (WBRT), surgical resection, ste-
Indianapolis, IN, USA
reotactic radiosurgery, systematic chemotherapy,
S. K. Asante targeted therapy, and immunotherapy. Poor
Department of Neurosurgery, Houston Methodist
Hospital, Houston, TX, USA blood-brain barrier (BBB) permeability can

https://doi.org/10.1007/978-3-030-42958-4_2
20 M. Vishnoi et al.
limit the effectiveness of systemic chemother- Seed and Soil

apy to effectively treat brain metastasis [5].
Combined targeted and immunotherapeutic The realization that the profile of metastatic sites
approaches can produce shrinkage of brain did not reflect a stochastic distribution of the seed
metastasis, can slow tumor growth, and can pre- based on passive blood flow and target organ
vent or/delay neurologic symptoms [1]. In brain mass suggested that specific cellular and molecu-
metastatic clinical trials, multimodal combina- lar mechanisms were actively involved in regu-
tion therapies provide more survival benefit to lating metastasis. Stephen Pagett first proposed
patients than individual treatments; however, that this phenomenon is governed by the unique
posttreatment toxicity may adversely affect match of the metastatic cancer seed with a condu-
patients’ quality of life (QOL) [1, 6]. Therefore, cive soil—the “seed and soil” hypothesis [7]. To
it is imperative to understand the complexity of achieve successful metastasis, cancer cells must
the brain metastatic cascade and translate these shed from their primary site, survive and self-
findings in clinical settings to develop effective renew in the circulation (blood/lymph), intrava-
therapies with the ultimate goal of increasing sate, and colonize distant organs where they must
patients’ QOL and survival. In this review, we survive and grow (Fig. 2.1) [8–10].
will discuss the molecular and genetic proper- The heterogeneous populations of tumor cells
ties of the tumor progenitor cells responsible for that comprise the primary tumor possess distinct
brain metastatic seeding and colonization, as molecular and cellular phenotypes evidenced by
well as their modulation through tumor-host their differential proliferative, invasive, angio-
niche interactions, the neuro-inflammatory cas- genic, and metastatic abilities [11, 12]. The meta-
cade, and neovascularization. static cascade exerts further selection pressure
Primary tumor
Early Late Brain metastasis
dissemination dissemination
Colonization
EMT
migration
Primary Asymptomatic
etc.
dissemination progression Metastatic Seeding
dissemination
MET
Intravasation Extravasation
cf DNA
Exosomes
Homotypic CTC clusters Single CTC Heterotypic CTC clusters
Endothelial cell Fibroblast/mesenchymal cell Monocytes/eosinophil

Epithelial cell Circulating tumor cell Pericyte
Red blood cell Macrophages Circulating free DNA
Extracellular matrix Lymph node

Exosomes/vesicles Bone marrow
Blood vessel Brain
Fig. 2.1 Steps of brain metastatic cascade

2 Basic Biology of Brain Metastasis 21
that influences cell proliferation, quiescence, described by Elizabeth Hay in the context of
adhesion, invasiveness, plasticity, cell-surface embryogenesis [17].
(growth and hormone) receptors, and immunoge- At the molecular level, EMT is driven by tran-
nicity that ultimately defines the metastatic scription factors such as ZEB1/2, SNAIL, SLUG,
potential [8, 9]. The metastatic “seeds” mobilize and TWIST1 and signaling through HGF, TGF-β,
and invade the lymphatic or vasculature system EGF, PDGF, Notch1, Wnt, PI3k/AKT, and
where they disseminate as single cells or cell Hedgehog pathways that together promote motil-
clusters (tumor emboli) [9, 13, 14]. These cells ity, migration, and invasion of tumor cells
then are often home to, and interact with, condu- (Fig. 2.2) [18–25]. For example, TWIST1 is criti-
cive microenvironments at distant organs (soil) cal for mammary epithelial carcinoma cell
where stromal and host factors govern their colo- extravasation and is implicated in metastatic
nization, survival, and growth. Therefore, organ- capacity for numerous cancers [26].
specific colonization and macro-metastases Downregulation of E-cadherin (an epithelial cel-
formation are highly complex processes that lular adhesion protein) and upregulation of
depend upon specific “homeostatic mechanisms” N-cadherin (the so-called “cadherin switch”)
and interactions with extracellular matrix (ECM) accompany EMT allowing a cell typically held in
proteins and cells (immune, stromal, fibroblasts tight apposition to become mobile and correlate
cells, etc.) that comprise the target organ micro- with the metastatic potential of cancers metasta-
environmental niche [8, 9, 15, 16]. The unique sizing to the brain [27, 28]. This is followed by
properties of the brain environment (BBB and degradation of the epithelial basement membrane
neural niches) and how they impact BM forma- and invasion through the endothelial basement
tion and growth are discussed more in detail membrane, and then transit into the blood vessel
below. [29–31]. In addition to promoting invasiveness,
EMT promotes malignant phenotypes through
effects on immunosuppression, treatment resis-
he Early Dissemination Phase
T tance, and cancer stem cells (CSCs) [25].
of Brain Metastasis Therefore, EMT contributes broadly to BM for-
mation through production of metastatic “seeds,”
Epithelial-Mesenchymal activation of malignant cellular properties, and
and Mesenchymal-Epithelial reprogramming of the tumor microenvironment.
Transitions: EMT/MET To successfully generate metastases, dissemi-
Implied in the “seed and soil” hypothesis and nated tumor cells must survive in the blood
critical to the manifestation of a metastasis is the stream (see below), extravasate from the circula-
capacity of a primary cancer cell or aggregate to tion, colonize, and grow in distant organs. Despite
navigate an imposing biological gauntlet, roughly the presumed importance of EMT in the initial
divided into discrete stages: (i) separation from dissemination of metastatic cancer cells, meta-
the primary tumor mass and invasion into and static tumors frequently retain epithelial features
survival in the blood stream (intravasation), (ii) of the primary tumor [25]. These conflicting
exit from the blood stream to achieve coloniza- observations are reconciled by recognizing that
tion within a distant organ (extravasation), and the final phase of the metastatic cascade (extrava-
(iii) survival and growth in a distant organ sation, colonization, and macro-metastatic
(Fig. 2.1). A major advance in the conceptualiza- growth) requires reversal of the mesenchymal to
tion of metastasis came with the realization that epithelial phenotype; a process termed the
the cellular phenotypes required of the first stage mesenchymal-epithelial transition (MET) [32–34].
(intravasation) of the metastatic cascade recapit- Activators of EMT signaling are lacking at sites
ulated the features of a developmentally and mor- of metastatic colonization, including the brain,
phogenetically recognized phenomenon termed which promotes MET and macro- metastatic
epithelial mesenchymal transition (EMT) first growth [8, 25, 35]. This cross-talk between
Primary tumor site Brain tumor site

Signaling pathways
HGF, TGF-β, EGF, PDGF
Notch1, Wnt, Pl3k/AKT
Hedghog
EMT MET
Hybrid E/M state
Migration
Epithelial markers Transcription factors miRNA Stem-cell markers Mesenchymal markers

E-cadherin Twist1/2 mir-21 Nanog N-cadherin
Claudin Snail mir-200 Sox-2 Fibronectin
Occludin Zeb1/2 mir-205 Oct4 FSP-1
Desmoplakin Slug mir-34a LGR5 Vimentin
Mucin-1 bHLH mir-145 CD133 Vitronectin
Cytokeratin β-catenin Let-7 SMA
FGFR2
MMPs
Fig. 2.2 Molecular mechanisms of epithelial (E)-mesenchymal (M) transition during brain metastasis
extravasated cancer cells and the microenviron- Cancer stem cells (CSCs) are operationally
ment of distant organs underscores the impor- defined by properties of proliferation, self-
tance of the metastatic niche or “soil” for renewal, multi-lineage differentiation, and,
successful generation of metastases. Here, we importantly, the capacity to recapitulate the can-
will discuss the mechanism of cross-talk between cer phenotype in vivo [37]. The correlation of
metastatic cancer cells and the brain that specifi- specific molecular markers with CSC phenotypes
cally contributes to BM formation. While the pre- has facilitated the investigation of the role of
cise mechanisms of EMT and MET that regulate CSCs in BMs. For instance, in breast cancer the
BM formation have yet to be elucidated, it is sig- CD44 hi/CD24 low CSC phenotype is responsi-
nificant that both processes promote the pheno- ble for maintaining self-renewal and proliferation
types of CSCs, the putative “seeds” for BM through Notch signaling and drives metastatic
formation [7, 25]. progression in the brain [22, 38, 39]. On the other
hand, the chemokine CXCR4/12 signaling axis
ancer Stem Cells
C provides microenvironment cues to CSCs for
Many cancers possess a subpopulation of cancer proper homing and brain colonization [40]. Of
stem cells (CSCs) that play critical roles in note, targeting CSC phenotypes CD44 hi/CD24,
tumorigenesis, treatment resistance, and progres- CD133, and BMI1 or inhibiting CXCR4/12 and
sion and are commonly considered the “seed” for the Notch signaling axis effectively eradicates
metastasis [11]. Although they comprise a minor- brain metastatic spread and improves therapeutic
ity population of tumor cells, CSCs are of great efficacy [39, 41, 42]. In concert with MET and
clinical importance by virtue of their increased local angiogenesis, they drive growth of macro-
resistance to treatment and putative role in for- scopic brain tumors [43, 44]. Another critical
mation and/or growth of BMs (Fig. 2.3) [36]. observation is that cancer cells can switch
Localized tumor Residual disease Micrometastasis Macrometastasis Brain tumor relapse
Secondary
Primary resistance Genotypic/ resistance
Survival, renewal phenotypic Resistant tumor
EMT/stem cell & proliferation alterations cell progression
First-line therapy Asymptomatic progression Second-line therapy
Tissue biopsy
Tissue biopsy Personalized
Clinical imaging/
- Molecular profiling Serial liquid biopsy screening therapy
radiation
- Immunohistochemistry
Liquid biopsy 3D organoids
- Candidate drug targets
- Molecular profiling & clonal evolution

- Early-cancer screening - Prognostic & diagnostic biomarkers Microfluidic device
- Prognostic & diagnostic - Monitoring clinical remission & therapy Surgical resection Liquid biopsy
biomarkers Systemic resistance Monitoring secondary
localized therapy Xenografts clinical remission
Fig. 2.3 Molecular landscape of brain metastasis and its therapeutic implications
between non-CSC and CSC phenotypes in CTCs may also derive not only from the primary
response to microenvironmental cues such as tumor site but also from distant macro-
hypoxia [44–46]. This plasticity has profound metastases—a mechanism termed “self-seeding”
implications for context-dependent identification (see Fig. 2.1).
and assessment of CSC burden and development CTCs are a minority heterogeneous cancer
of CSC-targeted therapies. Regardless of how cell population that can be isolated from patient
metastatic cells acquire stem-like properties, they blood by various techniques such as flow cytom-
must all navigate and survive a journey through etry, magnetic beads, and microfluidic devices
the blood stream. The recent refinement of meth- and identified based on immune phenotyping,
ods to identify and characterize circulating tumor cell size, and deformability [48–51].
cells (CTC) has shed further light on the pheno- CellSearch™, which captures EpCAM-positive
types and mechanisms employed by CSCs to epithelial-derived CTCs, is currently the only
colonize the brain (Fig. 2.3). Food and Drug Administration (FDA)-approved
platform for CTC analysis, although it excludes
potential EpCAM-negative CTCs that may be a
he “Liquid” Phase of Brain
T significant contributor to BM formation [39].
Metastasis Disseminated CTCs intravasate and migrate into
the blood circulation and survive as single cell or
irculating Tumor Cells (CTCs)
C cluster/emboli. CTC clusters have survival
and Dormant Cancer Cells (DCCs) advantages as they more effectively resist anoi-
As putative metastatic “seeds,” circulating tumor kis, bloodstream shear forces, environmental or
cells (CTCs) drive metastasis and disease recur- oxidative stresses, and immune surveillance [52–
rence [47, 48]. CTCs may colonize distant sites 54]. Higher CTC counts in peripheral blood
and rapidly progress to form macro-metastases or correlate with disease burden and worse patient
remain as dormant cancer cells (DCCs) in per- survival in various malignancies such as mela-
missive pre-metastatic niches that after months or noma and breast, lung, prostate, and pancreatic
years are triggered to form macro-metastases. cancers [54–56].
Although CTCs are not routinely identified Exosomes

in the majority of BM patients, BM formation Exosomes are small membrane bound extracel-
presumably requires the existence of CTCs at lular vesicles secreted by cancer cells that contain
some point prior to their clinical manifestation. DNA, RNA, proteins, and lipids [62]. Exosomes
More than two CTCs were detected in only function locally within primary and metastatic
5.9% of patients with oligo-metastatic NSCLC tumors as well as remotely through vascular dis-
to the brain; the frequency of more than three semination and cellular uptake at metastatic sites
CTCs in BM patients with systemic metastases [63]. They are increasingly analyzed in liquid
and other tumor types ranges widely from 0% to biopsies since they inform tumor growth, evolu-
25% [55]. These data underscore several impor- tion, and pathogenesis and in BMs are responsi-
tant considerations: (i) CTC dissemination and ble for inducing a plethora of biological
thus detection may be intermittent with periods processes, such as EMT, angiogenesis, metasta-
of dormant residence in other sites, and (ii) in sis, therapy resistance, and epigenetic/stem-cell
addition to unidirectional production of CTCs regulation (Figs. 2.1 and 2.3) [64]. In breast can-
from the primary cancer, CTCs may also arise cer, expressions of mir-122 and mir-210 were
from metastatic deposits including BMs, the so- associated with brain metastasis [65, 66]. In mel-
called “self-seeding” mechanism. The identifi- anoma, CD46 receptors are responsible for
cation of CTCs in cerebrospinal fluid (CSF) uptake of tumor-associated exosomes in BBB
from BM patients with concurrent leptomenin- endothelial cells [67].
geal disease (LMD) may represent a form of An important function of exosomes in BM
“self-seeding” [57]. biology is their capacity to generate organotropic
CTCs are heterogeneous and specific subpop- pre-metastatic niches conducive to DCC growth
ulations may have unique tropism for colonizing or CTC homing, colonization, and proliferation.
the brain [9, 48, 58, 59]. Using an expanded CTC In experimental studies, the brain preferentially
isolation protocol, Boral et al. demonstrated that takes up exosomes from neurotropic metastatic
inclusion of EpCAM-negative CTCs with CSC cancer cell lines through specific direct interac-
markers markedly increased CTC yields in breast tions with CD31+ BBB endothelial cells [68].
carcinoma patients [60]. Stratifying these meta- Exosomal organotropism also appears to be
static patients based on the presence of BMs, they related to specific integrin profiles with ITGB3
identified a 121-gene signature associated with highly upregulated in brain tropic exosomes [69].
BMs [60]. Other studies have shown that a spe- Remarkably, “educating” mouse hosts with exo-
cific subpopulation of EpCAM-negative CTCs somes redirects the organotropism of cancer cell
from breast cancer patients has a unique propen- lines to reflect patterns of exosomal uptake.
sity for forming BMs in experimental models These data suggest that targeting brain-specific
[61]. These studies indicate that identifiable sub- exosomes may be a useful future strategy to miti-
populations of CTCs may have specific capacity gate BM formation.
to generate BMs that could theoretically be tar- In addition to roles in organotropism, exo-
geted systemically to prevent BMs. Clinical and somes are implicated in promoting immunosup-
biological relevance of CTCs is an area of ongo- pressive “havens” for DCCs, angiogenesis that
ing investigation, but one that appears to have triggers progression and growth of micrometasta-
great promise to inform prognosis, treatment ses, and disruption of the BBB [62, 63, 70]. Of
responses, metastatic risk, and even new thera- note, experimental evidence implicates exosomal
peutic approaches. The utility of liquid biopsies microRNAs (miRNAs) generated from astro-
for BMs extends beyond the detection of CTCs, cytes and BM cells during tumor growth through
allowing profiling of exosomes and circulating- reversible epigenetic downregulation of PTEN
tumor DNA (ctDNA) (see Fig. 2.3). and conversion of resident microglia from the
M1 to M2 immunosuppressive phenotype [71, he Final Metastatic Phase: Brain

T
72]. These observations demonstrate the impor- Colonization, Growth, and the Role
tant roles of both primary tumor-derived and of the Brain Microenvironment
local neural cell-derived exosomes in orchestrat-
ing the complex processes of BM tropism and In the final stage of BM formation, CTCs and/or
growth. Exosome-based targeted therapies there- DCCs, which have colonized permissive pre-
fore may be a useful strategy to mitigate BM for- metastatic niches, engage the brain microenvi-
mation and progression [63]. ronment (BME) and through reciprocal
interactions undergo macro-metastatic growth.
ctDNA The complex interactions between BM cells and
Circulating-tumor DNA (ctDNA) is released resident neural cells (astrocytes, neurons, and
into biological fluids by apoptotic or necrotic microglia), infiltrative immune cells, brain micro-
cancer cells. ctDNA is detected in most sys- vasculature, extracellular matrix proteins, meta-
temic cancers with increased levels correspond- bolic changes, cytokine signaling, and even
ing with metastasis [73]. In breast and melanoma synaptic inputs result in reprogramming of BM
carcinomas, two cancers with high propensity cells and BME to facilitate BM survival and
for BMs, ctDNA is detectable in over 80% of growth. An additional important element of the
cases [73]. Levels of ctDNA are associated with BME is the blood-brain barrier (BBB) and subse-
tumor burden and patient survival [64, 74, 75]. quent formation of a blood-tumor barrier (BTB),
To our knowledge no data exist demonstrating critical to the process of CTC extravasation,
an association between ctDNA and BM inci- immune cell infiltration, and systemic delivery of
dence or a pathogenic role in BM genesis. therapeutic agents. With selected examples we
However, analysis of ctDNA from cerebrospinal will address clinically relevant highlights of these
fluid (CSF) is emerging as a useful marker for interactions.
patients with parenchymal BMs and leptomen-
ingeal disease that may be more sensitive and he Blood-Brain Barrier (BBB)
T
specific than plasma-derived ctDNA [76, 77]. and Blood-Tumor Barrier (BTB)
For instance, in patients with central nervous The blood-brain barrier (BBB) is a highly spe-
system (CNS)-restricted metastatic disease, cialized semipermeable structure consisting of
CSF ctDNA was detected in 58% versus 0% endothelial cells, pericytes, and astrocytes, which
from plasma and importantly, changes in CSF form tight junctions that restrict access to the
ctDNA detection corresponded with clinical brain from the circulation [1]. The neurovascular
treatment responses [77]. In another study, unit of the BBB maintains homeostatic environ-
genomic mutations were identified after mental conditions for normal neuronal function
sequencing of CSF DNA in 63% (20 of 32) of and provides a barrier to CTC extravasation that
patients with parenchymal CNS metastases, must be overcome for BM formation (reviewed
while detection of ctDNA in CSF has been in Ref. [81]). Three molecules, cyclooxygenase
reported for 75–100% of patients with LMD COX2 (also known as PTGS2), the epidermal
[76, 78]. These studies indicate the potential for growth factor receptor (EGFR) ligand HBEGF,
ctDNA to serve as a biomarker for tracking and the α2,6-sialyltransferase ST6GALNAC5
tumor progression and treatment response [75, have been identified as mediators of cancer
79, 80]. Multicenter large cohort studies are cell extravasation across the BBB [82].
required to evaluate the evolutionary changes of ST6GALNAC5 promotes adhesion of tumor
ctDNA over the course of metastatic cancer cells to brain endothelial cells, whereas COX2
treatment and their correlation with disease sta- and HBEGF promote cell migration across the
tus and patient survival. BBB [82]. In addition, matrix metalloproteinases
(MMPs) and vascular endothelial growth factor typically decrease cytotoxic T-lymphocyte (CTL)
(VEGF) facilitate extravasation, seeding, and function [105].
micrometastasis formation through ECM destruc- Harter et al. profiled the quantity and topog-
tion and creation of a vascular niche [1, 83–87]. raphy of all TILs (CD3+) and specific subpopu-
In the BM peri-tumoral region, the BBB is lations of T-reg cells (FoxP3+) and CTLs
modified to generate a so-called blood-tumor (CD8+) and PD-1/PD-L1 expression in BMs in
barrier (BTB) characterized by increased local both mixed tumor and breast carcinoma-
permeability. Changes in BBB characteristic of restricted cohorts [105]. TILs and their subpop-
the BTB are mediated by alterations in endothe- ulations were detected in all BM types but with
lial cell tight junctions and pericyte function, different frequencies (highest in renal cell carci-
and are associated with neuroinflammation and noma) and patterns of distribution (diffuse in
changes in ECM components [1]. The molecu- melanoma, stromal in carcinomas). In contrast
lar mechanisms underlying permeability to other studies, where expansion of cytotoxic
changes in the BTB include upregulation of T-lymphocytes and infiltration of T-cells corre-
VEGF and downregulation of zona occludens late with patient survival, none of the TIL or
(ZO) and vascular endothelial cell adhesion PD-L1/PD-1 metrics were associated with
molecule (VE-CAM) in endothelial cells, patient survival [107–109]. By contrast, the
altered expression of desmin and CD13 in peri- presence of a peri-tumoral and to a lesser extent
cytes, and elaboration of other molecules stromal mononuclear infiltrate and lower PD-1/
including membrane transporters, tumor necro- PD-L1 expression in lung adenocarcinoma BM
sis factor (TNF) receptors, claudin-5, and angio- patients predicted better survival after resection
poietin-2 [1, 88–92]. Of clinical relevance, these [110]. In another study of NSCLC patients, dis-
changes in permeability result in heterogeneous parate responses of the primary and BM lesions
uptake that may enhance uptake of drugs and to PD-1 blockade mirrored a decrease in
antibodies normally restricted by the intact BBB BM-specific PD-1 expression in paired primary
[1, 54, 93–96]. and BM samples [102]. In paired breast cancer
primary and BM samples, TILs are decreased in
Immune BM Microenvironment BMs as was the proportion of “adaptive”
BMs generate an inflammatory and immunosup- immune phenotypes (TIL+/PD-L1+) expected
pressive microenvironment that promotes tumor to be responsive to ICIs [103, 104, 111]. In mel-
growth and treatment resistance [97]. The anoma BMs, increased immune cell infiltration
immune BM microenvironment involves com- corresponded with increased PD-L1+, survival,
plex interactions between tumor and resident and enrichment of oxidative phosphorylation
neural cells and infiltrating cells of lymphoid compared with non-CNS metastases [100].
(cytotoxic-CD4+, helper-CD4+, T-regulatory Overall the melanoma BMs had reduced
[T-reg] cells, and natural killer) and myeloid immune cell infiltrates, and gene expression
(dendritic/antigen presenting cells, macrophages, analysis revealed an immunosuppressive pheno-
and myeloid-derived suppressor cells [MDSCs]) type compared with non-CNS metastases. Of
lineage [98–101]. Intense interest in tumor- note, the metabolic signature positively corre-
infiltrating lymphocytes (TILs) has been fostered lated with patient survival, and preclinical mod-
by the success of immune checkpoint inhibitors els demonstrated that inhibition of oxidative
(ICIs) in treatment of systemic melanoma, and, phosphorylation was a promising therapeutic
more recently, other cancers with a propensity for target of MAPK-resistant melanoma BMs [100].
BMs including NSCLC and breast cancer [102– In addition to TILs, other immune cells includ-
105]. In fact, recent trials have demonstrated ing myeloid cells are implicated in BM growth
variable activity of ICIs against BMs [98, 106]. [72]. The association between reduction of
CTLA4 and PD-L1/PD-1 inhibitors block tumor- peripheral myeloid-derived suppressor cells
mediated immunosuppressive mechanisms that (MDSCs) and BM incidence in lung cancer
patients treated with combination systemic beva- occurrence of an abscopal response is relatively
cizumab and TKIs suggests that MDSCs may rare, further investigations into its precise mech-
play a role in the immunosuppressive BM anisms are expected to provide insight into more
microenvironment. Experimental studies of
effective strategies to activate the immune sys-
mouse mammary carcinoma BMs also demon- tem to improve response for systemic and CNS-
strated that MDSCs generate a “pre-metastatic based cancers.
niche” for BM formation [90]. T-regulatory
(T-reg) cells suppress immune reactions by secre- M Cross-Talk with the Brain Metastasis
B
tion of factors such as TGF-β and IL-10 and Microenvironment
higher T-reg cell burden in tumors and peripheral In addition to interactions with the vascular
blood is associated with poor clinical outcomes BBB/BTB niche and infiltrating immune cells
[112, 113]. In lung adenocarcinoma, FOXP3+ described above, cross-talk with resident neural
T-reg cells are detected in BMs albeit at lower cells also plays an important role in BM biology
numbers than in primary tumors [102]. Finally, (reviewed in Refs. [97, 99, 124]). The brain is
resident microglia and systemically derived mac- generally a hostile microenvironment for extrav-
rophages are implicated in early stages of BM asated cancer cells, the majority of which die;
formation and contribute to the immunosuppres- however, those that survive as dormant or
sive microenvironment [114]. In summary, the actively propagating cells appear uniquely able
complex and immune BM microenvironment to co-opt or adapt to the conditions in the brain
generates an immunosuppressive state and plays microenvironment [81]. For instance, cancer
critical roles in BM formation from the pre-met- cells that grow in the brain activate unique brain-
astatic niches to macro-metastatic growth. enriched gene expression profiles, and undergo
Further elucidation of the diversity of immu- metabolic reprogramming so that they can effec-
nosuppressive mechanisms in BMs is needed to tively utilize non-glucose energy sources, like
develop more effective immunotherapy and the brain (reviewed in Refs. [125, 126]).
strategies to reprogram the immune microenvi- Expression of Serpins on BM cells counteracts
ronment of BMs to facilitate responses to with the cell-death and anti-migratory effects of
immunotherapy. brain-derived plasmin necessary for BM cell sur-
As noted above, BMs are “cold tumors” and vival and engagement with brain microvascular
thereby less responsive to immunotherapy [1]. cells for local invasion [127]. While neural cells
Therefore, techniques to activate the immune can impede BM growth, specific interactions
microenvironment in BMs have great clinical with neural cells have also been shown to pro-
significance. For example, the abscopal effect is mote BM survival and growth. For instance,
a presumed immune-mediated mechanism astrocyte interactions promote BMs through gap
whereby local radiation to a single lesion result- junction-mediated transfer of cGAMP and astro-
ing in release of tumor antigens and T-cell expan- cyte-derived exosomal miRNA-mediated sup-
sion can activate a dramatic generalized pression of PTEN function [72, 128]. Similarly,
antitumor response distant from the site of radia- the activated state of microglia can either inhibit
tion [115–117]. An experimental melanoma BM or promote BM growth [81]. BM cell secreted
model demonstrated an abscopal effect with exosomal miRNAs can reprogram microglia to
combined irradiation and PD-L1 blockade simi- promote BM growth through immunosuppres-
lar to the reported clinical potentiation of the sive mechanisms [71]. Finally, based on the
abscopal effect with concurrent ICI therapy increasingly recognized impact of peripheral
[118–121]. As several reports suggest, it may be innervation in cancer metastasis and CNS neural
possible to harness the abscopal effect to treat activity to promote glioma proliferation, future
BMs through targeting a systemic lesion or con- studies should be directed to understanding how
versely activate a systemic response through electrical activity may influence BM physiology
local irradiation of BMs [122, 123]. While the [129–131].
Molecular Heterogeneity shared by paired primary and BM specimens,

and Selection for Brain Metastasis these observations suggest that clonal selection
during BM formation may be required for effec-
Given the complexity of mechanisms and envi- tive metastatic outgrowth and therapeutic resis-
ronmental selection pressures summarized above, tance [133, 138, 142].
it is not surprising that primary cancers and their Activation of specific oncogenic signaling
brain metastases exhibit extensive molecular het- pathways occurs in BMs in concert with the evo-
erogeneity. Since the seminal publication by lution of genomic changes. In primary mela-
Gerlinger et al. in metastatic renal cell carcinoma, noma, lung and breast cancer patients, more than
intra-tumor molecular heterogeneity and 50% of brain metastatic tissue contain clinically
branched evolution have been recognized to con- relevant oncogenic alterations in PTEN, PIK3CA,
tribute to the genesis, progression, and treatment EGFR, and HER2 genes and cancer hot spot
resistance of many cancers [5, 132–134]. regions that activate PI3K–AKT–MTOR and
Genomic instability and selective evolution are EGFR/HER2 pathways involved in tumor cell
principle mechanisms driving heterogeneity at growth and proliferation [5, 132]. Primary tumors
the genetic, epigenetic, and transcriptional levels treated with systemic therapy such as PI3K/AKT/
[132, 133, 135]. Multiregional tumor biopsy mTOR, CDK, and HER2/EGFR inhibitors are
sampling, research autopsies, spatial and tempo- more inclined to develop brain metastasis [138].
ral liquid biopsies, and single-cell sequencing are In squamous cell lung cancers (SQCLC), PI3K-
emerging approaches that will help decode the aberrant tumors were associated with high meta-
complex architecture of tumor, specifically as it static tumor burden and increased incidence of
relates to brain metastasis [132, 133, 135–137]. brain metastasis [143]. However, colorectal can-
For BMs, studies of paired primary and meta- cer (CRC) shows less genetic heterogeneity
static lesions reveal several clinically relevant (APC, KRAS, FBXW7, PIK3CA, BRAF,
insights including (i) a high proportion of BMs SMAD4, and ACVR2A mutations) with greater
possess mutations distinct from those of the pri- genetic concordance between matched primary
mary site, (ii) BMs from individual patients share and brain metastatic tumors [144].
mutations distinct from those detected in the pri- Overall, brain metastases exhibit a branched
mary cancer, and (iii) BMs exhibit activation of evolution pattern reflecting primary tumor muta-
oncogenic signaling pathways (e.g., PI3K/Akt/ tion profiles and acquisition of additional unique
mTOR) distinct from those present in the primary molecular profiles with respect to other non-CNS
cancer [5, 138–140]. These observations suggest metastases. Additionally, molecular profiles of
that BMs may arise from unique cell subpopula- intracranial sites within individual patients sug-
tions within the original cancer and/or that gest a high degree of homogeneity. This genomic
selection pressure for specific mutations and phe- concordance may provide guidance for system-
notypes drive successful BM formation and atic personalized therapy and facilitate our
growth. understanding of mechanisms involved in brain
Genomic studies indicate that BMs retain metastasis.
ancestral mutations of their primary cancer but
acquire additional unique mutations through
branched evolution [136, 138, 139, 141]. In the Spinal Metastasis
largest study to date of paired primary and meta-
static cancer samples, Brastianos et al. deter- Metastatic spinal cord compression is considered
mined that BMs share mutations with the primary an oncological emergency that may require
cancer but develop unique or “private mutations” immediate treatment either through surgical
in all cases, of which 53% represent potential decompression, emergency radiotherapy, or a
actionable targets unique to their CNS disease combination of the two. This occurs in 3–5% of
[138]. As further shown by EGFR mutations cancer patients, with breast, lung, and prostate
being the most frequent source [145]. The major- [147, 157]. Anaplastic lymphoma kinase (ALK)
ity of metastases affects the bone first and cause gene mutations are associated with aggressive
compression through direct mass effect or patho- features in NSCLC including early CNS metas-
logical fracture. Even more rare are intradural tasis and higher rates of intramedullary spinal
extramedullary and intramedullary metastases cord metastasis [146, 157, 163]. While rare, the
accounting for less than 6% and 1–2% of spinal consequences of spinal extra- and intramedul-
metastasis, respectively [146–148]. The inci- lary metastases are devastating and warrant fur-
dence of intramedullary spinal metastasis may be ther study of their basic biology to develop more
increasing perhaps with extended overall sur- effective therapies. See the “Spinal Metastases”
vival. Additionally, metastasis to the spine is gen- section of this book for in-depth coverage of this
erally a poor prognostic sign of overall patient topic.
survival with median survival of only 8 months in
patients treated for intramedullary renal cell
metastasis [149]. Conclusion
By virtue of their different recipient tissue
microenvironments, it is not surprising that the Brain metastasis is a devastating disease with
cellular and molecular mechanisms that pro- increasing incidence. The increased rate is due
mote bone metastasis, and thereby osseous- to a lack of prognostic and diagnostic biomark-
based spinal cord compression, differ from ers at early disease stages. Systemic, longitudi-
those driving BMs (reviewed in Refs. [150–153]). nal blood-based liquid biopsy (CTCs, cell-free
Given the rarity of both extra- and intra-axial DNA, exosomes, secretory proteins, etc.),
spine metastasis, studies of their specific mech- alongside molecular imaging approaches, may
anisms are scarce. Presumably, extramedullary provide novel biomarkers for designing early
spinal metastases result from local leptomenin- diagnostic tools (see Fig. 2.3). In brain meta-
geal growth of CSF-disseminated cells. Like static patients, surgical resection is a key part
BMs, molecular analysis of leptomeningeal of clinical management and provides an imme-
cancer cells reveals mutations shared with and diate opportunity for tumor molecular charac-
unique to the primary cancer site that can be terization for determining effective therapies.
monitored through analysis of ctDNA [154, These studies can also assist in identifying
155]. By contrast, intramedullary spinal metas- therapeutic targets to eliminate residual dis-
tases are more likely to originate through mech- ease or recurrence in brain metastatic patients
anisms similar to those that regulate BMs. with other primary cancers. Poor prognosis of
Intramedullary spinal cord metastasis (ISCM) is brain metastatic patients is also related to drug
exceedingly rare with incidence of ~2% in sys- resistance and tumor heterogeneity between
temic cancers [147, 156, 157]. It is most com- primary and brain metastasis tumors. In the era
monly seen with lung and breast cancers but has of precision medicine and individualized ther-
also been reported for colon cancer, Merkel cell apy, deciphering the tumor heterogeneity based
carcinoma, renal cell carcinoma, gastric cancer, on spatiotemporal selection is clinically imper-
ovarian cancer, and thyroid cancer [148, 149, ative. Multidisciplinary approaches are neces-
158–162]. As with BMs, the increasing success sary to fill in the gaps in knowledge regarding
of systemic therapies may be contributing to the the molecular landscape of brain metastasis.
increased incidence of ISCM [148]. Lung can- Preclinical models such as microfluidic device,
cers frequently metastasize to the CNS, but organotypic 3D culture, and patient-derived
intramedullary spine metastasis is detected in xenografts may clarify both the interplay
only 1.65% of 1215 autopsy cases and 1.8% of between metastatic cell and brain tumor micro-
NSCLC patients; and these were highly associ- environment and the brain metastatic cascade
ated with concomitant BMs suggesting common (Fig. 2.3). These emerging tools overcome tra-
mechanisms for their colonization and growth ditional cell-based technologies as they have
the potential to monitor real-time cancer pro- progression: in the wake of the << seed and soil >>
gression and personalize therapy for patients. theory. Med Oncol (Northwood, London, England).
2017;34(11):185.
Further, the a dvancement in future multimodal 15. Fidler IJ, Yano S, Zhang RD, Fujimaki T,
studies will open new paradigms to understand Bucana CD. The seed and soil hypothesis: vas-
the realm of brain metastasis and improve cularisation and brain metastases. Lancet Oncol.
patient outcomes. 2002;3(1):53–7.
16. Paget S. The distribution of secondary growths in
cancer of the breast. 1889. Cancer Metastasis Rev.
1989;8(2):98–101.
17. Hay ED. An overview of epithelio-mesenchymal
References transformation. Acta Anat (Basel). 1995;154(1):
8–20.
1. Achrol AS, Rennert RC, Anders C, Soffietti R, 18. Thiery JP. Epithelial-mesenchymal transi-
Ahluwalia MS, Nayak L, et al. Brain metastases. Nat tions in tumour progression. Nat Rev Cancer.
Rev Dis Primers. 2019;5(1):5. 2002;2(6):442–54.
2. Martin AM, Cagney DN, Catalano PJ, Warren LE, 19. Jechlinger M, Grunert S, Beug H. Mechanisms in
Bellon JR, Punglia RS, et al. Brain metastases in epithelial plasticity and metastasis: insights from 3D
newly diagnosed breast cancer: a population-based cultures and expression profiling. J Mammary Gland
study. JAMA Oncol. 2017;3(8):1069–77. Biol Neoplasia. 2002;7(4):415–32.
3. Toyokawa G, Seto T, Takenoyama M, Ichinose 20. Shi Y, Massague J. Mechanisms of TGF-beta sig-
Y. Insights into brain metastasis in patients with naling from cell membrane to the nucleus. Cell.
ALK+ lung cancer: is the brain truly a sanctuary? 2003;113(6):685–700.
Cancer Metastasis Rev. 2015;34(4):797–805. 21. Kokudo T, Suzuki Y, Yoshimatsu Y, Yamazaki
4. Hall WA, Djalilian HR, Nussbaum ES, Cho T, Watabe T, Miyazono K. Snail is required for
KH. Long-term survival with metastatic cancer TGFbeta-induced endothelial-mesenchymal tran-
to the brain. Med Oncol (Northwood, London, sition of embryonic stem cell-derived endothelial
England). 2000;17(4):279–86. cells. J Cell Sci. 2008;121(Pt 20):3317–24.
5. Dagogo-Jack I, Carter SL, Brastianos PK. Brain 22. Kahn SA, Wang X, Nitta RT, Gholamin S, Theruvath
metastasis: clinical implications of branched evolu- J, Hutter G, et al. Notch1 regulates the initiation of
tion. Trends Cancer. 2016;2(7):332–7. metastasis and self-renewal of Group 3 medulloblas-
6. Andrews DW, Scott CB, Sperduto PW, Flanders toma. Nat Commun. 2018;9(1):4121.
AE, Gaspar LE, Schell MC, et al. Whole brain 23. Rahmathulla G, Toms SA, Weil RJ. The molec-
radiation therapy with or without stereotactic ular biology of brain metastasis. J Oncol.
radiosurgery boost for patients with one to three 2012;2012:723541.
brain metastases: phase III results of the RTOG 24. Talbot LJ, Bhattacharya SD, Kuo PC. Epithelial-
9508 randomised trial. Lancet (London, England). mesenchymal transition, the tumor micro-
2004;363(9422):1665–72. environment, and metastatic behavior of
7. Ramakrishna R, Rostomily R. Seed, soil, and epithelial malignancies. Int J Biochem Mol Biol.
beyond: the basic biology of brain metastasis. Surg 2012;3(2):117–36.
Neurol Int. 2013;4(Suppl 4):S256–64. 25. Dongre A, Weinberg RA. New insights into the
8. Valastyan S, Weinberg RA. Tumor metastasis: mechanisms of epithelial-mesenchymal transition
molecular insights and evolving paradigms. Cell. and implications for cancer. Nat Rev Mol Cell Biol.
2011;147(2):275–92. 2019;20(2):69–84.
9. Pantel K, Brakenhoff RH. Dissecting the metastatic 26. Yang J, Mani SA, Donaher JL, Ramaswamy S,
cascade. Nat Rev Cancer. 2004;4(6):448–56. Itzykson RA, Come C, et al. Twist, a master regu-
10. Metastasis JE. Neoplastic Diseases: A Treatise on lator of morphogenesis, plays an essential role in
Tumors. 2nd ed. W.B. Saunders Company; 1922; tumor metastasis. Cell. 2004;117(7):927–39.
76–88. 27. Jiang WG. E-cadherin and its associated protein
11. Prasetyanti PR, Medema JP. Intra-tumor heterogene- catenins, cancer invasion and metastasis. Br J Surg.
ity from a cancer stem cell perspective. Mol Cancer. 1996;83(4):437–46.
2017;16(1):41. 28. Kafka A, Tomas D, Beros V, Pecina HI, Zeljko M,
12. McGranahan N, Swanton C. Clonal heterogeneity Pecina-Slaus N. Brain metastases from lung cancer
and tumor evolution: past, present, and the future. show increased expression of DVL1, DVL3 and
Cell. 2017;168(4):613–28. beta-catenin and down-regulation of E-cadherin. Int
13. Chambers AF, Groom AC, MacDonald J Mol Sci. 2014;15(6):10635–51.
IC. Dissemination and growth of cancer cells in 29. Barsky SH, Siegal GP, Jannotta F, Liotta LA. Loss of
metastatic sites. Nat Rev Cancer. 2002;2(8):563–72. basement membrane components by invasive tumors
14. Amelot A, Terrier LM, Mazeron JJ, Valery CA, but not by their benign counterparts. Lab Invest.
Cornu P, Carpentier A, et al. Timeline metastatic 1983;49(2):140–7.
30. Langley RR, Fidler IJ. The biology of brain metasta- tions sustain glioblastoma cell dedifferentiation
sis. Clin Chem. 2013;59(1):180–9. to a stem-like phenotype through survivin: pos-
31. Hagedorn HG, Bachmeier BE, Nerlich sible involvement in radioresistance. Cell Death Dis.
AG. Synthesis and degradation of basement mem- 2014;5:e1543.
branes and extracellular matrix and their regulation 46. Auffinger B, Tobias AL, Han Y, Lee G, Guo D, Dey
by TGF-beta in invasive carcinomas (review). Int J M, et al. Conversion of differentiated cancer cells
Oncol. 2001;18(4):669–81. into cancer stem-like cells in a glioblastoma model
32. Chaffer CL, Thompson EW, Williams after primary chemotherapy. Cell Death Differ.
ED. Mesenchymal to epithelial transition in 2014;21(7):1119–31.
development and disease. Cells Tissues Organs. 47. Dasgupta A, Lim AR, Ghajar CM. Circulating and
2007;185(1–3):7–19. disseminated tumor cells: harbingers or initiators of
33. Hugo H, Ackland ML, Blick T, Lawrence MG, metastasis? Mol Oncol. 2017;11(1):40–61.
Clements JA, Williams ED, et al. Epithelial-- 48. Joosse SA, Gorges TM, Pantel K. Biology, detec-
mesenchymal and mesenchymal--epithelial tran- tion, and clinical implications of circulating tumor
sitions in carcinoma progression. J Cell Physiol. cells. EMBO Mol Med. 2015;7(1):1–11.
2007;213(2):374–83. 49. Alix-Panabieres C, Pantel K. Challenges in cir-
34. Yao D, Dai C, Peng S. Mechanism of the culating tumour cell research. Nat Rev Cancer.
mesenchymal-epithelial transition and its relation- 2014;14(9):623–31.
ship with metastatic tumor formation. Mol Cancer 50. Riethdorf S, Pantel K. Disseminated tumor cells
Res. 2011;9(12):1608–20. in bone marrow and circulating tumor cells in
35. Franchino F, Ruda R, Soffietti R. Mechanisms and blood of breast cancer patients: current state of
therapy for cancer metastasis to the brain. Front detection and characterization. Pathobiology.
Oncol. 2018;8:161. 2008;75(2):140–8.
36. Nolte SM, Venugopal C, McFarlane N, Morozova 51. Yu M, Stott S, Toner M, Maheswaran S, Haber
O, Hallett RM, O’Farrell E, et al. A cancer DA. Circulating tumor cells: approaches to
stem cell model for studying brain metastases isolation and characterization. J Cell Biol.
from primary lung cancer. J Natl Cancer Inst. 2011;192(3):373–82.
2013;105(8):551–62. 52. Giuliano M, Shaikh A, Lo HC, Arpino G, De Placido
37. Tan BT, Park CY, Ailles LE, Weissman IL. The can- S, Zhang XH, et al. Perspective on circulating tumor
cer stem cell hypothesis: a work in progress. Lab cell clusters: why it takes a village to metastasize.
Invest. 2006;86(12):1203–7. Cancer Res. 2018;78:845.
38. Strizzi L, Hardy KM, Seftor EA, Costa FF, 53. Aceto N, Bardia A, Miyamoto DT, Donaldson MC,
Kirschmann DA, Seftor RE, et al. Development and Wittner BS, Spencer JA, et al. Circulating tumor cell
cancer: at the crossroads of Nodal and Notch signal- clusters are oligoclonal precursors of breast cancer
ing. Cancer Res. 2009;69(18):7131–4. metastasis. Cell. 2014;158(5):1110–22.
39. Zhang L, Ridgway LD, Wetzel MD, Ngo J, Yin W, 54. Adkins CE, Mohammad AS, Terrell-Hall
Kumar D, et al. The identification and character- TB, Dolan EL, Shah N, Sechrest E, et al.
ization of breast cancer CTCs competent for brain Characterization of passive permeability at the
metastasis. Sci Transl Med. 2013;5(180):180ra48. blood-tumor barrier in five preclinical models
40. Wurth R, Bajetto A, Harrison JK, Barbieri F, Florio of brain metastases of breast cancer. Clin Exp
T. CXCL12 modulation of CXCR4 and CXCR7 Metastasis. 2016;33(4):373–83.
activity in human glioblastoma stem-like cells and 55. Hanssen A, Riebensahm C, Mohme M, Joosse SA,
regulation of the tumor microenvironment. Front Velthaus JL, Berger LA, et al. Frequency of cir-
Cell Neurosci. 2014;8:144. culating tumor cells (CTC) in patients with brain
41. Chen J, Li Y, Yu TS, McKay RM, Burns DK, Kernie metastases: implications as a risk assessment
SG, et al. A restricted cell population propagates marker in oligo-metastatic disease. Cancers (Basel).
glioblastoma growth after chemotherapy. Nature. 2018;10(12):527.
2012;488(7412):522–6. 56. Klinac D, Gray ES, Freeman JB, Reid A, Bowyer S,
42. Beier D, Schulz JB, Beier CP. Chemoresistance of Millward M, et al. Monitoring changes in circulating
glioblastoma cancer stem cells--much more complex tumour cells as a prognostic indicator of overall sur-
than expected. Mol Cancer. 2011;10:128. vival and treatment response in patients with meta-
43. Chu JE, Allan AL. The role of cancer stem cells static melanoma. BMC Cancer. 2014;14:423.
in the organ tropism of breast cancer metastasis: 57. Lin X, Fleisher M, Rosenblum M, Lin O, Boire
a mechanistic balance between the “seed” and the A, Briggs S, et al. Cerebrospinal fluid circulating
“soil”? Int J Breast Cancer. 2012;2012:209748. tumor cells: a novel tool to diagnose leptomeningeal
44. Lee G, Hall RR 3rd, Ahmed AU. Cancer stem cells: metastases from epithelial tumors. Neuro Oncol.
cellular plasticity, niche, and its clinical relevance. J 2017;19(9):1248–54.
Stem Cell Res Ther. 2016;6(10):pii: 363. 58. Uhr JW, Pantel K. Controversies in clinical
45. Dahan P, Martinez Gala J, Delmas C, Monferran cancer dormancy. Proc Natl Acad Sci U S A.
S, Malric L, Zentkowski D, et al. Ionizing radia- 2011;108(30):12396–400.
59. Riethdorf S, Wikman H, Pantel K. Review: biologi- DNA in early- and late-stage human malignancies.
cal relevance of disseminated tumor cells in cancer Sci Transl Med. 2014;6(224):224ra24.
patients. Int J Cancer. 2008;123(9):1991–2006. 74. Heitzer E, Haque IS, Roberts CES, Speicher
60. Boral D, Vishnoi M, Liu HN, Yin W, Sprouse ML, MR. Current and future perspectives of liquid biop-
Scamardo A, et al. Molecular characterization of sies in genomics-driven oncology. Nat Rev Genet.
breast cancer CTCs associated with brain metastasis. 2019;20(2):71–88.
Nat Commun. 2017;8(1):196. 75. Siravegna G, Geuna E, Mussolin B, Crisafulli G,
61. Wang H, Zhang C, Zhang J, Kong L, Zhu H, Yu Bartolini A, Galizia D, et al. Genotyping tumour
J. The prognosis analysis of different metasta- DNA in cerebrospinal fluid and plasma of a HER2-
sis pattern in patients with different breast can- positive breast cancer patient with brain metastases.
cer subtypes: a SEER based study. Oncotarget. ESMO Open. 2017;2(4):e000253.
2017;8(16):26368–79. 76. Boire A, Brandsma D, Brastianos PK, Le Rhun E,
62. Liu Y, Cao X. Organotropic metastasis: role of tumor Ahluwalia M, Junk L, et al. Liquid biopsy in cen-
exosomes. Cell Res. 2016;26(2):149–50. tral nervous system metastases: a RANO review and
63. Weidle HU, Birzele F, Kollmorgen G, RÜGer R. The proposals for clinical applications. Neuro Oncol.
multiple roles of exosomes in metastasis. Cancer 2019;21:571.
Genomics Proteomics. 2017;14(1):1–16. 77. De Mattos-Arruda L, Mayor R, Ng CK, Weigelt B,
64. Shankar GM, Balaj L, Stott SL, Nahed B, Carter Martinez-Ricarte F, Torrejon D, et al. Cerebrospinal
BS. Liquid biopsy for brain tumors. Expert Rev Mol fluid-derived circulating tumour DNA better repre-
Diagn. 2017;17(10):943–7. sents the genomic alterations of brain tumours than
65. Fong MY, Zhou W, Liu L, Alontaga AY, Chandra plasma. Nat Commun. 2015;6:8839.
M, Ashby J, et al. Breast-cancer-secreted miR- 78. Pentsova EI, Shah RH, Tang J, Boire A, You D, Briggs
122 reprograms glucose metabolism in premeta- S, et al. Evaluating cancer of the central nervous sys-
static niche to promote metastasis. Nat Cell Biol. tem through next-generation sequencing of cerebro-
2015;17(2):183–94. spinal fluid. J Clin Oncol. 2016;34(20):2404–15.
66. Camacho L, Guerrero P, Marchetti D. MicroRNA 79. Huang WT, Lu NM, Hsu WY, Chang SE, Atkins A,
and protein profiling of brain metastasis com- Mei R, et al. CSF-ctDNA SMSEQ analysis to tailor
petent cell- derived exosomes. PLoS One. the treatment of a patient with brain metastases: a
2013;8(9):e73790. case report. Case Rep Oncol. 2018;11(1):68–74.
67. Kuroda H, Tachikawa M, Yagi Y, Umetsu M, Nurdin 80. Siravegna G, Marsoni S, Siena S, Bardelli
A, Miyauchi E, et al. Cluster of differentiation 46 A. Integrating liquid biopsies into the management
is the major receptor in human blood-brain barrier of cancer. Nat Rev Clin Oncol. 2017;14(9):531–48.
endothelial cells for uptake of exosomes derived 81. Wilhelm I, Fazakas C, Molnar K, Vegh AG, Hasko J,
from brain-metastatic melanoma cells (SK-Mel-28). Krizbai IA. Foe or friend? Janus-faces of the neuro-
Mol Pharm. 2019;16:292–304. vascular unit in the formation of brain metastases. J
68. Hoshino A, Costa-Silva B, Shen TL, Rodrigues G, Cereb Blood Flow Metab. 2018;38(4):563–87.
Hashimoto A, Tesic Mark M, et al. Tumour exosome 82. Bos PD, Zhang XH, Nadal C, Shu W, Gomis
integrins determine organotropic metastasis. Nature. RR, Nguyen DX, et al. Genes that mediate
2015;527(7578):329–35. breast cancer metastasis to the brain. Nature.
69. Hoshino A, Costa-Silva B, Shen T-L, Rodrigues G, 2009;459(7249):1005–9.
Hashimoto A, Mark MT, et al. Tumour exosome 83. Fan J, Cai B, Zeng M, Hao Y, Giancotti FG, Fu
integrins determine organotropic metastasis. Nature. BM. Integrin beta4 signaling promotes mammary
2015;527(7578):329–35. tumor cell adhesion to brain microvascular endo-
70. Tominaga N, Kosaka N, Ono M, Katsuda T, thelium by inducing ErbB2-mediated secretion of
Yoshioka Y, Tamura K, et al. Brain metastatic cancer VEGF. Ann Biomed Eng. 2011;39(8):2223–41.
cells release microRNA-181c-containing extracellu- 84. Kusters B, Leenders WP, Wesseling P, Smits D,
lar vesicles capable of destructing blood-brain bar- Verrijp K, Ruiter DJ, et al. Vascular endothelial
rier. Nat Commun. 2015;6:6716. growth factor-A(165) induces progression of mela-
71. Xing F, Liu Y, Wu SY, Wu K, Sharma S, Mo YY, noma brain metastases without induction of sprout-
et al. Loss of XIST in breast cancer activates MSN- ing angiogenesis. Cancer Res. 2002;62(2):341–5.
c-Met and reprograms microglia via exosomal 85. Izraely S, Sagi-Assif O, Klein A, Meshel T, Tsarfaty
miRNA to promote brain metastasis. Cancer Res. G, Pasmanik-Chor M, et al. The metastatic micro-
2018;78(15):4316–30. environment: brain-residing melanoma metasta-
72. Zhang L, Zhang S, Yao J, Lowery FJ, Zhang Q, sis and dormant micrometastasis. Int J Cancer.
Huang WC, et al. Microenvironment-induced PTEN 2012;131(5):1071–82.
loss by exosomal microRNA primes brain metastasis 86. Gorantla V, Kirkwood JM, Tawbi HA. Melanoma
outgrowth. Nature. 2015;527(7576):100–4. brain metastases: an unmet challenge in the era of
73. Bettegowda C, Sausen M, Leary RJ, Kinde I, Wang active therapy. Curr Oncol Rep. 2013;15(5):483–91.
Y, Agrawal N, et al. Detection of circulating tumor 87. Eichler AF, Chung E, Kodack DP, Loeffler JS,
Fukumura D, Jain RK. The biology of brain
metastases- translation to new therapies. Nat Rev lular players in cancer-associated brain metastasis:
Clin Oncol. 2011;8(6):344–56. translational approach to understand molecular
88. Avraham HK, Jiang S, Fu Y, Nakshatri H, Ovadia mechanism of tumor progression. Cancer Metastasis
H, Avraham S. Angiopoietin-2 mediates blood- Rev. 2018;37(4):791–804.
brain barrier impairment and colonization of triple- 102. Kim R, Keam B, Kim S, Kim M, Kim SH, Kim
negative breast cancer cells in brain. J Pathol. JW, et al. Differences in tumor microenvironments
2014;232(3):369–81. between primary lung tumors and brain metastases
89. Ma SC, Li Q, Peng JY, Zhouwen JL, Diao JF, Niu in lung cancer patients: therapeutic implications
JX, et al. Claudin-5 regulates blood-brain bar- for immune checkpoint inhibitors. BMC Cancer.
rier permeability by modifying brain microvascu- 2019;19(1):19.
lar endothelial cell proliferation, migration, and 103. Sobottka B, Pestalozzi B, Fink D, Moch H, Varga
adhesion to prevent lung cancer metastasis. CNS Z. Similar lymphocytic infiltration pattern in pri-
Neurosci Ther. 2017;23(12):947–60. mary breast cancer and their corresponding distant
90. Connell JJ, Chatain G, Cornelissen B, Vallis KA, metastases. Oncoimmunology. 2016;5(6):e1153208.
Hamilton A, Seymour L, et al. Selective permeabi- 104. Ogiya R, Niikura N, Kumaki N, Yasojima H, Iwasa
lization of the blood-brain barrier at sites of metas- T, Kanbayashi C, et al. Comparison of immune
tasis. J Natl Cancer Inst. 2013;105(21):1634–43. microenvironments between primary tumors and
91. Lyle LT, Lockman PR, Adkins CE, Mohammad brain metastases in patients with breast cancer.
AS, Sechrest E, Hua E, et al. Alterations in peri- Oncotarget. 2017;8(61):103671–81.
cyte subpopulations are associated with elevated 105. Harter PN, Bernatz S, Scholz A, Zeiner PS, Zinke
blood-tumor barrier permeability in experimental J, Kiyose M, et al. Distribution and prognostic rel-
brain metastasis of breast cancer. Clin Cancer Res. evance of tumor-infiltrating lymphocytes (TILs) and
2016;22(21):5287–99. PD-1/PD-L1 immune checkpoints in human brain
92. Yonemori K, Tsuta K, Ono M, Shimizu C, Hirakawa metastases. Oncotarget. 2015;6(38):40836–49.
A, Hasegawa T, et al. Disruption of the blood brain 106. Harary M, Reardon DA, Iorgulescu JB. Efficacy
barrier by brain metastases of triple-negative and and safety of immune checkpoint blockade for brain
basal-type breast cancer but not HER2/neu-positive metastases. CNS Oncol. 2019;8:CNS33.
breast cancer. Cancer. 2010;116(2):302–8. 107. Kono K, Mimura K, Kiessling R. Immunogenic
93. Kodack DP, Askoxylakis V, Ferraro GB, Fukumura tumor cell death induced by chemoradiotherapy:
D, Jain RK. Emerging strategies for treating molecular mechanisms and a clinical translation.
brain metastases from breast cancer. Cancer Cell. Cell Death Dis. 2013;4:e688.
2015;27(2):163–75. 108. Kirilovsky A, Marliot F, El Sissy C, Haicheur N,
94. Fidler IJ. The role of the organ microenviron- Galon J, Pages F. Rational bases for the use of the
ment in brain metastasis. Semin Cancer Biol. Immunoscore in routine clinical settings as a prog-
2011;21(2):107–12. nostic and predictive biomarker in cancer patients.
95. Fidler IJ, Balasubramanian K, Lin Q, Kim SW, Kim Int Immunol. 2016;28(8):373–82.
SJ. The brain microenvironment and cancer metasta- 109. Berghoff AS, Fuchs E, Ricken G, Mlecnik B,

sis. Mol Cells. 2010;30(2):93–8. Bindea G, Spanberger T, et al. Density of tumor-
96. Shibahara I, Kanamori M, Watanabe T, Utsunomiya infiltrating lymphocytes correlates with extent
A, Suzuki H, Saito R, et al. Clinical features of of brain edema and overall survival time in
precocious, synchronous, and metachronous brain patients with brain metastases. Oncoimmunology.
metastases and the role of tumor resection. World 2016;5(1):e1057388.
Neurosurg. 2018;113:e1–9. 110. Teglasi V, Reiniger L, Fabian K, Pipek O, Csala I,
97. Doron H, Pukrop T, Erez N. A blazing landscape: Bago AG, et al. Evaluating the significance of den-
neuroinflammation shapes brain metastasis. Cancer sity, localization, and PD-1/PD-L1 immunopositiv-
Res. 2019;79(3):423–36. ity of mononuclear cells in the clinical course of
98. Kamath SD, Kumthekar PU. Immune checkpoint lung adenocarcinoma patients with brain metastasis.
inhibitors for the treatment of central nervous Neuro Oncol. 2017;19(8):1058–67.
system (CNS) metastatic disease. Front Oncol. 111. Teng MW, Ngiow SF, Ribas A, Smyth
2018;8:414. MJ. Classifying cancers based on T-cell infiltration
99. Quail DF, Joyce JA. The microenvironmen- and PD-L1. Cancer Res. 2015;75(11):2139–45.
tal landscape of brain tumors. Cancer Cell. 112. Xue D, Xia T, Wang J, Chong M, Wang S, Zhang
2017;31(3):326–41. C. Role of regulatory T cells and CD8(+) T lym-
100. Fischer GM, Jalali A, Kircher DA, Lee WC, phocytes in the dissemination of circulating tumor
McQuade JL, Haydu LE, et al. Molecular profil- cells in primary invasive breast cancer. Oncol Lett.
ing reveals unique immune and metabolic features 2018;16(3):3045–53.
of melanoma brain metastases. Cancer Discov. 113. Plaumann J, Engelhardt M, Awwad MHS,
2019;9:628–45. Echchannaoui H, Amman E, Raab MS, et al. IL-10
101. Raza M, Prasad P, Gupta P, Kumar N, Sharma T, inducible CD8(+) regulatory T-cells are enriched in
Rana M, et al. Perspectives on the role of brain cel- patients with multiple myeloma and impact the gen-
eration of antigen-specific T-cells. Cancer Immunol 128. Chen Q, Boire A, Jin X, Valiente M, Er EE, Lopez-
Immunother. 2018;67(11):1695–707. Soto A, et al. Carcinoma-astrocyte gap junctions
114. Wu SY, Watabe K. The roles of microglia/macro- promote brain metastasis by cGAMP transfer.
phages in tumor progression of brain cancer and Nature. 2016;533(7604):493–8.
metastatic disease. Front Biosci (Landmark Ed). 129. Venkatesh HS, Johung TB, Caretti V, Noll A, Tang
2017;22:1805–29. Y, Nagaraja S, et al. Neuronal activity promotes
115. Mole RH. Whole body irradiation; radiobiology or glioma growth through neuroligin-3 secretion. Cell.
medicine? Br J Radiol. 1953;26(305):234–41. 2015;161(4):803–16.
116. Demaria S, Ng B, Devitt ML, Babb JS, Kawashima 130. Futakuchi M, Singh RK. Animal model for mam-
N, Liebes L, et al. Ionizing radiation inhibition mary tumor growth in the bone microenvironment.
of distant untreated tumors (abscopal effect) is Breast Cancer. 2013;20(3):195–203.
immune mediated. Int J Radiat Oncol Biol Phys. 131. Allen JK, Armaiz-Pena GN, Nagaraja AS, Sadaoui
2004;58(3):862–70. NC, Ortiz T, Dood R, et al. Sustained adren-
117. Okwan-Duodu D, Pollack BP, Lawson D, Khan ergic signaling promotes intratumoral inner-
MK. Role of radiation therapy as immune activa- vation through BDNF induction. Cancer Res.
tor in the era of modern immunotherapy for meta- 2018;78(12):3233–42.
static malignant melanoma. Am J Clin Oncol. 132. Dagogo-Jack I, Shaw AT. Tumour heterogeneity and
2015;38(1):119–25. resistance to cancer therapies. Nat Rev Clin Oncol.
118. Brix N, Tiefenthaller A, Anders H, Belka C, Lauber 2018;15(2):81–94.
K. Abscopal, immunological effects of radiotherapy: 133. Alizadeh AA, Aranda V, Bardelli A, Blanpain C,
narrowing the gap between clinical and preclinical Bock C, Borowski C, et al. Toward understand-
experiences. Immunol Rev. 2017;280(1):249–79. ing and exploiting tumor heterogeneity. Nat Med.
119. Postow MA, Callahan MK, Barker CA, Yamada Y, 2015;21(8):846–53.
Yuan J, Kitano S, et al. Immunologic correlates of 134. Gerlinger M, Rowan AJ, Horswell S, Math M, Larkin
the abscopal effect in a patient with melanoma. N J, Endesfelder D, et al. Intratumor heterogeneity
Engl J Med. 2012;366(10):925–31. and branched evolution revealed by multiregion
120. Hiniker SM, Reddy SA, Maecker HT, sequencing. N Engl J Med. 2012;366(10):883–92.
Subrahmanyam PB, Rosenberg-Hasson Y, Swetter 135. Stanta G, Bonin S. Overview on clinical relevance of
SM, et al. A prospective clinical trial combining intra-tumor heterogeneity. Front Med. 2018;5:85.
radiation therapy with systemic immunotherapy in 136. Echeverria GV, Powell E, Seth S, Ge Z, Carugo A,
metastatic melanoma. Int J Radiat Oncol Biol Phys. Bristow C, et al. High-resolution clonal mapping of
2016;96(3):578–88. multi-organ metastasis in triple negative breast can-
121. Pfannenstiel LW, McNeilly C, Xiang C, Kang cer. Nat Commun. 2018;9(1):5079.
K, Diaz-Montero CM, Yu JS, et al. Combination 137. Stanta G, Jahn SW, Bonin S, Hoefler G. Tumour het-
PD-1 blockade and irradiation of brain metastasis erogeneity: principles and practical consequences.
induces an effective abscopal effect in melanoma. Virchows Arch. 2016;469(4):371–84.
Oncoimmunology. 2019;8(1):e1507669. 138. Brastianos PK, Carter SL, Santagata S, Cahill DP,
122. Stamell EF, Wolchok JD, Gnjatic S, Lee NY, Taylor-Weiner A, Jones RT, et al. Genomic char-
Brownell I. The abscopal effect associated with a acterization of brain metastases reveals branched
systemic anti-melanoma immune response. Int J evolution and potential therapeutic targets. Cancer
Radiat Oncol Biol Phys. 2013;85(2):293–5. Discov. 2015;5(11):1164–77.
123. Hamilton AJ, Seid J, Verdecchia K, Chuba 139. Han CH, Brastianos PK. Genetic characterization of
P. Abscopal effect after radiosurgery for solitary brain metastases in the era of targeted therapy. Front
brain metastasis from non-small cell lung cancer. Oncol. 2017;7:230.
Cureus. 2018;10(12):e3777. 140. Neagu MR, Gill CM, Batchelor TT, Brastianos
124. Hoshide R, Jandial R. The role of the neural PK. Genomic profiling of brain metastases: cur-
niche in brain metastasis. Clin Exp Metastasis. rent knowledge and new frontiers. Chin Clin Oncol.
2017;34(6–7):369–76. 2015;4(2):22.
125. Schild T, Low V, Blenis J, Gomes AP. Unique
141. Bertucci F, Finetti P, Guille A, Adelaide J, Garnier
metabolic adaptations dictate distal organ- S, Carbuccia N, et al. Comparative genomic analysis
specific metastatic colonization. Cancer Cell. of primary tumors and metastases in breast cancer.
2018;33(3):347–54. Oncotarget. 2016;7(19):27208–19.
126. Rondeau G, Abedinpour P, Desai P, Baron VT, 142. Liao L, Ji X, Ge M, Zhan Q, Huang R, Liang X,
Borgstrom P, Welsh J. Effects of different tissue et al. Characterization of genetic alterations in brain
microenvironments on gene expression in breast metastases from non-small cell lung cancer. FEBS
cancer cells. PLoS One. 2014;9(7):e101160. Open Bio. 2018;8(9):1544–52.
127. Valiente M, Obenauf AC, Jin X, Chen Q, Zhang XH, 143. Paik PK, Shen R, Won H, Rekhtman N, Wang L,
Lee DJ, et al. Serpins promote cancer cell survival Sima CS, et al. Next-generation sequencing of stage
and vascular co-option in brain metastasis. Cell. IV squamous cell lung cancers reveals an associa-
2014;156(5):1002–16. tion of PI3K aberrations and evidence of clonal het-
erogeneity in patients with brain metastases. Cancer of tumor cells in cerebrospinal fluid and matched pri-
Discov. 2015;5(6):610–21. mary tumors from metastatic breast cancer patients
144. Jesinghaus M, Wolf T, Pfarr N, Muckenhuber A, with leptomeningeal carcinomatosis. Cancer Res.
Ahadova A, Warth A, et al. Distinctive spatiotem- 2013;73(23):7134–43.
poral stability of somatic mutations in metastasized 155. Cheng H, Perez-Soler R. Leptomeningeal metas-
microsatellite-stable colorectal cancer. Am J Surg tases in non-small-cell lung cancer. Lancet Oncol.
Pathol. 2015;39(8):1140–7. 2018;19(1):e43–55.
145. National Collaborating Centre for Cancer. National 156. Potti A, Abdel-Raheem M, Levitt R, Schell DA,
Institute for Health and Clinical Excellence: guid- Mehdi SA. Intramedullary spinal cord metasta-
ance. Metastatic spinal cord compression: diagnoses (ISCM) and non-small cell lung carcinoma
sis and management of patients at risk of or with (NSCLC): clinical patterns, diagnosis and thera-
metastatic spinal cord compression. Cardiff (UK): peutic considerations. Lung Cancer (Amsterdam,
National Collaborating Centre for Cancer (UK) Netherlands). 2001;31(2–3):319–23.
National Collaborating Centre for Cancer; 2008. 157. Okamoto H, Shinkai T, Matsuno Y, Saijo
146. Karsy M, Guan J, Sivakumar W, Neil JA, Schmidt N. Intradural parenchymal involvement in the spi-
MH, Mahan MA. The genetic basis of intradural nal subarachnoid space associated with primary lung
spinal tumors and its impact on clinical treatment. cancer. Cancer. 1993;72(9):2583–8.
Neurosurg Focus. 2015;39(2):E3. 158. Haykal T, Towfiq B. Merkel cell carcinoma with
147. Costigan DA, Winkelman MD. Intramedullary spi- intramedullary spinal cord metastasis: a very rare
nal cord metastasis. A clinicopathological study of clinical finding. Clin Case Rep. 2018;6(6):1181–2.
13 cases. J Neurosurg. 1985;62(2):227–33. 159. Perez-Suarez J, Barrio-Fernandez P, Ibanez-Plagaro
148. Payer S, Mende KC, Westphal M, Eicker FJ, Ribas-Arino T, Calvo-Calleja P, Mostaza-
SO. Intramedullary spinal cord metastases: an Saavedra AL. Intramedullary spinal cord metas-
increasingly common diagnosis. Neurosurg Focus. tasis from gastric adenocarcinoma: case report
2015;39(2):E15. and review of literature. Neurocirugia (Astur).
149. Weng Y, Zhan R, Shen J, Pan J, Jiang H, Huang K, 2016;27(1):28–32.
et al. Intramedullary spinal cord metastasis from 160. Kaballo MA, Brennan DD, El Bassiouni M, Skehan
renal cell carcinoma: a systematic review of the lit- SJ, Gupta RK. Intramedullary spinal cord metasta-
erature. Biomed Res Int. 2018;2018:7485020. sis from colonic carcinoma presenting as Brown-
150. Zheng H, Li W, Kang Y. Tumor-stroma interactions Sequard syndrome: a case report. J Med Case
in bone metastasis: molecular mechanisms and ther- Reports. 2011;5:342.
apeutic implications. Cold Spring Harb Symp Quant 161. Isoya E, Saruhash Y, Katsuura A, Takahashi S,
Biol. 2016;81:151–61. Matsusue Y, Hukuda S. Intramedullary spinal
151. Theriault RL, Theriault RL. Biology of bone metas- cord metastasis of ovarian tumor. Spinal Cord.
tases. Cancer Control. 2012;19(2):92–101. 2004;42(8):485–7.
152. Futakuchi M, Fukamachi K, Suzui M. Heterogeneity 162. Zhou Z, Li Y, Yan X, Wang X, Chen S, Xiao
of tumor cells in the bone microenvironment: mech- J. Characteristics of a thyroid carcinoma cell
anisms and therapeutic targets for bone metastasis line derived from spinal metastasis. Biosci Rep.
of prostate or breast cancer. Adv Drug Deliv Rev. 2016;36(6):e00426.
2016;99(Pt B):206–11. 163. Gainor JF, Ou SH, Logan J, Borges LF, Shaw
153. Battafarano G, Rossi M, Marampon F, Del Fattore AT. The central nervous system as a sanctuary site in
A. Cellular and molecular mediators of bone meta- ALK-positive non-small-cell lung cancer. J Thorac
static lesions. Int J Mol Sci. 2018;19(6):1709. Oncol. 2013;8(12):1570–3.
154. Magbanua MJ, Melisko M, Roy R, Sosa EV,
Hauranieh L, Kablanian A, et al. Molecular profiling
Preclinical Models of Brain
Metastasis
3
Lucía Zhu and Manuel Valiente
odels for Brain Metastasis

M engineered with different reporters, including
Research: An Overview those compatible with non-invasive imaging (e.g.,
luciferase, Luc, for bioluminescence) and/or his-
The complexity of the multistep process of tology (e.g., green fluorescence protein, GFP).
metastases cannot be fully recapitulated in vitro. Labeled cancer cells were then inoculated in mice
Consequently, the use of mice as the experimen- using different routes such as intracardiac (IC)
tal model of choice is broadly accepted. The injections through the left ventricle, intracarotid,
study of brain metastasis in preclinical models or intracranial approaches [1–3]. Intravascular
includes several steps that, in principle, are com- injection is the preferred method since it incorpo-
mon to metastases in other organs (e.g., the abil- rates the strong selective step of the extravasation
ity of cancer cells to migrate toward and through the BBB. The advantage of intracarotid
intravasate into capillaries at the primary tumor injection is the reduction of the incidence of extra-
as well as the survival of tumor cells while in cir- cranial metastases. However, at the same time, this
culation). Given the interest of this book, we will procedure requires surgery and thus increases the
consider aspects of metastatic dissemination of time to develop the experimental procedure.
particular interest in the brain. Preclinical models Consequently, intracardiac injection of human
have been used to study these specific steps cancer cells has been the method of choice to
within the metastatic cascade that involve extrav- induce experimental brain metastasis. Frequently,
asation through the blood-brain barrier (BBB), inoculation of metastatic cancer cells recovered
survival of extravasated metastasis initiating from pleural fluids or lymph nodes, the so-called
cells, reactivation of proliferation to re-grow the parental cell line (P), into mouse circulation does
tumor in the brain as well as the interaction with not yield a significant number of mice with brain
the surrounding microenvironment. metastases [1, 3]. This parental (P) cell line is
In order to study brain metastasis in the labora- highly heterogeneous and may or may not contain
tory, researchers obtained cancer cells from cellular clones that could have the ability to target
patients, usually from pleural fluids or lymph node the brain. In order to enrich those cancer cell
metastases (Fig. 3.1a). These cancer cells were clones with the ability to grow in the brain, paren-
tal cells are inoculated in mice IC and when metas-
tases are detected in specific organs, the metastatic
L. Zhu · M. Valiente (*) lesion is dissected out and grown in vitro. This
Brain Metastasis Group, Spanish National Cancer process of positive selection has to be repeated
Research Centre (CNIO), Madrid, Spain between 3 and 5 times to enrich those variants
e-mail: mvaliente@cnio.es

https://doi.org/10.1007/978-3-030-42958-4_3
38 L. Zhu and M. Valiente
a
Metastasis
Metastasis
1ry tumor Parental cells
GFP+/ Luc+ BrM1 BrM2 BrM3
b c d e
Mouse BrM BrM cells
BrM cells Metastases
cancer cells
Brain
Endothelial slice
cells
Vessels
GEMM Astrocytes
(microenvironment)
3D cultures
f g
MT/ret
Braf CA;Cdkn2alox/lox;Ptenlox/lox;myrAkt1 Rb1lox/lox;Trp53lox/lox

Melanoma Small cell lung cancer PDX
Fig. 3.1 Models for brain metastasis research. (a) Organotypic brain cultures allow modelling of initial (d)
Schema representing the generation of brain metastatic or advanced (e) stages of brain colonization. This prepara-
cell lines (BrM). (b) Syngeneic BrM cell lines could be tion is a useful resource to analyze interactions with the
used to evaluate brain metastasis in an immunocompetent microenvironment and it is compatible with genetic and
host. This experimental model also allows interrogation of pharmacologic manipulations. (f) Available GEMMs that
genetic modifications induced in specific components of have been described to generate spontaneous brain metas-
the microenvironment by using genetically engineered tasis. (g) Human brain metastasis can be cultured in vitro
mouse models (GEMMs). (c) Artificial blood–brain bar- or inoculated in immunosuppressed mice to establish
rier (BBB) assay could be used to evaluate mediators of brain metastasis patient-derived xenografts (PDX)
permeability as well as penetration of drugs. (d, e)
present in the P cell line with an increased ability microenvironment by inoculating them into genet-
to target the brain. These organotropic cell lines ically engineered mouse models (GEMMs)
are termed brain metastatic (BrM) [1, 3–5] (Fig. 3.1b). The use of these experimental models
(Fig. 3.1a). This approach has been broadly to study the last step of metastasis, brain coloniza-
applied to generate not only human BrM cell lines tion, has generated a significant amount of knowl-
but also mouse BrM cell lines from the main edge about the underlying biology by reporting
sources of brain metastases, including breast, lung, multiple mediators of brain metastasis that have
renal cancer, melanoma and colorectal cancer been validated in human samples [1, 3–5, 10–15].
among others, and representative of the most fre- Few of them have been translated into experimen-
quent oncogenomic profiles from each tumor type tal therapeutic interventions with positive results,
[3, 4, 6–10]. In addition, mouse BrM cell lines which later have been translated into clinical trials
could be also used to study the contribution of the [3, 10].
3 Preclinical Models of Brain Metastasis 39
In spite of the success of organotropic models, Given recent experimental protocols applied to
alternative and complementary approaches must be other brain tumors [26], it is highly desirable that
incorporated to preclinical research. For instance, this clinically relevant model gets incorporated
models that generate spontaneous brain metastasis into brain metastasis research.
from orthotopic injections or from spontaneously Recent clinical trials using whole brain radia-
developed primary tumors are highly needed. The tion therapy (WBRT) have questioned the inter-
significant inefficiency, the time required for detect- est of this approach, given the limited benefit for
ing brain metastasis, and the limitation imposed by patients and the negative impact on neurocogni-
the faster growth of the primary tumor are all cave- tion [27–30].
ats that have prevented their use [16–18]. In addi- Although limited scientific reports have
tion, in order to incorporate the higher degree of addressed the efficacy of WBRT to challenge
genomic complexity in human cancer, it is manda- brain metastasis viability, their conclusions
tory to incorporate human brain metastasis through include the limited therapeutic benefit on estab-
patient-derived xenografts (PDX) models [19–23]. lished metastases.
However, their main caveats are that they require As demonstrated clinically with the use of pre-
immunosuppressed hosts and they are not easy ventive WBRT on small-cell lung cancer (SCLC)
models to incorporate genetic manipulations. patients [31–33], experimental models have con-
In general, the field has been studying naive firmed that treating micrometastasis is more
brain metastases when patients are usually heav- effective than treating established metastases [34,
ily treated with neurosurgery, radiation, chemo- 35]. In a triple-negative breast cancer (TNBC)
therapy, targeted therapies, and immunotherapies. model, 88% reduction of micrometastases was
The next generation of brain metastasis preclini- observed upon delivery of a fractionated dose
cal models should include relevant therapies to consisting of ten sessions of 3 Gy each. In con-
validate the knowledge generated with naive trast, only 55% tumor reduction was observed in
models and to address critical questions includ- macrometastases [34]. Similarly, when a single
ing treatment resistance. dose of radiation was applied 5 days after cancer
In addition, surrogates of the BBB have been stud- cell injection, a 70% reduction of brain metasta-
ied not only to functionally validate molecular media- ses was reported [35]. However, if radiation was
tors required to cross the vascular barrier [1] but also to delayed 3 weeks and applied once brain metasta-
test drug permeability [3] (Fig. 3.1c). Brain organo- ses from the breast cancer cell line MDA-IBC3
typic cultures in which BrM cells are plated on the were detected, responses were minimal [35].
surface (Fig. 3.1d) or are already present after process- Modelling responses to WBRT using in vitro
ing brains with established metastases (Fig. 3.1e) offer approaches suggest that clonogenic growth (onco-
a good alternative to evaluate scientific hypothesis spheres) faithfully predict the low responses
before testing them in vivo [3, 4, 10, 24, 25]. The main found in vivo [35]. In fact, c-Met is among the
advantages of organotypic cultures are that they con- enriched genes in oncospheres [36]. When its
tain the brain microenvironment, which allows more expression is targeted, clonogenic growth, which
in-depth studies, and that they are compatible with is not sensitive to radiation, becomes affected. In
both human and mice tissues, in which both genetic vivo, targeting c-Met sensitizes MDA-435 to radi-
and pharmacologic approaches could be tested. ation not only in the brain but also in extracranial
tumors, which are intrinsically more sensitive to
the application of this therapy [36]. Results from
Local Therapies in Experimental these works suggest that the brain microenviron-
Models of Brain Metastasis ment might offer clues to the resistance of brain
metastasis to radiation. Interestingly, when WBRT
In spite of the broad use of neurosurgery to treat was applied to a naive brain before inoculation of
patients with brain metastasis, this approach has cancer cells, tumor cells inoculated afterward
not been incorporated into experimental models. experienced superior growth ability [37].
Similarly, breast cancer cells obtained from brains Systemic Therapy in Experimental
treated with radiation that were later cultured Models of Brain Metastasis:
ex vivo did not reproduce their initial resistance Chemotherapy and Targeted
in vivo [34]. Furthermore, upon reinjection into Therapies
mice, the resistance of cancer cells to WBRT
reappeared [34]. Mathematical models predicted The penetration of many systemic chemothera-
that response of brain metastases to radiation peutic agents into the brain has been proved to be
could be improved by doses more than 20 Gy limited despite the assumption that the BBB is
[35]. However, an experimental protocol of 30 Gy disrupted in brain metastasis and modified into a
fractionated in ten doses of 3 Gy is enough to dis- blood-tumor barrier (BTB). Paclitaxel and doxo-
rupt the generation of Dcx+ immature neurons rubicin, two potent chemotherapies used in can-
from neural stem cells [34], discarding the possi- cer, did not reach therapeutic levels in two
bility of providing higher doses, given the associ- experimental breast cancer brain metastasis mod-
ated neurotoxicity. Alternative approaches to els and were ineffective in treating brain metasta-
minimize the impact of radiation on neurocogni- ses, despite higher accumulation of these two
tion have been validated experimentally. Using agents in the lesions compared to normal brain
metastasis-free mice subjected to WBRT or tissue [40]. This increased permeability of the
WBRT with hippocampal sparing (HSI, hippo- BTB has been associated with alterations in peri-
campal sparing irradiation), radiation-induced cyte subpopulations, specifically an increase of
toxicity was studied at both cellular and behav- pericytes expressing desmin, as shown in differ-
ioral levels [38]. All mice (control, WBRT, ent experimental brain metastases derived from
WBRT + HSI) did well in non-specific neurocog- breast cancer, including triple-negative, HER2+
nitive tests, while differed in those involving the and inflammatory breast cancer [41]. However,
hippocampus. Specifically, an increased deficit in these drug concentrations remain insufficient to
spatial memory was detected given that 40% of exert cytotoxic effects compared to that observed
mice receiving WBRT failed the object placement in peripheral metastases derived from the same
task, while only 14% do so in the non-irradiated model [40], proving that BBB-permeable agents
and HSI groups. If more time is given to perform are needed to target cancer cells in this secondary
the analysis, further challenging memory, 70% of organ. In this regard, temozolomide, a well-
the animals that received WBRT failed versus known alkylating agent used for the treatment of
45% of those receiving HSI and 33% of controls. primary brain tumors that penetrates the BBB,
Interestingly, hippocampal tests that do not has been shown to be effective in preventing
involve neurogenesis were not altered upon brain metastasis from a TNBC brain metastasis
WBRT [38]. Behavioral tests correlated with cel- model expressing low levels of MGMT [42];
lular findings, including increased cell death and these results have not been successfully trans-
absence of proliferation in the dentate gyrus, lated into patients [43]. However, these clinical
which has increased levels of microglia [38]. studies have included temozolomide therapy for
Experimental models recapitulate the lack of established macrometastases, so the use of this
major benefit with WBRT reported by recent therapy as a preventive strategy has not been
clinical studies and suggest that alternative explored yet.
approaches to deliver radiation could be better, as The BBB not only imposes a limitation to
confirmed by the application of stereotactic chemotherapeutic agents but also other drugs
radiosurgery (SRS) [39]. Nonetheless, identifica- targeting specific molecular alterations from
tion of the molecular mediators of radio resis- key oncogenic signaling pathways in cancer.
tance associated with WBRT in vivo and the Side-by-
side assessment of drug efficacy of
development of radio sensitizers will facilitate a two PI3K/mTOR inhibitors (brain-permeable
more personalized approach to its application GNE-317 and nonpermeable GDC-0980) by
based on potential biomarkers. in vivo two-photon microscopy in an experimen-
tal melanoma brain metastasis model showed breast cancer models in a preventive scenario
effective targeting of brain metastases only by [47]; however, established intracranial lesions
the brain-penetrating inhibitor [44]. BKM120, from other models are resistant to trastuzumab
another selective PI3K inhibitor shown to be and lapatinib treatment while orthotopic implan-
BBB- permeable, was effective in reducing tation (i.e. fat pad) of the same cells does respond
brain metastasis incidence in 50% of the sample to both treatments [48]. Efforts to overcome this
population when several HER2+ human breast resistance have resulted in combination therapies
cancer cell lines were implanted orthotopi- of anti-VEGFR2 antibody DC101 together with
cally or injected intravenously [45], suggesting trastuzumab and/or lapatinib, resulting in more
that targeting the PI3K-AKT-mTOR pathway than fourfold survival benefit of the triple combi-
with brain- penetrating small molecules could nation treatment compared to untreated control
be an effective treatment for brain metastasis mice [48]. In this same line, targeting of other
(Table 3.1). tyrosine kinases related to the pathway like HER3
Around 18% of patients diagnosed with brain with the monoclonal antibody LJM716 reduces
metastasis are eligible for targeted therapies, spe- brain metastases and increases survival
cifically those harboring molecular alterations in significantly in a HER2+ breast cancer model
their primary tumor: HER2+ breast cancer, compared to treatment with trastuzumab or per-
EGFR-mutant and ALK-translocated lung can- tuzumab alone, which do not give any benefit
cer, and BRAF-mutant melanomas, all of which compared to the untreated control group [49],
have shown positive intracranial response to dif- reflecting the need of targeting oncogenic path-
ferent targeted agents that are both under clinical ways through several mediators for overcoming
development or Food and Drug Administration treatment-derived drug resistance (Table 3.1).
(FDA)-approved [46]. Preclinically, these results The use of EGFR tyrosine kinase inhibitors
have been recapitulated with different experi- (TKIs) for patients with advanced EGFR-mutant
mental mouse models. Lapatinib has been shown non-small-cell lung cancer (NSCLC) has resulted
to delay brain metastases growth in some HER2+ in positive intracranial response apart from inhib-
Table 3.1 Use of preclinical models to test targeted therapies

BBB
Compound Target permeability Preclinical model Setting Result Ref
GNE317/GDC-0980 PI3K/mTOR Yes/No Melanoma (A2058) Interventive +/− [44]
BKM120 PI3K Yes HER2+ breast cancer Preventive + [45]
(MDA-MB-453/BT474)
Lapatinib HER2 Yes HER2+ breast cancer Preventive + [47]
(MDA-MB-231-BR-
HER2)
Lapatinib + trastuzumab HER2 Yes/? HER2+ breast cancer Interventive −/− [48]
(BT474)
Trastuzumab/pertuzumab HER2 ?/? HER2+ breast cancer Interventive − [48]
(BT474)
Lapatinib/trastuzumab + HER2/ Yes/?/? HER2+ breast cancer Interventive + [48]
DC101 VEGFR2 (BT474)
Trastuzumab/pertuzumab HER2/HER3 Yes/?/? HER2+ breast cancer Interventive + [49]
+ LJM716 (BT474)
Rociletinib/osimertinib EGFRMUT No/Yes EGFRMUT lung cancer Interventive −/+ [51]
(PC9)
Crizotinib/alectinib ALK No/yes EML4-ALK variant 5a Interventive −/+ [55]
lung cancer
(A925LPE3)
Entrectinib ALK/ROS1/ Yes EML4-ALK Interventive + [56]
TRK (NCI-H2228)
iting extracranial disease thus increasing overall Melanoma brain metastasis patients also ben-

survival [50]. However, preclinical studies, efit from targeted therapies, mainly BRAF V600E
including therapies for brain metastases from this TKIs dabrafenib and vemurafenib [57–59].
particular primary tumor, are scarce. Osimertinib, Preclinical models of brain metastasis, including
a third-generation EGFR TKI selective for these therapies, are limited. Several BRAF
EGFR-TKI-sensitizing mutation (EGFRm) and V600E mutated melanoma human melanoma
T790M resistance mutations approved in 2017 cells have been shown to generate experimental
for clinical use, showed greater penetration of the brain metastases [60]. Vemurafenib-resistant
BBB than gefitinib, rociletinib, or afatinib [51]. It melanoma cells generated in vitro show distinct
induced sustained tumor regression in an expression profile to vemurafenib-sensitive cells
EGFRm-NSCLC brain metastasis experimental but do not change their ability to colonize the
model at clinically relevant therapeutic doses brain despite their increased ability to metasta-
while rociletinib did not (Table 3.1), and could size to the lung and the liver [60]. Since 50% of
potentially overcome resistance to previous treat- melanoma brain metastasis results from BRAF-
ment with EGFR-TKIs as shown by patients V600E- mutated primaries, new experimental
included in the AURA phase I/II study models incorporating this molecular alteration
(NCT01802632) [51]. and targeted therapies are needed to study meta-
ALK-translocated lung cancer patients have static spread to the brain in the skin cancer with
shown positive responses to the first-generation highest death rates.
TKI crizotinib, although intracranial response
was only achieved with BBB-permeable next-
generation TKIs like ceritinib, brigatinib, and nbiased Screens for Brain
U
alectinib due to suboptimal accumulation of Metastasis Mediators
crizotinib in the brain [52–54]. These responses
have been faithfully recapitulated preclinically In Vitro Transcriptomics
with an EML4-ALK variant 5a lung adenocarci-
noma brain metastasis model sensitive to both Metastatic colonization is a multistep process
crizotinib and alectinib at the primary tumor site, that enriches disseminated cancer cells through
but resistant to crizotinib and sensitive to alec- positive and negative selection from an initial
tinib in the brain [55] (Table 3.1). In spite of cellular pool derived from the primary tumor.
these advances, progression-free survival (PFS) Consequently, metastatic lesions will be richer in
of patients receiving these TKIs does not exceed cancer cells with all the attributes required to
15 months. Drug resistance developed through reach and colonize the target organ. This has
prolonged treatment thus remains as an unmet been the rationale for the development of organo-
need and novel small molecule inhibitors target- tropic metastatic derivatives that are established
ing resistant ALK-dependent brain metastases through multiple rounds of in vivo selection
are necessary. Studies with next-generation ALK (Fig. 3.1a).
inhibitors such as lorlatinib and brigatinib are In order to dissect brain tropism at the molecu-
promising. Entrectinib, an orally bioavailable lar level, comparisons between P cell lines, without
potent inhibitor of ALK, ROS1 and TRK family the ability to target the brain, and BrM cells were
kinases, has been reported to induce significant performed. This approach has been applied to
reduction of intracranially implanted tumors breast cancer [1] and lung cancer models of brain
from EML4-ALK rearranged NSCLC increas- metastasis [5] (Fig. 3.2a). Transcriptomic analysis
ing mice survival in more than 70% [56] of P versus BrM cells growing in vitro reflects sig-
(Table 3.1). Future clinical trials could open the nificant differences between them. Although the
way to a drug potentially suited to treat brain overlap of differentially expressed genes among
metastases from several molecularly defined pri- different models is more limited [4], upregulated
mary tumors. genes, potential mediators of brain metastasis tro-
pism, or downregulated genes, potential brain For instance, a 17-gene signature named brain
metastasis suppressors, were successfully validated metastasis signature (BrMS), obtained by com-
in functional experiments using in vivo brain paring two different ER−/HER2− breast adeno-
metastasis assays and in human samples, where carcinoma models tropic to the brain
their increased levels at the primary tumor correlate (MDA231-BrM and CN34-BrM) respect to their
with a higher risk of brain metastasis incidence. parental cell lines, was sufficient to predict brain
Many of the genes found with this approach medi- relapse when applied to three independent patient
ated the ability to cross the BBB [1, 12, 13, 15] or cohorts [1]. Among BrMS genes present in can-
interactions with the brain microenvironment [4, cer cells, the α2,6-sialyltransferase encoded by
10, 11] (Fig. 3.2a). ST6GALNAC5 was selected. Mechanistic studies
human PURB
a Parental BrM b ONECUT2
Brain ESRRG
TCF4
CTSS
ST6GALNAC5 HOXB9
LEF1 sc
Breast PCDH7 Lung
cancer NS cancer
Bone
Lungs
c
Brain met BrM LOF candidate
No brain mets P GOF candidate
Molecular
dissection
1ry 1ry
Melanoma (brain met.) (no brain met.)
Breast
cancer Rationale
therapeutic
Lung cancer Brain met. candidates strategy
Fig. 3.2 Use of preclinical models to dissect the molecu- metastases when human cancer cells were analyzed but
lar regulation of brain metastasis. (a) Parental and brain also allowed evaluating the tumor microenvironment by
metastatic derivatives (BrM) have been interrogated analyzing mouse genes. (c) Evaluation of human samples,
in vitro. Analysis of differentially expressed genes shows including primary tumors and brain metastasis, can allow
not only cancer-type specific but also commonly deregu- identification of candidate genes that may contribute to
lated mediators of the disease (LEF1, PCDH7, NS) brain metastasis formation. In order to functionally vali-
involved in a variety of mechanisms required for brain date candidate genes, loss of function (LOF) and gain of
colonization. (b) Brain metastases have been interrogated function (GOF) approaches can be applied using preclini-
in situ and compared with orthotopic and subcutaneous cal models. These mechanistic assays in experimental
tumors and metastases growing in other organs. These models will help improve therapeutic strategies
studies not only identified potential mediators of brain
proved that cancer cell surface decoration with from cancer cells growing in the brain also differs
2–6 sialyl groups was required to increase the from the one obtained from orthotopic tumors
ability to cross the BBB [1]. [61], epigenetic mechanisms may play a critical
In contrast to breast cancer, lung cancer usu- role in reprogramming cancer cells during the
ally disseminates fast. A Wnt-dependent program adaptation to the brain microenvironment.
is responsible for facilitating the aggressive dis- Reprogramming of cancer cells growing in the
semination of lung cancer to multiple organs brain involves the upregulation of neuronal genes
including the brain [5]. Two lung adenocarci- [61]. This emerging expression pattern was sug-
noma models (H2030-BrM and PC9-BrM) tropic gested to be regulated by various transcription
to the brain were used to identify key components factors, including PURB, ONECUT2, ESRRG,
of the Wnt pathway. LEF1 increases the ability of and TCF4, that show reduced promoter methyla-
BrM cells to grow in spheres, which is a surro- tion in brain metastatic lesions.
gate of metastasis-initiating capabilities, while A similar approach comparing different
HOXB9 is required for a superior migratory organotropic cell lines including a lung meta-
behavior that is necessary for brain colonization static (LM) derivative, a bone metastatic deriva-
[5] (Fig. 3.2a). Although both requirements are tive (BoM), and a BrM one derived from the
critical for brain metastasis, they are equally same parental ER−/HER2− breast cancer cell line
important for bone metastasis [5]. (MDA231) was used to evaluate in situ differen-
tial expression patterns of proteases and their
inhibitors specifically [15]. Transcriptomic dif-
In Situ Transcriptomics ferences among metastatic cells in different
organs are amplified along the process of organ
In vitro unbiased screens to identify mediators of colonization, suggesting that the transcriptome
brain metastasis have been complemented with of cancer cells reflects organ adaptation [15].
the analysis of transcriptomes in situ [15, 61–63] These approaches also allow scoring the micro-
(Fig. 3.2b). The rationale of this alternative environment by excluding human genes derived
approach is that there may be important media- from human cancer cells. Attending to the
tors of brain metastasis not permanently but tran- expression of mouse genes, the three organs
siently induced in cancer cells tropic to the brain. evaluated (brain, bones, and lungs) cluster inde-
In fact, these analyses confirm that there are tran- pendently. However, the brain differs signifi-
scriptomic modifications only manifested when cantly more from lungs or bones than these two
the cancer cell is studied in a given organ. Breast, organs among themselves. When cancer cells
lung, melanoma, and colon cancer cells were initiate organ colonization to form micrometas-
grown as either subcutaneous tumors, at the tases, they do not significantly alter the expres-
orthotopic location according to the origin of sion pattern of the organ compared to the naive
the cancer cell, or in the brain after intracarotid one without metastasis. In contrast, at late stages
injection [61]. Differentially expressed genes (macrometastases), the organ transcriptome is
show that the transcriptome of cancer cells does significantly altered in the lungs, bones, and
not change significantly when grown at the sub- brain. Again, the degree of transcriptomic
cutaneous location or in the orthotopic location. changes in lungs and bones is more discrete than
However, when the same cancer cells are grown that in the brain [15]. This could reflect the abun-
in the brain, their transcriptomic profile diverges dance of specific barriers in the brain that limit
from those obtained at other locations (in vitro, the growth of incoming metastatic cells com-
subcutaneous, orthotopic) and become more sim- pared to other secondary organs more similar to
ilar to other cancer cell lines from different tumor the primary tumor that may only require a lim-
types also obtained from the brain. Changes in ited adaptation of cancer cells to thrive.
gene expression correlate with altered methy- Although the main findings of unbiased tran-
lome patterns. Since the methylome obtained scriptomic screens applied to brain metastasis
experimental models have been validated in finding confirms that miRNA enriched in EVs
patient samples [1, 5, 15], the inverse approach secreted from primary tumors could influence
has not been equally investigated. Evaluating vascular barriers to facilitate extravasation of
candidates obtained from unbiased screens in cancer cells [71]. In addition, miR-509 down-
human samples using experimental models regulation in human brain metastasis as well as
would allow testing their functional contribution experimental brain organotropic breast cancer
to brain colonization as well as to dissect the cell lines allows maintenance of high expression
underlying molecular regulation (Fig. 3.2c). Both levels of RhoC, which is required to produce
considerations are key to rationalize more spe- MMP9, an enzyme targeting endothelial cell-
cific and effective therapies. Although the limited junctions of the BBB, and TNFα [72], which
number of studies that have compared human and plays an important role for increased BBB-
experimental transcriptomic screens found permeability in sepsis [73].
reduced overlap in terms of specific genes, path- miRNAs continue to be required once meta-
ways were partially conserved. This suggests that static cells have crossed the BBB. Re-initiation
experimental brain metastasis models are valu- of the secondary tumor requires stem cell-like
able platforms for the identification of novel properties [74], which could be provided by the
mediators of the disease and to test them expression of pluripotency factors. Among them
functionally. KLF4 is required for the initiation of breast can-
cer brain metastasis. To maintain high expres-
sion levels of KLF4, CD24−/CD44+/ESA+ brain
Noncoding RNA metastasis cancer stem cells downregulate
miR-7 [64]. In addition, miRNAs from the
In parallel to transcriptomic analyses, expression microenvironment also play an important role in
profiles of small noncoding RNAs, mainly miR- colonization. Reactive astrocytes, which closely
NAs, have been developed to identify mediators interact with cancer cells, are highly secretory
of brain metastasis. Unbiased screens comparing cells known to produce EVs [75]. miR-19a-con-
organotropic cell lines in vitro [64, 65], their taining EVs produced by astrocytes are trans-
exosome content [66], and human samples have ferred to cancer cells. miR-19a downregulates
been performed [67–70]. Differentially regu- PTEN expression leading to the attraction of
lated miRNAs between primary tumors with or CCR2+ macrophages/microglia as a conse-
without brain relapse or directly at brain metas- quence of the increased production of CCL2
tases [67, 68], as well as liquid biopsies from the from PTENlow cancer cells [76].
cerebrospinal fluid (CSF) [70], have been evalu- The brain microenvironment could be also
ated to validate the importance of selected modulated by cancer cells residing at the primary
candidates. tumor through the production of miR-122-
miRNAs functionally validated in experi- contained EVs. miR-122 targets enzymes
mental models include modulators of extravasa- involved in glucose metabolism. Decreased lev-
tion through the BBB. High levels of miR-181c els of PKM2 and GLUT1 induced by miR-122
contained in extracellular vesicles (EVs) from lead to the reduction of glucose uptake and con-
breast cancer cell lines metastatic to the brain sumption by brain astrocytes, which increases the
are responsible for downregulating the expres- available extracellular pool of this nutrient, thus
sion of PDPK1, which is an essential factor for benefiting incoming cancer cells [77].
actin dynamics by mediating the phosphoryla- Although mesenchymal traits are required at
tion of cofilin. Defective actin dynamics impairs various steps of the metastatic process, some
intracellular trafficking of multiple proteins experimental models show an additional step that
required for the maintenance of brain endothe- takes place upon organ colonization. The process
lial cell intercellular junctions such as tight of mesenchymal to epithelial transition (MET) is
junction proteins and N-cadherin [66]. This regulated by miR-200s family [78]. Liquid biop-
sies from the CSF of patients with parenchymal This process is accompanied by a progressive
or leptomeningeal metastases could be separated increase in expression and activity of the ret
from noncancerous biopsies by a combination of transgene, leading to hyperactivation of the
several miRNAs contained in this family, includ- MAPK-related pathway [16]. The PI3K-AKT-
ing miR-10b, miR-21, miR-200a/c, and miR-141 mTOR pathway has been shown as a viable ther-
[70]. miR-141 is required to mediate MET in apeutic target in several brain metastasis
breast cancer brain metastasis [65]. preclinical models pharmacologically [44, 45].
These transcriptomic screens should be com- Genetically, a melanoma mouse model with acti-
plemented with others that have interrogated the vated AKT1 in the context of BRAF V600E and
epigenome [79–81] and the proteome [82–90]. silenced INK4A-ARF, generated spontaneous
Comparative analysis of omic approaches will brain metastases recapitulating the human dis-
offer a more accurate view of the regulatory ease, and this metastatic capacity was augmented
mechanisms and pathways that are key in experi- by additional PTEN silencing [17] (Fig. 3.1f).
mental models, where investigational therapies This model allows functional validation and
can be tested, and in humans. characterization of PI3K-AKT-mTOR pathway
as key in brain metastasis biology. Although lung
cancer is the most common source of brain
dvanced Modeling of Brain
A metastases, GEMMs of lung cancer scoring inci-
Metastasis in Mice dence of metastatic spread to this secondary
organ are scarce. A GEMM of small-cell lung
Preclinical models extensively used for studying cancer (SCLC), a subtype of lung cancer with
brain metastasis include cell line-derived xeno- high incidence of brain metastasis, has been
transplants, generally based on organotropic reported to generate spontaneous intracranial
human cell lines that preferentially target the brain lesions from neuroendocrine lung tumors that
and are implanted intracardiac or intracranially in were engineered by conditional somatic inactiva-
immunodeficient mice. Syngeneic mouse cell tion of Rb1 and Trp53 in lung epithelial cells
lines with brain tropism have been used to address [18]. These tumors gave rise to extrapulmonary
the interaction of cancer cells with the brain micrometastases including the brain and resembled
environment or the immune system [1, 4–7, 10, human SCLC both morphologically and immun-
15, 41, 76, 91, 92]. However, these models of ophenotypically [18] (Fig. 3.1f), which allows
induced brain metastasis have limitations when more reliable translation of preclinical results
recapitulating the course of the human disease, into clinical approaches. GEMMs that faithfully
where brain metastases are spontaneously gener- recapitulate the human disease will open new
ated in the presence of a primary tumor and gener- scenarios for brain metastasis research such as
ally other extracranial metastases. the study of prevention. Mouse models represent-
ing primary tumors with high incidence of brain
metastasis like non-small-cell lung cancer,
pontaneous Models of Brain
S HER2+ and triple-negative breast cancer are
Metastasis urgently needed.
Genetically engineered mouse models (GEMMs)

that result in spontaneous brain metastases are Patient-Derived Xenografts
limited. Two genetic mouse models of melanoma
based on different oncogenic drivers have been The use of patient-derived xenografts (PDXs)
reported (Fig. 3.1f). The MT/ret transgenic for modeling brain metastasis during the past
mouse model resembles the process of malignant few years [19, 20, 23, 93] has opened new pos-
transformation in human melanoma, resulting in sibilities for personalized medicine to be applied
metastases to distant organs including the brain. to patients with cancer dissemination to the brain
(Fig. 3.1g). PDXs from patients’ brain metasta- but also a valuable resource for evaluating bio-
ses from different primary sources (non-small- markers that predict response to therapy in the
cell lung cancer (NSCLC) [20], several subtypes context of brain metastasis.
of breast cancer [19, 23], and melanoma [93])
have been used to establish preclinical mouse
models by engraftment of cells derived from Future Challenges
fresh surgical samples in immunodeficient mice.
In all studies, PDXs show highly similar histo- Despite the efforts in improving currently avail-
pathological features, genetic or functional prop- able experimental models for brain metastasis,
erties when compared to the parental human whether these models faithfully recapitulate the
brain metastasis, thus proving that PDXs are a human disease is a matter of continuous debate.
reliable resource for recapitulating the human Intracardiac, intracarotid, or intracranial injec-
disease. Based on these similarities, PDXs have tion of brain tropic human or syngeneic cell lines
been used for evaluating the efficacy of targeted are still the most commonly used preclinical
therapies or to perform low-throughput drug models for studying the biology of the disease
screenings. In vitro tumor spheres from PDXs and developing novel therapeutic strategies for
from NSCLC brain metastases that maintain brain metastasis patients. Spontaneous brain
their in vivo brain metastatic potential have been metastases from GEMMs are still limited.
established for this purpose [20]. Five PDX- Available GEMMs [16–18] generate aggressive
derived tumor spheres were screened for 20 primary or extracranial metastases, thus impos-
agents targeting commonly altered oncogenic ing an additional limitation since brain macrome-
pathways in NSCLC such as EGFR, MET, Mtor, tastases are rare and clinically relevant stages of
and VEGFR. Efficacy of these agents varied the disease cannot be easily observed in these
among the different samples, indicating that models. Most PDXs maintain pathological fea-
each one relies on different oncogenic altera- tures of the parental tumor—their increased het-
tions and that personalized approaches based on erogeneity clinically allows more personalized
PDXs will improve current therapies by predict- approaches. CRISPR/Cas9 technology will
ing drug responses. In vivo, inhibition of the improve available models by introducing specific
PI3K/mTOR pathway using a combined treat- genomic alterations detected in human brain
ment with the PI3K inhibitor BKM120 and the metastasis to dissect their functional contribu-
mTOR inhibitor RAD001 (both able to penetrate tion and test their importance as a therapeutic tar-
the brain) resulted in durable tumor regressions get. On the other hand, since most patients have
in 3/5 PDXs of HER2+ breast cancer brain been treated with multiple lines of therapy before
metastases [23], suggesting the potential effi- brain metastases occur, experimental models that
cacy of this combined therapy in the respective incorporate them will allow developing more
donor patients. In this same study, whole-exome realistic experimental studies, which will be fur-
sequencing of the PDXs and matched tumor ther improved by the addition of local therapies
samples from the donor patients showed that such as neurosurgery and radiotherapy.
each PDX and its matched patient sample shared
almost identical genetic alterations regarding Acknowledgments Research at the Brain Metastasis
copy-number variations and somatic mutation Group is supported by MINECO grants MINECO-Retos
SAF2017-89643-R (M.V.), Bristol-Myers Squibb-
rate. Interestingly, the two non-responding PDXs Melanoma Research Alliance Young Investigator Award
and their matched patient specimens showed 2017 (498103) (M.V.), Beug Foundation’s Prize for
hypermutated genomes with enriched mutation Metastasis Research 2017 (M.V.), Fundación Ramón
frequencies in DNA-repair genes, suggesting Areces (CIVP19S8163) (M.V.), Worldwide Cancer
Research (19-0177) (M.V.), H2020-FETOPEN (828972)
that genomic instability is correlated with ther- (M.V.), Clinic and Laboratory Integration Program CRI
apy resistance. Based on these observations, Award 2018 (54545) (M.V.), AECC Coordinated
PDXs are not only a useful tool for drug testing, Translational Groups 2017 (GCTRA16015SEOA) (M.V.),
and La Caixa-Severo Ochoa International PhD Program 13. Jilaveanu LB, Parisi F, Barr ML, Zito CR, Cruz-

Fellowship (L.Z.). M.V. is a Ramón y Cajal Investigator Munoz W, Kerbel RS, et al. PLEKHA5 as a biomarker
(RYC-2013-13365) and EMBO YIP (4053). and potential mediator of melanoma brain metastasis.
Clin Cancer Res. 2015;21(9):2138–47.
14. Wrage M, Hagmann W, Kemming D, Uzunoglu FG,
Riethdorf S, Effenberger K, et al. Identification of
References HERC5 and its potential role in NSCLC progression.
Int J Cancer. 2015;136(10):2264–72.
1. Bos PD, Zhang XH-F, Nadal C, Shu W, Gomis 15. Sevenich L, Bowman RL, Mason SD, Quail DF,
RR, Nguyen DX, et al. Genes that mediate Rapaport F, Elie BT, et al. Analysis of tumour- and
breast cancer metastasis to the brain. Nature. stroma-supplied proteolytic networks reveals a brain-
2009;459(7249):1005–9. metastasis-promoting role for cathepsin S. Nat Cell
2. Schackert G, Fidler IJ. Site-specific metastasis of Biol. 2014;16(9):876–88.
mouse melanomas and a fibrosarcoma in the brain 16. Kato M, Takahashi M, Akhand AA, Liu W, Dai
or meninges of syngeneic animals. Cancer Res. Y, Shimizu S, et al. Transgenic mouse model
1988;48(12):3478–84. for skin malignant melanoma. Oncogene.
3. Priego N, Zhu L, Monteiro C, Mulders M, Wasilewski 1998;17(14):1885–8.
D, Bindeman W, et al. STAT3 labels a subpopulation 17. Cho JH, Robinson JP, Arave RA, Burnett WJ, Kircher
of reactive astrocytes required for brain metastasis. DA, Chen G, et al. AKT1 activation promotes
Nat Med. 2018;24(7):1024–35. development of melanoma metastases. Cell Rep.
4. Valiente M, Obenauf AC, Jin X, Chen Q, Zhang 2015;13(5):898–905.
XH-F, Lee DJ, et al. Serpins promote cancer cell sur- 18. Meuwissen R, Linn SC, Linnoila RI, Zevenhoven
vival and vascular co-option in brain metastasis. Cell. J, Mooi WJ, Berns A. Induction of small cell lung
2014;156(5):1002–16. cancer by somatic inactivation of both Trp53 and
5. Nguyen DX, Chiang AC, Zhang XH-F, Kim JY, Kris Rb1 in a conditional mouse model. Cancer Cell.
MG, Ladanyi M, et al. WNT/TCF signaling through 2003;4(3):181–9.
LEF1 and HOXB9 mediates lung adenocarcinoma 19. Contreras-Zárate MJ, Ormond DR, Gillen AE, Hanna
metastasis. Cell. 2009;138(1):51–62. C, Day NL, Serkova NJ, et al. Development of novel
6. Morsi A, Gaziel-Sovran A, Cruz-Munoz W, Kerbel patient-derived xenografts from breast cancer brain
RS, Golfinos JG, Hernando E, et al. Development metastases. Front Oncol. 2017;7:252.
and characterization of a clinically relevant mouse 20. Lee HW, Lee J-I, Lee SJ, Cho HJ, Song HJ, Jeong DE,
model of melanoma brain metastasis. Pigment Cell et al. Patient-derived xenografts from non-small cell
Melanoma Res. 2013;26(5):743–5. lung cancer brain metastases are valuable translational
7. Schwartz H, Blacher E, Amer M, Livneh N, platforms for the development of personalized tar-
Abramovitz L, Klein A, et al. Incipient melanoma geted therapy. Clin Cancer Res. 2015;21(5):1172–82.
brain metastases instigate astrogliosis and neuroin- 21. Wall BA, Yu LJ, Khan A, Haffty B, Goydos JS, Chen
flammation. Cancer Res. 2016;76(15):4359–71. S. Riluzole is a radio-sensitizing agent in an in vivo
8. Vanharanta S, Shu W, Brenet F, Hakimi AA, Heguy model of brain metastasis derived from GRM1
A, Viale A, et al. Epigenetic expansion of VHL-HIF expressing human melanoma cells. Pigment Cell
signal output drives multiorgan metastasis in renal Melanoma Res. 2015;28(1):105–9.
cancer. Nat Med. 2013;19(1):50–6. 22. Ding L, Ellis MJ, Li S, Larson DE, Chen K, Wallis
9. Yagiz K, Rodriguez-Aguirre ME, Lopez Espinoza JW, et al. Genome remodelling in a basal-like
F, Montellano TT, Mendoza D, Mitchell LA, et al. A breast cancer metastasis and xenograft. Nature.
retroviral replicating vector encoding cytosine deami- 2010;464(7291):999–1005.
nase and 5-FC induces immune memory in meta- 23. Ni J, Ramkissoon SH, Xie S, Goel S, Stover DG,
static colorectal cancer models. Mol Ther Oncolytics. Guo H, et al. Combination inhibition of PI3K and
2018;8:14–26. mTORC1 yields durable remissions in mice bear-
10. Chen Q, Boire A, Jin X, Valiente M, Er EE, Lopez- ing orthotopic patient-derived xenografts of HER2-
Soto A, et al. Carcinoma-astrocyte gap junctions pro- positive breast cancer brain metastases. Nat Med.
mote brain metastasis by cGAMP transfer. Nature. 2016;22(7):723–6.
2016;533(7604):493–8. 24. Pukrop T, Dehghani F, Chuang H-N, Lohaus R,

11. Martínez-Aranda A, Hernández V, Guney E, Muixí L, Bayanga K, Heermann S, et al. Microglia promote
Foj R, Baixeras N, et al. FN14 and GRP94 expression colonization of brain tissue by breast cancer cells in a
are prognostic/predictive biomarkers of brain metas- Wnt-dependent way. Glia. 2010;58(12):1477–89.
tasis outcome that open up new therapeutic strategies. 25. Chuang H-N, van Rossum D, Sieger D, Siam L,
Oncotarget. 2015;6(42):44254–73. Klemm F, Bleckmann A, et al. Carcinoma cells mis-
12. Li B, Wang C, Zhang Y, Zhao XY, Huang B, Wu use the host tissue damage response to invade the
PF, et al. Elevated PLGF contributes to small- brain. Glia. 2013;61(8):1331–46.
cell lung cancer brain metastasis. Oncogene. 26. Morrissy AS, Garzia L, Shih DJH, Zuyderduyn S,
2013;32(24):2952–62. Huang X, Skowron P, et al. Divergent clonal selection
dominates medulloblastoma at recurrence. Nature. tool for identifying new strategies to limit neurocog-
2016;529(7586):351–7. nitive decline. Sci Rep. 2015;5:14384.
27. Mulvenna P, Nankivell M, Barton R, Faivre-Finn C, 39. Yamamoto M, Serizawa T, Shuto T, Akabane A,

Wilson P, McColl E, et al. Dexamethasone and sup- Higuchi Y, Kawagishi J, et al. Stereotactic radio-
portive care with or without whole brain radiotherapy surgery for patients with multiple brain metastases
in treating patients with non-small-cell lung cancer (JLGK0901): a multi-institutional prospective obser-
with brain metastases unsuitable for resection or vational study. Lancet Oncol. 2014;15(4):387–95.
stereotactic radiotherapy (QUARTZ): results from 40. Lockman PR, Mittapalli RK, Taskar KS, Rudraraju V,
a phase 3, non-inferiority, randomised trial. Lancet. Gril B, Bohn KA, et al. Heterogeneous blood-tumor
2016;388(10055):2004–14. barrier permeability determines drug efficacy in
28. Kocher M, Soffietti R, Abacioglu U, Villà S, Fauchon experimental brain metastases of breast cancer. Clin
F, Baumert BG, et al. Adjuvant whole-brain radiother- Cancer Res. 2010;16(23):5664–78.
apy versus observation after radiosurgery or surgical 41. Lyle LT, Lockman PR, Adkins CE, Mohammad

resection of one to three cerebral metastases: results AS, Sechrest E, Hua E, et al. Alterations in peri-
of the EORTC 22952-26001 study. J Clin Oncol. cyte subpopulations are associated with elevated
2011;29(2):134–41. blood-tumor barrier permeability in experimental
29. Frisk G, Tinge B, Ekberg S, Eloranta S, Bäcklund brain metastasis of breast cancer. Clin Cancer Res.
LM, Lidbrink E, et al. Survival and level of care 2016;22(21):5287–99.
among breast cancer patients with brain metastases 42. Palmieri D, Duchnowska R, Woditschka S, Hua

treated with whole brain radiotherapy. Breast Cancer E, Qian Y, Biernat W, et al. Profound prevention of
Res Treat. 2017;166(3):887–96. experimental brain metastases of breast cancer by
30. Jiang T, Su C, Li X, Zhao C, Zhou F, Ren S, et al. temozolomide in an MGMT-dependent manner. Clin
EGFR TKIs plus WBRT demonstrated no survival Cancer Res. 2014;20(10):2727–39.
benefit other than that of TKIs alone in patients with 43. Cao KI, Lebas N, Gerber S, Levy C, Le Scodan R,
NSCLC and EGFR mutation and brain metastases. J Bourgier C, et al. Phase II randomized study of
Thorac Oncol. 2016;11(10):1718–28. whole-brain radiation therapy with or without concur-
31. Aupérin A, Arriagada R, Pignon JP, Le Péchoux C, rent temozolomide for brain metastases from breast
Gregor A, Stephens RJ, et al. Prophylactic cranial cancer. Ann Oncol. 2015;26(1):89–94.
irradiation for patients with small-cell lung cancer in 44. Osswald M, Blaes J, Liao Y, Solecki G, Gömmel M,
complete remission. Prophylactic Cranial Irradiation Berghoff AS, et al. Impact of blood-brain barrier integ-
Overview Collaborative Group. N Engl J Med. rity on tumor growth and therapy response in brain
1999;341(7):476–84. metastases. Clin Cancer Res. 2016;22(24):6078–87.
32. Slotman B, Faivre-Finn C, Kramer G, Rankin E, Snee 45. Nanni P, Nicoletti G, Palladini A, Croci S, Murgo
M, Hatton M, et al. Prophylactic cranial irradiation A, Ianzano ML, et al. Multiorgan metastasis of
in extensive small-cell lung cancer. N Engl J Med. human HER-2+ breast cancer in Rag2-/-;Il2rg-/-
2007;357(7):664–72. mice and treatment with PI3K inhibitor. PLoS One.
33. Takahashi T, Yamanaka T, Seto T, Harada H, Nokihara 2012;7(6):e39626.
H, Saka H, et al. Prophylactic cranial irradiation versus 46. Valiente M, Ahluwalia MS, Boire A, Brastianos PK,
observation in patients with extensive-disease small- Goldberg SB, Lee EQ, et al. The evolving landscape
cell lung cancer: a multicentre, randomised, open- of brain metastasis. Trends Cancer. 2018;4(3):176–96.
label, phase 3 trial. Lancet Oncol. 2017;18(5):663–71. 47. Gril B, Palmieri D, Bronder JL, Herring JM, Vega-
34. Smart D, Garcia-Glaessner A, Palmieri D, Wong-
Valle E, Feigenbaum L, et al. Effect of lapatinib on
Goodrich SJ, Kramp T, Gril B, et al. Analysis of the outgrowth of metastatic breast cancer cells to the
radiation therapy in a model of triple-negative brain. J Natl Cancer Inst. 2008;100(15):1092–103.
breast cancer brain metastasis. Clin Exp Metastasis. 48. Kodack DP, Chung E, Yamashita H, Incio J,

2015;32(7):717–27. Duyverman AMMJ, Song Y, et al. Combined target-
35. Smith DL, Debeb BG, Thames HD, Woodward
ing of HER2 and VEGFR2 for effective treatment of
WA. Computational modeling of micrometastatic HER2-amplified breast cancer brain metastases. Proc
breast cancer radiation dose response. Int J Radiat Natl Acad Sci U S A. 2012;109(45):E3119–27.
Oncol Biol Phys. 2016;96(1):179–87. 49. Kodack DP, Askoxylakis V, Ferraro GB, Sheng Q,
36. Yang H, Lee HW, Kim Y, Lee Y, Choi Y-S, Kim KH, Badeaux M, Goel S, et al. The brain microenviron-
et al. Radiosensitization of brain metastasis by target- ment mediates resistance in luminal breast cancer to
ing c-MET. Lab Invest. 2013;93(3):344–53. PI3K inhibition through HER3 activation. Sci Transl
37. Hamilton AM, Wong SM, Wong E, Foster PJ. Cranial Med. 2017;9(391):pii: eaal4682.
irradiation increases tumor growth in experimen- 50. Porta R, Sánchez-Torres JM, Paz-Ares L, Massutí
tal breast cancer brain metastasis. NMR Biomed. B, Reguart N, Mayo C, et al. Brain metastases from
2018;31(5):e3907. lung cancer responding to erlotinib: the importance of
38. Tomé WA, Gökhan Ş, Brodin NP, Gulinello ME, EGFR mutation. Eur Respir J. 2011;37(3):624–31.
Heard J, Mehler MF, et al. A mouse model replicating 51. Ballard P, Yates JWT, Yang Z, Kim D-W, Yang

hippocampal sparing cranial irradiation in humans: a JC-H, Cantarini M, et al. Preclinical comparison of
osimertinib with other EGFR-TKIs in EGFR-mutant during brain metastasis formation. Neuropathology.
NSCLC brain metastases models, and early evidence 2009;29(4):389–97.
of clinical brain metastases activity. Clin Cancer Res. 63. Sato R, Nakano T, Hosonaga M, Sampetrean O,

2016;22(20):5130–40. Harigai R, Sasaki T, et al. RNA sequencing analy-
52. Crinò L, Ahn M-J, De Marinis F, Groen HJM,
sis reveals interactions between breast cancer
Wakelee H, Hida T, et al. Multicenter phase II study or melanoma cells and the tissue microenviron-
of whole-body and intracranial activity with ceritinib ment during brain metastasis. Biomed Res Int.
in patients with ALK-rearranged non-small-cell lung 2017;2017:8032910.
cancer previously treated with chemotherapy and 64. Okuda H, Xing F, Pandey PR, Sharma S, Watabe M,
crizotinib: results from ASCEND-2. J Clin Oncol. Pai SK, et al. miR-7 suppresses brain metastasis of
2016;34(24):2866–73. breast cancer stem-like cells by modulating KLF4.
53. Gadgeel SM, Shaw AT, Govindan R, Gandhi L,
Cancer Res. 2013;73(4):1434–44.
Socinski MA, Camidge DR, et al. Pooled analysis of 65. Debeb BG, Lacerda L, Anfossi S, Diagaradjane P,
CNS response to alectinib in two studies of pretreated Chu K, Bambhroliya A, et al. miR-141-mediated reg-
patients with ALK-positive non-small-cell lung can- ulation of brain metastasis from breast cancer. J Natl
cer. J Clin Oncol. 2016;34(34):4079–85. Cancer Inst. 2016;108(8).
54. Kim D-W, Tiseo M, Ahn M-J, Reckamp KL,
66. Tominaga N, Kosaka N, Ono M, Katsuda T, Yoshioka
Hansen KH, Kim S-W, et al. Brigatinib in patients Y, Tamura K, et al. Brain metastatic cancer cells
with crizotinib- refractory anaplastic lymphoma release microRNA-181c-containing extracellular ves-
kinase-positive non-small-cell lung cancer: a ran- icles capable of destructing blood-brain barrier. Nat
domized, multicenter phase II trial. J Clin Oncol. Commun. 2015;6:6716.
2017;35(22):2490–8. 67. Hanniford D, Zhong J, Koetz L, Gaziel-Sovran A,
55. Nanjo S, Nakagawa T, Takeuchi S, Kita K, Fukuda Lackaye DJ, Shang S, et al. A miRNA-based signa-
K, Nakada M, et al. In vivo imaging models of bone ture detected in primary melanoma tissue predicts
and brain metastases and pleural carcinomatosis with development of brain metastasis. Clin Cancer Res.
a novel human EML4-ALK lung cancer cell line. 2015;21(21):4903–12.
Cancer Sci. 2015;106(3):244–52. 68. Zhao S, Yu J, Wang L. Machine learning based predic-
56. Ardini E, Menichincheri M, Banfi P, Bosotti R, De tion of brain metastasis of patients with IIIA-N2 lung
Ponti C, Pulci R, et al. Entrectinib, a Pan-TRK, ROS1, adenocarcinoma by a three-miRNA signature. Transl
and ALK inhibitor with activity in multiple molecu- Oncol. 2018;11(1):157–67.
larly defined cancer indications. Mol Cancer Ther. 69. Li Z, Peng Z, Gu S, Zheng J, Feng D, Qin Q,

2016;15(4):628–39. et al. Global analysis of miRNA-mRNA interac-
57. Long GV, Trefzer U, Davies MA, Kefford RF,
tion network in breast cancer with brain metastasis.
Ascierto PA, Chapman PB, et al. Dabrafenib in Anticancer Res. 2017;37(8):4455–68.
patients with Val600Glu or Val600Lys BRAF-mutant 70. Teplyuk NM, Mollenhauer B, Gabriely G, Giese A,
melanoma metastatic to the brain (BREAK-MB): a Kim E, Smolsky M, et al. MicroRNAs in cerebrospi-
multicentre, open-label, phase 2 trial. Lancet Oncol. nal fluid identify glioblastoma and metastatic brain
2012;13(11):1087–95. cancers and reflect disease activity. Neuro Oncol.
58. McArthur GA, Maio M, Arance A, Nathan P, Blank 2012;14(6):689–700.
C, Avril MF, et al. Vemurafenib in metastatic mela- 71. Zhou W, Fong MY, Min Y, Somlo G, Liu L, Palomares
noma patients with brain metastases: an open-label, MR, et al. Cancer-secreted miR-105 destroys vascu-
single-arm, phase 2, multicentre study. Ann Oncol. lar endothelial barriers to promote metastasis. Cancer
2017;28(3):634–41. Cell. 2014;25(4):501–15.
59. Davies MA, Saiag P, Robert C, Grob J-J, Flaherty KT, 72. Xing F, Sharma S, Liu Y, Mo YY, Wu K, Zhang
Arance A, et al. Dabrafenib plus trametinib in patients YY, et al. miR-509 suppresses brain metastasis of
with BRAF(V600)-mutant melanoma brain metasta- breast cancer cells by modulating RhoC and TNF-α.
ses (COMBI-MB): a multicentre, multicohort, open- Oncogene. 2015;34(37):4890–900.
label, phase 2 trial. Lancet Oncol. 2017;18(7):863–73. 73. Tsao N, Hsu HP, Wu CM, Liu CC, Lei HY. Tumour
60. Zubrilov I, Sagi-Assif O, Izraely S, Meshel T,
necrosis factor-alpha causes an increase in blood-
Ben- Menahem S, Ginat R, et al. Vemurafenib brain barrier permeability during sepsis. J Med
resistance selects for highly malignant brain and Microbiol. 2001;50(9):812–21.
lung-metastasizing melanoma cells. Cancer Lett. 74. Oskarsson T, Batlle E, Massagué J. Metastatic stem
2015;361(1):86–96. cells: sources, niches, and vital pathways. Cell Stem
61. Park ES, Kim SJ, Kim SW, Yoon S-L, Leem S-H, Kim Cell. 2014;14(3):306–21.
S-B, et al. Cross-species hybridization of microarrays 75. Wasilewski D, Priego N, Fustero-Torre C, Valiente
for studying tumor transcriptome of brain metastasis. M. Reactive astrocytes in brain metastasis. Front
Proc Natl Acad Sci U S A. 2011;108(42):17456–61. Oncol. 2017;7:298.
62. Saito N, Hatori T, Aoki K, Hayashi M, Hirata Y, Sato 76. Zhang L, Zhang S, Yao J, Lowery FJ, Zhang Q, Huang
K, et al. Dynamics of global gene expression changes W-C, et al. Microenvironment-induced PTEN loss by
exosomal microRNA primes brain metastasis out- 85. Li F, Glinskii OV, Zhou J, Wilson LS, Barnes S,
growth. Nature. 2015;527(7576):100–4. Anthony DC, et al. Identification and analysis of signal-
77. Fong MY, Zhou W, Liu L, Alontaga AY, Chandra ing networks potentially involved in breast carcinoma
M, Ashby J, et al. Breast-cancer-secreted miR- metastasis to the brain. PLoS One. 2011;6(7):e21977.
122 reprograms glucose metabolism in premeta- 86. Dun MD, Chalkley RJ, Faulkner S, Keene S, Avery-
static niche to promote metastasis. Nat Cell Biol. Kiejda KA, Scott RJ, et al. Proteotranscriptomic
2015;17(2):183–94. profiling of 231-BR breast cancer cells: identifica-
78. Korpal M, Kang Y. The emerging role of miR-200 tion of potential biomarkers and therapeutic tar-
family of microRNAs in epithelial- mesenchymal gets for brain metastasis. Mol Cell Proteomics.
transition and cancer metastasis. RNA Biol. 2015;14(9):2316–30.
2008;5(3):115–9. 87. Improta G, Zupa A, Fillmore H, Deng J, Aieta M,
79. Pangeni RP, Channathodiyil P, Huen DS, Eagles LW, Musto P, et al. Protein pathway activation mapping of
Johal BK, Pasha D, et al. The GALNT9, BNC1 and brain metastasis from lung and breast cancers reveals
CCDC8 genes are frequently epigenetically dysregu- organ type specific drug target activation. J Proteome
lated in breast tumours that metastasise to the brain. Res. 2011;10(7):3089–97.
Clin Epigenetics. 2015;7:57. 88. Zila N, Bileck A, Muqaku B, Janker L, Eichhoff

80. Marzese DM, Scolyer RA, Huynh JL, Huang SK, OM, Cheng PF, et al. Proteomics-based insights into
Hirose H, Chong KK, et al. Epigenome-wide DNA mitogen-activated protein kinase inhibitor resistance
methylation landscape of melanoma progression to of cerebral melanoma metastases. Clin Proteomics.
brain metastasis reveals aberrations on homeobox D 2018;15:13.
cluster associated with prognosis. Hum Mol Genet. 89. Hoshino A, Costa-Silva B, Shen T-L, Rodrigues G,
2014;23(1):226–38. Hashimoto A, Tesic Mark M, et al. Tumour exosome
81. Salhia B, Kiefer J, Ross JTD, Metapally R, Martinez integrins determine organotropic metastasis. Nature.
RA, Johnson KN, et al. Integrated genomic and epig- 2015;527(7578):329–35.
enomic analysis of breast cancer brain metastasis. 90. Chen EI, Hewel J, Krueger JS, Tiraby C, Weber

PLoS One. 2014;9(1):e85448. MR, Kralli A, et al. Adaptation of energy metabo-
82.
Sanz-Pamplona R, García-García J, Franco S, lism in breast cancer brain metastases. Cancer Res.
Messeguer X, Driouch K, Oliva B, et al. A taxonomy 2007;67(4):1472–86.
of organ-specific breast cancer metastases based on 91. Pencheva N, Buss CG, Posada J, Merghoub T,

a protein-protein interaction network. Mol Biosyst. Tavazoie SF. Broad-spectrum therapeutic suppression
2012;8(8):2085–96. of metastatic melanoma through nuclear hormone
83. Martín B, Aragüés R, Sanz R, Oliva B, Boluda S, receptor activation. Cell. 2014;156(5):986–1001.
Martínez A, et al. Biological pathways contributing to 92. Malladi S, Macalinao DG, Jin X, He L, Basnet H,
organ-specific phenotype of brain metastatic cells. J Zou Y, et al. Metastatic latency and immune eva-
Proteome Res. 2008;7(3):908–20. sion through autocrine inhibition of WNT. Cell.
84. Mustafa DAM, Pedrosa RMSM, Smid M, van der 2016;165(1):45–60.
Weiden M, de Weerd V, Nigg AL, et al. T lymphocytes 93. Krepler C, Sproesser K, Brafford P, Beqiri M, Garman
facilitate brain metastasis of breast cancer by induc- B, Xiao M, et al. A comprehensive patient-derived
ing Guanylate-Binding Protein 1 expression. Acta xenograft collection representing the heterogeneity of
Neuropathol. 2018;135(4):581–99. melanoma. Cell Rep. 2017;21(7):1953–67.
Pathology of Brain Metastases
4
David J. Pisapia
Neuropathological assessment of brain metasta- the differential diagnosis. Metastases tend to

ses first and foremost establishes that a given be well circumscribed rather than infiltrative
intraparenchymal or meningeal lesion in fact rep- lesions, and they are usually T1-hypo to isoin-
resents metastatic disease and that all potential tense and enhancing [1]. While multiplicity
histological mimics have been ruled out. Second, may also be a clue, up to 72% of patients with
determining the site of origin is paramount, not metastatic disease to the brain may present
only if a primary site of disease has yet to be clin- with a solitary mass [2]. Metastases overall are
ically or radiologically identified, but also in more common within the supratentorial com-
order to rule out a second malignancy in a patient partment, in watershed areas, and at the gray-
with a known primary. Finally, the pathologist white junction [3]. Approximately 80% are
must ensure that appropriate molecular charac- located in the cerebral hemispheres and 15% in
terization of the metastasis is performed, when the cerebellum [4]. Certain tumor types may
necessary, following guidelines that have been demonstrate anatomical predilection within the
established for the relevant primary tumor type. brain; for example, evidence suggests that
We will consider each of these components to renal cell carcinoma has a proclivity to involve
pathological assessment in this chapter. the intraventricular compartment and choroid
plexus [5]. Associations between particular
primary tumor types and the vascular territory
Identifying Metastatic Disease in which metastases are more likely to be found
have been noted. For example, melanoma is
Intraoperative Consultation relatively less common in the cerebellum and
breast cancer is less common in the posterior
Initial histological assessment of potential cerebral arterial vascular distribution [6].
metastatic disease often first occurs during Once it is established that metastatic disease
intraoperative consultation. The neuropatholo- lies within the radiological differential diagno-
gist is typically guided by preoperative MRI sis, imaging of the chest, abdomen, and pelvis
findings suggesting that a metastasis is within may be performed prior to a brain biopsy or
resection in an effort to identify the primary
malignancy and/or a more readily accessible
metastatic deposit for biopsy. Among patients
D. J. Pisapia (*)
Department of Pathology and Laboratory Medicine,
with metastatic cancer, cancers with higher rates
Weill Cornell Medicine, New York, NY, USA of brain involvement include melanoma (28%),
e-mail: djp2002@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_4
54 D. J. Pisapia
small-cell carcinoma of the lung, lung adeno- Additional histological ambiguity may be
carcinoma, and non-small-cell carcinoma of the present in several circumstances. Tumors with a
lung (not otherwise specified), each with high nuclear to cytoplasmic ratio and small blue
approximately 25% of patients developing brain cell morphology present a diagnostic challenge
metastases [7]. Other cancers with a relatively at the time of frozen section with a differential
high proportion of patients experiencing dissem- diagnosis that includes metastatic small-cell
ination to the brain include squamous cell carci- carcinoma, lymphoma, glioblastoma with small
noma of the lung (15%), renal cell carcinoma cell features, central nervous system (CNS)
(11%), carcinoma of the breast (8%), testicular embryonal tumors, and metastatic primitive
cancer (7%), and esophageal carcinoma (5%, neuroectodermal tumors derived from extracra-
among patients with metastatic disease) [7]. nial sites (Fig. 4.2). Complicating the picture for
With routine intraoperative squash prepara- dural-based metastases is the possibility of
tion and frozen section analysis, the histological atypical or anaplastic meningioma, which may
distinction between metastatic carcinoma and also demonstrate epithelioid characteristics.
primary glial neoplasms is usually trivial Moreover, the rare phenomenon of meningio-
(Fig. 4.1). On squash preparation, carcinoma mas secondarily harboring metastases has been
cells tend to cluster, display distinct cytoplasmic well documented, and neuropathologists should
borders and may show prominent nucleoli be alert to the possibility of two distinct cell
(Fig. 4.1a, b). In the case of adenocarcinomas, populations appearing on the slide [8, 9].
intracellular vacuoles with mucin production Complicating the picture for patients with tumor
may be seen. In contrast, most primary gliomas predisposition syndromes such as von Hippel-
demonstrate spindled nuclei with fibrillar pro- Lindau syndrome (VHL) is an increased pretest
cesses (Fig. 4.1f). On frozen section, the archi- probability for both metastatic disease and
tectural features are valuable in distinguishing intracranial primaries. For VHL patients with a
between the often pushing border between cerebellar lesion, the differential diagnosis may
native brain parenchyma and carcinoma cells include both hemangioblastoma and metastatic
and the insidiously infiltrative nature of neo- renal cell carcinoma. Moreover, there may exist
plastic glial cells (Fig. 4.2). Moreover, gland histological overlap on frozen section between
formation, squamous nests, or other epithelial the stromal cells of hemangioblastoma and the
characteristics are usually apparent on frozen cytoplasmic clearing of renal cell carcinoma,
sections. Challenges arise in the case of glial clear cell type.
neoplasms, which may on occasion display epi- Metastatic melanoma deserves special men-
thelioid characteristics (Figs. 4.1e and 4.3). In tion in this discussion. While some of these
particular, epithelioid glioblastoma and pleo- tumors may demonstrate melanin pigment as a
morphic xanthoastrocytoma may sometimes clue to the frozen section pathologist (Fig. 4.1c),
mimic carcinoma, with dyshesive, plump cells melanin production is by no means specific for
predominating on squash preparation (Fig. 4.3a) metastatic melanoma. In particular, for lesions
or frozen section. At the same time, on occasion arising within the extra-axial, meningeal com-
a seemingly diffusely infiltrative, perivascular partment one must also consider primary mela-
distribution of markedly atypical carcinoma nocytic lesions of the CNS. As discussed later,
cells, such as those seen in pleomorphic carci- with the advent of next generation sequencing
noma of the lung, can appear similar to an epi- techniques it is becoming easier to resolve this
thelioid glial malignancy (Fig. 4.3c). differential in difficult cases. Finally, for paraspi-
Immunohistochemical staining for GFAP (e.g., nal tumors, or those mass lesions associated with
Fig. 4.3b) or TTF1 (e.g., Fig. 4.3d) may help cranial nerves, melanotic schwannoma may also
resolve the differential (see also caveat with enter into the differential diagnosis of pigmented
respect to TTF1 staining discussed below). lesions. The distinction between metastasis and
4 Pathology of Brain Metastases 55
Fig. 4.1 Intraoperative squash preparation of lesional tis- ated with the meningeal compartment. (d) The round
sue often represents the first histological encounter with regular nuclei and stippled chromatin pattern seen here is
potential metastases. (a) H&E-stained squash prep of characteristic of neuroendocrine neoplasms, particularly
fresh tissue reveals clusters of polygonal epithelial cells pituitary adenoma for the neuropathologists; however,
with a clumped and cohesive smearing pattern. Cells may metastatic neuroendocrine neoplasms may also enter the
also demonstrate prominent nucleoli and intracytoplasmic differential here. (e) In this case of an astrocytoma with
vacuolization. This case represents metastatic adenocarci- epithelioid and gemistocytic features, the histological dis-
noma of the lung. (b) A case of squamous cell carcinoma tinction with metastasis may become more challenging.
of lung origin, also demonstrating cellular cohesion. (c) (f) Compare the rounded cytoplasmic features of panel E
Melanin pigment is clearly visible on this squash prep with the more characteristic fibrillar processes of less
placing metastatic melanoma at the top of the differential unusual neoplastic glial cells, seen here in a pilocytic
diagnosis for intraparenchymal lesions, as well as primary astrocytoma. All panels show H&E-stained squash
melanocytic neoplasms of the CNS if the lesion is associ- preparations
56 D. J. Pisapia
Fig. 4.2 Intraoperative frozen section of small-cell carci- ing characteristic of small-cell carcinoma, are better
noma of lung. (a) H&E-stained frozen section demon- appreciated here on an H&E-stained permanent section.
strates a population of small blue cells with scant (c–f) Immunohistochemical staining confirms the diagno-
cytoplasm forming a well-demarcated border with adja- sis with tumor cells showing labeling for synaptophysin
cent brain parenchyma. The architectural features includ- (c) and cytokeratin 7 (d), an absence of labeling for glial
ing the border itself is helpful since the cytologic fibrillary acidic protein (GFAP), which highlights the
differential may be broad on frozen section and could adjacent reactive brain parenchyma (e). TTF1 staining is
include lymphoma as well as glioblastoma with small cell also positive in tumor cells (f)
features. (b) Cytologic features, such as the nuclear mold-
primary peripheral nerve sheath tumors or pri- cell carcinoma, and primary intracranial terato-
mary meningeal melanocytic lesions is obviously mas can harbor secondary malignant carcino-
crucial to properly stage the patient’s disease and mas [10, 11].
determine further clinical management. Finally, in the sellar and suprasellar region,
Rarer, but certainly relevant to the discus- the neuropathologist is occasionally confronted
sion, are de facto epithelial malignancies that with a differential diagnosis of pituitary adenoma
arise as primaries within the intracranial com- versus metastatic carcinoma (Fig. 4.1d); assess-
partment. For example, epidermoid cysts may ment for metastatic neuroendocrine carcinomas
undergo malignant transformation to squamous often necessitates immunohistochemical stains.
Fig. 4.3 Histological mimics of carcinoma such as epi- appearing to be negative. What are likely reactive astro-
thelioid glioblastoma may present diagnostic challenges. cytic processes additionally can be seen in the back-
(a) Intraoperative squash preparation of glioblastoma ground. (c) Occasionally, highly pleomorphic carcinoma
often includes highly pleomorphic cells. In this example, cells may mimic glioma cells. Here, in a case of pleomor-
some cells demonstrate epithelioid features. Occasional phic adenocarcinoma of lung origin, some cells appear to
glial processes are also present. (b) Immunohistochemical be diffusely infiltrating the brain. (d) Immunohistochemical
staining for glial acidic fibrillary protein (GFAP) reveals staining demonstrates TTF-1 immunoreactivity in these
positive labeling in only some tumor cells, with others cells, corroborating their origin from the lung
As an example from our own practice, a pituitary Permanent Histology

tumor histologically compatible with pituitary and Immunohistochemical
adenoma on hematoxylin and eosin (H&E)- Assessment of Brain Metastases
stained sections demonstrated strong nuclear
expression for Nkx3.1, confirming a prostatic While a history of known primary malignancy is
metastasis in a man with known prostate cancer often provided to the pathologist, it is essential to
with neuroendocrine features. confirm that the metastatic lesion is consistent
58 D. J. Pisapia
a b
CK7+ CK7− Pancreatic Nuclear markers
Biliary
Gastric
CK20+ Pancreatic Urothelial GATA3
Biliary
Gastric Colonic* Breast
Pulmonary TTF1
Urothelial
Thyroid
Salivary PAX8
Uterine
Pulmonary Ovarian ER
Prostate
Breast Small cell
Hepatocellular Nkx3.1
CK20−
Thyroid Adrenocortical Renal cell

Salivary Renal cell*** Prostate
Small cell Hepatocellular CDX2
Uterine
Ovarian** Adrenocortical
Colonic
Fig. 4.4 Immunohistochemical workup of metastases to properly referred to as Nkx2.1 or NK2 homeobox 1),
determine site of origin. (a) The cytokeratin profile of epi- GATA3 GATA binding protein 3, PAX8 paired box 8, ER
thelial tumor cells can inform the likely site of origin. In estrogen receptor, encoded by ESR1, Nkx3.1 NK3
particular, the pattern of CK7 and CK20 reactivity is often homeobox 1, CDX2 caudal type homeobox 2. *Higher
used to narrow down the differential diagnosis. (b) Further stage and/or rectal location may correlate with increased
immunostaining with more specific markers such as tran- CK7 expression. **Whereas ovarian serous adenocarci-
scription factors that show specificity for different cell noma is negative for CK20, mucinous ovarian neoplasms
types is now routinely used for diagnostic purposes. may exhibit CK20 labeling. ***Reactivity for CK7 is
TTF1, thyroid transcription factor one (while TTF1 is the typically negative in clear cell renal cell carcinoma, but
term commonly used for this protein, the encoding gene is positive in chromophobe renal cell carcinoma
with the known primary or whether it represents degrees of sensitivity and specificity to confirm a
a manifestation of a distinct primary neoplasm. site of origin. Antibodies recognizing cytokeratin
In one study, the percentage of cancer patients (CK) intermediate filaments of differing molecular
ultimately demonstrating synchronous distinct weights are typically applied since many carcino-
neoplasms was shown to be 15–17% [12]. mas demonstrate characteristic patterns of CK
If hematoxylin and eosin (H&E)-stained reactivity (Fig. 4.4a). For example, while colon
slides are available from a prior biopsy or resec- adenocarcinoma typically labels for CK20, pul-
tion at the primary site, the histological features monary adenocarcinoma demonstrates predomi-
of the tumor may be directly compared to the nant staining for CK7. Those carcinomas arising
metastasis, and in some cases morphological from a peri-diaphragmatic location, for example,
similarity serves as the gold standard to confirm pancreatic carcinoma or gastric carcinoma, often
the suspected site of origin. Even if a tumor is display reactivity for both CK7 and CK20. A host
determined to be originating from a particular of transcription factors with protein expression
organ by immunohistochemical means (as dis- localized to the nucleus are also used to character-
cussed below), morphological assessment is ize the likely site of origin. Examples include
important to determine which primary lesion Nkx3.1 (prostate) [13], TTF-1 (lung, thyroid) [14],
seeded the metastasis in a patient with multiple CDX-2 (colon) [15], and PAX-8 (renal, Mullerian
primary lesions in that organ (e.g., in a patient malignancies, and thyroid) [16] (Fig. 4.4b).
with multiple lung or breast primaries). The utility of GATA binding protein 3
(GATA3) illustrates a few important principles
regarding clinical use of transcription factor
Immunohistochemistry immunohistochemistry. While GATA3 had been
identified as playing a role in T-cell lymphocyte
Most pathology laboratories employ a battery of development since the early 1990s [17] and its
immunohistochemical stains, each with variable association with ER-positive breast cancer cell
lines was demonstrated as early as 1999 [18], its giomas are classically positive for epithelial
emergence as a clinically useful marker with rou- membrane antigen (EMA) and SSTR2 but nega-
tine and widespread use in pathology laboratories tive for CK.
is much more recent (within the last 5 years) Stains must be interpreted in the appropriate
[19–21]. It is now recognized that GATA3 is a context and pathologists must have familiarity
sensitive marker for most breast and urothelial with the limitations, cross-reactivities, and stain-
carcinomas. However, iterative reevaluation of its ing characteristics of each antibody clone used,
specificity is subject to comprehensive tissue including its performance characteristics in a par-
studies (evaluation over large numbers of tumor ticular laboratory. For example, CK AE1/AE3
samples using tissue microarrays) or simply antibody cocktails (commonly referred to as pan
through close attention to clinical findings in cytokeratin) may show extensive cross reactivity
unusual circumstances. Indeed, in addition to to astroglial antigens [26]. In rare cases, it may
breast and urothelial malignancies, robust therefore be necessary to employ an antibody
GATA3 expression is detected in a majority of such as Cam5.2 to reliably distinguish between
paragangliomas (including pheochromocytoma), carcinoma and glioma. The morphological fea-
basal cell, adnexal, and squamous carcinomas of tures in combination with immunohistochemical
this skin, chromophobe renal cell carcinoma, staining are usually sufficient to resolve aberrant
choriocarcinoma, and mesothelioma, as well as pan-cytokeratin staining. It should be noted that
less commonly in many other tumor types [21]. certain antibody clones for TTF1 are known to
In an educational example from our own neu- react more promiscuously in glial tumors [27],
ropathology service, a patient with a history of and bona fide TTF1 expression is increasingly
breast cancer and an intrasellar tumor was found recognized in a range of mostly midline primary
to have a GATA3-postive epithelial neoplasm intracranial neoplasms, including pituicytomas,
upon resection that was concerning for breast other neoplastic processes of the posterior pitu-
metastasis. Further investigation revealed that itary gland [28], and chordoid glioma of the third
GATA3 labeling is in fact characteristic of ventricle, a lesion potentially associated with the
gonadotroph- lineage pituitary adenomas as in organum vasculosum of the stria terminalis [29].
this case (possibly due to cross reactivity with Beyond strict identification of a site of origin,
GATA2 protein), a finding that was later corrobo- ancillary immunohistochemical staining may be
rated by other investigators [22]. Thus, as new necessary for additional diagnostic, prognostic, or
“markers” are incorporated into the lab, it is theragnostic purposes. In breast metastases, it is
essential to consider each case within its proper standard to assess the hormone receptor status for
clinical context and to refrain from making estrogen receptor (ER) and progesterone receptor
assumptions about specificity or expected stain- (PR) as well as for the status of HER2 expression.
ing patterns in poorly characterized tumor types. These staining patterns are associated with recog-
Other commonly used markers for the workup nized molecular subtypes of breast cancer includ-
of metastatic disease include S100, HMB45, ing luminal (typically hormone receptor positive
A103, and SOX10 (melanoma) [23], p40 (squa- and HER2 negative), HER2-enriched (hormone
mous cell carcinoma (SqCC), including SqCC of receptor negative and HER2 positive), and triple
the lung) [24], and HepPar1, arginase-1, and negative breast carcinoma including basal-like
glypican-3 (hepatocellular carcinoma) [25]. tumors (hormone receptor and HER2 negative)
Again, because any one particular antibody [30]. In cases with ambiguous HER2 immunohis-
may not demonstrate sufficient specificity, anti- tochemical labeling, FISH is employed to resolve
body panels may be required to confirm a site of the amplification status of this gene. At our insti-
origin. Finally, antibodies may be necessary to tution, we employ guidelines developed by the
exclude the histological mimics mentioned American Society of Clinical Oncology and the
above. For example, epithelioid astrocytic tumors College of American Pathologists for the report-
typically display labeling for glial fibrillary ing of ER, PR, and HER2 stains (Fig. 4.5e–h) [31,
acidic protein (GFAP; Fig. 4.3a, b) while menin- 32]. In prostatic adenocarcinomas, we also
60 D. J. Pisapia
Fig. 4.5 Sample immunohistochemical workup including H&E again demonstrates a nest of pleomorphic epithelial
diagnostic and treatment-relevant biomarkers. (a–d) A cells, this time with an area of central tumor necrosis (e).
case of metastatic adenocarcinoma of the lung. In this case, The immunohistochemical profile indicates positive label-
the H&E stained section demonstrates highly pleomorphic ing for estrogen receptor (ER) in the vast majority of cells
epithelial cells with abundant cytoplasm and prominent (f) and progesterone receptor (PR) in a minority of cells
nucleoli (a). The immunohistochemical profile supports a (g). Because the threshold for considering a tumor to be
diagnosis of lung adenocarcinoma with positive nuclear positive for PR is low (≥1%), this tumor is considered
staining for TTF1 (b) and cytoplasmic staining for CK7 positive for both ER and PR. The tumor also demonstrates
(c). Moreover, the tumor demonstrates significant labeling strong circumferential membrane staining that is complete,
for PDL1 (d), making this patient a candidate for check- intense and within >10% of tumor cells, and it is therefore
point inhibitor therapy with antibodies directed against the considered positive for HER2 by immunohistochemistry
PD1/PDL1 T cell signaling mechanism. (e–h) In this case (h). An equivocal result by IHC would be followed up with
of metastatic ductal adenocarcinoma of the breast, the assessment using fluorescent in situ hybridization (FISH)
Fig. 4.5 (continued)
routinely assess for neuroendocrine differentia- susceptibility to immune checkpoint inhibitors is

tion using antibodies for synaptophysin and chro- likely to be increasingly incorporated into the
mogranin, which may alter treatment strategies routine antigen characterization of brain metasta-
[33]. Immunostaining for MLH1, PSM2, MSH2, ses (e.g., Fig. 4.5a–d). Indeed, it has been demon-
and MSH6 may be used to assess for the expres- strated that there may be significant discordance
sion of mismatch repair proteins in colorectal car- between the PD-L1 labeling characteristics
cinomas and other tumors [34]. between primary and metastatic lesions, arguing
Finally, immunohistochemical assessment of for the de novo assessment of this biomarker in
PD-L1 as a marker of potential therapeutic metastatic foci [35].
62 D. J. Pisapia
olecular Assessment of Brain

M STAT6-NAB2 fusions as seen in solitary fibrous
Metastases tumor, also referred to as solitary fibrous tumor/
hemangiopericytoma in the central nervous sys-
The molecular characterization of neoplastic dis- tem). Finally, immunohistochemistry can assess
ease, including in metastatic lesions, has for overexpression of particular protein products,
increased dramatically over the last decade. As such as in the setting of ERBB2 (HER2) amplifi-
recurrent alterations become incorporated into cation in breast cancer and ALK rearrangement in
diagnostic schema as entity-defining parameters lung adenocarcinomas.
for tumor subtypes, especially to guide targeted Molecular assessment is essential, particularly
therapies, multiple clinical assays have been if the metastatic lesion represents the first tissue
developed to interrogate molecular alterations in sampling of a neoplastic process. For lung adeno-
tumor tissue; most of these are readily applied to carcinomas in such circumstances, we routinely
formalin fixed paraffin embedded samples. assess for EGFR, KRAS, ALK, and ROS1 altera-
While many laboratories at academic medical tions using a combination of immunohistochem-
centers offer a comprehensive menu of molecular istry, FISH, and targeted sequencing assays. In
tests, some clinical care settings may use com- particular, the 143-gene NGS assay employed in
mercial laboratories. The advent of next genera- our Clinical Genomics Laboratory can detect all
tion sequencing technologies with massively of these alterations in a single assay—EGFR and
parallel strategies has enabled the assessment of KRAS hotspot mutations are detected in the
larger panels of genes with relatively low input deoxyribonucleic acid (DNA)-based portion of
tissue requirements. At our institution, we offer the assay while ROS1, ALK, and MET exon 14
single gene assays (utilizing immunohistochemi- skipping are found by the RNA-based compo-
cal proxies, FISH, or PCR-based assays), a nent. In addition to MMR protein assessment
50-gene targeted sequencing panel, and a larger with immunohistochemistry for adenocarcinoma
targeted panel that interrogates 143 genes. This of the colon, microsatellite instability may be
latter assay assesses oncogenic hotspot loci, assessed via molecular means. Moreover, we can
whole exon assessment of a set of tumor suppres- identify KRAS and NRAS alterations via our tar-
sors, copy number alteration status for a subset of geted sequencing panel.
genes, and also features an RNA-based compo- Rarely, molecular assays are used to resolve
nent for the detection of selected fusion transcripts diagnoses in cases with ambiguous histological
(e.g., for the detection of ROS1 and ALK rear- and/or immunohistochemical features. For exam-
rangements). We additionally offer a clinically ple, returning to the differential diagnosis of
validated whole exome sequencing test that inter- melanocytic lesions, while BRAF and NRAS
rogates 22,000 genes and reports somatic altera- mutations are enriched in melanomas derived
tion calls relative to coincident germline from cutaneous sites, GNAQ and GNA11 muta-
sequencing derived from a peripheral blood sam- tions are much more common in primary mela-
ple; a targeted solid tumor gene panel that inter- nocytic lesions of the CNS as well as in uveal
rogates 500 genes is also currently under melanoma [36]. In contrast, melanotic schwan-
validation. nomas, which occasionally enter into this differ-
Examples of immunohistochemical proxies ential diagnosis in paraspinal locations, are
for molecular alterations include antibodies characterized by PRKAR1A mutations (akin to
directed against mutated epitopes (e.g., BRAF those seen in the context of Carney syndrome)
V600E). Antibodies may also be used to detect and may additionally show loss of heterozygosity
aberrant localization of protein epitopes that at this locus [37]. Finally, even if a mutation char-
result from fusion gene products (e.g., STAT6 acteristic of primary melanocytic CNS tumors is
localization to the nucleus in the setting of identified, a comprehensive dermatologic and
ophthalmologic examination as well as PET/CT the brain is likely attributable to the fact that the
evaluation should be performed. epigenetic footprint of a tumor encodes infor-
In a second example of molecular diagnostics, mation concerning both developmental as well
poorly differentiated neoplasms with ambiguous as oncogenic pathways for a particular cell pop-
histological features or poorly sampled lesions ulation [42, 43].
may occasionally prompt consideration of both Since the presence of brain metastases de
primary CNS tumors as well as metastases from facto represents advanced stage disease, clini-
distant sites. For example, if a tumor were found cians are more frequently using larger sequenc-
to harbor loss of chromosome 10, gain of chromo- ing panels in an effort to detect targetable
some 7, deletion of CDKN2A/B, TERT promoter alterations that may not have been present (or
mutation, and EGFR amplification (all of which may not have been assessed) in the patient’s
are alterations characteristic of glioblastoma), the primary site of disease. Targetable alterations
molecular results would suggest glioblastoma may be identified using next generation
over a metastatic carcinoma or sarcoma. If the sequencing of brain metastatic tissue in patients
molecular profile of a primary tumor is known, it without significant molecular information
can be compared to the profile of a putative regarding their primary site of disease. Salient
metastasis in an effort to molecularly prove deri- examples from our own service include a
vation from the primary. patient with metastatic melanoma whose tumor
Indeed, it is likely that molecular testing will was found to harbor a BRAF V600E mutation
be increasingly used to detect the clonal rela- (Fig. 4.6a, b), and a second patient with meta-
tionship between a given metastasis and its pre- static adenocarcinoma of the colon that was
sumed primary. In one study that matched 86 found to have ERBB2 (HER2) amplification in
paired primary tumors with brain metastases, addition to several additional nontargetable but
four metastases were found to be clonally unre- characteristic alterations, namely, KRAS, APC,
lated to the sampled primary tumor and were and TP53 mutations (Fig. 4.6c–e). In the study
hypothesized to have arisen from a clonally dis- of matched primary and brain metastatic sam-
tinct neoplasm within the same primary organ ples referenced earlier, 53% (46/86) of cases
due to a high-risk oncogenic field effect (three harbored at least one potentially actionable
of these were lung carcinomas that occurred in alteration in brain metastatic samples that was
the setting of smoking exposure, and one was a not identified in the paired primary tumor sam-
breast cancer arising in the setting of a germline ple. These alterations included those affecting
BRCA1 mutation) [38]. the PI3K/AKT/mTOR pathway, such as PTEN
Just as epigenetic profiling of primary neo- and PIK3CA, as well as ERBB2 (HER2) and
plasms of the CNS has recently demonstrated EGFR alterations, indicating susceptibility to
great promise as a diagnostic aide [39], there tyrosine kinase inhibition [38]. In addition to
have been improved outcomes using methyla- individual molecular alterations, the overall
tion profiling for the identification of metasta- tumor mutational burden (TMB) may alter clin-
ses of unknown primary and applying ical management [44]. A tumor’s TMB can be
cancer-specific treatments, rather than more gleaned from broader NGS panels—its rela-
generalized empirical treatment strategies [40]. tionship to predicted neoantigen production,
In one study, therapeutically relevant subtypes immune system regulation, and response to
of melanoma, breast, and lung cancers meta- immunotherapy (such as checkpoint inhibitors)
static to brain were successfully classified on remains to be fully elucidated [45, 46].
the basis of methylation profiling [41]. The In addition to DNA-based and epigenetic
robustness of epigenetic profiling in classifying assessment of metastases, other modalities includ-
both primary and secondary malignancies of ing RNA-sequencing paradigms, metabolomic
64 D. J. Pisapia
B
Tier 1 Variants
Variant Type VAF CN

BRAF c.1799T>A, p.Val600Glu missense 43.0% N/A
E
Tier 1 Variants
Variant Type VAF CN

KRAS c.35G>T, p.Gly12Val Missense 57.4% N/A
Tier 2 Variants
Variant Type VAF CN

APC c.4242_4245deIAAGT, p.Ser1415fs Frameshift deletion 89.9% N/A
TP53 c.560-1G>A, p.? Splice site 63.7% N/A
ERBB2 Amplification N/A 42.9
Fig. 4.6 Sample molecular workup of metastatic disease and CK20 positivity (not shown), corroborating the site of
using next generation sequencing. (a) In this case of meta- origin as colon. (d) Next generation sequencing demon-
static melanoma, the H&E depicts sheets of epithelioid strated several mutations that are characteristic of colonic
cells with large nuclei, abundant cytoplasm, and focal adenocarcinoma including KRAS, APC, and TP53 muta-
dark melanin pigment. (b) Next generation sequencing tions. In addition, the sequencing panel is able to detect a
revealed a BRAF V600E point mutation. (c) H&E- subset of copy number alterations and in this case showed
staining shows islands of carcinoma cell surrounded by amplification of ERBB2 (HER2), a potentially targetable
areas of necrosis. Immunostaining demonstrated CDX2 alteration
Table 4.1 Commonly assessed biomarkers in brain metastases

Primary site Commonly assessed proteins Commonly assessed genes
Breast GATA3, GCDFP-15, Mammaglobin, ER, PR, HER2, ERBB2/HER2, BRCA1,
CK7 BRCA2
Lung, adenocarcinoma TTF1, Napsin-A, CK7, PDL1 EGFR, KRAS, ALK, ROS1
Melanoma HMB45, S100, SOX10, Melan-A BRAF, NRAS, TERT
Colon, adenocarcinoma CK20, CDX2, MMR proteins (MLH1, PMS2, APC, BRAF, RAS
MSH2, and MSH6)
Renal cell carcinoma, clear cell PAX8, CD10, CAIX VHL, BAP1, PBRM1
type
assessment, and exosomal assessment of brain (e.g., BRAF V600E) (also see Table 4.1).
tumors have the potential to reveal new clinically Interestingly, as the sampling of brain metastases
useful biomarkers and/or therapeutic targets. For becomes more clinically relevant for therapeutic
example, recent evidence indicates that patients planning (even when the basic diagnosis and pri-
with BRAF-mutated melanoma may benefit from mary organ of origin are already known) the use
inhibition of mitochondrial respiration in combi- of less invasive strategies to interrogate metasta-
nation with BRAF inhibitors; a hypothesis gener- ses such as cell-free DNA or CSF-based sequenc-
ated using a combination of RNA sequencing, ing techniques will likely become more prevalent.
metabolomic, and pharmacogenetic data [47, 48]. Looking ahead, the pathological evaluation of
Better understanding of the biology underlying brain metastases is rapidly changing, driven pri-
tumor derived exosomes has identified mecha- marily by the ongoing revolution in molecular
nisms of metastatic spread as well as potential use genetics, high throughout sequencing technolo-
of exosomes for diagnosis (e.g., in liquid biop- gies, and multiparametric “omic” assessment of
sies), novel therapeutic targets, and potentially tissue samples.
improved drug delivery [49, 50].
References
Conclusion
1. Osborn AG, Salzman KL, Jhaveri MD, Barkovich
The management of patients with brain metasta- J. Diagnostic imaging: brain. 3rd ed. Philadelphia:
ses inherently begins with diagnostic assessment. Elsevier; 2016.
2. Stark AM, Stohring C, Hedderich J, Held-Feindt J,
While clinicoradiological information is some- Mehdorn HM. Surgical treatment for brain metastases:
times sufficiently diagnostic (or the goals of care prognostic factors and survival in 309 patients with
do not necessitate a tissue diagnosis), tissue regard to patient age. J Clin Neurosci. 2011;18:34–8.
assessment is usually crucial not only to exclude https://doi.org/10.1016/j.jocn.2010.03.046.
3. Hwang TL, Close TP, Grego JM, Brannon WL,
other neoplastic etiologies, but to define the pri- Gonzales F. Predilection of brain metastasis in gray
mary site of origin and to further refine the and white matter junction and vascular border zones.
molecular assessment of the metastasis, even in Cancer. 1996;77:1551–5. https://doi.org/10.1002/
cases when the primary lesion has already been (SICI)1097-0142(19960415)77:8<1551::AID-
CNCR19>3.0.CO;2-Z.
extensively characterized. Increasingly, immuno- 4. Fink KR, Fink JR. Imaging of brain metasta-
histochemical and molecular testing algorithms ses. Surg Neurol Int. 2013;4:S209–19. https://doi.
are being employed to identify subtypes of can- org/10.4103/2152-7806.111298.
cer (e.g., ALK-rearranged lung adenocarcinoma 5. Shapira Y, Hadelsberg UP, Kanner AA, Ram Z,
Roth J. The ventricular system and choroid plexus
or HER2-enriched breast carcinoma) and charac- as a primary site for renal cell carcinoma metasta-
terize potential therapeutically relevant proteins sis. Acta Neurochir. 2014;156:1469–74. https://doi.
(e.g., PD-L1) or targetable molecular alterations org/10.1007/s00701-014-2108-7.
66 D. J. Pisapia
6. Mampre D, et al. Propensity for different vascular sensitive than traditional markers mammaglobin and
distributions and cerebral edema of intraparenchy- GCDFP15 for identifying breast cancer in surgical
mal brain metastases from different primary can- and cytology specimens of metastatic and matched
cers. J Neurooncol. 2019;143:115–22. https://doi. primary tumors. Appl Immunohistochem Mol
org/10.1007/s11060-019-03142-x. Morphol. 2016;24:229–37. https://doi.org/10.1097/
7. Cagney DN, et al. Incidence and prognosis of patients PAI.0000000000000186.
with brain metastases at diagnosis of systemic malig- 21. Miettinen M, et al. GATA3: a multispecific but poten-
nancy: a population-based study. Neuro Oncol. tially useful marker in surgical pathology: a system-
2017;19:1511–21. https://doi.org/10.1093/neuonc/ atic analysis of 2500 epithelial and nonepithelial
nox077. tumors. Am J Surg Pathol. 2014;38:13–22. https://doi.
8. Klotz S, et al. Clinical neuropathology image 6-2018: org/10.1097/PAS.0b013e3182a0218f.
metastasis of breast carcinoma to meningioma. Clin 22. Mete O, Kefeli M, Caliskan S, Asa SL. GATA3

Neuropathol. 2018;37:252–3. https://doi.org/10.5414/ immunoreactivity expands the transcription fac-
NP301150. tor profile of pituitary neuroendocrine tumors. Mod
9. Takei H, Powell SZ. Tumor-to-tumor Pathol. 2019;32:484–9. https://doi.org/10.1038/
metastasis to the central nervous system. s41379-018-0167-7.
Neuropathology. 2009;29:303–8. https://doi. 23.
Ordonez NG. Value of melanocytic-associated
org/10.1111/j.1440-1789.2008.00952.x. immunohistochemical markers in the diagnosis of
10. Hamlat A, et al. Malignant transformation of
malignant melanoma: a review and update. Hum
intra-cranial epithelial cysts: systematic article Pathol. 2014;45:191–205. https://doi.org/10.1016/j.
review. J Neurooncol. 2005;74:187–94. https://doi. humpath.2013.02.007.
org/10.1007/s11060-004-5175-4. 24. Tatsumori T, et al. p40 is the best marker for diag-
11. Freilich RJ, Thompson SJ, Walker RW, Rosenblum nosing pulmonary squamous cell carcinoma:
MK. Adenocarcinomatous transformation of comparison with p63, cytokeratin 5/6, desmo-
intracranial germ cell tumors. Am J Surg Pathol. collin-3, and sox2. Appl Immunohistochem Mol
1995;19:537–44. Morphol. 2014;22:377–82. https://doi.org/10.1097/
12. Weir HK, Johnson CJ, Thompson TD. The effect of PAI.0b013e3182980544.
multiple primary rules on population-based cancer 25. Nguyen T, et al. Comparison of 5 immunohistochemi-
survival. Cancer Causes Control. 2013;24:1231–42. cal markers of hepatocellular differentiation for the
https://doi.org/10.1007/s10552-013-0203-3. diagnosis of hepatocellular carcinoma. Arch Pathol
13. Gurel B, et al. NKX3.1 as a marker of prostatic Lab Med. 2015;139:1028–34. https://doi.org/10.5858/
origin in metastatic tumors. Am J Surg Pathol. arpa.2014-0479-OA.
2010;34:1097–105. https://doi.org/10.1097/ 26. Fanburg-Smith JC, Majidi M, Miettinen M. Keratin
PAS.0b013e3181e6cbf3. expression in schwannoma; a study of 115 retro-
14. Srodon M, Westra WH. Immunohistochemical stain- peritoneal and 22 peripheral schwannomas. Mod
ing for thyroid transcription factor-1: a helpful aid in Pathol. 2006;19:115–21. https://doi.org/10.1038/
discerning primary site of tumor origin in patients modpathol.3800489.
with brain metastases. Hum Pathol. 2002;33:642–5. 27. Pratt D, et al. Re-evaluating TTF-1 immunohisto-

15. Werling RW, Yaziji H, Bacchi CE, Gown AM. CDX2, chemistry in diffuse gliomas: expression is clone-
a highly sensitive and specific marker of adenocarci- dependent and associated with tumor location.
nomas of intestinal origin: an immunohistochemical Clin Neuropathol. 2017;36:263–71. https://doi.
survey of 476 primary and metastatic carcinomas. Am org/10.5414/NP301047.
J Surg Pathol. 2003;27:303–10. 28. Shibuya M. Welcoming the new WHO classifica-

16. Ordonez NG. Value of PAX 8 immunostaining in tion of pituitary tumors 2017: revolution in TTF-
tumor diagnosis: a review and update. Adv Anat 1-positive posterior pituitary tumors. Brain Tumor
Pathol. 2012;19:140–51. https://doi.org/10.1097/ Pathol. 2018;35:62–70. https://doi.org/10.1007/
PAP.0b013e318253465d. s10014-018-0311-6.
17. Ho IC, et al. Human GATA-3: a lineage-restricted tran- 29. Bielle F, et al. Chordoid gliomas of the third ventricle
scription factor that regulates the expression of the T share TTF-1 expression with organum vasculosum of
cell receptor alpha gene. EMBO J. 1991;10:1187–92. the lamina terminalis. Am J Surg Pathol. 2015;39:948–
18.
Hoch RV, Thompson DA, Baker RJ, Weigel 56. https://doi.org/10.1097/PAS.0000000000000421.
RJ. GATA-3 is expressed in association with estrogen 30. Cancer Genome Atlas Network. Comprehensive

receptor in breast cancer. Int J Cancer. 1999;84:122–8. molecular portraits of human breast tumours.
19.
Liu H, Shi J, Prichard JW, Gong Y, Lin Nature. 2012;490:61–70. https://doi.org/10.1038/
F. Immunohistochemical evaluation of GATA-3 nature11412.
expression in ER-negative breast carcinomas. 31. Lambein K, Van Bockstal M, Denys H, Libbrecht
Am J Clin Pathol. 2014;141:648–55. https://doi. L. 2013 update of the American Society of Clinical
org/10.1309/AJCP0Q9UQTEESLHN. Oncology/College of American Pathologists guide-
20. Sangoi AR, Shrestha B, Yang G, Mego O, Beck
line for human epidermal growth factor receptor 2
AH. The novel marker GATA3 is significantly more testing: impact on immunohistochemistry-negative
breast cancers. J Clin Oncol. 2014;32:1856–7. https:// analysis. Lancet Oncol. 2016;17:1386–95. https://doi.
doi.org/10.1200/JCO.2013.54.2530. org/10.1016/S1470-2045(16)30297-2.
32. Hammond ME, Hayes DF, Wolff AC, Mangu PB, 41. Orozco JIJ, et al. Epigenetic profiling for the molec-
Temin S. American society of clinical oncology/ ular classification of metastatic brain tumors. Nat
college of american pathologists guideline recom- Commun. 2018;9:4627. https://doi.org/10.1038/
mendations for immunohistochemical testing of s41467-018-06715-y.
estrogen and progesterone receptors in breast cancer. 42. Moss J, et al. Comprehensive human cell-type methyl-
J Oncol Pract. 2010;6:195–7. https://doi.org/10.1200/ ation atlas reveals origins of circulating cell-free DNA
JOP.777003. in health and disease. Nat Commun. 2018;9:5068.
33. Aggarwal R, et al. Clinical and genomic character- https://doi.org/10.1038/s41467-018-07466-6.
ization of treatment-emergent small-cell neuroendo- 43. Slieker RC, et al. DNA methylation landscapes

crine prostate cancer: a multi-institutional prospective of human fetal development. PLoS Genet.
study. J Clin Oncol. 2018;36:2492–503. https://doi. 2015;11:e1005583. https://doi.org/10.1371/journal.
org/10.1200/JCO.2017.77.6880. pgen.1005583.
34. Sepulveda AR, et al. Molecular biomarkers for the 44. Goodman AM, et al. Tumor mutational burden as an
evaluation of colorectal cancer. Am J Clin Pathol. independent predictor of response to immunotherapy in
2017. https://doi.org/10.1093/ajcp/aqw209. diverse cancers. Mol Cancer Ther. 2017;16:2598–608.
35. Burgess EF, et al. Discordance of high PD-L1 expres- https://doi.org/10.1158/1535-7163.MCT-17-0386.
sion in primary and metastatic urothelial carcinoma 45. Leibold AT, Monaco GN, Dey M. The role of the
lesions. Urol Oncol. 2019;37:299.e219–25. https:// immune system in brain metastasis. Curr Neurobiol.
doi.org/10.1016/j.urolonc.2019.01.002. 2019;10:33–48.
36. van de Nes J, et al. Targeted next generation sequenc- 46. Mansfield AS, et al. Contraction of T cell richness in
ing reveals unique mutation profile of primary lung cancer brain metastases. Sci Rep. 2018;8:2171.
melanocytic tumors of the central nervous sys- https://doi.org/10.1038/s41598-018-20622-8.
tem. J Neurooncol. 2016;127:435–44. https://doi. 47. Sundstrom T, et al. Inhibition of mitochondrial respi-
org/10.1007/s11060-015-2052-2. ration prevents BRAF-mutant melanoma brain metas-
37. Wang L, et al. Consistent copy number changes and tasis. Acta Neuropathol Commun. 2019;7:55. https://
recurrent PRKAR1A mutations distinguish Melanotic doi.org/10.1186/s40478-019-0712-8.
Schwannomas from Melanomas: SNP-array and next 48. Fischer GM, et al. Molecular profiling reveals unique
generation sequencing analysis. Genes Chromosomes immune and metabolic features of melanoma brain
Cancer. 2015;54:463–71. https://doi.org/10.1002/ metastases. Cancer Discov. 2019;9:628–45. https://
gcc.22254. doi.org/10.1158/2159-8290.CD-18-1489.
38. Brastianos PK, et al. Genomic characterization of brain 49. Wortzel I, Dror S, Kenific CM, Lyden D. Exosome-
metastases reveals branched evolution and potential mediated metastasis: communication from a distance.
therapeutic targets. Cancer Discov. 2015;5:1164–77. Dev Cell. 2019;49:347–60. https://doi.org/10.1016/j.
https://doi.org/10.1158/2159-8290.CD-15-0369. devcel.2019.04.011.
39. Capper D, et al. DNA methylation-based clas-
50. Gonçalo Rodrigues, Ayuko Hoshino, Candia M.
sification of central nervous system tumours. Kenific, Irina R. Matei, Loïc Steiner, Daniela Freitas,
Nature. 2018;555:469–74. https://doi.org/10.1038/ et al. Tumour exosomal CEMIP protein promotes can-
nature26000. cer cell colonization in brain metastasis. Nature Cell
40. Moran S, et al. Epigenetic profiling to classify can- Biology. 2019;21(11):1403–12.
cer of unknown primary: a multicentre, retrospective
Role of Precision Medicine
in Patients with CNS Metastasis
5
Albert Eusik Kim and Priscilla K. Brastianos
Introduction depending on the primary malignancy [1].

Treatment options are limited and involve a mul-
Brain metastases (BMs) are the most common tidisciplinary approach including surgical resec-
central nervous system (CNS) malignancy, and are tion, radiotherapy, and systemic treatment.
widely felt to represent a grim prognosis. Historically, patients with BM were treated with
Progression of intracranial disease is the cause of whole brain radiotherapy (WBRT); however,
death in up to 50% of patients with clinically sig- recent data in specific clinical scenarios where
nificant BM [1]. The reported incidence of BM is there are effective systemic treatment options
10–30% in all adults with cancer, and up to 40% of suggest that deferring WBRT may be reasonable
patients with metastatic cancer [1]. However, these due to a lack of overall survival benefit and the
estimates likely underestimate the true incidence associated neurotoxicity. At present, treatment
in the current era of modern cancer therapies. Over for BM is often case-specific and dependent on
the past decade, the incidence of BM has risen due many factors, such as performance status of the
to improved diagnostic testing that facilitates patient, as well as the number, location of, and
detection of asymptomatic BM and increased primary tumor type of BM [3]. Surgical resection
patient survival through better tolerated and more followed by radiotherapy is generally the stan-
effective treatment strategies [1]. Lung cancer dard of care for solitary or large (>3 cm) symp-
(39–56%), breast cancer (13–30%), and mela- tomatic lesions [1, 3]. Stereotactic radiosurgery
noma (6–11%) are among the most likely systemic (SRS) alone is frequently used for oligometa-
cancers to cross into the CNS [2]. Less common, static disease, which is commonly defined as up
but still reported, are gastrointestinal cancers to four BMs [1, 3]. Hippocampal-sparing WBRT,
(3–8%) and renal cell carcinoma (2–4%) [2]. which may have a lower risk of neurocognitive
Prognosis for BM is poor, with a median sur- side effects [1], can be considered in patients
vival ranging from 3 to 27 months after detection, with multiple disseminated BMs and leptomenin-
geal spread of disease.
A. E. Kim It is generally recommended that patients with
Department of Neurology, Medical Oncology, active extracranial disease receive systemic ther-
Massachusetts General Hospital, Boston, MA, USA apy after local brain therapy, as surgery and radia-
P. K. Brastianos (*) tion alone are not curative. Differential responses
CNS Metastasis Center, Department of Medicine, to these treatments for intracranial and e xtracranial
Massachusetts General Hospital, Harvard Medical
School, Boston, MA, USA disease are often observed, where systemic dis-
e-mail: PBRASTIANOS@mgh.harvard.edu ease is adequately controlled with progression of

https://doi.org/10.1007/978-3-030-42958-4_5
70 A. E. Kim and P. K. Brastianos
intracranial tumor burden [4]. The reasons behind mutations. Recent studies, however, have dem-
this differential response are multifactorial and onstrated significant genomic heterogeneity
not completely understood. One reason may be between BM and the paired primary tumor [5].
inadequate penetration of these systemic thera- In a study of 86 patients in which BM, primary
pies [4]. However, even with the use of new agents tumors, and normal tissue were analyzed by
with known intracranial efficacy, the majority of whole exome sequencing, 46 (53%) patients had
patients progress in the brain. This issue illus- distinct, potentially actionable mutations in the
trates an incomplete understanding of BM tumor BM not detected in the paired primary tumor
biology and the drivers that mediate blood-brain [5]. The vast majority of BMs, however, are
barrier (BBB) penetration and CNS proliferation. clonally related to the primary tumor, as only
This is due, in part, to a relative paucity of clinical 4/86 (4.6%) specimens were shown to be unre-
trials evaluating systemic therapies in BM, due lated to the primary lesion [5]. Similarly, distal
largely to the exclusion of patients with BM from extracranial and regional lymph node metasta-
clinical trials due to perceived poor prognosis. ses were also found to be clonally related to the
Another barrier is the lack of understanding of the primary tumor, but highly divergent from BM
genomic drivers behind development of BM and [5]. These findings suggest that branched evolu-
longitudinal changes in tumor genomics and tion, or the divergent propagation of multiple
physiology during treatment. Direct tissue analy- subclonal populations arising from a common
sis to understand these changes can be challeng- ancestor [6, 7], likely explains genomic differ-
ing due to the surgical risk associated with tissue ences between the primary tumor and different
sampling or inoperable location within the brain. metastases as well as the phenomenon of locore-
Noninvasive methods of genomic profiling of BM gional genomic heterogeneity. During branched
are currently under development and detailed in evolution, tumors will acquire hundreds, if not
this review. thousands, of genetic alterations, a minority of
In the current era of precision medicine, which is driver mutations that confer a selective
choice of treatment for many systemic cancers growth advantage to clones harboring the muta-
has become increasingly personalized and depen- tion [7]. These advantageous mutations allow
dent on the molecular or genomic characteriza- for the development and proliferation of sub-
tion of systemic cancer. To this end, improved clonal populations.
control of both intracranial and extracranial The exact genomic signatures required for
tumor burden has been observed with targeted CNS metastases and proliferation are still unclear.
therapy and immunotherapy. In this review, we Interestingly, spatially and temporally separated
present current efforts to characterize the genomic BM from the same patient possess a more homog-
drivers and heterogeneity of BM, as compared to enous genomic signature when compared to each
the primary tumor, using modern sequencing other as opposed to the primary tumor [5], sug-
techniques. A better understanding of these gesting that specific genomic alterations may be
genomic alterations will lead to more precise tai- integral for the brain metastatic process. To this
loring of current treatments and new therapeutic end, several studies have shown that upregulation
approaches. Additionally, we will present current of specific pathways such as phosphatidylinositol-
knowledge of targeted therapies for BM of sys- 4,5-bisphosphate 3-kinase (Pi3K) [8], epidermal
temic cancers of different histologies. growth factor receptor (EGFR) [8], or human
EGFR 2 (HER2) [9] is associated with cancer
cells crossing into the blood-brain barrier (BBB)
enetic Heterogeneity in Brain
G and proliferating within the CNS. Furthermore,
Metastases alterations in the cyclin-dependent kinase (CDK)
pathways, such as CDKN2A loss and CDK4/6
Selection of targeted therapy for BM has tradi- amplification, have also been implicated in CNS
tionally relied on genomic analysis of the initial metastases [5]. The exact role that these genomic
primary tumor resection to identify actionable alterations play in BM pathogenesis is not known
5 Role of Precision Medicine in Patients with CNS Metastasis 71
at this time, and remains an active area of ter understand the breadth of genomic heteroge-
research. For example, are these genetic altera- neity in BM and will result in further refinement
tions simply related to the underlying histology of current treatment strategies.
of the primary tumor, or is dysregulation of these
pathways necessary for CNS spread and prolif-
eration? In support of the latter, a recent study enomic Profiling of Brain
G
demonstrated loss of phosphatase and tensin Metastases
homolog (PTEN), a tumor suppressor gene,
expression in human tumor cells with normal The recent introduction of targeted therapies and
PTEN expression after dissemination to the brain checkpoint inhibitors has resulted in unprece-
but not to other organs [10]. Furthermore, the dented durable responses for many systemic can-
PTEN deficient level in BM tumor cells was cers, including those with a high propensity for
restored after leaving the brain microenviron- BMs, such as melanoma, non-small-cell lung
ment. This finding seems to indicate that certain cancer, and breast cancer. As such, cancer treat-
genomic changes are needed for CNS prolifera- ment has become increasingly personalized and
tion, a topic worthy of further prospective study dependent on the molecular and genomic traits of
for confirmation. each patient’s cancer. Similarly, identification of
Divergent evolution of BM has important these actionable mutations within BM holds
therapeutic implications. This genomic heteroge- great potential to drastically alter outcomes.
neity likely explains the divergent response seen Unfortunately, determining the exact genomic
in intracranial and extracranial disease burden in signature for BM can be unwieldy as this fre-
response to targeted therapies. In many cases, quently entails direct tissue analysis. As BM
actionable mutations for CNS metastases may often possesses targetable mutations not present
only be present in BM. As BMs are not always in the primary tumor or distal extracranial metas-
resected for diagnostic purposes due to the mor- tases [5], genomic analysis of these extracranial
bidity associated with tissue sampling, CNS ther- sites can miss these genomic alterations and thus
apeutic strategies are often made from analysis of targeted therapy opportunities for BM. This clini-
the primary tumor or extracranial metastasis. cal conundrum illustrates a critical need for non-
This assumption can result in sampling bias, invasive and clinically practical methods to
given frequent BM genomic divergence from capture intracranial molecular profiling. Such a
extracranial tissue samples. If available, action- biomarker would provide a better understanding
able targetable alterations for BM purposes of temporal evolution of BM, inform choice of
should be assessed from BM tissue analysis. It treatment, and aid in early identification of drug-
should be noted that whether specific systemic resistant mutations.
targeted therapies hold prophylactic or durable Molecular analysis of circulating tumor DNA
therapeutic efficacy for BM is unknown at this (ctDNA) in plasma is currently used for several
time. It is possible that reprogramming of the systemic cancers as a noninvasive tool for
cancer cell transcriptome by the CNS microenvi- genomic profiling and monitoring treatment
ronment may impact efficacy of systemic thera- response [11–13]. However, tumor DNA was
pies in BM. This question requires further study found to be either absent or only present in small
to fully answer. amounts in the plasma of patients with primary
As BM tissue analysis or serial brain biopsies brain tumors or solid tumor BM [12]. In such
are not always feasible, continued development cases, molecular analysis of ctDNA isolated from
of noninvasive techniques that shed light on cerebrospinal fluid (CSF) is emerging as a prom-
genomic and physiologic changes as a result of ising biomarker. The fraction of cell-free ctDNA
treatment are critical. Several such methods, such in the CSF is higher than in plasma due to the
as liquid biopsies, circulating tumor cells, or cell- relative absence of background normal DNA in
free deoxyribonucleic acid (DNA), are described CSF [13]. This allows for the detection of somatic
further below. Such techniques may help us bet- mutations in the CSF with moderate sequence
coverage, whereas plasma ctDNA sequencing [15]. Larger tumor size, lymphovascular space
requires very deep sequence coverage to achieve invasion, and hilar lymph node involvement are
similar sensitivities for detecting mutations associated with an increased risk of BMs [16].
occurring at low allele frequencies. Additionally, Unfortunately, despite an aggressive multimodal-
mutations present only in BM and not in the ity treatment approach combining platinum-
extracranial tumors were represented in CSF based chemotherapy, radiation, and surgery,
ctDNA [12]. Lastly, tumor DNA burden in CSF prognosis remains poor. The reported 1-year
ctDNA was observed to change during treatment mortality rate after developing BM ranges from
[12]. Mutant allelic frequency of CSF ctDNA 81% to 90% [14]. In addition, approximately
decreased with tumor response to treatments and 40–50% of patients with complete initial
increased with progression. While current meth- responses to therapy will develop BM [17]. Over
ods using CSF ctDNA for detection of all types the past decade, NSCLC management has been
of mutations still require optimization, the above revolutionized by the identification of oncogenic
data suggests that CSF ctDNA may soon develop driver mutations in anaplastic lymphoma kinase
into a clinical tool for BM genomic analysis. (ALK) and epidermal growth factor receptor
Additional biomarkers that reflect the BM (EGFR) and the development of targeted thera-
genomic signature are currently under develop- pies, resulting in unprecedented response rates.
ment. One such example is an exosome, an extra-
cellular vesicle released from the cell upon fusion
of an intermediate endocytic compartment with SCLC: EGFR Tyrosine Kinase
N
the plasma membrane. These vesicles are felt to Inhibitor
be a conduit for intercellular communication and
may contain genomic data consistent with a Activating mutations in EGFR are generally
tumor’s molecular properties. The burgeoning found in NSCLC patients with the following
field of radiogenomics, or the relationship characteristics: female gender, age <35 years,
between an imaging-derived phenotype and Asian descent (in about 40%), history of never or
genomic data, may also be a promising way to light-smoking and adenomatous histology [18].
noninvasively monitor for genomic alterations. In such patients, EGFR mutation testing is rec-
Using these correlations with serial imaging may ommended. EGFR mutations render these tumors
shed light on alterations in tumor biology as a sensitive to EGFR tyrosine kinase inhibitors
result of treatment. If optimized, radiogenomics (TKIs), which results in significantly improved
may assist in the early detection of drug-resistant outcomes when compared to platinum-based
mutations and thus inform a change to a more combination chemotherapy [19]. For patients
efficacious treatment regimen. Both fields are without a non-squamous histology EGFR muta-
largely in their infancy, and currently associated tion testing is not recommended due to extremely
with significant limitations. low likelihood of positivity, unless they are non-
smokers [18].
First- and second-generation EGFR TKIs
Non-Small-Cell Lung Cancer selectively target the EGFR receptor through
competitive, reversible binding at the tyrosine
Non-small-cell lung cancer (NSCLC) is the lead- kinase domain, and are currently first-line ther-
ing cause of cancer mortality worldwide, apy for EGFR-mutant NSCLC [19, 20]. Erlotinib
accounting for 18.2% of total deaths from cancer and gefitinib are among the most commonly used
[14]. Furthermore, NSCLC, adenocarcinoma in EGFR first-generation TKIs. However, the major-
particular, is the most common primary malig- ity of patients with initial response to EGFR
nancy to metastasize to the brain [3]. TKIs had disease progression due to an acquired
Approximately 25–30% of NSCLC patients will resistance within 1–2 years [21]. The develop-
develop BM during the course of their disease ment of an additional EGFR mutation, most com-
monly the threonine-to-methionine substitution SCLC: Anaplastic Lymphoma Kinase

N
at position 790 on exon 20 (T790M), is respon- (ALK) Tyrosine Kinase Inhibitors
sible for approximately 60% of this acquired
resistance [22]. Third generation TKIs, such as The discovery of the ALK gene rearrangement
osimertinib and rociletinib, have shown promis- and development of genetically driven therapies
ing activity for these resistant EGFR-mutant targeting this aberration have led to tremendous
types [23]. progress in treating NSCLC. The most common
Presently, data on the efficacy of EGFR TKIs rearrangement arises from a fusion between ALK
in treating NSCLC BMs is hopeful, but limited. and the echinoderm microtubule-like protein 4
Barriers to an accurate evaluation are the lack of (EML4) gene. This results in an oncogenic tyro-
clinical trials studying targeted therapies in BM, sine kinase with constitutive activity, and is found
and regional genomic heterogeneity—as an in up to 5% of NSCLC [29]. BM is a relatively
EGFR-mutant status in the primary tumor is not common occurrence in ALK-rearranged NSCLC,
always present in BM. Nonetheless, available with incidence quoted at 23.8% at time of diag-
data suggests that these agents likely have some nosis and 58.4% at 3 years [30]. As with EGFR,
CNS activity. Recent preclinical data demon- ALK translocations are associated with younger
strates intracranial activity of afatinib, a second- age, history of light or no smoking, and adeno-
generation EGFR TKI and an irreversible ErbB carcinoma histology [31]. Consequently, testing
family inhibitor [24]. Post-hoc subgroup analysis for ALK is highly recommended for such patients
from the LUX-Lung 3 and LUX-Lung 6 studies, [2], as the presence of an ALK-mutation is corre-
which allowed patients with asymptomatic BM lated with response to ALK TKIs.
to be enrolled, showed survival benefit from Crizotinib, a first-generation ALK TKI that
treatment with afatinib compared to platinum- also has activity against MET and ROS1 [31], is
based chemotherapy. Progression free survival superior to standard-of-care chemotherapy for
(PFS) (8.2 vs. 5.4 months) and objective response management of systemic ALK-rearranged
rate (ORR) (70–75% vs. 20–28%) were signifi- NSCLC [32]. While assessing ALK TKIs for
cantly better with afatinib than platinum-based CNS efficacy is limited due to exclusion of BMs
chemotherapy [25]. Another small phase II pro- from many randomized clinical trials, crizotinib
spective trial exploring EGFR TKIs in BM likely holds some CNS efficacy. In the PROFILE
reported an 83% ORR with first-generation TKIs 1005 and 1007 studies, patients with untreated
[26]; however, other studies have reported more asymptomatic BMs were included in a pooled
modest responses [27]. For acquired resistance, a retrospective analysis. For these patients, intra-
recent study demonstrated superior BBB penetra- cranial disease control rate was noted to be 56%
tion with osimertinib than with gefitinib or afa- at 12 weeks, with a median time to CNS
tinib, as well as sustained BM regression in an progression of 7 months [31]. In PROFILE 1014,
EGFR-mutant mouse model [28]. a randomized phase III trial of crizotinib versus
Taken together, EGFR TKIs, especially platinum-based chemotherapy, patients with sta-
osimertinib, appear to have positive CNS activity. ble treated BMs were allowed to enroll with CNS
How to apply these findings in the context of sur- efficacy as a secondary endpoint. In this cohort,
gical resection and radiotherapy still remains CNS disease control rate for patients with BM
unclear. It seems reasonable to incorporate EGFR was significantly higher with crizotinib at
TKIs up front in asymptomatic BMs and to con- 12 weeks (85% vs. 45%) and median PFS was
sider delaying surgery or radiation until BM pro- significantly longer (9 vs. 4 months) [33].
gression to minimize adverse effects. Further Second-generation ALK TKIs are promising
prospective trials evaluating EGFR TKIs and options for ALK-rearranged NSCLC patients
sequential approaches with brain radiotherapy to who develop resistance to crizotinib, and are also
optimize CNS efficacy and minimize radiation- felt to have improved CNS efficacy. Of these
induced neurotoxicity are needed. agents, alectinib and ceritinib are among those
with the strongest evidence for BM. Preliminary with greater than 50% PD-L1 expression, sug-
findings from the J-ALEX study, a Japanese gesting that PD-L1 expression may be a predic-
phase III trial that recruited ALK-inhibitor naïve tive biomarker for response [41].
patients with ALK-rearranged NSCLC, reported Many immunotherapy trials for NSCLC, to
that the alectinib cohort had yet to reach median date, have excluded patients with active brain
PFS, while the crizotinib cohort’s median PFS metastases. However, a recent early analysis of a
was 10.2 months [34]. Two other phase II studies phase II trial investigating activity and safety of
with alectinib demonstrated CNS response rates pembrolizumab in NSCLC and melanoma
up to 75% and median CNS disease response patients with untreated or progressive BMs
durations of 10–11 months [35, 36]. In the showed encouraging results. Patients with
ASCEND-1 study, 94 patients with ALK- NSCLC had tumor tissue positive for PD-L1
rearranged NSCLC BM were retrospectively expression. In this study, 33% (6 of 18) of
analyzed. Of this cohort, 79% of ALK TKI-naïve NSCLC patients had durable intracranial
and 65% of ALK TKI-pretreated patients had response without high-grade adverse events [42].
intracranial response to ceritinib [37]. Newer Further randomized prospective studies are
ALK TKIs such as lorlatinib and brigatinib likely needed to investigate these promising options for
have even better brain efficacy. As with first- brain metastases.
generation ALK TKIs, further work is needed to
determine utility of these treatments in combina-
tion with radiotherapy with the intent of maxi- Breast Cancer
mizing CNS efficacy.
Breast cancer is the most common cancer in
women and the second-leading cause of cancer-
NSCLC: Immunotherapy related death in women [3]. It is also the second
most common cancer to metastasize to the brain,
Immune checkpoint inhibitors have emerged as after NSCLC [1]. The exact incidence of BM
an option for patients with advanced NSCLC from breast cancer in the current era of modern
without an actionable driver mutation (i.e., EGFR therapies is not clearly defined; however, it is
and ALK), or for those with actionable mutations estimated that between 10% and 45% of breast
that have progressed on next-generation targeted cancer patients will be affected by BM during
agents [38]. Immune checkpoints, which refer to their disease course, depending on breast tumor
inhibitory pathways that modulate the physio- subtype [43]. This number will likely increase as
logic immune response to minimize collateral overall survival improves with newer, more dura-
damage and thus maintain self-tolerance, are co- ble, therapies.
opted by tumors. For example, the interaction of As expected, prognosis for BM in breast can-
programmed death 1 (PD-1) receptor on activated cer remains poor. A large retrospective study
T cells with programmed death ligand 1 (PD-L1) identified older age, Karnofsky Performance
on tumor cells leads to T-cell inactivation, which Status (KPS), and tumor subtype as prognostic
prevents the immune system from attacking the factors [44]. Within breast cancer, there are four
tumor cell [38]. Nivolumab and pembrolizumab main tumor subtypes. Basal subtype [estrogen
are anti-PD1 monoclonal antibodies that have receptor (ER), progesterone receptor (PR), and
been shown to improve survival outcomes in HER2 negative; also referred to as “triple nega-
patients with metastatic NSCLC without action- tive”] has the worst prognosis, with a median OS
able mutations, as compared to docetaxel-based of 5 months after developing BM [44]. Luminal
chemotherapy [39, 40]. Furthermore, pembroli- A (ER- and/or PR-positive, HER2-negative, low
zumab demonstrated PFS and overall survival levels of Ki-67) are generally low-grade tumors
(OS) superiority to platinum-based chemother- with the best prognosis [44]. Other subtypes
apy as first-line therapy in patients with NSCLC include luminal B (ER- and/or PR-positive, and
either HER2-positive or HER2-negative with is associated with an increased risk of BM, as

high levels of Ki-67) and HER2-enriched (ER/ approximately 30–50% of patients with HER2-
PR-negative and HER2-positive). Patients with positive breast cancer will develop BM during
triple-negative and HER2-enriched breast cancer their disease course [48]. The propensity of
are at highest risk of CNS metastases [44]. HER2-positive breast cancer for CNS relapse
Current management of BM from breast cancer may be related to improved survival of patients
is similar to those of other primary cancers, and with HER2- directed therapy, the limited CNS
includes consideration of systemic therapies in penetration of HER2-directed agents, and perhaps
addition to surgical resection and radiation. the neurotropism of HER2-positive breast cancer
In this section, we describe current targeted [48]. As with other types of primary tumors, tem-
therapies for breast cancer. Triple negative breast poral and spatial genomic heterogeneity are seen
cancer (TNBC) is especially challenging to treat with breast cancer BM. A retrospective study
due to lack of clinically actionable genomic alter- showed that 24% of 182 patients with HER2-
ations and nondurable response to systemic che- positive primary breast cancer had HER2-negative
motherapy [45]. For this cohort, there has been a metastatic disease [49]. There is also evidence to
growing pool of novel targets as gene sequencing suggest that BM commonly occurs in patients
has become more readily accessible. One promis- with HER2-positive breast cancer that is other-
ing target for TNBC is poly ADP-ribose poly- wise systemically well controlled with HER2-
merase (PARP), a family of proteins involved in directed therapy [48]. As with other types of
DNA repair and genomic stability. Histologic systemic cancers, these findings illustrate the
studies have shown similarities between the path- necessity of repeat genomic analysis on BM tis-
ological and clinical features of TNBC- and sue if clinically feasible.
BRCA-associated cancers [45]. Interestingly, Like most other monoclonal antibodies,
BRCA-1 and BRCA-2 mutant cell lines have trastuzumab, which targets the HER2 receptor,
been shown to be exquisitely sensitive to PARP has limited CNS activity due to its inability to
inhibition [46]. Several PARP inhibitors (i.e., cross the intact BBB [48]. Consequently, adju-
olaparib and veliparib) are currently being evalu- vant radiation with trastuzumab, pertuzumab,
ated in the adjuvant, neoadjuvant, and metastatic and T-DM1 are all being investigated as options
setting for the subset of TNBC with BRCA-1 or for HER2-positive BM. A recent pharmacoki-
BRCA-2 mutations. netic study demonstrated improved CNS pene-
tration of trastuzumab after BBB disruption by
radiation. The ratio of the CSF to plasma levels
reast Cancer: HER2 Antibodies
B of trastuzumab improved significantly from
and TKIs 1:420 before radiotherapy to 1:76 after radio-
therapy [50]. Pertuzumab, another monoclonal
HER2 is a member of the human epidermal antibody against the HER2 receptor, likely has
growth factor receptor family, which consists of some synergistic CNS antitumor efficacy in
four membrane-bound receptor tyrosine kinase combination with trastuzumab and docetaxel, as
implicated in multiple signaling cascades that shown in the CLEOPATRA trial, a randomized
mediate cell proliferation and apoptosis. This pro- phase III placebo-controlled trial of pertuzumab
tein is overexpressed in 20% of all breast cancer in metastatic HER2-positive breast cancer. The
patients [47]. HER2-directed therapies, such as median time to development of BMs as first site
trastuzumab, lapatinib, pertuzumab, and T-DM1 of disease progression was significantly longer
(ado-trastuzumab emtansine, an antibody- drug in the pertuzumab arm compared to the placebo
conjugate consisting of trastuzumab linked to the arm (15.0 vs. 11.9 months), and the median OS
cytotoxic agent DM1), significantly improve PFS was 56.5 months in the pertuzumab arm, com-
and OS of patients with HER2-positive metastatic pared to 40.8 months in the placebo arm [51].
breast cancer. Furthermore, HER2-overexpression Other small case series have also demonstrated
some efficacy for pertuzumab-containing regi- OraL EveROlimus-3 (BOLERO-3) trial showed
mens in BM from HER2-positive breast cancer that triple therapy with everolimus, trastuzumab,
BM [52, 53]. Finally, several retrospective stud- and vinorelbine was superior to placebo, trastu-
ies indicate some potential activity for the anti- zumab, and vinorelbine in trastuzumab-resistant
body-cytotoxin conjugate T-DM1 in CNS advanced HER2+ breast cancer [61]. Another
disease [54], but clear prospective evidence is large phase III trial showed that everolimus com-
lacking. bined with an aromatase inhibitor improved PFS
Lapatinib is a dual small-molecule HER2 and in heavily pretreated hormone receptor-positive
EGFR TKI that has shown some ability to cross advanced breast cancer [62]. While these trials
a disrupted BBB. A novel PET imaging study excluded brain metastases, these results may per-
using radiolabeled lapatinib demonstrated haps be generalized to BM as everolimus has
increased levels of lapatinib in brain metastases been demonstrated to possess CNS penetration in
as compared to normal brain tissue [55]. patients with primary brain tumors [63]. Clinical
Lapatinib has demonstrated partial response of trials evaluating the role of everolimus and other
CNS disease to a modest degree as adjuvant therapies targeting the Pi3K and mTOR signaling
monotherapy (CNS ORR 6% [56]) and in com- pathways in management of breast cancer BM
bination with capecitabine (CNS ORR 20–38% are ongoing.
in pretreated patients [57, 58]). This CNS antitu- Alterations in the CDK pathway are common
mor efficacy is augmented in treatment-naïve in breast cancer brain metastases [5]. Activation
patients with HER2-positive breast cancer (CNS of CDK4 and CDK6 by cyclin D results in cell
ORR 65% [59]). Neratinib, an irreversible proliferation by facilitating G1 phase progres-
HER1, HER2, and HER4 TKI, also may have sion and transition from G1 to S phase in the
CNS efficacy in HER2-positive metastatic dis- cell cycle [48]. CDK inhibitors, such as riboci-
ease. The NEfERTT trial, a randomized phase III clib, palbociclib, and abemaciclib, have demon-
trial of patients with metastatic HER2-positive strated success in hormone-receptor positive
breast cancer, noted significantly lower rates of breast cancer [64]. Recent preclinical studies
CNS progression and delayed time to CNS have shown good CNS penetration of abemaci-
metastases with the neratinib-paclitaxel combi- clib, and some efficacy for breast cancer BM as
nation than with trastuzumab-paclitaxel, demonstrated by several case series [65].
although the two groups had similar OS [60]. Current trials are further investigating the effi-
Further studies evaluating these regimens are cacy of these agents.
ongoing.
Melanoma
Breast Cancer: Additional Mutations
Melanoma is the third most common systemic
Sequencing studies of BM from breast cancer cancer to metastasize to the brain [3].
demonstrated that actionable mutations in the Approximately 50% of patients with stage IV
phosphoinositide 3-kinase/protein kinase B/ melanoma will develop BM during the course of
rapamycin (Pi3K/AKT/mTOR) pathways are their disease [1, 3]. As with other systemic malig-
common [5]. This pathway regulates several cel- nancies, prognosis of BM in metastatic mela-
lular functions in cancer, most notably cell noma is poor due to significant neurologic
growth and proliferation. Increased activation of morbidity. Median OS after the diagnosis of BM
this pathway is one hypothesized mechanism of has historically been about 4.7 months, although
resistance to hormonal therapy. Everolimus, an a recent retrospective analysis reported improve-
mTOR inhibitor, is currently being studied for ment of median OS to 7.7 months with the recent
breast cancer BM. The breast cancer trials of use of targeted therapies [66].
Melanoma: Mitogen-Activated was 7.5 months [69]. Interestingly, genomic

Protein Kinase (MAPK) Pathway sequencing analysis of BRAF-inhibitor resistant
BM revealed genomic alterations resulting in
Approximately 50% of patients with metastatic activation of the Pi3K/AKT pathway [70].
melanoma will have an activating mutation in Mitogen-activated protein (MEK) kinase is
BRAF, a serine/threonine protein kinase within downstream of BRAF in the MAPK pathway, and
the MAPK signaling pathway [67]. BRAF is a is frequently activated by members of the Pi3K
key regulator of cell growth, division, and differ- pathway as a resistance mechanism from BRAF
entiation, and when inactive can result in down- inhibition. To prevent resistance, BRAF inhibi-
stream constitutive activation of the MAPK tors are frequently combined with MEK inhibi-
pathway. This provides a basis for the mutational tors, such as trametinib and cobimetinib, in
activation and uncontrolled tumor growth for metastatic melanoma. When BRAF inhibitors
multiple cancers, and thus a potential target for were combined with MEK inhibitors, treatment
selective inhibition. efficacy was further potentiated in patients with
In melanoma, the most common BRAF muta- BRAF mutant extracranial metastatic melanoma,
tion is the substitution of valine for glutamic acid as evidenced by improved PFS (2 years) and OS
(V600E), comprising nearly 90% of all BRAF (3 years) [71–73]. Dual BRAF and MEK inhibi-
mutations in melanoma [67]. The second most tions for brain metastases are currently being
common BRAF alteration is the valine for lysine evaluated in clinical trials.
substitution (V600K), which represents 5–6% of
cases. BRAF-mutant melanomas are generally
more aggressive and may confer a higher risk of Melanoma: Pi3K/AKT/mTOR Pathway
developing BM [67]. There are currently two
FDA-approved BRAF inhibitors for systemic Genomic analysis of 16 pairs of patient-matched
melanoma: vemurafenib and dabrafenib [67]. melanoma brain metastases and extracranial
BRAF inhibitors have markedly improved OS for metastases demonstrated increased activation of
patients with BRAF-mutant metastatic mela- the Pi3K/AKT/mTOR pathway specific to BM
noma. This response, however, is not usually [8]. Preclinical and animal studies using a Pi3K
durable [66]. As with other systemic tumors, cur- inhibitor, BKM120, demonstrate growth inhibi-
rent BM genomic sequencing studies indicate tion rates of up to 80% and induced apoptosis
that the development of treatment-resistant in vitro and inhibition of tumor growth of human
genomic alterations contributes to treatment brain metastatic melanoma cells within brains of
failure. nude mice [74]. These findings suggest that an
Evidence for dabrafenib and vemurafenib in alteration in the Pi3K pathway, for reasons
BM efficacy is limited, as many large phase III unknown at present, may make a tumor more at
trials excluded CNS disease. Nonetheless, these risk for CNS spread and proliferation.
agents likely hold some CNS efficacy. The Furthermore, Pi3K inhibitors may be a potential
BREAK-MB trial, a multicenter phase II trial therapeutic option worthy of prospective clinical
with 172 patients with BRAF-mutant melanoma trials for metastatic melanoma.
with at least one asymptomatic brain metastasis,
showed that dabrafenib had activity for patients
with either untreated or pretreated BM. For both Melanoma: Immunotherapy
groups, there was a response rate of >30% with
improvement in OS and PFS [68]. In a retrospec- Unprecedented treatment advances for patients
tive study of 27 patients, vemurafenib resulted in with advanced-stage melanoma have occurred
an intracranial response rate of 71%. The median recently with the advent of immunotherapy.
intracranial PFS was 4.6 months and median OS High-dose interleukin-2 had early success [75],
but was frequently associated with severe toxici- 85% local BM control and a median OS of
ties and was consequently limited only to patients 11.8 months with nivolumab and SRS to BM
with excellent performance status. Ipilimumab, [84]. Two recent phase II studies, specifically tai-
an anti-cytotoxic T-lymphocyte-associated pro- lored for patients with melanoma BM, provide
tein 4 (CTLA-4) antibody, received FDA approval even stronger evidence of checkpoint inhibitor
in 2011 after a landmark study in 2010 demon- efficacy. One study tested ipilimumab and
strated improved patient outcomes in unresect- nivolumab in 74 patients with at least one mea-
able stage III or IV melanoma [76]. Systemic surable, nonirradiated, asymptomatic BM. Here,
response rates for ipilimumab have ranged from the rate of intracranial clinical benefit (57%) was
10% to 15%, with improved response in those concordant to that of extracranial benefit (56%)
with BRAF-wild type melanoma [77, 78]. About with a 20% complete response rate and 30% par-
20% of patients with response to ipilimumab tial response rate intracranially [85]. Another
were long-term survivors, measured on the order study with a similar cohort found that combina-
of years [79, 80]. Soon afterward, two anti-PD-1 tion ipilimumab and nivolumab had an intracra-
antibodies, nivolumab and pembrolizumab, were nial response rate of 46% (16 of 35) and
approved by the FDA for metastatic melanoma. single-agent nivolumab resulted in an intracra-
Subsequent clinical testing with PD-1 checkpoint nial response rate of 20% (5 of 25) [86]. Similar
blockade demonstrated improved outcomes with to prior trials, the combination of ipilimumab and
less toxicity as compared to ipilimumab [81]. nivolumab was associated with more high-grade
Nivolumab, with a PFS of 6.9 months, was more adverse events (54% vs. 16% for nivolumab
effective than ipilimumab monotherapy, which monotherapy) [86].
displayed a median PFS of 2.9 months [81]. Despite these promising results, predictive
Additionally, pembrolizumab or nivolumab biomarkers of response are desperately needed
monotherapy were associated with ORR ranging for more precise tailoring of existing therapies,
from 33% to 57%, with the majority of responses especially given the high risk of adverse events.
being durable [77, 82]. In a recent phase III trial Genomic sequencing of melanoma BMs are
of patients with advanced melanoma without being analyzed with the hope of identifying
BM, the combination of nivolumab and ipilim- mutational profiles associated with better
umab achieved a median PFS of 11.5 months, prognoses.
superior to either monotherapy, but was also
associated with more high-grade toxicity (59%
for combination ipilimumab/nivolumab vs. 21% Conclusion
with nivolumab) [77].
More data are emerging that checkpoint inhib- Brain metastases represent an understudied and
itors likely possess some efficacy within the underserved area within oncology. This entity is
CNS. In a phase II study of ipilimumab in 72 associated with poor prognosis, due to significant
melanoma patients with BM, the disease control neurologic morbidity and current lack of durable
rate was 24% in patients who were neurologi- CNS-directed therapies. Consequently, better
cally asymptomatic and not on corticosteroids. treatments for brain metastases are critically nec-
One- and two-year survival rates were 31% and essary, as incidence is rising as therapies for sys-
26% in this cohort [80]. Furthermore, there is temic cancer improve. One major reason for
increasing data that suggests improved OS when current treatment difficulties is the paucity of
SRS is used with checkpoint inhibitors. One ret- clinical trials evaluating systemic treatments for
rospective analysis found that the 2-year survival brain metastases, due largely to exclusion of
rate of those receiving SRS plus ipilimumab was patients with CNS disease. Recently, next-
47.2%, compared with 19.7% in those who generation targeted agents and immunotherapies
received SRS alone [83]. Another retrospective have demonstrated improved tolerability and
study of 26 patients with melanoma BM noted an promising response rates for CNS disease.
Current trials evaluating these therapeutic strate- barrier permeability determines drug efficacy in
experimental brain metastases of breast cancer. Clin
gies specifically for brain metastases are under- Cancer Res. 2010;16(23):5664–78.
way and desperately needed to optimize 5. Brastianos PK, Carter SL, Santagata S, Cahill DP,
treatment. Taylor-Wener A, Jones RT, et al. Genomic character-
Another breakthrough for brain metastases ization of brain metastases reveals branched evolu-
tion and potential therapeutic targets. Cancer Discov.
has been the recognition of spatial and temporal 2015;5(11):1164–77.
genomic heterogeneity across different meta- 6. Navin N, Kendall J, Troge J, Andrews P, Rodgers L,
static sites. Recent genomic analyses have dem- McIndoo J, et al. Tumor evolution inferred by single-
onstrated the presence of actionable driver cell sequencing. Nature. 2011;472:90–4.
7. Gerlinger M, Rowan AJ, Horswell S, Larkin J,
mutations within brain metastases not present in Endesfelder D, Gronroos E, et al. Intratumor hetero-
the paired primary tumor. This genomic heterogeneity and branched evolution revealed by multire-
geneity likely contributes to the clinically gion sequencing. N Engl J Med. 2012;366:883–92.
observed divergent response seen between intra- 8. Chen G, Chakravarti N, Aardalen K, Lazar AJ, Tetzlaff
MT, Wubbenhorst B, et al. Molecular profiling of
cranial and extracranial disease burden. As brain patient-matched brain and extracranial melanoma
metastasis tissue analysis is not always feasible, metastases implicates the PI3K pathway as a thera-
noninvasive methods to obtain genomic informa- peutic target. Clin Cancer Res. 2014;20(21):5537–46.
tion are necessary to guide personalized genomic- 9. Bos PD, Zhang XH, Nadal C, Shu W, Gomis
RR, Nguyen DX, et al. Genes that mediate
directed therapy for brain metastases. Novel breast cancer metastasis to the brain. Nature.
approaches such as cell-free circulating tumor 2009;459(7249):1005–9.
DNA in the CSF and radiogenomics are under 10. Zhang L, Zhang S, Yao J, Lowery FJ, Zhang Q, Huang
development and promising. These methods, if W, et al. Microenvironment-induced PTEN loss by
exosomal microRNA primes brain metastasis out-
optimized for clinical use, may be repeated dur- growth. Nature. 2015;527:100–4.
ing a treatment course to help determine response 11. Dawson SJ, Tsui DW, Murtaza M, Biggs H, Rueda
and to assist in the early detection of drug- OM, Chin SF, et al. Analysis of circulating tumor
resistant mutations. Such biomarkers would be a DNA to monitor metastatic breast cancer. N Engl J
Med. 2013;368(13):1199–209.
critical step forward in better understanding the 12. De Mattos-Arruda L, Mayor R, Ng CK, Weigelt B,
temporal evolution of brain metastases and Martinez-Ricarte F, Torrejon D, et al. Cerebrospinal
informing choice of treatment. fluid-derived circulating tumor DNA better represents
the genomic alterations of brain tumors than plasma.
Nat Commun. 2015;6:8839.
Author Disclosures AE Kim has nothing to disclose.
13. Murtaza M, Dawson SJ, Tsui DW, Gale D, Forshew T,
PK Brastianos has consulted for Genentech-Roche,
Piskorz AM, et al. Non-invasive analysis of acquired
Lilly, Angiochem, and Tesaro, has received honoraria
resistance to cancer therapy by sequencing of plasma
from Genentech-Roche and Merck, and research funding
DNA. Nature. 2013;497(7447):108–12.
and/or clinical trial support (to MGH) from Pfizer and
14. Jemal A, Bray F, Center MM, Ferlay J, Ward E,

Merck.
Forman D. Global cancer statistics. CA Cancer J Clin.
2011;61(2):69–90.
15. Mamon HJ, Yeap BY, Janne PA, Reblando J, Shrager
S, Jaklitsch MT. High risk of brain metastases in sur-
References gically staged IIIA non-small-cell lung cancer patients
treated with surgery, chemotherapy, and radiation. J
1. Brastianos PK, Curry WT, Oh KS. Clinical discussion Clin Oncol. 2005;23(7):1530–7.
and review of the management of brain metastases. J 16. Hubbs JL, Boyd JA, Hollis D, Chino JP, Saynak M,
Natl Compr Canc Netw. 2013;11:1153–64. Kelsey CR. Factors associated with the develop-
2. Berghoff AS, Bartsch R, Wohrer A, Streubel B, ment of brain metastases: analysis of 975 patients
Birner P, Kros JM, et al. Predictive molecular mark- with early stage nonsmall cell lung cancer. Cancer.
ers in metastases to the central nervous system: recent 2010;116(21):5038–46.
advances and future avenues. Acta Neuropathol. 17. Chen AM, Jahan TM, Jablons DM, Garcia J, Larson
2014;128(6):879–91. DA. Risk of cerebral metastases and neurological
3. Nayak L, Lee EQ, Wen PY. Epidemiology of brain death after pathological complete response to neoad-
metastases. Curr Oncol Rep. 2012;14(1):48–54. juvant therapy for locally advanced nonsmall-cell lung
4. Lockman PR, Mittapalli RK, Taskar KS, Rudraraju V, cancer: clinical implications for the subsequent man-
Gril B, Bohn KA, et al. Heterogenous blood-tumor agement of the brain. Cancer. 2007;109(8):1668–75.
18. Sholl LM, Yeap BY, Iafrate AJ, Holmes-Tisch AJ, 30. Rangachari D, Yamaguchi N, VanderLaan PA, Folch
Chou YP, Wu MT, et al. Lung adenocarcinoma with E, Mahadevan A, Floyd SR, et al. Brain metas-
EGFR amplification has distinct clinicopathologic tases in patients with EGFR-mutated or ALK-
and molecular features in never-smokers. Cancer Res. rearranged non-small cell lung cancers. Lung Cancer.
2009;69(21):8341–8. 2015;88(1):108–11.
19. Mok TS, Wu Y-L, Thongprasert S, Yang C-H, Chu 31. Kwak EL, Bang YJ, Camidge DR, Shaw AT, Solomon
D-T, Saijo N, et al. Gefitinib or carboplatin-paclitaxel B, Maki RG, et al. Anaplastic lymphoma kinase inhi-
in pulmonary adenocarcinoma. N Engl J Med. bition in non-small-cell lung cancer. N Engl J Med.
2009;361:947–57. 2010;363(18):1693–703.
20. Sequist LV, Yang JC, Yamamoto N, O’Byrne K, Hirsh 32. Shaw AT, Kim DW, Nakagawa K, Seto T, Crino L,
V, Mok T, et al. Phase III study of afatinib or cispla- Ahn MK, et al. Crizotinib versus chemotherapy in
tin plus pemetrexed in patients with metastatic lung advanced ALK-positive lung cancer. N Engl J Med.
adenocarcinoma with EGFR mutations. J Clin Oncol. 2013;368(25):2385–94.
2013;31(27):3327–34. 33. Costa DB, Shaw AT, Ou SH, Solomon BJ, Riely GJ,
21.
Sequist LV, Waltman BA, Dias-Santagata D, Ahn MJ, et al. Clinical experience with crizotinib in
Digumarthy S, Turke AB, Fidias P, et al. Genotypic patients with advanced ALK-rearranged non-small-
and histological evolution of lung cancers acquir- cell lung cancer and brain metastases. J Clin Oncol.
ing resistance to EGFR inhibitors. Sci Transl Med. 2015;33(17):1881–8.
2011;3(75):75ra26. 34. Hida T, Nokihara H, Kondo M, Kim YH, Azuma K,
22. Yu HA, Arcila ME, Rekhtman N, Sima CS, Zakowski Seto T, et al. Alectinib versus crizotinib in patients
MF, Pao W, et al. Analysis of tumor specimens at the with ALK-positive non-small-cell lung cancer
time of acquired resistance to EGFR-TKI therapy in (J-ALEX): an open-label, randomised phase 3 trial.
155 patients with EGFR-mutant lung cancers. Clin Lancet. 2017;390(10089):29–39.
Cancer Res. 2013;19(8):2240–7. 35. Shaw AT, Gandhi L, Gadgeel S, Riely GJ, Cetnar J,
23. Janne PA, Yang JC, Kim DW, Planchard D, Ohe Y, West H, et al. Alectinib in ALK-positive, crizotinib-
Ramalingam SS, et al. AZD9291 in EGFR inhibitor- resistant, non-small-cell lung cancer: a single
resistant non-small-cell lung cancer. N Engl J Med. group, multicenter, phase 2 trial. Lancet Oncol.
2015;372(18):1689–99. 2016;17(2):234–42.
24. Hoffknecht P, Tufman A, Wehler T, Pelzer T, Wiewrodt 36. Ou SH, Ahn JS, De Petris L, Govindan R, Yang JC,
R, Schutz M, et al. Efficacy of the irreversible ErbB Hughes B, et al. Alectinib in crizotinib-refractory
family blocker afatinib in epidermal growth factor ALK-rearranged non-small-cell lung cancer: a phase
receptor (EGFR) tyrosine kinase inhibitor (TKI)- II global study. J Clin Oncol. 2016;34(7):661–8.
pretreated non-small-cell lung cancer patients with 37. Kim D-W, Mehra R, Tan D, Felip E, Chow L,

brain metastases or leptomeningeal disease. J Thorac Camidge P, et al. Activity and safety of ceritinib in
Oncol. 2015;10(1):156–63. patients with ALK-rearranged non-small-cell lung
25. Schuler M, Wu YL, Hirsh V, O’Byrne K, Yamamoto N, cancer (ASCEND-1): updated results from the mul-
Mok T, et al. First-line afatinib versus chemotherapy in ticentre, open-label, phase 1 trial. Lancet Oncol.
patients with non-small cell lung cancer and common 2016;17(4):452–63.
epidermal growth factor receptor gene mutations and 38. Raju S, Joseph R, Sehgal S. Review of checkpoint
brain metastases. J Thorac Oncol. 2016;11(3):380–90. immunotherapy for the management of non-small
26. Park SJ, Kim HT, Lee DH, Kim KP, Kim SW, Suh cell lung cancer. Immunotargets Ther. 2018;7:63–75.
C, et al. Efficacy of epidermal growth factor receptor 39. Borghaei H, Paz-Ares L, Horn L, Spigel DR, Steins
tyrosine kinase inhibitors for brain metastasis in non- M, Ready NE, et al. Nivolumab versus docetaxel in
small cell lung cancer patients harboring either exon advanced nonsqaumous non-small-cell lung cancer. N
19 or 21 mutation. Lung Cancer. 2012;77(3):556–60. Engl J Med. 2015;373(17):1627–39.
27. Porta R, Sanchez-Torres JM, Paz-Ares L, Massuti 40. Herbst RS, Baas P, Kim DW, Felip E, Perez-Gracia JL,
B, Reguart N, Mayo C, et al. Brain metastases from Han JY, et al. Pembrolizumab versus docetaxel for pre-
lung cancer responding to erlotinib: the importance of viously treated, PD-L1-positive advanced non-small-
EGFR mutation. Eur Respir J. 2011;37(3):624–31. cell lung cancer (KEYNOTE-010): a randomized
28. Ballard P, Yates JW, Yang Z, Kim DW, Yang JC, controlled trial. Lancet. 2016;387(10027):1540–50.
Cantarini M, et al. Preclinical comparison of osimer- 41. Reck M, Rodriguez-Abreu D, Robinson AG, Hui R,
tinib with other EGFR-TKIs in EGFR-mutant Csoszi T, Fulop A, et al. Pembrolizumab versus che-
NSCLC brain metastases models, and early evidence motherapy for PD-L1-positive non-small-cell lung
of clinical brain metastases activity. Clin Cancer Res. cancer. N Engl J Med. 2016;375(19):1823–33.
2016;22(20):5130–40. 42. Goldberg SB, Gettinger SN, Mahajan A, Chiang

29. Guerin A, Sasane M, Zhang J, Culver KW, Dea K, AC, Herbst RS, Sznol M, et al. Pembrolizumab for
Nitulescu R, et al. Brain metastases in patients with patients with melanoma or non-small-cell lung can-
ALK+ non-small cell lung cancer: clinical symptoms, cer and untreated brain metastases: early analysis of
treatment patterns, and economic burden. J Med a non-randomised, open-label, phase 2 trial. Lancet
Econ. 2015;18(4):312–22. Oncol. 2016;17(7):976–83.
43. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau HER2-positive breast cancer. Clin Cancer Res.
FD, Lai P, Sawaya RE. Incidence proportions of brain 2009;15(4):1452–9.
metastases in patients diagnosed (1973 to 2001) in the 57. Sutherland S, Ashley S, Miles D, Chan S, Wardley A,
Metropolitan Detroit Cancer Surveillance System. J Davidson N, et al. Treatment of HER2-positive meta-
Clin Oncol. 2004;22(14):2865–72. static breast cancer with lapatinib and capecitabine in
44. Sperduto PW, Chao ST, Sneed PK, Luo X, Suh J, the lapatinib expanded access programme, including
Roberge D, et al. Diagnosis-specific prognostic fac- efficacy in brain metastases – the UK experience. Br J
tors, indexes, and treatment outcomes for patients Cancer. 2010;102(6):995–1002.
with newly diagnosed brain metastases: a multi- 58. Lin NU, Eierman W, Greil R, Campone M, Kaufman
institutional analysis of 4,259 patients. Int J Radiat B, Steplewski K, et al. Randomized phase II study of
Oncol Biol Phys. 2010;77(3):655–61. lapatinib plus capecitabine or lapatinib plus topotecan
45. Cleator S, Heller W, Coombes RC. Triple-negative for patients with HER2-positive breast cancer brain
breast cancer: therapeutic options. Lancet Oncol. metastases. J Neurooncol. 2011;105(3):613–20.
2007;8(3):235–44. 59. Bachelot T, Romieu G, Campone M, Dieras V, Cropet
46. Farmer H, McCabe N, Lord CJ, Tutt AN, Johnson C, Dalenc F, et al. Lapatinib plus capecitabine in
DA, Richardson TB, et al. Targeting the DNA repair patients with previously untreated brain metasta-
defect in BRCA mutant cells as a therapeutic strategy. ses from HER-2 positive metastatic breast cancer
Nature. 2005;434(7035):917–21. (LANDSCAPE): a single-group phase II study.
47. Lin NU, Winer EP. Brain metastases: the HER2 para- Lancet Oncol. 2013;14(1):64–71.
digm. Clin Cancer Res. 2007;13(6):1648–55. 60. Awada A, Colomer R, Inoue K, Bondarenko I, Badwe
48. Venur VA, Leone JP. Targeted therapies for brain
RA, Demetriou G, et al. Neratinib plus paclitaxel vs
metastases from breast cancer. Int J Mol Sci. trastuzumab plus paclitaxel in previously untreated
2016;17(9):E1543. metastatic ERBB2-positive breast cancer: the
49. Niikura N, Liu J, Hayashi N, Mittendorf EA, Gong Y, NEfERT-T randomized clinical trial. JAMA Oncol.
Palla SL, et al. Loss of human epidermal growth fac- 2016;2(12):1557–64.
tor receptor 2 (HER2) expression in metastatic sites of 61. Andre F, O’Regan R, Ozguroglu M, Toi M, Xu B,
HER2-overexpressing primary breast tumors. J Clin Jerusalem G, et al. Everolimus for women with
Oncol. 2012;30(6):593–9. trastuzumab-resistant, HER2-positive, advanced
50. Stemmler HJ, Schmitt M, Willems A, Bernhard H, breast cancer (BOLERO-3): a randomized, double-
Harbeck N, Heinemann V. Ratio of trastuzumab levels blind, placebo-controlled phase 3 trial. Lancet Oncol.
in serum and cerebrospinal fluid is altered in HER2- 2014;15(6):580–91.
positive breast cancer patients with brain metastases 62. Baselga J, Campone M, Piccart M, Burris HA III, Rugo
and impairment of blood-brain barrier. Anticancer HS, Sahmoud T, et al. Everolimus in postmenopausal
Drugs. 2007;18(1):23–8. hormone-receptor-positive advanced breast cancer. N
51. Swain SM, Baselga J, Kim SB, Ro J, Semiglazov V, Engl J Med. 2012;366(6):520–9.
Campone M, et al. Pertuzumab, trastuzumab, and 63. Franz DN, Belousouva E, Sparagana S, Bebin EM,
docetaxel in HER2-positive metastatic breast cancer. Frost M, Kuperman R, et al. Efficacy and safety of
N Engl J Med. 2015;372(8):724–34. everolimus for subependymal giant cell astrocytomas
52. Senda N, Yamaguchi A, Nishimura H, Shiozaki T, associated with tuberous sclerosis complex (EXIST-1):
Tsuyuki S. Pertuzumab, trastuzumab, and docetaxel a multi-centre, randomized, placebo-controlled phase
reduced the recurrence of brain metastasis from 3 trial. Lancet. 2013;381(9861):125–32.
breast cancer: a case report. Breast Cancer Tokyo Jpn. 64. Turner NC, Ro J, Andre F, Loi S, Verma S,

2016;23(2):323–8. Iwata H, et al. Palbociclib in hormone-receptor-
53. Koumarianou A, Kontopoulou C, Kouloulias V,
positive advanced breast cancer. N Engl J Med.
Tsionou C. Durable clinical benefit of pertuzumab 2015;373(3):209–19.
in a young patient with BRCA2 mutation and HER2- 65. Sahebjam S, Rhun EL, Kulanthiavel P, Turner PK,
overexpression breast cancer involving the brain. Klise S, Wang HT, et al. Assessment of concentra-
Case Rep Oncol Med. 2016;2016:5718104. tions of abemaciclib and its major active metabolites
54. Bartsch R, Berghoff AS, Vogl U, Rudas M, Bergen E, in plasma, CSF, and brain tumor tissue in patients
Dubsky P, et al. Activity of T-DM1 in HER2-positive with brain metastases secondary to hormone recep-
breast cancer brain metastases. Clin Exp Metastasis. tor positive (HR+) breast cancer. J Clin Oncol.
2015;32(7):729–37. 2016;34(15_suppl):526.
55. Saleem A, Searle GE, Kenny LM, Huiban M,
66. Dagogo-Jack I, Gill CM, Cahill DP, Santagata

Kozlowski K, Waldman AD, et al. Lapatinib access S, Brastianos PK. Treatment of brain metasta-
into normal brain and brain metastases in patients ses in the modern genomic era. Pharmacol Ther.
with Her-2 overexpressing breast cancer. EJNMMI 2017;170:64–72.
Res. 2015;5:30. 67. Long GV, Menzies AM, Nagrial AM, Haydu LE,

56. Lin NU, Dieras V, Paul D, Lossignol D, Christodoulou Hamilton AL, Mann GJ, et al. Prognostic and clinico-
C, Stemmler HJ, et al. Multicenter phase II study pathologic associations of oncogenic BRAF in meta-
of lapatinib in patients with brain metastases from static melanoma. J Clin Oncol. 2011;29(10):1239–46.

ipilimumab in patients with metastatic melanoma. N
Ascierto PA, Chapman PB, et al. Dabrafenib in Engl J Med. 2010;363(8):711–23.
patients with Val600Glu or Val600Lys BRAF-mutant 77. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ,
melanoma metastatic to the brain (BREAK-MB): a Cowey CL, Lao CD, et al. Combined nivolumab and
multi-center, open-label, phase 2 trial. Lancet Oncol. ipilimumab or monotherapy in untreated melanoma.
2012;13(11):1087–95. N Engl J Med. 2015;373(1):23–34.
69. Harding JJ, Catalanotti F, Munhoz RR, Cheng DT, 78. Postow MA, Chesney J, Pavlick AC, Robert C,

Yaqubie A, Kelly N, et al. A retrospective evalu- Grossmann K, McDermott D, et al. Nivolumab and
ation of vemurafenib as treatment for BRAF- ipilimumab versus ipilimumab in untreated mela-
mutant melanoma brain metastases. Oncologist. noma. N Engl J Med. 2015;372(21):2006–17.
2015;20(7):789–97. 79. Tazi K, Hathaway A, Chiuzan C, Shirai K. Survival of
70. Davies MA, Stemke-Hale K, Lin E, Tellez C, Deng melanoma patients with brain metastases treated with
W, Gopal YN, et al. Integrated molecular and clinical ipilimumab and stereotactic radiosurgery. Cancer
analysis of AKT activation in metastatic melanoma. Med. 2015;4(1):1–6.
Clin Cancer Res. 2009;15(24):7358–46. 80. Margolin K, Ernstoff MS, Hamid O, Lawrence D,
71. Long GV, Stroyakovskiy D, Gogas H, Levchenko
McDermott D, Puzanov I, et al. Ipilimumab in patients
E, de Braud F, Larkin J, et al. Combined BRAF and with melanoma and brain metastases: an open-label,
MEK inhibition versus BRAF inhibition alone in mel- phase II trial. Lancet Oncol. 2012;13(5):459–65.
anoma. N Engl J Med. 2014;371(20):1877–88. 81. Specenier P. Nivolumab in melanoma. Expert Rev
72. Long GV, Stroyakovskiy D, Gogas H, Levchenko
Anticancer Ther. 2016;16(12):1247–61.
E, de Braud F, Larkin J, et al. Dabrafenib and tra- 82. Robert C, Schachter J, Long GV, Arance A, Grob
metinib versus dabrafenib and placebo for Val600 JJ, Mortier L, et al. Pembrolizumab versus ipi-
BRAF-mutant melanoma: a multicenter, double- limumab in advanced melanoma. N Engl J Med.
blind, phase 3 randomised controlled trial. Lancet. 2015;372(26):2521–32.
2015;386(9992):444–51. 83. Knisely JPS, Yu JB, Flanigan J, Sznol M, Kluger
73. Robert C, Karaszewska B, Schachter J, Rutkowski HM, Chiang VLS. Radiosurgery for melanoma brain
P, Mackiewicz A, Stroiakovski D, et al. Improved metastases in the ipilimumab era and the possibility of
overall survival in melanoma with combined dab- longer survival. J Neurosurg. 2012;117(2):227–33.
rafenib and trametinib. N Engl J Med. 2015;371(1): 84. Ahmed KA, Stallworth DG, Kim Y, Johnstone PA,
30–9. Harrison LB, Caudell JJ, et al. Clinical outcomes
74. Meier FE, Niessner H, Schmitz J, Schmid A,
of melanoma brain metastases treated with stereo-
Calaminus C, Pichler B, et al. The PI3K inhibitor tactic radiation and anti-PD-1 therapy. Ann Oncol.
BKM120 has potent antitumor activity in melanoma 2016;27(3):434–41.
brain metastases in vitro and in vivo. J Clin Oncol. 85. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS,
2013;31(15_suppl):e20050. Moschos SJ, et al. Combination nivolumab and ipili-
75. Krieg C, Letourneau S, Pantaleo G, Boyman
mumab in melanoma metastatic to the brain. N Engl J
O. Improved IL-2 immunotherapy by selec- Med. 2018;379:722–30.
tive stimulation of IL-2 receptor on lymphocytes 86. Long GV, Atkinson V, Lo S, Sandhu S, Guminski AD,
and endothelial cells. Proc Natl Acad Sci U S A. Brown MP, et al. Combination nivolumab and ipilim-
2010;107(26):11906–11. umab or nivolumab alone in melanoma brain metas-
76. Hodi FS, O’Day SJ, McDermott DF, Weber RW,
tases: a multicenter randomised phase 2 study. Lancet
Sosman JA, Haanen JB, et al. Improved survival with Oncol. 2018;19(5):672–81.
Classification of Brain Metastases
6
Paul W. Sperduto
Introduction radiation therapy (WBRT), chemotherapy, tar-

geted drug therapies, and immunotherapies].
Brain metastases are a common and complex Furthermore, four prospective randomized trials
conundrum for cancer care. An estimated 300,000 have shown WBRT adds no survival benefit over
patients are diagnosed each year with brain SRS alone in SRS-eligible patients [4–7] and, on
metastases in the United States [1] and that inci- the other end of the prognostic spectrum, there is
dence is growing due to advances in treatment evidence that supportive care may be as effective
that result in patients living longer and thus at as WBRT [8]. Accordingly, WBRT is used less
prolonged risk for development of brain metasta- commonly than in the past.
ses [2]. It is a complex problem because of the
marked heterogeneity of this patient population:
brain metastases may arise from a wide variety of Classification Systems
tumor types and subtypes. Furthermore, these
patients may have already received a plethora of These concerns led to efforts to better understand
different treatments for their cancer or may pres- prognosis. The purpose of a prognostic index is
ent with brain metastases at the time of initial to predict outcome before, not after, treatment. It
diagnosis. This heterogeneity has long plagued is important to distinguish prognostic from pre-
interpretation of clinical trials involving this dictive factors. A prognostic factor identifies
patient population because it was essentially good versus bad outcome irrespective of the
impossible to sufficiently stratify studies to ver- treatment used, whereas a predictive factor iden-
ify similar groups of patients were being com- tifies good versus bad outcome for a specific
pared [3]. Interpretation of clinical trials and treatment. Gaspar et al. published the Radiation
efforts to estimate prognosis are further compli- Therapy Oncology Group (RTOG) Recursive
cated by the plethora of possible combinations of Partitioning Analysis for brain metastases
currently available treatment options [surgery, (Table 6.1) in 1997 [9]. This prognostic index
stereotactic radiosurgery (SRS), whole brain consisted of three classes: I (age < 65, Karnofsky
performance score (KPS) ≥70, controlled pri-
mary tumor, no extracranial metastases), II (all
patients not in class I or III), and III (KPS < 70),
P. W. Sperduto (*) which correlated with median survival of 7.7,
Minneapolis Radiation Oncology & University 4.5, and 2.3 months, respectively, at that time.
of Minnesota, Minneapolis, MN, USA Weltman et al. published the score index for
e-mail: psperduto@mropa.com

https://doi.org/10.1007/978-3-030-42958-4_6
84 P. W. Sperduto
Table 6.1 Radiation Therapy Oncology Group (RTOG) Table 6.3 Basic score for brain metastases (BSBM)
recursive partitioning analysis (RPA) for patients with
Score
brain metastases
0 1
Median KPS 50–70 80–100
Class Criteria survival Control of primary tumor No Yes
Class I Age < 65 yrs, KPS ≥ 70, 7.1 mo Extracranial metastases Yes No
controlled primary tumor, and
no extracranial metastases Data from Ref. [11]
Median survival (MS) by BSBM: BSBM 3 (MS >32 mo),
Class II All patients not in Class I or III 4.2 mo
BSBM 2 (MS 13.1 mo), BSBM 1 (MS 3.3 mo), BSBM 0
Class III KPS < 70 2.1 mo (MS 1.9 mo)
Data from Ref. [9] KPS Karnofsky performance status
KPS Karnofsky performance status
patients, thus making the results of those trials

Table 6.2 Score index for radiosurgery (SIR) worthwhile, relevant, and interpretable.
Score Our group has published a series of articles
0 1 2 developing and refining a diagnosis-specific
Age (years) ≥60 51–59 ≤50 prognostic index, the graded prognostic assess-
KPS ≤50 60–70 80–100 ment (GPA), for patients with brain metastases.
Systemic disease Progressive Stable CR or NED The GPA was first published in 2008 [14] based
Number of lesions ≥3 2 1
on 1960 patients from five randomized
Volume of largest >13 5–13 <5
lesion (mL) Radiation Therapy Oncology Group (RTOG)
Data from Ref. [10]
trials (7916, 8528, 8905, 9104, and 9508).
Median survival (MS) by SIR score: SIR 1–3 (MS 2.91 Analysis showed four prognostic factors (age,
mo), SIR 4–7 (MS 7.00 mo), SIR 8–10 (MS 31.38 mo) KPS, extracranial metastases, and number of
KPS Karnofsky performance status, CR complete brain metastases) were significant for survival.
response, NED no evidence of disease
Those prognostic factors were weighted in pro-
portion to their regression coefficients and
radiosurgery (SIR) (Table 6.2) in 2000 [10]. This scaled such that patients with the best/worst
index used the sum of scores (0–2) for each of prognosis would have a GPA of 4.0/0.0, respec-
five prognostic factors (age, KPS, status of sys- tively. In 2010, we refined the GPA based on an
temic disease, number of brain metastases, and analysis of a retrospective multi-institutional
the volume of the largest metastasis). Lorenzoni database of 4259 patients. That study found
et al. published the basic score for brain metasta- survival varies by diagnosis and diagnosis-spe-
ses (BSBM) (Table 6.3) in 2004 [11]. This index cific prognostic factors [15]. The Breast-GPA
is based on the sum of scores (0–1) for three was then further refined using tumor subtype
prognostic factors (KPS, control of primary [16] and a summary report was published [17].
tumor, and extracranial metastases). In 2012, More recently, the GPA indices for lung cancer,
Sloan-Barnholtz published a nomogram melanoma, and renal cell carcinoma have been
(Fig. 6.1) in an effort to further individualize updated using molecular and other clinical fac-
prognosis [12]. In 2014, Kondziolka published an tors with new data from patients (2,186 lung
interesting survey study in which experts in the cancer and 823 melanoma patients) diagnosed
field were asked to estimate survival for a series since 2005 including molecular factors. The
of patients given all relevant clinical parameters. Lung-molGPA incorporates EGFR and ALK
This study showed that even experts cannot pre- gene status [18, 19] and similarly the mela-
dict outcomes with certainty for all patients [13]. noma-molGPA incorporates BRAF status [20,
All prognostic indices have limitations but can 21]. The original melanoma-GPA found only
provide guidance for clinical decision-making two factors to be significant (KPS and the num-
and are essential for stratification of clinical trials ber of brain metastases), whereas the updated
so that those trials are comparing comparable melanoma-molGPA found five factors (BRAF
6 Classification of Brain Metastases 85
0 10 20 30 40 50 60 70 80 90 100
Points
BO OO LL LSM SMM
Site and histology
BA LA OA OSQ LQ LSQ
Tumor uncontrolled
Status of primary disease
Tumor controlled
Brain & other sites

Metastatic spread
Brain alone
GR Biopsy only
Surgery status
CR PR None
Age
10 30 50 60 70 80 90
<70
KPS
>=70
Multiple
Number of brain lessions
Single
Total points
0 20 40 60 80 100 120 140 160 180 200 220 240
Predicted 6-month survival prob.

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1
Predicted 12-month survival prob.

0.65 0.6 0.5 0.4 0.3 0.2 0.1 0.05 0.01
Predicted median survival days

600 500 400 300 250 200 150 120 100 90 80 70 60
Fig. 6.1 Nomogram for 6-month and 12-month survival lung and squamous cell, OA other and adenocarcinoma,
probability and median survival prediction for RTOG OSQ other and squamous cell, SMM skin-melanoma, OO
brain metastases patients. Abbreviations for site and his- other and other. Surgery: PR partial resection, CR com-
tology: BA breast and adenocarcinoma, BO breast and plete resection, GR gross resection. (Reprinted from
other, LA lung and adenocarcinoma, LL lung and large Sloan-Barnholtz-Sloan et al. [12], with permission from
cell, LO lung and other, LSM lung and small cell, LSQ Oxford University Press)
status, KPS, age, extracranial metastases, and cranial metastases, and the number of brain
number of brain metastases) to be significant. metastases [22, 23].
The renal GPA has also been updated. Data Table 6.4 shows the median survival time for
from 711 renal cell carcinoma patients with patients with brain metastases by diagnosis-
brain metastases, diagnosed between 2006 and specific GPA. Table 6.5 shows the diagnosis-
2016, showed four prognostic factors to be sig- specific definition of the updated GPA indices
nificant for survival: KPS, hemoglobin, extra- and a user-friendly worksheet to facilitate cal-
86 P. W. Sperduto
Table 6.4 Median survival time for patients with brain metastases by diagnosis specific—graded prognostic assess-
ment score
DS-GPA
Overall 0–1.0 1.5–2.0 2.5–3.0 3.5–4.0
MST (95% CI) MST (95% CI) MST (95% CI) MST (95% CI) MST (95% CI)
Diagnosis N n (%) n (%) n (%) n (%) p (log-rank)
NSCLC 15 (14–17) 7 (6–9) 14 (12–15) 26 (23–31) 47 (37-NE) <0.001
1521 337 (22%) 664 (44%) 455 (30%) 65 (4%)
SCLC 5 (4–6) 3 (2–3) 5 (4–7) 8 (6–9) 17 (5–27) <0.001
281 65 (23%) 119 (42%) 84 (30%) 13 (5%)
Melanoma 10 (9–11) 5 (4–7) 8 (7–9) 16 (13–19) 34 (24–50) <0.001
823 136 (17%) 386 (47%) 256 (31%) 45 (5%)
RCC 12 (11–13) 4 (3–5) 12 (9–14) 17 (13–21) 35 (20–41) <0.001
669 170 (25%) 178 (27%) 204 (30%) 117 (17%)
Breast cancer 14 (12–16) 3 (3–4) 8 (6–9) 15 (13–16) 25 (23–27) <0.001
400 23 (6%) 104 (26%) 140 (35%) 133 (33%)
GI cancer 5 (4–6) 3 (2–5) 4 (3–7) 7 (5–12) 14 (10–27) <0.001
209 76 (36%) 65 (31%) 50 (24%) 18 (9%)
Other 6 (5–7) – – – – –
450
The top row in each cell is the median survival time (MST) in months and its associated 95% CI. The bottom row is the
frequency and percentage of patients with the corresponding DS-GPA category for a given diagnosis. Abbreviations:
DS-GPA Diagnosis specific-graded prognostic assessment, NSCLC non-small cell lung cancer (adenocarcinoma),
SCLC small cell lung cancer, RCC renal cell carcinoma, GI gastrointestinal, NE not estimable
Table 6.5 GPA worksheet to estimate survival from brain metastases by diagnosis
Non-small cell/small cell
lung cancer GPA scoring criteria Patient
0 0.5 1.0 Score
Age ≥70 <70 n/a –
KPS ≤70 80 90–100 –
ECM Present Absent –
#BM >4 1–4 n/a –
Gene status EGFR neg/unk and ALK n/a EGFR pos or –
neg/unk ALK pos
Sum total = –
Adenocarcinoma MS by GPA: GPA 0–1.0 = 6.9; 1.5–2.0 = 13.7; 2.5–3.0 = 26.5; 3.5–4.0 = 46.8
Non-adenocarcinoma MS by GPA: GPA 0–1.0 = 5.3; 1.5–2.0 = 9.8; 2.5–3.0 = 12.8
Melanoma 0 0.5 1.0 Score
Age ≥70 <70 n/a –
KPS <70 80 90–100 –
ECM Present n/a Absent –
#BM >4 2–4 1 –
Gene status BRAF neg/unk BRAF n/a –
pos
Sum total = –
MS (mo) by GPA: 0–1.0 = 4.9, 1.5–2.0 = 8.3, 2.5–3.0 = 15.8, 3.5–4.0 = 34.1
Breast cancer 0 0.5 1.0 1.5 2.0 Score
KPS ≤50 60 70–80 90–100 n/a –
Subtype Basal n/a LumA HER2 LumB –
Age ≥60 <60 n/a n/a n/a –
Sum total = –
Subtype: Basal = triple negative (ER/PR/HER2-neg)
LumA = Luminal A (ER/PR-pos, HER2-neg)
LumB = Luminal B (triple positive, ER/PR/HER2-pos)
HER2 = HER2-pos, ER/PR-neg
Non-small cell/small cell
lung cancer GPA scoring criteria Patient
MS (mo) by GPA: 0–1.0 = 3.4, 1.5–2.0 = 7.7, 2.5–3.0 = 15.1, 3.5–4.0 = 25.3
Renal cell carcinoma 0 0.5 1.0 2.0 Score
KPS <80 80 90–100 –
ECM Present Absent –
Hgb ≤11 11.1–12.5 >12.5 –
#BM >4 1–4 –
Sum Total = –
MS (mo) by GPA: 0–1.0 = 3.3, 1.5–2.0 = 7.3, 2.5–3.0 = 11.3, 3.5–4.0 = 14.8
GI cancers 0 1 2 3 4 Score
KPS <70 70 80 90 100 –
MS (mo) by GPA: 0–1.0 = 3.1, 2.0 = 4.4, 3.0 = 6.9, 4.0 = 13.5
Data from Refs. [17, 19, 21]
Abbreviations: GPA graded prognostic assessment, KPS Karnofsky performance score, ECM extracranial metastases,
#BM number of brain metastases, ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth fac-
tor receptor 2, MS median survival in months, neg/unk negative or unknown
culation of the graded prognostic assessment ized and the past philosophy of fatalistic futility
by diagnosis and estimate survival for patients should be abandoned. (2) On the other hand, as
with brain metastases. A free online/smart shown in Table 6.4, if a patient has a GPA of
phone application is available at brainmetgpa. 0–1.0, regardless of diagnosis, their expected sur-
com, which further simplifies the calculation of vival is poor. For these patients, supportive care,
the GPA. as suggested by the QUARTZ Trial [8], may be
Table 6.6 shows a multivariate analysis of risk the best option. (3) For patients with GPA scores
of death and median survival by treatment above 1.0, the median survival time (Table 6.4)
(excluding drug therapies) and diagnosis. It is varies more by diagnosis and more aggressive
important to understand these data are retrospec- treatment strategies may be appropriate, but these
tive in nature with the selection bias inherent in retrospective data do not provide a basis for
all retrospective studies so one should not con- assuming that longer survival is a consequence of
clude that one treatment is better than another more aggressive treatment. Indeed, the survival
based on these data. Figure 6.2 shows Kaplan– by treatment data shown in Table 6.4 is certainly
Meier curves for survival for six diagnoses by fraught with selection bias and should not be
GPA, demonstrating excellent separation blindly applied or expected. Nonetheless, these
between groups. data reflect patterns of care for patients with brain
The diagnosis-specific GPA indices presented metastases. (4) Performance status is prognostic
here define how survival has improved for brain in every diagnosis. Clinicians should take the
metastasis patients over the past four decades. time to accurately assess and document their
This progress mirrors the progress seen in sur- patients’ performance status. (5) Table 6.5 shows
vival for patients with the same diagnoses who the number of brain metastases is a significant
do not have brain metastases. These data hold prognostic factor for lung cancer, melanoma, and
several implications for clinical management and renal cell carcinoma, but not for breast or gastro-
research involving patients with brain metasta- intestinal cancers. Patients should not be denied
ses: (1) There is marked heterogeneity in out- treatment because of the number of brain metas-
comes for patients with brain metastases and tases. (6) Extracranial metastases are only prog-
these outcomes vary not only by diagnosis but nostic in lung cancer and melanoma but not in
also by diagnosis-specific prognostic factors, as breast cancer, renal cell carcinoma, or gastroin-
detailed herein. Because of this heterogeneity, we testinal cancers. The implication here is that
should not treat all patients with brain metastases those patients with nonlung, nonmelanoma
the same way—treatment should be individual- malignancies should not be denied aggressive
88 P. W. Sperduto
Table 6.6 Multivariable analysis of risk of death and median survivala by treatment and diagnosis
Treatment
WBRT + S+ S + WBRT
WBRT SRS SRS S + SRS WBRT + SRS
Diagnosis Statistics
NSCLC Risk of death 1.0 1.08 1.20 0.66b 0.78 0.79
n = 1,521 (HR)
95% CI 0.92–1.27 0.94–1.54 0.50–0.88 0.58–1.06 0.40–1.58
p-value 0.35 0.15 <0.01 0.11 0.51
Median survivala 13 14 10 32 20 20
n (%) 342 (22%) 767 (50%) 139 (9%) 114 (7%) 76 (5%) 13 (1%)
SCLC Risk of death 1.0 0.97 0.24b 0.00 0.42b 0.00
n = 281 (HR)
95% CI 0.41–2.26 0.10–0.59 NA 0.25–0.73 NA
p-value 0.94 0.002 0.99 0.002 0.98
n (%) 229 (81%) 13 (5%) 21 (7%) 1 (0.4%) 16 (6%) 1 (0.4%)
Melanoma Risk of death 1.0 0.69b 0.62b 0.50b 0.54b 0.70
n = 823 (HR)
95% CI 0.54–0.89 0.45–0.86 0.36–0.69 0.35–0.84 0.36–1.36
p-value < 0.01 <0.01 <0.01 <0.01 0.29
n (%) 91 (11%) 464 (56%) 73 (9%) 95 (12%) 34 (4%) 12 (1%)
Renal cell Risk of death 1.00 0.84 0.78 0.38 0.64 1.29
n = 711 (HR)
95% CI 0.62–1.12 0.51–1.19 0.25–0.59 0.38–1.08 0.45–3.68
p-value 0.23 0.25 <0.01 0.09 0.64
n (%) 90 (12%) 410 (58%) 41 (6%) 70 (10%) 23 (3%) 4 (1%)
Breast Risk of death 1.0 1.07 0.74 0.59 0.72 0.47b
cancer (HR)
n = 400 95% CI 0.66–1.73 0.47–1.16 0.28–1.23 0.43–1.21 0.23–0.96
p-value 0.80 0.18 0.16 0.72 0.04
n (%) 131 (33%) 115 (29%) 86 (22%) 19 (5%) 28 (7%) 20 (5%)
GI cancer Risk of death 1.0 0.72 0.69 2.30 0.33b 0.39b
n = 209 (HR)
95% CI 0.40–1.28 0.39–1.22 0.43–12.4 0.19–0.56 0.17–0.90
p-value 0.26 0.21 0.33 <0.001 0.03
n (%) 95 (45%) 35 (17%) 35 (17%) 2 (1%) 34 (16%) 8 (4%)
Data from Refs. [17, 19, 21]
Diagnoses: NSCLC non-small-cell lung cancer (adenocarcinoma), SCLC small-cell lung cancer, GI gastrointestinal
Treatments: S surgery, WBRT whole brain radiation therapy, SRS stereotactic radiosurgery
Statistics: Risk of death: hazard ratio (HR) normalized to patients treated with whole brain radiation therapy alone
(HR = 1.0) and calculated by multivariable Cox regression, adjusted for DS-GPA and stratified by institution
a
Median survival in months based on one-sample Kaplan–Meier method
b
Statistically significantly better than WBRT alone; 95% confidence interval
Fig. 6.2 Kaplan–Meier
curves for survival by GPA a b
for six diagnoses: breast
cancer, non-–small-cell lung
cancer, small-cell lung cancer,
melanoma, renal cell
carcinoma, gastrointestinal
cancers. (a) Initial MRI
shows largest of three brain
metastases, December 06,
2006. (b) Gamma Knife plan
for right frontal brain
metastasis, December 13,
2006. (c) Gamma Knife plan
for left frontal brain
2006. (d) Gamma Knife plan
c d
for left occipital brain
2006. (e) MRI 9 months after
GK shows marked radiation
necrosis and edema,
September 26, 2007. (f) MRI
18 months after GK shows
resolving radiation necrosis,
May 23, 2008. (g) MRI
21 months after GK shows
minimal residual
enhancement, October 23,
2008. (h) MRI 10.7 years after
GK shows no evidence of e f
disease, August 02, 2017.
(From Sperduto et al. [24].
Creative Commons Attribution
License CC-BY 3.0)
g h
90 P. W. Sperduto
treatment for their brain metastases because they melanoma by excisional biopsy of a posterior
have extracranial metastases. (7) Age is strongly scalp lesion on September 15, 2005. This
prognostic in lung cancer and weakly prognostic malignant melanoma was histopathologically
in breast cancer and melanoma but not prognostic staged as Clark’s Level IV, Breslow depth at
in renal cell carcinoma or gastrointestinal can- least 6 mm, with angiolymphatic invasion and
cers. Thus, age should not be used as a rationale positive deep and peripheral margins. Brain
to withhold aggressive treatment for nonlung MRI for initial radiologic staging on September
malignancies. (8) Because lung cancer and brain 27, 2005, showed multiple scalp lesions but no
metastases from lung cancer are so common, evidence of parenchymal brain metastases. PET
those patients have masked our understanding of scan on September 27, 2005, showed hypermet-
the distinct course for patients with nonlung abolic activity only in the left neck. On October
malignancies and brain metastases, as demon- 11, 2005, she underwent a left modified radical
strated by points 5, 6, and 7 above. (9) Tumor neck dissection and wide local excision of the
subtype in breast cancer is of paramount impor- scalp lesion. Pathology confirmed metastatic
tance and prognostic significance but it is not as melanoma in 3 of 28 lymph nodes with exten-
prognostic as the Breast-GPA index. (10) A dis- sion into the adjacent soft tissues in two areas.
proportionate number of patients with gastroin- Pathology from the scalp excision showed a
testinal cancers present with GPA of 0–1.0. maximum tumor depth of 1.9 cm and the deep
Whether this is due to lack of screening MRI in margin remained positive. She underwent two
these patients versus other biological reasons additional scalp excisions and the deep margin
remains unclear but the finding should serve as a remained positive. Her stage was T4bN2bM0,
reminder that brain metastases are not uncom- stage IIIC. She received 64 Gy radiation ther-
mon in GI cancer patients. On-going research apy to the left neck and scalp, completed on
will better elucidate prognosis for these patients January 20, 2006. She then received three cycles
and the GI-GPA will be updated accordingly. of cisplatinum, interferon, and vinblastine fol-
(11) Clinicians may use the worksheet in lowed by interleukin-2, completed in March
Table 6.5 or go to brainmetgpa.com, a free user- 2006. She did well without evidence of recur-
friendly smart-phone application to calculate rence until November 2006 when she under-
their patient’s GPA score and estimate survival went a debridement of necrotic tissue in the
[12]. The GPA may be used for purposes of strati- scalp lesion. PET scan on December 5, 2006,
fication in clinical trials dealing with patients showed a 0.7 cm hypermetabolic nodule in the
with brain metastases. retroperitoneum consistent with metastatic
All prognostic indices are imperfect and can- recurrence. Brain MRI on December 6, 2006,
not always predict the outcome for an individual showed three brain metastases (2.5 cm right
patient. The following case study is remarkable caudate, 1.1 cm left parieto- occipital, and
for the patient’s outcome because it demon- 0.7 cm left posterior frontal) (Fig. 6.2a), which
strates not only the application of the GPA in a were not present on the prior scan performed on
clinical setting but also the potential pitfalls of June 22, 2006.
prognostic indices for such a heterogeneous Whole brain radiation therapy was not given
patient population. (and has not been given) due to the prior scalp
radiation. She underwent SRS (Gamma Knife)
on December 13, 2006, to all three lesions: right
Case Study caudate, 20 Gy to a volume 8.4 cm3 (Fig. 6.2b);
left posterior frontal 24 Gy to a volume of
A 36-year-old white female marathon runner 0.47 cm3 (Fig. 6.2c); and left parieto-occipital,
presented in August 2005 with a right neck 24 Gy to a volume of 1.6 cm3 (Fig. 6.2d). She
mass. Fine needle aspiration initially confirmed underwent SABR to the pelvic soft tissue
a malignancy, later confirmed as a malignant metastasis (25 Gy × 5 over two weeks, completed
on February 23, 2007). Between March and June between January 1, 2006, and December 31,
2007, she received four cycles of carboplatin, 2015. Notably, the patient presented here was
paclitaxel, and temozolomide treatment. In diagnosed in 2006, so she is a contemporary of
September 2007, she developed headaches, nau- the patients in the melanoma-molGPA update
sea, vomiting, and confusion. MRI on September study. The study showed five prognostic factors
26, 2007, showed a marked increase in enhance- significant for survival (Table 6.5).
ment and edema in the right frontal lobe consis- Overall median survival for melanoma
tent with radiation necrosis (Fig. 6.2e). Due to patients with brain metastases has improved from
increased headaches and possible radiation 6 to 10 months since the 1980s, and the median
necrosis, the temozolomide was discontinued. survival by melanoma-molGPA groups for GPA
She has received no treatment since September of 0–1.0, 1.5–2.0, 2.5–3.0, and 3.5–4.0 was 4.9,
2007. The edema was treated with steroids, which 8.3, 15.8, and 34.1 months, respectively. The
were gradually tapered off over four months. patient presented here had a melanoma-GPA of
Brain MRI on May 23, 2008, showed improve- 3.0 on a 4.0 scale on both the original and updated
ment with central necrosis of the previously GPA indices, correlating with an estimated sur-
solid-appearing lesion (Fig. 6.2f). Brain MRI on vival of 8.8 and 15.8 months, respectively. This
October 23, 2008, showed further resolution of patient is disease-free and asymptomatic with a
the enhancement/necrosis with minimal residual perfect FACT-Brain QOL score 13 years after the
enhancement (Fig. 6.2g). Serial imaging since diagnosis of multiple brain metastases. Clearly,
that time has shown no evidence of recurrent prognostic indices are imperfect but nonetheless
tumor or necrosis. provide our best estimate of survival for these
She remains clinically and radiographically patients.
free of disease 13 years after the diagnosis of
multiple brain metastases and more than 10 years
after completion of treatment. Brain MRI on Summary
August 2, 2017, showed no change in the mini-
mal residual enhancement/scar tissue (Fig. 6.2h) Patients with brain metastases are a heteroge-
and PET scan on August 2, 2017, showed no evi- neous population and outcomes vary widely by
dence of disease. She has remained asymptom- diagnosis and diagnosis-specific prognostic fac-
atic for over a decade and continues to run tors. Because of this heterogeneity and the pleth-
marathons, as recently as October 14, 2017. In ora of available treatment options, it is difficult to
November 2017, she completed the FACT-Brain estimate survival. These problems have compli-
questionnaire, a patient-reported QOL tool to cated clinical decision-making as well as inter-
reassess brain cognition. Her FACT-BR score pretation of clinical trials. The graded prognostic
was perfect (200 on a scale of 200), 11 years after assessment (GPA) is a diagnosis-specific prog-
diagnosis of her brain metastases. Notably, this nostic index that has been updated to reflect the
patient never underwent craniotomy or whole current treatment era by incorporating diagnosis-
brain radiation therapy and thus avoided the specific prognostic factors including molecular
related long-term neurocognitive toxicity of these factors such as tumor subtype and gene status.
interventions. The GPA is useful for clinical decision-making
To fully appreciate this patient’s remarkable as physicians determine whether and what treat-
outcome, it is appropriate to review how her out- ment is appropriate for these patients. It can also
come compares to the best available evidence of be useful to stratify clinical trials to ensure those
survival for melanoma patients with brain metas- trials are comparing comparable patients, which
tases. We recently updated and published the is especially important in such a heterogeneous
melanoma-molGPA [20, 21] based on a multi- patient population. Without accurate stratifica-
institutional retrospective study of 483 mela- tion, the results of clinical trials are uninterpreta-
noma patients with brain metastases diagnosed ble and a waste of resources.
92 P. W. Sperduto
Acknowledgments This work has been a collaborative 9. Gaspar LE, Scott C, Rotman M, et al. Recursive
multi-institutional effort. The faculty and residents of the partitioning analysis (RPA) of prognostic factors in
following institutions have selflessly contributed time and three Radiation Therapy Oncology Group (RTOG)
energy to one or more of the studies on the graded prog- brain metastases trials. Int J Radiat Oncol Biol Phys.
nostic Assessment: MD Anderson, Memorial Sloan 1997;37:745–51.
Kettering Cancer Center, Mayo Clinic, University of 10.
Weltman E, Salvajoli JV, Brandt RA, et al.
California San Francisco, Mayo Clinic, Massachusetts Radiosurgery for brain metastases: a score index for
General Hospital, Dana Farber Cancer Institute, Duke predicting prognosis. Int J Radiat Oncol Biol Phys.
University, Yale University, University of Colorado 2000;46:1155–61.
Denver, Cleveland Clinic, University of Wisconsin 11. Lorenzoni J, Devriendt D, Massager N, et al.

Madison, McGill University and Centre Hospitalier de l’ Radiosurgery for treatment of brain metastases:
Université de Montreal, University of Maryland, estimation of patient eligibility using three strati-
University of Alabama Birmingham, and the University of fication systems. Int J Radiat Oncol Biol Phys.
Minnesota. This work would not have been possible with- 2004;60:218–24.
out the tireless work of these dedicated colleagues. Special 12. Sloan-Barnholtz-Sloan JS, Yu C, Sloan AE, et al.

recognition is appropriate for Ryan Shanley who has pro- A nomogramfor individualized estimation of sur-
vided his statistical wisdom for nearly a decade. vival among patients with brain metastasis. Neuro-
Oncology. 2012;14:910–8.
13. Kondziolka D, Parry PV, Lunsford DL, et al. The
accuracy of predicting survival in individual patients
References with cancer. J Neurosurg. 2014;120:24–30.
14. Sperduto PW, Berkey B, Gaspar LE, Mehta M,

1. Gavrilovic IT, Posner JB. Brain metastases: epi- Curran W. A new prognostic index and compari-
demiology and pathophysiology. J Neuro-Oncol. son to three other indices for patients with brain
2005;75(1):5–14. metastases: an analysis of 1960 patients in the
2. Park DM, Posner JB. Management of intracranial RTOG database. Int J Radiat Oncol Biol Phys.
metastases: history. In: Sawaya R, editor. Intracranial 2008;70:510–4.
metastases: current management strategies. Oxford, 15. Sperduto PW, Chao ST, Sneed PK, et al. Diagnosis-
England: Blackwell Publishing Ltd; 2004. p. 3–19. specific prognostic factors, indexes, and treat-
3. Yamamoto M, Serizawa T, Shuto T, et al. Stereotactic ment outcomes for patients with newly diagnosed
radiosurgery for patients with multiple brain brain metastases: a multi-institutional analysis
metastases (JLGK0901): a multi-institutional of 4,259 patients. Int J Radiat Oncol Biol Phys.
prospective observational study. Lancet Oncol. 2010;77:655–61.
2014;15(4):387–95. 16. Sperduto PW, Kased N, Roberge D, et al. The effect
4. Aoyama H, Shirato H, Tago M, et al. Stereotactic of tumor subtype on survival and the graded prog-
radiosurgery plus whole-brain radiation therapy vs nostic assessment (GPA) for patients with breast
stereotactic radiosurgery alone for treatment of brain cancer and brain metastases. Int J Radiat Oncol
metastases, a randomized controlled trial. JAMA. Biol Phys. 2011;82:2111. https://doi.org/10.1016/j.
2006;295:2483–91. ijrobp.2011.02.027.
5. Chang EL, Wefel JS, Hess KR, et al. Neurocognition 17. Sperduto PW, Kased N, Roberge D, et al. Summary
in patients with brain metastases treated with radio- report on the graded prognostic assessment: an accu-
surgery or radiosurgery plus whole-brain irradia- rate and facile diagnosis-specific tool to estimate sur-
tion: a randomized controlled trial. Lancet Oncol. vival for patients with brain metastases. J Clin Onc.
2009;10:1037–44. 2011;30:419–25.
6. Kocher M, Soffietti R, Abacioglu U, et al. Adjuavant 18. Sperduto PW, Yang TJ, Beal K, et al. The effect of
whole-brain radiotherapy versus observation after gene alterations and tyrosine kinase inhibition on sur-
radiosurgery or surgical resection of one to three cere- vival and cause of death in patients with adenocarci-
bral metastases: resuts of the EORTC 22952-26001 noma of the lung and brain metastases. Int J Radiat
study. J Clin Oncol. 2011;29:134–41. Oncol Biol Phys. 2016;96(2):406–13.
7. Brown PD, Jaeckle K, Ballman KV, et al. Effect of 19. Sperduto PW, Yang TJ, Beal K, et al. Improved
radiosurgery alone vs radiosurgery with whole brain survival and prognostic ability in lung cancer
radiation therapy on cognitive function in patients patients with brain metastases: an update of the
with 1 to 3 brain metastases. JAMA. 2016;4:401–9. graded prognostic assessment for lung cancer using
8. Mulvenna P, Nankivell M, Barton R, et al. molecular markers (Lung-molGPA). JAMA Oncol.
Dexamethasone and supportive care with or without 2017;3(6):827–31.
whole brain radiotherapy in treating patients with 20. Sperduto PW, Jiang W, Brown PD, et al. The prog-
non-small cell lung cancer with brain metastases nostic value of BRAF, cKIT and NRAS mutations in
unsuitable for resection or stereotactic radiotherapy melanoma patients with brain metastases. Int J Radiat
(QUARTZ): results from a phase 3, non-inferiority, Oncol Biol Phys. 2017;98(5):1069–77.
randomized trial. Lancet. 2016;388:2004–14.
21. Sperduto PW, Jiang W, Brown PD, et al. Estimating 23. Sperduto PW, Deegan BJ, Li J, et al. Estimating

survival in melanoma patients with brain metastases: survival for renal cell carcinoma patients with brain
an update of the graded prognostic assessment for mel- metastases: an update of the renal graded prog-
anoma using molecular markers (Melanoma-molGPA). nostic assessment (renal-GPA). Neuro-Oncology.
Int J Radiat Oncol Biol Phys. 2017;99(4):812–6. 2018;20:1652.
22. Sperduto PW, Deegan BJ, Li J, et al. The effect of 24. Sperduto W, King DM, Watanabe Y, et al. Case report
targeted therapies on prognostic factors, patterns of of extended survival and quality of life in a melanoma
care and survival in patients with renal cell carcinoma patient with multiple brain metastases and review
and brain metastases. Int J Radiat Oncol Biol Phys. of literature. Cureus. 2017;9(12):e1947. https://doi.
2018;101(4):845–53. org/10.7759/cureus.1947.
The Role of Advanced Imaging
in the Management of Brain
7
Metastases
Eaton Lin and Gloria C. Chiang
Introduction In the postradiation period, the primary role of

imaging is to differentiate recurrent or progres-
Nearly 20–40% of patients with cancer develop sive metastatic disease in the brain from radiation
brain metastases [1]. Approximately half of brain injury. Radiotherapy leads to breakdown of the
metastases are solitary at initial presentation [2], blood-brain barrier, which can increase contrast
and they can be asymptomatic 25–40% of the enhancement and vasogenic edema, make treated
time [1]. In the pretreatment period, the primary lesions appear larger, and mimic tumor progres-
role of imaging is to detect and diagnose brain sion on conventional magnetic resonance (MR)
metastases, by differentiating them from other imaging. A study of more than 500 metastases
neoplastic lesions, including primary brain found that almost one-third of metastases treated
tumors and nonneoplastic lesions. with radiosurgery showed an apparent increase in
Treatment of brain metastases may include a tumor volume on postcontrast T1-weighted
combination of systemic chemotherapy, surgery, imaging, typically 6 weeks to 15 months after
and radiotherapy. Multiple forms of radiotherapy SRS [10]. Various multimodal imaging tech-
exist, including stereotactic radiosurgery (SRS) niques, including MR perfusion, MR spectros-
and whole-brain radiation [3], but stereotactic copy, diffusion-weighted imaging, and positron
radiosurgery is often favored due to the risk of emission tomography (PET), have been studied
cognitive dysfunction following whole-brain to help differentiate true disease progression
radiation. However, SRS leads to a higher risk of from radiation injury.
radiation injury/necrosis, with a reported relative Recently, there have been exciting advances in
risk of 19 [4]. Within 6–24 months of SRS, radia- the use of systemic immunotherapy to treat brain
tion injury occurs in 5–34% of cases [5–7] and metastases. A multicenter phase 2 study of com-
can even be seen in patients more than 5 years bined nivolumab and ipilimumab in metastatic
after radiotherapy [8]. Approximately 10% have melanoma to the brain found a clinical benefit in
symptomatic radiation injury that may require 57%, with a complete response in 26% of patients
surgery [9]. [11]. Pembrolizumab was studied in a small
cohort of patients with metastatic non-small-cell
lung cancer and was found to show a clinical
E. Lin · G. C. Chiang (*) response in 33% of patients [12]. Patients who
Department of Radiology, Division of receive immunotherapy may develop a transient
Neuroradiology, NewYork-Presbyterian/Weill Cornell increase in the size of contrast-enhancing lesions
Medicine, New York, NY, USA and associated vasogenic edema, which is similar
e-mail: gcc9004@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_7
96 E. Lin and G. C. Chiang
to MR imaging findings of patients who undergo gadolinium-based contrast bolus through brain tis-
radiotherapy. Furthermore, those who receive sue. DCE-MR is widely referred to as “permeabil-
immunotherapy may have even greater risk of ity magnetic resonance imaging (MRI)” because
developing radiation-related changes following the time-intensity curve reflects both tissue perfu-
SRS, with an odds ratio of 2.4 [13]. The immuno- sion and vessel permeability. The volume transfer
therapy response assessment for neuro-oncology coefficient (Ktrans), a measure of gadolinium
(iRANO) guidelines therefore require patients leakage from the intravascular to the extravascular
who have apparent tumor progression on imag- space, is commonly used to reflect permeability.
ing, within 6 months of receiving immunother- The advantages of DCE-MR over DSC-MR are
apy, to undergo repeat imaging in 3 months to higher spatial resolution and decreased sensitivity
confirm tumor progression [14]. Multimodal to susceptibility effects, due to T1-weighted imag-
advanced imaging techniques may play a role in ing rather than T2∗-weighted imaging.
this cohort as well. Arterial spin labeling (ASL) is a third MR per-
The goal of this chapter is to review the fusion technique, which has the advantage of not
advanced imaging techniques used in the man- requiring injection of an exogenous contrast agent.
agement of brain metastases, in both the pretreat- Rather, ASL-MRI uses radiofrequency pulses to
ment and posttreatment periods. “label” endogenous arterial blood water, which
passes into the capillary bed of the brain tissue of
interest. The difference between “labeled” images
MR Perfusion and unlabeled images is used to derive the cerebral
blood flow (CBF). However, ASL-MRI has the
SC-, DCE-, and ASL-MR Perfusion
D main limitation of a low signal-to-noise ratio.
Techniques
Dynamic susceptibility contrast (DSC) MR per- retreatment Imaging for Differential

P
fusion is the most commonly used advanced Diagnosis
imaging technique in daily clinical management
of brain metastases. This technique involves In the pretreatment period, it is important to dif-
injecting a bolus of gadolinium contrast intrave- ferentiate brain metastases from a primary brain
nously, then monitoring the contrast bolus tumor, such as glioblastoma, since the latter
through a region of brain tissue with dynamic would require more aggressive resection. When
T2∗-weighted MR images. Since gadolinium is there is a solitary brain metastasis, this differen-
paramagnetic, the passage of the contrast bolus tiation can be difficult because both metastases
decreases the signal intensity in the region of the and high-grade gliomas can have avid enhance-
brain tissue being imaged, and the change in sig- ment and central necrosis. DSC-MRI has been
nal intensity over time can be represented on a shown to accurately differentiate between the
time-intensity curve. The area under this curve two by using CBV in the peritumoral region; the
(AUC) is used to derive the cerebral blood vol- surrounding T2-hyperintense vasogenic edema
ume (CBV), which is a commonly used measure of metastases has significantly lower relative
of tumor vascularity or angiogenesis. The relative CBV than the T2-hyperintense nonenhancing
CBV can be obtained by comparing the CBV of tumor of high-grade gliomas [15, 16]. This
the region of the presumed metastasis with the method has been shown to have a sensitivity of
CBV from an area of uninvolved brain, often in 77% and specificity of 96% in differentiating
the contralateral hemisphere. metastasis from glioma [17].
T1-weighted dynamic contrast-enhanced Peak height and percentage of signal intensity
(DCE) MR perfusion is similar to DSC-MR, but recovery of the time-intensity curve with DSC-
uses dynamic T1-weighted images to track the MRI have also been reported to differentiate
7 The Role of Advanced Imaging in the Management of Brain Metastases 97
solitary metastasis from high-grade glioma and colon cancer; there was no significant differ-
[18] (Fig. 7.1). In fact, percent signal recovery ence in permeability parameters between glio-
was shown to have the higher accuracy than MR blastoma and hypervascular melanoma metastases
spectroscopy (MRS) in differentiating metasta- [20]. In the pretreatment period, when differential
sis from lymphoma and high-grade glioma with diagnosis is key, DSC-MR is believed to have
an area-under-the-curve of 0.97 [19]. Combining higher diagnostic potential than DCE-MR.
percent signal recovery with MRS further One paper using ASL-MR reported an area-
increased the area-under-the-curve to 0.99 and under-the-curve of 0.84 in differentiating metasta-
increased the specificity from 83% to 100% [19]. ses from gliomas, finding lower CBF in both the
Differentiating high-grade gliomas from brain enhancing portion and the surrounding T2 hyper-
metastases using DCE-MR has also been intensity of metastases compared to high-grade
attempted, although DCE-MR can differentiate gliomas [21]. Although promising, the diagnostic
only glioblastoma from hypovascular brain utility of ASL-MRI is still not as high as DSC-MR
metastases, such as non-small-cell lung, breast, in the pretreatment setting.
Cho
a b d I:28.8
2.0 NAA
I:22.4
Cr
1.5 18.0
1.0
Cr2
I:7.34
0.5
0.0
ppm
4 3 2 1
c PERFUSION e NAA
I:44.5
100
50
-50
4
Cho
-100 I:26.4 Cr
I:23.1
-150
-200
2 Cr2
-250 I:14.7
-300
-350
-400 0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 ppm
1 2 t 4 3 2 1
Fig. 7.1 Dynamic susceptibility contrast (DSC) MR per- to baseline of the signal-intensity curve (c), compatible
fusion and proton MR spectroscopy to diagnose a solitary with a metastasis. Single-voxel proton MR spectroscopy
brain metastasis in a patient with lung adenocarcinoma. A demonstrated an elevated choline-to-NAA ratio (d), com-
solitary-enhancing lesion in the right superior frontal pared to the unaffected side (e), also compatible with a
gyrus (a), seen on postcontrast MRI, demonstrates ele- neoplastic lesion, such as a metastasis
vated cerebral blood volume (b) and an incomplete return
Posttreatment MR Spectroscopy
In the posttreatment period, MR perfusion is used Technique

to differentiate recurrent brain metastasis from
radiation injury, since radiation can break down Proton magnetic resonance spectroscopy (MRS)
the blood-brain barrier and result in an apparent is another noninvasive imaging technique that
increase in contrast enhancement on conven- can be used to differentiate intracranial meta-
tional MRI. A 2009 paper using DSC-MR found static disease from other tumors and radiation-
that patients with recurrent metastases after related changes. MRS involves measuring
gamma knife radiosurgery had a significantly metabolite levels within brain tissue. Typical
lower percentage of signal intensity recovery, metabolites assessed include N-acetylaspartate
higher relative CBV, and higher relative peak (NAA), a marker of neuronal integrity, choline
height compared to an area of radiation injury (Cho), a marker of cell membrane turnover, lac-
[22]. A study that used both DSC-MR and tate (Lac), a marker of anaerobic metabolism,
DCE-MR reported 62% sensitivity and 81% and lipid, a marker of necrosis. Creatine (Cr), an
specificity for Ktrans and 74% sensitivity and energy metabolite, is typically used as an internal
82% specificity for CBV in differentiating recur- control against which other metabolite peaks are
rent metastatic disease and radiation injury [23]. compared.
A prospective study in a combined cohort of glio- Single-voxel MRS methods involve compar-
mas and brain metastases found that DCE-MR ing metabolite levels in a prescribed region of
outperformed 2-(18F) fluoro-2-deoxy-d-glucose interest over the presumed metastasis or peritu-
(FDG) positron emission tomography (PET) in moral region with a region of interest over unin-
detecting recurrent metastasis, with plasma vol- volved brain. Two-dimensional multislice [28]
ume showing the highest area-under-the-curve of and three-dimensional [29] multivoxel or spec-
0.87, resulting in 92% sensitivity and 77% speci- troscopic imaging techniques have also been
ficity [24]. Another study found that ASL-MR described. These multivoxel MRS techniques
was more accurate than both FDG-PET and thal- have better spatial resolution, since smaller vox-
lium single-photon emission computed tomogra- els are used, and cover larger portions of the
phy (SPECT) in differentiating recurrent brain. Since multiple voxels cover the region of
metastasis from radiation injury, with accuracies interest, tumor heterogeneity can also be better
of 87, 73, and 53%, respectively, and specificities assessed. However, multivoxel MRS has longer
of 100, 75, and 63%, respectively [25]. acquisition times, due to the millimolar-range
MR perfusion has also been found to be prog- concentration of these metabolites, and require
nostic of subsequent treatment response. A longi- additional postprocessing time and expertise.
tudinal increase in Ktrans of 15% on DCE-MR Both techniques require careful voxel placement
showed 78% sensitivity and 85% specificity for to avoid contaminating the metabolites of inter-
predicting progression of metastatic disease est with lipid signal from the calvarial marrow,
4 weeks after SRS [26]. A decrease in relative particularly at the skull base, and scalp soft
CBF on ASL-MR after SRS correctly predicted tissues.
tumor response [27].
Taken together, MR perfusion is widely used
in both the pretreatment and posttreatment man- Pretreatment
agement of brain metastases, to both diagnose
brain metastases and differentiate recurrent dis- Like MR perfusion, MRS has been studied to dif-
ease from radiation injury. Of the three methods ferentiate metastases from other brain lesions.
described, DSC-MR is the most widely cited and There is some evidence that metastases show
commonly used technique in daily clinical elevated levels of Cho, Lac, glutamate/glutamine,
practice. and myo-inositol, as well as decreased NAA
[30]. One paper reported that MRS, when added MRS rather than FDG-PET after gamma knife
to conventional MR sequences, increased the rate radiosurgery [39, 40].
of correct diagnosis of intracranial masses, Longitudinal MRS can be used to monitor
including metastases, from 55% to 71% [31]. treatment efficacy. A high Cho peak can be seen
Lipid levels, a marker of necrosis, appear to dif- in viable tumor before treatment, and a decrease
ferentiate between high-grade gliomas and in this Cho peak with an increase in the lipid
metastases [31], with lipid peak-area ratios peak after SRS is suggestive of tumor necrosis
resulting in 80% sensitivity and specificity [32]. after treatment [41]. Although a 2016 meta-anal-
However, another paper reported that MRS could ysis confirmed that Cho-to-NAA and Cho-to-Cr
not accurately distinguish metastasis from high- ratios are useful in differentiating recurrent
grade glioma, reporting an AUC of 60% [33]. metastasis from radiation injury [42], the accura-
Similar to MR perfusion techniques, MRS may cies of these ratios likely decrease when the
be most effective in assessing the peritumoral interrogated lesion includes a combination of
region, with lower Cho-to-Cr ratios seen with tumor and treatment-related changes, as often
metastases compared to high-grade gliomas [15, occurs in clinical practice. Indeed, a paper found
16]. MRS may even help differentiate among dif- that the Cho-to-Cr and lipid/lactate-to-Cho ratios
ferent types of metastases, with higher mobile could accurately distinguish lesions that con-
lipid content in colonic metastases [34] and lower sisted of pure tumor from pure radiation necro-
Cho-to-Cr ratios in non-small-cell lung cancer sis, but specimens with mixed tumor and
compared to breast and melanoma metastases radiation necrosis were more difficult to diag-
[35]. Although MRS is available in most tertiary nose [43].
care centers, the additional scanning and postpro-
cessing time required makes this technique less
commonly used compared to MR perfusion. Diffusion-Weighted Imaging
Nevertheless, it can often confirm findings seen
with MR perfusion or be useful in patients who Technique
cannot receive intravenous contrast, such as preg-
nant patients. Diffusion-weighted magnetic resonance imaging
(DWI) generates image contrast based upon dif-
ferences in Brownian motion, the random ther-
Posttreatment mal movement of molecules in fluid. A DWI
sequence will generate several different images
Since brain metastases treated with radiotherapy and maps; the most relevant to this discussion are
often increase in size on conventional postcon- the apparent diffusion coefficient (ADC) and iso-
trast MR imaging, MRS has also been studied to tropic or trace diffusion maps. Isotropic diffusion
differentiate recurrent metastases and radiation maps are the first-line images used for clinical
injury. Weybright et al. reported that Cho-to- diagnosis, while the ADC images provide a more
NAA, Cho-to-Cr, and NAA-to-Cr ratios on MRS specific assessment of diffusion characteristics
accurately distinguished recurrent metastasis by mathematically removing inherent T2 effects.
from radiation injury, correctly classifying 96% Areas of relatively free water molecular move-
of patients [36], while Elias et al. reported sensi- ment, such as in normal cerebrospinal fluid
tivities of 86 and 93% and specificities of 90% spaces, will have low signal intensity on isotropic
and 70% using Cho-to-NAA and NAA-to-Cr diffusion maps and high signal intensity on ADC,
ratios [37]. In patients with pathologically proven whereas areas of restricted diffusion are hyperin-
recurrent metastases at surgery after gamma tense on isotropic diffusion maps and hypoin-
knife radiosurgery, MRS was found to have a tense on ADC.
positive predictive value of 82% [38]. Chernov A common advanced DWI technique is diffu-
et al. reported higher accuracies using multivoxel sion tensor imaging (DTI), which employs a
greater number of gradient directions to enable DWI assessments of peritumoral edema have
assessment of diffusion directionality; this can demonstrated more promising results as a poten-
convey useful structural information such as tial differentiating factor. A key histological dif-
white matter tract orientation. The two main ference between these two entities is that
parameters derived from DTI data are mean dif- glioblastomas grow in an infiltrative manner and
fusivity (MD), which is analogous to ADC, and invade surrounding tissues, whereas metastases
fractional anisotropy (FA), which is an index of are typically expansive and displace surrounding
diffusion asymmetry or directionality within a tissues. Given that tumor hypercellularity corre-
voxel. lates with diffusion restriction, many authors
Brownian motion is affected by microenviron- have postulated that the area of nonenhancing
mental architecture, temperature, and a variety of peritumoral edema in glioblastomas may dem-
other factors, including several pathologic states. onstrate greater diffusion restriction—due to
The most well-known and common clinical use infiltration with malignant cells—when com-
of DWI is for detection of restricted diffusion in pared with peritumoral edema of metastases,
acute infarction. However, restricted diffusion which is comprised predominantly of vasogenic
can also be seen in a variety of other pathologic edema.
settings, including abscesses, encephalitis, hem- One study supporting this theory demon-
orrhage, status epilepticus, demyelinating dis- strated a gradient of ADC values in the peritu-
ease, epidermoid cysts, toxic/metabolic moral edema of glioblastomas, with progressively
conditions, and hypercellular tumors. The corre- increasing ADC values further from the enhanc-
lation between tumor cellularity and diffusion ing tumor, corresponding to progressively
restriction has generated interest in its utility in decreased extent of nonenhancing infiltrative
tumor diagnosis and management [44–46]. tumor [61]. No such gradient was evident within
Resulting studies have correlated ADC values to peritumoral edema for brain metastases. Several
tumor grade, histology, and treatment response additional studies have demonstrated signifi-
[47–51]. cantly increased peritumoral MD for metastases
compared with glioblastomas [53, 62, 63] and
significantly higher ADC or minimum ADC val-
Pretreatment ues for metastases [16, 60, 64], although a few
studies still showed inconclusive results [56, 65,
Investigations into the diagnostic role of DWI in 66]. A meta-analysis involving 14 studies with
the pretreatment setting have largely focused on 1143 patients demonstrated moderate perfor-
the differentiation of solitary brain metastases mance for DWI and DTI in differentiating metas-
and high-grade gliomas. There is considerable tases from glioblastomas, particularly in analyses
overlap between the signal characteristics and of peritumoral edema [67]. This meta-analysis
enhancement patterns of these two entities, and showed a pooled sensitivity of 72.6% and a
conventional imaging alone is often unreliable. pooled specificity of 77.0% for studies evaluating
Studies investigating DWI characteristics within enhancing tumor, compared with a pooled sensi-
the area of contrast-enhancing tumor have thus tivity of 84.7% and pooled specificity of 84.0%
far been incongruous. Numerous studies have for studies evaluating MD and ADC in perien-
shown discordant FA differences for CNS hancing area.
metastases and glioblastomas [52–57]. DWI is also helpful for differentiating CNS
Similarly, some studies have shown signifi- metastases from many other neoplastic entities.
cantly lower MD and ADC values for metasta- For example, when compared with CNS metasta-
ses compared with glioblastomas [58, 59], ses, ADC values are significantly higher in
whereas others have found no statistically sig- hemangioblastomas [68] and significantly lower
nificant difference [54, 60]. in primary CNS lymphomas [44]. Among CNS
metastases, DWI may have some utility in histo- diagnosis and monitoring. The commonly used
logic differentiation. For example, in lung metas- metric to assess metabolic activity is the stan-
tases, lower ADC values are associated with dardized uptake value (SUV), a ratio of the con-
poorly differentiated adenocarcinomas and with centration of radioactivity in tissue to the injected
small-cell carcinomas [69]. DWI also has predic- dose per kilogram of the patient’s body weight.
tive value for biomarkers in CNS metastases, The SUV within a region-of-interest placed over
with lower ADC values in EGFR mutation- the tumor can then be compared to a reference
positive CNS metastases from lung adenocarci- region, providing a semiquantitative measure of
noma and ER/PR-positive metastases from breast metabolic activity; either normal-appearing white
cancer [70–73]. matter or gray matter is typically used as a refer-
ence region.
Since the cortex and deep gray matter of the
Posttreatment normal brain is highly metabolically active, it can
be difficult to identify tumors, including metasta-
In the posttreatment setting, DWI is helpful for ses, which are located in or near these areas on
differentiating progression and pseudoprogres- FDG-PET, leading to decreased tumor-to-
sion. Pseudoprogression can be seen in up to background ratios. Amino acid tracers, on the other
33% of metastases treated with stereotactic radio- hand, show high tumor-to-background ratios due to
surgery (SRS) [10, 74, 75]. These subacute low uptake of these tracers in normal brain tissue.
posttreatment-related changes may be difficult or Radiolabeled amino acid tracers are taken up by
impossible to differentiate from true progressive membrane-associated carrier proteins, which are
disease based on enhancement patterns and con- upregulated in tumor cells, and then accumulate
ventional imaging. However, an increase in inside the tumor cells. Three amino acid tracers that
ADC—presumably reflecting decreased tumor have been widely studied in the setting of brain
cellularity—when comparing pre- and post-SRS tumors include 11C-methyl-l-methionine (MET),
imaging is suggestive of pseudoprogression O-(2-[18F]fluoroethyl)-l-tyrosine (FET), and
rather than true progressive disease, although 3,4-dihydroxy-6-[18F]-fluoro-l-phenylalanine
interval decreases in ADC are less reliable for (FDOPA). While MET has a short half-life of
predicting true progression [76–80]. 20 minutes, both FET and FDOPA have the added
advantage of having a longer half-life of 110 min-
utes, making it more widely available to clinical
Positron Emission Tomography practices without their own cyclotron.
Technique and Radiotracers
Pretreatment
FDG-PET has been widely used in the imaging
of metastatic disease. As a glucose analog, FDG Unlike MR perfusion and MRS, FDG-PET does
is actively transported into the cell, phosphory- not play a significant role in the initial diagnosis
lated by hexokinase in the glycolytic pathway, of brain metastases because the background of
and then trapped within the cell. The FDG that is normal high cortical metabolism decreases its
trapped can then be imaged, and the uptake on sensitivity, particularly for lesions less than
PET serves as a proxy for glucose utilization and 1 cm in size [81–83]. Reported sensitivities of
metabolic activity. Since brain metastases are FDG-PET for detecting brain metastases range
typically more metabolically active than nontu- from 27% to 50% [84–86]. Furthermore, some
moral brain regions, they demonstrate greater metastases may be hypometabolic, such as
FDG uptake than normal surrounding brain mucinous adenocarcinoma and renal cell carci-
regions, providing a means for noninvasive tumor noma. These may be better detected with amino
acid tracers. One study using MET-PET reported study of almost 400 patients using FET-PET
that 80% of the brain metastases that did not found no significant difference between gliomas
show increased uptake on FDG-PET did show and metastases [89].
increased uptake on MET-PET [87]. MET-PET
may also be used for SRS treatment planning
because it more accurately delineates the mar- Posttreatment
gins of the metastases, resulting in smaller irra-
diation volumes and longer median survival In the posttreatment period, FDG-PET is widely
time [88]. used to differentiate recurrent metastasis and
PET also has a limited role in differentiating radiation injury, with a sensitivity and specific-
brain metastases from high-grade gliomas. A ity of 71 and 80%, respectively [90] (Fig. 7.2).
a b c
d e f
Fig. 7.2 A patient with metastatic breast cancer, who had sity seen on the b1000 image (c) and hypointensity on the
undergone multiple rounds of systemic chemotherapy and ADC map (f), likely reflecting hypercellularity. The
stereotactic radiosurgery to brain metastases, was sent for lesions demonstrated susceptibility hypointensity (d),
evaluation after presenting with aphasia. The enhancing with the lesions predominantly showing hypointensity on
lesions seen on postcontrast MRI (a) demonstrated quantitative susceptibility mapping (QSM) (e), suggest of
marked uptake on FDG-PET (b), which were found to mineralization posttreatment, but two foci of QSM hyper-
represent viable metastatic disease on biopsy. There is intensity in the left temporal lobe (e), suggestive of intral-
also associated reduced diffusion, with signal hyperinten- esional hemorrhage
A decrease in SUV on FDG-PET can also be The main practical limitations of PET in the
used to monitor effectiveness of a drug against management of brain metastases are the added
metastatic disease in clinical trials [91]. Dual- cost, time, and radiation exposure. FDG uptake
phase FDG-PET, using early and delayed imag- can also be difficult to discern adjacent to areas
ing, may increase sensitivity and specificity to of normal metabolically active cortex and deep
95 and 100%, respectively, with an overall accu- gray matter. Overall, it appears to be less accurate
racy of 96% [92]. Dual-phase FDG-PET takes than MR perfusion in the posttreatment setting
advantage of the different time-activity curves [24]. The amino acid PET tracers, though higher
between tumor, normal brain tissue and postin accuracy, remain investigational and require
treatment inflammatory cells. An increase in the access to radiochemistry laboratories and a
maximum SUV of the lesion relative to normal cyclotron for 11C-compounds.
gray matter suggests tumor rather than inflam- The use of MRI coregistration with FDG-PET
matory change. The drawback is that delayed appears to increase sensitivity for detecting recur-
imaging occurs at least 2 hours after initial rent metastasis after SRS from 71% to 86%, but
imaging, which is difficult to maintain in a busy without a significant increase in specificity (80%)
clinical practice. [90]. The advent of integrated PET-MRI scanners
Amino acid tracers, though used in clinical tri- addresses some of the aforementioned issues by
als and in research studies, are not yet approved reducing radiation exposure (compared to a
by the Food and Drug Administration (FDA) for PET-CT), reducing imaging time (PET and MRI
clinical use in the United States. Nevertheless, simultaneously acquired), and improving anatom-
studies have reported high accuracy in differenti- ical resolution for precise localization of metabolic
ating recurrent metastasis from radiation injury uptake. However, specificity remains lacking.
with MET-PET, FET-PET, and FDOPA-PET
[93–97]. Combined imaging with MET-PET and
FET-PET showed sensitivity and specificity of 91 Susceptibility-Weighted Imaging
and 100% [95], while FDOPA-PET showed sen-
sitivities and specificities of 81–90% and 84–92% Technique
[96, 97], outperforming MR perfusion, which has
a specificity of only 68% [96]. Susceptibility-weighted imaging (SWI) is a high-
Dynamic FET-PET, which allows assessment resolution, velocity-corrected gradient echo MRI
of parameters such as time-to-peak and slope of sequence. A key feature of this imaging sequence
the time-activity curves, also found high diagnosis that it does not refocus spins dephased by mag-
tic accuracies in the range of 80–90% [98–100], netic field inhomogeneities. Compounds that
but requires longer acquisition times of have ferromagnetic, paramagnetic, and diamag-
40–50 minutes. In the setting of checkpoint netic properties can all interact with the local
inhibitor therapy for melanoma metastases, magnetic field; these field distortions appear
dynamic FET has been shown to be particularly hypointense on SWI.
useful [101]. With the advent of machine-learning From a clinical standpoint, SWI is markedly
algorithms, a 2018 paper found that FET-PET sensitive for hemorrhage due to the paramagnetic
textural features had slightly higher accuracy properties of hemoglobin derivatives such as
than textural analysis of contrast-enhanced MRI deoxyhemoglobin, methemoglobin, and hemosid-
in differentiating radiation injury and recurrent erin. SWI is significantly more sensitive for hem-
brain metastases, 83% versus 81% [102]. This orrhage than other MR sequences such as T2∗- or
paper found that FET-PET was more sensitive at T1-weighted imaging [103]. SWI is also extremely
88%, and MRI was more specific at 90%, but sensitive for dystrophic calcifications and bone
combined accuracy reached 89%, with 85% sen- minerals due to their diamagnetic properties.
sitivity and 96% specificity. While computed tomography is sometimes used
to further characterize areas of susceptibility iden- fied only on SWI. Accordingly, SWI still remains
tified on SWI, differences in local phase altera- complementary in identification of CNS metasta-
tions between paramagnetic and diamagnetic ses, in addition to its role in characterizing pres-
compounds will often allow for distinction ence or absence of associated hemorrhage. SWI
between calcium and hemoglobin derivates on the is also useful for detection of metastases in
filtered phase component of SWI. Newer methods patients with impaired renal function, contrast
of quantifying these magnetic field distortions— allergies, or other contraindications for contrast-
such as quantitative susceptibility mapping—also enhanced imaging.
offer more reliable differentiation between cal-
cium and hemorrhage [104, 105].
Posttreatment
Pretreatment The role of SWI in posttreatment imaging is also

largely confined to hemorrhage detection. In par-
SWI has a mostly supplementary role in the ticular, radiation-induced cavernous malforma-
imaging of CNS metastases, primarily being used tions and cerebral microbleeds of other etiologies
in the detection of lesional hemorrhage. are relatively common following radiation ther-
Classically, CNS metastases from melanoma, apy to the brain [112, 113], and SWI is the most
renal cell carcinoma, choriocarcinoma, and thy- sensitive imaging sequence for identifying these
roid carcinoma have a high propensity for hemor- microbleeds [114, 115]. Radiation-induced cav-
rhage, although metastases from many other ernous malformations have slightly higher pro-
primaries may also potentially bleed, particularly spective hemorrhage risk when compared with
metastases from lung and breast carcinoma due nonradiation cavernous malformations [116],
to their frequency. Presence or absence of hemor- and cerebral microbleeds carry important
rhage is an important feature in imaging charac- prognostic implications such as an increased risk
terization of CNS metastases, as the presence of for cognitive dysfunction [117]. Accordingly,
hemorrhage portends poorer prognosis [106]. their identification remains an important part of
Given the exquisite sensitivity of SWI for posttreatment imaging characterization. While
hemorrhage, some authors—particularly of ear- stereotactic radiosurgery presents no significant
lier studies on SWI—have hypothesized that additional risk of lesional hemorrhage in CNS
early hemorrhagic metastases or micrometasta- metastases [118, 119], smaller series have specu-
ses may be first detectable or better imaged on lated that radiosurgery may induce breakdown of
SWI [107, 108]. However, more recent studies fragile tumor vessels and contribute to rare hem-
following nonenhancing foci of susceptibility in orrhagic posttreatment complications [120].
patients with hemorrhagic brain metastases have Regardless of causal relationships, lesional hem-
not shown these foci to evolve into true meta- orrhage remains a clinically significant finding in
static lesions [109], and incidence of hemorrhage the posttreatment period best assessed with SWI.
has been shown to be significantly lower or
absent in micrometastases [110]. The sensitivity
of SWI for hemorrhagic metastases also does not Future Directions
rival that of contrast enhanced T1 imaging. A
study comparing the sensitivity of SWI with The field of neuroimaging continues to undergo
other MR imaging sequences for melanoma rapid development, and many new techniques offer
metastases demonstrated much higher sensitivity the potential for more detailed characterization of
of postcontrast T1 imaging (99.7%) compared the molecular, cellular, and structural components
with SWI (61.0%) [111]. Nonetheless, 2% of the of brain metastases. The techniques discussed in
lesions in this study were more conspicuous on this section are not yet under widespread clinical
SWI, and 1 out of the 712 lesions was first identi- use, and further research is necessary. However,
these novel imaging techniques hold promise for upregulation during early stages of metastatic
future improvements in diagnosis and characteriza- lesions [134–137]. The administration of anti-
tion of brain metastases. VCAM-1 antibodies conjugated to SPIOs enables
highly sensitive in vivo imaging of areas of
VCAM-1 upregulation [138]. This technique has
Molecular/Cellular MRI demonstrated VCAM-1 upregulation not only in
established gadolinium-enhancing brain metasta-
MR imaging has traditionally been limited to ses, but also in nonenhancing micrometastases,
macroscopic anatomical evaluation, which reveals with a proportional increase in VCAM-1 expres-
manifestations seen in later stages of disease. For sion upon tumor progression [139]. While further
example, contrast-enhanced MRI is the gold stan- research is needed, it is estimated that this tech-
dard for imaging brain metastases, with contrast nique can already detect brain metastases approx-
enhancement dependent upon the blood- brain imately 300 μm in diameter, allowing diagnosis at
barrier breakdown occurring in later-stage lesions. substantially earlier stages than with gadolinium-
However, newer techniques enable in vivo imag- enhanced MRI and opening doors to therapeutic
ing of molecular and cellular manifestations that techniques targeting earlier stage disease.
occur earlier in the disease process, such as early
inflammation and angiogenesis.
Cellular visualization on MR image requires Postcontrast T1 Mapping
cell labeling with a detectable agent. Currently,
the most commonly used agents are superpara- The distinction between radiation necrosis and
magnetic iron oxide (SPIO) nanoparticles, which recurrent tumor in the postradiation setting
are generally composed of an iron core, a poly- remains a common diagnostic dilemma. While
mer coating, and functional moieties. Iron oxide MR perfusion and FDG-PET can be useful dif-
nanoparticles can be administered intravenously, ferentiating tools, there are limitations to this
with eventual uptake by phagocytic cells allow- techniques, including low spatial resolution, deg-
ing for imaging of inflammation. Alternatively, radation by susceptibility artifacts, need for high-
iron oxide nanoparticles can also be used to label velocity bolus injections, and interinstitutional
cells in culture prior to injection or transplanta- variations in technique and analysis. Based on
tion, allowing for cell tracking. A variety of cell early research, postcontrast T1 mapping—a tech-
types have been labeled and tracked using this nique free of these limitations—shows potential
technique, including stem cells [121], dendritic as an alternative for differentiating radionecrosis
cells [122–124], T-lymphocytes [125], and can- and recurrent tumor.
cer cells [126–128]. Subsequently, imaging with T1 mapping enables quantitative evaluation of
an MR pulse sequence sensitive to iron—such as T1 relaxation times of tissues. Postcontrast T1
T2∗ imaging—enables visualization of iron mapping can be used to exploit differences in vas-
nanoparticle locations. cularity and contrast enhancement kinetics
For imaging brain metastases, one promising between radiation necrosis and recurrent tumor.
avenue within molecular and cellular MRI has Vasculature in both tumor and radionecrosis
been based on the imaging of endothelial vascular exhibits increased permeability, contributing to
cell adhesion molecule-1 (VCAM-1). The vascu- contrast enhancement on gadolinium-enhanced
lar endothelium of the brain is profoundly reactive MRI. However, tumor microvasculature is nor-
to pathological stimuli, with surface molecules mally characterized by abundant neoangiogenesis
such as VCAM-1 mediating the adhesion and with intact vascular lumens [140–143], contribut-
migration of lymphocytes in areas of inflamma- ing to brisk early contrast accumulation and rapid
tion [129–131]. Metastatic lesions are closely clearance of contrast [144], whereas areas of
associated with existing cerebral vasculature radionecrosis have damaged vascular lumens and
[132, 133], and there is evidence of VCAM-1 absent neovascularization [140], contributing to
slow accumulation and slow clearance of contrast portation of 18F-ML-10 into apoptotic cells, but
[145]. Studies of delayed contrast enhanced MRIs not into viable or necrotic cells [157]. The detec-
(performed greater than 1 hour after contrast tion of apoptosis posttreatment provides early
injection) have shown contrast clearance in areas evidence of response to therapy, at a time when
of tumor and contrast accumulation in areas of FDG-PET results are often confounded by post-
enhancing nontumor tissue [146]. In a study of treatment inflammation and persistent tumor cell
postcontrast T1 mapping in patients with brain uptake. In patients undergoing whole brain radia-
metastases postradiosurgery, T1 values were mea- tion therapy, significant correlation has been
sured for enhancing lesions 5 and 60 minutes after demonstrated between early posttreatment 18F-
contrast administration. The authors demonstrated ML- 10 findings and subsequent anatomic
that the difference between these two values could changes seen on MR image 6–8 weeks after com-
distinguish between recurrent tumor and radione- pletion of therapy [158].
crosis, with an AUC of 0.97, sensitivity of 81.5%, There are many additional experimental PET
and specificity of 96.5% [147]. agents with more niche roles; for example,
68
Ga-DOTATATE has an affinity for somatostatin
receptors, which are highly expressed in neuro-
Novel PET Agents endocrine neoplasms and meningiomas, among
other tumors. While research with this radiotracer
While FDG is the principal oncologic PET imag- for brain metastases remains minimal, it has been
ing agent, new radiotracers continue to expand suggested to have greater sensitivity than tradi-
the diagnostic capabilities of PET for imaging tional imaging methods for detecting metastases
brain metastases. Amino acid radiotracers are the from neuroendocrine tumors and medullary thy-
most extensively studied alternative agents (dis- roid carcinoma [159, 160].
cussed more thoroughly earlier in the chapter). In summary, imaging currently plays a major
F-18 fluorothymidine (FLT) is another PET agent role in the management of brain metastases, both
which—like amino acid tracers—demonstrates pre- and posttreatment. As more specific MR and
lower uptake in normal cerebral parenchyma PET techniques and tracers continue to be devel-
compared with FDG, allowing for better lesion- oped, multimodal imaging will serve to detect
to-background contrast [148, 149]. FLT is a thy- and monitor the various molecular processes
midine analog retained in proliferating tissue and underlying intracranial metastatic disease
tumor via thymidine salvage pathways in cellular development.
proliferation [150]. FLT-PET has demonstrated
promising results for assessment of treatment
response in malignant gliomas and extracranial References
melanoma [149, 151–154]. There has been less
research using FLT-PET for monitoring brain 1. Soffietti R, Cornu P, Delattre JY, Grant R, Graus F,
metastases, but preliminary studies have shown Grisold W, et al. EFNS guidelines on diagnosis and
treatment of brain metastases: report of an EFNS
potential in FLT-PET response assessment for
Task Force. Eur J Neurol. 2006;13:674–81.
breast and melanoma brain metastases [155, 2. Delattre JY, Krol G, Thaler HT, Posner
156]. JB. Distribution of brain metastases. Arch Neurol.
Another promising PET agent is 2-(5-fluoro- 1988;45:741–4.
3. Arvold ND, et al. Updates in the management of
pentyl)-2-methyl-malonic acid ( F-ML-10)—a
18
brain metastases. Neuro Oncol. 2016;18:1043–65.
PET probe designed for selective detection of 4. Lamba N, Muskens IS, DiRisio AC, Meijer L,
apoptosis. While apoptosis is easily detected Briceno V, Edrees H, et al. Stereotactic radiosurgery
in vitro, in vivo assessment is challenging. 18F- versus whole-brain radiotherapy after intracranial
metastasis resection: a systematic review and meta-
ML-10 targets a complex set of cell membrane
analysis. Radiat Oncol. 2017;12:106.
alterations that occur during the apoptotic pro- 5. Kohutek ZA, Yamada Y, Chan TA, Brennan CW,
cess, resulting in selective transmembrane trans- Tabar V, Gutin PH, et al. Long-term risk of radio-
necrosis and imaging changes after stereotactic analysis and perfusion-weighted MRI. Clin Radiol.
radiosurgery for brain metastases. J Neurooncol. 2010;65:15–20.
2015;125:149–56. 18. Cha S, Lupo JM, Chen MH, Lamborn KR, McDermott
6. Sneed PK, Mendez J, Vemer-van den Hoek JG, MW, Berger MS, et al. Differentiation of glioblas-
Seymour ZA, Ma L, Molinaro AM, et al. Adverse toma multiforme and single brain metastasis by
radiation effect after stereotactic radiosurgery for peak height and percentage of signal intensity recov-
brain metastases: incidence, time course, and risk ery derived from dynamic susceptibility-weighted
factors. J Neurosurg. 2015;123:373–86. contrast-
enhanced perfusion MR imaging. AJNR
7. Schuttrumpf LH, Niyazi M, Nachbichler SB, Am J Neuroradiol. 2007;28:1078–84.
Manapov F, Jansen N, Siefert A, et al. Prognostic 19. Vallee A, Guillevin C, Wager M, Delwail V, Guillevin
factors for survival and radiation necrosis after R, Vallee J-N. Added value of spectroscopy to perfu-
stereotactic radiosurgery alone or in combination sion MRI in the differential diagnostic performance
with whole brain radiation therapy for 1–3 cerebral of common malignant brain tumors. AJNR Am J
metastases. Radiat Oncol. 2014;9:105. Neuroradiol. 2018;39:1423–31.
8. Fujimoto D, von Eyben R, Gibbs IC, Chang SD, Li 20. Jung BC, Arevalo-Perez J, Lyo JK, Holodny AI,
G, Harsha GR, et al. Imaging changes over 18 months Karimi S, Young RJ, et al. Comparison of glio-
following stereotactic radiosurgery for brain metas- blastomas and brain metastases using dynamic
tases: both late radiation necrosis and tumor progres- contrast-enhanced perfusion MRI. J Neuroimaging.
sion can occur. J Neurooncol. 2018;136:207–12. 2016;26:240–6.
9. Dequesada IM, Quisling RG, Yachnis A, Friedman 21. Sunwoo L, Yun TJ, You SH, Yoo RE, Kang KM,
WA. Can standard magnetic resonance imaging Choi SH, et al. Differentiation of glioblastoma from
reliably distinguish recurrent tumor from radiation brain metastasis: qualitative and quantitative analy-
necrosis after radiosurgery for brain metastases? sis using arterial spin labeling MR imaging. PLoS
A radiographic-pathological study. Neurosurgery. One. 2016;11:e0166662.
2008;63:898–903. 22. Barajas RF, Chang JS, Sneed PK, Segal MR,
10. Patel TR, McHugh BJ, Bi WL, Minja FJ, Knisely JP, McDermott MW, Cha S. Distinguishing recurrent
Chiang VL. A comprehensive review of MR imaging intra-axial metastatic tumor from radiation necro-
changes following radiosurgery to 500 brain metas- sis following gamma knife radiosurgery using
tases. AJNR Am J Neuroradiol. 2011;32:1885–92. dynamic susceptibility-weighted contrast-enhanced
11. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi perfusion MR imaging. AJNR Am J Neuroradiol.
FS, Moschos SJ, et al. Combined nivolumab and 2009;30:367–72.
ipilimumab in melanoma metastatic to the brain. N 23. Jakubovic R, Sahgal A, Soliman H, Milwid R,
Engl J Med. 2018;379:722–30. Zhang L, Eilaghi A, et al. Magnetic resonance
12. Goldberg SB, Gettinger SN, Mahajan A, Chiang imaging- based tumour perfusion parameters are
AC, Herbst RS, Sznol M, et al. Pembrolizumab for biomarkers predicting response after radiation
patients with melanoma or non-small-cell lung can- to brain metastases. Clin Oncol (R Coll Radiol).
cer and untreated brain metastases: early analysis of 2014;26:704–12.
a non-randomised, open-label, phase 2 trial. Lancet 24. Hatzoglou V, Yang TJ, Omura A, et al. A prospec-
Oncol. 2016;17:976–83. tive trial of dynamic contrast-enhanced MRI
13. Colaco RJ, Martin P, Kluger HM, Yu JB, Chiang perfusion and fluorine-18 FDG PET-CT in dif-
VL. Does immunotherapy increase the rate of radia- ferentiating brain tumor progression from radia-
tion necrosis after radiosurgical treatment of brain tion injury after cranial irradiation. Neuro Oncol.
metastases? J Neurosurg. 2016;125:17–23. 2016;18:873–80.
14. Okada H, Weller M, Huang R, Finocchiaro G, 25. Lai G, Mahadevan A, Hackney D, Warnke PC,
Gilbert MR, Wick W, et al. Immunotherapy response Nigim F, Kasper E, et al. Diagnostic accuracy of
assessment in neuro-oncology: a report of the RANO PET, SPECT, and arterial spin-labeling in differen-
working group. Lancet Oncol. 2015;16:e534–42. tiating tumor recurrence from necrosis in cerebral
15. Law M, Cha S, Knopp EA, Johnson G, Arnett J, Litt metastasis after stereotactic radiosurgery. AJNR Am
AW. High-grade gliomas and solitary metastases: dif- J Neuroradiol. 2015;36:2250–5.
ferentiation by using perfusion and proton spectro- 26. Almeida-Freitas DB, Pinho MC, Otaduy MC, et al.
scopic MR imaging. Radiology. 2002;222:715–21. Assessment of irradiated brain metastases using
16. Chiang IC, Kuo YT, Lu CY, Yeung KW, Lin WC, dynamic contrast-enhanced magnetic resonance
Sheu FO, et al. Distinction between high-grade glio- imaging. Neuroradiology. 2014;56:437–43.
mas and solitary metastases using peritumoral 3-T 27. Weber MA, Thilmann C, Lichy MP, Gunther M,
magnetic resonance spectroscopy, diffusion, and per- Delorme S, Zuna I, et al. Assessment of irradiated
fusion imagings. Neuroradiology. 2004;46:619–27. brain metastases by means of arterial spin-labeling
17. Hakyemez B, Erdogan C, Gokalp G, Dusak A, and dynamic susceptibility-weighted contrast-
Parlak M. Solitary metastases and high-grade glio- enhanced perfusion MRI: initial results. Invest
mas: radiological differentiation by morphometric Radiol. 2004;39:277–87.
28. Balmaceda C, Critchell D, Mao X, et al. Multisection tance of multi-voxel proton MRS. Minim Invasive
1
H magnetic resonance spectroscopic imaging Neurosurg. 2005;48:228–34.
assessment of glioma response to chemotherapy. J 41. Kimura T, Sako K, Tanaka K, Gotoh T, Yoshida H,
Neurooncol. 2006;76:185–91. Aburano T, et al. Evaluation of the response of meta-
29. Vigneron D, Bollen A, McDermott M, et al. Three- static brain tumors to stereotactic radiosurgery by
dimensional magnetic resonance spectroscopic proton magnetic resonance spectroscopy, 201TlCl
imaging of histologically confirmed brain tumors. single-photon emission computerized tomogra-
Magn Reson Imaging. 2001;19:89–101. phy, and gadolinium-enhanced magnetic resonance
30. Fan G, Sun B, Wu Z, Guo Q, Guo Y. In vivo single- imaging. J Neurosurg. 2004;100:835–41.
voxel proton MR spectroscopy in the differentiation 42. Chuang MT, Liu YS, Tsai YS, Chen YC, Wang
of high-grade gliomas and solitary metastases. Clin CK. Differentiating radiation-induced necrosis
Radiol. 2004;59:77–85. from recurrent brain tumor using MR perfusion
31. Moller-Hartmann W, Herminghaus S, Krings T, and spectroscopy: a meta-analysis. PLoS One.
Marquardt G, Lanfermann H, Pilatus U, et al. 2016;11:e0141438.
Clinical application of proton magnetic resonance 43. Rock JP, Hearshen D, Scarpace L, Croteau D,
spectroscopy in the diagnosis of intracranial mass Gutierrez J, Fisher JL, et al. Correlations between
lesions. Neuroradiology. 2002;44:371–81. magnetic resonance spectroscopy and image-guided
32. Opstad KS, Murphy MM, Wilkins PR, Bell BA, histopathology, with special attention to radiation
Griffiths JR, Howe FA. Differentiation of metas- necrosis. Neurosurgery. 2002;51:912–9.
tases from high-grade gliomas using short echo 44. Yamasaki F, Kurisu K, Satoh K, Arita K, Sugiyama
time 1H spectroscopy. J Magn Reson Imaging. K, Ohtaki M, et al. Apparent diffusion coefficient
2004;20:187–92. of human brain tumors at MR imaging. Radiology.
33. Devos A, Lukas L, Suykens JA, Vanhamme L, Tate 2005;235:985–91.
AR, Howe FA, et al. Classification of brain tumours 45. Kinuko K, Inoue Y, Nakayama K, Shakudo M,
using short echo time 1H MR spectra. J Magn Reson. Morino M, Ohata K, et al. The role of diffusion-
2004;170:164–75. weighted imaging in patients with brain tumors.
34. Chernov MF, Ono Y, Kubo O, Hori T. Comparison of AJNR Am J Neuroradiol. 2001;22:1081–8.
1H-MRS detected metabolic characteristics in single 46. Sugahara T, Korogi Y, Kochi M, Ikushima I,
metastatic brain tumors of different origin. Brain Shigematu Y, Hirai T, et al. Usefulness of diffusion-
Tumor Pathol. 2006;23:35–40. weighted MRI with echo-planar technique in the
35. Huang BY, Kwock L, Castillo M, Smith evaluation of cellularity in gliomas. J Magn Reson
JK. Association of choline levels and tumor perfu- Imaging. 1999;9:53–60.
sion in brain metastases assessed with proton MR 47. Wieduwilt MJ, Valles F, Issa S, Behler CM, Hwang
spectroscopy and dynamic susceptibility contrast- J, McDermott M, et al. Immunochemotherapy with
enhanced perfusion weighted MRI. Technol Cancer intensive consolidation for primary CNS lymphoma:
Res Treat. 2010;9:327–37. a pilot study and prognostic assessment by diffusion-
36. Weybright P, Sundgren PC, Maly P, Hassan DG, weighted MRI. Cancer Res Treat. 2012;18:1146–55.
Nan B, Rohrer S, et al. Differentiation between brain 48. Lee EJ, Lee SK, Agid R, Bae JM, Keller A,
tumor recurrence and radiation injury using MR Terbrugge K. Preoperative grading of presumptive
spectroscopy. Am J Roentgenol. 2005;185:1471–6. low-grade astrocytomas on MR imaging: diagnostic
37. Elias AE, Carlos RC, Smith EA, Frechtling D, value of minimum apparent diffusion coefficient.
George B, Maly P, et al. MR spectroscopy using nor- AJNR Am J Neuroradiol. 2008;29:1872–7.
malized and non-normalized metabolite ratios for 49. Guo AC, Cummings TJ, Dash RC, Provenzale
differentiating recurrent brain tumor from radiation JM. Lymphomas and high-grade astrocytomas: com-
injury. Acad Radiol. 2011;18:1101–8. parison of water diffusibility and histologic charac-
38. Truong MT, St Clair EG, Donahue BR, Rush SC, teristics. Radiology. 2002;224:177–83.
Miller DC, Formenti SC, et al. Results of surgi- 50. Lee KC, Moffat BA, Schott AF, Layman R,
cal resection for progression of brain metastases Ellingworth S, Juliar R, et al. Prospective early
previously treated by gamma knife radiosurgery. response imaging biomarker for neoadjuvant
Neurosurgery. 2006;59:86–97. breast cancer chemotherapy. Cancer Res Treat.
39. Chernov MF, Hayashi M, Izawa M, Usukura M, 2007;13:443–50.
Yoshida S, Ono Y, et al. Multivoxel proton MRS 51. Huang WY, Wen JB, Wu G, Yin B, Li JJ, Geng
for differentiation of radiation-induced necrosis DY. Diffusion-weighted imaging of metastatic
and tumor recurrence after gamma knife radio- brain tumors: comparison with histologic type
surgery for brain metastases. Brain Tumor Pathol. and tumor cellularity. AJNR Am J Neuroradiol.
2006;23:19–27. 2006;37:2010–8.
40. Chernov M, Hayashi M, Izawa M, Ochiai T, Usukura 52. Wang W, Steward CE, Desmond PM. Diffusion ten-
M, Abe K, et al. Differentiation of the radiation- sor imaging in glioblastoma multiforme and brain
induced necrosis and tumor recurrence after gamma metastases: the role of p, q, L, and fractional anisot-
knife radiosurgery for brain metastases: impor- ropy. AJNR Am J Neuroradiol. 2009;30:203–8.
53. Bauer AH, Erly W, Moser FG, Maya M, Nael toma multiforme from metastatic brain tumor using
K. Differentiation of solitary brain metastasis from proton magnetic resonance spectroscopy, diffu-
glioblastoma multiforme: a predictive multiparamet- sion and perfusion metrics at 3 T. Cancer Imaging.
ric approach using combined MR diffusion and per- 2012;12:423–36.
fusion. Neuroradiology. 2015;57:697–703. 67. Suh CH, Kim HS, Jung SC, Kim SJ. Diffusion-
54. Holly KS, Barker BJ, Murcia D, Bennett R, Kalakoti weighted imaging and diffusion tensor imaging for
P, Ledbetter C. High-grade gliomas exhibit higher differentiating high-grade glioma from solitary brain
peritumoral fractional anisotropy and lower mean metastasis: a systematic review and meta-analysis.
diffusivity than intracranial metastases. Front Surg. AJNR Am J Neuroradiol. 2018;39:1208–14.
2017;4:18. 68. She D, Yang X, Xing Z, Cao D. Differentiating
55. Wang S, Kim S, Chawla S, Wolf RL, Zhang WG, hemangioblastomas from brain metastases using
O’Rourke DM, et al. Differentiation between glio- diffusion-weighted imaging and dynamic susceptibil-
blastomas and solitary brain metastases using diffu- ity contrast-enhanced perfusion-weighted MR imag-
sion tensor imaging. Neuroimage. 2009;44:653–60. ing. AJNR Am J Neuroradiol. 2016;37:1844–50.
56. Wang S, Kim SJ, Poptani H, Woo JH, Mohan S, Jin 69. Hayashida Y, Hirai T, Morishita S, Kitajima M,
R, et al. Diagnostic utility of diffusion tensor imaging Murakami R, Korogi Y, et al. Diffusion-weighted
in differentiating glioblastomas from brain metasta- imaging of metastatic brain tumors: comparison
ses. AJNR Am J Neuroradiol. 2014;35:928–34. with histologic type and tumor cellularity. AJNR Am
57. Bette S, Huber T, Wiestler B, Beockh-Behrens J Neuroradiol. 2006;27:1419–25.
T, Gempt J, Ringel F, et al. Analysis of fractional 70. Jung WS, Park CH, Hong CK, Suh SH, Ahn
anisotropy facilitates differentiation of glioblas- SJ. Diffusion-weighted imaging of brain metastasis
toma and brain metastases in a clinical setting. Eur J from lung cancer: correlation of MRI parameters
Radiol. 2016;85:2181–7. with the histologic type and gene mutation status.
58. Byrnes TJ, Barrick TR, Bell BA, Clark CA. Diffusion AJNR Am J Neuroradiol. 2018;39:273–9.
tensor imaging discriminates between glioblastoma 71. Ahn SJ, Park M, Bang S, Cho E, Ahn SG, Suh
and cerebral metastases in vivo. NMR Biomed. SH, et al. Apparent diffusion coefficient histo-
2011;24:54–60. gram in breast cancer brain metastases may predict
59. Caravan I, Ciortea CA, Contis A, Lebovici their biological subtype and progression. Sci Rep.
A. Diagnostic value of apparent diffusion coefficient 2018;8:12767.
in differentiating between high-grade gliomas and 72. Kim SH, Cha ES, Kim HS, Kang BJ, Choi JJ, Jung
brain metastases. Acta Radiol. 2018;59:599–605. JH, et al. Diffusion-weighted imaging of breast
60. Lee EJ, TerBrugge K, Mikulis D, Choi DS, Bae JM, cancer: correlation of the apparent diffusion coef-
Lee SK. Diagnostic value of peritumoral minimum ficient value with prognostic factors. J Magn Reson
apparent diffusion coefficient for differentiation of Imaging. 2009;30:615–20.
glioblastoma multiforme from solitary metastatic 73. Martincich L, Deantoni V, Bertotto I, Redana S,
lesions. AJR Am J Roentgenol. 2011;196:71–6. Kubatzki F, Sarotto I, et al. Correlations between
61. Lemercier P, Paz Maya S, Patrie JT, Flors L, Leiva- diffusion-weighted imaging and breast cancer bio-
Salinas C. Gradient of apparent diffusion coefficient markers. Eur Radiol. 2012;22:1519–28.
values in peritumoral edema helps in differentiation 74. Tsao M, Xu W, Sahgal A. A meta-analysis evaluating
of glioblastoma from solitary metastatic lesions. stereotactic radiosurgery, whole-brain radiotherapy,
AJR Am J Roentgenol. 2014;203:163–9. or both for patients presenting with a limited number
62. Lu S, Ahn D, Johnson G, Cha S. Peritumoral dif- of brain metastases. Cancer. 2012;118:2486–93.
fusion tensor imaging of high-grade gliomas and 75. Yamamoto M, Serizawa T, Shuto T, Akabane A,
metastatic brain tumors. AJNR Am J Neuroradiol. Higuchi Y, Kawagishi J, et al. Stereotactic radio-
2003;24:937–41. surgery for patients with multiple brain metasta-
63. Lu S, Ahn D, Johnson G, Law M, Zagzag D, ses (JLGK0901): a multi-institutional prospective
Grossman RI. Diffusion-tensor MR imaging of observational study. Lancet Oncol. 2014;15:387–95.
intracranial neoplasia and associated peritumoral 76. Farjam R, Tsien CI, Feng FY, Gomez-Hassan D,
edema: introduction of the tumor infiltration index. Hayman JA, Lawrence TS, et al. Investigation of the
Radiology. 2004;232:221–8. diffusion abnormality index as a new imaging bio-
64. Pavlisa G, Rados M, Pavlisa G, Pavic L, Potocki K, marker for early assessment of brain tumor response
Mayer D. The differences of water diffusion between to radiation therapy. Neuro Oncol. 2014;16:131–9.
brain tissue infiltrated by tumor and peritumoral 77. Lee CC, Wintermark M, Xu Z, Yen CP, Schlesinger
vasogenic edema. Clin Imaging. 2009;33:96–101. D, Sheehan JP. Application of diffusion-weighted
65. Kono K, Inoue Y, Nakayama K, Shakudo M, Mornio magnetic resonance imaging to predict the intra-
M, Ohata K, et al. The role of diffusion-weighted cranial metastatic tumor response to gamma knife
imaging in patients with brain tumors. AJNR Am J radiosurgery. J Neurooncol. 2014;118:351–61.
Neuroradiol. 2001;22:1081–8. 78. Tomura N, Narita K, Izumi J, Suzuki A, Anbai A,
66. Tsougos I, Svolos P, Kousi E, Fountas K, Theodorou Otani T, et al. Diffusion changes in a tumor and peri-
K, Fezoulidis I, et al. Differentiation of glioblas- tumoral tissue after stereotactic irradiation for brain
tumors: possible prediction of treatment response. J 91. Shankar LK, Hoffman JM, Bacharach S, et al.
Comput Assist Tomogr. 2006;30:496–500. Consensus recommendations for the use of 18F-
79. Huang CF, Chou HH, Tu HT, Yang MS, Lee JK, FDG PET as an indicator of therapeutic response in
Lin LY. Diffusion magnetic resonance imaging as patients in National Cancer Institute Trials. J Nucl
an evaluation of the response of brain metastases Med. 2006;47:1059–66.
treated by stereotactic radiosurgery. Surg Neurol. 92. Horky LL, Hsiao EM, Weiss SE, Drappatz J,
2008;69:62–8. Gerbaudo VH. Dual phase FDG-PET imaging of
80. Knitter JR, Erly WK, Stea BD, Lemole GM, brain metastases provides superior assessment
Germano IM, Doshi AH, et al. Interval change in of recurrence versus post-treatment necrosis. J
diffusion and perfusion MRI parameters for the Neurooncol. 2011;103:137–46.
assessment of pseudoprogression in cerebral metas- 93. Terakawa Y, Tsuyuguchi N, Iwai Y, et al. Diagnostic
tases treated with stereotactic radiation. AJR Am J accuracy of 11C-methionine PET for differentiation
Roentgenol. 2018;211:168–75. of recurrent brain tumors from radiation necrosis
81. Marom EM, McAdams HP, Erasmus JJ, Goodman after radiotherapy. J Nucl Med. 2008;49:694–9.
PC, Culhane DK, Coleman RE, et al. Staging 94. Tsuyuguchi N, Sunada I, Iwai Y, Yamanaka K,
non-small cell lung cancer with whole-body Tanaka K, Takami T, et al. Methionine positron
PET. Radiology. 1999;212:803–9. emission tomography of recurrent metastatic brain
82. Ohno Y, Koyama H, Nogami M, Takenaka D, tumor and radiation necrosis after stereotactic
Yoshikawa T, Yoshimura M, et al. Whole-body MR radiosurgery: is a differential diagnosis possible? J
imaging vs FDG-PET: comparison of accuracy of Neurosurg. 2003;98:1056–64.
M-stage diagnosis for lung cancer patients. J Magn 95. Grosu AL, Astner ST, Riedel E, et al. An interindivid-
Reson Imaging. 2007;26:498–509. ual comparison of O-(2-[18F]fluoroethyl)-L-tyrosine
83. Rohren EM, Provenzale JM, Barboriak DP, Coleman (FET)- and L-[methyl-11C]methionine (MET)-PET
RE. Screening for cerebral metastases with FDG in patients with brain gliomas and metastases. Int J
PET in patients undergoing whole-body staging of Radiat Oncol Biol Phys. 2011;81:1049–58.
non-central nervous system malignancy. Radiology. 96. Cicone F, Minniti G, Romano A, et al. Accuracy of
2003;226:181–7. F-DOPA PET and perfusion-MRI for differentiat-
84. Kitajima K, Nakamoto Y, Okizuka H, Onishi Y, ing radionecrotic from progressive brain metastases
Senda M, Suganama N, et al. Accuracy of whole- after radiosurgery. Eur J Nucl Med Mol Imaging.
body FDG-PET/CT for detecting brain metastases 2015;42:103–11.
from non-central nervous system tumors. Ann Nucl 97. Lizarraga KJ, Allen-Auerbach M, Czernin J,
Med. 2008;22:595–602. DeSalles A, Yong WH, Phelps ME, et al. 18F-FDOPA
85. Kruger S, Mottaghy FM, Buck AK, et al. Brain PET for differentiating recurrent or progressive
metastasis in lung cancer. Comparison of cerebral brain metastatic tumors from late or delayed radia-
MRI and 18F-FDG-PET/CT for diagnosis in the ini- tion injury after radiation treatment. J Nucl Med.
tial staging. Nuklearmedizin. 2011;50:101–6. 2014;55:303–6.
86. Brink I, Schumacher T, Mix M, Ruhland S, Stoelben 98. Galldiks N, Stoffels G, Filss CP, et al. Role of
E, Digel W, et al. Impact of [18F]FDG-PET on the O-(2-(18)F-fluoroethyl)-L-tyrosine PET for differ-
primary staging of small-cell lung cancer. Eur J Nucl entiation of local recurrent brain metastases from
Med Mol Imaging. 2004;31:1614–20. radiation necrosis. J Nucl Med. 2012;53:1367–74.
87. Chung JK, Kim YK, Kim SK, Lee YJ, Paek S, Yeo 99. Ceccon G, Lohmann P, Stoffels G, Judov N, Filss
JS, et al. Usefulness of 11C-methionine PET in the CP, Rapp M, et al. Dynamic O-(2-18F-fluoroethyl)-
evaluation of brain lesions that are hypo- or iso- L-tyrosine positron emission tomography differenti-
metabolic on 18F-FDG PET. Eur J Nucl Med Mol ates brain metastasis recurrence from radiation injury
Imaging. 2002;29:176–82. after radiotherapy. Neuro Oncol. 2017;19:281–8.
88. Momose T, Nariai T, Kawabe T, et al. Clinical 100. Romagna A. Suspected recurrence of brain metasta-
benefit of 11C methionine PET imaging as a plan- ses after focused high dose radiotherapy: can [(18)F]
ning modality for radiosurgery of previously irra- FET-PET overcome diagnostic uncertainties? Radiat
diated recurrent brain metastases. Clin Nucl Med. Oncol. 2016;11:139.
2014;39:939–43. 101. Kebir S, Rauschenbach L, Galldiks N, Schlaak M,
89. Hutterer M, Nowosielski M, Putzer D, et al. Hattingen E, Landsberg J, et al. Dynamic O-(2-[18F]
[18F]-fluoro-ethyl-L-tyrosine PET: a valuable diag- fluoroethyl)-L-tyrosine PET imaging for the detec-
nostic tool in neuro-oncology, but not all that glitters tion of checkpoint inhibitor-related pseudoprogres-
is glioma. Neuro Oncol. 2013;15:341–51. sion in melanoma brain metastases. Neuro Oncol.
90. Chao ST, Suh JH, Raja S, Lee S-Y, Barnett G. The 2016;18:1462–4.
sensitivity and specificity of FDG PET in distin- 102. Lohmann P, Kocher M, Ceccon G, et al. Combined
guishing recurrent brain tumor from radionecrosis in FET PET/MRI radiomics differentiates radiation
patients treated with stereotactic radiosurgery. Int J injury from recurrent brain metastasis. Neuroimage
Cancer. 2001;96:191–7. Clin. 2018;20:537–42.
103. Zhang W, Ma XX, Ji YM, Kang XS, Li sensitive magnetic resonance imaging with 3.0 tesla.
CF. Haemorrhage detection in brain metastases of Yonago Acta Med. 2013;56(1):7–12.
lung cancer patients using magnetic resonance imag- 116. Cutsforth-Gregory JK, Lanzino G, Link MJ, Brown
ing. J Int Med Res. 2009;37(4):1139–44. RD Jr, Flemming KD. Characterization of radiation-
104. de Rochefort L, Brown R, Prince MR, Wang induced cavernous malformations and comparison
Y. Quantitative MR susceptibility mapping using with a nonradiation cavernous malformation cohort.
piece-wise constant regularized inversion of the mag- J Neurosurg. 2015;122(5):1214–22.
netic field. Magn Reson Med. 2008;60(4):1003–9. 117. Roddy E, Sear K, Felton E, Tamrazi B, Gauvain
105. Schweser F, Deistung A, Lehr BW, Reichenbach K, Torkildson J, et al. Presence of cerebral micro-
JR. Differentiation between diamagnetic and para- bleeds is associated with worse executive function
magnetic cerebral lesions based on magnetic suscep- in pediatric brain tumor survivors. Neuro Oncol.
tibility mapping. Med Phys. 2010;37(10):5165–78. 2016;18(11):1548–58.
106. Hamilton R, Krauze M, Romkes M, Omolo B, 118. Ghia AJ, Tward JD, Anker CJ, Boucher KM, Jensen
Konstantinopoulos P, Reinhart T, et al. Pathologic RL, Shrieve DC. Radiosurgery for melanoma brain
and gene expression features of metastatic melano- metastases: the impact of hemorrhage on local con-
mas to the brain. Cancer. 2013;119(15):2737–46. trol. J Radiosurg SBRT. 2014;3(1):43–50.
107. Gaviani P, Mullins ME, Braga TA, Hedley-Whyte 119. Redmond AJ, Diluna ML, Herbert R, Moliterno JA,
ET, Halpern EF, Schaefer PS, et al. Improved detec- Desai R, Knisely JP, et al. Gamma knife surgery for
tion of metastatic melanoma by T2∗-weighted imag- the treatment of melanoma metastases: the effect of
ing. Am J Neuroradiol. 2006;27(3):605–8. intratumoral hemorrhage on survival. J Neurosurg.
108. Sehgal V, Delproposto Z, Haddar D, Haacke EM, 2008;109:99–105.
Sloan AE, Zamorano LJ, et al. Susceptibility- 120. Kalfas F, Ronchini N, Godowicz TT, Cavazzani P,
weighted imaging to visualize blood products and Severi P. Peritumoral and intratumoral hemorrhage
improve tumor contrast in the study of brain masses. after stereotactic radiosurgery for renal cell carci-
J Magn Reson Imaging. 2006;24(1):41–51. noma metastasis to the brain. J Radiosurg SBRT.
109. Gramsch C, Goricks SL, Behrens F, Zimmer L, 2011;1(2):163–8.
Schadendorf D, Krasny A, et al. Isolated cere- 121. Sykova E, Jendelova P. In vivo tracking of stem
bral susceptibility artefacts in patients with malig- cells in brain and spinal cord injury. Prog Brain Res.
nant melanoma: metastasis or not? Eur Radiol. 2007;161:367–83.
2013;23:2622–7. 122. Zhang X, de Chickera SN, Willert C, Economopoulos
110. Franceschi AM, Moschos SJ, Anders CK, Glaubiger V, Noad J, Rohani R, et al. Cellular magnetic reso-
S, Collichio FA, Lee CB, et al. Utility of susceptibil- nance imaging of monocyte-derived dendritic cell
ity weighted imaging (SWI) in the detection of brain migration from healthy donors and cancer patients
hemorrhagic metastases from breast cancer and mel- as assessed in a scid mouse model. Cytotherapy.
anoma. J Comput Assist Tomogr. 2016;40(5):803–5. 2011;13(10):1234–48.
111. Deike-Hofmann K, Thunemann D, Breckwoldt MO, 123. de Chickera S, Willert C, Mallet C, Foley R, Foster
Schwarz D, Radbruch A, Enk A, et al. Sensitivity P, Dekaban GA. Cellular MRI as a suitable, sensitive
of different MRI sequences in the early detec- non-invasive modality for correlating in vivo migra-
tion of melanoma brain metastases. PLoS One. tory efficiencies of different dendritic cell popula-
2018;13(3):e0193946. tions with subsequent immunological outcomes. Int
112. Roongpiboonsopit D, Kuijf HJ, Charidimou A, Immunol. 2012;24(1):29–41.
Xiong L, Vashkevich A, Martinez-Ramirez S, et al. 124. Dekaban GA, Snir J, Shrum B, de Chickera S,
Evolution of cerebral microbleeds after cranial irra- Willert C, Merrill M, et al. Semiquantitation of
diation in medulloblastoma patients. Neurology. mouse dendritic cell migration in vivo using cellular
2017;88(8):789–96. MRI. J Immunother. 2009;32(3):240–51.
113. Passos J, Nzwalo H, Valente M, Marques J, Azevedo 125. Shapiro EM, Medford-Davis LN, Fahmy TM, Dunbar
A, Netto E, et al. Microbleeds and cavernomas after CE, Koretsky AP. Antibody-mediated cell labeling of
radiotherapy for paediatric primary brain tumors. J peripheral T cells with micron-sized iron oxide particles
Neurol Sci. 2017;372:413–6. (MPIOs) allows single cell detection by MRI. Contrast
114. Nandjgam RNK, Viswanathan A, Delgado P, Media Mol Imaging. 2007;2(3):147–53.
Skehan ME, Smith EE, Rosand J, et al. MR imag- 126. Foster PJ, Dunn EA, Karl KE, Snir JA, Nycz CM,
ing detection of cerebral microbleeds: effect of Harvey AJ, et al. Cellular magnetic resonance imag-
susceptibility-weighted imaging, section thick- ing: in vivo imaging of melanoma cells in lymph
ness, and field strength. AJNR Am J Neuroradiol. nodes of mice. Neoplasia. 2008;10(3):207–16.
2009;30(2):338–43. 127. Perera M, Ribot EJ, Percy DB, McFadden C,

115. Tanjino T, Kanasaki Y, Tahara T, Michimoto K, Simedrea C, Palmieri D, et al. In vivo magnetic
Kodani K, Kakite S, et al. Radiation-induced micro- resonance imaging for investigating the devel-
bleeds after cranial irradiation: evaluation by phase- opment and distribution of experimental brain
metastases due to breast cancer. Transl Oncol. 142. Aronen HJ, Gazit IE, Louis DN, Buchbinder BR,
2012;5(3):217–25. Pardo FS, Weisskoff RM, et al. Cerebral blood vol-
128. Ribot EJ, Foster PJ. In vivo MRI discrimination ume maps of gliomas: comparison with tumor grade
between live and lysed iron-labelled cells using and histologic findings. Radiology. 1994;191:41–51.
balanced steady state free precession. Eur Radiol. 143. Cha S, Johnson G, Wadghiri YZ, Jin O, Babb J,
2012;22(9):2027–34. Zagzag D, et al. Dynamic, contrast-enhanced perfu-
129. Canella B, Raine CS. The adhesion molecule and sion MRI in mouse gliomas correlation with histopa-
cytokine profile of multiple sclerosis lesions. Ann thology. Magn Reson Med. 2003;49:848–55.
Neurol. 1995;37:424–35. 144. Hompland T, Gulliksrud K, Ellingsen C, Rofstad
130. Polman CH, O’Connor PW, Havrdova E, Hutchinson EK. Assessment of the interstitial fluid pressure
M, Kappos L, Miller DH, et al. A randomized, of tumors by dynamic contrast-enhanced mag-
placebo-controlled trial of natalizumab for relapsing netic resonance imaging with contrast agents
multiple sclerosis. N Engl J Med. 2006;354:889–910. of different molecular weights. Acta Oncol.
131. Piraino PS, Yednock TA, Freedman SB, Pleiss MA, 2013;52:627–35.
Vandevert C, Thorsett ED, et al. Prolonged reversal 145. Wong CS, Van der Kogel AJ. Mechanisms of radia-
of chronic experimental allergic encephalomyelitis tion injury to the central nervous system implications
using a small molecule inhibitor of alpha4 integrin. J for neuroprotection. Mol Interv. 2004;4:273–84.
Neuroimmunol. 2002;131:147–59. 146. Zach L, Guez D, Last D, Daniels D, Grober Y,
132. Carbonell WS, Ansorge O, Sibson N, Muschel Nissim O, et al. Delayed contrast extravasation MRI:
R. The vascular basement membrane as “soil” in a new paradigm in neuro-oncology. Neuro Oncol.
brain metastasis. PLoS One. 2009;4:e5857. 2015;17:457–65.
133. Kusters B, Leenders WP, Wesseling P, Smits D, 147. Wang B, Zhang Y, Zhao B, Zhao P, Ge M, Gao M,
Verrijp K, Ruiter DJ, et al. Vascular endothelial et al. Postcontrast T1 mapping for differential diag-
growth factor-A(165) induces progression of mela- nosis of recurrence and radionecrosis after gamma
noma brain metastases without induction of sprout- knife radiosurgery for brain metastasis. AJNR Am J
ing angiogenesis. Cancer Res. 2002;62:341–5. Neuroradiol. 2018;39(6):1025–31.
134. Laubli H, Borsig L. Selecting as mediators of lung 148. Chen W, Cloughesy T, Kamdar N, Satyamurthy N,
metastasis. Cancer Microenviron. 2010;3:97–105. Bergsneider M, Liau L, et al. Imaging proliferation
135. Ludwig RJ, Boehme B, Podda M, Henschler R, in brain tumors with 18F-FLT PET: comparison with
Jager E, Tandi C, et al. Endothelial P-selectin as a 18F-FDG. J Nucl Med. 2005;46(6):945–52.
target of heparin action in experimental melanoma 149. Chen W, Delaloye S, Silverman DH, Geist C,
lung metastasis. Cancer Res. 2004;64:2743–50. Czernin J, Sayre J, et al. Predicting treatment
136. Khatib AM, Kontogiannea M, Fallavollita L, response of malignant gliomas to bevacizumab and
Jamison B, Meterissian S, Brodt P. Rapid induction irinotecan by imaging proliferation with [18F] fluo-
of cytokine and E-selectin expression in the liver rothymidine positron emission tomography: a pilot
in response to metastatic tumor cells. Cancer Res. study. J Clin Oncol. 2007;25(30):4714–21.
1999;59:1356–61. 150. Shields AF, Grierson JR, Dohmen BM, Machulla HJ,
137. Vidal-Vanaclocha F, Fantuzzi G, Mendoza L, Stayanoff JC, Lawhorn-Crews JM, et al. Imaging
Fuentes AM, Anasagasti MJ, Martin J, et al. IL-18 proliferation in vivo with [F-18]FLT and positron
regulates IL-1beta-dependent hepatic melanoma emission tomography. Nat Med. 1998;4(11):1334–6.
metastasis via vascular cell adhesion molecule-1. 151. Ribas A, Benz MR, Allen-Auerbach MS, Radu
Proc Natl Acad Sci U S A. 2000;97:734–9. C, Chmielowski B, Seja E, et al. Imaging of
138. McAteer MA, Sibson NR, von Zur Muhlen C, CTLA4 blockade-induced cell replication with
Schneider JE, Lowe AS, Warrick N, et al. In vivo (18)F-FLT PET in patients with advanced mela-
magnetic resonance imaging of acute brain inflam- noma treated with tremelimumab. J Nucl Med.
mation using microparticles of iron oxide. Nat Med. 2010;51(3):340–6.
2007;13:1253–8. 152. Schwarzenberg J, Czernin J, Cloughesy TF,
139. Serres S, Soto MS, Hamilton A, McAteer MA, Ellingson BM, Pope WB, Geist C, et al. 3′-Deoxy-
Carbonell WS, Robson MD, et al. Molecular 3′-18F-fluorothymidine PET and MRI for early
MRI enables early and sensitive detection of survival predictions in patients with recurrent malig-
brain metastases. Proc Natl Acad Sci U S A. nant glioma treated with bevacizumab. J Nucl Med.
2012;109(17):6674–9. 2012;53(1):29–36.
140. Zach L, Guez D, Last D, Daniels D, Grober Y, 153. Wardak M, Schiepers C, Dahlbom M, Cloughesy T,
Nissim O, et al. Delayed contrast extravasation Chen W, Satyamurthy N, et al. Discriminant analy-
MRI for depicting tumor and non-tumoral tissues sis of (1)(8)F-fluorothymidine kinetic parameters to
in primary and metastatic brain tumors. PLoS One. predict survival in patients with recurrent high-grade
2012;7:e52008. glioma. Clin Cancer Res. 2011;17(20):6553–62.
141. Jain RK. Molecular regulation of vessel maturation. 154. Schiepers C, Dahlbom M, Chen W, Cloughesy T,
Nat Med. 2003;9(6):685–93. Czernin J, Phelps ME, et al. Kinetics of 3′-deoxy-
3′-18F-fluorothymidine during treatment monitor- small-molecule detector of apoptosis. Cell Res.

ing of recurrent high-grade glioma. J Nucl Med. 2009;19(5):625–37.
2010;51(5):720–7. 158. Allen AM, Ben-Ami M, Reshef A, Steinmetz A,
155. Nguyen NC, Yee MK, Tuchay AM, Kirkwood JM, Kundel Y, Inbar E, et al. Assessment of response of
Tawbi H, Mountz JM. Targeted therapy and immu- brain metastases to radiotherapy by PET imaging
notherapy response assessment with F-18 fluoro- of apoptosis with 18F-ML-10. Eur J Nucl Med Mol
thymidine positron-emission tomography/magnetic Imaging. 2012;39:1400–8.
resonance imaging in melanoma brain metastasis: a 159. Duarte PS, Marin JFG, De Carvalho JWA, Sapienza
pilot study. Front Oncol. 2018;8:18. MR, Buchpiguel CA. Brain metastasis of medul-
156. O’Sullivan CC, Lindenberg M, Bryla C, Patronas lary thyroid carcinoma without macroscopic cal-
N, Peer CJ, Amiri-Kordestani L, et al. ANG 1005 cification detected first on 68Ga-DOTATATE and
for breast cancer brain metastases: correlation then on 18F-Fluoride PET/CT. Clin Nucl Med.
between 18F-FLT-PET after first cycle and MRI 2018;43(8):623–4.
in response assessment. Breast Cancer Res Treat. 160. Carreras C, Kulkarni HR, Baum RP. Rare metastases
2016;160(1):51–9. detected by 68Ga-somatostatin receptor PET/CT in
157. Cohen A, Shirvan A, Levin G, Grimberg H, patients with neuroendocrine tumors. Recent Results
Reshef A, Ziv I. From the Gla domain to a novel Cancer Res. 2013;194:379–84.
Part II
Evaluation of CNS Metastasis
Clinical Presentation of Central
Nervous System Metastases
8
Laura E. Donovan and Rajiv S. Magge
Introduction in part due to improved imaging techniques as

well as more effective systemic therapies result-
Central nervous system (CNS) metastases are ing in longer overall survival [3]. The CNS is
associated with significant morbidity and mor- considered a sanctuary site for disease. While
tality and remain one of the most challenging there have been advances in the treatment of
complications of systemic cancer. While intra- certain types of CNS metastases with targeted
parenchymal brain metastases represent the most therapies or checkpoint inhibitors, the majority
common site of CNS disease, other potential of chemotherapeutic agents have limited blood-
locations in the brain include the pituitary gland, brain barrier penetration [7, 8]. Survival var-
ventricular system and choroid plexus, as well ies greatly depending on the underlying cancer
as the spinal cord and leptomeninges [1]. In this subtype, burden of systemic disease, and other
chapter we provide an overview of the clinical patient-associated factors such as age and perfor-
presentation of CNS metastases including diag- mance status [9].
nostic workup and initial management. Brain metastases can present at any point
along the disease course. The Surveillance,
Epidemiology, and End Results (SEER) data-
Brain Metastases base recently added information regarding the
presence or absence of brain metastases at the
Brain metastases are the most common intra- time of initial diagnosis. Based on these data, the
cranial malignancy, occurring ten times more incidence proportion of brain metastases in all
frequently than primary brain tumors [2]. The patients with newly diagnosed cancer was calcu-
reported incidence of brain metastases varies, lated to be about 2%. Brain metastases at diagno-
ranging from 6% to 30% across various studies sis were most common (>10%) in small cell and
[3–6]. The incidence is thought to be increasing, non-small cell lung cancer regardless of cancer
stage. Conversely, among all patients with breast
cancer, melanoma, and renal cancer, the incidence
L. E. Donovan at diagnosis was relatively low (0.4%, 0.7%, and
Department of Neurology, Columbia University 1.5% respectively). Compared to patients with
Irving Medical Center, New York, NY, USA any stage cancer diagnosis, patients with sys-
R. S. Magge (*) temic metastases at baseline carried an increased
Department of Neurology, NewYork-Presbyterian/ incidence of brain metastases at 12.1%. In this
Weill Cornell Medicine, New York, NY, USA population, the incidence of brain metastases was
e-mail: ram9116@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_8
118 L. E. Donovan and R. S. Magge
highest in patients with melanoma (28.2%), lung medial motor cortex often affect the leg, while
adenocarcinoma (26.8%), small cell lung cancer more lateral lesions tend to involve the arm and
(23.5%), and renal cancer (10.8%) [10]. face to a larger degree [19].
The presentation of brain metastases varies The temporal lobes include the hippocampus,
dramatically, ranging from incidentally discov- limbic system, portions of the visual pathways,
ered, asymptomatic lesions found during a stag- and Wernicke’s area. Temporal lobe metastases,
ing workup to acute neurologic decompensation particularly bilateral lesions, can present with
requiring emergent intervention, particularly in short-term memory impairment. If the domi-
the case of hemorrhagic metastases. Depending nant hemisphere is affected, Wernicke’s apha-
on the location, number, size, and degree of sur- sia, characterized by an inability to comprehend
rounding edema, they can present with a diversity language (also known as receptive aphasia),
of symptoms [11]. can result. On exam, a contralateral superior
quadrantanopia may be detected if the optic
tracts are involved; however, this is not always
Focal Neurologic Deficits reported by the patient. Seizures are also very
common, particularly with medial temporal lobe
Focal neurologic deficits are the presenting symp- lesions [11, 19].
tom in 20–75% of patients with brain metastases Patients with right parietal lesions often pres-
[11, 12]. The specific deficit depends on the loca- ent with visual spatial disturbance, specifically
tion of the tumor. Intraparenchymal metastases left neglect. This may manifest itself as bumping
are most often found along the grey-white junc- into things on the left or, in more extreme cases,
tion or in watershed regions. This is thought to neglecting the left side completely. Patients may
reflect hematogenous dissemination of disease report forgetting to close the car door on the
with seeding of distal capillaries by tumor micro- left or improperly clothing the left side of their
emboli [13, 14]. While some studies suggest the body. Often there is a lack of awareness of the
majority of brain metastases (70–80%) are supra- deficit, or anosognosia, seen with non-dominant
tentorial, other autopsy studies have found nearly parietal lesions. Left parietal lesions can present
equal rates of disease in the posterior fossa and with acalculia. Contralateral hemisensory loss or
cerebellum [15, 16]. Compared to other cancers, visual field deficits, specifically an inferior qua-
breast and lung cancer metastases seem to have drantanopia, can also be seen. Occipital lesions
a predilection for the cerebellum [17]. Although also present with a contralateral visual field cut,
limited by small sample size, a recent study quan- typically involving the entire contralateral hemi-
tifying the spatial distribution of brain metastases field. Complex visual hallucinations have also
found that metastases were more common along been reported [11, 19].
branches of the anterior cerebral artery, particu- Infratentorial disease can present with ataxia
larly in the paracingulate gyrus [18]. or gait impairment. Cerebellar hemispheric
Supratentorial metastases can involve any lobe lesions can cause ipsilateral dysmetria and
of the brain. Patients with symptomatic tumors incoordination. Lesions affecting the cerebellar
in the frontal lobes can present with contralat- vermis are more likely to contribute to truncal
eral hemiparesis as well as personality changes instability instead of classic dysmetria. Given
ranging from abulia to disinhibition. When the the high density of motor and sensory path-
dominant hemisphere is involved, a Broca’s-type ways as well as cranial nerve nuclei that run
aphasia, characterized by difficulty expressing through the brainstem, even small lesions can
language, can occur. Due to the spatial arrange- be highly symptomatic. Brainstem lesions can
ment of motor function along the homunculus, cause contralateral hemiparesis and hemisen-
weakness from cortical lesions may be very sory loss of the face, arm, and leg. If the lower
specific, such as isolated hand weakness from pons (below the facial nucleus) or medulla are
a metastasis in the hand knob. Lesions in the affected, patients may present with crossed find-
8 Clinical Presentation of Central Nervous System Metastases 119
ings including ipsilateral weakness of the face or vomiting [24]. However, a prospective study
and contralateral weakness in the body [19]. of over 100 patients at Memorial Sloan Kettering
In the setting of intratumoral hemorrhage, Cancer Center with brain tumors (both primary
these deficits may be acute in onset; however, and metastatic) found that the majority (77%)
in many patients, they progress over the course described a tension-type headache that was most
of days to weeks. Progressive focal neurologic often bifrontal or ipsilateral. Unlike classic ten-
deficits in any patient with known systemic can- sion-type headaches, these were more frequently
cer should trigger additional workup for CNS associated with nausea (40%) and worsened with
metastases. bending over (32%). In this series, the classic
morning headache was uncommon [25].
Headaches are also very common in the
Cognitive Impairment general population, with an annual prevalence
approaching 60% [26]. In a cancer patient with
While not often considered a true focal neuro- an underlying headache disorder, a change in the
logic deficit, cognitive impairment is also com- frequency, severity, or character of their typical
mon in patients with brain metastases [11]. This headaches should prompt additional evaluation to
can manifest as disorientation, confusion, mem- exclude brain metastases.
ory impairment, and/or executive dysfunction.
One study evaluating whole brain radiotherapy
in patients with lung cancer found that 65% of Seizures
patients with brain metastases had cognitive
dysfunction prior to treatment [20]. In patients Up to one-third of patients with brain metastases
with primary brain tumors, cognitive impair- present with seizures. In one retrospective study
ment is one of the leading causes of disability of over 500 patients with surgically resected
and caregiver distress. In caregivers of patients metastases, multiple lesions, temporal and
with brain metastases, cognitive impairment was occipital locations, and bone involvement were
associated with worse coping strategies, which all associated with preoperative seizures. Large
can negatively impact quality of life [21]. While tumors (>5 cm) and those in locations other than
delirium or acute mental status changes are com- the frontal lobes were associated with uncon-
mon in cancer patients, this is a less common pre- trolled seizures preoperatively (defined as requir-
sentation of brain metastases. In a series of 132 ing more than one antiepileptic drug (AED)).
patients requiring neurology consults for altered Headaches and cognitive dysfunction were also
mental status, brain metastases were the underly- commonly seen with seizures. In this cohort,
ing etiology in only 15% of cases [22]. subtotal resection, >3 metastatic lesions, tempo-
ral lobe location, local recurrence, and no post-
operative chemotherapy were all associated with
Headaches seizures in the postoperative setting [27, 28].
While some studies have suggested the pres-
Headaches are another common symptom of brain ence or absence of seizures has no impact on
metastases, reported by approximately 25–60% overall survival with brain metastases, they can
of patients, particularly in the setting of multiple significantly impair quality of life. Each state
lesions [1, 11]. These can result from increased has laws limiting driving after seizures. Patients
intracranial pressure (ICP) as well as traction also need to be maintained on AEDs, sometimes
on the dura which contains pain fibers [23]. The indefinitely. Poorly controlled seizures are asso-
classic headache resulting from a brain tumor is ciated with worse outcomes in patients with brain
focal, worse in the morning, and exacerbated by metastases [29].
lying flat or Valsalva maneuvers. These head- Numerous studies have demonstrated no ben-
aches may also be associated with nausea and/ efit to prophylactic AEDs in the primary preven-
tion of seizures, with an increased risk of adverse rhage into the pituitary gland. While this is of
events [30, 31]. For this reason, the American concern with pituitary adenomas, it is rarely seen
Academy of Neurology recommends against pro- with metastases [39].
phylactic AED use for patients with brain tumors,
including metastases [32]. Despite this, prophy-
lactic AED use remains common in practice [33]. Leptomeningeal Disease
Many of the original studies focused on older
AEDs with more side effects, while newer drugs The leptomeninges include the pia mater, sub-
such as levetiracetam are often better tolerated arachnoid space, and arachnoid membrane
with a more favorable risk-benefit profile [34, 35]. surrounding the brain and spinal cord [19].
There are also data to suggest that primary pro- Metastases to this space are typically a late-stage
phylaxis may be beneficial in a high-risk subset of complication of cancer. While leptomeningeal
patients or in the perioperative period to decrease disease (LMD) is most common in adenocarci-
the rate of early postoperative seizures [36, 37]. nomas and hematologic malignancies, almost
However, randomized controlled trials are limited any cancer can metastasize to the leptomeninges
and this remains an area of controversy. [40, 41]. As the cerebral spinal fluid (CSF) flows
throughout the entire leptomeningeal space, bath-
ing the brain and spinal cord, the presentation
Uncommon Intracranial Metastases of LMD is highly variable and can range from
symptomatic hydrocephalus to isolated cranial
Pituitary Metastases neuropathies, multifocal deficits, and/or seizures.
Metastases to the pituitary gland are rare, When LMD involves the cerebral leptomenin-
accounting for 0.14–3.6% of intracranial metas- ges, patients often present with signs of elevated
tases, although in autopsy series, the incidence ICP. Leptomeningeal metastases can interfere
has been reported as high as 28%. Breast and with CSF reabsorption through the arachnoid
lung cancer are the most common cancers to granulations, causing hydrocephalus, or limit
metastasize to the pituitary gland, but many other ventricular compliance such as in the setting of
cancers have been reported. Unlike adenomas, diffuse subarachnoid tumor, resulting in elevated
which affect the anterior pituitary gland, metas- ICP without radiographic hydrocephalus [42].
tases tend to have a predilection for the posterior Patients often present with positional headaches,
pituitary [38, 39]. worse in the morning or when bending over.
Over 80% of pituitary metastases are asymp- These can be associated with nausea or vomit-
tomatic. In patients who present with symptoms, ing and sometimes with neck pain and stiffness
visual impairment has been reported in almost [43]. Vision changes including blurry vision or
50% of cases. The most common visual field horizontal diplopia from a partial cranial nerve
deficit seen with pituitary lesions is a bitemporal VI palsy may also be seen. As ICP increases,
hemianopia due to compression of the optic chi- patients may become increasingly lethargic [41].
asm, which overlies the pituitary gland. Endocrine Other alterations in consciousness include sei-
dysfunction, specifically diabetes insipidus (DI) zures or abrupt unresponsiveness precipitated by
and panhypopituitarism, was reported in over changing position, a phenomenon known as pres-
one-third of cases each. Patients with diabetes sure or plateau waves [42].
insipidus often present with increased thirst and Cranial nerve involvement from leptomen-
urine output. Panhypopituitarism can be more ingeal disease can manifest as vision changes,
difficult to diagnose as symptoms may be non- numbness over the face, facial weakness, hearing
specific including fatigue, lethargy, and ortho- loss, tinnitus, or hoarseness [42, 43]. Involvement
stasis. Headaches were also common, occurring of the spinal cord and cauda equina nerve roots
in 35% of patients. Pituitary apoplexy is a life- can contribute to radicular pain, bowel or blad-
threatening emergency characterized by hemor- der dysfunction, or focal numbness or weakness
in the legs [41, 42]. A combination of symptoms resonance imaging (MRI) [12]. For patients pre-
affecting multiple levels of the neuro-axis should senting with focal neurologic complaints, imag-
raise suspicion for LMD in a patient with meta- ing can be focused to the area of highest concern,
static cancer [1]. such as the brain alone or a particular spinal
level. In the case of patients with parenchymal
brain metastases identified on imaging, full CNS
Spinal Metastases staging is not always necessary if the patient is
otherwise asymptomatic. For patients present-
Tumors involving the spine are divided into three ing with leptomeningeal disease, workup should
categories based on location: extradural, intradu- include complete imaging of the neuro- axis
ral extramedullary, and intradural intramedullary. including brain and total spine, with and with-
The vast majority of metastases are extradural out contrast. When there is clinical suspicion for
[44]. Extradural tumors often arise from the ver- LMD but negative imaging, the gold standard for
tebral bodies, most commonly in the thoracic diagnosis is a lumbar puncture for CSF analysis.
spine, and extend into the extradural space [42, Multiple lumbar punctures may be necessary as
44]. Initially, these lesions may present with the sensitivity of CSF cytology does not exceed
severe back pain. Pain is often severe, worse at 90% until after three studies [48]. Extradural spi-
night, and may wake the patient from sleep. Both nal metastases arising from the vertebrae rarely
extradural and intradural extramedullary lesions occur in isolation, so imaging the entire spine is
can present with cord compression. As the spinal recommended [49]. Once CNS metastases are
cord becomes compressed, patients can develop identified, systemic restaging is recommended
focal neurologic deficits including weakness, as this has implications for both prognosis and
numbness, bowel or bladder dysfunction, or gait treatment options.
impairment [43]. Approximately 5% of patients The initial management of a patient with
with metastatic cancer initially present with cord symptomatic brain metastases includes high-dose
compression [45]. dexamethasone to decrease edema and reduce
Intramedullary metastases are rare, with an symptom burden. Steroids may not be necessary
incidence of <2%. Although they may be the in asymptomatic brain metastases without signif-
presenting symptom of disease, intramedullary icant edema. Treatment options for patients with
metastases are typically seen in the setting of brain metastases have evolved and may include a
known brain metastases or leptomeningeal dis- combination of radiation, surgery, chemotherapy,
ease [46]. Patients may present with spinal cord immunotherapy, or targeted agents. These will
syndromes, such as a Brown-Sequard syndrome, be discussed extensively in the later chapters of
characterized by ipsilateral weakness and vibra- this book; however, the appropriate approach to
tory/proprioceptive loss and contralateral loss of the management of each patient depends on the
pinprick and temperature below the level of the burden of CNS disease, the extent of systemic
lesion. Pain, weakness, and sensory changes are disease, and the options available for systemic
the most commonly reported symptoms; how- treatment [2, 50, 51].
ever, bowel or bladder dysfunction and spasticity
can also be seen. Typically patients have a rela-
tively rapid decline as the lesion increases in size,
References
but it is possible for diagnosis to be delayed [47].
1. Lee EQ. Nervous system metastases from systemic can-
cer. Continuum. 2015;21(2 Neuro-Oncology):415–28.
Workup and Management 2. Patchell RA. The management of brain metastases.
Cancer Treat Rev. 2003;29(6):533–40.
3. Davis FG, Dolecek TA, McCarthy BJ, Villano
The imaging modality of choice for CNS JL. Toward determining the lifetime occurrence of
metastases is gadolinium-enhanced magnetic metastatic brain tumors estimated from 2007 United
States cancer incidence data. Neuro-Oncology. tion therapy in brain metastases. J Clin Oncol.
2012;14(9):1171–7. 2003;21(13):2529–36.
4. Nathoo N, Chahlavi A, Barnett GH, Toms 21. Saria MG, Courchesne N, Evangelista L, et al.

SA. Pathobiology of brain metastases. J Clin Pathol. Cognitive dysfunction in patients with brain metas-
2005;58(3):237–42. tases: influences on caregiver resilience and coping.
5. Nayak L, Lee EQ, Wen PY. Epidemiology of brain Support Care Cancer. 2017;25(4):1247–56.
metastases. Curr Oncol Rep. 2012;14(1):48–54. 22. Tuma R, DeAngelis LM. Altered mental status in
6. Martin AM, Cagney DN, Catalano PJ, et al. patients with cancer. Arch Neurol. 2000;57(12):
Brain metastases in newly diagnosed breast can- 1727–31.
cer: a population- based study. JAMA Oncol. 23. Taylor LP. Mechanism of brain tumor headache.

2017;3(8):1069–77. Headache. 2014;54(4):772–5.
7. Deeken JF, Loscher W. The blood-brain barrier and 24.
Headache Classification Committee of the
cancer: transporters, treatment, and Trojan horses. International Headache Society (IHS). The interna-
Clin Cancer Res. 2007;13(6):1663–74. tional classification of headache disorders, 3rd edn.
8. Lockman PR, Mittapalli RK, Taskar KS, et al. Cephalalgia. 2018;38(1):1–211.
Heterogeneous blood-tumor barrier permeability 25. Forsyth PA, Posner JB. Headaches in patients with
determines drug efficacy in experimental brain metas- brain tumors: a study of 111 patients. Neurology.
tases of breast cancer. Clin Cancer Res. 2010;16(23): 1993;43(9):1678–83.
5664–78. 26. Stovner LJ, Andree C. Prevalence of headache in

9. Sperduto PW, Chao ST, Sneed PK, et al. Diagnosis- Europe: a review for the Eurolight project. J Headache
specific prognostic factors, indexes, and treatment Pain. 2010;11(4):289–99.
outcomes for patients with newly diagnosed brain 27. Pojskic M, Bopp MHA, Schymalla M, Nimsky C,
metastases: a multi-institutional analysis of 4,259 Carl B. Retrospective study of 229 surgically treated
patients. Int J Radiat Oncol Biol Phys. 2010;77(3): patients with brain metastases: prognostic factors,
655–61. outcome and comparison of recursive partitioning
10. Cagney DN, Martin AM, Catalano PJ, et al. Incidence analysis and diagnosis-specific graded prognostic
and prognosis of patients with brain metastases at assessment. Surg Neurol Int. 2017;8:259.
diagnosis of systemic malignancy: a population-based 28. Wu A, Weingart JD, Gallia GL, et al. Risk factors
study. Neuro-Oncology. 2017;19(11):1511–21. for preoperative seizures and loss of seizure control
11. Noh T, Walbert T. Brain metastasis: clinical manifes- in patients undergoing surgery for metastatic brain
tations, symptom management, and palliative care. tumors. World Neurosurg. 2017;104:120–8.
Handb Clin Neurol. 2018;149:75–88. 29. Cacho-Diaz B, San-Juan D, Salmeron K, Boyzo C,
12. Kaal EC, Taphoorn MJ, Vecht CJ. Symptomatic man- Lorenzana-Mendoza N. Choice of antiepileptic drugs
agement and imaging of brain metastases. J Neuro- affects the outcome in cancer patients with seizures.
Oncol. 2005;75(1):15–20. Clin Transl Oncol. 2018;20(12):1571–6.
13. Massague J, Obenauf AC. Metastatic coloniza-
30. Lobos-Urbina D, Kittsteiner-Manubens L, Pena J. Is
tion by circulating tumour cells. Nature. 2016; primary prevention with antiepileptic drugs effec-
529(7586):298–306. tive in brain tumors or brain metastases? Medwave.
14. Hwang TL, Close TP, Grego JM, Brannon WL,
2017;17(Suppl1):e6871.
Gonzales F. Predilection of brain metastasis in gray 31. Wu AS, Trinh VT, Suki D, et al. A prospective ran-
and white matter junction and vascular border zones. domized trial of perioperative seizure prophylaxis
Cancer. 1996;77(8):1551–5. in patients with intraparenchymal brain tumors. J
15. Tsukada Y, Fouad A, Pickren JW, Lane WW. Central Neurosurg. 2013;118(4):873–83.
nervous system metastasis from breast carcinoma. 32. Stevens GH. Antiepileptic therapy in patients with
Autopsy study. Cancer. 1983;52(12):2349–54. central nervous system malignancies. Curr Neurol
16. Ghia A, Tome WA, Thomas S, et al. Distribution of Neurosci Rep. 2006;6(4):311–8.
brain metastases in relation to the hippocampus: 33. Dewan MC, Thompson RC, Kalkanis SN, Barker

implications for neurocognitive functional preserva- FG 2nd, Hadjipanayis CG. Prophylactic antiepileptic
tion. Int J Radiat Oncol Biol Phys. 2007;68(4):971–7. drug administration following brain tumor resection:
17. Bender ET, Tome WA. Distribution of brain metasta- results of a recent AANS/CNS Section on Tumors
ses: implications for non-uniform dose prescriptions. Survey. J Neurosurg. 2017;126(6):1772–8.
Br J Radiol. 2011;84(1003):649–58. 34. Zachenhofer I, Donat M, Oberndorfer S, Roessler
18. Yanagihara TK, Lee A, Wang TJC. Quantitative
K. Perioperative levetiracetam for prevention of sei-
analysis of the spatial distribution of metastatic brain zures in supratentorial brain tumor surgery. J Neuro-
lesions. Tomography. 2017;3(1):16–22. Oncol. 2011;101(1):101–6.
19. Blumenfeld H. Neuroanatomy through clinical cases. 35. Gokhale S, Khan SA, Agrawal A, Friedman AH,

2nd ed. Sunderland: Sinauer Associates; 2010. McDonagh DL. Levetiracetam seizure prophy-
20. Mehta MP, Rodrigus P, Terhaard CH, et al. Survival laxis in craniotomy patients at high risk for post-
and neurologic outcomes in a randomized trial operative seizures. Asian J Neurosurg. 2013;8(4):
of motexafin gadolinium and whole-brain radia- 169–73.
36. Goldlust SA, Hsu M, Lassman AB, Panageas KS, 43. Mendez JS, DeAngelis LM. Metastatic complications
Avila EK. Seizure prophylaxis and melanoma brain of cancer involving the central and peripheral nervous
metastases. J Neuro-Oncol. 2012;108(1):109–14. systems. Neurol Clin. 2018;36(3):579–98.
37. Joiner EF, Youngerman BE, Hudson TS, et al.
44. Van Goethem JW, van den Hauwe L, Ozsarlak O,
Effectiveness of perioperative antiepileptic drug pro- De Schepper AM, Parizel PM. Spinal tumors. Eur J
phylaxis for early and late seizures following onco- Radiol. 2004;50(2):159–76.
logic neurosurgery: a meta-analysis. J Neurosurg. 45. Schiff D, O’Neill BP, Suman VJ. Spinal epidural

2018;130:1274–82. metastasis as the initial manifestation of malignancy:
38. He W, Chen F, Dalm B, Kirby PA, Greenlee
clinical features and diagnostic approach. Neurology.
JD. Metastatic involvement of the pituitary gland: a 1997;49(2):452–6.
systematic review with pooled individual patient data 46. Schiff D, O’Neill BP. Intramedullary spinal cord

analysis. Pituitary. 2015;18(1):159–68. metastases: clinical features and treatment outcome.
39. Javanbakht A, D’Apuzzo M, Badie B, Salehian
Neurology. 1996;47(4):906–12.
B. Pituitary metastasis: a rare condition. Endocr 47. Lee SS, Kim MK, Sym SJ, et al. Intramedullary spi-
Connect. 2018;7:1049–57. nal cord metastases: a single-institution experience. J
40. Mittica G, Senetta R, Richiardi L, et al. Meningeal Neuro-Oncol. 2007;84(1):85–9.
carcinomatosis underdiagnosis and overestima- 48. Glantz MJ, Cole BF, Glantz LK, et al. Cerebrospinal
tion: incidence in a large consecutive and unselected fluid cytology in patients with cancer: minimizing
population of breast cancer patients. BMC Cancer. false-negative results. Cancer. 1998;82(4):733–9.
2015;15:1021. 49. Shah LM, Salzman KL. Imaging of spinal metastatic
41. Nayar G, Ejikeme T, Chongsathidkiet P, et al.
disease. Int J Surg Oncol. 2011;2011:769753.
Leptomeningeal disease: current diagnostic and thera- 50. Lin X, DeAngelis LM. Treatment of brain metastases.
peutic strategies. Oncotarget. 2017;8(42):73312–28. J Clin Oncol. 2015;33(30):3475–84.
42. DeAngelis LM, Posner JB, Posner JB. Neurologic 51. Aizer AA, Lee EQ. Brain metastases. Neurol Clin.
complications of cancer. 2nd ed. Oxford/New York: 2018;36(3):557–77.
Oxford University Press; 2009.
Management of Seizures in Brain
Metastases
9
Ankush Bhatia and Edward K. Avila
Epidemiology, Incidence, monly cortical lesions in the frontal lobe, parietal

and Etiology of Seizures in Patients lobe, and temporal lobe. This is undoubtedly due
with Brain Metastases to the inherent epileptogenicity of the cortical
gray matter [2]. Occipital lobe seizures are seen
Managing seizures is an integral aspect of neuro- less frequently. Masses near the fissure of Rolando
oncological care in patients with brain metasta- are more prone to seizures, but lesions in the pitu-
ses. Seizure has been reported as the presenting itary or posterior fossa are rarely associated with
symptom in brain metastases in up to 20% of seizures unless they invade supratentorially [19].
patients [1, 2]. A recent systematic review There is increased risk of seizures with a higher
reported the incidence of seizures at 14.8%, number of metastatic lesions.
although this has been cited as high as 40% in the Clinicians should carefully consider the etiol-
literature [1–18]. Regardless of the exact percent- ogy of seizures in a patient with brain metastases.
age, seizures are a common problem in this popu- In addition to the metastasis itself acting as a
lation. The incidence varies depending on tumor focus for seizure, other possibilities include lep-
type: In a retrospective series of 470 patients with tomeningeal or dural metastases, metabolic con-
brain metastases, the likelihood of seizure was ditions, cerebral infarction or hemorrhage,
highest in melanoma (67%) and lowest in breast infections, and treatment-related causes.
cancer (16%). Other common tumor types were Table 9.1 identifies some of the potential etiolo-
lung (29%), gastrointestinal (21%), and unknown gies of seizures in patients with metastatic brain
primary (25%). The high incidence of seizure tumors. Cancer patients are at higher risk for
with melanoma brain metastases is thought to be metabolic encephalopathies such as hyponatre-
due to the tendency toward hemorrhagic conver- mia or hypoglycemia, opportunistic infections, or
sion, which can be epileptogenic. side effects of therapy. Paraneoplastic encephali-
Tumor location is another important factor that tis is another potential cause of seizures in
can impact the frequency of seizures in patients patients with systemic cancer.
with brain metastases. Seizures are almost exclu- Much research has attempted to clarify the
sively due to supratentorial disease, most com- various factors that contribute to seizure develop-
ment in patients with brain metastases. While the
mechanism of tumor-associated epilepsy remains
A. Bhatia · E. K. Avila (*) poorly understood, theories focus on peritumoral
Department of Neurology, Memorial Sloan Kettering amino acid disturbances, local metabolic imbal-
Cancer Center, New York, NY, USA ances, cerebral edema, pH abnormalities, and
e-mail: avilae@mskcc.org

https://doi.org/10.1007/978-3-030-42958-4_9
126 A. Bhatia and E. K. Avila
Table 9.1 Possible etiologies of seizures in the patientby cortical irritation from invasion of cortical
with brain metastases brain parenchyma, adjacent leptomeningeal
Metastatic central nervous deposits, or local edema. Tumor-related seizures
system neoplasms Treatment-related causes
are often repetitive or stereotyped, preceded by
Parenchymal metastases Radiation therapy
Dural-based metastases Acute an aura and followed by a postictal phase. The
Leptomeningeal Early-delayed International League Against Epilepsy (ILAE)
metastases recently proposed a new classification of seizure
Late-delayed types that classifies focal seizures based on
Toxic/metabolic conditions Chemotherapy whether consciousness is altered during the epi-
Hyponatremia Antimetabolites
sode [22].
Hypoglycemia Methotrexate
Hypoxia Cytarabine
Focal seizures with retained awareness (previ-
Hypocalcemia l-asparaginase ously known as simple partial seizures) are fur-
Hypomagnesemia Vincaakaloids ther separated based on the semiology and
Topoisomerase epileptic region on the cortex. For example, a sei-
inhibitors zure in the occipital cortex may manifest as flash-
Cerebral infarction Alkylators ing lights in the opposite field of vision, whereas
Ifosfamide
a seizure that begins in the motor cortex may
Cerebral hemorrhage Nitrosoureas
Cisplatin cause rhythmic jerking movements of the contra-
Infections Bevacizumab lateral face, arm, and leg. A parietal cortex sei-
Bacterial zure can disrupt spatial perception, while a mass
Listeria monocytogenes Opioids in the dominant frontal cortex can disrupt lan-
Viral Meperidine guage and cause aphasic seizures. Temporal lobe
Cytomegalovirus Antiemetics seizures may begin with auras such as an abnor-
Herpes simplex Phenothiazines
mal taste, smell, or gastrointestinal symptoms.
Fungal Butyrophenones
Cryptococcus neoformans Antibiotics Patients may experience only auras, which are
Aspergillus fumigatus Penicillins focal seizures that can cause symptoms, but not
Candida species Fluoroquinolones impair consciousness. Auras can be present for
Parasites Imipenem-cilastatin months and eventually progress to a generalized
Toxoplasma gondii Paraneoplastic disease seizure [23]. Patients may either return to normal
∗∗Edited and updated from (Table 4–9, pg. 108, from immediately after the event or have a prolonged
DeAngelis/Posner book “Neurological Complications of postictal period of worsened neurological func-
Cancer” 2nd edition. Edited and updated with permission
from Oxford University Press tion, corresponding to where the seizure origi-
nated in the brain. Notably, a patient with a focal
altered immunologic activity (Fig. 9.1) [20, 21]. motor seizure of the arm may suffer from postic-
Further understanding of these mechanisms may tal weakness that can last for minutes to hours,
elucidate why some patients with seizures also known as Todd’s paralysis.
become refractory to antiepileptic drugs even Focal seizures with impaired consciousness,
after removal of the metastatic lesion. formerly known as complex partial seizures,
occur in patients who have alteration of aware-
ness during the event. During these seizures,
Clinical Manifestations patients may be alert but not respond to environ-
mental stimuli; they may engage in repetitive
Seizures in patients with brain metastases are behaviors like facial grimacing, chewing, or lip
usually focal but can appear generalized if the smacking, otherwise known as automatisms.
focal discharge is asymptomatic. The presence of Hostile or aggressive behavior can also occur in
an aura (warning sign), or specific ictal and peri- patients who have a focus in the deep frontal
ictal phenomena, typically reflects the tumor’s lobe. Similarly, patients may have an aura, ictal
location within the brain. The ictus can be caused period, followed by a postictal period.
9 Management of Seizures in Brain Metastases 127
Impaired pH buffering
Alkaline pH Acidic pH
Ischemia
+++ −−−
Excitatory Inhibitory
neuronal pool neuronal pool
NMDA & Immunological
? GABA factors K+, Na+ channels
GABA K+, Na+, Ca2+

currents
ACh Tumor associated epilepsy
currents
Amino acid receptors lonic Electrical

imbalance changes
Conformational lonic buffering

changes by glial cells
GABA 5HT Cell membrane changes
Amino acids
EAAs
Shared Structural changes
release Fluidity parameters
Lipid production
Enzyme disturbances
& peroxidation
Glutathione, triglyceride, and glycosphingolipid
pathway disturbances
Fig. 9.1 Summary of possible causative and influencing Whittle IR. The pathogenesis of tumour associated epi-
mechanisms on tumor-associated epilepsy. The rich inter- lepsy. Acta Neurochir (Wien). 2000;142(1):1–15.
play of the varied factors and many plausible routes for Reprinted with permission from Springer)
seizure causation is highlighted. (From: Beaumont A,
Any type of focal seizure can progress to gen- lightheadedness, presyncope, or even frank syn-
eralization, which often involve tonic-clonic cope and are typically associated with positional
movements. The tonic phase begins with a sud- changes (e.g., sitting to standing). These episodes
den loss of consciousness followed by stiffening occur in the setting of increased intracranial pres-
of the arms, legs, chest, and back, which can then sure (ICP) or venous obstruction, even in the
evolve into jerking of muscles for minutes. The absence of headache. Key physical exam findings
clonic phase is characterized by tongue biting as that would suggest increased ICP include papill-
well as frothy and bloody sputum. The postictal edema on funduscopic exam and hydrocephalus
phase begins after movement has ceased; the on imaging. Seizures have been reported in up to
patient usually enters a deep sleep that can last 25% of patients with leptomeningeal disease.
several minutes before the patient gradually Patients with seizures should be treated with anti-
wakens. seizure drugs, while patients with increased ICP
Of note, plateau waves (or pressure waves) in should be treated with steroids and/or neurosur-
a patient with elevated ICP, often due to lepto- gical intervention.
meningeal disease, can frequently mimic seizure; Brain tumor patients can also develop status epi-
however, the two diagnoses should not be con- lepticus, which may be either focal or g eneralized,
fused as treatment differs significantly. Plateau convulsive or nonconvulsive. Management of status
waves are events that can involve dizziness, epilepticus is discussed later in this chapter.
Diagnostic Evaluation also not routinely needed for those without

clinical evidence of a seizure. However, EEG is
The diagnostic evaluation of a brain metastasis essential for diagnosing nonconvulsive seizures
patient with a seizure requires careful history tak- and nonconvulsive status epilepticus (NCSE)
ing, detailed neurological exam, electroencepha- and should be considered in all patients with
lography (EEG), and neuroimaging [19, 24]. The brain metastases who have altered mental
history of present illness should include a reliable status.
description of the event not only from the patient Monitoring patients during the diagnostic
but also from an eyewitness, understanding that workup can be done with a routine EEG (ambula-
one is not always present. Questions should focus tory or inpatient) or with prolonged video EEG
on possible triggers or precipitants of seizure that monitoring while inpatient. During a routine
can lower the seizure threshold, such as strong EEG, electrical activity is recorded from elec-
emotions, exercise, and alcohol use [25]. An trodes placed on the scalp in standard positions
accurate description of the seizure involves for a short amount of time—generally 30 min.
descriptions of the events leading up to the sei- The sensitivity of detecting IEDs is low in a rou-
zure, the ictal phenomena, and the postictal state. tine EEG and can be increased with prolonged
The history summarizes the various seizure semi- monitoring overnight with video EEG monitor-
ologies, past antiepileptic use with associated ing [27]. Sensitivity can also be increased when
side effects, birth history, and any history of CNS seizure frequency increases and timing of EEG is
infections. The neurological exam of a seizure closer to last seizure or if seizures are provoked
patient is usually normal unless a structural CNS by hyperventilation, photic stimulation, sleep
lesion causing localizing signs such as Todd’s deprivation, or medication withdrawal. However,
paralysis is present. Comprehensive laboratory treatment should never be delayed if the clinician
testing including AED levels, complete blood believes that seizure is the most likely diagnosis.
count, electrolytes, glucose, calcium, magne- Interpretation of EEG findings is best done by an
sium, renal function, liver function, and urinaly- experienced clinician with specific training in
sis should all be completed in case there are any EEG. Lateralized periodic discharges (LPDs,
reversible metabolic abnormalities lowering the previously known as PLEDs) are commonly seen
seizure threshold [26]. in patients with rapidly growing cerebral malig-
EEG is an important diagnostic tool for the nancies, which cause acute cortical injury. LPDs
evaluation of a seizure patient. EEG can help are defined by lateralized, persistent spikes, sharp
support the diagnosis of epilepsy, localize the waves, or sharply contoured slow waves that
origin of epileptic activity, and at times assist in occur repetitively [28]. Focal slowing or general-
determining the underlying epileptic syndrome. ized slowing of the EEG rhythm is nonspecific
However, there are limitations of EEG. For and can be seen in patients with multiple sys-
example, intermittent EEG changes and interic- temic issues, a postictal period, or an underlying
tal epileptiform discharges (IEDs) can be infre- structural lesion that is not necessarily
quent and not always present during the epileptogenic.
recording of a patient who has had a prior sei- Neuroimaging is vital to the evaluation of sei-
zure. Additionally, epileptic activity from brain zure in a patient with suspected or known brain
metastases that are small or deep in the brain metastases [29]. Computed tomography (CT) is
may not have a concordant EEG finding as usually the first imaging modality obtained in a
these microvolt signals may not be visible on patient with a new-onset seizure because it is
scalp recordings. Clinicians must thus under- available quickly and can exclude certain
stand the strengths and weaknesses of EEG to neurological emergencies such as hemorrhage

diagnose a patient with epilepsy. An EEG may [29, 30]. Contrast-enhanced magnetic resonance
not be required in a patient who has had a clini- imaging (MRI) is a more sensitive imaging
cally obvious seizure with full recovery. EEG is modality for the detection of brain metastases
than CT and is the neuroimaging modality of drug-drug interactions (Table 9.2) [34–37].
choice [31]. Better spatial resolution and soft- Levetiracetam is often prescribed in the general
tissue contrast allows for visualization of smaller population because it is well-tolerated; however
brain metastases and leptomeningeal disease. some patients can have neuropsychiatric side
Positron-emission tomography (PET) and func- effects such as irritability, agitation, anxiety,
tional MRI are additional neuroimaging modali- and depression [37]. Clinicians should remain
ties that can be used for presurgical evaluation of vigilant as patients with frontal lobe brain
patients with brain metastases [32]. metastases are at higher risk for these neuropsy-
Lumbar puncture should be performed if there chiatric side effects.
is suspicion for a CNS infection or leptomenin- Multidrug regimens should be avoided if pos-
geal metastasis. Appropriate neuroimaging with sible since monotherapy will increase the likeli-
CT or MRI should be performed before lumbar hood of compliance, provide a wider therapeutic
puncture to rule out any space-occupying lesion window, and be more cost-effective over time.
that may render the procedure unsafe. Single-drug therapy also minimizes potential
interactions with chemotherapy and other drug-
drug interactions. Data from patients with pri-
Treatment mary brain tumors suggest approximately 50% of
patients with tumor-related epilepsy will respond
Seizures in patients with brain metastases con- adequately to a single AED [38]. If a patient
tribute to morbidity and mortality and should be experiences recurrent seizures, the initial AED
aggressively treated when they occur [33]. The dose should be maximized, and appropriate
two mainstays of treatment include antiepileptic serum levels should be checked before switching
drug therapy and tumor-directed therapy. Anti- or adding a second anti-seizure drug. Lacosamide
seizure drug therapy is usually first to be admin- has been shown to be efficacious as an adjunctive
istered while plans are made for tumor-directed agent in patients with medically refractory epi-
therapy. lepsy and primary brain tumors [39, 40].
Levetiracetam
Anti-seizure Drug Therapy One retrospective study examined the role of
levetiracetam in patients with metastatic brain
Every patient who has a seizure due to a brain tumors—of the 13 patients treated with leveti-
metastasis should be treated with anti-seizure racetam as monotherapy (6 patients) or adjunc-
medications due to the high risk of recurrent sei- tive therapy (7 patients) with a median dose of
zure. There are no randomized trials that have 1000 mg/day, the median seizure frequency
established superiority of one agent over another decreased to 0 per week, suggesting complete
agent in this population. AEDs should be chosen seizure control. Only 3 of 13 patients reported
with the goal of controlling seizure at the lowest somnolence or headache as the most common
effective dose while minimizing toxicity. Certain adverse event [41]. There does not appear to be
AEDs require monitoring serum levels at recom- any significant interaction with other drugs or
mended intervals. AED interactions with chemo- chemotherapy, which is why levetiracetam is
therapy regimens should also be considered often the first drug used for neuro-oncologic
before prescribing. patients. It is also conveniently available in
AEDs with no or minimal hepatic enzyme- both oral and intravenous forms. Brivaracetam
inducing or enzyme-inhibiting properties, such is a newer formulation that is advertised to have
as levetiracetam, brivaracetam, pregabalin, similar efficacy without the psychiatric side
lamotrigine, lacosamide, and topiramate, are effects of levetiracetam; however, there have
generally preferred in initial treatment due to been no studies of this drug in brain
a favorable side effect profile and minimal metastases.
Table 9.2 Antiepileptic drugs

Average dose (serum Mechanism of
therapeutic range) Metabolism action Common adverse effects
Enzyme-inducing AEDs
Phenytoin 20 mg/kg load, then Hepatic Sodium Rash, osteomalacia, Stevens-
3–5 mg/kg daily or channel Johnson syndrome, gum
twice daily (10–20 ug/ hyperplasia, hirsutism
mL)
Carbamazepine 800–2400 mg, two to Hepatic Sodium Drowsiness, diplopia, rash,
four times a day channel Stevens-Johnson syndrome,
(8–12 ug/mL) leukopenia, hyponatremia
Phenobarbital 10 mg/kg load, then 75% hepatic; 25% GABA Drowsiness, Stevens-Johnson
1–3 mg/kg/d (15–40 ug/ renal syndrome, frozen shoulder,
mL) rash, ataxia, mood change
Oxcarbazepine 900–2400 mg two to 80% hepatic Sodium Hyponatremia, diplopia,
four times a day channel headache, drowsiness
Nonenzyme-inducing AEDs
Valproic acid 10–60 mg/kg three to Hepatic GABA, Hair loss, weight gain,
four times a day sodium pancreatitis, thrombocytopenia,
(60–100 ug/mL); channel platelet dysfunction, tremor,
intravenous infusion rate parkinsonism, extrapyramidal
is 20 mg/min, same dose syndrome
as oral
Gabapentin 900–4800 mg daily in Renal GABA Drowsiness, rapid titration,
three to four doses ataxia, weight gain
Pregabalin 150–600 mg/day Unknown Calcium Drowsiness, dizziness, ataxia
channel
Topiramate 100–400 mg twice a day 30–50% hepatic; Sodium Cognitive impairment,
50–70% renal channel, paresthesias, slow titration,
GABA, weight loss, renal calculi
AMPA/kainate
Levetiracetam 500–2000 mg twice a Enzymatic hydrolysis Synaptic Agitation, psychosis,
day vesicle protein drowsiness, glaucoma
binding
Brevitaracetam 50–100 mg twice a day Hepatic and Synaptic Drowsiness, ataxia, nystagmus,
extrahepatic amidase vesicle protein hypersomnia
mediated hydrolysis binding
Lamotrigine 300–500 mg twice a day 85% hepatic Sodium Drowsiness, rash, particularly
channel with concurrent valproate, slow
titration
Zonisamide 200–600 mg once or >90% hepatic Calcium, Drowsiness, headache, weight
twice a day (10–30 ug/ sodium loss, renal calculi, slow titration
mL) channel
Lacosamide 200–400 mg/day Hepatic Sodium Dizziness, headache, diplopia,
demethylation channel blurred vision
Clobazam 5–40 mg/day Hepatic GABA agonist Sedation, cognitive effects,
N-demethylation drowsiness
∗∗Edited and updated from Table 4–10, pg. 110, DeAngelis/Posner book, Neurological Complications of Cancer with
permission from Oxford University Press
Phenytoin seizures in patients with cancer, however, due to

Phenytoin is very effective in controlling seizures its activity as a CYP3A4 inducer and potential
in brain metastases and is often preferred in the interaction with chemotherapy. It also has several
setting of status epilepticus as it can be conve- side effects including elevated liver function
niently loaded intravenously or given orally. It is tests, osteomalacia, ataxia, nystagmus, myopa-
often not preferred in routine management of thy, and myelotoxicity.
Zonisamide Topiramate
Zonisamide has not been specifically studied in Topiramate has been studied in primary brain
brain metastases, but it has been investigated in tumors (n = 47) as an adjunctive therapy or
other brain tumors. A study of six patients with monotherapy (mean dose, 240 mg/day). It has
glial brain tumors showed an 83% response rate been reported to result in a 76% seizure reduction
and 69% reduction in seizure frequency [42]. of greater than 50% with only 8% of patients
Limiting side effects include renal calculi, sexual experiencing side effects that led to discontinua-
dysfunction, and drowsiness. The drug does not tion in 6% [46]. Notable side effects include som-
appear to interfere with the metabolism of other nolence, fatigue, psychomotor slowing,
drugs that utilize the cytochrome P-450 enzyme confusion, weight loss, glaucoma, and kidney
system. stones. Little is known regarding its interaction
with anticancer agents.
Oxcarbazepine and Carbamazepine
A retrospective study analyzed oxcarbazepine
(mean dosage, 1162.5 mg/day) to assess effi- Tumor-Directed Therapy
cacy and tolerability compared to phenobarbital
and carbamazepine in patients with brain Treatment of brain metastases with surgery, radi-
metastases. The results showed significantly ation therapy, and/or chemotherapy may also
fewer side effects with oxcarbazepine com- improve seizure activity. Lesionectomy of the
pared to these other drugs with equivalent effi- suspected epileptogenic zone has been shown to
cacy [43]. Patients on therapeutic doses of be efficacious in non-brain tumor patients, which
carbamazepine may complain of intermittent has been extrapolated to brain tumor patients.
diplopia as well as drowsiness. Carbamazepine But while several studies have examined the role
can also cause leukopenia, which can be espe- of surgery in control of epilepsy, few of these
cially concerning in patients who are receiving studies were specifically focused on brain metas-
myelosuppressive chemotherapy. The drug has tases; there are no standardized surgical
also been reported to be associated with approaches for seizure control in brain tumor
SIADH, aseptic meningitis, and rash. While patients. Additionally, it is difficult to compare
oxcarbazepine can rarely cause rash and seda- studies focused on seizure surgery and brain
tion, it is thought to have a more favorable side tumors due to variable histology, pathology, and
effect profile than carbamazepine. tumor locations. It is hypothesized that seizures
in tumor-associated epilepsy do not necessarily
Gabapentin and Pregabalin originate from the mass lesion but rather from the
Gabapentin was studied as an adjunctive anti- adjacent brain tissue, and therefore tumor-
epileptic in four patients with metastatic brain associated epilepsy may differ from idiopathic
tumors, and there was reported seizure resolu- epilepsy [47].
tion in half of patients [44]. Similarly, another Three operative strategies exist for brain
study showed that pregabalin (median dose, tumor resection for patients with seizures: (1)
300 mg) had a greater than 50% reduction in focal tumor resection, (2) radical tumor resection
seizure frequency in patients with primary brain without electrocorticography, and (3) radical
tumors [45]. Gabapentin and pregabalin appear tumor removal with electrocorticography.
to be quite safe, and these drugs are widely used Despite these options, resection without electro-
in cancer patients to treat chemotherapy-induced corticography may not eliminate the epilepto-
peripheral neuropathy. There is little to no inter- genic focus. Likewise, many patients require
action with other agents; however, side effects antiepileptic drugs before, during, and after
include somnolence, dizziness, ataxia, fatigue, tumor resection. Several studies have demon-
and weight gain. strated seizure frequency reduction after treatment
of primary brain tumors with chemotherapy, Untoward Effects of Anticonvulsants

although none of these studies specifically
included brain metastases [48–53]. Drug Interactions
Interactions between antiepileptic drugs and che-
motherapy are complex and mainly revolve
Prophylaxis around the cytochrome P-450 (CYP) system
(Table 9.3). Phenobarbital, phenytoin, and carba-
The data for prophylactic anticonvulsants in mazepine, for example, are three antiepileptic
patients with brain metastases are limited. A drugs known to be strong CYP3A4 inducers and
recent meta-analysis found only one study can significantly decrease the levels of vincristine,
which met their inclusion criteria, the most paclitaxel, irinotecan, teniposide, methotrexate,
strenuous of which was baseline information on and busulfan. Valproic acid has several complex
study participants, including subgroups of those interactions with certain chemotherapies, as it is
with brain metastasis [54]. The recommenda- one of the few cytochrome enzyme- inhibiting
tion was level 3 for adults with brain metastasis AEDs with highly protein-bound properties, in
who do not experience a seizure due to their addition to potential CYP2A6 induction. Highly
metastatic brain disease; routine AED prophy- protein-bound AEDs or chemotherapy agents—
laxis was not recommended. The recommenda- including phenytoin, phenobarbital, valproic acid,
tion was based on a study included in the cisplatin, etoposide, and teniposide—can interact
meta-analysis, which used phenytoin or pheno- with each other, affecting free and bound levels of
barbital as a prophylactic AED [55]. Since sei- both drugs. It should be noted that some patients
zure incidence was not significantly different in who take oral AEDs may have difficulty tolerat-
the treatment versus nontreatment group, the ing them while on highly emetogenic chemother-
authors cited adverse effects of AEDs as a rea- apy regimens. Clinicians should be cautious of
son against their use. heightened toxicity from the increased amount of
However, newer AEDs such as levetiracetam unbound drug, especially in patients who are
and lacosamide have gained popularity in the brain cachectic or malnourished. Chemotherapeutic
tumor population and are thought to be useful and agents such as methotrexate, doxorubicin, and
safe. In a retrospective study of patients with fre- cisplatin can decrease AED levels of valproic
quent, weekly seizures, levetiracetam use in meta- acid, carbamazepine, and phenytoin.
static lesions was tolerated well [41]. Seizure In the last decade, we have seen a substantial
frequency was reduced in all patients with meta- increase in the effectiveness of tyrosine kinase
static lesions to less than 50% of pre-levetiracetam inhibitors (TKI) for the treatment of brain metas-
baseline. There may be some subgroups of brain tases from various systemic malignancies. It is
metastasis patients who benefit from prophylactic necessary to pay close attention to drug interac-
AED use. A group of patients with metastatic mel- tions between antiepileptic drugs and TKIs mov-
anoma brain metastasis were evaluated for prophy- ing forward. CYP3A4-inducing AEDs can
lactic AED use over a 2-year period [9]. Seizure significantly increase the clearance and reduce
risk was studied relative to brain metastasis charac- the AUC of TKIs, specifically crizotinib,
teristics—hemorrhage and multiple supratentorial dasatinib, imatinib, and lapatinib [56]. There are
metastases were associated with increased seizure other TKIs that are 3A4 inhibitors; however,
risk. Univariate analysis revealed AED prophylaxis there is very little data reported on the metabo-
was significantly associated with a decreased sei- lism of AEDs in this context.
zure risk. Limitations of the study included its ret-
rospective nature and small patient cohort. Side Effects
However, it suggests that AED prophylaxis may be All antiepileptic drugs have potential side effects,
beneficial in some subgroups of patients with brain summarized in Table 9.2. The most common side
metastasis. effects across all AEDs are lethargy and cognitive
Table 9.3 Pharmacological aspects of antiepileptic drugs and interactions with chemotherapy agents
AED IV? CYP inducer PB (%) AED effect on chemo Chemo effect on AED
PHB Yes 1A2, 2A6, 2B6, 50 Thi↓ Nit↓ Vbl↓ Vnc↓ Mtx↓ Tmz
2C9, 3A4, 2C19 Pac↓ 9AC↓ Ten↓ Pro↑ Prd↓
Dox↓ Tam↓ Ifo↓
PHT Yes 2B6, 2C9, 2C19, 90 Pro↑ Pac↓ Bus↓ Top↓Vbl↓ Vnc↓ Mtx↓ Pro↑ Cis↓ Nit↓ Eto↓ Dox
3A4, 1A2 Mtx↓ Iri↓ 9AC↓ Ten↓ Dex↓ SrI↓ ↓ Dac↓ Vbl↓ Ble↓ Dex ↓↑ Car↓
5FU↑ Cpc↑ Tam↑ Tmz
CBZ No 1A2, 2B6, 2C9, 75 MTX↓ Pac↓ Vbl↓ Vnc↓ Ten↓ Cis↓ Dox↓ Tmz
2C19, 3A4 9AC↓ Srl↓ Pro↑
OXC No 3A4 40 – Tmz
VPA Yes 2A6 (inhib. 2C9, 90 – Mtx↓ Dox↓ Cis↓
2C19, 3A4)
TPX No 3A4 30 – Tmz
ZNS No (Inhib. 2E1) 50 – –
LTG No No 50 Mtx –
GBP No No <5 – –
PGB No No <5 – –
LVT Yes No <5 – –
LCS Yes No <5 – –
5FU 5-fluorouracil, 9AC 9-aminocampothecin, AED antiepileptic drug, Ble bleomycin, Bus busulfan, Ca calcium chan-
nel, Car carboplatin, CBZ carbamazepine, chemo chemotherapy, Cis cisplatin, cog cognitive/behavioral, Cpc
capecitabine, CYP cytochrome P-450, Dac dacarbazine, Dex dexamethasone, Dox doxorubicin, Eto etoposide, GABA
Ƴ-aminobutyric acid, GBP gabapentin, Ifo ifosfamide, inhib. enzyme inhibition, Iri irinotecan, IV intravenous, K kid-
ney, L liver, LCS lacosamide, LTG lamotrigine, LVT Levetiracetam, MAOI monoamine oxidase inhibitor, Mech mecha-
nism, Met metabolism, Mtx methotrexate, Na sodium channel, Nit nitrosourea, NMDA N-methyl-D-aspartate, n/v
nausea and vomiting, OXC oxcarbazepine, Pac paclitaxel, PB protein binding, PGB pregabalin, PHB phenobarbital
(and primidone), PHT phenytoin, Prd prednisone, Pro procarbazine, Srl sirolimus (and temsirolimus), SV synaptic
vesicle, Tam tamoxifen, Ten teniposide, Thi thiotepa, Tmz temozolomide, Top topotecan, TPX topiramate, Vbl vinblas-
tine, Vnc vincristine, VPA valproic acid, ZNS zonisamide
Boldface in table body denotes strong enzyme activity
∗∗Edited and updated from Table 2 of Avila and Graber, Curr Neurol Neurosci Rep (2010) 10:60–67 with permission
from Springer Nature
dysfunction, even if levels are within therapeutic a medical emergency and often requires aggres-
range. These side effects are often enhanced sive intensive care with intubation and general
when patients have several brain metastases. Side anesthesia [57]; one approach is illustrated in
effects of specific antiepileptic drugs among indi- Table 9.4.
vidual patients can vary, and their use often Nonconvulsive status epilepticus (NCSE) in
requires an individualized approach. See patients with brain metastases may be underdi-
Table 9.2 for a more detailed list of side effects of agnosed as patients are often altered or coma-
various antiepileptics. tose with no overt signs of seizure activity. One
study suggests an increased risk of mortality
within 2 months in patients with metastatic dis-
Convulsive and Nonconvulsive ease and progressing brain lesions [60]. In
Status Epilepticus another series, 8% of comatose patients were
found to be in electrographic status epilepticus
Status epilepticus is defined as either continu- [59]. In any comatose patient with risk factors
ous or intermittent seizures without recovery of for seizures and subtle motor or oculomotor
consciousness between seizures [57]. Status movements, electroencephalogram is recom-
epilepticus can be either convulsive or noncon- mended to definitively rule out electrographic
vulsive [58, 59]. Convulsive status epilepticus is seizures.
Table 9.4 Protocol for the treatment of convulsive status epilepticus at Memorial Sloan Kettering Cancer Center
First 5 min
ABCs
Diagnose status epilepticus
Obtain IV access
Begin ECG monitoring
Fingerstick for glucose—correct if necessary
Draw blood for BMP, Mg, Ca, Ph, CBC, LFT, AED levels (PHB, PHT, VPA, CBZ), toxicology screen
Call Neurology consult
6–10 min
Thiamine 100 mg IV; 50 ml of D50 IV in appropriate clinical setting
Lorazepam 4 mg IV over 2 min; if necessary, repeat once every 5 min. If no IV access, give diazepam 20 mg
rectally or midazolam 10 mg intranasally or intramuscularly
10–20 min
Add fosphenytoin 20 mg/kg IV at 50 mg/min with BP and ECG monitoring. Can re-bolus fosphenytoin 10 mg/kg if
seizures persist. Maintain level 15–20 μg/mL
20–60 min
If seizures persist, intubate and start phenobarbital IV 20 mg/kg at 50–100 mg/min
If still seizing, can add or switch (PHB) to midazolam: load 0.0.2 mg/kg; repeat 0.2–0.4 mg/kg boluses every 5 min
until seizures stop, up to a maximum total loading dose of 2 mg/kg. Initial rate 0.1 mg/kg/h. Continuous IV range
0.05–2 mg/kg/h
Or
Propofol: Load 1 mg/kg; repeat 1–2 mg/kg boluses every 3–5 min until seizures stop, up to a maximum total loading
dose of 10 mg/kg. Initial rate 2 mg/kg/hour. Dose range 1–15 mg/kg/hour
After 60 min
If seizures persist, use anesthetics
Continuous IV propofol: Load 1 mg/kg; initial 2 mg/kg/hr. Titrate until burst suppression
Will need to arrange continuous EEG monitoring (preferably as soon as the patient does not awaken rapidly)
Another possible consideration for fourth line treatment is valproate 40 mg/kg over 10 min. Can re-bolus 20 mg/kg
over 5 min
If bacterial meningitis is suspected, start ceftriaxone, vancomycin, and ampicillin (can start along with treatment for
SE). Start acyclovir if HSV encephalitis is suspected. Perform LP when stable
∗∗Edited and updated from Table 4–11, pg. 111, DeAngelis/Posner, Neurological complications of Cancer, with per-
mission from Oxford University Press
Driving quality of life. Driving restrictions can signifi-

cantly impact a patient’s ability to maintain
Placing driving restrictions on patients with sei- employment, attend social activities, and/or par-
zures from brain metastases is a highly contro- ticipate in school. Data are limited in determining
versial topic. The development of efficacious which patients with seizures can safely drive, and
seizure medications over the last several decades therefore regulations vary considerably in the
has reversed historical conviction that no patient USA from state to state. Clinicians should con-
with seizures—controlled or uncontrolled— sult the regulations of their respective state or
should be allowed to drive. There is consensus, country in which they practice before advising
however, that patients with uncontrolled seizures patients. It should be emphasized, however, that
should not drive given their increased risk of clinician judgment supersedes any state
motor vehicle accidents and subsequent property regulation.
damage, as well as potential for injury to self or While no studies have specifically looked at
others. For those patients with controlled sei- patients with brain metastases, data extrapolated
zures, clinicians must balance the risk of public from other epilepsy studies suggest that the most
safety with patient autonomy and preservation of reliable predictor of risk of seizure while driving
is the seizure-free interval. Limited data have with ovarian carcinoma: prognostic factors and out-
advocated for a seizure-free interval ranging any- come. J Neuro-Oncol. 2004;66(3):313–25.
7. Coia LR, Aaronson N, Linggood R, Loeffler J,
where from 3 to 12 months [61–65]. Priestman TJ. A report of the consensus workshop
panel on the treatment of brain metastases. Int J
Radiat Oncol Biol Phys. 1992;23(1):223–7.
Seizures at the End of Life 8. Glantz MJ, Cole BF, Friedberg MH, Lathi E, Choy
H, Furie K, et al. A randomized, blinded, placebo-
controlled trial of divalproex sodium prophylaxis in
Seizures are common at the end of life in patients adults with newly diagnosed brain tumors. Neurology.
with brain metastases. For those who are able to 1996;46(4):985–91.
swallow medications and have a previous history 9. Goldlust SA, Hsu M, Lassman AB, Panageas KS,
Avila EK. Seizure prophylaxis and melanoma brain
of seizures, patients should continue their anti- metastases. J Neuro-Oncol. 2012;108(1):109–14.
epileptic drugs. However, clinicians should be 10. Lee MH, Kong DS, Seol HJ, Nam DH, Lee JI. Risk of
aware of the often inevitable depressed mental seizure and its clinical implication in the patients with
status and the need to convert oral antiepileptics cerebral metastasis from lung cancer. Acta Neurochir.
2013;155(10):1833–7.
to non-oral routes. In patients who cannot safely 11. Miabi Z. Metastatic brain tumors: a retrospective

swallow, seizure management will depend on the review in East Azarbyjan (Tabriz). Acta Med Iran.
location of care support. If the patient requires 2011;49(2):115–7.
inpatient care, intravenous access can be main- 12. Mongan JP, Fadul CE, Cole BF, Zaki BI, Suriawinata
AA, Ripple GH, et al. Brain metastases from colorec-
tained and parenteral antiepileptics can be con- tal cancer: risk factors, incidence, and the pos-
tinued. For those at home, subcutaneous/ sible role of chemokines. Clin Colorectal Cancer.
sublingual lorazepam or buccal clonazepam can 2009;8(2):100–5.
be used to control seizures. Rectal diazepam and 13. Posner JB, Chernik NL. Intracranial metastases from
systemic cancer. Adv Neurol. 1978;19:579–92.
rectal/subcutaneous phenobarbital are another 14. Raizer JJ, Hwu WJ, Panageas KS, Wilton A, Baldwin
option in the home setting. Initiation of dexa- DE, Bailey E, et al. Brain and leptomeningeal
methasone should be considered for patients with metastases from cutaneous melanoma: survival out-
seizures caused by increased intracranial pres- comes based on clinical features. Neuro-Oncology.
2008;10(2):199–207.
sure from mass effect. 15. Simionescu MD. Metastatic tumors of the brain: a
follow-up study of 195 patients with neurosurgical
considerations. J Neurosurg. 1960;17:361–73.
16. Tremont-Lukats IW, Bobustuc G, Lagos GK, Lolas
References K, Kyritsis AP, Puduvalli VK. Brain metastasis from
prostate carcinoma: The M. D. Anderson Cancer
1. Cohen N, Strauss G, Lew R, Silver D, Recht Center experience. Cancer. 2003;98(2):363–8.
L. Should prophylactic anticonvulsants be admin- 17. Wong J, Hird A, Zhang L, Tsao M, Sinclair E,

istered to patients with newly-diagnosed cerebral Barnes E, et al. Symptoms and quality of life in
metastases? A retrospective analysis. J Clin Oncol. cancer patients with brain metastases following pal-
1988;6(10):1621–4. liative radiotherapy. Int J Radiat Oncol Biol Phys.
2. Lynam LM, Lyons MK, Drazkowski JF, Sirven JI, 2009;75(4):1125–31.
Noe KH, Zimmerman RS, et al. Frequency of sei- 18. Zacest AC, Besser M, Stevens G, Thompson JF,

zures in patients with newly diagnosed brain tumors: McCarthy WH, Culjak G. Surgical management of
a retrospective review. Clin Neurol Neurosurg. cerebral metastases from melanoma: outcome in
2007;109(7):634–8. 147 patients treated at a single institution over two
3. Byrne TN, Cascino TL, Posner JB. Brain metastasis decades. J Neurosurg. 2002;96(3):552–8.
from melanoma. J Neuro-Oncol. 1983;1(4):313–7. 19. van Breemen MS, Wilms EB, Vecht CJ. Epilepsy

4. Chan V, Sahgal A, Egeto P, Schweizer T, Das in patients with brain tumours: epidemiology,
S. Incidence of seizure in adult patients with mechanisms, and management. Lancet Neurol.
intracranial metastatic disease. J Neuro-Oncol. 2007;6(5):421–30.
2017;131(3):619–24. 20. Beaumont A, Whittle IR. The pathogenesis of

5. Chang L, Chen YL, Kao MC. Intracranial metastasis tumour associated epilepsy. Acta Neurochir.
of hepatocellular carcinoma: review of 45 cases. Surg 2000;142(1):1–15.
Neurol. 2004;62(2):172–7. 21. You G, Sha Z, Jiang T. The pathogenesis of tumor-
6. Cohen ZR, Suki D, Weinberg JS, Marmor E, Lang FF, related epilepsy and its implications for clinical treat-
Gershenson DM, et al. Brain metastases in patients ment. Seizure. 2012;21(3):153–9.
22. Fisher RS, Cross JH, French JA, Higurashi N, Hirsch tumour-related epilepsy: an Italian multicen-
E, Jansen FE, et al. Operational classification of tre prospective observational study. Eur J Neurol.
seizure types by the International League Against 2017;24(10):1283–9.
Epilepsy: Position Paper of the ILAE Commission 38. van Breemen MS, Rijsman RM, Taphoorn MJ,

for Classification and Terminology. Epilepsia. Walchenbach R, Zwinkels H, Vecht CJ. Efficacy of
2017;58(4):522–30. anti-epileptic drugs in patients with gliomas and sei-
23. Theodore WH. The postictal state: effects of age
zures. J Neurol. 2009;256(9):1519–26.
and underlying brain dysfunction. Epilepsy Behav. 39. Maschio M, Zarabla A, Maialetti A, Fabi A, Vidiri
2010;19(2):118–20. A, Villani V, et al. Quality of life, mood and seizure
24. Avila EK, Graber J. Seizures and epilepsy in cancer control in patients with brain tumor related epilepsy
patients. Curr Neurol Neurosci Rep. 2010;10(1):60–7. treated with lacosamide as add-on therapy: a prospec-
25. Fisher RS, Harding G, Erba G, Barkley GL, Wilkins tive explorative study with a historical control group.
A. Epilepsy Foundation of America Working Epilepsy Behav. 2017;73:83–9.
G. Photic- and pattern-induced seizures: a review for 40. Ruda R, Pellerino A, Franchino F, Bertolotti C,
the Epilepsy Foundation of America Working Group. Bruno F, Mo F, et al. Lacosamide in patients
Epilepsia. 2005;46(9):1426–41. with gliomas and uncontrolled seizures: results
26. Riggs JE. Neurologic manifestations of electrolyte from an observational study. J Neuro-Oncol.
disturbances. Neurol Clin. 2002;20(1):227–39.. vii 2018;136(1):105–14.
27. Burkholder DB, Britton JW, Rajasekaran V, Fabris 41.
Newton HB, Dalton J, Goldlust S, Pearl
RR, Cherian PJ, Kelly-Williams KM, et al. Routine D. Retrospective analysis of the efficacy and tolerabil-
vs extended outpatient EEG for the detection of ity of levetiracetam in patients with metastatic brain
interictal epileptiform discharges. Neurology. tumors. J Neuro-Oncol. 2007;84(3):293–6.
2016;86(16):1524–30. 42. Maschio M, Dinapoli L, Saveriano F, Pompili A,

28. Hirsch LJ, LaRoche SM, Gaspard N, Gerard E,
Carapella CM, Vidiri A, et al. Efficacy and tolerabil-
Svoronos A, Herman ST, et al. American Clinical ity of zonisamide as add-on in brain tumor-related
Neurophysiology Society’s standardized critical care epilepsy: preliminary report. Acta Neurol Scand.
EEG terminology: 2012 version. J Clin Neurophysiol. 2009;120(3):210–2.
2013;30(1):1–27. 43. Maschio M, Dinapoli L, Vidiri A, Pace A, Fabi A,
29. Recommendations for neuroimaging of patients
Pompili A, et al. The role side effects play in the
with epilepsy. Commission on Neuroimaging of the choice of antiepileptic therapy in brain tumor-related
International League Against Epilepsy. Epilepsia. epilepsy: a comparative study on traditional antiepi-
1997;38(11):1255–6. leptic drugs versus oxcarbazepine. J Exp Clin Cancer
30. Duncan JS. Imaging and epilepsy. Brain. 1997;120(Pt Res. 2009;28:60.
2):339–77. 44. Perry JR, Sawka C. Add-on gabapentin for refractory
31. Bernal B, Altman NR. Evidence-based medicine:
seizures in patients with brain tumours. Can J Neurol
neuroimaging of seizures. Neuroimaging Clin N Am. Sci. 1996;23(2):128–31.
2003;13(2):211–24. 45. Novy J, Stupp R, Rossetti AO. Pregabalin in

32. Swartz BE, Tomiyasu U, Delgado-Escueta AV,
patients with primary brain tumors and seizures: a
Mandelkern M, Khonsari A. Neuroimaging in tem- preliminary observation. Clin Neurol Neurosurg.
poral lobe epilepsy: test sensitivity and relationships 2009;111(2):171–3.
to pathology and postoperative outcome. Epilepsia. 46. Maschio M, Dinapoli L, Zarabla A, Pompili A,

1992;33(4):624–34. Carapella CM, Pace A, et al. Outcome and tolerabil-
33. Rossetti AO, Stupp R. Epilepsy in brain tumor
ity of topiramate in brain tumor associated epilepsy. J
patients. Curr Opin Neurol. 2010;23(6):603–9. Neuro-Oncol. 2008;86(1):61–70.
34. Usery JB, Michael LM 2nd, Sills AK, Finch CK. A 47. Berger MS, Ghatan S, Haglund MM, Dobbins J,

prospective evaluation and literature review of leve- Ojemann GA. Low-grade gliomas associated with
tiracetam use in patients with brain tumors and sei- intractable epilepsy: seizure outcome utilizing elec-
zures. J Neuro-Oncol. 2010;99(2):251–60. trocorticography during tumor resection. J Neurosurg.
35. Maschio M, Dinapoli L, Gomellini S, Ferraresi V, 1993;79(1):62–9.
Sperati F, Vidiri A, et al. Antiepileptics in brain metas- 48. Blonski M, Taillandier L, Herbet G, Maldonado IL,
tases: safety, efficacy and impact on life expectancy. J Beauchesne P, Fabbro M, et al. Combination of neo-
Neuro-Oncol. 2010;98(1):109–16. adjuvant chemotherapy followed by surgical resection
36. Saria MG, Corle C, Hu J, Rudnick JD, Phuphanich as a new strategy for WHO grade II gliomas: a study
S, Mrugala MM, et al. Retrospective analysis of the of cognitive status and quality of life. J Neuro-Oncol.
tolerability and activity of lacosamide in patients 2012;106(2):353–66.
with brain tumors: clinical article. J Neurosurg. 49. Koekkoek JA, Dirven L, Heimans JJ, Postma TJ,

2013;118(6):1183–7. Vos MJ, Reijneveld JC, et al. Seizure reduction in a
37. Bedetti C, Romoli M, Maschio M, Di Bonaventura low-grade glioma: more than a beneficial side effect
C, Nardi Cesarini E, Eusebi P, et al. Neuropsychiatric of temozolomide. J Neurol Neurosurg Psychiatry.
adverse events of antiepileptic drugs in brain 2015;86(4):366–73.
50. Taillandier L, Duffau H. Epilepsy and insular grade II 58. Husain AM, Horn GJ, Jacobson MP. Non-convulsive
gliomas: an interdisciplinary point of view from a ret- status epilepticus: usefulness of clinical features in
rospective monocentric series of 46 cases. Neurosurg selecting patients for urgent EEG. J Neurol Neurosurg
Focus. 2009;27(2):E8. Psychiatry. 2003;74(2):189–91.
51. Sherman JH, Moldovan K, Yeoh HK, Starke RM, 59. Towne AR, Waterhouse EJ, Boggs JG, Garnett LK,
Pouratian N, Shaffrey ME, et al. Impact of temo- Brown AJ, Smith JR Jr, et al. Prevalence of non-
zolomide chemotherapy on seizure frequency in convulsive status epilepticus in comatose patients.
patients with low-grade gliomas. J Neurosurg. Neurology. 2000;54(2):340–5.
2011;114(6):1617–21. 60. Marcuse LV, Lancman G, Demopoulos A, Fields

52. Pace A, Vidiri A, Galie E, Carosi M, Telera S,
M. Nonconvulsive status epilepticus in patients with
Cianciulli AM, et al. Temozolomide chemobrain tumors. Seizure. 2014;23(7):542–7.
therapy for progressive low-grade glioma: clini- 61. Ma BB, Bloch J, Krumholz A, Hopp JL, Foreman
cal benefits and radiological response. Ann Oncol. PJ, Soderstrom CA, et al. Regulating drivers with
2003;14(12):1722–6. epilepsy in Maryland: results of the application
53. Brada M, Viviers L, Abson C, Hines F, Britton J, of a United States consensus guideline. Epilepsia.
Ashley S, et al. Phase II study of primary temozolo- 2017;58(8):1389–97.
mide chemotherapy in patients with WHO grade II 62. Consensus statements, sample statutory provisions,
gliomas. Ann Oncol. 2003;14(12):1715–21. and model regulations regarding driver licensing
54.
Mikkelsen T, Paleologos NA, Robinson PD, and epilepsy. American Academy of Neurology,
Ammirati M, Andrews DW, Asher AL, et al. The role American Epilepsy Society, and Epilepsy Foundation
of prophylactic anticonvulsants in the management of of America. Epilepsia. 1994;35(3):696–705.
brain metastases: a systematic review and evidence- 63. Sheth SG, Krauss G, Krumholz A, Li G. Mortality in
based clinical practice guideline. J Neuro-Oncol. epilepsy: driving fatalities vs other causes of death in
2010;96(1):97–102. patients with epilepsy. Neurology. 2004;63(6):1002–7.
55. Forsyth PA, Weaver S, Fulton D, Brasher PM,
64. Krauss GL, Krumholz A, Carter RC, Li G, Kaplan
Sutherland G, Stewart D, et al. Prophylactic anticon- P. Risk factors for seizure-related motor vehi-
vulsants in patients with brain tumour. Can J Neurol cle crashes in patients with epilepsy. Neurology.
Sci. 2003;30(2):106–12. 1999;52(7):1324–9.
56. Benit CP, Vecht CJ. Seizures and cancer: drug inter- 65. Drazkowski JF, Fisher RS, Sirven JI, Demaerschalk
actions of anticonvulsants with chemotherapeutic BM, Uber-Zak L, Hentz JG, et al. Seizure-related
agents, tyrosine kinase inhibitors and glucocorticoids. motor vehicle crashes in Arizona before and after
Neuro-Oncol Pract. 2016;3(4):245–60. reducing the driving restriction from 12 to 3 months.
57. Chen JW, Wasterlain CG. Status epilepticus: patho- Mayo Clin Proc. 2003;78(7):819–25.
physiology and management in adults. Lancet Neurol.
2006;5(3):246–56.
Cerebrovascular Complications
in Patients with Cancer
10
Jaclyn E. Burch and Alan Z. Segal
Introduction Ischemic Stroke
The hypercoagulable state associated with cancer, Epidemiology

known since the descriptions of Trousseau in the
nineteenth century, has become a well-recognized Over 15 million Americans have or have had can-
cause of venous thromboembolism. It has only cer, making up approximately 3.5% of the popula-
more recently become apparent that cancer tion. In 2018, an estimated 1.7 million patients will
increases the risk of arterial thromboembolism as be diagnosed with new cases of cancer in the USA
well. In patients with cancer, cerebrovascular dis- [2]. In general, the risk of developing ischemic
ease is the second leading cause of lesions of the stroke is increased after a new diagnosis of most
central nervous system (CNS), only behind metas- cancer types, with the risk highest in the first sev-
tases [1]. The most frequent cause of cerebrovas- eral months after diagnosis. In a Swedish study,
cular disease in this population is stroke, both the overall risk of ischemic stroke in the first
ischemic and hemorrhagic; this chapter will focus 6 months after diagnosis was increased 1.6-fold
primarily on these entities. While patients with [3]. Incidence of stroke seems to correlate with the
cancer remain at risk for conventional mechanisms aggressiveness of the underlying cancer type, with
of cerebrovascular disease, cancer patients have the highest rates occurring in patients with newly
unique risk factors including complications of diagnosed lung, pancreas, colorectal, and nervous
coagulation disorders, direct tumor effects, toxic- system cancers and leukemia. The increased risk
ity of cancer treatment, and increased risk of infec- also correlates with stage of cancer—patients with
tion in the setting of immunosuppression. These metastatic disease at baseline demonstrate an ele-
features make this population distinct from the vated risk of ischemic stroke when compared to
general population and should be considered care- more localized disease [3–5]. Additionally, a cor-
fully during their evaluation and care. relation also exists between patients who present
with acute stroke and a concurrent or subsequent
diagnosis of cancer. Notably, a study using the
J. E. Burch
Department of Neurology, University of Rochester Nationwide Inpatient Sample found that approxi-
Medical Center, Rochester, NY, USA mately one in ten patients who presented with an
A. Z. Segal (*) acute ischemic stroke also had a diagnosis of can-
Department of Clinical Neurology, NewYork- cer [6]. Stroke can also be the presenting symptom
Presbyterian/Weill Cornell Medicine, of cancer, with occult malignancy being diagnosed
New York, NY, USA in as many as 0.4–3% at the time of presentation
e-mail: azs2001@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_10
140 J. E. Burch and A. Z. Segal
[7, 8]. Finally, in a Norwegian study, 4.3% of role in the increased risk of stroke from conven-
patients were diagnosed with cancer after initial tional mechanisms. This is due in part to shared
stroke, with a median time from stroke to cancer risk factors between cancer and stroke, (e.g.,
diagnosis of 14 months [9]. smoking) as well as cancer-specific mechanisms.
Historically, there has been a lack of consen- Patients with cancer have increased levels of sys-
sus regarding whether cancer itself is a separate temic inflammation, which is thought to result in
risk factor for stroke as both share common risk higher rates of atherosclerotic plaque formation
factors such as age, smoking, and obesity. [16]. Prior radiation exposure in patients with
However, recent studies have demonstrated that head and neck cancer predisposes to the develop-
this increased risk persists despite controlling for ment of radiation-induced vasculopathy, also
age and other cardiovascular risk factors [5]. In increasing risk of stroke from large vessel disease
addition, a higher risk of experiencing cerebral [10]. Patients with cancer may also have higher
ischemic events was found even in patients whose rates of atrial fibrillation. For example, women
cancers are not traditionally related to a history of diagnosed with atrial fibrillation were more likely
tobacco use, such as non-Hodgkin lymphoma [3, to go on to be diagnosed with malignant cancer
5]. Otherwise, the prevalence of traditional vas- [17]. Another study found that patients were
cular risk factors appears to be similar in patients more likely to have atrial fibrillation at the time
with stroke when comparing those with and with- of their initial cancer diagnosis [18].
out cancer [10–13]. The rate of cryptogenic stroke is higher in
patients with cancer [13, 15] and can be associ-
ated with a higher D-dimer, infarcts in multiple
Clinical Presentation vascular territories, as well as metastatic disease
at the time of the stroke [19]. The mechanism pos-
Typically, the clinical presentation of acute stroke tulated to unify these facts is stroke secondary to
is the acute onset of focal neurologic symptoms, nonbacterial thrombotic endocarditis (NBTE),
similar to patients in the general population. In a which is characterized by noninfectious fibrin
retrospective study evaluating ischemic stroke deposition on the heart valves and is thought to be
patients at Memorial Sloan Kettering Cancer increased in the setting of cancer-related hyperco-
Center, the most common presenting symptoms agulability. NBTE was found to be the most com-
were hemiparesis (78%), speech disturbance mon cause of symptomatic stroke in an autopsy
(51%), and visual field deficits (26%) [10]. study [1]. However, its presence can be difficult to
Another common symptom is encephalopathy, establish antemortem, and the clinical tools avail-
particularly because cancer patients have higher able to do so, including echocardiography, have
likelihood of embolic infarcts in multiple arterial relatively low sensitivity [20, 21]. Cerebral intra-
territories [14]. They may also have other areas of vascular coagulation, characterized by multiple
systemic thrombosis at the time of presentation, thrombotic cerebral infarcts without an embolic
including deep vein thrombosis (DVT) and pul- source, has also been reported as a rare cause of
monary embolism (PE) [11, 15]. Finally, stroke multifocal infarcts, usually occurring in patients
may also be incidentally found as a result of sur- with advanced disseminated disease [1].
veillance and screening brain imaging. Tumor emboli have also been described and
are another potential cause of embolic strokes
[22–24]. A patent foramen ovale (PFO), present
Pathophysiology in approximately 25% of the population, can
allow paradoxical embolism which is an impor-
Stroke Mechanisms tant source of stroke to consider in the cancer
As mentioned previously, patients with cancer population where increased rates of DVT are
have unique mechanisms of stroke. However, seen. Immunosuppression and invasive proce-
cancer and its treatment also play an important dures, such as indwelling catheters, may also pre-
10 Cerebrovascular Complications in Patients with Cancer 141
dispose patients to the development of infective strokes can cause contrast enhancement at the
endocarditis with resultant septic emboli [1]. subacute stage in the absence of any neoplasm.
As it may be impossible to differentiate between
Cancer-Related Hypercoagulability the two based on baseline imaging, follow-up
The hypercoagulable state induced in patients with imaging to assess interval change may be neces-
cancer plays a role in several of the mechanisms of sary. Rarely, biopsy is necessary to fully diagnose
stroke. This state is complex and not entirely the etiology. As in the general stroke population,
understood but is most certainly multifactorial. the pattern of infarct on imaging can clarify the
These effects include the host response to the can- underlying mechanism. For example, the distri-
cer, which includes the production of acute-phase bution of infarcts in multiple vascular territories
reactants, paraproteins, as well as inflammation suggests an embolic source such as atrial fibrilla-
and necrosis [25]. When white blood cells, like tion, NBTE, as well as septic, paradoxical, or
monocytes or macrophages, interact with cancer tumor emboli. Vascular imaging can be important
cells, certain factors such as tumor necrosis factor, to evaluate for sources of large vessel disease,
interleukin-1, and interleukin-6 are released. This especially if there is history of head and neck
leads to endothelial vessel wall damage, causing radiation, and to evaluate for evidence of mycotic
increased thrombogenicity [26]. Cancer cells also aneurysm in the setting of potential bacterial
produce factors promoting thrombosis themselves, endocarditis. Echocardiography can evaluate the
including tissue factor and cancer procoagulant, a structure of the heart, evaluating for stigmata of
protein that can directly activate factor X [25]. atrial fibrillation, the presence of a PFO, and any
Patients with malignancy are commonly noted marantic or septic endocarditis. Notably, trans-
to have abnormalities in laboratory tests of coagu- thoracic echocardiography (TTE) is limited in its
lation. Schwarzbach et al. found that patients with ability to detect marantic endocarditis given the
cancer had significantly higher D-dimer levels at typically small size of these vegetations.
the time of stroke [15]. These tests are often lim- Transesophageal echocardiography (TEE) is
ited in their ability to assist with diagnosis because superior to TTE for detecting a cardiac source of
the values for the common coagulation factors, stroke. A retrospective study by Merkler et al.
such as D-dimer, can overlap with those of cancer evaluated the diagnostic yield of TTE and TEE in
patients without stroke, lacking the sensitivity and cancer patients with ischemic stroke. Of those
specificity to direct therapy [27]. Cancer-mediated who were suspected to have a cardiac source,
hypercoagulability likely plays its most promi- TTE demonstrated a definite or possible cardiac
nent role during the first few months after cancer source in only 24% of patients, compared to 76%
diagnosis (when it is most severe) as well as in of patients who were evaluated with TEE [20].
aggressive metastatic disease. Compounding this, However, this increase in the diagnostic yield
some antitumor therapy may lead to hypercoagu- needs to be weighed carefully against the more
lability, including platinum compounds, high- invasive nature of the procedure, as well as the
dose fluorouracil, tamoxifen, mitomycin, and patient’s clinical stability and other comorbidi-
growth factors; the mechanism of this toxicity is ties. Rhythm monitoring, with ECGs and telem-
not well understood [26, 28]. etry, remains important, especially if the stroke
appears embolic in nature. There are often abnor-
malities in coagulation factors such as D-dimer,
Diagnosis although they are also often nonspecific in the
cancer patient and may not significantly alter
Magnetic resonance imaging (MRI) is crucial in management [27]. Physical examination should
the diagnosis of ischemic stroke, particularly in include evaluation for evidence of septic emboli
distinguishing between neoplastic and vascular and DVT. Limb venous Dopplers can be consid-
etiologies. Contrast enhancement is most often a ered, particularly if there is evidence of PFO on
sign of a neoplastic etiology; however, ischemic echocardiography.
Management worthwhile considering endovascular therapy

even if expected survival is limited.
Acute Treatment
The mainstay of acute treatment for ischemic Secondary Prevention
stroke is administration of IV tissue plasminogen Prevention of recurrent stroke hinges on identify-
activator (tPA) and consideration for endovascu- ing treatable mechanisms such as atrial fibrillation
lar therapy. Patients with cancer often have con- or carotid stenosis. However, a large proportion of
traindications for IV tPA such as coagulation strokes remain cryptogenic, especially in the can-
abnormalities, recent surgery, or systemic bleed- cer population. While mechanisms secondary to
ing. Per the 2018 American Stroke Association cancer-related hypercoagulability like NBTE may
guidelines for management of acute ischemic be suspected, these can be difficult to diagnose
stroke, patients with systemic malignancy may clinically. Furthermore, cancer-related hypercoag-
benefit from IV tPA in the absence of other con- ulability may theoretically play a role in exacer-
traindications [29]. Providers should also con- bating more conventional stroke mechanisms,
sider whether other laboratory testing such as such as large artery atherosclerosis, which is typi-
complete blood count for platelets or a coagula- cally treated with antiplatelet agents [36].
tion panel should be obtained prior to administra- The standard treatment for venous thrombo-
tion of thrombolytics. It should be noted that a embolism in patients with cancer is anticoagula-
structural gastrointestinal malignancy contributes tion, with prior studies demonstrating a benefit
to a higher risk of bleeding and is a contraindica- with low-molecular-weight heparin (LMWH)
tion to administration of IV tPA in the guidelines [37]. Recently, a study comparing the treatment
[29]. No prospective trials have specifically evalu- of venous thromboembolism with an oral factor
ated IV tPA in cancer patients; however, several Xa inhibitor (edoxaban) versus LMWH demon-
studies have demonstrated that rates of intracere- strated non-inferiority [38]. A non-inferiority
bral hemorrhage are similar to those found in non- trial of another factor Xa inhibitor, apixaban, ver-
cancer patients from the original IV tPA trials sus LMWH is ongoing and will be completed
[30–32]. Similarly, unstudied are patients with soon [39]. The treatment of ischemic stroke in
primary or metastatic brain tumors. A few case patients with cancer is less clear, and overall
series have surprisingly demonstrated relatively there is a lack of data to guide treatment. A few
low hemorrhage rates in this group [33, 34]. studies have provided a rationale for the use of
However, intratumoral hemorrhage remains a anticoagulation in the ischemic stroke associated
concern, and the presence of known intracranial with cancer-mediated hypercoagulability, dem-
intra-axial tumors is considered a contraindica- onstrating a reduction in levels of D-dimer with
tion for IV tPA in the 2018 guidelines. However, improved survival [40, 41]. Regarding the spe-
it may be reasonable to consider tPA in patients cific choice in anticoagulant, a retrospective
with small extra-axial tumors, such as meningi- study comparing enoxaparin with oral warfarin
oma [29]. Endovascular therapy is another treat- found no statistical difference for either recurrent
ment option to be considered in the acute setting, stroke or major bleeding [14].
although again, there are limited data in cancer However, anticoagulation carries its own risks
patients. A recent case series demonstrated a ben- and is often contraindicated in patients with coag-
eficial outcome for patients with cancer treated ulopathies and thrombocytopenia. Furthermore,
with endovascular therapy [35]. Several retro- patients with cancer already face an increased
spective studies have shown similar rates of mor- bleeding risk (10% risk of clinically relevant
tality and ICH after endovascular therapy in bleeding in the absence of anticoagulation), and
patients with cancer when compared to the gen- many already carry a history of prior major bleed-
eral population, although prospective data are ing [42]. LMWH, which has been the preferred
lacking [30, 31]. As patients with a large vessel choice of anticoagulant for oncologists for the
occlusion can have severe deficits, it may still be management of venous thromboembolism [42],
has additional drawbacks. In a recent feasibility was associated with increased recurrent stroke
study by Navi et al., cancer patients with ischemic rate [45]. High rates of recurrent stroke were also
stroke were randomized to secondary prevention seen in a Korean case-control study, where 28%
with enoxaparin versus aspirin. Sixty percent of of cancer patients had recurrent stroke within 1
patients in the LMWH group crossed over to the year of the index stroke [46].
aspirin group primarily because of injection-
related discomfort or cost [43]. Survival
The alternative to anticoagulation for second- Overall, strokes tend to be more severe in patients
ary prevention is treatment with antiplatelet with cancer, with a worse functional status and a
agents, primarily aspirin. Aspirin incurs a lower higher likelihood of developing early deteriora-
bleeding risk than anticoagulation and is often tion [47]. In a study by Cestari et al., the median
easier to administer and less expensive. There survival after ischemic stroke was 4.5 months.
may also be a cancer-specific rationale for aspi- Survival was worse if the stroke was embolic—
rin—platelets are postulated to promote cancer median survival was 2.6 months in patients with
growth and metastasis [44]. There are limited embolic appearing stroke, while it extended to
data comparing the use of antiplatelets to antico- 9.8 months in those with strokes of other etiolo-
agulation for the secondary prevention of isch- gies. In a more recent study by Navi et al., the
emic stroke in cancer patients. In a retrospective median survival for those with active cancer and
study by Cestari et al., there did not appear to be an identified stroke mechanism was 147 days,
a survival benefit for either treatment modality; a while only 55 days for those whose stroke was
more recent retrospective study at Memorial cryptogenic [13]. Predictors of worse outcome or
Sloan Kettering Cancer Center found no differ- mortality include stroke severity, metastatic dis-
ence in treatment effect with regard to recurrent ease, diabetes, cryptogenic etiology of stroke,
thromboembolic events [10, 45]. In the feasibility and elevated levels of C-reactive protein or
study by Navi et al., 20 out of 49 eligible patients D-dimer [13, 47, 48].
were enrolled, with a 60% crossover rate from
the enoxaparin group to the aspirin group. This
pilot did demonstrate feasibility, but with the rec- Intracranial Hemorrhage
ommendation that future trials use an oral antico-
agulant given the low rates of recruitment and Epidemiology
high crossover from LMWH. Data are still lack-
ing to definitively identify which agent would be In an autopsy study, cerebrovascular disease was
best for secondary prevention of stroke in the the second most common cause of complications
cancer population. involving the CNS, behind only intracranial
metastasis. About half of these cases were intra-
cranial hemorrhage (ICH) [1]. Several studies
Prognosis have sought to establish the incidence of ICH in
patients with both systemic and intracranial
Risk of Recurrence tumors, ranging between 2 and 14% of patients
Patients with cancer are at a higher risk of recur- [1, 49–51]. ICH in cancer patients is most fre-
rent thromboembolic events, including recurrent quently intraparenchymal/intracerebral (includ-
stroke. In a retrospective study at Memorial Sloan ing intratumoral hemorrhage). However, ICH in
Kettering Cancer Center, 31% of patients with cancer can be associated with hemorrhage into
cancer had a recurrent thromboembolic event at any intracranial compartment, with order of fre-
3 months after their index stroke. This includes quency from most common to least being intra-
13% with a recurrent acute ischemic stroke, parenchymal hemorrhage (IPH), subarachnoid
which is threefold higher than what is seen in the hemorrhage (SAH), subdural hemorrhage
general population. Adenocarcinoma histology (SDH), and epidural hemorrhage (EDH). In a
recent study, it was found that close to half (44%) trauma, hypertensive hemorrhage, hemorrhagic
of cancer patients with ICH experience hemor- conversion of ischemic stroke, venous thrombo-
rhage at multiple foci [52]. The most frequent sis, aneurysmal rupture (including mycotic and
causes of symptomatic ICH in patients with can- neoplastic aneurysms), and posterior reversible
cer are intratumoral hemorrhage (ITH) and encephalopathy syndrome (PRES). Acute leuke-
coagulopathy, which occur most commonly in mia can be associated with hyperleukocytosis
patients with solid tumors and hematologic (considered peripheral blast count >100,000/
malignancy, respectively. The hemorrhage loca- mm3), which can cause leukostasis with associ-
tion largely depends on the location of the tumor. ated local vessel destruction and rare parenchy-
Dural or skull-based tumors, such as prostate mal hemorrhage. Other causes, such as cerebral
metastases, can be associated with SDH or EDH amyloid angiopathy, are possible but uncommon
[1, 53]. [1, 27, 52].
Certain tumor types are more likely to be Coagulopathy can be seen in cancer patients
associated with ITH. Of solid tumors, melanoma, for many reasons. These include thrombocyto-
lung, breast, and renal cell cancer are most com- penia in the setting of hematologic malignancy,
monly associated with ICH [52]. This is partly treatment-related myelotoxicity, or tumor mar-
explained by a higher incidence of intracranial row infiltration. Coagulopathy can also be seen
metastasis among these cancer types [54, 55]. In secondary to platelet abnormalities and/or
addition, other solid tumors such as thyroid can- coagulation factor abnormalities secondary to
cer, hepatocellular carcinoma, and choriocarci- liver failure and poor nutrition with associated
noma are less common causes of intracranial vitamin K deficiency. Disseminated intravascu-
metastases, but seem to have a higher propensity lar coagulation (DIC) may also cause coagu-
to be associated with ITH [1, 50]. The individual lopathy. The mechanism in cancer is not
histology of these tumors is likely important; entirely understood, but is thought to be caused
increased intratumoral vascularization (resulting by inappropriate consumption of platelets and
in dilated, thin-walled vessels that are more likely coagulation factors secondary to excess pro-
to rupture) and necrosis predispose intracranial duction of thrombin. DIC is also seen in the
metastases to bleed [50]. Primary brain tumors, setting of sepsis, which bone marrow trans-
in particular glioblastoma, are associated with plant patients are particularly vulnerable to [1,
ITH. This is likely multifactorial due to angio- 27, 58, 59].
genesis, distal vessel necrosis, and dilation, dis- The anti-angiogenic agent bevacizumab, a
tention, and erosion of vessels by tumor growth monoclonal antibody that targets vascular endo-
[50, 56]. Oligodendrogliomas, although less thelial growth factor (VEGF), can also increase
common, seem to have an increased likelihood risk of hemorrhage. For a period of time, patients
for ITH, potentially related to infiltration of reti- with intracranial metastases were excluded from
form capillaries [57]. trials of bevacizumab after an early case report of
a young man with hepatocellular carcinoma
developed fatal ICH while being treated with
Mechanisms and Pathophysiology bevacizumab [60]. However, a study evaluating
bevacizumab at Memorial Sloan Kettering
The most commonly seen mechanisms of ICH in Cancer Center found similar rates of ICH in
cancer patients are ITH, seen in solid tumor bevacizumab-treated patients with and without
types, and coagulopathy, seen in hematological intracranial tumors, when compared to patients
malignancy. However, it is not uncommon for the not receiving bevacizumab [61]. Brain radiation
etiology of ICH to be multifactorial, including can cause delayed vascular abnormalities, includ-
when ITH is complicated by concomitant coagu- ing cavernous malformations, with subsequent
lopathy. Other less common causes include head ICH [58].
Clinical Presentation in areas outside the hemorrhage is also sugges-

tive of an intracranial metastasis underlying the
In a study performed at MSKCC, only about half hemorrhage. Diffusion restriction in setting of
(56.5%) of patients found to have ICH at autopsy hemorrhagic conversion of ischemic stroke as
were previously symptomatic [1]. However, in a well as multifocal lobar microhemorrhages (seen
more recent study assessing patients with radio- in patients with cerebral amyloid angiopathy)
graphic ICH, most patients (94%) were symp- can also be evaluated with MRI. Cerebral venous
tomatic at the time the imaging was performed sinus thrombosis may also be considered on the
[52]. This difference likely stems from the fact differential, which is discussed later in this
that the diagnosis of ICH in the cancer patient chapter.
occurs along a spectrum and can range from inci-
dentally found hemosiderin deposition to more
significant hemorrhage resulting in mass effect Management
and elevated intracranial pressure [62]. Patients
with primary brain tumors were more likely to An exhaustive discussion of the management of
demonstrate foci of asymptomatic hemorrhage acute ICH is outside the scope of this chapter,
when compared to patients with solid or hemato- and the standard guidelines for management of
poietic tumors [52]; this may be because primary ICH still apply [66, 67]. In the cancer population,
brain tumor patients often have pre-existing neu- steroids are used much more frequently in the
rologic deficits and also undergo more frequent management of ICH, as was seen in a retrospec-
surveillance imaging. Notably, only 5% of cancer tive study where 75% of patients were treated
patients were found to have acute hypertension with steroids after a diagnosis of ICH [52].
(defined as SBP >180 or DBP >120) at the time Steroids should be considered for the treatment
of ICH presentation. The most common present- of mass effect associated with vasogenic edema
ing symptoms were hemiparesis, headache, in the setting of ITH. For hemorrhage in the set-
impairment of consciousness, and seizure. ting of a coagulopathy, treatment should include
Relatively fewer patients (6%) presented with addressing the underlying hematologic abnor-
coma [63]. mality. In the case of DIC, the appropriate treat-
ment is unclear but generally includes treating
any underlying factors, such as sepsis, and may
Diagnosis include replacement of clotting factors with fresh
frozen plasma, cryoprecipitate, and platelets [27].
In cases where there is suspicion for ICH, Craniotomy can be considered for resection in
patients should be evaluated with a non-contrast the case of ITH, as surgical treatment has been
CT head (CTH). If there are no contraindications shown to be beneficial in the setting of a symp-
to contrast, a CT angiogram can later be obtained tomatic intracranial metastasis; this may be fol-
to evaluate for any vascular malformations. lowed by radiation [52, 68, 69]. If the lesion is
When clinically stable, a contrast-enhanced MRI not resectable or if there are multiple lesions,
should be performed in order to evaluate for then treatment with whole brain radiation or ste-
underlying tumor. Features suggestive of under- reotactic radiosurgery for palliation may be
lying tumor include multiple hemorrhage sites, considered [52]. In the setting of SDH or EDH,
areas of hemorrhage at the gray-white junction, surgical decompression would fall under the
more heterogeneous signal patterns, multiple same indications as it would in the general popu-
stages of a hematoma development in a single lation, albeit with careful consideration of safety
lesion, as well as delayed evolution of the hema- in the setting of any coagulopathies [66]. SAH
toma, excessive and persistent edema at the site should also be managed as per standard guide-
of the hematoma, and lack or absence of a well- lines with added consideration for increased like-
defined hemosiderin rim [64, 65]. Enhancement lihood of mycotic aneurysms given the increased
risk of infection in this population [67]. These by treatment with anticoagulation. Therapeutic
aneurysms are often not amenable to surgery and anticoagulation in patients with intracranial
are generally treated with antibiotics [52]. tumors, including both primary brain tumors and
Neoplastic aneurysms are rare, and although intracranial metastases, generally appears to be
treatment has not been clearly established, che- safe. Several studies have shown similar rates of
motherapy and radiation may be considered [70]. ICH and mortality compared to patients not treated
Finally, intermittent pneumatic compression with anticoagulation [39, 52, 62, 71–74].
devices for the legs should be initiated on the day
of hospital admission, while early treatment with
chemoprophylaxis for deep venous thrombosis CVST
with enoxaparin or heparin should be considered
1 to 4 days from onset of ICH if there is no fur- Cerebral venous sinus thrombosis (CVST) results
ther bleeding [66]. from an occlusion within the venous system of
the brain. The risk of CVST, like systemic deep
venous thrombosis, is increased in patients with
Prognosis cancer. In a recent case-control study, the risk of
CVST in cancer patients appeared to be about
The outcomes for cancer patients with ICH are fivefold higher. The risk was highest in the first
largely dictated by the prognosis of the underly- year after diagnosis and was also elevated in
ing malignancy. The 1-year mortality was 78% in patients with hematologic malignancies, as com-
a retrospective study by Navi et al. Median sur- pared to those with a solid malignancy. In fact,
vival was 3 months but was better for those with the risk of developing CVST was about 90 times
primary brain tumors (5.9 months), as compared higher in the first year after diagnosis of a hema-
to a 2.1-month median survival for those with tologic malignancy, when compared to the gen-
solid tumors and 1.5 months for those with hema- eral population [75]. The mechanism for this
tologic malignancies. Survival was also affected heightened risk is likely similar to the elevated
by the underlying mechanism of the hemorrhage, incidence of ischemic stroke shortly after diagno-
with longer survival seen in those suffering ICH sis, reflecting increased prothrombotic changes
secondary to ITH (3.7 months) or a combination in the setting of increased cancer activity, as well
of ITH and coagulopathy (1.8 months), as comas adverse effects of surgery and certain chemo-
pared to those with ICH secondary to coagulopa- therapies. In particular, l-asparaginase is thought
thy alone, where median survival was only to be associated with the development of CVST
0.3 months. Only 15% of patients were com- and may account in part for some of the increased
pletely independent at discharge, with 22% dying risk of CVST seen in hematologic malignancies
during admission. Patients with solid tumors had [27]. Mechanisms specific to the cancer patient
the best functional outcome at time of discharge, include cancer-induced hypercoagulability,
with 53% completely or partially independent. tumor compression or invasion, local infection,
Patients with hematologic malignancies were and chemotherapy-related toxicities [76]. The
most likely to die during the hospitalization, superior sagittal sinus is most frequently affected
while patients with primary brain tumors had the [75, 76], but thrombosis can also occur in the
lowest rate of in-hospital mortality [52]. other sinuses such as the inner cerebral or super-
ficial veins [77].
The most common presentation of CVST is
afety of Anticoagulation in Patients
S headache. Other common symptoms include sei-
with Intracranial Tumors zures, nausea/vomiting, altered level of con-
sciousness, and focal neurologic findings such as
Venous thromboembolism is commonly seen in weakness or aphasia [75, 76]. Increased density of
the cancer population, which is primarily managed the transverse or sagittal sinuses can indicate
thrombosis on non-contrast CTH, but the sensitiv- 7. Uemura J, Kimura K, Sibazaki K, Inoue T, Iguchi
Y, Yamashita S. Acute stroke patients have occult
ity of this sign is not high, and if there is clinical malignancy more often than expected. Eur Neurol.
concern for CVST, additional imaging should be 2010;64:140–4.
pursued [78]. On MRI, the clot may be acutely 8. Cocho D, Gendre J, Boltes A, et al. Predictors of
isointense on T1 and hypointense on T2 [78]. occult cancer in acute ischemic stroke patients. J
Stroke Cerebrovasc Dis. 2015;24:1324–8.
MRI may also demonstrate sequelae of the clot, 9. Selvik HA, Thomassen L, Bjerkreim AT,
such as intraparenchymal hemorrhage or infarc- Naess H. Cancer-associated stroke: the Bergen
tion. An infarction with associated hemorrhage NORSTROKE Study. Cerebrovasc Dis Extra.
that crosses vascular distributions or is near a 2015;5:107–13.
10. Cestari DM, Weine DM, Panageas KS, Segal AZ,
venous sinus is suggestive of CVST [77, 78]. CT DeAngelis LM. Stroke in patients with cancer: inci-
venography will show a filling defect, which in dence and etiology. Neurology. 2004;62:2025–30.
the sagittal sinus is referred to as the “empty delta 11. Zhang YY, Chan DK, Cordato D, Shen Q, Sheng
sign.” MR venography (MRV) will similarly dem- AZ. Stroke risk factor, pattern and outcome in
patients with cancer. Acta Neurol Scand. 2006;114:
onstrate a loss of flow in the area of the clot, with 378–83.
contrast-enhanced MRV better demonstrating the 12. Zhang YY, Cordato D, Shen Q, Sheng AZ, Hung WT,
venous structures than time-of-flight MRV [78]. Chan DK. Risk factor, pattern, etiology and outcome
Treatment should be aimed at the underlying in ischemic stroke patients with cancer: a nested case-
control study. Cerebrovasc Dis. 2007;23:181–7.
cause of CVST. If due to tumor compression or 13. Navi BB, Singer S, Merkler AE, et al. Cryptogenic sub-
invasion, then cancer treatment, either by chemo- type predicts reduced survival among cancer patients
therapy, radiation, or surgery, may be possible with ischemic stroke. Stroke. 2014;45:2292–7.
[39]. There are a limited number of randomized 14. Jang H, Lee JJ, Lee MJ, et al. Comparison of enoxapa-
rin and warfarin for secondary prevention of cancer-
trials of treatment of CVST in the general popula- associated stroke. J Oncol. 2015;2015:502089.
tion, but treatment guidelines generally recom- 15. Schwarzbach CJ, Schaefer A, Ebert A, et al. Stroke
mend anticoagulation, even in the setting of ICH, and cancer: the importance of cancer-associated
barring any other contraindications [78]. hypercoagulation as a possible stroke etiology. Stroke.
2012;43:3029–34.
16.
Tapia-Vieyra JV, Delgado-Coello B, Mas-Oliva
J. Atherosclerosis and cancer; a resemblance with far-
References reaching implications. Arch Med Res. 2017;48:12–26.
17. Conen D, Wong JA, Sandhu RK, et al. Risk of malig-
1. Graus F, Rogers LR, Posner JB. Cerebrovascular nant cancer among women with new-onset atrial
complications in patients with cancer. Medicine. fibrillation. JAMA Cardiol. 2016;1:389–96.
1985;64:16–35. 18. Guzzetti S, Costantino G, Vernocchi A, Sada S,

2. Noone AM, Howlader N, Krapcho M, Miller D, Brest Fundaro C. First diagnosis of colorectal or breast can-
A, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR, cer and prevalence of atrial fibrillation. Intern Emerg
Chen HS, Feuer EJ, Cronin KA, editors. SEER can- Med. 2008;3:227–31.
cer statistics review, 1975–2015. Bethesda: National 19. Kim SG, Hong JM, Kim HY, et al. Ischemic stroke
Cancer Institute; 2018.. https://seer.cancer.gov/ in cancer patients with and without conventional
csr/1975_2015/. Based on November 2017 SEER data mechanisms: a multicenter study in Korea. Stroke.
submission, posted to the SEER web site, April 2018 2010;41:798–801.
3. Zoller B, Ji J, Sundquist J, Sundquist K. Risk of 20. Merkler AE, Navi BB, Singer S, et al. Diagnostic
haemorrhagic and ischaemic stroke in patients with yield of echocardiography in cancer patients with
cancer: a nationwide follow-up study from Sweden. ischemic stroke. J Neuro-Oncol. 2015;123:115–21.
Eur J Cancer. 2012;48:1875–83. 21. Dutta T, Karas MG, Segal AZ, Kizer JR. Yield of
4. Navi BB, Reiner AS, Kamel H, et al. Association transesophageal echocardiography for nonbacterial
between incident cancer and subsequent stroke. Ann thrombotic endocarditis and other cardiac sources of
Neurol. 2015;77:291–300. embolism in cancer patients with cerebral ischemia.
5. Navi BB, Reiner AS, Kamel H, et al. Risk of arterial Am J Cardiol. 2006;97:894–8.
thromboembolism in patients with cancer. J Am Coll 22. Lefkovitz NW, Roessmann U, Kori SH. Major

Cardiol. 2017;70:926–38. cerebral infarction from tumor embolus. Stroke.
6. Sanossian N, Djabiras C, Mack WJ, Ovbiagele 1986;17:555–7.
B. Trends in cancer diagnoses among inpatients 23. Behrendt CE, Ruiz RB. Cerebral ischemic events

hospitalized with stroke. J Stroke Cerebrovasc Dis. in patients with advanced lung or prostate cancer.
2013;22:1146–50. Neuroepidemiology. 2005;24:230–6.
24. Navi BB, Kawaguchi K, Hriljac I, Lavi E, DeAngelis 42. Kamphuisen PW, Beyer-Westendorf J. Bleeding com-
LM, Jamieson DG. Multifocal stroke from tumor plications during anticoagulant treatment in patients
emboli. Arch Neurol. 2009;66:1174–5. with cancer. Thromb Res. 2014;133:S49–55.
25. Prandoni P, Falanga A, Piccioli A. Cancer and venous 43. Navi BB, Marshall RS, Bobrow D, et al. Enoxaparin
thromboembolism. Lancet Oncol. 2005;6:401–10. vs aspirin in patients with cancer and ischemic stroke:
26. Bick RL. Cancer-associated thrombosis. N Engl J
the TEACH Pilot Randomized Clinical Trial. JAMA
Med. 2003;349:109–11. Neurol. 2018;75:379–81.
27. Rogers LR. Cerebrovascular complications in patients 44. Li N. Platelets in cancer metastasis: to help the “vil-
with cancer. Semin Neurol. 2010;30:311–9. lain” to do evil. Int J Cancer. 2016;138:2078–87.
28. Li SH, Chen WH, Tang Y, et al. Incidence of isch- 45. Navi BB, Singer S, Merkler AE, et al. Recurrent

emic stroke post-chemotherapy: a retrospective thromboembolic events after ischemic stroke in
review of 10,963 patients. Clin Neurol Neurosurg. patients with cancer. Neurology. 2014;83:26–33.
2006;108:150–6. 46. Kim JM, Jung KH, Park KH, Lee ST, Chu K, Roh
29. Powers WJ, Rabinstein AA, Ackerson T, et al.
JK. Clinical manifestation of cancer related stroke:
Guidelines for the early management of patients retrospective case-control study. J Neuro-Oncol.
with acute ischemic stroke: a guideline for health- 2013;111:295–301.
care professionals from the American Heart 47. Kneihsl M, Enzinger C, Wunsch G, et al. Poor short-
Association/American Stroke Association. Stroke. term outcome in patients with ischaemic stroke and
2018;49:e46–e110. active cancer. J Neurol. 2016;263:150–6.
30. Masrur S, Abdullah AR, Smith EE, et al. Risk of throm- 48. Shin YW, Lee ST, Jung KH, et al. Predictors of sur-
bolytic therapy for acute ischemic stroke in patients vival for patients with cancer after cryptogenic stroke.
with current malignancy. J Stroke Cerebrovasc Dis. J Neuro-Oncol. 2016;128:277–84.
2011;20:124–30. 49. Kondziolka D, Bernstein M, Resch L, et al.

31. Murthy SB, Karanth S, Shah S, et al. Thrombolysis Significance of hemorrhage into brain tumors: clini-
for acute ischemic stroke in patients with cancer: a copathological study. J Neurosurg. 1987;67:852–7.
population study. Stroke. 2013;44:3573–6. 50. Lieu AS, Hwang SL, Howng SL, Chai CY. Brain
32. Tissue plasminogen activator for acute ischemic
tumors with hemorrhage. J Formos Med Assoc.
stroke. The National Institute of Neurological 1999;98:365–7.
Disorders and Stroke rt-PA Stroke Study Group. N 51. Schrader B, Barth H, Lang EW, et al. Spontaneous
Engl J Med. 1995;333:1581–7. intracranial haematomas caused by neoplasms. Acta
33. Etgen T, Steinich I, Gsottschneider L. Thrombolysis Neurochir. 2000;142:979–85.
for ischemic stroke in patients with brain tumors. J 52. Navi BB, Reichman JS, Berlin D, et al. Intracerebral
Stroke Cerebrovasc Dis. 2014;23:361–6. and subarachnoid hemorrhage in patients with cancer.
34. Murthy SB, Moradiya Y, Shah S, Shastri A, Bershad Neurology. 2010;74:494–501.
EM, Suarez JI. In-hospital outcomes of thrombolysis 53. Reichman J, Singer S, Navi B, et al. Subdural

for acute ischemic stroke in patients with primary hematoma in patients with cancer. Neurosurgery.
brain tumors. J Clin Neurosci. 2015;22:474–8. 2012;71:74–9.
35. Merkler AE, Marcus JR, Gupta A, et al. Endovascular 54. Schouten LJ, Rutten J, Huveneers HAM, Twijnstra
therapy for acute stroke in patients with cancer. A. Incidence of brain metastases in a cohort of
Neurohospitalist. 2014;4:133–5. patients with carcinoma of the breast, colon, kidney,
36. Navi BB, Iadecola C. Ischemic stroke in cancer
and lung and melanoma. Cancer. 2002;94:2698–705.
patients: a review of an underappreciated pathology. 55. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau
Ann Neurol. 2018;83:873–83. FD, Lai P, Sawaya RE. Incidence proportions of brain
37. Lee AYY, Levine MN, Baker RI, et al. Low-molecular- metastases in patients diagnosed (1973 to 2001) in the
weight heparin versus a coumarin for the prevention Metropolitan Detroit Cancer Surveillance System. J
of recurrent venous thromboembolism in patients Clin Oncol. 2004;22:2865–72.
with cancer. N Engl J Med. 2003;349:146–53. 56. Kaya B, Cicek O, Erdi F, et al. Intratumoral

38. Raskob GE, van Es N, Verhamme P, et al. Edoxaban hemorrhage-related differences in the expression of
for the treatment of cancer-associated venous throm- vascular endothelial growth factor, basic fibroblast
boembolism. N Engl J Med. 2018;378:615–24. growth factor and thioredoxin reductase 1 in human
39. Agnelli G, Becattini C, Bauersachs R, et al.
glioblastoma. Mol Clin Oncol. 2016;5:343–6.
Apixaban versus dalteparin for the treatment of 57. Liwnicz BH, Wu SZ, Tew JM Jr. The relationship
acute venous thromboembolism in patients with between the capillary structure and hemorrhage in
cancer: the Caravaggio Study. Thromb Haemost. gliomas. J Neurosurg. 1987;66:536–41.
2018;118:1668–78. 58. Rogers LR. Cerebrovascular complications in cancer
40. Seok JM, Kim SG, Kim JW, et al. Coagulopathy
patients. Neurol Clin N Am. 2003;21:167–92.
and embolic signal in cancer patients with ischemic 59. Velander AJ, DeAngelis LM, Navi BB. Intracranial
stroke. Ann Neurol. 2010;68:213–9. hemorrhage in patients with cancer. Curr Atheroscler
41. Lee MJ, Chung JW, Ahn MJ, et al. Hypercoagulability Rep. 2012;14:373–81.
and mortality of patients with stroke and active cancer: 60. Gordon MS, Margolin K, Talpaz M, et al. Phase

the OASIS-CANCER Study. J Stroke. 2017;19:77–87. I safety and pharmacokinetic study of recombi-
nant human anti-vascular endothelial growth fac- 69. Licata B, Turazzi S. Bleeding cerebral neoplasms

tor in patients with advanced cancer. J Clin Oncol. with symptomatic hematoma. J Neurosurg Sci.
2001;19:843–50. 2003;47:201–10.
61. Khasraw M, Holodny A, Goldlust SA, DeAngelis
70. Omofoye OA, Barnett R, Lau W, Trembath D, Jordan
LM. Intracranial hemorrhage in patients with can- JD, Sasaki-Adams DM. Neoplastic cerebral aneu-
cer treated with bevacizumab: the Memorial Sloan- rysm from metastatic nonsmall cell lung carcinoma:
Kettering experience. Ann Oncol. 2012;23:458–63. case report and literature review. Neurosurgery.
62. Weinstock MJ, Uhlmann EJ, Zwicker JI. Intracranial 2018;83:E221–5.
hemorrhage in cancer patients treated with anticoagu- 71. Choucair AK, Silver P, Levin VA. Risk of intracra-
lation. Thromb Res. 2016;140:S60–5. nial hemorrhage in glioma patients receiving anti-
63. Kamel H, Navi BB, Hemphill JC 3rd. A rule to iden- coagulant therapy for venous thromboembolism. J
tify patients who require magnetic resonance imag- Neurosurg. 1987;66:357–8.
ing after intracerebral hemorrhage. Neurocrit Care. 72. Yust-Katz S, Mandel JJ, Wu J, et al. Venous throm-
2013;18:59–63. boembolism (VTE) and glioblastoma. J Neuro-Oncol.
64.
Atlas SW, Grossman RI, Gomori JM, et al. 2015;124:87–94.
Hemorrhagic intracranial malignant neoplasms: spin- 73. Schiff D, DeAngelis LM. Therapy of venous throm-
echo MR imaging. Neuroradiology. 1987;164:71–7. boembolism in patients with brain metastases. Cancer.
65. Tung GA, Julius BD, Rogg JM. MRI of intracerebral 1994;73:493–8.
hematoma: value of vasogenic edema ratio for pre- 74. Donato J, Campigotto F, Uhlmann EJ, et al.

dicting the cause. Neuroradiology. 2003;45:357–62. Intracranial hemorrhage in patients with brain metas-
66. Hemphill JC 3rd, Greenberg SM, Anderson CS,
tases treated with therapeutic enoxaparin: a matched
et al. Guidelines for the management of sponta- cohort study. Blood. 2015;126:494–9.
neous intracerebral hemorrhage: a guideline for 75. Silvis SM, Hiltunen S, Lindgren E, et al. Cancer and
healthcare professionals from the American Heart risk of cerebral venous thrombosis: a case-control
Association/American Stroke Association. Stroke. study. J Thromb Haemost. 2018;16:90–5.
2015;46:2032–60. 76. Raizer JJ, DeAngelis LM. Cerebral sinus thrombo-
67. Connolly ES Jr, Rabinstein AA, Carhuapoma JR,
sis diagnosed by MRI and MR venography in cancer
et al. Guidelines for the management of aneurys- patients. Neurology. 2000;54:1222–6.
mal subarachnoid hemorrhage: a guideline for 77. Grisold W, Oberndorfer S, Struhal W. Stroke and can-
healthcare professionals from the American Heart cer: a review. Acta Neurol Scand. 2009;119:1–16.
Association/American Stroke Association. Stroke. 78. Saposnik G, Barinagarrementeria F, Brown RD Jr,
2012;43:1711–37. et al. Diagnosis and management of cerebral venous
68. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized thrombosis: a statement for healthcare profession-
trial of surgery in the treatment of single metastases to als from the American Heart Association/American
the brain. N Engl J Med. 1990;322:494–500. Stroke Association. Stroke. 2011;42:1158–92.
Mood Disorders in Patients
with CNS Metastases
11
Kaleena Chilcote
Abbreviations patients’ well-being. The National Comprehen-

sive Cancer Network (NCCN) has developed
AEDs Antiepileptic drugs guidelines to assist clinicians in assessing and
CBT Cognitive behavioral therapy managing patient distress. This guideline defines
CNS Central nervous system distress as “a multifactorial unpleasant experi-
DBT Dialectical behavior therapy ence of a psychological (i.e., cognitive, behav-
DSM-V Diagnostic and Statistical Manual of ioral, emotional), social, spiritual, and/or physical
Mental Disorders, fifth edition nature that may interfere with the ability to cope
HADS Hospital Anxiety and Depression effectively with cancer, its physical symptoms,
Scale and its treatment.” By this definition, multiple
HPA axis Hypothalamic-pituitary-adrenal axis factors contribute to one’s sense of well-being as
IL-2 Interleukin-2 well as to the development of mood and anxiety
IL-6 Interleukin-6 disorders, and all of these factors warrant moni-
MBSR Mindfulness-based stress reduction toring and intervention when appropriate [1]. As
MDD Major depressive disorder data is collected from studies using this definition
NCCN National Comprehensive Cancer and guidelines, the importance of addressing
Network mood disorders in patients with cancer is becom-
NSAIDs Nonsteroidal anti-inflammatory drugs ing more apparent. In patients with intracranial
PHQ-9 Patient Health Questionnaire-9 involvement, quality of life is more closely tied
to a patient’s sense of emotional well-being than
physical well-being [2]. Comorbid mood disor-
Introduction ders are associated with increased patient dis-
tress, lower quality of life, higher healthcare
Patient distress is becoming more widely costs, caregiver burden, other maladaptive
assessed as national agencies and credentialing health behaviors, and poorer cancer-related out-
bodies highlight the importance of monitoring comes.
K. Chilcote (*)
Psychosocial Oncology, Department of Behavioral
Medicine and Psychiatry, West Virginia University
Cancer Institute, Morgantown, WV, USA
e-mail: kaleena.chilcote@hsc.wvu.edu

https://doi.org/10.1007/978-3-030-42958-4_11
152 K. Chilcote
Depressive Disorders development of mood symptoms [12]. Depressive

disorders are most frequently associated with
Etiology lesions of the frontal and temporal lobes, though
there is no clear connection between depression
Multiple factors have been studied as potential and lesion location [13, 14]. Several syndromes
causes of depression with evidence suggesting caused by pathway disruption can present with
that development and perpetuation of depression symptoms that overlap with mood disorders. A
are multifactorial. There is increasing evidence dysexecutive syndrome with frontal lobe lesions
linking depression and inflammation in the body. impacting the dorsolateral prefrontal circuit pres-
Similarly, there is evidence linking cancer and ents with impairments in executive functioning
inflammation, perhaps providing a link between (perseveration, difficulty managing multiple and
higher rates of depression observed in patients new tasks). Patients may also experience psycho-
with cancer, particularly those cancer types asso- motor slowing, flattened affect, and impairments
ciated with more systemic inflammation. Cancer in self-care that resemble depression. Disinhibition
cells can produce multiple pro-inflammatory syndrome occurs with frontal lobe lesions impact-
mediators, including cytokines, chemokines, ing the orbitofrontal circuit and presents with
growth factors, and transcription factors. Cell emotional lability, impulsivity, and impaired
death resulting from cancer treatments, like radi- judgment that can mimic mood disorders, includ-
ation therapy and chemotherapy, leads to producing depression or a bipolar illness. Lesions of the
tion of cytokines that can trigger a cascade of anterior cingulate circuit can lead to apathy, which
immune responses [3, 4]. There is also a correla- also commonly mimics depression [15]. Multiple
tion between depression and elevated levels of primary psychiatric diagnoses have ties to dys-
interleukin-6 (IL-6) [5, 6]. In a study of women function in these circuits as well. This includes
with breast cancer, there was a clear association attention-deficit hyperactivity disorder, obsessive
between major depressive disorder and elevated compulsive disorder, Tourette syndrome,
IL-6 levels as well as consistent abnormalities on Huntington’s disease, and schizophrenia [15–17].
dexamethasone suppression testing, which sug-
gests a link between IL-6 and the hypothalamic-
pituitary-adrenal (HPA) axis [7]. The HPA access Epidemiology
has been studied extensively in its relation to
mood disorders. Growing evidence exists about The prevalence rates for depressive disorders in
how cancer might relate to HPA dysfunction. For patients with cancer vary and are often related to
example, women with ovarian cancer have been factors such as cancer type, disease stage, treat-
found to have higher evening cortisol levels than ment modalities, time from diagnosis, physical
controls [8]. The degree of causation in this rela- symptom burden, and patient demographics [18–
tionship remains unknown. Patients receiving 21]. For example, in patients with breast cancer,
immunotherapy with IL-2 and/or interferon-alfa predictors of depression include being in the year
were found to have lower levels of tryptophan, a following diagnosis, younger age, receiving adju-
precursor for serotonin, which suggests that cyto- vant chemotherapy, experiencing an impact on
kines might have a direct impact on the produc- fertility, and physical side effects from treatment
tion of neurotransmitters implicated in mood [22]. These variables, along with inconsistent
regulation [9]. These shared mechanisms between ways of defining and measuring depression, have
depression and cancer raise questions about the made it difficult to fully appreciate the impact of
potential interplay of these disorders and how depression on this patient population as a whole.
depression can impact cancer occurrence and Depressive symptoms as well as mixed anxiety/
progression [10, 11]. depressive symptom states have been found to be
Lesions involving the brain can disrupt impor- more common in certain cancer types, including
tant structures and pathways that also lead to the stomach, pancreatic, oropharyngeal, lung, and
11 Mood Disorders in Patients with CNS Metastases 153
gynecologic, and those with intracranial involve- are less closely tied to physical symptom burden.
ment [20, 23]. To date, studies of prevalence gen- This includes a deeper assessment of sadness,
erally focus on the impact of primary cancer type, tearfulness, social withdrawal, worthlessness,
and there is limited information specifically guilt, and suicidal ideation [28]. One must also
assessing the impact of central nervous system keep an open mind regarding other possible
(CNS) metastases. Overall, approximately 25% causes or contributors to the patient’s symptoms.
of patients with cancer have a depressive disorder Persistent depressive disorder is another
that warrants treatment, representing at least a depressive disorder that has been studied less for-
threefold increase compared to the general popu- mally in the cancer population but should remain
lation [20, 24–26]. on the differential diagnosis. With persistent
depressive disorder, formerly called dysthymia,
patients experience a depressed mood more days
Differential Diagnosis than not for a period of at least 2 years. They also
experience other symptoms of depression but
The term “depression” now has a wide range of have fewer requirements in order to meet criteria
meanings, varying from more social uses to when compared to MDD. Patients can experience
severely impairing symptoms that warrant inten- major depressive episodes superimposed on per-
sive treatment. The Diagnostic and Statistical sistent depressive disorder. This should be con-
Manual of Mental Disorders, currently in its fifth sidered in patients with periods of symptom
edition (DSM-V), provides a framework for con- exacerbation that improve but never fully resolve
ceptualizing mental health diagnoses and details between episodes [27]. There are no studies spe-
widely accepted diagnostic criteria for clinical cifically examining persistent depressive disorder
syndromes. Major depressive disorder (MDD) is in patients with CNS metastases and limited data
the most commonly referenced depressive disor- on the general cancer population.
der. To meet criteria for MDD, patients must have When depressive symptoms occur exclusively
at least five symptoms present for at least 2 weeks in the context of a stressor and cause impairment
with subsequent impairments in daily function- in daily life or functioning, an adjustment disor-
ing. Symptoms of depression include depressed der would be the most appropriate diagnosis [27].
mood or predominant irritability, decreased inter- This is common in patients who have cancer and
est in activities, significant change in appetite often warrants treatment approaches similar to
and/or weight, significant change in sleep, psy- that of MDD.
chomotor agitation or retardation, low energy/ For patients whose symptoms of depression
fatigue, feelings of worthlessness or guilt, are directly due to a substance or other medical
impaired attention and concentration, and sui- problem, the appropriate diagnosis may be sub-
cidal ideation. The depression also cannot be due stance-/medication-induced depressive disorder
to the effects of a substance, illicit or prescribed, or depressive disorder due to a general medical
or other medical condition [27]. condition [27]. Substances can be illicit, pre-
When working with patients with medical ill- scribed, over-the-counter, and/or supplements
ness, particularly cancer, it can be a challenge to and include intentional and accidental ingestions.
differentiate a physical complaint related to the Depression in a patient with at least one CNS
illness from a somatic manifestation of a mood metastatic lesion would be appropriately diag-
disorder. Consider a patient with cancer who suf- nosed in this category if the lesion itself is
fers from nausea leading to weight loss, impaired believed to be causing the symptoms.
sleep and irritability while on steroids, fatigue, The differential diagnosis for depression in
and difficulty with concentration and short-term patients with cancer is broad, and the etiology is
memory loss since starting chemotherapy. When often multifactorial. Factors that might contribute
providing a diagnosis for patients with cancer, to a depression-type picture and should be con-
greater stress might be placed on symptoms that sidered are as follows.
154 K. Chilcote
the care team and continue to impact patients

Potential contributors to depressed mood in into the future. In a study of 154 patients with
patients with cancer cancer who experienced delirium while hospital-
• Depressive disorder ized, 53.5% recalled their delirium, and the
• Bipolar disorder majority of these patients recalled this experi-
• Substance/Medication use ence as being highly distressing after resolution
• Alcohol [37]. Up to 90% of patients with cancer have
• AEDs delirium at the end of life [32]. When depression
• Interferon-alfa, IL-2 and delirium occur together, priority should be
• Corticosteroids given to addressing the causes of delirium, which
• Vitamin D deficiency are typically multifactorial in patients with can-
• Malnutrition cer [25, 34].
• Hypothyroidism In patients presenting with predominant cog-
• Low testosterone nitive complaints and possible mood disorder, it
• Pain is important to consider an underlying cognitive
• Cancer-related fatigue disorder in addition to other causes. A gradual
• Sleep disorders onset of impairments can often allow patients to
• Apathy compensate in day-to-day functioning. With the
• Demoralization increasing demands that come with a cancer
• Delirium diagnosis and treatment, such as managing new
• Dementia medications and frequent appointments, underly-
ing symptoms can be unmasked and become
more impairing.
Hypoactive delirium often masquerades as a
depressive disorder. Symptoms can include
blunted affect, emotional lability including tear- Bipolar Disorders
fulness, apathy, decreased involvement in daily
activities, apparent lack of motivation, low Bipolar disorders are differentiated from depres-
energy, decreased PO intake, decreased physical sive disorders by the presence of at least one epi-
activity, and impairments in attention/concentra- sode of hypomania or mania in a person’s lifetime.
tion. A waxing and waning course, alterations in Although depressive episodes typically occur at
level of consciousness, and perceptual distur- higher rates than manic episodes, history of a
bances can be helpful in distinguishing delirium depressive episode is not a requirement for a diag-
from depression. Risk factors for delirium in nosis of a bipolar disorder. As with depressive dis-
cancer include a number of factors commonly orders, the DSM-V identifies multiple diagnoses
associated with patients with CNS metastases: that help further classify the symptom profile and
history of delirium, advanced age, premorbid guide treatment decisions. These include bipolar I
cognitive impairment, intracranial disease disorder, bipolar II disorder, cyclothymic disor-
involvement, leptomeningeal disease, low albu- der, substance-/medication-induced bipolar disor-
min, dehydration, infection, hypoxia, recent sur- der, and bipolar disorder due to another medical
gery, cytokine release syndrome, comorbid bone condition. Hypomanic and manic episodes differ
or liver metastases, and use of steroids, benzodi- in their severity with hypomania lasting fewer
azepines, and opioids [29–34]. There is limited days and having a noticeable, but less impairing,
evidence to guide the management of agitated impact on daily functioning. Symptoms may
delirium associated with new immunotherapy include grandiosity, decreased need for sleep,
approaches [35, 36]. Delirium can be distressing increased and pressured speech, racing thoughts,
for patients, family members, and members of distractibility, an increase in goal-directed activi-
ties, and involvement in activities that are likely to Suicidality

have negative outcomes (i.e., risky financial deci-
sions, spending sprees, driving very fast, sexual Suicidal thoughts, attempts, and completions are
indiscretion, etc.) [27]. more common in patients with cancer compared to
Much as in depressive disorders, patients with a those without. Rates have also been found to be
bipolar illness are at an elevated risk of negative higher in the cancer population when compared to
health outcomes when compared to the general those with other medical illness, even when con-
population. This includes some factors that are trolling for expected prognosis [40]. Rates vary
associated with cancer, such as tobacco and alcohol widely across studies and highlight the challenges
use. However, there is no evidence that a patient of studying this heterogeneous patient population
with a bipolar illness is at a higher risk of develop- [41]. In general, the risk factors for suicidality that
ing cancer than others. There is also limited data apply to the general population also apply to
specifically looking at cancer-related outcomes in patients with cancer. Risk factors specific to
patients with an underlying bipolar disorder. patients with cancer include hopelessness indepen-
There are examples of hypomania/mania dent of depression, impaired physical functioning,
being caused by a medical condition. Some of the poor health overall, increasing stage of disease, and
most well-studied include stroke, traumatic brain specific primary cancer types such as CNS malig-
injury, multiple sclerosis, and disorders of adre- nancy [42–45]. There are mixed results on the
nal functioning [27, 38]. impact of gender in this population as a whole [41].
Some medications and other substances can There are no studies looking specifically at
lead to hypomania/mania and might be part of a suicidality in patients with CNS metastases, but
patient’s treatment while targeting cancer. Perhaps advanced stage of disease and involvement of a
the most well-known example is that of corticoste- primary CNS lesion both suggest that this popu-
roids. As previously mentioned, interferon- alfa lation is at increased risk. The highest rates occur
also has rarely caused mania and should be moni- close to the time of diagnosis [46]. Although
tored. Treatment-related mood symptoms are dis- there is consensus that suicidality generally
cussed later in this chapter. decreases over time following cancer diagnosis,
providers should always keep in mind that suicid-
ality can occur at any time. In a study of more
Anxiety Disorders than 720,000 breast cancer survivors, participants
continued to demonstrate elevated risk of suicide
Although the focus of this chapter is on mood compared to the general population, even
disorders, we cannot discuss mood disorders 25 years after cancer diagnosis [42]. Similarly,
without some mention of anxiety. Like depres- multiple studies show continued elevated risks in
sion, anxiety presents in patients with cancer at adult survivors of childhood cancers [43].
significantly higher rates than in the general pop-
ulation. When depression and anxiety symptoms
occur together, they are associated with more ood Symptoms Related to Cancer
M
severe depression, less robust response to treat- Treatments
ment, lower quality of life, poorer adherence to
mental health treatments, slower recovery, The side effects of specific chemotherapy agents
higher suicide rates, and higher overall health- will not be discussed in this chapter, but it should
care costs [23]. Studies also suggest that patients be noted that numerous neuropsychiatric side
with brain metastases have higher rates of anxi- effects are possible with cancer treatments. In
ety than depression, particularly at specific fact, receiving chemotherapy independently cor-
points in treatment, such as prior to initiating relates with rates of depression in the breast can-
radiation therapy [39]. cer population, regardless of the agent being used
156 K. Chilcote
[47]. This reinforces the importance of monitor- age experience “behavioral symptoms” that
ing for mood disorders in all patients receiving might include depression, anxiety, mood lability,
treatment. and agitation. One percent of adults developed
psychotic symptoms [57]. On the other hand,
many AEDs function as mood stabilizers and can
Hormonal Agents be beneficial in treating mood disorders.
The use of hormonal agents also increases the

risk for depression. There are clear links between Steroids
hormones and depressive symptoms in healthy
individuals. For example, mood disorders in Glucocorticoids have a clear association with the
women can have cyclical patterns related to men- onset of multiple psychiatric side effects includ-
ses, and women are at higher risk for depression ing depression, hypomania/mania, suicidal ide-
in the postpartum period and surrounding meno- ation, psychosis, delirium, and sleep changes
pause. There is mixed evidence about tamoxi- [38, 56]. Onset is often within the first couple of
fen’s effect on depression risk [47–50]. Perhaps weeks and dose-dependent but can occur after
unsurprisingly, patients with other risks factors long-term use. A diagnosis of primary bipolar
for depression have higher rates of developing disorder does not increase the risk of steroid-
depression while on tamoxifen [50]. Increased related mania. However, patients who have a
depressive symptoms also correlate with other history of this response to steroids are at an
physical symptoms, such as hot flashes and sex- increased risk, and prophylaxis with a mood sta-
ual dysfunction, both of which are more common bilizer for future treatments should be consid-
in women on tamoxifen compared to those who ered. One should not underestimate the impact
were not [51, 52]. steroid-related sleep impairments can have on a
patient’s functioning and sense of well-being.
This should be monitored closely and treated
Immunotherapy aggressively.
As immunotherapies become more commonly

used, there is increasing data about the neuropsy- Radiation Therapy
chiatric side effects, particularly in the acute
phase. Interferon-alfa is one of the most well- Chapter 29 of this book discusses the potential
known examples of a medication causing depres- neuropsychiatric impacts of radiation therapy in
sion and has warnings for the risk of suicidality. depth. These potential adverse outcomes cannot
Depression occurs in up to 58% of patients be overlooked. In a study of 170 patients with
receiving this medication. It should also be noted brain metastases undergoing whole brain radia-
that there is a lower, but still significant, risk of tion, self-reported measures of postradiation
mania associated with interferon-alfa use [53– symptoms showed a high prevalence of symptom
55]. IL-2 has also been associated with higher burden, most commonly fatigue, poor sense of
depression rates [56]. well-being, anxiety, drowsiness, and poor appe-
tite. They also found that symptoms tend to clus-
ter together—anxiety and depression are frequent
Antiepileptic Drugs covariables [58]. Distress measures show similar
patterns in patients undergoing whole brain or
The antiepileptic drug class (AEDs) as a whole hypofractionated stereotactic radiotherapy com-
has warnings about increased risk of depression, pared to those without brain mets undergoing
with rates varying between medications [56]. radiation to the breast [39]. Fatigue, a common
Clinical studies for oral levetiracetam show 13% side effect of radiation, can mimic depression in
of adults and 38% of those less than 18 years of this phase of treatment.
When a medication suspected of contributing Morbidity

to a mood disorder is an integral part of a patient’s
cancer treatment, it is often not feasible to dis- Cancer-related morbidity can similarly be
continue the medication. Providers should con- impacted by the presence of a mood disorder.
sider lowering the dose of the offending agent or Depression rates correlate with levels of anxi-
transitioning to another agent in the same class, if ety, fatigue, and pain [63, 64]. Distress is also
possible. It is important to consider the benefit of associated with other maladaptive behaviors,
psychiatric medications as adjuvant therapy, some of which have their own associated cancer
behavioral strategies, and lifestyle changes. risks, such as tobacco use [65]. Current depres-
sion is a risk factor for future psychiatric comor-
bidities, which can negatively impact a patient’s
I mpact of Mood Disorders progress. Patients with depression during hospi-
on Cancer-Related Outcomes talization following hematopoietic cell trans-
plant were found to have higher rates of
Engagement in Treatment post-traumatic stress disorder and lower quality
of life ratings at their 6-month follow-up visits
Psychosocial stress has been linked to multiple [66]. In a study of 154 patients admitted to the
factors that potentially play a role in cancer devel- hospital for surgery for thoracic and head and
opment or progression including inflammation, neck cancers, depression and fear of cancer
oxidative stress, decreased immune surveillance, recurrence were strongly associated with higher
and dysfunction of the HPA axis [10, 11]. nicotine relapse rates [67]. El-Jawahri et al.
Unsurprisingly, clinicians and researchers are compared 1116 patients with depression prior to
interested in how these relationships can impact allogeneic hematopoietic cell transplantation to
cancer-related outcomes in patients who struggle 6317 patients without pretransplant depression
with mood disorders. Studies show that there are and found higher rates of grade 2–4 acute graft-
differences in how patients make decisions related versus-host disease, lower overall survival rates,
to their cancer treatment. For example, in a study and fewer days alive and out of the hospital in
of women with breast cancer conducted by the first 100 days posttransplant in patients with
Colleoni et al., only 51.3% of women with comor- premorbid depression [68].
bid depression accepted the recommendation of
adjuvant chemotherapy compared to 92.2% of
women without depression [59]. Treatment adher- Mortality
ence rates also differ. Studies have found that
patients with depression are up to three times more It is challenging to study the impact of mood
likely to be nonadherent with medication recom- disorders on cancer-related mortality given the
mendations from their medical team [60]. Guilt is high number of confounding factors. However,
a common feeling in patients with cancer who studies have found that patients with higher
may fear they are a burden on others or somehow depressive symptom burden have shorter sur-
deserve illness because of a perception of previous vival times [69–73]. Also, having depression
wrongdoings—this has been found to be an inde- prior to cancer diagnosis correlates with lower
pendent risk factor for treatment nonadherence survival compared to those without precancer
[61]. As the treatment paradigm in cancer contin- depression. This difference is especially
ues to shift toward managing a chronic disease, prominent for patients with depression and

long-term follow-up and chronic medication use precancer physical limitations [74]. The etiol-
become more important. Kaul et al. noted that ogy of this relationship is likely multifactorial
young adult cancer survivors are approximately with potential impact from cancer treatment
twice as likely to report medication nonadherence nonadherence or maladaptive behaviors like

as their peers and that mental distress is a signifi- comorbid substance use as discussed previ-
cant risk factor for this behavior [62]. ously [10, 69].
158 K. Chilcote
Healthcare Utilization and Costs patients with earlier-stage disease. Studies specif-

ically looking at caregivers of patients with CNS
The impact of comorbid mood disorders and can- metastases are limited, but do show increased
cer can also be felt on a systems level. With a rates of depression and anxiety symptoms [81]. In
shifting focus toward patient satisfaction, we see addition to the vital role that caregivers play as
that depressive symptom severity inversely cor- part of the treatment team, evidence also reveals
relates with satisfaction in medical care [75]. an association between high levels of caregiver
Patient distress levels also correlate with the distress and high levels of patient distress.
number of reported concerns during an outpatient
oncology visit [76]. This translates to increased
time spent with members of the treatment team, Role of Screening
either through longer visits, more frequent visits,
or increased utilization of urgent and emergency Studies reveal that healthcare providers often fail
services [77]. Studies have clearly shown that to recognize patients who are experiencing emo-
mental health issues lead to higher healthcare tional distress, highlighting the importance of
costs as a whole. Implementation of appropriate routine screening for all patients [82]. As men-
treatment strategies that target mood disorders tioned, confounding factors related to cancer, can-
and anxiety lowers those costs [77, 78]. Also, cer treatment, and medical comorbidities can
studies show that proactive involvement of psy- make screening for mood disorders more chal-
chotherapy, particularly cognitive behavioral lenging. Many instruments are available, includ-
therapy (CBT) skills, can lead to higher quality ing some that have been validated for use
of life reports, fewer psychiatric symptoms, and specifically in patients with cancer, though no
lower healthcare costs, even in patients who did screening tools have been validated specifically in
not report elevated levels of distress at the time of patients with CNS metastases. This validation
diagnosis [78, 79]. This underscores the impor- occurs by comparing outcomes on the screening
tance of addressing mental health needs in all instrument with those of a gold standard tool,
patients. such as a standardized structured clinical inter-
view [83]. Identifying the most appropriate
screening tool requires assessing several factors
Interactions with Caregivers including the symptoms of primary interest,
patient population, clinic work flow, procedures
Caregivers can serve a wide range of functions, for who will administer and follow-up when a
providing emotional, cognitive, spiritual, physi- patient screens positive, available technology for
cal, and social support. The presence of brain administration and/or interpretation, available
metastases often corresponds with increasing care time, etc. Systematic reviews of English instru-
demands as patients develop new or worsening ments completed by Luckett et al. and Vodermaier
symptoms that impact daily life. The concept of et al. provide additional information on individual
caregiver burden has become a focus of research screening tools [84–86].
as patients with cancer live longer and the care- The Patient Health Questionnaire-9 (PHQ-9) is
giver role has correspondingly become more a self-report instrument with nine items that reflect
fluid, transitioning in focus from end-of-life care the diagnostic criteria for MDD outlined in the
to that of long-term survivorship. Being a care- DSM-V. Patients rate the severity of their symp-
giver correlates with higher levels of anxiety, toms in the past 2 weeks on a scale from 0 for “not
depression, social isolation, and concerns about at all” to 3 for “nearly every day” [87]. This was
financial stress and stigma related to the cancer developed for use in primary care and since vali-
[80]. Studies have shown that caregivers of those dated for use in patients with cancer [88].
with advanced cancer have higher rates of depres- The Hospital Anxiety and Depression Scale
sion and anxiety compared to those caring for (HADS) is a 14-item self-report tool commonly
used in research and clinical settings to screen for demonstrate benefit for patients in specific popu-
anxiety and depression symptoms in patients lations, data is limited in regard to patients with
with medical illness. This has been validated in a CNS metastases in particular.
wide range of patient populations, including Cognitive behavioral therapy (CBT), origi-
those with cancer, and has proven to be particu- nally developed to target depression, is a widely
larly reliable in screening for depression in this used form of psychotherapy. It focuses on identi-
population [89]. fying dysfunctional patterns of cognition, which
The NCCN Distress Thermometer has been often occur automatically and without awareness,
validated for use in patients with intracranial in order to change one’s emotional response and
tumors [82, 90]. It serves as a screening tool by behavior [93]. Evidence exists for using CBT in
asking patients to rate their distress on a scale patients with cancer to target many symptoms,
from 0 to 10 with 10 representing the highest including depression, fear of cancer recurrence,
level of distress. Patients also have the opportu- pain intensity, and fatigue [94–96].
nity to select areas in which they would like addi- Mindfulness-based stress reduction (MBSR),
tional support and/or resources by checking off developed by Jon Kabat-Zinn, has helped con-
topics on a Problem List. Areas include practical tribute to the rise in popularity of “mindfulness”
problems, family issues, emotional stress, spiri- practices in popular culture. Mindfulness is a
tual concerns, and physical ailments [1]. Although form of meditation that refers to a purposeful and
this instrument can gather information about a sustained focus on one’s self and the immediate
wider range of issues compared to the others dis- situation and/or surroundings to help bring focus
cussed, results are less easily correlated with spe- and clarity [97]. When incorporated into formal
cific diagnoses, and studies show that the distress treatment, this can involve multiple strategies,
detected correlates with anxiety more than such as individual meditation, guided medita-
depression [91, 92]. tions in person or through the use of pre-recorded
audio, body scans, and yoga [98]. This has been
studied in patients with cancer and found to be
Treatment Strategies helpful for many symptoms including overall
anxiety, fear of cancer recurrence, quality of life,
Comorbid mood disorders are best treated with a depression, cognitive symptoms, and physical
multidisciplinary approach that addresses patient tension [94, 98–100]. There is mixed evidence
needs while taking into account their inherent about the longevity of these benefits [99, 100].
strengths and weakness and the environment in Providers who teach these skills suggest they be
which they spend their time. Although therapy incorporated as a lifestyle change rather than a
and medication have independently been shown time-limited therapy.
to be effective for both unipolar and bipolar mood Motivational interviewing relies on a collab-
disorders, a comprehensive approach utilizing orative relationship between patient and provider
both tools should be encouraged. to help illicit and build upon one’s motivations for
change while honoring patient autonomy [101].
Although this style has been most studied in
Psychotherapy patients with substance use disorders, it is being
applied more widely over time. In patients with
Gathering comprehensive data on the effective- cancer, potential targets include optimizing diet,
ness of different therapy modalities for patients exercise, and lifestyle factors that impact sleep
with cancer has its challenges. Studies vary con- and fatigue, pain, mood, and substance misuse,
siderably in regard to the targeted symptoms, uti- among other aspects of daily life [102–104].
lized treatment modality, training of those Similarly, dialectical behavior therapy (DBT)
delivering the treatment, and the means of assess- has seen a significant broadening of applications
ing effectiveness [41]. While numerous studies since the original skills training manual was pub-
160 K. Chilcote
lished in 1993 [105]. Originally developed to depressive disorders. There is also less evidence
treat patients with borderline personality disor- specifically related to patients with CNS metasta-
der, this therapy modality focuses on four sets of ses. Despite this paucity in formal evidence, anti-
skills: mindfulness, interpersonal effectiveness, depressants are routinely used to manage both
emotion regulation, and distress tolerance [106]. depression and anxiety symptoms in this patient
This modality typically requires a greater time population. In fact, rates of medication use for
commitment each week, but should be strongly depression and anxiety in patients with cancer in
encouraged. the USA are typically about two times that of the
Additional therapy modalities have been general population, and these medications are
developed specifically to assist patients with used more frequently as disease progresses [115].
chronic medical illness and those facing the end Choosing an appropriate medication to target
of life. Dignity therapy was developed to help depression in patients with metastatic cancer
patients find meaning and hope as they approach requires attention to a number of factors:
death [107]. Meaning-centered psychotherapy,
both as individual and group modalities, is simi- 1. Primary symptom of interest: See Table 11.1
larly focused on assisting patients in finding and for information on the most commonly used
sustaining meaning [108–110]. antidepressants and considerations for their
use. Of note, there is limited evidence for the
use of stimulants as monotherapy to treat
Medications depression. If this is considered, it would be
wise to involve a psychiatric provider to assist
Before considering medication management to with proper use.
target mood disorders, it is important to evalu- 2. Other potential targets: While effective in
ate and address other contributing factors. The treating depression, antidepressants have
impact of comorbid substance use disorders other effects that might be beneficial and
should not be overlooked, and incorporating should be considered. Sleep, appetite, nausea,
screening for substance use is an integral part hot flashes, sexual dysfunction, and neuro-
of mental health care. Impairments in sleep pathic pain are the most common targets. See
correlate with depression risk, and treating Table 11.1 for examples. In addition to those
sleep disorders can result in lower depression listed, trazodone is an antidepressant that is
symptoms [111, 112]. Rates of sleep apnea are used off-label for insomnia. With less risk of
higher in patients with cancer compared to the tolerance or withdrawal and limited risk for a
general public, and sleep-disordered breathing paradoxical reaction more common in patients
correlates with increased mortality in cancer with CNS pathology, trazodone is often
patients, specifically [113]. It has also been viewed as superior to benzodiazepines for this
found to correlate to increased rates of cancer purpose. Primary caution is with orthostasis.
development, though there are many confound- 3. Potential problematic side effects: Patients
ing factors [114]. All patients should be with intracranial pathology are often more
screened for malnutrition, nutritional deficien- sensitive to medication side effects. In gen-
cies, and hypothyroidism. eral, starting at low doses and titrating slowly
is the best approach. It should be noted that
Antidepressants all serotonergic antidepressants have some
There is a robust body of evidence for using anti- risk for osteoporosis with long-term use, gas-
depressants to treat depression, including specifi- trointestinal bleeding through antiplatelet
cally for patients with cancer. There is less activity, and hyponatremia. Bupropion, which
evidence available to help guide treatment in acts by increasing norepinephrine and dopa-
patients who have symptoms of depression but mine, can be quite beneficial for some patients
do not fully meet diagnostic criteria for one of the by increasing daytime motivation/energy,
Table 11.1 Most commonly used antidepressants and considerations for use in patients with cancer
Primary mechanism of action Reasons to consider Cautions with use
Selective serotonin Inhibition of 5-HT reuptake Considered first-line Risk of headaches, GI upset,
reuptake inhibitors Generally well-tolerated sexual dysfunction
(SSRIs)
Citalopram
Escitalopram
Fluoxetine
Fluvoxamine
Paroxetine
Sertraline
Serotonin- Inhibition of 5-HT and Helpful for neuropathic Risk of HTN
norepinephrine reuptake norepinephrine reuptake pain
inhibitors (SNRIs) Activating impact of NE Discontinuation syndrome is
can increase motivation more prominent and requires
and daytime energy slower taper
Desvenlafaxine Venlafaxine for hot
Duloxetine flashes
Venlafaxine
Bupropion Inhibition of norepinephrine Helpful for smoking Risk of HTN, seizures
and dopamine reuptake cessation
Activating impact can Can exacerbate anxiety
increase motivation and
daytime energy
Off-label use for Can cause appetite
attentional issues suppression and weight loss
Low risk of sexual side Caution in psychotic
effects disorders
Less weight gain
Mirtazapine Inhibition of 5-HT2 and 5-HT3 Antiemetic properties Risk of dry mouth, weight
gain
Increased serotonin and Increases appetite Rare risk of neutropenia
norepinephrine through Sedating impact helpful through bone marrow
alpha-2 adrenergic antagonism for sleep suppression
Tricyclic antidepressants Inhibition of 5-HT and Helpful for neuropathic Anticholinergic, anti-
(TCAs) norepinephrine reuptake pain muscarinic, and anti-alpha
Amitriptyline Sedating impact helpful adrenergic side effects
Desipramine for sleep
Doxepin
Imipramine
Nortriptyline
improving attention, and aiding in smoking lead to the development of bothersome dry
cessation. However, it should be used with mouth and constipation as well as potentially
caution in patients with CNS metastases or more problematic effects like urinary reten-
primary brain tumors due to a dose-depen- tion, bowel ileus or obstruction, dental caries
dent risk of seizures [116]. When combining impacting oral intake, and cognitive impair-
medications, one should keep in mind the ment. There are also additional risks when
additive effects of side effect profiles. Use of combining multiple serotonergic medica-
anticholinergic medications is a common tions, such as tramadol, fentanyl, triptans, and
example in patients with cancer. As part of antiemetic agents, in addition to antidepres-
chemotherapy, pain, nausea, and psychiatric sants. Serotonin syndrome can present with
medication regimens, these medications can autonomic instability, altered mental status,
162 K. Chilcote
tremor, hyperreflexia, and myoclonus and can interactions than other psychotropics and should
progress to seizures, coma, or death if not be watched closely. Medication nonadherence
recognized and treated. can also be more detrimental. For example,
4. Drug-drug interactions: Providers should lamotrigine is classically known for its risk of the
always assess for possible drug-drug interac- life-threatening Stevens-Johnson syndrome dur-
tions before prescribing a new medication. ing dose titration. If a patient misses approxi-
When working with patients who have cancer, mately 5 consecutive days’ dosing, regardless of
it is important to consider what agents are the reason for this nonadherence, the dose must
typically used in the cancer treatment standard be re-titrated from the beginning of the titration
of care and make decisions accordingly. There schedule, which can have adverse effects on a
are numerous potential interactions between patient’s mood and behavior. Lithium can be a
psychiatric medications and other medica- powerful mood stabilizer but it is very reliant on
tions commonly used in cancer treatment. The consistent body water status. Lithium toxicity,
most frequently discussed drug-drug interac- which can be fatal, occurs more frequently with
tion in this category is that of tamoxifen and dehydration, infection, and multiple medication
paroxetine, a selective serotonin reuptake interactions, including the use of low-dose non-
inhibitor (SSRI). Tamoxifen is an inactive steroidal anti- inflammatory drugs (NSAIDs).
prodrug metabolized through the liver by Management with lithium in the context of can-
cyp2D6 into its active metabolites. Multiple cer requires close monitoring and should involve
antidepressants are inhibitors of this enzyme a psychiatric provider.
and pose a theoretical risk of decreasing the Antipsychotic medications also have mood
effectiveness of tamoxifen. Interestingly, stabilizing properties. Although most are used for
studies have not shown this to be true in clini- psychotic disorders and bipolar mania, there is
cal practice. In the largest study to date, Haque evidence to support off-label use for many indi-
et al. found that there was no correlation cations benefitting patients with cancer. This can
between antidepressant use and cancer recur- include use as an antiemetic, benzodiazepine-
rence or contralateral breast cancer diagnosis sparing sleep aid, appetite stimulant in failure to
in patients taking both an antidepressant and thrive, treatment for agitation or severe irritabil-
tamoxifen [117]. The risks and benefits of ity related to intracranial disease, and to treat
using this combination should be considered steroid-related mood disorders, anxiety, and
for each individual case. insomnia [120].
5 . Mechanism of delivery: Patients with cancer
often have temporary difficulty taking medi-
cations by mouth. In the USA, parenteral for- Conclusion/Summary
mulations are not as readily available [118,
119]. Patients may also have surgical inter- Patients with CNS metastases are at an increased
ventions or other medical issues that impact risk for mood disorders. This correlation is multi-
bioavailability of medication. Psychiatric pro- factorial, with contributions from shared mecha-
viders can be of assistance in these challeng- nisms on a cellular level, involvement of specific
ing cases. brain regions linked to the processing and gener-
ation of emotions, and side effects of cancer
Mood Stabilizers treatment to name a few. Comorbid mood disor-
There are multiple mood stabilizers that can be ders are linked to a number of poor cancer-related
used in the treatment of bipolar illness. If a outcomes and problematic behaviors, including
patient is currently stable on a psychiatric medi- medication nonadherence, comorbid substance
cation, it is advisable to avoid changes in this misuse, higher healthcare utilization and costs,
regimen as much as possible. This class of medi- and even mortality. Screening and early interven-
cation typically has more significant drug-drug tions are important and often involve collabora-
tion with mental health professionals to provide meta-analysis of published case studies. Expert Rev
Neurother. 2010;10(10):1529–36.
medications, psychotherapy, and other behav- 13. Valentine AD. Central nervous system tumors. In:
ioral strategies. Although a wide range of treat- Holland JC, Breitbart WS, Butow PN, Jacobsen
ment strategies are used in clinical practice, the PB, Loscalzo MJ, McCorkle R, editors. Psycho-
body of literature for this specific patient popula- oncology. New York: Oxford University Press;
2015. p. 87–91.
tion is small. Additional research is needed to 14. Rooney AG, Carson A, Grant R. Depression in cere-
provide evidence-based management recommen- bral glioma patients: a systematic review of observa-
dations for patients with CNS metastases. tional studies. J Natl Cancer Inst. 2011;102(1):61–76.
15. Bonelli RM, Cummings JL. Frontal-subcortical
circuitry and behavior. Dialogues Clin Neurosci.
2007;9(2):141–51.
References 16. Saxena S, Brody AL, Schwartz JM, Baxter
JR. Neuroimaging and frontal-subcortical circuitry
1. Holland JC, Bultz BD. The NCCN guideline for in obsessive-compulsive disorder. Br J Psychiatry
distress management: a case for making distress Suppl. 1998;35:26–37.
the sixth vital sign. J Natl Compr Cancer Netw. 17. Whiteside SP, Port JD, Abramowitz JS. A
2007;5(1):3–7. meta-analysis of functional neuroimaging in
2. Pelletier G, Verhoef MJ, Khatri N, Hagen N. Quality obsessive-compulsive disorder. Psychiatry Res.
of life in brain tumor patients: the relative contribu- 2004;132(1):69–79.
tions of depression, fatigue, emotional distress, and 18. Traeger L, Cannon S, Keating NL, Pirl WF, Lathan
existential issues. J Neuro-Oncol. 2002;57(1):41–9. C, Martin MY, et al. Race by sex differences in
3. Young K, Singh G. Biological mechanisms of cancer- depression symptoms and psychosocial service use
induced depression. Front Psych. 2018;9:299. among non-Hispanic black and white patients with
4. Jehn CF, Kuehnhardt D, Bartholomae A, Pfeiffer S, lung cancer. J Clin Oncol. 2014;32(2):107–13.
Krebs M, Regierer AC. Biomarkers of depression in 19. Zabora J, BrintzenhofeSzoc K, Curbow B, Hooker
cancer patients. Cancer. 2006;107:2723–9. C, Piantadosi S. The prevalence of psycho-
5. Breitbart W, Rosenfeld B, Tobias K, Pessin H, Ku logical distress by cancer site. Psychooncology.
GY, Yuan J, et al. Depression, cytokines, and pancre- 2001;10(1):19–28.
atic cancer. Psychooncology. 2014;23(3):339–45. 20. Linden W, Vodemaier A, Mackenzie R, Greig
6. Musselman DL, Miller AH, Porter MR, Manatunga D. Anxiety and depression after cancer diagnosis:
A, Gao F, Penna S, et al. Higher than normal plasma prevalence rates by cancer type, gender, and age. J
interleukin-6 concentrations in cancer patients with Affect Disord. 2012;14(2–3):343–51.
depression: preliminary findings. Am J Psychiatry. 21. Fann JR, Thomas-Rich AM, Katon WJ, Cowley D,
2001;158(8):1252–7. Pepping M, McGregor BA, et al. Major depression
7. Soygur H, Palaoglu O, Akarsu ES, Cankurtaran ES, after breast cancer: a review of epidemiology and
Ozalp E, Turhan L, et al. Interleukin-6 levels and HPA treatment. Gen Hosp Psychiatry. 2008;30(2):112–26.
axis activation in breast cancer patients with major 22. McFarland DC, Shaffer KM, Tiersten A, Holland
depressive disorder. Prog Neuropsychopharmacol J. Physical symptom burden and its association with
Biol Psychiatry. 2007;31(6):1242–7. distress, anxiety, and depression in breast cancer.
8. Lutgendorf SK, Weinrib AZ, Penedo F, Russell D, Psychosomatics. 2018;59(5):464–71.
DeGeest K, Costanzo ES, et al. Interleukin-6, cor- 23. Brintzenhofe-Szoc KM, Levin TT, Li Y, Kissane
tisol and depressive symptoms in ovarian cancer DW, Zabora JR. Mixed anxiety/depression symp-
patients. J Clin Oncol. 2008;26(29):4820–7. toms in a large cancer cohort: prevalence by cancer
9. Capuron L, Ravand A, Neveu PJ, Miller AH, Mues type. Psychosomatics. 2009;50(4):383–91.
M, Dantzer R. Association between decreased serum 24. Mitchell AJ, Chan M, Bhatti H, Halton M, Grassi
tryptophan concentrations and depressive symptoms L, Johansen C. Prevalence of depression, anxiety,
in cancer patients undergoing cytokine therapy. Mol and adjustment disorder in oncological, haemato-
Psychiatry. 2002;7(5):468–73. logical, and palliative-care settings: a meta-analysis
10. Bortolato B, Hyphantis TN, Valpione S, Perini G, of 94 interview-based studies. Lancet Oncol.
Maes M, Morris G, et al. Depression in cancer: the 2011;12(2):160–74.
many biobehavioral pathways driving tumor pro- 25. Fitzgerald P, Miller K, Li M, Rodin G. Depressive
gression. Cancer Treat Rev. 2017;52:58–70. disorders. In: Holland JC, Breitbart WS, Butow PN,
11. Spiegel D, Giese-Davis J. Depression and can- Jacobsen PB, Loscalzo MJ, McCorkle R, editors.
cer: mechanisms and disease progression. Biol Psycho-oncology. New York: Oxford University
Psychiatry. 2003;54(3):269–82. Press; 2015. p. 281–8.
12. Madhusoodanan S, Opler MG, Moise D, Gordon J, 26. Mitchel AJ, Chan M, Bhatti H, Halton M, Grassi L,
Danan DM, Sinha A, et al. Brain tumor location and Johnsen C, et al. Prevalence of depression, anxiety,
psychiatric symptoms: is there any association? A and adjustment disorder in oncological, haemato-
164 K. Chilcote
logical, and palliative-care settings: a meta-analysis national population-based study of 723,810 women.
of 94 interview-based studies. Lancet Oncol. J Natl Cancer Inst. 2006;98(19):1416–9.
2011;12(2):160–74. 43. Recklitis CJ, Diller LR, Li X, Najita J, Robison LL,
27. American Psychiatric Association. Diagnostic Zeltzer L. Suicide ideation in adult survivors of child-
and statistical manual of mental disorders. 5th ed. hood cancer: a report from the Childhood Cancer
Arlington: American Psychiatric Association; 2013. Survivor Study. J Clin Oncol. 2010;28(4):655–61.
28. Endicott J. Measurement of depression in patients 44. Chochinov HM, Wilson KG, Enns M, Lander
with cancer. Cancer. 1984;154(10):2243–7. S. Depression, hopelessness, and suicidal ide-
29. Kaplan JG, DeSouza TG, Shafran B, Pack D, Fuks ation in the terminally ill. Psychosomatics.
J, Portenoy R. Leptomeningeal metastases: compari- 1998;39(4):366–70.
son of clinical features and laboratory data of solid 45. Llorente MD, Burke M, Gregory GR, Bosworth HB,
tumors, lymphomas and leukemias. J Neuro-Oncol. Grambow SC, Horner RD, et al. Prostate cancer:
1990;9(3):222–9. a significant risk factor for late-life suicide. Am J
30. Ljubisavljjevic V, Kelly B. Risk factors for devel- Geriatr Psychiatry. 2012;12:195–201.
opment of delirium among oncology patients. Gen 46. Johnson TV, Garlow SJ, Brawley OW, Master
Hosp Psychiatry. 2003;25(5):345–52. VA. Peak window of suicides occurs within the first
31. Uchida M, Okuyama T, Ito Y, Nakaquchi T, Miyazaki month of diagnosis: implications for clinical oncol-
M, Sakamoto M, et al. Prevalence, course, and fac- ogy. Psychooncology. 2012;21(4):351–6.
tors associated with delirium in elderly patients with 47. Lee KC, Ray GT, Hunkeler EM, Finley
advanced cancer: a longitudinal observational study. PR. Tamoxifen treatment and new-onset depres-
Jpn J Clin Oncol. 2015;45(10):934–40. sion in breast cancer patients. Psychosomatics.
32. Lawlor PG, Gagnon B, Mancini IL, Pereira JL, 2007;48(3):205–10.
Hanson J, Suarez-Almazor ME, et al. Occurrence, 48. Thompson DS, Spanier CA, Vogel VG. The relation-
causes, and outcome of delirium in patients with ship between tamoxifen, estrogen, and depressive
advanced cancer: a prospective study. Arch Intern symptoms. Breast J. 1999;5(6):375–82.
Med. 2000;160(6):786–94. 49. Cathcart CK, Jones SE, Pumroy CS, Peters GN,
33. Gaudreau JD, Gagnon P, Harel F, Roy MA, Knox SM, Cheek JH. Clinical recognition and man-
Tremblay A. Psychoactive medications and risk agement of depression in node negative breast can-
of delirium in hospitalized cancer patients. J Clin cer patients treated with tamoxifen. Breast Cancer
Oncol. 2005;23(27):6712–8. Res Treat. 1993;27(3):277–81.
34. Tuma R, DeAngelis LM. Altered mental sta- 50. Day R, Ganz PA, Constantino JP. Tamoxifen and
tus in patients with cancer. Arch Neurol. depression: more evidence from the National
2000;57(12):1727–31. Surgical Adjuvant Breast and Bowel Project’s Breast
35. Gust J, Taraseviciute A, Turtle CJ. Neurotoxicity Cancer Prevention Randomized Study. J Natl Cancer
associated with CD19-targeted CAR-T cell thera- Inst. 2001;93(21):1615–23.
pies. CNS Drugs. 2018;32(12):1091–101. 51. Day R, Ganz PA, Constantino JP, Cronin WM,
36. Hay KA. Cytokine release syndrome and neuro- Wickerham DL, Fisher B. Health-related qual-
toxicity after CD19 chimeric antigen receptor- ity of life and tamoxifen in breast cancer preven-
modified CAR-T cell therapy. Br J Haematol. tion: a report from the National Surgical Adjuvant
2018;183(3):364–74. Breast and Bowel Project P-1 Study. J Clin Oncol.
37. Breitbart W, Gibson C, Tremblay A. The delir- 1999;17(9):2659–69.
ium experience: delirium recall and delirium- 52. Leon-Ferre RA, Majithia N, Loprinzi
related distress in hospitalized patients with CL. Management of hot flashes in women with
cancer, their spouses/caregivers, and their nurses. breast cancer receiving ovarian function suppres-
Psychosomatics. 2002;43(3):183–94. sion. Cancer Treat Rev. 2017;52:82–90.
38. Taylor DM, Barnes TE, Young AH. The Maudsley 53. Peginterferon alfa-2b [package insert]. Kenilworth:
prescribing guidelines in psychiatry. 13th ed. Wiley Schering Corporation; 2001.
Blackwell: Hoboken; 2018. 54. Lim C, Olson J, Zaman A, Phelps J, Ingram
39. Cordes MC, Scherwath A, Tahera A, Cole AM, Ernst KD. Prevalence and impact of manic traits in
G, Oppitz K, et al. Distress, anxiety and depression depressed patients initiating interferon therapy for
in patients with brain metastases before and after chronic hepatitis C infection. J Clin Gastroenterol.
radiotherapy. BMC Cancer. 2014;14:731–42. 2010;44(7):e141–6.
40. Miller M, Mogun H, Azrael D, Hempstead K, 55. Constant A, Castera L, Dantzer R, Couzigou P, de
Solomon DH. Cancer and the risk of suicide in older Ledinghen V, Demotes-Mainard J, et al. Mood alter-
Americans. J Clin Oncol. 2008;26(29):4720–4. ations during interferon-alfa therapy in patients with
41. Robson A, Scrutton F, Wilkinson L, MacLeod chronic hepatitis C: evidence for an overlap between
F. The risk of suicide in cancer patients: a review manic/hypomanic and depressive symptoms. J Clin
of the literature. Psychooncology. 2010;19(12): Psychiatry. 2005;66(8):1050–7.
1250–8. 56. Patten SB, Barbui C. Drug-induced depression:
42. Schairer C, Brown LM, Chen BE, Howard R, Lynch a systematic review to inform clinical practice.
CF, Hall P, et al. Suicide after breast cancer: an inter- Psychother Psychosom. 2004;73(4):207–15.
57. Levetiracetam [package insert]. Smyrna: UCB Inc; 71. Satin JR, Linden W, Phillips MJ. Depression as
2017. a predictor of disease progression and mortal-
58. Chow E, Fan G, Hadi S, Wong J, Kirou-Mauro A, ity in cancer patients: a meta-analysis. Cancer.
Filipczak L. Symptom clusters in cancer patients 2009;115(22):5349–61.
with brain metastases. Clin Oncol. 2008;20(1): 72. Hjerl K, Andersen EW, Keiding N, Mouridsen
76–82. HT, Mortensen PB, Jorgensen T. Depression as
59. Colleoni M, Mandala M, Peruzzotti G, Robertson a prognostic factor for breast cancer mortality.
C, Bredart A, Goldhirsch A. Depression and degree Psychosomatics. 2003;44(1):24–30.
of acceptance of adjuvant cytotoxic drugs. Lancet. 73. Lloyd-Williams M, Shiels C, Taylor F, Dennis
2000;356:1326–7. M. Depression – an independent predictor of early
60. DiMatteo MR, Lepper HS, Croghan TW. Depression death in patients with advanced cancer. J Affect
is a risk factor for noncompliance with medical treat- Disord. 2009;113(1–2):127–32.
ment: meta-analysis of the effects of anxiety and 74. Stommel M, Given BA, Given CW. Depression and
depression on patient adherence. Arch Intern Med. functional status as predictors of death among can-
2000;160(14):2101–7. cer patients. Cancer. 2002;94(10):2719–27.
61. Ayres A, Hoon PW, Franzoni JB, Matheny KB, 75. Bui QU, Ositir GV, Kuo YF, Freeman J, Goodwin
Cotanch PH, Takayanagi S. Influence of mood and JS. Relationship of depression to patient satisfaction:
adjustment to cancer on compliance with chemo- findings from the barriers to breast cancer study.
therapy among breast cancer patients. J Psychosom Breast Cancer Res Treat. 2005;89(1):23–8.
Res. 1994;38(5):393–402. 76. Goebel S, Stark AM, Kaup L, von Harscher M,
62. Kaul S, Avila JC, Mehta HB, Rodriguez AM, Kuo Mehdorn HM. Distress in patients with newly
YF, Kirchhoff AC. Cost-related medication non- diagnosed brain tumours. Psycho-Oncology.
adherence among adolescent and young adult cancer 2011;20:623–30.
survivors. Cancer. 2017;123(14):2726–34. 77. Bultz BD, Holland JC. Emotional distress in patients
63. Reddick BK, Nanda JP, Campbell L, Ryman DG, with cancer: the sixth vital sign. Commun Oncol.
Gaston-Johansson F. Examining the influence of 2006;3:311–4.
coping with pain on depression, anxiety, and fatigue 78. Carlson LE, Bultz BD. Efficacy and medical
among women with breast cancer. J Psychosoc cost offset of psychosocial interventions in can-
Oncol. 2005;23(2–3):137–57. cer care: making the case for economic analyses.
64. Reuter K, Classen CC, Roscoe JA, Morrow GR, Psychooncology. 2004;13(12):837–49.
Kirshner JJ, Rosenbluth R, et al. Association of cop- 79. Simpson JSA, Carlson LE, Trew M. Impact of a
ing style, pain, age and depression with fatigue in group psychosocial intervention on health care
women with primary breast cancer. Psychooncology. utilization by breast cancer patients. Cancer Pract.
2006;15(9):772–9. 2001;9(1):19–26.
65. Kaul S, Avila JC, Mutambudzi M, Russell H, 80. Rossi Ferrario S, Zotti AM, Massara G, Nuvolone
Kirchhoff AC, Schwartz CL. Mental distress and G. A comparative assessment of psychological and
health care use among survivors of adolescent and psychosocial characteristics of cancer patients and
young adult cancer: a cross-sectional analysis of their caregivers. Psychooncology. 2003;12(1):1–7.
the National Health Interview Survey. Cancer. 81. Saria MG, Courchesne NS, Evangelista L, Carter
2017;123(5):869–78. JL, MacManus DA, Gorman MK, et al. Anxiety and
66. El-Jawahri A, Vandusen HB, Traeger LN, Fishbein depression associated with burden in caregivers of
JN, Keenan T, Gallagher ER, et al. Quality of life patients with brain metastases. Oncol Nurs Forum.
and mood predict posttraumatic stress disorder after 2017;44(3):306–15.
hematopoietic stem cell transplantation. Cancer. 82. Keir ST, Calhoun-Eagan RD, Swartz JJ, Saleh OA,
2016;122(5):806–12. Friedman HS. Screening for distress in patients with
67. Simmons VN, Litvin EB, Jacobsen PB, Patel RD, brain cancer using the NCCN’s rapid screening mea-
McCaffrey JC, Oliver JA, et al. Predictors of smok- sure. Psychooncology. 2008;17(6):621–5.
ing relapse in patients with thoracic cancer or head 83. Jacobsen PB, Donovan KA. Assessment and screen-
and neck cancer. Cancer. 2013;119(7):1420–7. ing for anxiety and depression. In: Holland JC,
68. El-Jawahri A, Chen YB, Brazauskas R, He N, Lee SJ, Breitbart WS, Butow PN, Jacobsen PB, Loscalzo MJ,
Kknight JM, et al. Impact of pre-transplant depres- McCorkle R, editors. Psycho-oncology. New York:
sion on outcomes of allogeneic and autologous Oxford University Press; 2015. p. 378–83.
hematopoietic stem cell transplantation. Cancer. 84. Luckette T, Butow PN, King MT, Oguchi M, Heading
2017;123(10):1826–38. G, Hackl NA, et al. A review and recommendations
69. Brown KW, Levy AR, Rosberger Z, Edgar for optimal outcome measures of anxiety, depres-
L. Psychological distress and cancer survival: a sion and general distress in studies evaluating psy-
follow-up 10 years after diagnosis. Psychosom Med. chosocial interventions for English-speaking adults
2003;65(4):636–43. with heterogeneous cancer diagnoses. Support Care
70. Faller H, Bulzebruck H, Drings P, Lang H. Coping, Cancer. 2010;18(10):1241–62.
distress, and survival among patients with lung can- 85. Vodermaier A, Linden W, Siu C. Screening for
cer. Arch Gen Psychiatry. 1999;56(8):756–62. emotional distress in cancer patients: a systematic
166 K. Chilcote
review of assessment instruments. J Natl Cancer I. tematic review and meta-analysis. Acta Oncol.
2009;101(21):1464–88. 2017;56(12):1665–76.
86. Mitchell AJ, Meader N, Davies E, Clover K, Carter 100. Carlson LE, Tamagawa R, Stephen J, Drysdale E,
GL, Loscalzo MJ, et al. Meta-analysis of screening Zhong L, Speca M. Randomized-controlled trial
and case finding tools for depression in cancer: evi- of mindfulness-based cancer recovery versus sup-
dence based recommendations for clinical practice on portive expressive group therapy among distressed
behalf of the Depression in Cancer Care Consensus breast cancer survivors (MINDSET): long-term
Group. J Affect Disord. 2012;140(2):149–60. follow-up results. Psychooncology. 2016;25(7):
87. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: 750–9.
validity of a brief depression severity measure. J Gen 101. Miller WR, Rollnick S. Motivational interview-
Intern Med. 2001;16(9):606–13. ing: helping people change. 3rd ed. New York: The
88. Thekkumpurath P, Walker J, Butcher I, Hodges L, Guilford Press; 2013.
Kleiboer A, O’Connor M, et al. Screening for major 102. Spencer JC, Wheeler SB. A systematic review of
depression in cancer outpatients: the diagnostic motivational interviewing interventions in can-
accuracy of the 9-item patient health questionnaire. cer patients and survivors. Patient Educ Couns.
Cancer. 2011;117(1):218–27. 2016;99(7):1099–105.
89. Vodermaier A, Millman RD. Accuracy of the 103. Bennett JA, Lyons KS, Winters-Stone K, Nail
Hospital Anxiety and Depression Scale as a LM, Scherer J. Motivational interviewing to
screening tool in cancer patients: a systematic increase physical activity in long-term cancer sur-
review and meta- analysis. Support Care Cancer. vivors: a randomized controlled trial. Nurs Res.
2011;19(12):1899–906. 2007;56(1):18–27.
90. Goebel S, Mahdorn HM. Measurement of psycho- 104. Ream E, Gargaro G, Barsevick A, Richardson
logical distress in patients with intracranial tumours: A. Management of cancer-related fatigue during
the NCCN distress thermometer. J Neuro-Oncol. chemotherapy through telephone motivational inter-
2011;204(1):357–64. viewing: modeling and randomized exploratory trial.
91. Trask PC, Paterson A, Riba M, Brines B, Griffith Patient Educ Couns. 2015;98(2):199–206.
K, Parker P, et al. Assessment of psychological dis- 105. Cogwell Anderson R, Jensik K, Peloza D, Walker
tress in prospective bone marrow transplant patients. A. Use of the dialectical behavior therapy skills
Bone Marrow Transplant. 2002;29(11):917–25. and management of psychosocial stress with newly
92. Mitchell AJ. Pooled results from 38 analyses of the diagnosed breast cancer patients. Plast Surg Nurs.
accuracy of distress thermometer and other ultra- 2013;33(4):159–63.
short methods of detecting cancer-related mood dis- 106. Linehan MM. DBT skills training manual. 2nd ed.
orders. J Clin Oncol. 2007;25(29):4670–81. New York: The Guilford Press; 2015.
93. Beck JS. Cognitive behavior therapy. 2nd ed. 107. Chochinov HM. Dignity therapy. 1st ed. New York:
New York: The Guilford Press; 2011. Oxford University Press; 2012.
94. Chen D, Sun W, Liu N, Wang J, Zhao J, Zhang Y, 108. Breitbart WS, Poppito SR. Individual meaning-
et al. Fear of cancer recurrence: a systematic review centered psychotherapy for patients with advanced
of randomized, controlled trials. Oncol Nurs Forum. cancer: a treatment manual. 1st ed. New York:
2018;45(6):703–12. Oxford University Press; 2014.
95. Knoerl R, Lavoie Smith EM, Weisberg J. Chronic 109. Vos J, Vitali D. The effects of psychological
pain and cognitive behavioral therapy: an integrative meaning-centered therapies on quality of life and
review. West J Nurs Res. 2016;38(5):596–628. psychological stress: a meta-analysis. Palliat Support
96. Sandler CX, Goldstein D, Horsfield S, Bennett BK, Care. 2018;16(5):608–32.
Friedlander M, Bastick PA, et al. Randomized evalu- 110. Breitbart W, Pessin H, Rosenfeld B, Applebaum AJ,
ation of cognitive-behavioral therapy and graded Lichtenthal WG, Li Y, Saracino RM, Marziliano
exercise therapy for post-cancer fatigue. J Pain AM, Masterson M, Tobias K, Fenn N. Individual
Symptom Manag. 2017;54(1):74–84. meaning-centered psychotherapy for the treatment
97. Wolf C, Serpa JG. A clinician’s guide to teaching of psychological and existential distress: a random-
mindfulness: a comprehensive session-by-session ized controlled trial in patients with advanced can-
program for mental health professionals and health cer. Cancer. 2018;124(15):3231–9.
care providers. 1st ed. Oakland: New Harbinger 111. Campbell P, Tang N, McBeth J, Lewis M, Main
Publications; 2015. CJ, Croft PR, et al. The role of sleep problems in
98. Zhang MF, Wen Y, Liu WY, Peng LF, Wu XD, Liu the development of depression in those with per-
QW. Effectiveness of mindfulness-based therapy sistent pain: a prospective cohort study. Sleep.
for reducing anxiety and depression in patients 2013;36(11):1693–708.
with cancer: a meta-analysis. Medicine (Baltimore). 112. Berger AM, Wielgus K, Hertzog M, Fischer P, Farr
2015;94(4):e0897-0. L. Patterns of circadian activity rhythms and their
99. Haller H, Winkler MM, Klose P, Dobos G, Kummel relationships with fatigue and anxiety/depression in
S, Cramer H. Mindfulness-based interventions women treated with breast cancer adjuvant chemo-
for women with breast cancer: an updated sys- therapy. Support Care Cancer. 2010;18(1):105–14.
113. Nieto FJ, Peppard PE, Young T, Finn L, Hla KM, 117. Haque R, Shi J, Schottinger JE, Ahmed SA, Cheetham
Farre R. Sleep-disordered breathing and cancer mor- TC, Chung J, et al. Tamoxifen and antidepressant
tality: results from the Wisconsin Sleep Cohort Study. drug interaction among a cohort of 16887 breast can-
Am J Respir Crit Care Med. 2012;186(2):190–4. cer survivors. J Natl Cancer Inst. 2015;108(3):1–8.
114. Palamaner Subash Shantha G, Kumar AA, Cheskin 118. Stevens JR, Coffey J, Fojtik M, Kurtz K, Stern
LJ, Pancholy SB. Association between sleep- TA. The use of transdermal therapeutic sys-
disordered breathing, obstructive sleep apnea, and tems in psychiatric care: a primer on patches.
cancer incidence: a systematic review and meta- Psychosomatics. 2015;56(5):423–44.
analysis. Sleep Med. 2015;16(10):1289–94. 119. Kaminsky BM, Bostwick JR, Guthrie SK. Alternate
115. Hawkins NA, Soman A, Buchanan Lunsford N, routes of administration of antidepressant and
Leadbetter S, Rodriguez JL. Use of medications for antipsychotic medications. Ann Pharmacother.
treating anxiety and depression in cancer survivors in 2015;49(7):808–17.
the United States. J Clin Oncol. 2017;35(1):78–85. 120. Goldman LS, Goveas J. Olanzapine treat-
116. Bupropion hydrochloride [package insert]. ment of corticosteroid-induced mood disorders.
Greenville: GlaxoSmithKline; 2017. Psychosomatics. 2002;43(6):495–7.
Leptomeningeal Disease
and the Role of Intrathecal
12
Therapy
Fadi Saadeh and Adrienne Boire
Introduction most surface of the brain parenchyma lies the glia

limitans: a layer of astrocytic end feet that proj-
The meninges (meninx, Greek for membrane) are ects on the pia mater cells to create an additional
complex connective tissue structures that sur- protective barrier [6]. The glia limitans permis-
round the brain and spinal cord. Embryologically sively allows size-dependent passage of select
derived from meningeal mesenchyme, the menin- molecules from the CSF to the brain parenchyma
ges are divided into the leptomeninx (thin mem- [6–9].
brane) that houses the pia, arachnoid mater and Unlike the dura mater, supplied by the sys-
CSF, and the pachymeninx (tough membrane) or temic circulation, the leptomeningeal blood sup-
dura mater. The dura mater is a well-innervated, ply arises from the anterior, middle, and posterior
highly vascularized collagenous membrane that cerebral arteries before penetrating the brain
contains lymphatics [1, 2]. Beneath the dura lies parenchyma. The leptomeningeal space enjoys a
the multilayered arachnoid mater. This mem- somewhat complex relationship with the sys-
brane encases the CSF-filled subarachnoid space temic circulation. The leptomeninges reside
creating a cellular barrier through tight junctions behind the blood-CSF barrier, consisting of the
[3, 4]. Adjacent to the brain parenchyma, a one to choroid plexus epithelium (Fig. 12.1). CSF circu-
two cell-layer membrane, the pia mater, covers lating through the leptomeninges is absorbed via
the brain and spinal cord. Intimately associated the arachnoid granulations where it returns to the
with the nervous tissue, the pia extends into sulci venous system. Small molecules may enter and
and fissures, delves deep into the parenchyma, exit the parenchyma via perivascular (Virchow-
and reflects on subarachnoid vessels. Fibroblast- Robin) spaces [10]; the functional relevance of
like cells produce collagen bundles that along these pathways remains an area of active study.
with trabeculae connect the two layers of the lep- Spread of cancer cells into the leptomenin-
tomeninges [5]. Between the pia and the outer- geal space is described as leptomeningeal
metastasis (LM). Historically, this pathophysio-
logic entity has been described as “carcinoma-
F. Saadeh
Department of Neurology, Memorial Sloan Kettering tous meningitis,” “meningeal carcinomatosis,”
Cancer Center, New York, NY, USA and/or “leptomeningeal carcinomatosis.” We
A. Boire (*) prefer the more inclusive term leptomeningeal
Human Oncogenesis and Pathology Program, metastasis as it encompasses all malignancies
Department of Neurology, Memorial Sloan Kettering and remains agnostic to the role that inflamma-
Cancer Center, New York, NY, USA tion may play in this pathophysiology. LM
e-mail: boirea@mskcc.org

https://doi.org/10.1007/978-3-030-42958-4_12
170 F. Saadeh and A. Boire
Fig. 12.1 Arterial and Arterial blood

venous circulations
communicate through
compartments across the
Blood-brain-barrier Blood-CSF-barrier
blood-brain and
(vascular endothelium) (choroid epithelium)
blood-CSF barriers.
(Adapted from Malcolm
B Carpenter. Human
Neuroanatomy. 7th ed. Intracellular Interstitial CSF-filled
United States: compartment compartments compartments
Baltimore: Williams &
Wilkins, ©1976;
Carpenter’s Human
Neuroanatomy) Venules and Arachnoid villi
veins
Cerebral veins
Venous blood, dural

sinus and spinal veins
occurs in approximately 5–8% of patients with rising incidence of LM, which varies by primary
solid tumors and 5–15% of those with hemato- tumor type. The most common cancers that
logic malignancies [11]. Cancer cells may gain develop LM are breast cancer (12–34%) partic-
access to CSF compartments through four poten- ularly lobular carcinoma [16–18], lung cancer
tial routes: spread through Bateson’s plexus via especially NSCLC (3–5%) [19], acute non-
the venous circulation [12], pass through cho- lymphocytic leukemia (5–15%) [20, 21], non-
roid plexus via arterial circulation [13], direct Hodgkin lymphoma (6%) [22, 23], melanoma
invasion of spinal and cranial nerves [13] or (5–25%), GI malignancies (4–14%), and
brain parenchyma through direct penetration of unknown primaries (1–7%) [24–27]. Tumors
the glia limitans [14] (Fig. 12.2). Once within with lower LM predilection are mycosis fungoi-
the leptomeninges, cancer cells face an addi- des [28–30], multiple myeloma [31, 32], squa-
tional challenge—survival within the nutrient- mous cell carcinoma [33], thyroid cancer [34],
poor CSF. Employing animal models, we have rectal cancer [35], carcinoid [36], rhabdomyo-
recently found that cancer cells upregulate com- sarcoma [37], CLL [38], and neuroblastoma
plement C3. Focal generation of the split product [39]. However, it cannot be overstated that any
C3a leads to the loss of blood-CSF barrier integ- malignancy may seed the leptomeningeal space.
rity, enriching CSF composition [15].
rognosis of Leptomeningeal
P
he Incidence of LM in Different
T Metastasis in Different Cancers
Cancers
Median overall survival in LM patients is poor
The improvement in overall survival of patients and ranges between 6–8 weeks untreated and up
with metastasis at other sites and advances in to 4 months with treatment [40–43]. There is a
diagnostic techniques have contributed to the wide range of outcomes for treated patients. For
12 Leptomeningeal Disease and the Role of Intrathecal Therapy 171
Fig. 12.2 Metastatic cancer cells may employ four por- invasion along the nerve roots, and 4. invasion from the
tals of entry to access the leptomeningeal space: 1. hema- parenchyma
togenous arterial route, 2. hematogenous venous route, 3.
instance, reported overall survival for breast can- tion of characteristic findings on MR imaging of
cer patients ranges from 1.75 to 4.5 months with the brain and spinal cord, and (3) CSF examina-
a 1-year survival rate of 16–24%. Outcomes are tion. We therefore recommend that all patients
less favorable in lung cancer (average 3–6 months suspected of LM undergo formal neurologic
and 1-year survival of 19%) and melanoma examination, MR imaging of the brain and spine,
(1.7–2.5 months and 1-year survival rate of 7%) as well as CSF sampling.
[44–62]. Prognosis can be stratified according to
risk (refer to treatment section) with performance
status and systemic disease burden as robust Signs and Symptoms
prognostic factors. In the era of targeted thera-
pies, molecular subtypes play a major role in Clinical signs and symptoms in patients with
determining a patient’s prognosis; e.g., patients LM range from subacute to acute and present
with LM from HER-2-positive breast cancer within days to weeks. Leptomeningeal metasta-
demonstrate longer median overall survival as sis presents with protean manifestations; while
compared to their triple negative counterparts unsurprising given the ubiquity of the leptomen-
(5.2 vs. 2.5 months) [63]. ingeal space over the central nervous system,
certain signs and symptoms are characteristic
and should raise clinical suspicion. Multifocal
Diagnostic Scheme neurologic signs and symptoms in a patient with
or without a primary malignancy should raise
The diagnosis of LM rests on three pillars: (1) the suspicion of LM. For instance, 64% of
neurologic signs and/or symptoms consistent patients with LM usually present with multifocal
with leptomeningeal localization, (2) demonstra- signs and symptoms [27, 64].
Cerebral due to oculomotor, trochlear, or abducens nerve

Patients harboring LM most commonly present involvement [48, 64, 71]. LM deposits on the tri-
with headache [24, 27]. This can be due to ele- geminal nerve may elicit facial pain or numbness.
vated ICP or meningeal irritation. With the for- Involvement of the mandibular division causes the
mer, patients present with nausea, vomiting, and “numb chin syndrome” reported by up to 22% of
dizziness precipitated by a change in head posi- LM patients [73, 74]. Facial nerve involvement
tion (plateau waves) [65, 66]. Although no spe- typically causes a lower motor neuron palsy affect-
cific headache location or pattern is specific to a ing the upper and lower face; patients with meta-
LM diagnosis, severe episodic headaches consis- static cancer presenting with Bell’s palsy merit
tent with elevated and labile ICP should raise additional workup to address possible LM. Less
concern. Funduscopic examination can reveal common symptoms include sensorineural hearing
papilledema depending on the degree and dura- loss (vestibulocochlear nerve) in less than 5% of
tion of ICP elevation. Brain MRI may demon- patients and brainstem involvement (vagus and
strate hydrocephalus due to obstruction of CSF glossopharyngeal nerve) manifesting in dysar-
egress or infiltration of arachnoid villi. In severe thria, dysphagia, and/or hoarseness.
cases, tentorial herniation may occur, notably in
leptomeningeal leukemia [67]. Even without Spinal
hydrocephalus, CSF flow dynamics in LM may Leptomeningeal involvement at the level of the
remain abnormal [68]. Headaches can also be spine may result in radiculopathies presenting as
caused by direct meningeal irritation, which can lower extremity weakness, numbness, and absent
elicit afebrile neck stiffness and meningeal signs reflexes. A cauda equina syndrome (diminished
such as Kernig’s and Brudzinski’s signs. Lateral rectal tone, urinary retention, constipation, sad-
and midline cerebellar signs and symptoms can dle anesthesia) may be present. Asymptomatic
be present in up to 20% of LM patients including bladder enlargement is frequently found and is
vertigo, nausea, and gait disequilibrium [24, 27]. the most characteristic bladder pathology in LM
LM tumors may invade Virchow-Robin spaces patients [25].
into the parenchyma or remain perivascular caus-
ing disruption of the brain’s vasculature and elec-
trical activity [69]. A common symptom of LM is Diagnosis
a change in mental status including memory loss,
personality changes, and disorientation. While The above neurologic signs and symptoms are
these symptoms are not specific to LM, they often difficult to differentiate from treatment effects
could indicate underlying cerebral dysfunction, or primary disease. In the setting of such presenta-
undiagnosed/subclinical seizures, or hydroceph- tions, a high degree of suspicion for LM is appro-
alus. In some patients with LM, angiography priate and formal diagnosis is recommended.
shows partial occlusion or complete obliteration
that leads to transient ischemic attacks or strokes Neuroimaging
[70–72]. Up to 25% of LM patients develop sei- MRI is considered the most sensitive method for
zures, most commonly partial with secondary detecting LM [75]. With a specificity of 77% and
generalization. Seizures can be due to cortical sensitivity of 75% [76], in the appropriate clini-
irritation, local edema, or parenchymal invasion cal context, an MRI finding of leptomeningeal
[24, 48]. enhancement confirms LM diagnosis [77–79].
Neuroimaging findings in LM can be divided into
Cranial Nerves two groups [80]: (1) diagnostic features, includ-
The cranial nerves pass through the subarachnoid ing leptomeningeal enhancement, subependymal
space; symptoms involving greater than one cra- enhancement, and multiple nodules in the
nial nerve suggest LM. The most common cranial vertebral canal and ventricles, and (2) suggestive
nerve sign in LM patients is diplopia, which can be features, including nodular enhancement over the
cerebral cortex, metastatic lesion(s) approaching low ICP readings [29]. Importantly, low or zero
sulci and gyri, dural enhancement in the intracal- pressures can also be seen in patients with com-
varium or vertebral canal, bulky metastasis inside plete spinal blocks which occur late in the course
or in proximity to ventricles, direct invasion to of the disease.
the intracalvarium by head and neck malignancy,
cranial nerve enhancement, or communicating Cell Count
hydrocephalus [81]. Diagnostic features are non- CSF leukocyte count is typically increased in
specific and must be interpreted with caution. patients harboring LM. Leukocytic infiltrate is
Contrast-enhanced T1 images have the high- typically dominated by lymphocytes. However,
est specificity (93%) and sensitivity (59%) for other profiles may occur—eosinophils have been
detecting LM compared to other MRI sequences found in CSF samples of leptomeningeal metas-
[82]. MRI should not be limited to T1 tasis from lymphoma [88], Hodgkin’s disease
post-
contrast sequences. Consensus guidelines [89], and an unidentified epithelial tumor [90],
(EANO, ESMO) recommend that cerebral MRI while CSF basophils have been found with lepto-
should include axial T1-weighted, axial FLAIR, meningeal leukemia [91].
axial diffusion, axial T2-weighted, post-
gadolinium 3D T1-weighted, and post-Cytology
gadolinium 3D FLAIR sequences; spinal MRI With a sensitivity ranging from 45% to 100%,
should include sagittal T1-weighted sequences and a specificity of 95% [92], CSF cytology
without contrast and sagittal fat suppression remains the gold standard diagnostic test for
T2-weighted sequences combined with axial LM. Errors may be due to insufficient sample
T1-weighted images with contrast of regions of volume, delayed sample processing time, collec-
interest [83]. tion of less than two samples, and collection from
a location far from the symptomatic site [93]. To
CSF Analysis maximize cytology sensitivity, we recommend a
Although MRI is typically the first diagnostic test sample volume of 10 mL or more, brisk process-
performed, CSF examination is definitive. At a ing, and collection from a site close to the symp-
minimum, CSF examination should include mea- tomatic area. CSF cytology remains challenging
surement of opening pressure, cell count, cyto- due to the irregular shedding of cancer cells and
logical examination, and protein and glucose their limited presence in CSF [92–94].
concentrations.
Protein Concentration
Pressure CSF protein levels are elevated (>38 mg/dL) in
Elevated intracranial pressure is present in almost 60–80% of patients. This is usually attributed to
50% of LM patients and may be attributed to the breakdown of tumor and infiltrating cells
impairment of CSF drainage by obstructing along with a disruption of the blood-CSF barrier
malignant cells [84]. Before attributing elevated allowing serum protein to flow in [24, 25, 27].
ICP to LM, care must be taken to exclude other However, the composition of this CSF protein
causes, including elevated systemic venous pres- remains under study. Interpretation of CSF pro-
sure or respiratory disease [85]. Normal ICP lev- tein levels must account for the sample site—
els in the correct lateral recumbent position can ventricular taps through an Ommaya reservoir
range from 90 to 250 mm H2O [86]; measure- have lower normal protein concentration thresh-
ments obtained while prone or seated may be olds than cisternal or lumbar taps [25].
falsely elevated. In patients with LM, ICP levels
can range from 90 to 550 mm H2O [87] with most Glucose Concentration
values less than 150 mm H2O on first CSF glucose concentration is diminished
LP. Measurement should be done directly after (CSF:serum ratio < 0.6 or glucose <40 mg/dL) in
needle insertion to avoid CSF leakage and falsely about one-fourth to one-third of cases [70, 95].
Abnormally low CSF glucose may be the sole Table 12.1 Cerebrospinal fluid markers in leptomenin-
geal metastasis from different primary cancers
indicator of LM in the absence of any other CSF
abnormality [96, 97] and usually reflects diffuse Marker Primary disease
meningeal involvement [97]. However, low CSF Beta 2 Lymphoma
microglobulin
glucose (hypoglycorrhachia) may be found in AFP Germ cell
several other neurologic diseases [98] and is Beta Nonspecific
therefore sensitive but not specific. Several glucuronidase
causes of low glucose are postulated: (1) CEA Colon, ovarian, breast, bladder, lung
increased utilization of glucose by malignant CA-125 Ovarian
CA-15-3 Breast
cells in the leptomeninges due to their high
CA19-9 Adenocarcinoma
metabolism and correlation with high lactate lev-
CK-BB Small cell lung cancer
els [99], (2) increased utilization by cerebral cells GFAP Glioma
surrounding CSF, and (3) ineffective glucose HCG subunit Choriocarcinoma, embryonal, and
entry into CSF by impaired transport systems germ cell tumors
[25]. 5-HIAA Carcinoid
LDH isoenzyme Carcinoma
D
ther CSF Markers
O PSA Prostate
With a sensitivity ranging from 95 to 100% and a Protein S-100 Melanoma
specificity of about 100% in the absence of neu- HMB45 Melanoma
roimaging findings, immunocytochemical analy- Modified from Demopoulos, A; Posner, J. Cerebrospinal
sis has proven useful in diagnosing LM from fluid biochemical markers. In: UpToDate, Post, TW (Ed),
hematological malignancies [100]. Similarly, UpToDate, Waltham, MA, 2018. Rogers LR. Neurologic
Complications of Cancer, 2nd ed. Contemporary
flow cytometry of CSF is more useful in hemato-
Neurology Series. Neuro-Oncol. 2009;11:96–7
logical malignancies than LM from solid tumors
[101]. Detection of aneuploid or hyperdiploid
cells in the CSF due to abnormal chromosomal rate, 98.9%) [103]. PCR of specific mRNA from
migration and erratic cell division is a robust CSF has high sensitivity and may be considered
indicator of LM. These techniques have not been [104]. Neither of these approaches has transi-
proven useful in LM from solid tumors. tioned to clinical use. However, a rare cell cap-
ture technology to detect E-CAM-expressing
Tumor Markers circulating tumor cells (CTCs) [105, 106] is cur-
In the absence of cytological evidence of disease, rently employed at select cancer centers to detect
select tumor markers have diagnostic utility. LM [105, 107–109]. With a sensitivity of 93%
Detection of tumor markers is not universally and a specificity of 95%, detection of ≥1 CSF-
available in clinical laboratories, but it may be CTC/mL represents a robust marker for diagnos-
useful in certain cases. Generally, LM should be ing LM and should be considered during routine
high on the differential if the CSF tumor marker LM workup, if available [106].
concentration exceeds 2% of the serum value
(Table 12.1).
Other markers have been investigated as pos- Mutational Analysis
sible diagnostic tools in the absence of positive
cytology results. CSF vascular endothelial growth Importance of Mutational Analysis
factor (VEGF) was reported to be a useful bio-
marker in high-risk breast cancer, lung cancer, Once LM is confirmed, molecular characteriza-
and melanoma patients (sensitivity, 75%; speci- tion of the tumor cells becomes the next diagnos-
ficity, 97%) [102]. CSF microRNA analysis has tic priority. Several studies have shown that
also been studied as an early indicator of LM in different regions of the same tumor can harbor a
breast and lung cancer patients (true-positive genetically heterogeneous group of cells
[110–119]. In one study of parenchymal brain Response Measurement

metastases, whole exome sequencing revealed
additional oncogenic alterations distinct from In clinical practice, response to therapy is
those found in the primary tumor. Fifty-three per- assessed over 6–8 weeks with a combination of
cent of cases showed clinically targetable altera- clinical exam, MRI of the brain and spine, and
tions that were not detected in matched primary CSF examination. Standardized criteria have
tumors. In the case of LM, several studies dem- been proposed by the Response Assessment in
onstrate mutations within the LM that were not Neuro-Oncology (RANO) group. The proposal
detected in the original tumor [120]. In the era of includes standardized neurologic exam, CSF
targeted molecular therapy, such information is cytology, and neuroimaging [123, 124]. These
indispensable. While whole exome or even tar- criteria have not yet been validated in a prospec-
geted exome sequencing of CSF remains an area tive manner. Therefore, at present, response to
of active translational investigation, in many treatment remains a clinical assessment.
cases immunohistochemistry of CSF cytology Commonly, stable MRI scans with stable to
samples may be useful to detect sensitizing or improving CSF picture in a neurologically stable
resistance mutations, e.g., T790Min EGFR- patient is interpreted as a good clinical response.
driven NSCLC. It is recommended that clinicians Improvement in several (>1) modalities (MRI,
avail themselves of such resources. exam, CSF) is unusual but welcomed. Progressive
disease is evident by worsening in one or more of
these modalities.
Treatment
General Considerations Radiation Therapy
LM remains an incurable site of metastasis. As Radiotherapy in LM aims to alleviate symptoms

such symptomatic management and palliative by reducing the size of bulky masses blocking
care are crucial for treatment of all LM patients. CSF flow or compressing cranial nerves [84, 125,
Due to the brisk nature of LM progression and 126]. This may also improve medical therapy
disproportionate impact of neurologic symptoms penetration to residual disease [123, 127]. Sites
for patient quality of life, management of symp- of CSF flow obstruction as visualized by a radio-
tomatic lesions takes first priority in clinical man- nuclide CSF flow study could also be targeted by
agement. Once symptomatic lesions are treated, focal radiotherapy as an initial treatment.
further management is dictated by risk group. However, in practice, a ventriculoperitoneal
Patients with LM are stratified into two risk shunt (VPS) is typically required prior to such
groups. Patients in the poor-risk group have low treatments. WBRT is used in cases of extensive
KPS, multiple serious or fixed neurologic defi- nodular, symptomatic linear LM, or coexisting
cits, and extensive systemic cancer with few parenchymal lesions. While WBRT has not been
remaining therapeutic options. Patients in the associated with improved survival in LM patients,
good-risk group have a KPS of >60%, few or no it can improve patients’ quality of life [54–56,
neurologic deficits, minimal systemic disease, 59, 62]. Focal RT is typically given at a dose of
and/or available therapeutic options for treat- 30 Gy in three fractions or 20 Gy in five fractions
ment. In both groups, symptomatic lesions are to sites of symptomatic or bulky disease [128].
generally treated with radiation therapy (RT) However, the dose may be reduced to 20 Gy in
and/or surgical management, while chemother- two fractions in patients with better predicted
apy is used for the rest of the neuraxis [70, 121, survival (>12 months) to limit local side effects
122]. Patients in the good-risk group may [128, 129]. Extensive RT leads to substantial tox-
receive both local and systemic treatments, as icities including mucositis, esophagitis, myelo-
discussed below. suppression, and leukoencephalopathy. For a
typical heavily pre-treated solid tumor patient, mal cytarabine is preferred, while standard cyta-
such toxicities effectively preclude full craniospi- rabine is restricted to LM patients with liquid
nal radiation. The risk of leukoencephalopathy is malignancies. Liposomal cytarabine was discon-
high when systemic or intrathecal chemotherapy tinued in the USA in 2017 but may be available in
is combined with extensive RT, particularly with other countries. While standard cytarabine has a
the use of methotrexate. half-life of less than 4 h and can be eliminated
within 1–2 days, liposomal cytarabine may
remain therapeutic within the CSF for up to
Chemotherapy 28 days [132, 133]. When comparing DepoCyt to
IT MTX, one trial demonstrated no significant
Intrathecal Therapy difference in PFS [134]; another demonstrated
Although delivering treatment to the site of dis- delay to neurologic progression in DepoCyt-
ease is intuitively appealing, practical consider- treated patients [41, 135]. In a nonrandomized
ations limit the use of intrathecal chemotherapy trial, the combination of cytarabine and MTX
in many patients. To receive intrathecal chemo- demonstrated higher cytologic response and lon-
therapy, patients must demonstrate normal ICP ger median survival when compared to MTX
and CSF flow dynamics. Bulky disease will not alone, but no patient risk stratification was per-
be adequately treated with intrathecal formed [136].
approaches—intrathecal therapies only penetrate Thiotepa can also be used for the intrathecal
a few cell layers. If intrathecal chemotherapy is treatment of LM. It is highly lipid soluble and
indicated, it may be delivered through an hence has the shortest half-life of all IT agents.
Ommaya reservoir (intraventricularly) or into the Like MTX, it is administered twice weekly and
thecal sac via lumbar puncture. can cause myelosuppression [40, 137]. It is typi-
Methotrexate (MTX) is the most commonly cally employed as a second-line therapy in the
used intrathecal chemotherapy and can tran- case of MTX-refractory disease or MTX-induced
siently clear malignant cells from CSF in up to leukoencephalopathy. Since concurrent chemo-
61% of LM patients [40, 130]. With a CSF half- therapy (MTX) and RT can exacerbate side
life of 4.5 h, MTX is administered at 10–12 mg effects, MTX may be replaced by thiotepa for
twice weekly for 4 weeks as induction regimen. patients requiring concurrent RT. Clinical
In the event of clinical response, dosage is response to thiotepa is largely equivalent to IT
decreased to once weekly for 4–8 weeks fol- MTX [40, 138, 139].
lowed by biweekly maintenance therapy for sev-
eral months. The ideal duration of therapy with Systemic Therapy
MTX is unknown, but treatment beyond 6 months
may be unwarranted [130]. MTX is renally Untargeted
excreted after being absorbed by the choroid Many systemic chemotherapeutic agents can
plexus into the systemic circulation where it is achieve therapeutic concentrations in CSF. Systemic
bound to albumin [131]. Therefore, coadminis- therapy avoids the risk of Ommaya placement sur-
tration of drugs that displace MTX from albumin gery and catheter-related complications. For
should be done cautiously. Oral leukovorin, patients with CSF flow abnormalities, systemic
which dose not enter the CSF, is administered to chemotherapeutic agents may allow uniform distri-
counter systemic MTX toxicity. Other neurologic bution, even with bulky tumors [140].
toxicities due to MTX include delayed leukoen- High-dose MTX is the most commonly used
cephalopathy, aseptic meningitis, acute encepha- systemic agent in LM patients. However, clinical
lopathy, and transverse myelopathy. response after systemic high-dose MTX is mixed
Cytarabine may also be administered intrathe- [140, 141]. An important consideration is the
cally in two forms: standard and liposomal need for close inpatient monitoring, including
(DepoCyt). In patients with solid tumors, liposo- aggressive hydration and urinary alkalinization
followed by leucovorin rescue. Comparison of IT decreased leptomeningeal enhancement, nega-

therapy (MTX) to combined IT and systemic tive CSF cytology, and overall survival of 8 and
therapy in breast cancer patients with LM dem- 7.5 months, respectively [153]. A pilot study with
onstrated no survival benefit [142]. Additional a similar patient population reported median
neurologic complications were reported in the overall survival of 4.7 months and CNS response
intrathecal group. High-dose cytarabine can also rate of 70% [154]. When used alone in a mixed
be used systemically with a CSF concentration population of solid tumor LM, bevacizumab
reaching up to 22% of serum levels [143, 144]. resulted in median overall survival of 14 weeks
As with other systemic therapies for LM, high and CNS response rate of 13% [155].
doses carry significant toxicity. Efficacy of such Case studies report good response to BRAF
an approach in patients with solid tumor LM has inhibitors such as vemurafenib and dabrafenib in
not been demonstrated. patients with melanoma LM harboring the BRAF
Capecitabine is a fluoropyrimidine carbamate V600E mutation [156, 157]. This alteration con-
that is used as an oral substitute for 5-fluorouracil, stitutively activates the MAP kinase pathway,
a capecitabine precursor that is active in tumor while resistance to targeting this mutation is
sites. Despite limited information regarding mediated through MEK. While several trials have
capecitabine’s pharmacokinetics in the CSF, sev- successfully demonstrated the superiority of
eral observational studies have documented the BRAF and MEK inhibitors together as compared
effect of capecitabine on patients with LM [145, to BRAF alone in melanoma [157–159], patients
146]. A case series reported response to capecitabine with LM were excluded from these trials.
and trastuzumab combination therapy in patients Epidermal growth factor receptor (EGFR)
with breast cancer LM [147]. Compared to other tyrosine kinase inhibitors have been useful thera-
regimens, capecitabine is not associated with cen- pies in patients with EGFR-driven non-small cell
tral neurotoxicity and is generally well tolerated lung cancer (NSCLC). Erlotinib was shown to
[148]. However, in practice, a substantial propor- improve performance status in patients with LM
tion of breast cancer LM patients may have already from NSCLC [160–162]. Nevertheless, erlotinib
received capecitabine, limiting the utility of the failure was still seen in some studies of these
regimen. patients [163, 164]. Clinical and radiological
Temozolomide is an oral alkylating agent that improvement was seen in patients with LM from
has been employed in case studies of LM from lung adenocarcinoma following treatment with
solid tumors. These involved administration of gefitinib, especially at high doses [160, 165–169].
100 mg/m2 temozolomide daily every other week Since second- and third-generation TKIs have
for 4 weeks, with temporary disease stabilization better CNS penetration [170, 171], several trials
in two patients (median overall survival of have assessed these drugs in EGFR-mutant
43 days) [149]. NSCLC patients with LM. Afatinib, an inhibitor
of Her2 and EGFR kinases, showed 35% response
Targeted rate in EGFR-mutant NSCLC patients with LM
As with systemic malignancies, targeted thera- [172] and had efficacy in patients progressing on
pies are poised to revolutionize the management first-generation TKIs [173, 174]. Promising
of LM. Studies have shown high levels of VEGF results have been reported for osimertinib, a third-
in the CSF of LM patients, which correlated with generation TKI, in EGFR-mutant NSCLC patients
poor prognosis [102, 150, 151]. Angiogenesis with LM [175]; a phase 2 trial is currently recruit-
inhibition has been shown to prolong median ing patients harboring T790M-mutated NSCLC
overall survival in a preclinical model of breast with LM who failed initial EGFR TKI therapy
cancer LM [152]. Combined therapy of bevaci- (NCT03257124). Another new-generation TKI
zumab, etoposide, and cisplatin (BEEP) was currently in trials is AZD3759 which has superior
reported in two breast cancer patients with pro- CNS penetration and good tolerability in advanced
gressive LM after radiation. BEEP therapy led to NSCLC patients [176] (NCT02228369).
Similarly, ALK inhibitors have been employed with LM, and a phase 3 study is currently in prep-
for patients with ALK fusion gene-positive aration for this population (NCT03613181).
NSCLC and CNS metastasis. Prior trials com- Monoclonal antibodies against tumor-specific
bined all CNS metastasis, parenchymal or lepto- epitopes conjugated to radioisotopes like
meningeal; no studies have specifically targeted iodine-131 (131I) and yttrium-90 (90Y) have
LM patients. Alectinib has demonstrated favor- been employed to deliver brachytherapy to
able activity in LM from ALK fusion-positive tumors intrathecally. An early study by Moseley
NSLCLC [177–180]. et al. in patients with LM showed some clinical
efficacy of radioisotope-labeled HMFG1, an
antigen present on normal and neoplastic deriva-
Supportive Therapy tives of glandular epithelium [185, 186]. More
recent studies have reported the utility of intra-
Supportive therapy in LM aims at relieving neu- Ommaya injection of radiolabeled 131I-3F8
rologic symptoms to improve quality of life and131I-8H9, targeting tumor-associated antigens
[181]. Steroids can be used to reduce vasogenic GD2 and B7H3, respectively, in neuroblastoma
edema caused by the tumor and lessen neuro- with CNS involvement, including LM [187, 188].
logic symptoms. In addition, dexamethasone is Phase 2 and 3 trials are currently being planned
essential in the management of chemical- to evaluate the use of 131I-omburtamab, an 8H9
induced meningitis that may develop after IT target, for neuroblastoma patients with CNS
therapy, irrespective of the agent used [41, 42]. metastasis (NCT03275402).
LM can cause seizures in 10%–15% of
patients—transient symptoms should prompt
evaluation for ictal activity by EEG, and anti- Future Directions
epileptic drugs (AEDs) should be started.
Modern AEDs that do not induce CYP-450 Efforts to expand administration of systemic che-
enzyme activation such as levetiracetam, lacos- motherapeutic agents intrathecally have proven
amide, and zonisamide are preferred for treating somewhat disappointing. Such approaches have
patients with cancer [182]. included IT etoposide [189–191], topotecan
[192], busulfan [193], melphalan [194], nitro-
soureas [195], and dacarbazine [196]. For tumors
New Treatment Agents with molecularly established therapeutic targets,
repurposing currently available targeted agents
Immune therapy agents have been used for the intrathecally (like IT trastuzumab) has proven
treatment of many systemic malignancies and are more useful.
currently under investigation for the treatment of In the case of systemically administered tar-
LM. A phase 2 trial using nivolumab (PD-1 geted therapies with good CNS penetration such
inhibitor) and ipilimumab (CTLA-4 inhibitor) is as osimertinib, leptomeningeal responses have
currently recruiting patients with leptomeningeal been promising. This molecularly driven
metastasis from any solid tumor primary approach to treatment of LM suggests that we
(NCT02939300) [183]. Close monitoring is nec- are on the verge of a new paradigm in the man-
essary in these studies since immunotherapy can agement of LM—molecular characterization of
often result in life-threatening toxicities. For LM tumor cells prior to design of therapy.
instance, IT interferon alpha and interleukin-2 Numerous studies are currently underway to
trials did not move forward due to significant tox- capture the molecular phenotype of cancer cells
icities despite clinical response [150, 184]. within the leptomeninges. Approaches include
ANG1005 is a conjugated paclitaxel molecule sequencing of cell-free tumor DNA (ctDNA)
with enhanced BBB penetration. It has been [120] as well as flow cytometry-based investiga-
shown to be effective in breast cancer patients tions [197–200].
Once identified, molecular vulnerabilities of the clearance of interstitial solutes, including amy-
loid β. Sci Transl Med. 2012;4:147ra111.
leptomeningeal tumors must be targeted, and 9. Smith AJ, Yao X, Dix JA, Jin B-J, Verkman AS. Test
these treatments must be formally assessed in of the “glymphatic” hypothesis demonstrates diffu-
prospective clinical trials. A major impediment to sive and aquaporin-4-independent solute transport in
this has been the lack of response criteria as well rodent brain parenchyma. eLife. 2017;6:1.
10. Mestre H, Kostrikov S, Mehta RI, Nedergaard
as inability to reliably quantify the burden of dis- M. Perivascular spaces, glymphatic dysfunction,
ease. Much work is being done regarding to and small vessel disease. Clin Sci Lond Engl 1979.
investigate the use of circulating tumor cells and/ 2017;131:2257–74.
or flow cytometry to quantify LM disease burden 11. Beauchesne P. Intrathecal chemotherapy for treat-
ment of leptomeningeal dissemination of metastatic
[120, 197–200]. Together, these complimentary tumours. Lancet Oncol. 2010;11:871–9.
approaches will empower clinical and transla- 12. Glover RL, Brook AL, Welch MR. Teaching
tional researchers to make true progress in the NeuroImages: leptomeningeal lung carcinoma.
treatment of leptomeningeal metastasis. Neurology. 2014;82:e183–4.
13. Kokkoris CP. Leptomeningeal carcinomatosis.
How does cancer reach the pia-arachnoid? Cancer.
1983;51:154–60.
14. Boyle R, Thomas M, Adams JH. Diffuse involve-
References ment of the leptomeninges by tumour – a clinical
and pathological study of 63 cases. Postgrad Med J.
1. Protasoni M, Sangiorgi S, Cividini A, Culuvaris 1980;56:149–58.
GT, Tomei G, Dell’Orbo C, et al. The collagenic 15. Boire A, Zou Y, Shieh J, Macalinao DG, Pentsova E,
architecture of human dura mater. J Neurosurg. Massagué J. Complement component 3 adapts the
2011;114:1723–30. cerebrospinal fluid for leptomeningeal metastasis.
2. Absinta M, Ha S-K, Nair G, Sati P, Luciano NJ, Cell. 2017;168:1101–1113.e13.
Palisoc M, et al. Human and nonhuman primate 16. Lamovec J, Bracko M. Metastatic pattern of infil-
meninges harbor lymphatic vessels that can be visu- trating lobular carcinoma of the breast: an autopsy
alized noninvasively by MRI. eLife. 2017;6:1. study. J Surg Oncol. 1991;48:28–33.
3. Balin BJ, Broadwell RD, Salcman M, el-Kalliny 17. Lamovec J, Zidar A. Association of leptomeningeal
M. Avenues for entry of peripherally adminis- carcinomatosis in carcinoma of the breast with infil-
tered protein to the central nervous system in trating lobular carcinoma. An autopsy study. Arch
mouse, rat, and squirrel monkey. J Comp Neurol. Pathol Lab Med. 1991;115:507–10.
1986;251:260–80. 18. Altundag K, Bondy ML, Mirza NQ, Kau S-W,
4. Yasuda K, Cline C, Vogel P, Onciu M, Fatima S, Broglio K, Hortobagyi GN, et al. Clinicopathologic
Sorrentino BP, et al. Drug transporters on arachnoid characteristics and prognostic factors in 420 meta-
barrier cells contribute to the blood-cerebrospinal static breast cancer patients with central nervous sys-
fluid barrier. Drug Metab Dispos Biol Fate Chem. tem metastasis. Cancer. 2007;110:2640–7.
2013;41:923–31. 19. Cheng H, Perez-Soler R. Leptomeningeal metas-
5. Snyder JM, Hagan CE, Bolon B, Keene CD. 20 – tases in non-small-cell lung cancer. Lancet Oncol.
Nervous system. In: Treuting PM, Dintzis SM, 2018;19:e43–55.
Montine KS, editors. Comparative anatomy and 20. Dekker AW, Elderson A, Punt K, Sixma
histology [Internet]. 2nd ed. San Diego: Academic; JJ. Meningeal involvement in patients with acute
2018. p. 403–44.. [Cited 2018 Oct 10]. Available nonlymphocytic leukemia. Incidence, management,
from: http://www.sciencedirect.com/science/article/ and predictive factors. Cancer. 1985;56:2078–82.
pii/B9780128029008000208. 21. Peterson BA, Brunning RD, Bloomfield CD, Hurd
6. Hannocks M-J, Pizzo ME, Huppert J, Deshpande DD, Gau JA, Peng GT, et al. Central nervous system
T, Abbott NJ, Thorne RG, et al. Molecular charac- involvement in acute nonlymphocytic leukemia. A
terization of perivascular drainage pathways in the prospective study of adults in remission. Am J Med.
murine brain. J Cereb Blood Flow Metab Off J Int 1987;83:464–70.
Soc Cereb Blood Flow Metab. 2018;38:669–86. 22. Ersbøll J, Schultz HB, Thomsen BL, Keiding N,
7. Bedussi B, van Lier MGJTB, Bartstra JW, de Vos Nissen NI. Meningeal involvement in non-Hodgkin’s
J, Siebes M, VanBavel E, et al. Clearance from lymphoma: symptoms, incidence, risk factors and
the mouse brain by convection of interstitial fluid treatment. Scand J Haematol. 1985;35:487–96.
towards the ventricular system. Fluid Barrier CNS. 23. Hoerni-Simon G, Suchaud JP, Eghbali H, Coindre
2015;12:23. JM, Hoerni B. Secondary involvement of the cen-
8. Iliff JJ, Wang M, Liao Y, Plogg BA, Peng W, tral nervous system in malignant non-Hodgkin’s
Gundersen GA, et al. A paravascular pathway facili- lymphoma. A study of 30 cases in a series of 498
tates CSF flow through the brain parenchyma and patients. Oncology. 1987;44:98–101.
24. Kaplan JG, DeSouza TG, Farkash A, Shafran B, trexate and thiotepa in patients with previously
Pack D, Rehman F, et al. Leptomeningeal metasta- untreated neoplastic meningitis. Eastern Cooperative
ses: comparison of clinical features and laboratory Oncology Group. J Clin Oncol. 1993;11:561–9.
data of solid tumors, lymphomas and leukemias. J 41. Glantz MJ, Jaeckle KA, Chamberlain MC,
Neuro-Oncol. 1990;9:225–9. Phuphanich S, Recht L, Swinnen LJ, et al. A ran-
25. Rogers LR. Neurologic complications of can- domized controlled trial comparing intrathecal
cer, 2nd ed. Contemp Neurol Ser Neuro-Oncol. sustained-release cytarabine (DepoCyt) to intra-
2009;11:96–7. thecal methotrexate in patients with neoplastic
26. Kesari S, Batchelor TT. Leptomeningeal metastases. meningitis from solid tumors. Clin Cancer Res.
Neurol Clin. 2003;21:25–66. 1999;5:3394–402.
27. Clarke JL, Perez HR, Jacks LM, Panageas KS, 42. Glantz MJ, LaFollette S, Jaeckle KA, Shapiro W,
Deangelis LM. Leptomeningeal metastases in the Swinnen L, Rozental JR, et al. Randomized trial of
MRI era. Neurology. 2010;74:1449–54. a slow-release versus a standard formulation of cyta-
28. Hauch TW, Shelbourne JD, Cohen HJ, Mason D, rabine for the intrathecal treatment of lymphomatous
Kremer WB. Meningeal mycosis fungoides: clini- meningitis. J Clin Oncol. 1999;17:3110–6.
cal and cellular characteristics. Ann Intern Med. 43. Hitchins RN, Bell DR, Woods RL, Levi JA. A pro-
1975;82:499–505. spective randomized trial of single-agent versus
29. Lundberg WB, Cadman EC, Skeel combination chemotherapy in meningeal carcino-
RT. Leptomeningeal mycosis fungoides. Cancer. matosis. J Clin Oncol Off J Am Soc Clin Oncol.
1976;38:2149–53. 1987;5:1655–62.
30. Wabulya A, Imitola J, Santagata S, Kesari S. Mycosis 44. Rudnicka H, Niwinska A, Murawska M. Breast can-
fungoides with leptomeningeal involvement. J Clin cer leptomeningeal metastasis – the role of multimo-
Oncol Off J Am Soc Clin Oncol. 2007;25:5658–61. dality treatment. J Neuro-Oncol. 2007;84:57–62.
31. Maldonado JE, Kyle RA, Ludwig J, Okazaki 45. Gauthier H, Guilhaume MN, Bidard FC, Pierga JY,
H. Meningeal myeloma. Arch Intern Med. Girre V, Cottu PH, et al. Survival of breast cancer
1970;126:660–3. patients with meningeal carcinomatosis. Ann Oncol.
32. Patriarca F, Zaja F, Silvestri F, Sperotto A, Scalise 2010;21:2183–7.
A, Gigli G, et al. Meningeal and cerebral involve- 46. Lee S, Ahn HK, Park YH, Nam DH, Lee JI, Park W,
ment in multiple myeloma patients. Ann Hematol. et al. Leptomeningeal metastases from breast cancer:
2001;80:758–62. intrinsic subtypes may affect unique clinical manifes-
33. Weed JC, Creasman WT. Meningeal carcinomato- tations. Breast Cancer Res Treat. 2011;129:809–17.
sis secondary to advanced squamous cell carcinoma 47. de Azevedo CRAS, Cruz MRS, Chinen LTD,
of the cervix: a case report. Meningeal metasta- Peres SV, Peterlevitz MA, de Azevedo Pereira AE,
sis of advenced cervical cancer. Gynecol Oncol. et al. Meningeal carcinomatosis in breast cancer:
1975;3:201–4. prognostic factors and outcome. J Neuro-Oncol.
34. Barnard RO, Parsons M. Carcinoma of the thyroid 2011;104:565–72.
with leptomeningeal dissemination following the 48. Lara-Medina F, Crismatt A, Villarreal-Garza
treatment of a toxic goitre with 131-I and methyl C, Alvarado-Miranda A, Flores-Hernández L,
thiouracil. Case with a co-existing intracranial der- González-Pinedo M, et al. Clinical features and
moid. J Neurol Sci. 1969;8:299–306. prognostic factors in patients with carcinomatous
35. Bresalier RS, Karlin DA. Meningeal metastasis meningitis secondary to breast cancer. Breast J.
from rectal carcinoma with elevated cerebrospinal 2012;18:233–41.
fluid carcinoembryonic antigen. Dis Colon Rectum. 49. Meattini I, Livi L, Saieva C, Franceschini D,
1979;22:216–7. Marrazzo L, Greto D, et al. Prognostic factors and
36. Nagourney RA, Hedaya R, Linnoila M, Schein clinical features in patients with leptominengeal
PS. Carcinoid carcinomatous meningitis. Ann Intern metastases from breast cancer: a single center expe-
Med. 1985;102:779–82. rience. J Chemother Florence Italy. 2012;24:279–84.
37. Berry MP, Jenkin RD. Parameningeal rhabdomyo- 50. Niwińska A, Rudnicka H, Murawska M. Breast
sarcoma in the young. Cancer. 1981;48:281–8. cancer leptomeningeal metastasis: propensity of
38. Cash J, Fehir KM, Pollack MS. Meningeal involve- breast cancer subtypes for leptomeninges and the
ment in early stage chronic lymphocytic leukemia. analysis of factors influencing survival. Med Oncol
Cancer. 1987;59:798–800. Northwood London England. 2013;30:408.
39. Matthay KK, Brisse H, Couanet D, Couturier J, 51. Yust-Katz S, Garciarena P, Liu D, Yuan Y, Ibrahim
Bénard J, Mosseri V, et al. Central nervous system N, Yerushalmi R, et al. Breast cancer and lepto-
metastases in neuroblastoma: radiologic, clini- meningeal disease (LMD): hormone receptor sta-
cal, and biologic features in 23 patients. Cancer. tus influences time to development of LMD and
2003;98:155–65. survival from LMD diagnosis. J Neuro-Oncol.
40. Grossman SA, Finkelstein DM, Ruckdeschel JC, 2013;114:229–35.
Trump DL, Moynihan T, Ettinger DS. Randomized 52. Le Rhun E, Taillibert S, Zairi F, Kotecki N, Devos P,
prospective comparison of intraventricular metho- Mailliez A, et al. A retrospective case series of 103
consecutive patients with leptomeningeal metastasis fluid absorption and severe intracranial hyperten-
and breast cancer. J Neuro-Oncol. 2013;113:83–92. sion. Acta Neurochir. 1987;86:93–7.
53. Niwińska A, Rudnicka H, Murawska M. Breast can- 66. Laas R, Arnold H. Compression of the outlets of the
cer leptomeningeal metastasis: the results of com- leptomeningeal veins – the cause of intracranial pla-
bined treatment and the comparison of methotrexate teau waves. Acta Neurochir. 1981;58:187–201.
and liposomal cytarabine as intra-cerebrospinal fluid 67. Sinniah D, Looi LM, Ortega JA, Siegel SE,
chemotherapy. Clin Breast Cancer. 2015;15:66–72. Landing B. Cerebellar coning and uncal herniation
54. Morris PG, Reiner AS, Szenberg OR, Clarke JL, in childhood acute leukaemia. Lancet Lond Engl.
Panageas KS, Perez HR, et al. Leptomeningeal 1982;2:702–4.
metastasis from non-small cell lung cancer: survival 68. Grossman SA, Trump DL, Chen DC, Thompson G,
and the impact of whole brain radiotherapy. J Thorac Camargo EE. Cerebrospinal fluid flow abnormalities
Oncol Off Publ Int Assoc Study Lung Cancer. in patients with neoplastic meningitis. An evalua-
2012;7:382–5. tion using 111indium-DTPA ventriculography. Am
55. Park JH, Kim YJ, Lee J-O, Lee K-W, Kim JH, Bang J Med. 1982;73:641–7.
S-M, et al. Clinical outcomes of leptomeningeal 69. Broderick JP, Cascino TL. Nonconvulsive status
metastasis in patients with non-small cell lung can- epilepticus in a patient with leptomeningeal cancer.
cer in the modern chemotherapy era. Lung Cancer Mayo Clin Proc. 1987;62:835–7.
Amsterdam Netherlands. 2012;76:387–92. 70. Klein P, Haley EC, Wooten GF, VandenBerg
56. Gwak H-S, Joo J, Kim S, Yoo H, Shin SH, Han SR. Focal cerebral infarctions associated with peri-
J-Y, et al. Analysis of treatment outcomes of intravascular tumor infiltrates in carcinomatous leptomen-
ventricular chemotherapy in 105 patients for lepto- ingeal metastases. Arch Neurol. 1989;46:1149–52.
meningeal carcinomatosis from non-small-cell lung 71. Wasserstrom WR, Glass JP, Posner JB. Diagnosis
cancer. J Thorac Oncol Off Publ Int Assoc Study and treatment of leptomeningeal metastases from
Lung Cancer. 2013;8:599–605. solid tumors: experience with 90 patients. Cancer.
57. Lee SJ, Lee J-I, Nam D-H, Ahn YC, Han JH, Sun 1982;49:759–72.
J-M, et al. Leptomeningeal carcinomatosis in non- 72. Latchaw RE, Gabrielsen TO, Seeger JF. Cerebral
small-cell lung cancer patients: impact on survival angiography in meningeal sarcomatosis and carcino-
and correlated prognostic factors. J Thorac Oncol Off matosis. Neuroradiology. 1974;8:131–9.
Publ Int Assoc Study Lung Cancer. 2013;8:185–91. 73. Shapiro WR, Posner JB, Ushio Y, Chemik NL,
58. Riess JW, Nagpal S, Iv M, Zeineh M, Gubens MA, Young DF. Treatment of meningeal neoplasms.
Ramchandran K, et al. Prolonged survival of patients Cancer Treat Rep. 1977;61:733–43.
with non-small-cell lung cancer with leptomeningeal 74. Lossos A, Siegal T. Numb chin syndrome in cancer
carcinomatosis in the modern treatment era. Clin patients: etiology, response to treatment, and prog-
Lung Cancer. 2014;15:202–6. nostic significance. Neurology. 1992;42:1181–4.
59. Kuiper JL, Hendriks LE, van der Wekken AJ, de 75. Chang EL, Lo S. Diagnosis and management of cen-
Langen AJ, Bahce I, Thunnissen E, et al. Treatment tral nervous system metastases from breast cancer.
and survival of patients with EGFR-mutated non- Oncologist. 2003;8:398–410.
small cell lung cancer and leptomeningeal metas- 76. Boogerd W, Dorresteijn LD, van Der Sande JJ, de
tasis: a retrospective cohort analysis. Lung Cancer Gast GC, Bruning PF. Response of leptomeningeal
Amsterdam Netherlands. 2015;89:255–61. metastases from breast cancer to hormonal therapy.
60. Harstad L, Hess KR, Groves MD. Prognostic factors Neurology. 2000;55:117–9.
and outcomes in patients with leptomeningeal mela- 77. Singh SK, Agris JM, Leeds NE, Ginsberg
nomatosis. Neuro-Oncology. 2008;10:1010–8. LE. Intracranial leptomeningeal metastases: compar-
61. Geukes Foppen MH, Brandsma D, Blank CU, van ison of depiction at FLAIR and contrast-enhanced
Thienen JV, Haanen JB, Boogerd W. Targeted treat- MR imaging. Radiology. 2000;217:50–3.
ment and immunotherapy in leptomeningeal metas- 78. Yousem DM, Patrone PM, Grossman
tases from melanoma. Ann Oncol Off J Eur Soc Med RI. Leptomeningeal metastases: MR evaluation. J
Oncol. 2016;27:1138–42. Comput Assist Tomogr. 1990;14:255–61.
62. Abouharb S, Ensor J, Loghin ME, Katz R, Moulder 79. Rodesch G, Avni EF, Parizel P, Detemmerman
SL, Esteva FJ, et al. Leptomeningeal disease and D, Szliwowski H, Brotchi J, et al. Schilder’s dis-
breast cancer: the importance of tumor subtype. ease: neuroradiological findings. J Neuroradiol.
Breast Cancer Res Treat. 2014;146:477–86. 1988;15:386–93.
63. Morikawa A, Jordan L, Rozner R, Patil S, Boire 80. Freilich RJ, Krol G, DeAngelis LM. Neuroimaging
A, Pentsova E, et al. Characteristics and outcomes and cerebrospinal fluid cytology in the diagno-
of patients with breast cancer with leptomeningeal sis of leptomeningeal metastasis. Ann Neurol.
metastasis. Clin Breast Cancer. 2017;17:23–8. 1995;38:51–7.
64. Clarke JL. Leptomeningeal metastasis from systemic 81. Pan Z, Yang G, He H, Yuan T, Wang Y, Li Y, et al.
cancer. Contin Minneap Minn. 2012;18:328–42. Leptomeningeal metastasis from solid tumors: clini-
65. Hansen K, Gjerris F, Sørensen PS. Absence of cal features and its diagnostic implication. Sci Rep.
hydrocephalus in spite of impaired cerebrospinal 2018;8:10445.
82. Singh SK, Leeds NE, Ginsberg LE. MR imag- 98. Bomgaars L, Chamberlain MC, Poplack DG, Blaney
ing of leptomeningeal metastases: comparison SM. Leptomeningeal metastases. In: Levin VA, edi-
of three sequences. AJNR Am J Neuroradiol. tor. Cancer in the nervous system. 2nd ed. New York:
2002;23:817–21. Oxford University Press; 2002. p. 375–96.
83. Le Rhun E, Weller M, Brandsma D, Van den Bent 99. Schold SC, Wasserstrom WR, Fleisher M, Schwartz
M, de Azambuja E, Henriksson R, et al. EANO- MK, Posner JB. Cerebrospinal fluid biochemical
ESMO Clinical Practice Guidelines for diagnosis, markers of central nervous system metastases. Ann
treatment and follow-up of patients with leptomen- Neurol. 1980;8:597–604.
ingeal metastasis from solid tumours. Ann Oncol. 100. Zeiser R, Burger JA, Bley TA, Windfuhr-Blum M,
2017;28:iv84–99. Schulte-Mönting J, Behringer DM. Clinical follow-
84. Bruno MK, Raizer J. Leptomeningeal metastases up indicates differential accuracy of magnetic reso-
from solid tumors (meningeal carcinomatosis). nance imaging and immunocytology of the cerebral
Cancer Treat Res. 2005;125:31–52. spinal fluid for the diagnosis of neoplastic menin-
85. Bragin DE, Statom G, Nemoto EM. Dynamic cere- gitis – a single centre experience. Br J Haematol.
brovascular and intracranial pressure reactivity 2004;124:762–8.
assessment of impaired cerebrovascular autoregu- 101. Bromberg JEC, Breems DA, Kraan J, Bikker G,
lation in intracranial hypertension. Acta Neurochir van der Holt B, Smitt PS, et al. CSF flow cytometry
Suppl. 2016;122:255–60. greatly improves diagnostic accuracy in CNS hema-
86. Corbett JJ, Mehta MP. Cerebrospinal fluid pressure tologic malignancies. Neurology. 2007;68:1674–9.
in normal obese subjects and patients with pseudotu- 102. Groves MD, Hess KR, Puduvalli VK, Colman H,
mor cerebri. Neurology. 1983;33:1386–8. Conrad CA, Gilbert MR, et al. Biomarkers of dis-
87. Olson ME, Chernik NL, Posner JB. Infiltration of ease: cerebrospinal fluid vascular endothelial growth
the leptomeninges by systemic cancer. A clinical and factor (VEGF) and stromal cell derived factor
pathologic study. Arch Neurol. 1974;30:122–37. (SDF)-1 levels in patients with neoplastic meningitis
88. King DK, Loh KK, Ayala AG, Gamble JF. Letter: (NM) due to breast cancer, lung cancer and mela-
eosinophilic meningitis and lymphomatous menin- noma. J Neuro-Oncol. 2009;94:229–34.
gitis. Ann Intern Med. 1975;82:228. 103. Teplyuk NM, Mollenhauer B, Gabriely G, Giese
89. Mulligan MJ, Vasu R, Grossi CE, Prasthofer EF, A, Kim E, Smolsky M, et al. MicroRNAs in cere-
Griffin FM, Kapila A, et al. Neoplastic meningitis brospinal fluid identify glioblastoma and metastatic
with eosinophilic pleocytosis in Hodgkin’s disease: brain cancers and reflect disease activity. Neuro-
a case with cerebellar dysfunction and a review of Oncology. 2012;14:689–700.
the literature. Am J Med Sci. 1988;296:322–6. 104. Raj GV, Moreno JG, Gomella LG. Utilization of
90. Conrad KA, Gross JL, Trojanowski polymerase chain reaction technology in the detec-
JQ. Leptomeningeal carcinomatosis presenting as tion of solid tumors. Cancer. 1998;82:1419–42.
eosinophilic meningitis. Acta Cytol. 1986;30:29–31. 105. Patel AS, Allen JE, Dicker DT, Peters KL, Sheehan
91. Budka H, Guseo A, Jellinger K, Mittermayer JM, Glantz MJ, et al. Identification and enumeration
K. Intermittent meningitic reaction with severe baso- of circulating tumor cells in the cerebrospinal fluid
philia and eosinophilia in CNS leukaemia. J Neurol of breast cancer patients with central nervous system
Sci. 1976;28:459–68. metastases. Oncotarget. 2011;2:752–60.
92. Enting RH. Leptomeningeal neoplasia: epidemiol- 106. Lin X, Fleisher M, Rosenblum M, Lin O, Boire
ogy, clinical presentation, CSF analysis and diag- A, Briggs S, et al. Cerebrospinal fluid circulating
nostic imaging. Cancer Treat Res. 2005;125:17–30. tumor cells: a novel tool to diagnose leptomeningeal
93. Glantz MJ, Cole BF, Glantz LK, Cobb J, Mills P, metastases from epithelial tumors. Neuro-Oncology.
Lekos A, et al. Cerebrospinal fluid cytology in 2017;19:1248–54.
patients with cancer: minimizing false-negative 107. Le Rhun E, Massin F, Tu Q, Bonneterre J,
results. Cancer. 1998;82:733–9. Bittencourt MDC, Faure GC. Development of a new
94. Chamberlain MC, Kormanik PA, Glantz MJ. A method for identification and quantification in cere-
comparison between ventricular and lumbar cerebrospinal fluid of malignant cells from breast carci-
brospinal fluid cytology in adult patients with noma leptomeningeal metastasis. BMC Clin Pathol.
leptomeningeal metastases. Neuro-Oncology. 2012;12:21.
2001;3:42–5. 108. Le Rhun E, Tu Q, De Carvalho Bittencourt M, Farre
95. De Vita VT, Canellos GP. Hypoglycorrhachia in I, Mortier L, Cai H, et al. Detection and quantifica-
meningeal carcinomatosis. Cancer. 1966;19:691–4. tion of CSF malignant cells by the CellSearch tech-
96. Kim P, Ashton D, Pollard JD. Isolated hypoglycor- nology in patients with melanoma leptomeningeal
rachia: leptomeningeal carcinomatosis causing sub- metastasis. Med Oncol Northwood London England.
acute confusion. J Clin Neurosci Off J Neurosurg 2013;30:538.
Soc Australas. 2005;12:841–3. 109. Nayak L, Fleisher M, Gonzalez-Espinoza R, Lin O,
97. Fishman RA. Cerebrospinal fluid in diseases of the Panageas K, Reiner A, et al. Rare cell capture tech-
nervous system. Philadelphia: Saunders; 1992. nology for the diagnosis of leptomeningeal metasta-
sis in solid tumors. Neurology. 2013;80:1598–605.. 125. Joshi A, Ghosh J, Noronha V, Parikh PM, Prabhash
discussion 1603 K. Leptomeningeal metastasis in solid tumors with
110. Campbell PJ, Yachida S, Mudie LJ, Stephens PJ, a special focus on lung cancer. Indian J Cancer.
Pleasance ED, Stebbings LA, et al. The patterns and 2014;51:410–3.
dynamics of genomic instability in metastatic pan- 126. Chang EL, Maor MH. Standard and novel radiother-
creatic cancer. Nature. 2010;467:1109–13. apeutic approaches to neoplastic meningitis. Curr
111. Gerlinger M, Rowan AJ, Horswell S, Math M, Oncol Rep. 2003;5:24–8.
Larkin J, Endesfelder D, et al. Intratumor heteroge- 127. Chamberlain MC, Kormanik P. Carcinoma meningi-
neity and branched evolution revealed by multire- tis secondary to non-small cell lung cancer: combined
gion sequencing. N Engl J Med. 2012;366:883–92. modality therapy. Arch Neurol. 1998;55:506–12.
112. Liu W, Laitinen S, Khan S, Vihinen M, Kowalski J, 128. Soffietti R, Cornu P, Delattre JY, Grant R, Graus F,
Yu G, et al. Copy number analysis indicates mono- Grisold W, et al. EFNS Guidelines on diagnosis and
clonal origin of lethal metastatic prostate cancer. Nat treatment of brain metastases: report of an EFNS
Med. 2009;15:559–65. Task Force. Eur J Neurol. 2006;13:674–81.
113. Navin N, Kendall J, Troge J, Andrews P, Rodgers 129. Bruna J, Gonzalez L, Miro J, Velasco R, Gil M,
L, McIndoo J, et al. Tumour evolution inferred by Tortosa A, et al. Leptomeningeal carcinomatosis:
single-cell sequencing. Nature. 2011;472:90–4. prognostic implications of clinical and cerebrospinal
114. Ding L, Ellis MJ, Li S, Larson DE, Chen K, Wallis fluid features. Cancer. 2009;115:381–9.
JW, et al. Genome remodelling in a basal-like 130. Siegal T, Lossos A, Pfeffer MR. Leptomeningeal
breast cancer metastasis and xenograft. Nature. metastases: analysis of 31 patients with sustained
2010;464:999–1005. off-therapy response following combined-modality
115. Xie T, Cho YB, Wang K, Huang D, Hong HK, Choi therapy. Neurology. 1994;44:1463–9.
Y-L, et al. Patterns of somatic alterations between 131. Rubin R, Owens E, Rall D. Transport of methotrexate
matched primary and metastatic colorectal tumors by the choroid plexus. Cancer Res. 1968;28:689–94.
characterized by whole-genome sequencing. 132. Fulton DS, Levin VA, Gutin PH, Edwards MS,
Genomics. 2014;104:234–41. Seager ML, Stewart J, et al. Intrathecal cytosine ara-
116. Shah SP, Morin RD, Khattra J, Prentice L, Pugh T, binoside for the treatment of meningeal metastases
Burleigh A, et al. Mutational evolution in a lobular from malignant brain tumors and systemic tumors.
breast tumour profiled at single nucleotide resolu- Cancer Chemother Pharmacol. 1982;8:285–91.
tion. Nature. 2009;461:809–13. 133. Esteva FJ, Soh LT, Holmes FA, Plunkett W, Meyers
117. Haffner MC, Mosbruger T, Esopi DM, Fedor H, CA, Forman AD, et al. Phase II trial and pharma-
Heaphy CM, Walker DA, et al. Tracking the clonal cokinetic evaluation of cytosine arabinoside for
origin of lethal prostate cancer. J Clin Invest. leptomeningeal metastases from breast cancer.

2013;123:4918–22. Cancer Chemother Pharmacol. 2000;46:382–6.
118. Carter SL, Cibulskis K, Helman E, McKenna A, 134. Glantz MJ, Van Horn A, Fisher R, Chamberlain
Shen H, Zack T, et al. Absolute quantification of MC. Route of intracerebrospinal fluid chemotherapy
somatic DNA alterations in human cancer. Nat administration and efficacy of therapy in neoplastic
Biotechnol. 2012;30:413–21. meningitis. Cancer. 2010;116:1947–52.
119. Nik-Zainal S, Alexandrov LB, Wedge DC, Van Loo 135. Cole BF, Glantz MJ, Jaeckle KA, Chamberlain MC,
P, Greenman CD, Raine K, et al. Mutational pro- Mackowiak JI. Quality-of-life-adjusted survival
cesses molding the genomes of 21 breast cancers. comparison of sustained-release cytosine arabino-
Cell. 2012;149:979–93. side versus intrathecal methotrexate for treatment
120. Pentsova EI, Shah RH, Tang J, Boire A, You D, of solid tumor neoplastic meningitis. Cancer.
Briggs S, et al. Evaluating cancer of the central ner- 2003;97:3053–60.
vous system through next-generation sequencing of 136. Kim D-Y, Lee K-W, Yun T, Park SR, Jung JY, Kim
cerebrospinal fluid. J Clin Oncol. 2016;34:2404–15. D-W, et al. Comparison of intrathecal chemotherapy
121. Jaeckle KA. Neoplastic meningitis from systemic for leptomeningeal carcinomatosis of a solid tumor:
malignancies: diagnosis, prognosis and treatment. methotrexate alone versus methotrexate in combina-
Semin Oncol. 2006;33:312–23. tion with cytosine arabinoside and hydrocortisone.
122. Taillibert S, Chamberlain MC. Leptomeningeal Jpn J Clin Oncol. 2003;33:608–12.
metastasis. Handb Clin Neurol. 2018;149:169–204. 137. Gutin PH, Levi JA, Wiernik PH, Walker
123. Wang N, Bertalan MS, Brastianos MD. Treatment of malignant meningeal disease with
PK. Leptomeningeal metastasis from systemic can- intrathecal thioTEPA: a phase II study. Cancer Treat
cer: review and update on management. Cancer. Rep. 1977;61:885–7.
2018;124:21–35. 138. Comte A, Jdid W, Guilhaume MN, Kriegel I,
124. Chamberlain M, Junck L, Brandsma D, Soffietti R, Piperno-Neumann S, Dieras V, et al. Survival of
Rudà R, Raizer J, et al. Leptomeningeal metasta- breast cancer patients with meningeal carcinoma-
ses: a RANO proposal for response criteria. Neuro- tosis treated by intrathecal thiotepa. J Neuro-Oncol.
Oncology. 2017;19:484–92. 2013;115:445–52.
139. Le Rhun E, Taillibert S, Devos P, Zairi F, Turpin A, 152. Reijneveld JC, Taphoorn MJ, Kerckhaert OA,
Rodrigues I, et al. Salvage intracerebrospinal fluid Drixler TA, Boogerd W, Voest EE. Angiostatin
thiotepa in breast cancer-related leptomeningeal prolongs the survival of mice with leptomeningeal
metastases: a retrospective case series. Anti-Cancer metastases. Eur J Clin Investig. 2003;33:76–81.
Drugs. 2013;24:1093–7. 153. Chen IC, Lin CH, Jan IS, Cheng AL, Lu
140. Glantz MJ, Cole BF, Recht L, Akerley W, Mills P, YS. Bevacizumab might potentiate the chemother-
Saris S, et al. High-dose intravenous methotrexate apeutic effect in breast cancer patients with lepto-
for patients with nonleukemic leptomeningeal can- meningeal carcinomatosis. J Formos Med Assoc.
cer: is intrathecal chemotherapy necessary? J Clin 2016;115:243–8.
Oncol. 1998;16:1561–7. 154. Wu PF, Lin CH, Kuo CH, Chen WW, Yeh DC, Liao
141. Tetef ML, Margolin KA, Doroshow JH, Akman S, HW, et al. A pilot study of bevacizumab combined
Leong LA, Morgan RJ, et al. Pharmacokinetics and with etoposide and cisplatin in breast cancer patients
toxicity of high-dose intravenous methotrexate in the with leptomeningeal carcinomatosis. BMC Cancer.
treatment of leptomeningeal carcinomatosis. Cancer 2015;15:299.
Chemother Pharmacol. 2000;46:19–26. 155. Groves MD. A pilot study of systemically admin-
142. Boogerd W, van den Bent MJ, Koehler PJ, Heimans isteres bevacizumab in patients with neoplastic
JJ, van der Sande JJ, Aaronson NK, et al. The rel- maningitis: imaging, clinical, CSF, and biomarker
evance of intraventricular chemotherapy for lepto- outcomes. Neuro-Oncology. 2011;13:85–91.
meningeal metastasis in breast cancer: a randomised 156. Simeone E, De Maio E, Sandomenico F, Fulciniti F,
study. Eur J Cancer. 2004;40:2726–33. Lastoria S, Aprea P, et al. Neoplastic leptomeningitis
143. Slevin ML, Piall EM, Aherne GW, Harvey VJ, presenting in a melanoma patient treated with dab-
Johnston A, Lister TA. Effect of dose and schedule rafenib (a V600EBRAF inhibitor): a case report. J
on pharmacokinetics of high-dose cytosine arabino- Med Case Rep. 2012;6:131.
side in plasma and cerebrospinal fluid. J Clin Oncol 157. Long GV, Stroyakovskiy D, Gogas H, Levchenko
Off J Am Soc Clin Oncol. 1983;1:546–51. E, de Braud F, Larkin J, et al. Dabrafenib and tra-
144. Lopez JA, Nassif E, Vannicola P, Krikorian JG, metinib versus dabrafenib and placebo for Val600
Agarwal RP. Central nervous system pharmacoki- BRAF-mutant melanoma: a multicentre, double-
netics of high-dose cytosine arabinoside. J Neuro- blind, phase 3 randomised controlled trial. Lancet
Oncol. 1985;3:119–24. London England. 2015;386:444–51.
145. Siegal T. Leptomeningeal metastases: rationale for 158. Robert C, Karaszewska B, Schachter J, Rutkowski P,
systemic chemotherapy or what is the role of intra- Mackiewicz A, Stroiakovski D, et al. Improved over-
CSF-chemotherapy? J Neuro-Oncol. 1998;38:151–7. all survival in melanoma with combined dabrafenib
146. Bokstein F, Lossos A, Siegal T. Leptomeningeal and trametinib. N Engl J Med. 2015;372:30–9.
metastases from solid tumors: a comparison of 159. Ascierto PA, McArthur GA, Dréno B, Atkinson V,
two prospective series treated with and without Liszkay G, Di Giacomo AM, et al. Cobimetinib com-
intra-cerebrospinal fluid chemotherapy. Cancer. bined with vemurafenib in advanced BRAF(V600)-
1998;82:1756–63. mutant melanoma (coBRIM): updated efficacy
147. Shigekawa T, Takeuchi H, Misumi M, Matsuura K, results from a randomised, double-blind, phase 3
Sano H, Fujiuchi N, et al. Successful treatment of trial. Lancet Oncol. 2016;17:1248–60.
leptomeningeal metastases from breast cancer using 160. Yi HG, Kim HJ, Kim YJ, Han S-W, Oh D-Y, Lee
the combination of trastuzumab and capecitabine: a S-H, et al. Epidermal growth factor receptor (EGFR)
case report. Breast Cancer. 2009;16:88–92. tyrosine kinase inhibitors (TKIs) are effective for
148. Ekenel M, Hormigo AM, Peak S, Deangelis LM, leptomeningeal metastasis from non-small cell
Abrey LE. Capecitabine therapy of central nervous lung cancer patients with sensitive EGFR muta-
system metastases from breast cancer. J Neuro- tion or other predictive factors of good response for
Oncol. 2007;85:223–7. EGFR TKI. Lung Cancer Amsterdam Netherlands.
149. Davis TH, Fadul CE, Glantz MJ, et al. Pilot phase 2009;65:80–4.
II trial of temozolomide for leptomeningeal metas- 161. Dhruva N, Socinski MA. Carcinomatous meningitis
tases: preliminary report. J Clin Oncol. 2003; in non-small-cell lung cancer: response to high-dose
22:460. erlotinib. J Clin Oncol Off J Am Soc Clin Oncol.
150. Herrlinger U, Wiendl H, Renninger M, Forschler 2009;27:e31–2.
H, Dichgans J, Weller M. Vascular endothelial 162. Clarke JL, Pao W, Wu N, Miller VA, Lassman
growth factor (VEGF) in leptomeningeal metasta- AB. High dose weekly erlotinib achieves thera-
sis: diagnostic and prognostic value. Br J Cancer. peutic concentrations in CSF and is effective in
2004;91:219–24. leptomeningeal metastases from epidermal growth
151. Reijneveld JC, Brandsma D, Boogerd W, Bonfrer factor receptor mutant lung cancer. J Neuro-Oncol.
JG, Kalmijn S, Voest EE, et al. CSF levels of 2010;99:283–6.
angiogenesis-related proteins in patients with lepto- 163. Cessot A, Blanchet B, Goldwasser F. Erlotinib treat-
meningeal metastases. Neurology. 2005;65:1120–2. ment of meningeal carcinomatosis in lung cancer:
more is better. Ann Oncol Off J Eur Soc Med Oncol. with leptomeningeal carcinomatosis. Lung Cancer
2014;25:2093–4. Amsterdam Netherlands. 2014;85:479–80.
164. Kawamura T, Hata A, Takeshita J, Fujita S, Hayashi 175. Yang JC-H, Cho BC, Kim D-W, Kim S-W, Lee J-S,
M, Tomii K, et al. High-dose erlotinib for refractory Su W-C, et al. Osimertinib for patients (pts) with lep-
leptomeningeal metastases after failure of standard- tomeningeal metastases (LM) from EGFR-mutant
dose EGFR-TKIs. Cancer Chemother Pharmacol. non-small cell lung cancer (NSCLC): updated results
2015;75:1261–6. from the BLOOM study. J Clin Oncol. 2017;35:2020.
165. Kanaji N, Bandoh S, Nagamura N, Kushida Y, Haba 176. Ahn M-J, Kim D-W, Cho BC, Kim S-W, Lee JS,
R, Ishida T. Significance of an epidermal growth Ahn J-S, et al. Activity and safety of AZD3759 in
factor receptor mutation in cerebrospinal fluid for EGFR-mutant non-small-cell lung cancer with CNS
carcinomatous meningitis. Intern Med Tokyo Japan. metastases (BLOOM): a phase 1, open-label, dose-
2007;46:1651–5. escalation and dose-expansion study. Lancet Respir
166. Sakai M, Ishikawa S, Ito H, Ozawa Y, Yamamoto T, Med. 2017;5:891–902.
Onizuka M, et al. Carcinomatous meningitis from 177. Gadgeel SM, Gandhi L, Riely GJ, Chiappori AA,
non-small-cell lung cancer responding to gefitinib. West HL, Azada MC, et al. Safety and activity of
Int J Clin Oncol. 2006;11:243–5. alectinib against systemic disease and brain metas-
167. Hashimoto N, Imaizumi K, Honda T, Kawabe T, tases in patients with crizotinib-resistant ALK-
Nagasaka T, Shimokata K, et al. Successful re-treatment rearranged non-small-cell lung cancer (AF-002JG):
with gefitinib for carcinomatous meningitis as disease results from the dose-finding portion of a phase 1/2
recurrence of non-small-cell lung cancer. Lung Cancer study. Lancet Oncol. 2014;15:1119–28.
Amsterdam Netherlands. 2006;53:387–90. 178. Gainor JF, Chi AS, Logan J, Hu R, Oh KS, Brastianos
168. So T, Inoue M, Chikaishi Y, Nose N, Sugio K, PK, et al. Alectinib dose escalation reinduces central
Yasumoto K. Gefitinib and a ventriculo-peritoneal nervous system responses in patients with anaplastic
shunt to manage carcinomatous meningitis from lymphoma kinase-positive non-small cell lung can-
non-small-cell lung cancer: report of two cases. Surg cer relapsing on standard dose alectinib. J Thorac
Today. 2009;39:598–602. Oncol Off Publ Int Assoc Study Lung Cancer.
169. Jackman DM, Holmes AJ, Lindeman N, Wen PY, 2016;11:256–60.
Kesari S, Borras AM, et al. Response and resis- 179. Gainor JF, Sherman CA, Willoughby K, Logan J,
tance in a non-small-cell lung cancer patient with Kennedy E, Brastianos PK, et al. Alectinib salvages
an epidermal growth factor receptor mutation and CNS relapses in ALK-positive lung cancer patients
leptomeningeal metastases treated with high-dose previously treated with crizotinib and ceritinib.
gefitinib. J Clin Oncol Off J Am Soc Clin Oncol. J Thorac Oncol Off Publ Int Assoc Study Lung
2006;24:4517–20. Cancer. 2015;10:232–6.
170. Burel-Vandenbos F, Ambrosetti D, Coutts M, 180. Ou S-HI, Sommers KR, Azada MC, Garon
Pedeutour F. EGFR mutation status in brain metas- EB. Alectinib induces a durable (>15 months)
tases of non-small cell lung carcinoma. J Neuro- complete response in an ALK-positive non-small
Oncol. 2013;111:1–10. cell lung cancer patient who progressed on crizo-
171. Elmeliegy MA, Carcaboso AM, Tagen M, Bai F, tinib with diffuse leptomeningeal carcinomatosis.
Stewart CF. Role of ATP-binding cassette and solute Oncologist. 2015;20:224–6.
carrier transporters in erlotinib CNS penetration and 181. Roth P, Weller M. Management of neoplastic menin-
intracellular accumulation. Clin Cancer Res Off J gitis. Chin Clin Oncol. 2015;4:26.
Am Assoc Cancer Res. 2011;17:89–99. 182. Weller M, Stupp R, Wick W. Epilepsy meets cancer:
172. Hoffknecht P, Tufman A, Wehler T, Pelzer T, when, why, and what to do about it? Lancet Oncol.
Wiewrodt R, Schütz M, et al. Efficacy of the irre- 2012;13:e375–82.
versible ErbB family blocker afatinib in epidermal 183. Long GV, Atkinson V, Lo S, Sandhu S, Guminski
growth factor receptor (EGFR) tyrosine kinase AD, Brown MP, et al. Combination nivolumab and
inhibitor (TKI)-pretreated non-small-cell lung can- ipilimumab or nivolumab alone in melanoma brain
cer patients with brain metastases or leptomeningeal metastases: a multicentre randomised phase 2 study.
disease. J Thorac Oncol Off Publ Int Assoc Study Lancet Oncol. 2018;19:672–81.
Lung Cancer. 2015;10:156–63. 184. Chamberlain MC. A phase II trial of intra-
173. Tamiya M, Shiroyama T, Nishihara T, Nishida cerebrospinal fluid alpha interferon in the treatment
T, Hayama M, Tanaka A, et al. Afatinib success- of neoplastic meningitis. Cancer. 2002;94:2675–80.
fully treated leptomeningeal metastasis during 185. Ward BG, Cruickshank DJ. Circulating tumor-
erlotinib treatment in a patient with EGFR-mutant associated antigen detected by the monoclonal anti-
(Exon18:G719S) lung adenocarcinoma as a body HMFG2 in human epithelial ovarian cancer. Int
second-line chemotherapy. Asia Pac J Clin Oncol. J Cancer. 1987;39:30–3.
2017;13:e531–3. 186. Moseley RP, Papanastassiou V, Zalutsky MR,
174. Lin C-H, Lin M-T, Kuo Y-W, Ho C-C. Afatinib Ashpole RD, Evans S, Bigner DD, et al.
combined with cetuximab for lung adenocarcinoma Immunoreactivity, pharmacokinetics and bone mar-
row dosimetry of intrathecal radioimmunoconju- neoplastic meningitis: a Pediatric Brain Tumor

gates. Int J Cancer. 1992;52:38–43. Consortium Study (PBTC-004). Clin Cancer Res
187. Kramer K, Humm JL, Souweidane MM, Zanzonico Off J Am Assoc Cancer Res. 2006;12:1540–6.
PB, Dunkel IJ, Gerald WL, et al. Phase I study of tar- 194. Friedman HS, Archer GE, McLendon RE, Schuster
geted radioimmunotherapy for leptomeningeal can- JM, Colvin OM, Guaspari A, et al. Intrathecal mel-
cers using intra-Ommaya 131-I-3F8. J Clin Oncol phalan therapy of human neoplastic meningitis in
Off J Am Soc Clin Oncol. 2007;25:5465–70. athymic nude rats. Cancer Res. 1994;54:4710–4.
188. Kramer K, Kushner BH, Modak S, Pandit-Taskar N, 195. Kochi M, Kuratsu J, Mihara Y, Takaki S, Inoue N,
Smith-Jones P, Zanzonico P, et al. Compartmental Sueyoshi N, et al. Neurotoxicity and pharmacokinet-
intrathecal radioimmunotherapy: results for treat- ics of intrathecal perfusion of ACNU in dogs. Cancer
ment for metastatic CNS neuroblastoma. J Neuro- Res. 1990;50:3119–23.
Oncol. 2010;97:409–18. 196. Champagne MA, Silver HK. Intrathecal dacarbazine
189. Slavc I, Schuller E, Falger J, Günes M, Pillwein K, treatment of leptomeningeal malignant melanoma. J
Czech T, et al. Feasibility of long-term intraventricu- Natl Cancer Inst. 1992;84:1203–4.
lar therapy with mafosfamide (n = 26) and etoposide 197. Cordone I, Masi S, Summa V, Carosi M, Vidiri A,
(n = 11): experience in 26 children with dissemi- Fabi A, et al. Overexpression of syndecan-1, MUC-
nated malignant brain tumors. J Neuro-Oncol. 1, and putative stem cell markers in breast cancer
2003;64:239–47. leptomeningeal metastasis: a cerebrospinal fluid
190. Fleischhack G, Jaehde U, Bode U. Pharmacokinetics flow cytometry study. Breast Cancer Res BCR.
following intraventricular administration of chemo- 2017;19:46.
therapy in patients with neoplastic meningitis. Clin 198. Gold DR, Nadel RE, Vangelakos CG, Davis MJ,
Pharmacokinet. 2005;44:1–31. Livingston MY, Heath JE, et al. Pearls and oy-sters:
191. Chamberlain MC, Tsao-Wei DD, Groshen S. Phase the utility of cytology and flow cytometry in the
II trial of intracerebrospinal fluid etoposide in diagnosis of leptomeningeal leukemia. Neurology.
the treatment of neoplastic meningitis. Cancer. 2013;80:e156–9.
2006;106:2021–7. 199. Subirá D, Serrano C, Castañón S, Gonzalo R, Illán J,
192. Groves MD, Glantz MJ, Chamberlain MC, Pardo J, et al. Role of flow cytometry immunophe-
Baumgartner KE, Conrad CA, Hsu S, et al. A notyping in the diagnosis of leptomeningeal carcino-
multicenter phase II trial of intrathecal topotecan matosis. Neuro-Oncology. 2012;14:43–52.
in patients with meningeal malignancies. Neuro- 200. Subirá D, Simó M, Illán J, Serrano C, Castañón S,
Oncology. 2008;10:208–15. Gonzalo R, et al. Diagnostic and prognostic sig-
193. Gururangan S, Petros WP, Poussaint TY, Hancock nificance of flow cytometry immunophenotyping in
ML, Phillips PC, Friedman HS, et al. Phase I trial patients with leptomeningeal carcinomatosis. Clin
of intrathecal spartaject busulfan in children with Exp Metastasis. 2015;32:383–91.
Paraneoplastic Neurological
Disorders
13
Monica Weaver Buckley and John C. Probasco
Introduction inciting cancer often asymptomatic or occult. The

presentations of PNDs are diverse and variable,
Neurological complications are common in patients reflecting the potential involvement of multiple
with cancer and can arise from toxic, infectious, or areas of the nervous system in isolation or simulta-
metabolic abnormalities. Additionally, neurologi- neously. The presenting symptoms are dependent
cal complications can be due to the direct effect on the areas of the nervous system affected. PNDs
of the cancer from brain metastasis, spinal cord can affect the central nervous system (e.g., limbic
and nerve root compression, leptomeningeal dis- encephalitis, paraneoplastic cerebellar degenera-
ease, and side effects of medications. Patients with tion), spinal cord (e.g., necrotizing myelopathy,
cancer can also develop paraneoplastic neurologi- tractopathies), peripheral nervous system (e.g.,
cal disorders (PNDs). PNDs are syndromes that subacute sensory neuropathy), neuromuscular
involve nervous system organs physically remote junction (e.g., myasthenia gravis and Lambert-
from a malignant neoplasm or metastasis, and Eaton myasthenia syndrome), and muscle (e.g.,
PNDs can lead to significant disability and even necrotizing myopathies) [4]. It is important to note
death [1]. Most PNDs are subacute and progres- that these syndromes can also occur in the absence
sive, with onset of symptoms over the course of of cancer. For example, 70% of limbic encephalitis
weeks to months. PNDs can precede the detec- and 85–90% of myasthenia gravis are not associ-
tion of a cancer or its recurrence by years [2]. A ated with malignancy [5]. Therefore, presence of
patient presenting with subacute symptoms with these clinical syndromes does not necessarily indi-
neurological findings and risk factors including cate the presence of a malignancy.
personal history of smoking, cancer, or autoim-
mune disease or a family history which includes
cancer or autoimmune disease raises the suspicion Epidemiology
for a PND [1]. PNDs are thought to be the product
of immune cross-reactivity between tumor cells Symptomatic PNDs are rare and have variable
and normal components of the nervous system prevalence, affecting approximately 0.01–0.2%
(Fig. 13.1) [3]. This immune response can also of all patients with cancer; however, this may
be effective against a systemic cancer, with the be an underestimation [4]. The age of onset of
PNDs is variable, typically occurring in the sixth
M. W. Buckley · J. C. Probasco (*) to seventh decades. However, pediatric PNDs
Department of Neurology, Johns Hopkins University have been reported, often in association with
School of Medicine, Baltimore, MD, USA neuroblastoma [6, 7]. The prevalence of a PND
e-mail: jprobas1@jhmi.edu

https://doi.org/10.1007/978-3-030-42958-4_13
188 M. W. Buckley and J. C. Probasco
Antitumor immune response

Lymph node
DC
Tumor
T cell B cell
Blood brain barrier

Paraneoplastic neurological
disorder
b Central nervous system c Peripheral nervous system
Fig. 13.1 Pathogenesis of paraneoplastic neurological moting antitumor immunity. (b) T and B cells specific for
disorders (PNDs) onconeural antigens and onconeural autoantibodies cross
(a) Onconeural antigens are expressed by tumor cells. the blood-brain barrier to react with neuronal cells
Apoptosis and necrosis of tumor cells cause release of expressing onconeural antigens and trigger PNDs in the
onconeural antigens that are phagocytosed by antigen- central nervous system. (c) T and B cells specific for
presenting cells (APCs). In the lymph node (LN), den- onconeural antigens and onconeural autoantibodies react
dritic cells (DCs) present onconeural peptides to T and B with peripheral nerves, neuromuscular junction, or mus-
cells and activate the adaptive immune response, thus pro- cles and trigger PNDs
is dependent on the type of cancer, as certain docrine proteins (small-cell lung cancer, neuro-
malignancies have substantially higher incidence blastoma), contain mature or immature neuronal
of PNDs. For example, 50% of patients with the tissues (teratomas), involve immunoregulatory
rare osteosclerotic form of plasmacytoma pres- organs (thymoma), or produce immunoglobu-
ent with polyneuropathy, organomegaly, endo- lins (plasma-cell dyscrasias, B-cell lymphomas)
crinopathy, monoclonal gammopathy, and skin [4, 9]. Furthermore, many of these cancers fre-
changes (POEMS syndrome) and have demyelin- quently metastasize to regional lymph nodes,
ating peripheral neuropathy, while only 10–15% which promotes early recognition and prim-
of patients with a thymoma present with myas- ing of the immune response. Unlike paraneo-
thenia gravis [5]. Close neurological exams and plastic endocrine syndromes which generally
electrophysiological studies in asymptomatic present after the diagnosis of cancer, PNDs
patients with small-cell lung cancer have dem- are detected prior to the diagnosis of cancer in
onstrated subtle proximal weakness or delayed approximately 80% of cases [5]. Therefore, it is
conduction along peripheral nerves further sug- crucial to closely screen all patients presenting
gesting that the true incidence of PNDs may be with PNDs for malignancy given the possibility
higher. Furthermore, in one prospective, 5-year of detection and diagnosis of an occult cancer
study of patients with small-cell lung cancer, 9% at an early and potentially highly treatable stage
developed a PND [8]. [2]. Furthermore, prior research has suggested
Among patients with PNDs, overrepresented that patients who present with PNDs may have
cancers include cancers that express neuroen- improved survival and a more benign cancer
13 Paraneoplastic Neurological Disorders 189
course as the immune response that drives the classical PNDs, there are also well-characterized
PND may also target the tumor [10]. onconeural autoantibodies that are frequently
associated with malignancy and PNDs, includ-
ing anti-Hu, anti-Yo, anti-CRMP5/CV2, anti-Ri,
Diagnosis of Paraneoplastic anti-Ma1/Ma2, and anti-amphiphysin [6, 14].
Neurological Disorders Interestingly, certain autoantibodies found in
PNDs are more closely associated with specific
As neurological symptoms are common in cancers types as compared to specific neurologi-
patients with cancer and are most commonly due cal syndromes (Table 13.1) [15]. The association
to multiple etiologies including infection, elec- with cancer for certain classical PNDs and well-
trolyte abnormalities, medication side effects, or characterized autoantibodies is so specific that if
metastasis, it is important to distinguish between occult cancer is not identified at time of diagnosis
neurological syndromes that coincide with the of PND, it is recommended that the patient follow
presence of cancer and true PNDs. Additionally, up with surveillance imaging every 3–6 months
autoimmune neurological disorders can occur for 2–3 years [2]. For example, the presence
in the absence of malignancy and are frequently of anti-Yo antibodies and cerebellar degenera-
associated with the same autoantibodies against tion is highly suggestive for adenocarcinoma of
neuronal antigens [11]. PNDs can also present the ovary, uterus, fallopian tube, peritoneum, or
without identifiable antibodies against neuronal breast, and these cancers are found in 90% of
antigens, either due to lack of a humoral immune patients presenting with this classical PND and
response or technical limitations in identifying autoantibody [16, 17]. If another malignancy is
the autoantibody. Furthermore, patients can pres- identified, it is recommended that workup is pur-
ent with detectable autoantibodies but no asso- sued to diagnose a second more commonly asso-
ciated PND [12, 13]. The diagnosis of PNDs ciated tumor [14].
incorporates clinical presentation and neuro- PNDs have variable presentations reflecting
logical findings, detected cancer or cancer recur- involvement of multiple areas of the nervous
rence, imaging findings, electroencephalography system [15]. Nonclassical PNDs are neurologi-
(EEG), cerebral spinal fluid (CSF) analysis for cal syndromes that are not as closely associated
signs of inflammation, and electromyography with malignancy or specific onconeural antibod-
and nerve conduction studies (EMG/NCS) [14]. ies. Nonclassical PNDs include syndromes of the
PNDs are subdivided in “classical” and CNS (optic neuritis, brainstem encephalitis, and
“nonclassical” syndromes according to guide- stiff person syndrome), syndromes of the periph-
lines proposed in 2004 by an international panel eral nervous system (neuropathy, vasculitis, and
of neurologists with expertise in PNDs [14]. brachial neuritis), and syndromes of the neuro-
Classical paraneoplastic neurological disorders muscular junction and muscle (myasthenia gravis
are syndromes that are strongly associated with and acute necrotizing myopathy) [14].
certain types of cancer and specific autoantibod- Specific criteria are used to diagnose PNDs
ies. Classical PNDs include syndromes of the as it is important to distinguish between PNDs
CNS (limbic encephalitis, subacute cerebellar and neurological symptoms that coexist with
degeneration, encephalomyelitis, and opsoclonus- cancer. The 2004 consensus panel developed
myoclonus), syndromes of the peripheral ner- criteria to distinguish “definite” from “possible”
vous system (subacute sensory neuronopathy and “unlikely” PNDs. Definite criteria include
and chronic gastrointestinal pseudo-occlusion (1) classical syndrome associated with cancer
syndrome), and syndromes of the neuromuscu- diagnosis within 5 years, (2) nonclassical syn-
lar junction and muscle (Lambert-Eaton myas- drome associated with cancer that improves with
thenic syndrome and dermatomyositis) [14]. treatment of the cancer but no immunotherapy,
These PNDs are generally associated with typical (3) nonclassical syndrome with onconeural auto-
autoantibodies and malignancies. In addition to antibodies and cancer diagnosis, and (4) classi-
Table 13.1 Onconeural antibodies associated with cancer (well characterized or partially characterized antibodies)
190
Nonclassical PND
Antigen location Antibody Antigen Classical PND associated associated Associated cancer
Nuclear Anti-Hu Hu family of RNA binding proteins Limbic encephalitis Myelitis Small-cell lung
(ANNA-1) Extrapulmonary small cell
Subacute cerebellar Autonomic neuropathy Neuroblastoma
degeneration
Sensory neuronopathy Peripheral neuropathy Thymoma
Encephalitis
Anti-Ri (ANNA-2) NOVA family of RNA binding Subacute cerebellar Myelopathy Small-cell lung
proteins degeneration
Opsoclonus-myoclonus Peripheral neuropathy Breast
Encephalitis Neuroblastoma
Anti-Ma1/Ma2 Homologous 40 and 42 kDa Limbic encephalitis Encephalitis Testicular germ cell
neuronal nuclear proteins of Subacute cerebellar Hypothalamic Non-small cell lung
uncertain function degeneration encephalitis
Encephalomyelitis Colon
Breast
Anti-Sox-1 Transcription factor sex Lambert-Eaton Encephalitis Small-cell lung
(AGNA) determining region Y (SRY)-box-1 myasthenia
(SOX-1) protein
ANNA-3 170 kDa protein of unclear Limbic encephalitis Encephalitis Small-cell lung
significance Subacute cerebellar Brainstem encephalitis Gastrointestinal
degeneration Neuropathy
M. W. Buckley and J. C. Probasco
Cytoplasmic Anti-Yo (PCA-1) Cerebellar degeneration-related Subacute cerebellar Brainstem encephalitis Ovarian
protein 2 (cdr2) degeneration Myelopathy Fallopian tubes
Peripheral neuropathy Endometrium
Cervix
Breast
PCA-2 280 kDa of unknown identity or Subacute cerebellar Autonomic neuropathy Small-cell lung
function degeneration Peripheral neuropathy
Encephalitis
Anti-Tr (PCA-Tr) Delta/Notch-like epidermal growth Limbic encephalitis Autonomic neuropathy Hodgkin’s lymphoma
factor Subacute cerebellar Non-Hodgkin’s lymphoma
degeneration
Anti-CV2 Collapsin response-mediator Limbic encephalitis Encephalitis Small-cell lung
(CRMP5) protein 5 (CRMP5) Subacute cerebellar Cranial neuropathies Thymoma
degeneration
13 Paraneoplastic Neurological Disorders
Sensory neuronopathy Uveitis Uterine sarcoma

Optic neuritis
Myelopathy
Retinopathy
Peripheral neuropathy
Autonomic neuropathy
Anti-Zic 4 Zic proteins Subacute cerebellar Small-cell lung
degeneration
Anti-recoverin Recoverin Retinopathy Small-cell lung
Intracellular synaptic Anti-Amphiphysin Amphiphysin Limbic encephalitis Stiff person syndrome Breast
Subacute cerebellar Encephalitis Small-cell lung
degeneration
Sensory neuronopathy Myelopathy Thymoma
Anti-GAD 65 65 kDa glutamic acid Limbic encephalitis Stiff person syndrome Thymoma
decarboxylase enzyme Epilepsy Renal cell
Brainstem encephalitis Breast
Cerebellar ataxia Colon
Myelopathy
Extracellular synaptic Anti-mGluR1 Metabotropic glutamate receptor 1 Subacute cerebellar Hodgkin’s lymphoma
degeneration
Anti-mGluR5 Metabotropic glutamate receptor 5 Limbic encephalitis Hodgkin’s lymphoma
(continued)
191
192
Nonclassical PND
Antigen location Antibody Antigen Classical PND associated associated Associated cancer
Cell surface membrane Anti-AChR Muscle acetylcholine receptor Myasthenia gravis Thymoma
antigens Anti-ganglionic Neuronal acetylcholine receptor α3 Encephalopathy Small-cell lung
N-acetylcholine subunits Subacute Thymoma
receptor pandysautonomia
Anti-NMDAR NR1 subunits of N-methyl-D- Encephalitis Ovarian teratoma
aspartate receptor Anxiety Neuroblastoma
Psychosis Small-cell lung
Epilepsy Testicular germ cell
Extrapyramidal disorder
Central dysautonomia
Anti-GlyR Glycine receptor α1 Limbic encephalitis Stiff person syndrome Infrequent association
Ataxia
Hyperekplexia
Progressive
encephalomyelitis with
rigidity and myoclonus
Anti-VGCC N-type and P/Q-type voltage-gated Subacute cerebellar Lambert-Eaton myasthenia Small-cell lung
calcium channels degeneration Encephalitis Breast
Ovarian
Anti-VGKC Voltage-gated potassium channel Limbic encephalitis Encephalitis Small-cell lung
(VGKC) complex subunits Epilepsy Thymoma
Psychiatric symptoms Breast
Hypothalamic disorders Prostate
Anti-LGI1 Leucine-rich, glioma-inactivated Limbic encephalitis Faciobrachial dystonic Thymoma
protein interacts with Kv1 channels seizures
and AMPAR Abnormal sleep
Anti-CASPR2 Contactin-associated protein-like 2 Encephalitis Thymoma
of the neurexin IV superfamily Morvan syndrome
interacts with axonal Kv1 channels Neuromyotonia
M. W. Buckley and J. C. Probasco
Anti-DPPX Dipeptidyl-peptidase-like protein-6 Encephalitis Lymphoma
subunit of Kv4.2 potassium Psychiatric symptoms
channel Tremor
Nystagmus,
hyperekplexia
Ataxia
Progressive
encephalomyelitis with
rigidity and myoclonus
Anti-GABABR γ-Amino butyric acid-B receptor Limbic encephalitis Orolingual dyskinesia Small-cell lung
Breast
Anti-AMPAR Glutamate receptors 1 and 2 Limbic encephalitis Epilepsy Thymoma
Nystagmus Breast
Lung
13 Paraneoplastic Neurological Disorders
Anti-aquaporin 4 Aquaporin-4 channel Neuromyelitis optica Infrequent association

Encephalitis
Data from Probasco JC. Paraneoplastic Neurological Disorders. In: Johnsonton MV, Adams HP, Fatemi A. Neurobiology of Disease. 2nd ed. New York: Oxford University Press;
2016. p. 657–66
193
cal or nonclassical syndrome without cancer but associated with an atypical cancer, an attempt
with well-characterized onconeural antibodies should be completed to identify the onconeural
[14]. Probable PNDs are defined as (1) classi- protein on the atypical tumor (or to identify the
cal syndromes with high risk of cancer but no co-occurrence of a second more typical cancer)
cancer diagnosis or onconeural autoantibodies, [14, 20]. Interestingly, a high mutational burden
(2) neurological syndrome without cancer and in cancers is not associated with PNDs, and there
with partially characterized onconeural antibod- is no evidence to suggest that there are frequently
ies, and (3) nonclassical syndromes with cancer mutations in the genes for onconeural proteins
diagnosis within 2 years of neurological disorder in tumor cells [21]. Therefore, the observed
development but no onconeural autoantibodies immune responses are not due to infrequency
[14]. As previously mentioned, in patients with of the expression of relevant tumor antigens or
definite PND with no associated cancer or prob- mutations in genes encoding onconeural pro-
able PNDs, it is important to complete a dili- teins. Instead, autoimmunization occurs in an
gent screening for associated malignancies and inflammatory environment in response to the pro-
closely monitor with repeat screening as malig- duction of proteins by tumor cells that are usually
nancies may be identified years after the present- restricted to neural cells [9]. In fact, the presence
ing PND [2, 9, 18]. of onconeural protein expression on tumor cells
does not necessarily indicate that an immune
response will be generated against the onconeu-
Etiology and Pathogenesis ral protein; patients may instead develop T-cell
tolerance to onconeural proteins expressed in
The study of PNDs allows for exploration of the tumors, and tumors may be able to evade immune
hypothesis of immune surveillance and tolerance surveillance [22].
as immunological mechanisms form the link The onconeural antibodies produced can
between malignancy and development of these either be a driving mechanism of pathogenesis
syndromes. Nearly all PNDs result from genera- or a marker of immunological activity. It has
tion of immune response to onconeural antigens been suggested that cellular immunity is the
of tumors that cross-react with the nervous sys- main driver of PNDs. Onconeural antibodies are
tem. This section will discuss the immunological classified into three categories based on their
mechanisms driving the development of PNDs. association with malignancy: (1) molecularly
well characterized with strong association with
malignancy, (2) partially characterized with less
nconeural Antigens and Onconeural
O well-described association with malignancy, and
Antibodies (3) associated with both cancer- and non-cancer-
associated syndromes [5]. As previously dis-
One of the main findings supporting an antitumor cussed, the specific onconeural antibody is taken
immune response driving PNDs is the presence into consideration when diagnosing a patient
of high titers of onconeural-specific autoanti- with a definite or probable PND [14].
bodies in the serum and/or CSF of patients with
PNDs [9, 19]. Onconeural proteins are proteins
expressed by a tumor that are similar to pro- ctivation of Immune Response
A
teins that are otherwise only expressed by neu- Against Onconeural Proteins
ronal cells. These proteins are either expressed
intracellularly within the nucleus, nucleolus, Studies of patients with PNDs and immunohis-
or cytoplasm or are expressed on the plasma topathological studies of biopsied and autopsied
membrane [3]. Onconeural proteins are present neural tissues have shed light into the pathogenesis
in tumors of all patients with antibody-positive of PNDs (Fig. 13.1). The immunological mecha-
PND and detected cancer, and when a PND is nism priming the immune response involves the
tumor microenvironment and extracellular space demonstrated multifocal inflammatory changes

surrounding tumors, tumor draining lymph nodes with perivascular and parenchymal CD8+T cells
(LNs), and immune cells such as antigen-present- throughout the cerebellum, brainstem, and spinal
ing cells (APCs), including dendritic cells (DCs), cord in patients with cerebellar degeneration and
and CD4+specific T cells (Fig. 13.1) [3, 20, 23]. seropositive for anti-Yo [26]. Cytotoxic CD8+T
As tumor cells undergo apoptosis and necrosis, cells recognize MHC-I presented peptides on
intracellular onconeural peptides are released the surface of tumor cells and induce apoptosis
into a pro-inflammatory extracellular environ- or enzymatic lysis of tumor cells. Intracellular
ment. These tumor-derived onconeural pep- debris becomes available in the draining lymph
tides are taken up and processed by DCs in this node for binding by onconeural antibodies that
pro-inflammatory environment and taken to the may further amplify the immune response.
draining LN [20]. Following onconeural peptide Antibodies specific for onconeural antigens also
capture, DCs mature and enter T-cell-rich zones have tumoricidal potential through activation of
of the tumor-draining lymph nodes. There, DCs the complement cascade as well as Fc-receptor
present onconeural peptides to both CD4+and activation leading to cell death and antigen inter-
CD8+T cells. Onconeural peptide-specific helper nalization [26].
CD4+T cells are required for the activation of
similarly specific CD8+T cells. Prior studies sug-
gest that CD8+T-cell stimulation in the absence Intracellular Onconeural Proteins
of CD4+help leads to the death of the onconeural
antigen-specific CD8+T cells and promotes toler- Onconeural peptide-specific autoantibodies are
ance to these antigens [21, 23]. Activated helper a fundamental finding in patients with PNDs.
CD4+ T cells also provide signals to onconeural However, in studies involving the transfer of
peptide-specific B cells to proliferate and dif- autoantibodies from patients with PNDs into
ferentiate into onconeural antibody-producing animals, the animals frequently did not develop
plasma cells [20] (Fig. 13.1). These onconeu- any neurological abnormalities suggestive of the
ral antibodies can be directly pathogenic or be PNDs [3, 27]. These were the first studies sug-
markers of immunological activity. The cellular gesting that in certain PNDs, detected autoanti-
location of the target (intracellular or extracellu- bodies are makers of respective immune response
lar) tends to dictate whether the humoral or cel- rather than directly pathogenic. Instead, the driv-
lular immune response is the driving pathogenic ing immune response is thought to be a robust
response [20]. T-cell-mediated response against the neuronal
The priming of the antitumor immune response antigens.
can be effective at killing cancer cells and limit- In addition to the immune responses occur-
ing the development and growth of tumors. This ring in the tumor and the tumor draining lymph
is highlighted by the fact that patients with PNDs node, the systemic inflammatory state also causes
typically will have more successful cancer treat- changes within the nervous system that contrib-
ment and more benign cancer course as compared ute to the development of PNDs. In the setting
to patients without PNDs. Additionally, patients of inflammation, there are changes in the pro-
with detectable autoantibodies but no PND have cessing of self-peptides by neuronal cells; pro-
more limited cancer burden and improved prog- inflammatory cytokines such as interferon-γ
nosis [12, 13]. In patients with anti-Yo para- switch the proteasome in neuronal cells that gen-
neoplastic subacute cerebellar degeneration, erates peptides targeted for degradation into an
circulating onconeural peptide- specific CD8+T immunoproteasome [3]. The immunoproteasome
cells have been demonstrated to lyse target cells generates unique peptide fragments that are not
presenting the onconeural peptide on major his- recognized as “self-peptides.” These “non-self”
tocompatibility complex class I (MHC-I) mol- autoantigenic peptides are similar to the onco-
ecules [24, 25]. Autopsied CNS tissues have also neural antigens presented on distant tumor cells,
and these peptides are presented by MHC-I on release of acetylcholine into the neuromuscular
neuronal cells [3]. CD8+T cells primed in the junction causing the symptoms of muscle weak-
draining LN of the tumor migrate into the sys- ness [30]. Autoantibodies have been shown to
temic circulation and exit the systemic circula- be directly pathogenic as injection of polyclonal
tion to attack peripheral neurons or cross the IgG isolated from serum of patients with LEMS
blood-brain barrier into the CNS parenchyma [3, is sufficient to transfer the clinical syndrome to
23] (Fig. 13.1). The CD8+T cells recognize the laboratory mice [32]. In myasthenia gravis asso-
“non-self” autoantigenic peptides presented on ciated with anti-AChR antibodies, binding of
neuronal cells and cause neuronal damage and anti-AchR antibody leads to functional block-
cell loss. ade of the receptor from acetylcholine binding,
accelerated endocytosis and degradation of the
receptors, and overall decreased numbers of ace-
Cell Plasma Onconeural Proteins tylcholine receptors at the neuromuscular junc-
tion [3].
Onconeural autoantibodies can play a crucial Anti-NMDAR encephalitis is another PND
role in neurotoxicity in the setting of PNDs, with an autoantibody that is directly pathogenic.
particularly those associated with neural cell In anti-NMDAR encephalitis, autoantibodies are
plasma membrane receptors and channels. directed against the NR1 subunit of the NMDA
Autoantibodies targeting cell surface mem- glutamate receptor. Clinically, assay of the CSF
brane proteins can lead to neuronal dysfunction for the presence of anti-NMDAR antibodies is
and injury through several mechanisms [20]. more sensitive than serological testing as intra-
Autoantibodies targeting cell surface proteins can thecal synthesis of this autoantibody has been
act as agonists or antagonists and lead to cellular demonstrated [28, 33]. Autopsy studies have also
dysfunction by altering signaling through recep- demonstrated the crucial role of autoantibodies in
tors and channels [28–30]. Autoantibodies can the pathogenesis of anti-NMDAR encephalitis as
also cause direct cellular damage through acti- significant deposits of IgG have been identified
vation of the complement cascade or Fc recep- throughout the CNS with predominance in the
tors leading to antibody-dependent cell-mediated hippocampus. Furthermore, B cells and antibody-
cytotoxicity. Finally, antibodies may lead to their secreting plasma cells are more frequently identi-
internalization and thus decrease the density of fied than T-cell infiltrates [3]. Studies in vitro and
a target receptor or channel on the cell surface in vivo using anti-NMDAR autoantibodies iso-
causing neuronal dysfunction [31]. Examples of lated from the sera or CSF of patients with anti-
PNDs with a direct pathogenic role of autoanti- NMDAR encephalitis have demonstrated that
bodies targeting extracellular onconeural pro- the mechanism of action is antibody-mediated
teins include antibodies against voltage-gated capping and internalization leading to decrease
calcium channels (anti- VGCC, Lambert-Eaton in the density and localization of NMDAR clus-
myasthenia syndrome), acetylcholine receptor ters [28]. Passive transfer of anti-NMDAR auto-
antibody (anti-AchR, myasthenia gravis), and antibodies causes transfer of disease symptoms
N-methyl-D-aspartate receptor (anti-NMDAR, to mice [34]. Additionally, decrease in the CSF
anti-NMDAR encephalitis) [3]. anti-NMDAR autoantibody concentration is cor-
PNDs involving the neuromuscular junction related with clinical improvement and response
and peripheral nerves have strong evidence for to treatment.
the pathogenic role of autoantibodies targeting
onconeural proteins. Lambert-Eaton myasthenic
syndrome is associated with anti-VGCC antibod- Distinct HLA Associations with PNDs
ies (typically P/Q type). Antibody binding leads
to impairment in postsynaptic signal transduction Studies have been completed to investigate how
with decreased calcium ion entry and reduced genetic susceptibility affects the development
of autoimmune encephalitis, both paraneoplas- ral autoantibody is more predictive of the type of
tic and non-paraneoplastic. Specific HLA genes malignancy than the clinical syndrome [15]. This
have been associated with various neurological section will not discuss dermatomyositis, which
autoimmune diseases such as muscle-specific is a classical PND and is characterized by an
kinase antibody-positive myasthenia gravis. inflammatory myopathy.
Distinct HLA subtypes were also found in
patients with anti-leucine-rich glioma-inactivated
1 (anti-LGI1) and anti-contactin-associated pro- Subacute Cerebellar Degeneration
tein 2 (anti-CASPR2) autoimmune encephalitis
[35–37]. This finding further supports the crucial Subacute cerebellar degeneration is one of the
role that CD4+ T lymphocytes play in the patho- most common PNDs. It is characterized initially
genesis of anti-LGI1 and anti-CASPR2 antibody by nonspecific symptoms including dizziness,
encephalitis. Interestingly, these same studies nausea, and vomiting with rapid progression to
indicated that anti-NMDAR encephalitis was not ataxia, diplopia, dysarthria, and dysphagia [26].
associated with any specific HLA alleles. Initial magnetic resonance imaging (MRI) may
There is evidence to support both cell-mediated be normal or show only subtle changes (see
and humoral autoantibody-mediated processes in Fig. 13.2i, j). As the disease progresses, MRI dem-
the pathogenesis of PNDs. The pathogenic role onstrates cerebellar atrophy [26]. The onconeural
of each appears to be partially dependent on the autoantibodies that are most associated with para-
cellular location of the onconeural protein target. neoplastic cerebellar degeneration include anti-
PNDs involving intracellular targets are driven Yo, which is typically seen with gynecological or
by a robust cytotoxic T-cell response, and the breast cancers, and anti-Tr, which is found with
presence of detectable autoantibodies is a marker Hodgkin’s lymphoma [16, 18, 39]. Patients with
of immune system activity. However, in PNDs small-cell lung cancer can also develop paraneo-
involving extracellular cell membrane proteins, plastic subacute cerebellar degeneration. It is often
antibodies may play a central role and have been coincident with other paraneoplastic syndromes,
shown to be directly pathogenic in some syn- and patients will frequently have onconeural auto-
dromes. However, as highlighted by the striking antibodies positive for anti-Hu or autoantibodies
HLA allelic association with anti-LGI1 and anti- directed at the voltage-gated potassium channel
CASPR2 encephalitis, T cells still play a crucial complex (VGKC) without further specifications
role in these syndromes. [40]. The pathological findings associated with
paraneoplastic subacute cerebellar degeneration
include a relatively specific and extensive loss
ommon Classical Paraneoplastic
C of Purkinje cells with associated inflammatory
Neurological Disorders infiltrates during the early stages of the disease
[4, 41]. Treatment includes management of the
Malignancy can trigger both classical and non- underlying cancer and immunotherapy; however,
classical PNDs. This section will discuss some of it is generally poorly responsive to immuno-
the common classical paraneoplastic neurological therapy, especially in those patients seropositive
syndromes that have been associated with overt for anti-Yo antibodies [42]. When evaluating a
or occult malignancy including clinical features, patient with subacute paraneoplastic cerebellar
diagnostic findings, onconeural autoantibodies, degeneration, it is important to recall that approx-
and closely mimicking neurological disorders. imately 50% of cases are not paraneoplastic
Of note, specific onconeural autoantibodies can in origin [42]. Other diagnostic considerations
be associated with multiple clinical disorders, include vitamin deficiency (thiamine, vitamin E),
and approximately 31% of patients with PNDs non-paraneoplastic autoimmune cerebellar ataxia,
have multiple identified autoantibodies [38]. alcohol toxicity, infectious or postinfectious cer-
Prior studies have shown that a specific onconeu- ebellitis, and Creutzfeldt-Jakob disease [4].
a b
c d
Fig. 13.2 MRI findings in paraneoplastic encephalitis FLAIR MRI of a patient with anti-NMDAR encephalitis
and subacute cerebellar degeneration demonstrating right hippocampal atrophy. (g) Acute-
(a and b) T2/FLAIR and T1 post-gadolinium brain phase T2/FLAIR MRI of a patient with anti-LGI1 enceph-
MRI of a patient with anti-Hu limbic encephalitis demon- alitis demonstrating bilateral medial temporal lobe T2
strating atrophy and non-enhancing T2 hyperintensities of hyperintensities. (h) Convalescent-phase T2/FLAIR MRI
the left more than the right hippocampus. (c and d) A of the same patient demonstrating left hippocampal atro-
patient with anti-Ma2 limbic encephalitis with T2 hyper- phy and persistent T2/FLAIR hyperintensities of the bilat-
intensities of the medial temporal lobe and hypothalamus eral medial temporal lobes. (i and j) Axial T2/FLAIR and
as well as enhancement of the hypothalamus. (e) Acute- sagittal brain T1 MRI images of a patient with anti-P/Q
phase T2/FLAIR MRI of a patient with anti-NMDAR calcium channel autoimmune cerebellar degeneration in
encephalitis with subtle T2 hyperintensities of the bilat- the convalescent phase, demonstrating marked cerebellar
eral medial temporal lobes. (f) Convalescent-phase T2/ atrophy
e f
g h
i j
Encephalitis tosis [43]. Thus, the typical workup includes

imaging, EEG, and CSF analysis. Early in the
Autoimmune encephalitis (AE) is a neurologi- disease course, MRI of the brain can be normal,
cal disorder that is characterized by subacute but PET scan may show hypermetabolism in the
development of short-term memory loss, confu- mesiotemporal regions (which is included in the
sion, hallucinations, mood changes, and/or sei- consensus clinical criteria for definite limbic
zures [43]. Limbic encephalitis (LE) is a subtype encephalitis), or areas of hypometabolism, such
of autoimmune encephalitis that is confined to as hypometabolism of the visual cortex in anti-
the limbic system including the hippocampus, NMDAR encephalitis [45]. MRI usually evolves
hypothalamus, and amygdala and is considered and shows fluid-attenuated inversion recovery
a classical PND. Autoimmune encephalitis is a (FLAIR) or T2 sequence hyperintensity in the
common non-prion cause of rapidly progressively medial temporal lobes (Fig. 13.2). EEG studies
dementia and can lead to irreversible dementia if may show irritability over the temporal lobes
it is not adequately treated [44]. with foci of epileptic activity, seizures, or focal
It is important to note that paraneoplastic or generalized slow activity. CSF analysis is typi-
encephalitis can present with numerous other cally pursued and often shows mild pleocytosis,
neurological and systemic findings such as elevated protein, elevated IgG level, and possibly
extensive encephalomyelitis, sleep disturbances, an onconeural autoantibody [43]. The CSF pleo-
hypothalamic- pituitary hormone deficits, and cytosis may only be present in the early stages
sensorimotor neuropathy. Proposed consensus of the disease and can resolve over the course of
clinical diagnostic criteria for possible autoim- weeks to months [9].
mune encephalitis incorporate subacute pro- Paraneoplastic encephalitis is often associated
gressive memory deficits, altered mental status, with onconeural antibodies, including antibod-
and one of the three other supporting findings ies that recognize intracellular neuronal antigens
among the following: T2 hyperintense lesions and cell surface antigens. Onconeural antibod-
on brain MRI possibly with associated enhance- ies that recognize intracellular antigens are more
ment, new-onset seizures, and/or CSF pleocy- closely associated with malignancies and include
anti-
Hu, anti-collapsin response-mediator pro- possible etiologies include infectious encephali-
tein-5 (anti-CRMP5), anti-amphiphysin, and tis such herpes simplex encephalitis, which can
anti-Ma2 [4, 46] (see Table 13.1). Encephalitis also trigger autoimmune anti-NMDAR encepha-
is a common PND in small-cell lung cancer, litis [48], as well as direct involvement of malig-
and approximately 50% of patients with SCLC nancy such as brain/leptomeningeal metastasis or
and limbic encephalitis have anti-Hu onconeu- low-grade glioma [4].
ral antibodies [47]. Anti-CRMP5 antibodies are
associated with testicular germ-cell tumors in
young men and with non-small cell lung cancer Subacute Sensory Neuronopathy
and breast cancer among older patients. These
types of limbic encephalitis can be more difficult Subacute sensory neuronopathy is characterized
to treat as the T-cell-driven response causes irre- by rapidly progressive asymmetric sensory defi-
versible neuronal damage [4]. cits that progress to include all sensory modali-
Paraneoplastic encephalitis can also be asso- ties leading to rapid impairment of ambulation
ciated with cell surface onconeural antibod- within 3 months. Initial symptoms include loss of
ies. Antibodies against cell surface onconeural vibration sense and joint position that is followed
antibodies include those directed against the by pain and temperature sensory loss, typically
VGKC complex proteins (such as anti-LGI1 more pronounced in the upper extremities than
and anti- CASPR2), anti-NMDAR, anti-alpha- lower extremities. In addition to loss of sensa-
amino-3-hydroxy-5-methyl-4-isoxazolepropionic tion, patients also experience severe burning pain
acid receptor (anti-AMPAR), and anti-gamma- and hyperesthesia. Clinically, patients frequently
butyric acid receptors (e.g., anti-GABABR) have loss of sensation in all sensory modalities,
(see Table 13.1). The most well-studied is anti- sensory ataxia, and absent reflexes but preserved
NMDAR encephalitis. Anti-NMDAR encephalitis strength. Diagnostic workup includes nerve con-
frequently starts with a viral-like prodrome fol- duction studies that demonstrate reduced/absent
lowed by prominent psychiatric symptoms includ- sensory nerve action potentials and CSF analysis
ing psychosis, catatonia, and agitation in addition with pleocytosis and elevated protein [49].
to memory loss, altered mental status, abnormal Autoantibodies associated with subacute sen-
movements, and seizures [28]. An autoantibody sory neuronopathy include anti-Hu, anti-CRMP5,
targeting the NR1 (GluN1) subunit of the NMDA and anti-amphiphysin. Pathologically, there is
receptor causes functional disruption by cross- destruction of sensory neuron cell bodies in the
linking and catalyzing internalization of receptors dorsal root ganglia with predominant CD8+ T-cell
[28, 29]. Anti-NMDAR encephalitis is associated infiltration [50, 51]. Malignancies that are typi-
with underlying tumor in 38% of patients, most cally associated include lung cancer (both small-
commonly ovarian teratoma [4, 33]. Other tumors cell lung cancer and bronchial carcinoma), breast
that can be associated include small-cell lung car- cancer, ovarian cancer, and Hodgkin’s lymphoma
cinoma, testicular teratoma, and sex cord-stromal [52]. The differential diagnosis for subacute sen-
tumors. Treatment of anti-NMDAR encephalitis is sory neuronopathy includes other disorders that
generally successful and involves removal/treat- cause primary degeneration of sensory neurons
ment of tumor if applicable and immunotherapy. in the dorsal root ganglia such as Sjögren’s syn-
Recovery is nearly complete in 75% of patients drome, HIV infection, cisplatin toxicity, and vita-
who receive timely treatment [33]. min B6 toxicity [51].
Although LE is a classical PND, approxi-
mately 70% of patients with limbic encephali-
tis do not have a malignancy diagnosed within Opsoclonus-Myoclonus Syndrome
5 years. The differential for patients with malig-
nancy presenting with a constellation of symp- Opsoclonus-myoclonus syndrome (OMS) is a
toms suggestive of encephalitis is broad. Other classical PND that is characterized by opsoclo-
nus, myoclonus, ataxia, and behavioral and sleep cancer (SCLC). LEMS is diagnosed based on
disturbances [4]. Clinically, opsoclonus is char- clinical signs and symptoms, EMG/NCS stud-
acterized by oscillations of the eyes with hori- ies, and autoantibody testing. Clinical findings
zontal, vertical, and torsional saccades. OMS is include progressive proximal muscle weak-
a clinical diagnosis and requires presence of at ness with autonomic dysfunction and areflexia.
least three of the four clinical findings: opsoclo- In contrast to myasthenia gravis, patients with
nus, myoclonus and/or ataxia, behavioral/sleep LEMS typically first note proximal leg weakness
disturbances, and presence of cancer or onconeu- that quickly progresses to involve the arms with
ral autoantibodies [53]. OMS is associated with later ocular and bulbar symptoms. Autonomic
malignancy in 39% of patients and idiopathic in dysfunction is common in LEMS, manifesting
61%. Onconeural antibodies are found in 11% of as dry mouth and erectile dysfunction. Other
patients, and humoral and cell-mediated immune autonomic findings such as gastrointestinal dys-
mechanisms are both crucial for the pathogenesis motility, cardiovascular dysfunction, and bladder
of OMS. Onconeural antibodies that are associ- dysfunction are typically due to coexistence of
ated with paraneoplastic OMS include anti-Ri, autoimmune dysautonomia, which is probably
anti-Hu, anti-Yo, anti-Ma1/Ma1, anti-NMDAR, paraneoplastic in origin. Patients have decreased
anti-amphiphysin, anti-CRMP-5/anti-CV2, and or absent deep tendon reflexes. Unlike myasthe-
anti-Zic2. In adults, paraneoplastic OMS is most nia gravis, strength and reflexes improve after
frequently associated with breast carcinoma, muscle contraction and exercise (characteris-
ovarian teratoma, and SCLC; in children, it is tic but not especially sensitive for diagnosis).
associated with neuroblastoma in 50% of patients Electrophysiological studies help aid diagno-
[53]. Management focuses on identification and sis and can distinguish between closely related
treatment of the underlying cancer as well as syndromes such as myasthenia gravis [54].
immunosuppressive therapies such as cortico- Autoantibodies recognizing P/Q-type VGCC
steroids, adrenocorticotropic hormone (ACTH), are detected in greater than 85% of patients with
intravenous immunoglobulin (IVIG), cyclophos- LEMS, and some patients also have antibodies
phamide, and rituximab [53]. OMS can also be to N-type and L-type VGCC [30]. Other diag-
triggered as part of a parainfectious or postin- nostic considerations include myasthenia gravis,
fectious autoimmune response to infections myopathies (such as inclusion body myositis),
like HIV, mycoplasma pneumonia, Salmonella and Guillain-Barre syndrome [33].
enterica, rotavirus, cytomegalovirus, human her- It is important to note that specific onconeural
pesvirus 6, and hepatitis C [4]. OMS may present antibodies can be associated with various clini-
with ataxia alone and delayed opsoclonus, and cal syndromes and cancers, and identification
thus patients may be misdiagnosed with subacute of >1 onconeural antibodies is associated with
cerebellar degeneration [53]. Lastly, drug toxic- increased risk of malignancy [38]. Management
ity from lithium, phenytoin, or amitriptyline may is focused on treatment of the malignancy and
present in a similar fashion and should be con- disease sequelae as well as immunosuppression.
sidered [4].
Treatment of Paraneoplastic
Lambert-Eaton Myasthenic Neurological Disorders
Syndrome
Treatment of PNDs first focuses on the evalu-
First described in 1953, Lambert Eaton myas- ation for occult malignancy as PNDs present
thenic syndrome (LEMS) is a rare classical PND, prior to diagnosis of cancer in approximately
estimated to affect approximately 0.48 persons 80% of patients. The specific autoantibody
per million. LEMS is associated with tumors detected can help guide the evaluation for occult
in 50–60% of cases, especially small-cell lung malignancy (see Table 13.1), and malignancy
is more likely to be found in patients with a Given the rarity of these syndromes, data from
cluster of autoantibodies [38]. In patients with randomized controlled trials are not available.
previously diagnosed malignancy, a PND may
herald cancer relapse [23]. The screening for
occult malignancy includes a careful physical Neurological Immune-Related
exam as well as various diagnostic studies such Adverse Events Following Immune
as computed tomography (CT) imaging of the Checkpoint Inhibitor Therapy
chest, abdomen, and pelvis. Gender-specific
cancer screening including mammography and Immune checkpoint inhibitors (ICIs) represent
pelvic ultrasound in women and testicular ultra- a major breakthrough in the treatment of several
sound and prostate- specific antigen testing in advanced cancers, and immune-related adverse
men represent important adjunctive testing [1, events associated with their use are important to
2]. Though considered a secondary or tertiary consider. Immune checkpoints are molecules that
screening modality depending on the cancer of play a crucial role in maintaining self-tolerance,
concern, FDG-PET imaging has a greater sensi- dampening excessive inflammation, and preventing
tivity for occult malignancy over CT if seroposi- autoimmunity (Fig. 13.3). By tipping the balance
tive for a paraneoplastic autoantibody [55]. The in favor of T-cell activation, ICSs improve tumor
development of a PND can precede the diagno- antigen presentation, amplify immune responses,
sis of malignancy by several years (presumably and disrupt tolerance but can also cause autoimmu-
because the immune response is effective at con- nity involving any organ system [56, 57].
trolling the malignancy), so serial evaluation and ICIs can cause neurological immune-related
close follow-up are crucial. Once a malignancy adverse events (nirAEs) by triggering the devel-
is identified, treatment of the detected cancer opment of immune responses against neuronal
alone can have a dramatic effect on the PND antigens. NirAEs following immune checkpoint
and potentially lead to its stabilization. This is inhibitor (ICI) treatment for cancer are similar
thought to be due to reduction of the theoretical in many respects to PNDs [58]. NirAEs affect up
antigen source as well as potentially the immu- to 1.5% of patients, with serious events in 0.2–
nosuppressant effects of chemotherapy [1, 6]. 0.8% of patients causing significant morbidity
As PNDs are triggered by the generation of and mortality [39, 59, 60]. Neurological adverse
immune responses to onconeural antigens of events include both CNS complications such as
tumors leading to attack of neuronal cells, immu- encephalitis, aseptic meningitis, posterior revers-
nomodulatory therapy is an important component ible encephalopathy syndrome, and hypophysitis
of treatment. Intravenous methylprednisolone and peripheral nervous system complications like
is a typical first-line step for the treatment of polyneuropathy, transverse myelitis, Guillain-
many PNDs. In PNDs associated with intracel- Barre syndrome, myasthenia gravis, and myositis.
lular antigens, cytotoxic and lymphocyte-specific The same autoantibodies can be found in nirAEs
medications (e.g., cyclophosphamide and myco- as in PNDs; however, in some cases the profile
phenolate, respectively) are used with the goal of autoantibodies may be different for nirAEs.
of reducing the cell-mediated immune response. Treatment of nirAEs focuses on early recognition,
Therapies aimed at depleting autoantibodies, such interruption of ICI treatment, and immunosup-
as plasmapheresis, are ineffective. Cell membrane pression with high-dose steroids or other immu-
protein-associated PNDs respond to antibody- nosuppressive medications [59]. Interestingly, the
directed first-line therapies such as intravenous prognosis of nirAEs tends to be more favorable
immunoglobulin therapy and plasmapheresis. than the corresponding PNDs. Given the undis-
Other immunosuppressants such as rituximab, puted efficacy of ICIs and expanding indications
cyclophosphamide, mycophenolate, and azathio- for treatment, the number of patients exposed to
prine are used as second-line agents in the acute ICIs will continue to increase, and nirAEs will
phase as well as for chronic management [1, 6]. likely become more common.
Fig. 13.3 Immune
checkpoint inhibitors CTLA-4 a
leading to T-cell DC T cell
activation
(a) CTLA-4 signaling pMHC TCR
is a negative regulator of
T-cell activation and acts
Activation
at the initial stage of
T-cell activation in the
CD80/86 CTLA4
LN. ICIs targeting
CTLA-4 block
interaction of CD80/86
with CTLA-4, allowing
CD80/86 to bind with DC
the costimulatory
molecule CD28 and pMHC TCR
promote T-cell Activation
activation. (b) PD-1 tumor
signaling is a negative CD80/86 CD28 immunity
regulator of T-cell nirAE
CTLA4
activation and acts at
With ICI
later stages of the

immune response in LN
peripheral tissues
including in the tumor
microenvironment. ICIs PD-1 b
targeting either PD-1 or Tumor cell T cell
PD-L1 block the
interaction of PD-1 and pMHC TCR
PD-L1, thus promoting
T-cell activation
Activation
PD-L1 PD-1
Tumor cell T cell
pMHC TCR
Activation
tumor
immunity
PD-1
nirAE
With ICI
PD-L1
Tumor microenvironment
Summary toms with neurological immune-related adverse

events related to immune checkpoint inhibitor
PNDs are rare, although likely under-recognized, therapy for a variety of cancers. PND syndromes
complications of cancer. These disorders can involve multiple areas of the nervous system,
occur at any stage of cancer and treatment; PNDs and patients can present with a wide variety of
are often the first clinical manifestation of cancer symptoms and clinical signs. Investigations of
and may herald a relapse in patients with known PNDs continue to provide further insight into
prior cancer history. They also share many symp- the immune system’s response to cancer; how-
ever, much is still unknown regarding their exact 16. Venkatraman A, Opal P. Paraneoplastic cerebellar

degeneration with anti-Yo antibodies – a review. Ann
mechanisms and the best diagnostic and treat- Clin Transl Neurol. 2016;3(8):655–63.
ment strategies. 17. Shams’Ili S, Grefkens J, De Leeuw B, Van den Bent
M, Hooijkaas H, Van der Holt B, et al. Paraneoplastic
cerebellar degeneration associated with antineu-
ronal antibodies: analysis of 50 patients. Brain.
References 2003;126(6):1409–18.
18. Peterson K, Rosenblum MK, Kotanides H, Posner
1. Mckeon A. Paraneoplastic and other autoimmune dis- JB. Paraneoplastic cerebellar degeneration. I. A clini-
orders of the central nervous system. Neurohospitalist. cal analysis of 55 anti-Yo antibody-positive patients.
2013;3(2):53–64. Neurology. 1992;42(10):1931–7.
2. Titulaer MJ, Soffietti R, Dalmau J, Gilhus NE, 19. Graus F, Cordon-Cardo C, Posner JB. Neuronal anti-
Giometto B, Graus F, et al. Screening for tumours in nuclear antibody in sensory neuronopathy from lung
paraneoplastic syndromes: report of an EFNS Task cancer. Neurology. 1985;35(4):538–43.
Force. Eur J Neurol. 2011;18(1):19–e3. 20. McKeon A, Pittock SJ. Paraneoplastic encepha-

3. Iorio R, Lennon VA. Neural antigen-specific autoim- lomyelopathies: pathology and mechanisms. Acta
mune disorders. Immunol Rev. 2012;248(1):104–21. Neuropathol. 2011;122(4):381–400.
4. Dalmau J, Rosenfeld MR. Paraneoplastic syndromes 21. Albert ML, Darnell RB. Paraneoplastic neurological
of the CNS. Lancet Neurol. 2008;7(4):327–40. degenerations: keys to tumour immunity. Nat Rev
5. Pelosof LC, Gerber DE. Paraneoplastic syndromes: Cancer. 2004;4:36–44.
an approach to diagnosis and treatment. Mayo Clin 22. Kazarian M, Laird-Offringa IA. Small-cell lung

Proc. 2010;85(9):838–54. cancer-associated autoantibodies: potential applica-
6. Didelot A, Honnorat JÔ. Paraneoplastic disorders of tions to cancer diagnosis, early detection, and therapy.
the central and peripheral nervous systems. Handb Mol Cancer. 2011;10:33.
Clin Neurol. 2014;121:1159–79. 23. Darnell RB, Roberts WK. Neuroimmunology of the
7. Wells EM, Dalmau J. Paraneoplastic neurologic paraneoplastic neurological degenerations. Curr Opin
disorders in children. Curr Neurol Neurosci Rep. Immunol. 2004;16(5):616–22.
2011;11(2):187–94. 24. Albert ML, Darnell JC, Bender A, Francisco LM,
8. Gozzard P, Woodhall M, Chapman C, Nibber A, Bhardwaj N, Darnell RB. Tumor-specific killer cells
Waters P, Vincent A, et al. Paraneoplastic neurologic in paraneoplastic cerebellar degeneration. Nat Med.
disorders in small cell lung carcinoma. Neurology. 1998;4(11):1321–4.
2015;85(3):235–9. 25. Albert ML, Austin LM, Darnell RB. Detection and
9. Darnell RB, Posner JB. Paraneoplastic syndromes treatment of activated T cells in the cerebrospinal
involving the nervous system. N Engl J Med [Internet]. fluid of patients with paraneoplastic cerebellar degen-
2003;349(16):1543–54. eration. Ann Neurol. 2000;47(1):9–17.
10. Graus F, Dalmau J, Reñé R, Tora M, Malats N,
26. Iorio R, Smitt PS. Paraneoplastic cerebellar degenera-
Verschuuren JJ, et al. Anti-Hu antibodies in patients tion. In: Essentials of cerebellum and cerebellar disor-
with small-cell lung cancer: association with com- ders: a primer for graduate students. Cham: Springer;
plete response to therapy and improved survival. J 2016. p. 587–93.
Clin Oncol. 1997;15(8):2866–72. 27. Graus F, Illa I, Agusti M, Ribalta T, Cruz-Sanchez
11. Linnoila J, Pittock SJ. Autoantibody-associated cen- F, Juarez C. Effect of intraventricular injection of an
tral nervous system neurologic disorders. Semin anti-Purkinje cell antibody (anti-Yo) in a Guinea pig
Neurol. 2016;36(4):382–96. model. J Neurol Sci. 1991;106(1):82–7.
12. Dalmau J, Furneaux HM, Cordon-Cardo C, Posner 28. Dalmau J, Gleichman AJ, Hughes EG, Rossi JE, Peng
JB. The expression of the Hu (paraneoplastic X, Lai M, et al. Anti-NMDA-receptor encephalitis:
encephalomyelitis/sensory neuronopathy) antigen case series and analysis of the effects of antibodies.
in human normal and tumor tissues. Am J Pathol. Lancet Neurol. 2008;7(12):1091–8.
1992;141(4):881–6. 29. Hughes EG, Peng X, Gleichman AJ, Lai M, Zhou
13. Darnell JC, Albert ML, Darnell RB. cdr2, A target L, Tsou R, et al. Cellular and synaptic mechanisms
antigen of naturally occurring human tumor immu- of anti-NMDA receptor encephalitis. J Neurosci.
nity, is widely expressed in gynecological tumors. 2010;30(17):5866–75.
Cancer Res. 2000;60(8):2136–9. 30. Anda V, Ennon AL, Homas T, Ryzer JK, Uy G,

14. Graus F, Delattre JY, Antoine JC, Dalmau J, Giometto Riesmann EG, et al. Calcium-channel antibodies
B, Grisold W, et al. Recommended diagnostic criteria in the Lambert-Eaton syndrome and other para-
for paraneoplastic neurological syndromes. J Neurol neoplastic syndromes. N Engl J Med. 1995;30(17):
Neurosurg Psychiatry. 2004;75(8):1135–40. 5866–75.
15. Pittock SJ, Kryzer TJ, Lennon VA. Paraneoplastic 31. Diamond B, Huerta PT, Mina-Osorio P, Kowal C,
antibodies coexist and predict cancer, not neurologi- Volpe BT. Losing your nerves? Maybe it’s the anti-
cal syndrome. Ann Neurol. 2004;56(5):715–9. bodies. Nat Rev Immunol. 2009;9:449–56.
32. Meriney SD, Tarr TB, Ojala KS, Wu M, Li Y, Lacomis 46. Lancaster E, Dalmau J. Neuronal autoantigens-

D, et al. Lambert–Eaton myasthenic syndrome: mouse pathogenesis, associated disorders and antibody test-
passive-transfer model illuminates disease pathology ing. Nat Rev Neurol. 2012;8(7):380–90.
and facilitates testing therapeutic leads. Ann N Y 47. Alamowitch S, Graus F, Uchuya M, Reñé R, Bescansa
Acad Sci. 2018;1421(1):73–81. E, Delattre JY. Limbic encephalitis and small cell
33. Titulaer MJ, McCracken L, Gabilondo I, Armangué lung cancer. Clinical and immunological features.
T, Glaser C, Iizuka T, et al. Treatment and prognostic Brain. 1997;120(6):923–8.
factors for long-term outcome in patients with anti- 48. Armangue T, Spatola M, Vlagea A, Mattozzi S,

NMDA receptor encephalitis: an observational cohort Cárceles-Cordon M, Martinez-Heras E, et al.
study. Lancet Neurol. 2013;12(2):157–65. Frequency, symptoms, risk factors, and outcomes of
34. Malviya M, Barman S, Golombeck KS, Planagumà autoimmune encephalitis after herpes simplex enceph-
J, Mannara F, Strutz-Seebohm N, et al. NMDAR alitis: a prospective observational study and retrospec-
encephalitis: passive transfer from man to mouse tive analysis. Lancet Neurol. 2018;17(9):760–72.
by a recombinant antibody. Ann Clin Transl Neurol. 49. Chalk CH, Windebank AJ, Kimmel DW, Mcmanis
2017;4(11):768–83. PG. The distinctive clinical features of paraneoplastic
35. Kim TJ, Lee ST, Moon J, Sunwoo JS, Byun JI, Lim JA, sensory neuronopathy. Can J Neurol Sci/J Can des Sci
et al. Anti-LGI1 encephalitis is associated with unique Neurol. 1992;19(3):346–51.
HLA subtypes. Ann Neurol. 2017;81(2):183–92. 50. Gazic B, Pisem J, Dolenc-Groselj L, Popovic

36. Binks S, Varley J, Lee W, Makuch M, Elliott K,
M. Paraneoplastic encephalomyelitis/sensory motor
Gelfand JM, et al. Distinct HLA associations of peripheral neuropathy – an autopsy case study. Folia
LGI1 and CASPR2-antibody diseases. Brain. Neuropathol. 2005;43(2):113–7.
2018;141(8):2263–71. 51. Camdessanché JP, Jousserand G, Ferraud K, Vial C,
37. Klein CJ, Lennon VA, Aston PA, McKeon A, O’Toole Petiot P, Honnorat J, et al. The pattern and diagnostic
O, Quek A, et al. Insights from LGI1 and CASPR2 criteria of sensory neuronopathy: a case-control study.
potassium channel complex autoantibody subtyping. Brain. 2009;132(7):1723–33.
JAMA Neurol. 2013;70(2):229–34. 52. Tarin D. Update on clinical and mechanistic aspects
38. Horta ES, Lennon VA, Lachance DH, Jenkins SM, of paraneoplastic syndromes. Cancer Metastasis Rev.
Smith CY, McKeon A, et al. Neural autoantibody 2013;32:707–21.
clusters aid diagnosis of cancer. Clin Cancer Res. 53. Oh S-Y, Kim J-S, Dieterich M. Update on opsoc-
2014;20(14):3862–9. lonus–myoclonus syndrome in adults. J Neurol
39. Bernal F, Shams’Ili S, Rojas I, Sanchez-Valle R, Saiz [Internet]. 2018;266(6):1541–8. [Cited 2018 Dec
A, Dalmau J, et al. Anti-Tr antibodies as markers of 5]; Available from: http://www.ncbi.nlm.nih.gov/
paraneoplastic cerebellar degeneration and Hodgkin’s pubmed/30483882
disease. Neurology. 2003;60(2):230–4. 54. Titulaer MJ, Lang B, Verschuuren JJGM. Lambert-
40. Sabater L, Höftberger R, Boronat A, Saiz A. Antibody Eaton myasthenic syndrome: from clinical charac-
repertoire in paraneoplastic cerebellar degen- teristics to therapeutic strategies. Lancet Neurol.
eration and small cell lung cancer. PLoS One. 2011;10(12):1098–107.
2013;8(3):e60438. 55. McKeon A, Apiwattanakul M, Lachance DH, Lennon
41. Verschnuren J, Chuang L, Rosenblum MK, Lieberman VA, Mandrekar JN, Boeve BF, et al. Positron emission
F, Pryor A, Posner JB, et al. Inflammatory infiltrates tomography-computed tomography in paraneoplastic
and complete absence of Purkinje cells in anti-Yo- neurologic disorders: systematic analysis and review.
associated paraneoplastic cerebellar degeneration. Arch Neurol. 2010;67(3):322–9.
Acta Neuropathol. 1996;91(5):519–25. 56. Hassel JC. Ipilimumab plus nivolumab for advanced
42. Jones AL, Flanagan EP, Pittock SJ, Mandrekar
melanoma. Lancet Oncol. 2016;17(11):1471–2.
JN, Eggers SD, Ahlskog JE, et al. Responses to 57. Jain P, Jain C, Velcheti V. Role of immune-checkpoint
and outcomes of treatment of autoimmune cer- inhibitors in lung cancer. Ther Adv Respir Dis.
ebellar ataxia in adults. JAMA Neurol. 2015;72(11): 2018;12:1–13.
1304–12. 58. Cuzzubbo S, Javeri F, Tissier M, Roumi A, Barlog C,
43. Graus F, Titulaer MJ, Balu R, Benseler S, Bien
Doridam J, et al. Neurological adverse events associ-
CG, Cellucci T, et al. A clinical approach to diag- ated with immune checkpoint inhibitors: review of the
nosis of autoimmune encephalitis. Lancet Neurol. literature. Eur J Cancer. 2017;73:1–8.
2016;15(4):391–404. 59.
Hottinger AF. Neurologic complications of
44. Paterson RW, Takada LT, Geschwind MD. Diagnosis immune checkpoint inhibitors. Curr Opin Neurol.
and treatment of rapidly progressive dementias. 2016;29(6):806–12.
Neurol Clin Pract. 2012;2(3):187–200. 60. Astaras C, de Micheli R, Moura B, Hundsberger T,
45. Probasco JC, Solnes L, Nalluri A, Cohen J, Jones KM, Hottinger AF. Neurological adverse events associ-
Zan E, et al. Decreased occipital lobe metabolism by ated with immune checkpoint inhibitors: diagno-
FDG-PET/CT. Neurol Neuroimmunol Neuroinflamm. sis and management. Curr Neurol Neurosci Rep.
2018;5(1):e413. 2018;18(1):1–9.
Systemic Therapy of Brain
Metastases: Lung Cancer
14
Adam Lauko, Vyshak Alva Venur,
and Manmeet S. Ahluwalia
Introduction lar underpinnings of NSCLC, there is a larger

role for targeted agents and immunotherapy in
Lung cancer is the most common cancer that the treatment of NSCLC brain metastases.
results in the metastatic disease to the brain [1]. Further, prognostication has evolved and the new
Small cell cancers are characterized by neuroen- Lung-molGPA (molecular graded prognostic
docrine feature and account for 15% of lung can- assessment) includes molecular markers associ-
cers seen [1–3]. Brain metastases from both ated with NSCLC. The latest GPA incorporates
non-small cell lung cancer (NSCLC) and small molecular alteration data for patients with
lung cancer are treated differently; small cell NSCLC and brain metastases; the clinical vari-
lung cancer is often treated with whole-brain ables in the new model include patient age, per-
radiation (WBRT) [4]. Both cancer types have a formance status, extracranial metastases, number
propensity for brain involvement—10–30% of of brain metastases, and, in patients with adeno-
NSCLC patients develop brain metastases during carcinoma, the presence of EGFR or ALK altera-
their course [5]. Ten to twenty-five percent of tions [6].
patients with stage IV NSCLC have brain
involvement at presentation. Until recently,
NSCLC brain metastases were managed with Targeted Therapy
surgery and radiation therapy such as WBRT and
increasingly stereotactic radiosurgery (SRS). The identification of targetable gene alterations
With our improved understanding of the molecu- has transformed the management of lung cancer,
allowing for personalized therapies with excel-
lent response rates, including in the setting of
brain metastasis. In the multicenter Lung Cancer
Mutation Consortium, targetable oncogenic driv-
ers have been identified in 64% of patients with
A. Lauko · M. S. Ahluwalia (*) non-small cell lung cancer (NSCLC) adenocarci-
Burkhardt Brain Tumor and Neuro-Oncology Center,
Cleveland Clinic, Cleveland, OH, USA noma [7]. Within squamous cell carcinoma
e-mail: ahluwam@ccf.org (NSCLC), up to 80% of tumors has a known
V. A. Venur oncogenic driver mutation; however, unlike in
Department of Medical Oncology, University of adenocarcinoma, targeted therapies to these
Washington Seattle Cancer Care Alliance, Fred oncogenic drivers have not reached the same
Hutchinson Cancer Research Center, level of clinical utility [8, 9]. The latest NSCLC
Seattle, WA, USA

https://doi.org/10.1007/978-3-030-42958-4_14
208 A. Lauko et al.
guidelines published by National Comprehensive to the first-generation EGFR tyrosine kinase

Cancer Network (NCCN), 2019 version 4, rec- inhibitors (TKIs) gefitinib and erlotinib [18–20].
ommend that the nine genes should be tested, The majority of these genetic aberrations are
including EGFR, KRAS, HER2, ALK, ROS1, either exon 19 deletions (60%) or L858R mis-
MET, BRAF, RET, and NTRK [10]. In patients sense substitutions (35%), which result in consti-
with small cell lung cancer, where approximately tutive activation of the receptor without ligand
15–20% of patients present with brain metasta- binding [21]. EGFR mutations are more likely to
ses, a different set of genetic alterations is be found in nonsmokers, and the incidence of
observed [11]. These include insulin-like growth EGFR mutations varies with ethnicity, with up to
factor-I receptor (IGF-IR), c-Kit overexpression, 50% frequency in adenocarcinomas in the Asian
mutations in vascular endothelial growth factor population compared to only 10–15% in
receptor (VEGFR), epidermal growth factor Caucasians [20]. The IPASS trial was the first to
receptor (EGFR) mutation and PTEN, as well as confirm clinical efficacy of the targeted EGFR
Myc overexpression [12]. However, none of the inhibitor gefitinib in a selected patient population
currently available targeted therapeutics agents with EGFR mutations [22].
have shown efficacy in small cell lung cancer. Two retrospective studies published in the late
The remainder of this section will focus primar- 2000s provided the first evidence for the potential
ily on agents that target EGFR and ALK muta- efficacy of EGFR inhibitors in NSCLC patients
tions, as they are the most clinically relevant and with brain metastases (NSCLCBM) harboring
have the most data regarding intracranial efficacy. EGFR mutations [23, 24]. Both articles found
Although still limited, preliminary data suggest that patients with NSCLC who received EGFR
promising results in targeting ROS1, MET, TKIs at any time after the diagnosis of BM sur-
BRAF, RET, and NTRK alterations as well. vived longer than those that did not. In addition,
other retrospective studies published around the
same time found intracranial response rates rang-
pidermal Growth Factor Receptor
E ing from 42% to 82% [25–27].
(EGFR) Failure to penetrate the blood-brain barrier
(BBB) is often cited as one of the primary rea-
EGFR belongs to a receptor tyrosine kinase fam- sons for the lack of intracranial efficacy of cyto-
ily that also includes human epidermal growth toxic chemotherapy in patients with
factor receptor 2 (HER2, also known as ERBB2), NSCLCBM. Deng et al. found that the BBB pen-
HER3 (ERBB3), and HER4 (ERBB4). The etration rate of erlotinib was 4.4% ± 3.2% and
receptor contains four extracellular domains, a that the concentration of erlotinib was higher in
transmembrane domain, a tyrosine kinase patients with a partial response compared to sta-
domain, and a carboxy tail. The binding of an ble or progressive disease [28]. Similarly, Zhao
activating ligand leads to EGFR dimerization and et al. found that gefitinib had a BBB permeation
transphosphorylation, triggering a vast array of rate of 1.3% ± 0.7% and that the presence of
signaling pathways leading to cell growth, prolif- brain metastases increased the BBB penetration
eration, survival, invasion, and angiogenesis [13]. (1.46% vs. 0.95%; p-value, 0.042) [29]. Pulsatile
EGFR is overexpressed in 15–50% of NSCLC dosing is a strategy that has been proposed for
and was viewed from the start as a promising increasing CSF concentration of TKIs. In one
therapeutic target; however, early trials with study testing this approach, an increased dose of
EGFR inhibitors in unselected populations dem- 1500 mg of erlotinib allowed a CSF concentra-
onstrated minimal clinical efficacy and no tion that had been shown to inhibit the growth of
improvement over cytotoxic chemotherapies cell lines harboring EGFR mutations in vitro [30,
[14–17]. Three subsequent papers published in 31]. In a phase 1 study of twice-weekly pulse
2004 demonstrated that activating mutations in dose and daily low-dose erlotinib in patients with
the EGFR gene were associated with sensitivity NSCLC, the authors noted stable disease in the
14 Systemic Therapy of Brain Metastases: Lung Cancer 209
11 patients with brain metastases. In addition, of there was no overall survival benefit of TKI vs.
the 19 patients on trial that had systemic progres- chemotherapy (HR 0.98, 95% CI 0.87–1.10);
sion, none of these patients had any radiographic however, there was a vastly improved PFS in
evidence of new or progressing brain metastases, patients taking TKIs (HR 0.37, 95% CI
suggesting that pulsatile dosing may improve 0.29–0.49).
intracranial efficacy of erlotinib [32]. Unfortunately, the response duration to first-
Some of the first prospective data on the intra- generation EGFR TKIs is often limited due to a
cranial efficacy of first-generation EGFR TKIs second EGFR mutation on exon 20, with a
originated from a phase 2 study in China that threonine-methionine substitution on codon 790
treated NSCLC patients with asymptomatic brain (T790M) [38, 39]. Other documented mecha-
metastases with erlotinib. The authors found that nisms of resistance include HER2 amplification;
patients with known EGFR mutations had mutations to MET, BRAF, and PIK3CA; or trans-
increased survival compared to wild-type patients formation to small cell lung cancer [40]. The
(18.4 months vs. 37.5 months, p = 0.02) [33]. average patient will develop resistance within
Welsh et al. treated patients concurrently with 12–16 months of starting treatment with a first-
erlotinib and WBRT and found that median sur- generation TKI [41].
vival time was greater in patients with EGFR First-generation inhibitors are reversible com-
mutations compared to wild-type EGFR [34]. In petitive ATP inhibitors that target only EGFR,
a prospective phase 2 clinical trial in 41 Japanese while second-generation inhibitors such as afa-
patients with EGFR-mutant NSCLCBM, gefitinib, neratinib, and dacomitinib are irreversible
tinib demonstrated a response rate of 87.8% with inhibitors that also target HER2 and HER4.
a PFS of 14.5 months (95% CI, 10.2–18.3 months) LUX-LUNG 3 and 6, both randomized phase 3
and OS of 21.9 months (95% CI, 18.5– studies of afatinib, enrolled patients with asymp-
30.3 months). Interestingly, this study found that tomatic brain metastases [42, 43]. Although a
exon 19 deletion was associated with better PFS subgroup analysis of the two trials showed
and OS, when compared with L858R mutations increased PFS compared to conventional chemo-
[35]. Finally, a phase 2 trial of icotinib, a first- therapy (8.2 vs. 5.4 months; HR, 0.50;
generation TKI approved in China for EGFR- p = 0.0297), the field quickly progressed to using
mutated NSCLC, demonstrated a significantly third-generation EGFR TKIs.
increased OS in patients with EGFR mutations Osimertinib is a third-generation TKI that is
compared to wild-type EGFR (median OS effective against T790M mutations. The drug
22 months vs. 7.5 months, respectively, binds covalently to the cysteine on codon 797,
p = 0.0001) when given concurrently with WBRT overcoming the enhanced ATP affinity from
[36]. The best evidence for intracranial efficacy T790M mutations [13]. Osimertinib is effective
of first-generation TKIs came from a phase 3 both as second-line treatment in patients who
study comparing icotinib alone vs. WBRT with progressed after treatment with first-generation
or without chemotherapy. While there was a sig- EGFR TKIs [44] and a first-line agent in EGFR-
nificantly improved intracranial PFS in the ico- mutant disease [45]. In a trial comparing osimer-
tinib monotherapy group (HR 0.44, p < 0.0001), tinib to first-generation EGFR TKIs for first-line
there was no difference in the overall survival treatment, only 6% of patients had CNS progres-
between the two groups. However, a higher num- sion in the osimertinib group compared to 15%
ber of patients in the WBRT group crossed over in the standard EGFR TKI group. In the same
to icotinib, making OS difficult to interpret [37]. trial, patients treated with osimertinib had sig-
Crossing over after progression has made it dif- nificantly improved intracranial PFS (HR 0.48;
ficult to identify any statistically significant 95% CI 0.26–0.86). In a subgroup analysis com-
improvement in OS with TKIs vs. platinum- paring osimertinib to pemetrexed plus carbopla-
based chemotherapy. In a meta-analysis of almost tin or cisplatin, the median PFS was longer in
3000 patients from eight phase 3 clinical trials, those receiving osimertinib, among the 144
210 A. Lauko et al.
patients with brain metastases (8.5 months vs. crizotinib demonstrated superior intracranial
4.2 months; HR 0.32; 95% CI 0.21–0.49) [46]. activity compared to standard platinum-based
Together, this data consistently supports better chemotherapy [54–56]. Of the 79 patients with
intracranial activity with osimertinib compared stable brain metastases enrolled, those treated
to first-
generation EGFR TKIs and cytotoxic with crizotinib had significantly better intracra-
chemotherapies. nial disease control at 12 and 24 weeks (12 weeks:
85% vs. 45%, p < 0.001; 24 weeks: 56% vs. 25%,
p = 0.006). In a pooled analysis of PROFILE
Anaplastic Lymphoma Kinase (ALK) 1005 and 1007, 275 patients with asymptomatic
brain metastases were analyzed, with a 56%
The ALK gene was first discovered in 1994 in a intracranial disease control rate at 12 weeks and a
subset of non-Hodgkin lymphoma where ALK median intracranial PFS of 7 months. However,
was fused to nucleophosmin (NPM) as a result of progression of preexisting or development of
a chromosomal translocation [47]. ALK encodes new intracranial lesions while receiving crizo-
a transmembrane receptor tyrosine kinase that tinib was a common manifestation of acquired
belongs to the insulin receptor superfamily; it resistance.
comprises an extracellular domain, a single-pass This led to the development of second-
transmembrane domain, and an intracellular generation ALK inhibitors—ceritinib, alectinib,
kinase domain. In the normal physiology of and brigatinib [57–59]. In a phase 1 clinical trial
human cells, ALK’s function is unclear [48]. of ceritinib, of 14 patients with measurable intra-
ALK rearrangements in the context of NSCLC cranial lesions at baseline, 7 achieved an intracra-
were first identified in 2007, where it was found nial response and 3 had stable disease [60]. In
that a small inversion within chromosome 2p ASCEND-2, a phase 2 study of ceritinib, of 100
resulted in a fusion gene with echinoderm patients with baseline brain metastases, there was
microtubule-associated protein-like 4 (EMLA4) a 45% intracranial response rate (95% CI, 23.1%
gene (EML4-ALK) [49]. ALK translocations are to 68.5%) and 80% intracranial control rate.
found in approximately 3–7% of patients with Alectinib has also shown promising intracranial
NSCLC and cluster more commonly in non- activity. In the phase 1/2 study AF-002JG,
smoker and light smokers [50–52]. Patients with patients with crizotinib-resistant ALK-rearranged
ALK rearrangements treated with platinum-based NSCLC were given alectinib. Of 21 patients with
chemotherapy had no overall survival difference CNS metastases at baseline, 11 had an objective
compared to wild-type patients; however, out- response (6 complete, 5 partial) [61]. In the phase
comes of NSCLC patients with EML4-ALK posi- 3 J-ALEX trial, alectinib was compared to crizo-
tivity rapidly improved with the development of tinib in ALK inhibitor-naïve patients; alectinib
ALK inhibitors [51]. was associated with significantly prolonged PFS
As discussed above, many of the initial clini- (HR = 0.08, 95% CI, 0.01–0.61) [62]. In another
cal trials with first-generation EGFR inhibitors multicenter phase 3 trial (ALEX) comparing
were agnostic to a patient’s EGFR mutation sta- alectinib and crizotinib, only 12% of patients in
tus, slowing the implementation of these inhibi- the alectinib group had CNS progression com-
tors in clinical practice. However, the lessons pared to 45% with crizotinib (HR = 0.16; 95%
learned from these trials allowed for a more logi- CI, 0.10–0.28; p < 0.001) [63]. In addition, com-
cal approach to ALK inhibitors, with many stud- plete CNS response was more likely in the alec-
ies including prospective tumor genotyping. The tinib group (45% vs. 9%, p < 0.001).
first drug developed in this class was crizotinib, The third-generation inhibitor lorlatinib is the
an oral small molecule inhibitor of ALK, MET, latest ALK inhibitor to demonstrate intracranial
and ROS tyrosine kinases [53]. In a randomized efficacy. Lorlatinib is a potent, brain penetrant
phase 3 clinical trial of ALK-positive advanced inhibitor of ALK and ROS1 and has broad ALK
NSCLC patients (PROFILE 1014), first-line mutational coverage. In a recently published
phase 2 clinical trial of patients treated with at interaction between the T-cell and cancer cell is
least one prior ALK inhibitor, 51 of 81 of patients complex and involves MHC class 1 and T-cell
had an intracranial response (63%; 95% CI 51.5– receptors as well as additional co-stimulatory and
73.4) [64]. The results of this trial ultimately led co-inhibitory interactions. CTLA-4 (cytotoxic
to accelerated approval of lorlatinib for patients T-lymphocyte-associated protein 4) and PD-1
that have progressed on another ALK inhibitor. (Programmed death-1) are two such co-inhibitory
signals. Ipilimumab is an anti-CTLA-4 antibody
which helps block the CTLA-4 co-inhibitory sig-
risten Rat Sarcoma 2 Viral Oncogene
K nals in the lymph nodes. There are several anti-
Homolog (KRAS) PD-1 and PD-L1 antibodies including
pembrolizumab and nivolumab, which block the
KRAS belongs to a family of GTPases that trans- PD-1/PD-L1 interaction in the tumor microenvi-
duce growth signals from multiple tyrosine ronment. A phase 3 clinical trial of pembroli-
kinases [8]. Activating KRAS mutations contrib- zumab has shown improved PFS in metastatic
ute to constitutive signaling and are present in non-small cell lung cancer, with superior out-
about 30% of NSCLC adenocarcinoma; they are comes in patients with >50% tumor PD-L1
more commonly found in smokers [65, 66]. The expression [71]. Nivolumab has been shown to
presence of a KRAS mutation has been associ- be efficacious in cisplatin-resistant NSCLC [72].
ated with worse prognosis compared to wild-type Atezolizumab, another PD-L1 antibody, has been
tumors [67]. Despite being one of the most com- shown to add survival benefit to platinum-based
mon mutations and one of the first identified, chemotherapy in small cell lung cancer [73].
effective targeting of KRAS mutations has been Several other studies have shown promising
therapeutically challenging. Direct KRAS inhibi- activity of immune checkpoint inhibitors in lung
tion with salirasib was unsuccessful with no cancer.
patients having any radiographic response [68]. Brain metastases from lung cancer express
Selumetinib, a MEK inhibitor, potentially inhib- PD-L1, but the extent of this expression can be
its and targets downstream of KRAS but demon- varied. A small study of 32 patients showed
strated no additional efficacy in a phase 3 trial PD-L1 expression in 22% of brain metastases,
when added to docetaxel [69]. Our group has and it was a predictor of poor overall survival
unpublished data suggesting the use of immuno- [74]. In another study of brain metastases from
therapy negates the poor outcomes traditionally small cell lung cancer, up to 75% were noted to
seen in patients with NSCLC brain metastasis have PD-L1 expression [75]. A larger study of 73
and KRAS mutations. Patients with KRAS muta- lung cancer patients with paired samples of pri-
tions treated with immunotherapy had improved mary and brain metastases evaluated the tumor
overall survival from the diagnosis of brain and tumor microenvironment PD-L1 expression
metastases compared to chemotherapy (unpub- [76]. Tumor microenvironment PD-L1 and tumor
lished Lauko and Ahluwalia). Although a similar cell PD-L1 expression were discordant between
result was seen in a recent meta-analysis, intra- primary tumor and brain metastases in 14%
cranial disease was not analyzed [70]. (10/73) and 26% (19/73) cases, respectively. This
suggested significant differences between brain
metastases and the primary tumor. The density of
Immunotherapy in Brain T-cell infiltration may also vary considerably in
Metastases from Lung Cancer brain metastases. The T-cells are usually a mix-
ture of the exhausted and activated subtypes. In a
The advent of therapeutic strategies to aid the study of 116 brain metastases (approximately
immune system in recognizing cancer cells as 50% from NSCLC), 99% of the tumor microen-
foreign and mounting a response against them vironment had T-cell infiltration [77]. There was
has been a major milestone in oncology. The no co-relation between T-cell infiltration and
212 A. Lauko et al.
PD-L1 expression or corticosteroid use. Dense programs (EAPs) for patients with mostly asymp-
T-cell infiltration was noted in more than 50% of tomatic, stable brain metastases. One such large
all brain metastases samples, including effector EAP was an Italian series of 409 patients with
CD3+ and CD8+ cells and memory cells. The asymptomatic or previously treated non-
density of T-cell infiltration had a positive impact squamous lung brain metastases who were
on overall survival. Additionally, microglia and treated with nivolumab [82]. The intracranial dis-
macrophages represent a unique and significant ease response rate was 40% with median overall
part of the tumor microenvironment in brain survival of 8.1 months. In a French EAP of 409
metastases [75, 78]. Tumor mutational burden is non-small cell lung cancer patients receiving
another important predictor of response to immu- nivolumab, 130 had asymptomatic or stable
notherapy. A small study of 20 patients with lung treated brain metastases [83]. The intracranial
cancer brain metastases showed an increase in partial response rate was 12% and the median
tumor mutational burden compared to the corre- overall survival was 6.6 months.
sponding primary site; however, the T-cell clones Most major clinical trials with immunother-
were less rich, suggesting a possible role of apy excluded patients with active and progressive
immune checkpoint inhibitors in activating the brain metastases. However, several allowed for
immune system [76]. In summary, the character- inclusion of patients with stable and treated
istics and tumor microenvironment of lung can- lesions. For example, the KEYNOTE 024 study
cer make it a good target for immune checkpoint was a phase 3 clinical trial of pembrolizumab
inhibitors. compared to standard of care chemotherapy in
The biggest challenge in drug delivery to brain newly diagnosed NSCLC with PD-L1 expression
metastases is the blood-brain barrier. In fact, the of at least 50% of the tumor [71]. This study
brain was historically thought to be an immune showed significant improvement in progression-
privileged site. Recent window of opportunity free survival in the entire cohort. The subgroup of
(“phase 0”) studies have evaluated changes in patients with brain metastases was small—18
T-cell activation and tumor microenvironment in patients in the pembrolizumab arm and 10 in the
high-grade glioma patients after administration standard chemotherapy arm; the survival was not
of one dose of nivolumab or pembrolizumab [79, significantly different between the groups.
80]. The surgical specimen obtained after expo- KEYNOTE 189, a clinical trial of the combina-
sure to either anti-PD1 antibodies showed upreg- tion of pembrolizumab and chemotherapy in
ulation of T-cell and interferon-Y gene expression, newly diagnosed NSCLC, enrolled the largest
focal induction of PD-1 expression in the tumor number of patients with brain metastases [84,
microenvironment, and enhanced clonal T-cell 85]. The pembrolizumab plus chemotherapy
expansion. This gives the best evidence of intra- combination arm had 73 patients with stable
cranial activity of anti-PD1 antibodies. The use brain metastases, while the chemotherapy alone
of systemic high-dose corticosteroids in patients arm had 35 patients with stable brain metastases.
with symptomatic brain metastases is another With a hazard ratio of 0.36, the subgroup analysis
potential hurdle for utilization of immunothera- noted improved overall survival with the combi-
peutic agents. This was demonstrated in a phase nation of pembrolizumab and chemotherapy.
2 clinical trial of ipilimumab in melanoma brain Another single center phase 2 trial evaluated
metastases, where patients receiving corticoste- the NSCLC brain metastasis response rates with
roids had a significantly lower intracranial dis- pembrolizumab [86] and enrolled patients with
ease control rate with ipilimumab (10% vs. 24%) asymptomatic but progressive or untreated, mea-
compared to asymptomatic patients not on corti- surable (5–20 mm) brain metastases. A recently
costeroids [81]. reported interim analysis described 39 patients
There are several retrospective series of treated with pembrolizumab (34 with
nivolumab or pembrolizumab reported in the lit- PD-L1 > 50% expression in the primary tumor
erature, often in the setting of expanded access and 5 without PD-L1 expression). The intracranial
response rate in the cohort with increased PD-L1 several ongoing clinical trials evaluating the
expression was 29.4% (10 of 34 patients), while optimal timing of radiation therapy and sys-
none of the patient in the PD-L1-negative group temic therapy for patients with brain metastases.
had an intracranial response. The median overall Targeted therapies are generally not combined
survival of the entire group was 8.9 months. The with radiation therapy due to the possibility of
PFS among patients with an intracranial response worse cutaneous toxicities.
or stable disease was 10.7 months, and 31% were
alive at 2 years, suggesting a durable response.
More studies like this are needed to better under- Conclusion
stand the intracranial activity of immunotherapy
in brain metastases from lung cancer. The management paradigm for brain metastases
from lung cancer is rapidly evolving. The treat-
ment plan must be customized and take into
Combination of Radiation Therapy account each patient’s overall prognosis, extra-
and Immunotherapy cranial disease status, available systemic therapy
options, neurologic symptoms, and brain metas-
Radiation therapy has long been the backbone tasis burden. Surgery is an option for patients
of management of brain metastases. Stereotactic with solitary, large, and symptomatic brain
radiosurgery (SRS) has replaced whole-brain metastases. Whole-brain radiation therapy is con-
radiation therapy (WBRT) in a majority of sidered for patients with numerous symptomatic
patients, although there is still a role for the lat- brain metastases and limited systemic options.
ter in patients with numerous symptomatic SRS is utilized in patients with few brain metas-
metastases. As initial studies with targeted and tases who are expected to live longer, in hopes of
immunotherapy agents have shown intracranial avoiding long-term cognitive decline from
efficacy, the clear next step is to consider com- whole-brain radiation therapy. Novel systemic
binations with radiation therapy. There are sev- therapies, mainly targeted agents and immune
eral theories suggesting synergy between checkpoint inhibitors, have shown promising
radiation therapy and immunotherapy, such as intracranial activity in early studies. They are
the abscopal effect and release of neo-antigens generally used in patients with predominantly
with radiation. Numerous retrospective studies extracranial disease and asymptomatic brain
combining immunotherapy and radiation have metastases. The combination and timing of
been published in patients with brain metastases immunotherapy with radiation therapy requires
from melanoma and NSCLC. One retrospective, further investigation. Innovative research contin-
single center study compared the efficacy of ues to identify pathways that drive metastatic
SRS after traditional chemotherapy or immuno- growth with hopes of developing novel, effective
therapy in patients with NSCLC brain metasta- agents to target them.
ses. No differences in overall survival,
progression-free survival, or response rates
were between the groups (46 patients in the che- References
motherapy group and 39 in the immunotherapy
group) [87]. A larger retrospective study of 260 1. Noone AM, Howlader N, Krapcho M, Miller D, Brest
A, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR,
patients, including 157 with NSCLC, evaluated Chen HS, Feuer EJ, Cronin KA, editors. SEER can-
the role of radiation therapy given concurrently cer statistics review, 1975–2015. Bethesda: National
or within 2 weeks of initiating immunotherapy; Cancer Institute; 2018.
this combination was compared to SRS mono- 2. Kohler BA, Ward E, McCarthy BJ, et al. Annual report
to the nation on the status of cancer, 1975–2007, fea-
therapy and SRS with stereotactic body radia- turing tumors of the brain and other nervous system.
tion [88]. Overall, there was no difference in J Natl Cancer Inst. 2011;103(9):714–36. https://doi.
overall survival between the groups. There are org/10.1093/jnci/djr077.
214 A. Lauko et al.
3. Herbst RS, Heymach JV, Lippman SM. Lung cancer. Oncol. 2004;22(5):785–94. https://doi.org/10.1200/
N Engl J Med. 2008;359(13):1367–80. https://doi. JCO.2004.07.215.
org/10.1056/NEJMra0802714. 16. Herbst RS, Prager D, Hermann R, et al. TRIBUTE:
4. Castrucci WA, Knisely JPS. An update on the treat- a phase III trial of erlotinib hydrochloride (OSI-774)
ment of CNS metastases in small cell lung cancer. combined with carboplatin and paclitaxel chemother-
Cancer J Sudbury Massachusetts. 2008;14(3):138–46. apy in advanced non-small-cell lung cancer. J Clin
https://doi.org/10.1097/PPO.0b013e318172d6e1. Oncol Off J Am Soc Clin Oncol. 2005;23(25):5892–
5. Sørensen JB, Hansen HH, Hansen M, Dombernowsky 9. https://doi.org/10.1200/JCO.2005.02.840.
P. Brain metastases in adenocarcinoma of the lung: 17. Gatzemeier U, Pluzanska A, Szczesna A, et al. Phase
frequency, risk groups, and prognosis. J Clin Oncol III study of erlotinib in combination with cisplatin and
Off J Am Soc Clin Oncol. 1988;6(9):1474–80. https:// gemcitabine in advanced non-small-cell lung cancer:
doi.org/10.1200/JCO.1988.6.9.1474. the Tarceva Lung Cancer Investigation Trial. J Clin
6. Sperduto PW, Yang TJ, Beal K, et al. Estimating sur- Oncol Off J Am Soc Clin Oncol. 2007;25(12):1545–
vival in patients with lung cancer and brain metasta- 52. https://doi.org/10.1200/JCO.2005.05.1474.
ses: an update of the graded prognostic assessment 18. Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in
for lung cancer using molecular markers (Lung- lung cancer: correlation with clinical response to gefi-
molGPA). JAMA Oncol. 2017;3(6):827–31. https:// tinib therapy. Science. 2004;304(5676):1497–500.
doi.org/10.1001/jamaoncol.2016.3834. https://doi.org/10.1126/science.1099314.
7. Kris MG, Johnson BE, Berry LD, et al. Using mul- 19. Lynch TJ, Bell DW, Sordella R, et al. Activating muta-
tiplexed assays of oncogenic drivers in lung cancers tions in the epidermal growth factor receptor under-
to select targeted drugs. JAMA. 2014;311(19):1998– lying responsiveness of non-small-cell lung cancer
2006. https://doi.org/10.1001/jama.2014.3741. to gefitinib. N Engl J Med. 2004;350(21):2129–39.
8. Alamgeer M, Ganju V, Watkins DN. Novel thera- https://doi.org/10.1056/NEJMoa040938.
peutic targets in non-small cell lung cancer. Curr 20. Pao W, Miller V, Zakowski M, et al. EGF receptor
Opin Pharmacol. 2013;13(3):394–401. https://doi. gene mutations are common in lung cancers from
org/10.1016/j.coph.2013.03.010. “never smokers” and are associated with sensitivity
9. Savas P, Hughes B, Solomon B. Targeted therapy in of tumors to gefitinib and erlotinib. Proc Natl Acad
lung cancer: IPASS and beyond, keeping abreast of Sci U S A. 2004;101(36):13306–11. https://doi.
the explosion of targeted therapies for lung cancer. org/10.1073/pnas.0405220101.
J Thorac Dis. 2013;5(Suppl 5):S579–92. https://doi. 21. Rosell R, Moran T, Queralt C, et al. Screening for
org/10.3978/j.issn.2072-1439.2013.08.52. epidermal growth factor receptor mutations in lung
10. Dong J, Li B, Lin D, Zhou Q, Huang D. Advances cancer. N Engl J Med. 2009;361(10):958–67. https://
in targeted therapy and immunotherapy for non- doi.org/10.1056/NEJMoa0904554.
small cell lung cancer based on accurate molecular 22. Mok TS, Wu Y-L, Thongprasert S, et al. Gefitinib or
typing. Front Pharmacol. 2019;10:230. https://doi. carboplatin-paclitaxel in pulmonary adenocarcinoma.
org/10.3389/fphar.2019.00230. N Engl J Med. 2009;361(10):947–57. https://doi.
11. Lekic M, Kovac V, Triller N, Knez L, Sadikov A, org/10.1056/NEJMoa0810699.
Cufer T. Outcome of small cell lung cancer (SCLC) 23. Eichler AF, Kahle KT, Wang DL, et al. EGFR muta-
patients with brain metastases in a routine clinical tion status and survival after diagnosis of brain metas-
setting. Radiol Oncol. 2012;46(1):54–9. https://doi. tasis in nonsmall cell lung cancer. Neuro-Oncology.
org/10.2478/v10019-012-0007-1. 2010;12(11):1193–9. https://doi.org/10.1093/neuonc/
12. Fischer B, Marinov M, Arcaro A. Targeting receptor noq076.
tyrosine kinase signalling in small cell lung cancer 24. Gow C-H, Chien C-R, Chang Y-L, et al. Radiotherapy
(SCLC): what have we learned so far? Cancer Treat in lung adenocarcinoma with brain metastases: effects
Rev. 2007;33(4):391–406. https://doi.org/10.1016/j. of activating epidermal growth factor receptor muta-
ctrv.2007.01.006. tions on clinical response. Clin Cancer Res Off J Am
13. Herbst RS, Morgensztern D, Boshoff C. The biol- Assoc Cancer Res. 2008;14(1):162–8. https://doi.
ogy and management of non-small cell lung can- org/10.1158/1078-0432.CCR-07-1468.
cer. Nature. 2018;553(7689):446–54. https://doi. 25. Hotta K, Kiura K, Ueoka H, et al. Effect of gefitinib
org/10.1038/nature25183. (“Iressa”, ZD1839) on brain metastases in patients
14. Giaccone G, Herbst RS, Manegold C, et al. Gefitinib with advanced non-small-cell lung cancer. Lung
in combination with gemcitabine and cisplatin in Cancer Amsterdam Netherlands. 2004;46(2):255–61.
advanced non-small-cell lung cancer: a phase III https://doi.org/10.1016/j.lungcan.2004.04.036.
trial – INTACT 1. J Clin Oncol Off J Am Soc Clin 26. Kim J-E, Lee DH, Choi Y, et al. Epidermal growth
Oncol. 2004;22(5):777–84. https://doi.org/10.1200/ factor receptor tyrosine kinase inhibitors as a first-
JCO.2004.08.001. line therapy for never-smokers with adenocarcinoma
15. Herbst RS, Giaccone G, Schiller JH, et al. Gefitinib of the lung having asymptomatic synchronous brain
in combination with paclitaxel and carboplatin in metastasis. Lung Cancer Amsterdam Netherlands.
advanced non-small-cell lung cancer: a phase III 2009;65(3):351–4. https://doi.org/10.1016/j.
trial – INTACT 2. J Clin Oncol Off J Am Soc Clin lungcan.2008.12.011.
27. Porta R, Sánchez-Torres JM, Paz-Ares L, et al.

38. Kobayashi S, Boggon TJ, Dayaram T, et al. EGFR
Brain metastases from lung cancer respond- mutation and resistance of non-small-cell lung can-
ing to erlotinib: the importance of EGFR muta- cer to gefitinib. N Engl J Med. 2005;352(8):786–92.
tion. Eur Respir J. 2011;37(3):624–31. https://doi. https://doi.org/10.1056/NEJMoa044238.
org/10.1183/09031936.00195609. 39. Sequist LV, Waltman BA, Dias-Santagata D, et al.
28. Deng Y, Feng W, Wu J, et al. The concentration of Genotypic and histological evolution of lung cancers
erlotinib in the cerebrospinal fluid of patients with acquiring resistance to EGFR inhibitors. Sci Transl
brain metastasis from non-small-cell lung cancer. Med. 2011;3(75):75ra26. https://doi.org/10.1126/
Mol Clin Oncol. 2014;2(1):116–20. https://doi. scitranslmed.3002003.
org/10.3892/mco.2013.190. 40. Camidge DR, Pao W, Sequist LV. Acquired resistance
29. Zhao J, Chen M, Zhong W, et al. Cerebrospinal fluid to TKIs in solid tumours: learning from lung cancer.
concentrations of gefitinib in patients with lung ade- Nat Rev Clin Oncol. 2014;11(8):473–81. https://doi.
nocarcinoma. Clin Lung Cancer. 2013;14(2):188–93. org/10.1038/nrclinonc.2014.104.
https://doi.org/10.1016/j.cllc.2012.06.004. 41. Maemondo M, Inoue A, Kobayashi K, et al.

30. Clarke JL, Pao W, Wu N, Miller VA, Lassman
Gefitinib or chemotherapy for non-small-cell
AB. High dose weekly erlotinib achieves therapeutic lung cancer with mutated EGFR. N Engl J Med.
concentrations in CSF and is effective in leptomenin- 2010;362(25):2380–8. https://doi.org/10.1056/
geal metastases from epidermal growth factor receptor NEJMoa0909530.
mutant lung cancer. J Neuro-Oncol. 2010;99(2):283– 42. Sequist LV, Yang JC-H, Yamamoto N, et al. Phase
6. https://doi.org/10.1007/s11060-010-0128-6. III study of afatinib or cisplatin plus pemetrexed
31. Grommes C, Oxnard GR, Kris MG, et al. “Pulsatile” in patients with metastatic lung adenocarcinoma
high-dose weekly erlotinib for CNS metastases with EGFR mutations. J Clin Oncol Off J Am Soc
from EGFR mutant non-small cell lung cancer. Clin Oncol. 2013;31(27):3327–34. https://doi.
Neuro-Oncology. 2011;13(12):1364–9. https://doi. org/10.1200/JCO.2012.44.2806.
org/10.1093/neuonc/nor121. 43. Wu Y-L, Zhou C, Hu C-P, et al. Afatinib versus cis-
32. Yu HA, Sima C, Feldman D, et al. Phase 1 study of platin plus gemcitabine for first-line treatment of
twice weekly pulse dose and daily low-dose erlotinib Asian patients with advanced non-small-cell lung
as initial treatment for patients with EGFR-mutant cancer harbouring EGFR mutations (LUX-Lung
lung cancers. Ann Oncol. 2017;28(2):278–84. https:// 6): an open-label, randomised phase 3 trial. Lancet
doi.org/10.1093/annonc/mdw556. Oncol. 2014;15(2):213–22. https://doi.org/10.1016/
33. Wu Y-L, Zhou C, Cheng Y, et al. Erlotinib as
S1470-2045(13)70604-1.
second- line treatment in patients with advanced 44. Jänne PA, Yang JC-H, Kim D-W, et al. AZD9291 in
non-small-cell lung cancer and asymptomatic brain EGFR inhibitor-resistant non-small-cell lung cancer.
metastases: a phase II study (CTONG–0803). Ann N Engl J Med. 2015;372(18):1689–99. https://doi.
Oncol. 2013;24(4):993–9. https://doi.org/10.1093/ org/10.1056/NEJMoa1411817.
annonc/mds529. 45. Soria J-C, Ohe Y, Vansteenkiste J, et al. Osimertinib
34. Welsh JW, Komaki R, Amini A, et al. Phase II trial in untreated EGFR-mutated advanced non–small-
of erlotinib plus concurrent whole-brain radiation cell lung cancer. N Engl J Med. 2017;378(2):113–25.
therapy for patients with brain metastases from non- https://doi.org/10.1056/NEJMoa1713137.
small-cell lung cancer. J Clin Oncol Off J Am Soc Clin 46. Mok TS, Wu Y-L, Ahn M-J, et al. Osimertinib or
Oncol. 2013;31(7):895–902. https://doi.org/10.1200/ platinum-pemetrexed in EGFR T790M-positive lung
JCO.2011.40.1174. cancer. N Engl J Med. 2017;376(7):629–40. https://
35. Iuchi T, Shingyoji M, Sakaida T, et al. Phase II
doi.org/10.1056/NEJMoa1612674.
trial of gefitinib alone without radiation therapy 47. Morris SW, Kirstein MN, Valentine MB, et al.

for Japanese patients with brain metastases from Fusion of a kinase gene, ALK, to a nucleolar protein
EGFR-mutant lung adenocarcinoma. Lung Cancer gene, NPM, in non-Hodgkin’s lymphoma. Science.
Amsterdam Netherlands. 2013;82(2):282–7. https:// 1995;267(5196):316–7.
doi.org/10.1016/j.lungcan.2013.08.016. 48. Lin JJ, Riely GJ, Shaw AT. Targeting ALK: precision
36. Fan Y, Huang Z, Fang L, et al. A phase II study of ico- medicine takes on drug resistance. Cancer Discov.
tinib and whole-brain radiotherapy in Chinese patients 2017;7(2):137–55. https://doi.org/10.1158/2159-
with brain metastases from non-small cell lung can- 8290.CD-16-1123.
cer. Cancer Chemother Pharmacol. 2015;76(3):517– 49. Soda M, Choi YL, Enomoto M, et al. Identification
23. https://doi.org/10.1007/s00280-015-2760-5. of the transforming EML4-ALK fusion gene in non-
37. Yang J-J, Zhou C, Huang Y, et al. Icotinib versus small-cell lung cancer. Nature. 2007;448(7153):561–
whole-brain irradiation in patients with EGFR-mutant 6. https://doi.org/10.1038/nature05945.
non-small-cell lung cancer and multiple brain metas- 50. Takeuchi K, Choi YL, Soda M, et al. Multiplex

tases (BRAIN): a multicentre, phase 3, open-label, reverse transcription-PCR screening for EML4-
parallel, randomised controlled trial. Lancet Respir ALK fusion transcripts. Clin Cancer Res Off J Am
Med. 2017;5(9):707–16. https://doi.org/10.1016/ Assoc Cancer Res. 2008;14(20):6618–24. https://doi.
S2213-2600(17)30262-X. org/10.1158/1078-0432.CCR-08-1018.
216 A. Lauko et al.
51. Shaw AT, Yeap BY, Mino-Kenudson M, et al. Clinical 62. Hida T, Nokihara H, Kondo M, et al. Alectinib ver-
features and outcome of patients with non-small-cell sus crizotinib in patients with ALK-positive non-
lung cancer who harbor EML4-ALK. J Clin Oncol small-cell lung cancer (J-ALEX): an open-label,
Off J Am Soc Clin Oncol. 2009;27(26):4247–53. randomised phase 3 trial. Lancet London England.
https://doi.org/10.1200/JCO.2009.22.6993. 2017;390(10089):29–39. https://doi.org/10.1016/
52. Wong DW-S, Leung EL-H, So KK-T, et al. The EML4- S0140-6736(17)30565-2.
ALK fusion gene is involved in various histologic 63. Peters S, Camidge DR, Shaw AT, et al. Alectinib ver-
types of lung cancers from nonsmokers with wild- sus crizotinib in untreated ALK-positive non-small-
type EGFR and KRAS. Cancer. 2009;115(8):1723– cell lung cancer. N Engl J Med. 2017;377(9):829–38.
33. https://doi.org/10.1002/cncr.24181. https://doi.org/10.1056/NEJMoa1704795.
53. Cui JJ, Tran-Dubé M, Shen H, et al. Structure
64. Solomon BJ, Besse B, Bauer TM, et al. Lorlatinib in
based drug design of crizotinib (PF-02341066), a patients with ALK-positive non-small-cell lung can-
potent and selective dual inhibitor of mesenchymal- cer: results from a global phase 2 study. Lancet Oncol.
epithelial transition factor (c-MET) kinase and ana- 2018;19(12):1654–67. https://doi.org/10.1016/
plastic lymphoma kinase (ALK). J Med Chem. S1470-2045(18)30649-1.
2011;54(18):6342–63. https://doi.org/10.1021/ 65. Guin S, Ru Y, Wynes MW, et al. Contributions of
jm2007613. KRAS and RAL in non-small-cell lung cancer growth
54. Solomon BJ, Mok T, Kim D-W, et al. First-line crizo- and progression. J Thorac Oncol Off Publ Int Assoc
tinib versus chemotherapy in ALK-positive lung can- Study Lung Cancer. 2013;8(12):1492–501. https://
cer. N Engl J Med. 2014;371(23):2167–77. https:// doi.org/10.1097/JTO.0000000000000007.
doi.org/10.1056/NEJMoa1408440. 66. Chan BA, Hughes BGM. Targeted therapy for

55. Solomon BJ, Cappuzzo F, Felip E, et al. Intracranial non-small cell lung cancer: current standards and
efficacy of crizotinib versus chemotherapy in patients the promise of the future. Transl Lung Cancer
with advanced ALK-Positive non-small-cell lung can- Res. 2015;4(1):36–54. https://doi.org/10.3978/j.
cer: results from PROFILE 1014. J Clin Oncol Off issn.2218-6751.2014.05.01.
J Am Soc Clin Oncol. 2016;34(24):2858–65. https:// 67. Wood K, Hensing T, Malik R, Salgia R. Prognostic
doi.org/10.1200/JCO.2015.63.5888. and predictive value in KRAS in non-small-cell lung
56. Shaw AT, Kim D-W, Nakagawa K, et al. Crizotinib cancer: a review. JAMA Oncol. 2016;2(6):805–12.
versus chemotherapy in advanced ALK-positive lung https://doi.org/10.1001/jamaoncol.2016.0405.
cancer. N Engl J Med. 2013;368(25):2385–94. https:// 68. Riely GJ, Johnson ML, Medina C, et al. A phase II
doi.org/10.1056/NEJMoa1214886. trial of Salirasib in patients with lung adenocarcino-
57. Shaw AT, Kim D-W, Mehra R, et al. Ceritinib in mas with KRAS mutations. J Thorac Oncol Off Publ
ALK-rearranged non-small-cell lung cancer. N Engl J Int Assoc Study Lung Cancer. 2011;6(8):1435–7.
Med. 2014;370(13):1189–97. https://doi.org/10.1056/ https://doi.org/10.1097/JTO.0b013e318223c099.
NEJMoa1311107. 69. Jänne PA, van den Heuvel MM, Barlesi F, et al.

58. Shaw AT, Gandhi L, Gadgeel S, et al. Alectinib in Selumetinib plus docetaxel compared with docetaxel
ALK-positive, crizotinib-resistant, non-small-cell alone and progression-free survival in patients with
lung cancer: a single-group, multicentre, phase 2 KRAS-mutant advanced non-small cell lung can-
trial. Lancet Oncol. 2016;17(2):234–42. https://doi. cer: the SELECT-1 randomized clinical trial. JAMA.
org/10.1016/S1470-2045(15)00488-X. 2017;317(18):1844–53. https://doi.org/10.1001/
59. Kim D-W, Tiseo M, Ahn M-J, et al. Brigatinib in jama.2017.3438.
patients with crizotinib-refractory anaplastic lym- 70. Kim JH, Kim HS, Kim BJ. Prognostic value of KRAS
phoma kinase-positive non-small-cell lung cancer: a mutation in advanced non-small-cell lung cancer
randomized, multicenter phase II trial. J Clin Oncol treated with immune checkpoint inhibitors: a meta-
Off J Am Soc Clin Oncol. 2017;35(22):2490–8. analysis and review. Oncotarget. 2017;8(29):48248–
https://doi.org/10.1200/JCO.2016.71.5904. 52. https://doi.org/10.18632/oncotarget.17594.
60. Shaw A, Mehra R, Tan DSW, et al. BM-32CERITINIB 71. Reck M, Rodríguez-Abreu D, Robinson AG, et al.
(LDK378) for treatment of patients with ALK- Pembrolizumab versus chemotherapy for PD-L1-
rearranged (ALK+) non-small cell lung cancer positive non-small-cell lung cancer. N Engl J Med.
(NSCLC) and BRAIN metastases (BM) in the 2016;375(19):1823–33. https://doi.org/10.1056/
ASCEND-1 trial. Neuro-Oncology. 2014;16(Suppl NEJMoa1606774.
5):v39. https://doi.org/10.1093/neuonc/nou240.32. 72. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab
61. Gadgeel SM, Gandhi L, Riely GJ, et al. Safety and versus docetaxel in advanced nonsquamous
activity of alectinib against systemic disease and brain non–small-cell lung cancer. N Engl J Med.
metastases in patients with crizotinib-resistant ALK- 2015;373(17):1627–39. https://doi.org/10.1056/
rearranged non-small-cell lung cancer (AF-002JG): NEJMoa1507643.
results from the dose-finding portion of a phase 1/2 73. Horn L, Mansfield AS, Szczęsna A, et al. First-line
study. Lancet Oncol. 2014;15(10):1119–28. https:// atezolizumab plus chemotherapy in extensive-
doi.org/10.1016/S1470-2045(14)70362-6. stage small-cell lung cancer. N Engl J Med.
2018;379(23):2220–9. https://doi.org/10.1056/ Oncol. 2012;13(5):459–65. https://doi.org/10.1016/

NEJMoa1809064. S1470-2045(12)70090-6.
74. Takamori S, Toyokawa G, Okamoto I, et al. Clinical 82. Crinò L, Bidoli P, Ulivi P, et al. P1.01-053 Italian
significance of PD-L1 expression in brain metas- Nivolumab Expanded Access Programme (EAP):
tases from non-small cell lung cancer. Anticancer data from patients with advanced non-squamous
Res. 2018;38(1):553–7. https://doi.org/10.21873/ NSCLC and brain metastases. J Thorac Oncol.
anticanres.12258. 2017;12(11):S1915. https://doi.org/10.1016/j.
75. Berghoff AS, Ricken G, Wilhelm D, et al. Tumor infil- jtho.2017.09.707.
trating lymphocytes and PD-L1 expression in brain 83. Molinier O, Audigier-Valette C, Cadranel J, et al.
metastases of small cell lung cancer (SCLC). J Neuro- OA 17.05 IFCT-1502 CLINIVO: real-life experience
Oncol. 2016;130(1):19–29. https://doi.org/10.1007/ with nivolumab in 600 patients (Pts) with advanced
s11060-016-2216-8. non-small cell lung cancer (NSCLC). J Thorac
76. Mansfield AS, Aubry MC, Moser JC, et al. Temporal Oncol. 2017;12(11):S1793. https://doi.org/10.1016/j.
and spatial discordance of programmed cell death- jtho.2017.09.430.
ligand 1 expression and lymphocyte tumor infiltration 84. Gandhi L, Rodríguez-Abreu D, Gadgeel S, et al.

between paired primary lesions and brain metastases Pembrolizumab plus chemotherapy in meta-
in lung cancer. Ann Oncol Off J Eur Soc Med Oncol. static non–small-cell lung cancer. N Engl J Med.
2016;27(10):1953–8. https://doi.org/10.1093/annonc/ 2018;378(22):2078–92. https://doi.org/10.1056/
mdw289. NEJMoa1801005.
77. Berghoff AS, Fuchs E, Ricken G, et al. Density
85. El Rassy E, Botticella A, Kattan J, Le Péchoux

of tumor-infiltrating lymphocytes correlates with C, Besse B, Hendriks L. Non-small cell lung can-
extent of brain edema and overall survival time in cer brain metastases and the immune system: from
patients with brain metastases. Oncoimmunology. brain metastases development to treatment. Cancer
2016;5(1):e1057388. https://doi.org/10.1080/21624 Treat Rev. 2018;68:69–79. https://doi.org/10.1016/j.
02X.2015.1057388. ctrv.2018.05.015.
78. Berghoff AS, Lassmann H, Preusser M, Höftberger 86. Goldberg SB, Gettinger SN, Mahajan A, et al.

R. Characterization of the inflammatory response Durability of brain metastasis response and overall
to solid cancer metastases in the human brain. Clin survival in patients with non-small cell lung cancer
Exp Metastasis. 2013;30(1):69–81. https://doi. (NSCLC) treated with pembrolizumab. J Clin Oncol.
org/10.1007/s10585-012-9510-4. 2018;36(15_suppl):2009. https://doi.org/10.1200/
79. Cloughesy TF, Mochizuki AY, Orpilla JR, et al.
JCO.2018.36.15_suppl.2009.
Neoadjuvant anti-PD-1 immunotherapy promotes 87. Singh C, Qian JM, Yu JB, Chiang VL. Local tumor
a survival benefit with intratumoral and systemic response and survival outcomes after combined stereo-
immune responses in recurrent glioblastoma. Nat tactic radiosurgery and immunotherapy in non-small
Med. 2019;25(3):477–86. https://doi.org/10.1038/ cell lung cancer with brain metastases. J Neurosurg.
s41591-018-0337-7. 2019;132(2):512–7. https://doi.org/10.3171/2018.10.
80. Schalper KA, Rodriguez-Ruiz ME, Diez-Valle R,
JNS181371.
et al. Neoadjuvant nivolumab modifies the tumor 88. Chen L, Douglass J, Kleinberg L, et al. Concurrent
immune microenvironment in resectable glioblas- immune checkpoint inhibitors and stereotactic radio-
toma. Nat Med. 2019;25(3):470–6. https://doi. surgery for brain metastases in non-small cell lung
org/10.1038/s41591-018-0339-5. cancer, melanoma, and renal cell carcinoma. Int
81.
Margolin K, Ernstoff MS, Hamid O, et al. J Radiat Oncol. 2018;100(4):916–25. https://doi.
Ipilimumab in patients with melanoma and brain org/10.1016/j.ijrobp.2017.11.041.
metastases: an open-label, phase 2 trial. Lancet
Systemic Therapy of Brain
Metastases: Breast Cancer
15
Leigh Klaus Swartz and Aki Morikawa
Introduction Epidemiology
Breast cancer is the most common malignancy Among women in the USA, breast cancer is the
diagnosed in women and remains a significant most commonly diagnosed malignancy and the
public health burden worldwide. Despite signifi- second most common solid tumor to metastasize
cant advances in the management of breast can- to the brain [3]. In patients with breast cancer, the
cer that have improved overall survival, the presence of brain metastases confers a poor prog-
prognosis of patients with breast cancer brain nosis. Population-based studies utilizing the
metastases (BCBM) remains poor [1]. Depending Surveillance, Epidemiology, and End Results
on a number of factors including tumor charac- (SEER) database of the National Cancer Institute
teristics and performance status, the mainstay of (NCI) have shown that up to 8% of patients have
treatment is neurosurgery and/or radiation ther- BCBM at initial presentation [1]. The incidence
apy. Ongoing research has vastly accelerated the of BCBM varies according to tumor subtype,
role for systemic therapy in the management of based on human epidermal growth factor recep-
BCBM; however, effective therapies remain lim- tor 2 (HER2) receptor overexpression and/or
ited. Active research areas include investigating gene amplification and the presence of estrogen
the underlying disease mechanisms, examining receptor (ER), and progesterone receptor (PR)
response to therapy, determining novel uses for staining by immunohistochemistry (IHC). The
systemic therapy, expanding targeted therapy, highest incidence of BCBM is among patients
optimizing drug formulations for penetration with HER2-positive and triple-negative breast
across the blood-brain barrier/brain-tumor bar- cancer (TNBC), which confer a respective 2.7-
rier (BBB/BTB), augmenting local therapy, and 1.4-fold higher risk as opposed to patients
researching screening strategies, and determining with ER-positive/PR-positive/HER2-negative
expanded outcome measures [2]. disease [4]. In fact, studies have demonstrated
that up to 55% of patients with HER2-positive
metastatic disease will develop BCBM [5].
Additional risk factors for developing subsequent
BCBM include younger age at diagnosis
L. K. Swartz · A. Morikawa (*) (between the ages of 20 and 39), shorter time to
Division of Hematology/Oncology, Department of
development of first metastasis, higher number of
Internal Medicine, University of Michigan,
Ann Arbor, MI, USA non-brain metastatic sites of disease, and higher
e-mail: morikawa@med.umich.edu tumor grade [4, 6].

https://doi.org/10.1007/978-3-030-42958-4_15
220 L. K. Swartz and A. Morikawa
Screening hoice and Timing of Initial Therapy

C
for BCBM
Despite increased incidence in certain popula-
tions, screening for BCBM is not currently rec- The decision to initiate therapy for BCBM is best
ommended at the time of breast cancer diagnosis approached with a multi-disciplinary team con-
in asymptomatic patients without suspicious neu- sisting of medical oncology, radiation oncology,
rologic exam findings. Clinicians are advised to neurosurgery, and palliative care. The standard
have a low threshold to order brain imaging in approach for patients with limited and extensive
HER2-positive metastatic breast cancer (MBC) brain metastases involves choosing among stereo-
patients with neurologic symptoms that may be tactic radiosurgery (SRS), WBRT, neurosurgery,
clinically indicative of CNS involvement [7]. systemic therapy, best supportive care, and pallia-
MRI is preferred over conventional CT, if not tive care. The presence of symptoms related to
contraindicated [8]. Conversely, for stage IIIC or BCBM is paramount to determining a treatment
higher melanoma, stage II or higher non-small plan. Local therapy is preferred over initial sys-
cell lung cancer (NSCLC), and any stage small temic therapy for symptomatic BCBM. Per the
cell lung cancer, the NCCN recommends screen- NCCN guidelines, consideration should first be
ing MRI brain at the time of diagnosis given the given to neurosurgery for biopsy purposes or for
propensity of these cancers to metastasize to the alleviation of mass effect or neurologic symptoms
brain [9–11]. In a retrospective review comparing caused by brain metastases [15]. For patients who
659 patients with NSCLC to 349 patients with are not deemed neurosurgical candidates due to
breast cancer, the study authors concluded that at either patient or disease factors, SRS, WBRT, sys-
time of diagnosis of brain metastases, breast can- temic therapy, and clinical trials should be consid-
cer patients are more likely to have neurologic ered. The principles of radiation and neurosurgery
symptoms, seizures, leptomeningeal involve- will be discussed elsewhere in this textbook.
ment, and brainstem involvement [12]. Patients
with BCBM were also more likely to receive
whole brain radiation therapy (WBRT) upfront Principles and Challenges
and to die from neurologic causes compared to of Systemic Therapy
patients with NSCLC [12]. Another retrospective
review of 100 patients with HER2-positive breast Currently, there are limited systemic therapy
cancer demonstrated that diagnosing BCBM options for patients with BCBM. Systemic ther-
before development of symptoms was associated apy is only rarely employed in the first-line man-
with increased survival, decreased use of whole agement of BCBM, as it has been traditionally
brain radiation therapy (WBRT), and fewer brain reserved as salvage therapy following local ther-
lesions [13]. Additionally, surveillance for apy or to jointly control extra-axial disease.
BCBM during treatment represents an ongoing Ongoing research is opening doors to additional
research focus due to variable penetrance of sys- uses of systemic therapy including as upfront
temic agents through the BBB/BTB. For exam- therapy, secondary prevention, radiosensitiza-
ple, in a retrospective review investigating the tion, and treatment of side effects from local
HER2-directed antibody-drug conjugate therapy. Systemic therapy does have advantages
trastuzumab- emtansine (T-DM1), which is in that it can concurrently treat both CNS and
known to not adequately cross the BBB, authors non-CNS disease, and may postpone or poten-
demonstrated that the first site of progression was tially obviate the need for radiation therapy [2,
the brain in 56% of HER2-positive patients [14]. 16–19]. The BBB/BTB also poses a significant
Such studies suggest the need for ongoing inves- clinical quandary, given limited penetration of
tigation into BCBM surveillance strategies in an many chemotherapies and lack of information
effort to improve patient quality of life, treatment regarding their CNS efficacy [20]. Many patients
options, and survival. with BCBM will also present with brain metasta-
ses after extensive exposure to prior chemothera-
15 Systemic Therapy of Brain Metastases: Breast Cancer 221
pies and radiation, which complicates the malignant melanoma who had not received prior
availability of additional systemic options. radiation therapy [22]. Similar to the first study,
Consideration must also be given to HER2 status patients received 100 mg/m2 cisplatin on day 1 and
when selecting systemic therapy, given numerous etoposide 1000 mg/m2 on days 1, 3, and 5 or days
advances in HER2-targeted therapy that have led 4, 6, and 8 of each 21-day cycle. Of the 56 patients
to relatively longer overall survival coupled with with BCBM, 21 patients (37.5%) achieved either a
high rates of CNS-only disease progression for complete or partial response.
patients who overexpress HER2 [5, 14]. The NCCN suggests the option of single-agent
capecitabine in patients with recurrent HER2-
negative BCBM, which is supported by drug
Systemic Therapy for BCBM uptake studies in BCBM tissue, limited case
reports, and studies that use capecitabine in com-
The NCCN suggests several systemic therapy bination with other agents [24–29]. As an oral for-
recommendations for recurrent BCBM including mulation, capecitabine represents a more
high-dose methotrexate, capecitabine, cisplatin convenient option than intravenous chemother-
plus etoposide, and capecitabine plus temozolo- apy. In a retrospective review of seven patients
mide [15]. Unfortunately, the data underlying with breast cancer, one of whom had leptomenin-
these recommendations originates from gener- geal carcinomatosis (LC) and two of whom had
ally weak studies consisting mostly of small both BCBM and LC, three patients had a com-
arm prospective studies, retrospective plete response, and three patients had stable dis-
single-
reviews, and case reports. The paucity of evi- ease with capecitabine monotherapy [28]. From
dence to support these therapies highlights the the time of initiation of capecitabine, progression-
need for more clinical trials designed to advance free survival (PFS) was 8 months and overall sur-
care for patients with BCBM. The recommenda- vival (OS) was 13 months [28]. Capecitabine plus
tion for high-dose methotrexate originates from a temozolomide was investigated in a phase I trial
retrospective review of 32 patients with CNS of 24 patients with BCBM, with one complete
metastases, of whom 29 (91%) had breast cancer; response and three partial responses (18% CNS
94% of patients received high-dose 3.5 g/m2 IV overall response rate (ORR)) [25]. Capecitabine
methotrexate [21]. Of all patients (including non- plus lapatinib was also investigated in a multi-
breast cancer patients), 56% had a partial center phase II trial of 242 patients with progres-
response or stable disease in response to therapy. sive HER2-positive BCBM [30]. In the
After excluding patients who also received con- single-agent lapatinib arm, CNS response rate
current radiation, systemic therapy with (RR) was 6%, while the lapatinib plus capecitabine
capecitabine, or intrathecal chemotherapy, 23 arm demonstrated a CNS RR of 20% [30].
patients were evaluated and 57% had a partial In addition to the systemic therapies suggested
response or stable disease after receiving solely by the NCCN, a number of alternate systemic
high-dose IV methotrexate [21]. options are being investigated including combina-
The combination of cisplatin plus etoposide for torial regimens, targeted therapy, novel formulations
patients with recurrent BCBM was analyzed in of chemotherapy, and new classes of therapeutics.
two separate studies that were published in the The majority of published trials were designed for
1990s [22, 23]. The first study evaluated 100 mg/ the treatment of recurrent BCBM after local ther-
m2 platinum on day 1 and etoposide 1000 mg/m2 apy. Bevacizumab, a humanized anti-VEGF mono-
on days 4, 6, and 8 of each 21-day cycle in a cohort clonal antibody, was studied in various
of 22 patients with BCBM [23]. The overall combinations in BCBM, including as part of a regi-
response rate was 55%, with 5 (23%) patients men with carboplatin and in the bevacizumab, eto-
achieving a complete response [23]. Hormone poside, and cisplatin (BEEP) regimen consisting of
receptor and HER2 status were not reported for the bevacizumab, etoposide, and cisplatin. The combi-
cohort. The second study was a larger, prospective nation of carboplatin plus bevacizumab was stud-
study of patients with brain metastases from breast ied in a phase II trial—38 patients with progressive
cancer, non-small cell lung cancer (NSCLC), and BCBM (30 HER-2 positive, 8 HER-2 negative)
were enrolled [31]. CNS ORR was 63%, and nanoparticles [43]. Etirinotecan pegol (pegylated
responses were observed in both HER2-positive irinotecan) is an extended release formulation of
and HER2-negative patients [31]. Another phase II the topoisomerase I inhibitor irinotecan that both
study evaluated the bevacizumab, etoposide, and increases exposure to and decreases the toxicity
cisplatin regimen (BEEP) in patients with BCBM of the active metabolite of irinotecan, SN-38
with any hormone receptor and HER2-status who [44]. The phase III BEACON trial compared
had progressed after WBRT [32]. In the analysis, etirinotecan pegol to physician’s choice and dem-
12 patients were evaluable for response assess- onstrated a significant improvement in OS of
ment, demonstrating 75% CNS ORR and PFS 10 months versus 4.8 months, respectively, in the
6.6 months [32]. Temozolomide was also investi- BCBM subgroup [44]. Of 852 patients enrolled
gated as a single agent and in various combinations in the study, 67 (8%) had a history of BCBM
alongside capecitabine, cisplatin, and liposomal [44]. Etirinotecan pegol is being actively studied
doxorubicin. As a single agent, temozolomide has in several BCBM clinical trials including the
shown poor CNS RR in BCBM. The CNS RR phase III ATTAIN study, which is designed to
ranged from 0% to 4% in four phase II trials of compare etirinotecan pegol to physician’s choice
temozolomide monotherapy in BCBM patients in patients with stable BCBM who have been pre-
[33–36]. The combination of temozolomide with viously treated with an anthracycline, taxane, and
other systemic agents has yielded more promising capecitabine (NCT02915744). A phase II study
results. A phase II study investigating temozolo- examining the efficacy of etirinotecan pegol in
mide plus cisplatin in patients with BCBM demon- patients with brain metastases from lung cancer
strated a 40% CNS RR [37]. Another phase II trial and breast cancer is also ongoing (NCT02312622).
examined the combination of temozolomide plus Nanoliposomal irinotecan, a nanoparticle formu-
liposomal doxorubicin in eight patients with lation that improves irinotecan pharmacokinet-
BCBM and demonstrated a CNS RR of 66% [38]. ics, is also undergoing active clinical investigation
Novel therapeutic agents are being investi- in a phase II trial for patients with progressive
gated including ixabepilone, patupilone, and HER2-negative BCBM after local therapy
sagopilone of the epothilone class of microtubule (NCT03328884) [45]. Novel drug conjugates,
inhibitors. Ixabepilone, alone or in combination such as ANG1005, a novel paclitaxel-peptide
with capecitabine, is currently approved for MBC conjugate, are being actively evaluated. In a
that is resistant to anthracyclines and taxanes; phase II trial for patients with recurrent BCBM
however, patients with BCBM were excluded and LC, patients received an infusion of
from the trials that led to FDA approval [39, 40]. ANG1005 every 3 weeks; HER2-positive patients
Unfortunately, results from other trials of agents were allowed to continue trastuzumab and/or per-
in the epothilone class have yielded discouraging tuzumab [46]. Seventy percent of patients dem-
results. Patupilone was studied in a multicenter onstrated an intracranial clinical benefit by
phase II trial in patients with BCBM. Cohort A RECIST criteria [46]. Another phase II trial (pub-
included patients who had received prior WBRT lication pending) investigated ANG1005 as a
and cohort B included patients with untreated single agent in HER2-negative disease and in
brain metastases or LC. Study authors concluded combination with trastuzumab in HER2-positive
that patupilone was ineffective in treating BCBM disease (NCT01480583).
with unacceptable GI toxicity [41]. Another drug The portfolio of targeted therapy in BCBM is
of the same class, sagopilone, showed disap- expanding to include inhibitors of CDK 4/6,
pointing results in a phase II trial of 15 patients poly-ADP ribose polymerase (PARP), and
with BCBM. The study showed a CNS RR of PIK3CA. The CDK 4/6 inhibitors palbociclib,
13.3% and PFS of only 1.4 months [42]. ribociclib, and abemaciclib are currently
In order to effectively penetrate the BBB/ approved to treat advanced ER-positive,
BTB, novel formulations of existing therapies are HER2-negative breast cancer; however, the trials
being extensively studied through techniques that led to FDA approval of ribociclib and abe-
such as liposomal drug delivery, pegylation, and maciclib excluded patients with CNS metastases
[47, 48]. The PALOMA-2 trial that led to the 7.8 months [52]. An alternate PARP inhibitor,
approval of palbociclib allowed for the inclusion veliparib, is currently being studied in combina-
of patients with stable BCBM, provided they had tion with cisplatin in patients with TNBC or
been treated with local, definitive therapy and BRCA-mutated breast cancer with or without
were asymptomatic from brain metastases [49]. BCBM (NCT02595905). PI3K (phosphoinosit-
Abemaciclib is being actively investigated in a ide 3-kinase) inhibitors represent another area
phase II trial in patients with brain metastases of active research, with ongoing trials for
from breast cancer, NSCLC, and melanoma advanced breast cancer [53]. The phase III
(NCT02308020). Abemaciclib has been showed SANDPIPER trial is studying the PI3K-
to cross the BBB in xenograft models and has inhibitor, taselisib, plus fulvestrant in patients
potential efficacy in BCBM [50]. Iniparib was with advanced breast cancer; however, patients
originally promoted as a PARP inhibitor but then with active or untreated BCBM are excluded
was shown to lack clinically relevant PARP inhi- from study participation (NCT02340221) [54].
bition [51]. The combination of iniparib and iri- Buparlisib (BKM120) plus capecitabine is being
notecan was studied in a failed phase II trial of investigated in a phase II study designed for
BCBM in patients with TNBC and demonstrated patients with BCBM (NCT02000882) (Tables
CNS RR of 12%, TTP 2.1 months, and OS 15.1 and 15.2).
Table 15.1 Selected phase II trials of systemic therapy alone

BCBM
Drug Combination patients Comments Result
Afatinib A: Afatinib A: 40 3 arm, randomized CNS RR A: 30%, B: 34.2%, C: 41.9%
B: +Vinorelbine B: 38 [55]
C: +Choice C: 43
Bevacizumab Carboplatin 38 – CNS RR 63% (composite) [31]
Etoposide and 16 (12 Progression after CNS RR 75% [32]
cisplatin evaluable) WBRT
Capecitabine Lapatinib 50 Prior RT and CNS RR 20% [30]
trastuzumab
Capecitabine Lapatinib 45 LANDSCAPE trial, CNS RR 66% (volumetric), TTP 5.5
first-line, no prior RT mos, OS 17 mos, Time to RT 8.3 mos
[16]
Cisplatin Temozolomide 15 – CNS RR 40% [37]
Lapatinib 22, 50 Prior RT CNS RR 20–38% [30, 56]
Iniparib Irinotecan 34 TBCRC 018, TNBC CNS RR 12%, TTP 2.1 mos, OS 7.8
only mos [52]
Lapatinib – 39, 242 – CNS RR 2.6–6%, OS 6.4 mos [30, 57]
Topotecan 22 Closed due to toxicity CNS RR 0% [56]
and lack of efficacy
Neratinib – 40 For recurrent brain mets CNS RR 8%, PFS 1.9 mos, OS 8.7 mos
[58]
Patupilone – 55 Group A: progressive Group A: CNS PFS 27% at 3 mos, OS
BCBM after WBRT 12.7 mos, ORR 9%
Group B: LC or Group B: CNS RR 0% [41]
untreated BCBM
Sagopilone – 15 For recurrent brain mets CNS RR 13.3%, PFS 1.4 mos, OS 5.3
mos [42]
Temozolomide – 4, 10, 19, 51 For recurrent brain mets CNS RR 0–4% [33–36]
Pegylated 8 – CNS RR 66% [38]
liposomal
doxorubicin
RR response rate, ORR overall response rate, OS median overall survival, PFS progression-free survival, TTP time to
progression, LC leptomeningeal carcinomatosis
Table 15.2 Selected ongoing clinical trials

Intervention Drug Phase Clinical trial ID
Systemic therapy ARRY-380, anti-HER2, trastuzumab I NCT01921335
alone
Autologous activated dendritic cells for intratumoral I NCT03638765
injection
T-DMI, T-DMI + metronomic temozolomide I, II NCT03190967
Abemaciclib II NCT02308020
Atezolizumab + pertuzumab + trastuzumab II NCT03417544
BKM120 (buparlisib) + capecitabine II NCT02000882
Cabozantinib II NCT02260531
Cisplatin ± veliparib II NCT02595905
Eribulin II NCT03637868, NCT02581839,
NCT03412955
Etirinotecan II NCT02312622
ANG1005 (paclitaxel-peptide II NCT01480583
conjugate) ± trastuzumab
HKI-272 (neratinib) II NCT01494662
Nanoliposomal irinotecan II NCT03328884
Palbociclib II NCT02774681
Pertuzumab + high-dose trastuzumab II NCT02536339
Pyrotinib + capecitabine II NCT03691051
Tucatinib + capecitabine + trastuzumab II NCT02614794
T-DMI II NCT03203616
Proteome-based immunotherapy II/III NCT01782274
NKTR-102 (etirinotecan pegol) versus physician’s III NCT02915744
choice
Concurrent with Sunitinib I NCT00981890
SRS Pembrolizumab I, II NCT03449238
Afatinib I, II NCT02768337
Atezolizumab II NCT03483012
Concurrent with Sorafenib I NCT01724606
WBRT
Bevacizumab + etoposide + cisplatin II NCT02185352
Lapatinib II NCT01622868
Source: clinicaltrials.gov, last accessed October 2018
HER2-Positive Therapy ous additional considerations in recurrent disease

Considerations in BCBM and with regard to systemic therapy. The nature
of HER2-positive disease, the BBB/BTB, and
HER2 overexpression is present in the tumors of HER2-targeted treatment necessitates consider-
15–20% of patients with breast cancer, and ation of the brain and body as separate compart-
approximately half of these patients will develop ments. Per ASCO HER2-positive guidelines,
brain metastases [7, 59]. The most widely used systemic therapy should not be changed if sys-
HER2-directed agent, trastuzumab, has demon- temic disease is controlled at the time of BCBM
strated poor penetration into the CNS, which may diagnosis [7]. In this scenario, local therapy
partially account for the increased incidence of should be employed to treat newly diagnosed
HER2-positive BCBM. Prior to the discovery of BCBM. If systemic disease is not controlled at
HER2-directed therapy, HER2-positive breast the time of BCBM diagnosis, standard HER2-
cancers carried the worst prognosis among breast directed therapy should be employed to treat dis-
cancer patients [60]. Timing and consideration of ease in the body and brain in addition to local
initial therapy are similar to patients without therapy for treating CNS disease [7]. The NCCN
HER2 overexpression; however, there are numer- guidelines recommend the following HER2-
targeted systemic therapy regimens for HER2- was a tolerable regimen with potential benefit
positive recurrent BCBM: capecitabine and [61]. In order to investigate prevention of BCBM,
lapatinib, capecitabine and neratinib, or pacli- the CEREBEL study (A Phase III, Randomized,
taxel and neratinib [15]. Lapatinib is an oral, Open-Label Study of Lapatinib Plus Capecitabine
reversible dual tyrosine kinase inhibitor (TKI) Versus Trastuzumab Plus Capecitabine in Patients
that blocks HER1/EGFR1 and HER2. Neratinib with Human Epidermal Growth Factor Receptor
is an oral, irreversible TKI with activity against 2-Positive Metastatic Breast Cancer) was
HER1, HER2, HER4, and EGFR. designed to compare the incidence of BCBM
Lapatinib has been shown to have drug uptake between patients receiving lapatinib plus
in BCBM tissue [29]. Lapatinib was initially capecitabine and trastuzumab plus capecitabine
investigated as a single-agent after prior radiation [62]. After enrollment of 540 patients, the study
therapy in multiple BCBM trials with discourag- was terminated early and was inconclusive in
ing results. Single-agent lapatinib only achieved regard to the primary endpoint [62]. The study
a CNS RR of 2.6–6.6% in two phase II trials [30, did not show a difference in BCBM incidence
57]. The LANDSCAPE trial (Lapatinib Plus between the treatment arms [62].
Capecitabine in Patients with Previously The NCCN recommendation for capecitabine
Untreated Brain Metastases from HER2-Positive and neratinib for recurrent HER2-positive BCBM
Metastatic Breast Cancer) was a single-arm derives from the Translational Breast Cancer
phase II trial that enrolled 45 patients who had Research Consortium (TBCRC) 022 trial, which
not received WBRT, capecitabine, or lapatinib is pending final publication [63]. This non-
and had at least 1 CNS lesion measuring 10 mm randomized phase II trial enrolled 39 patients
or greater in diameter [16]. Forty-three percent of with HER2-positive BCBM who had not previ-
patients in the study were asymptomatic at the ously received capecitabine or lapatinib [63].
time of the radiologic assessment which led to Patients were required to have at least 1 CNS
BCBM diagnosis [16]. Ninety-three percent of lesion measuring 10 mm or greater in diameter
patients had received trastuzumab-based chemo- prior to enrollment. At baseline, 65% of patients
therapy prior to study enrollment [16]. Per the had received prior WBRT [63]. Patients received
study protocol, patients were given lapatinib neratinib 250 mg daily by mouth and capecitabine
1250 mg daily and capecitabine 2000 mg/m2 750 mg/m2 by mouth on days 1–14 of each
days 1–4 of each 21-day cycle. The study demon- 21-day cycle. The final results of the trial have
strated a 65.9% CNS objective response rate at not yet been published, but results from the pre-
21 months; however, 49% of patients experienced liminary abstract indicate 49% of patients
grade 3–4 toxicity [16]. The most common tox- achieved VORR (≥50% reduction in volumetric
icities were diarrhea, palmar-plantar erythro- sum of CNS target lesions) [63]. As such, nera-
dysesthesia, fatigue, and rash [16]. Based on the tinib plus capecitabine represents a promising
results of the LANDSCAPE trial, ASCO sug- systemic therapy regimen for patients with
gests that initial therapy with lapatinib and HER2-positive BCBM.
capecitabine may be considered in patients who The NCCN provides a category 2B recommen-
have low-volume, asymptomatic BCBM who dation for paclitaxel and neratinib for recurrent
have not received prior radiation therapy [7]. Due HER2-positive BCBM [15]. This is based on the
to concerns regarding inferior BTB penetration NEfERT-T randomized clinical trial (Neratinib
of lapatinib and overlapping toxicity of lapatinib Plus Paclitaxel vs Trastuzumab Plus Paclitaxel in
and capecitabine, a phase I study was initiated to Previously Untreated Metastatic ERBB2-Positive
investigate intermittent high-dose lapatinib alter- Breast Cancer) that enrolled 479 women with
nating with capecitabine in HER2-positive CNS locally recurrent or metastatic HER2-positive
metastases [61]. Results suggested that intermit- breast cancer [64]. The study uses ERBB2 nomen-
tent high-dose lapatinib at a dose of 1500 mg clature as the equivalent of HER2. Patients were
BID 3 days on, 11 days off sequentially with randomized in 1:1 fashion to receive neratinib
capecitabine 1500 mg BID 7 days on, 7 days off 240 mg orally daily plus 80 mg/m2 IV paclitaxel
on days 1, 8, and 15 of each 28-day cycle or 4 mg/ Although trastuzumab has not been shown to
kg followed by 2 mg/kg IV every week of trastu- have adequate BBB penetration, T-DMI has
zumab plus paclitaxel [64]. Notably, paclitaxel is shown activity in HER2-positive BCBM in case
known to have poor penetration into the BBB reports [68, 69]. In a retrospective analysis of the
[65]. The trial excluded patients with active phase III EMILIA trial, authors examined the
BCBM, but did allow enrollment of patients with incidence of CNS metastases in the T-DMI cohort
CNS metastases or spinal cord involvement if versus the capecitabine plus lapatinib cohort
they were asymptomatic, had been previously [70]. Despite increased incidence of CNS metas-
treated with definitive radiation and/or neurosur- tases and CNS progression in the T-DMI cohort,
gery, and were not taking steroids or anticonvul- a significant improvement in OS was observed in
sants for 4 weeks prior to the study [64]. Like the the T-DMI cohort in patients who had asymptom-
CEREBEL study, NEfERT-T can be viewed as a atic, previously treated CNS metastases at base-
BCBM prevention trial. Of the 479 patients line [70]. This analysis supports the concept of
enrolled in the study, 18 had BCBM at the time of separate brain and body compartments in HER2-
study enrollment [64]. The study demonstrated a positive disease.
significant reduction in symptomatic or progres- Finally, alternate dosing strategies of existing
sive CNS events in the neratinib plus paclitaxel HER2-targeted agents are being studied. Similar
arm (8.3%) compared to the trastuzumab plus to the phase I high-dose lapatinib study, a trial is
paclitaxel arm (17.3%) which remained signifi- investigating the combination of pertuzumab plus
cant after adjusting for baseline BCBM [64]. high-dose trastuzumab in patients whose CNS
Similarly, study authors estimated 2-year inci- disease has progressed after radiation therapy
dence of CNS recurrence to be significantly (NCT02536339) [71]. Inadequate BBB penetra-
decreased in the neratinib plus paclitaxel arm tion of trastuzumab may be due to inadequate
compared to the trastuzumab plus paclitaxel arm dosing, and an impaired BBB after radiation ther-
(relative risk of 0.45) [64]. As demonstrated in apy has been shown to increase CNS concentra-
prior studies, the most common side effect of tions of trastuzumab [71].
neratinib was diarrhea, which occurred in 92.5%
of patients in the neratinib plus paclitaxel arm,
and can be managed with primary diarrheal pro- Systemic Therapy Concurrent
phylaxis [64]. Given these data, it is reasonable to with Radiation
consider neratinib plus paclitaxel for systemic
therapy in patients HER2-positive MBC, as it Numerous trials have studied the use of systemic
may delay onset of CNS metastases. therapy in conjunction with radiation therapy in
A number of other notable therapies for BCBM. Due to high recurrence rates after local
HER2-positive BCBM warrant further discus- therapy, research has focused on improving ini-
sion. Tucatinib is an oral TKI with selective tial therapy through immunotherapy, radiosensi-
activity against HER2, which has been studied in tization, adjuvant RT after surgical resection, and
two phase Ib trials in patients with HER2- alternative local techniques. Additional areas of
positive MBC with or without BCBM [66, 67]. active research include secondary chemopreven-
One study combined tucatinib with capecitabine tion after local therapy in an effort to prevent
and trastuzumab, and the other investigated the recurrence, as demonstrated in the CEREBEL
combination of tucatinib with T-DMI. Both and NEfERT-T trials [62, 64].
combinations demonstrated acceptable toxicity Earlier trials of chemotherapy with WBRT
and promising efficacy [66, 67]. A phase II trial were not as promising as newer studies with tar-
is ongoing, investigating the combination of geted therapeutics. For example, a phase II trial
tucatinib versus placebo with capecitabine plus investigating cisplatin and vinorelbine given con-
trastuzumab in a randomized, double-blind currently with 30 Gy WBRT showed a 3.7-month
approach (NCT02614794). PFS and 6.5-month OS [72]. However, CNS
response rate of 76% was consistent with con- Due to their ability to cross the BBB/BTB,
temporary trials [72]. Sorafenib, an oral TKI with PARP inhibitors are also being evaluated in con-
anti-VEGF activity, is being actively investigated junction with RT. A phase I trial examined velipa-
as a radiosensitizer when used concurrently with rib in combination with 30 or 37.5 Gy WBRT for
both SRS and WBRT. In a phase I trial of brain metastases from a variety of solid tumors
sorafenib with SRS for the treatment of 1–4 brain [77]. Of 81 patients enrolled in the study, 25 had
metastases, investigators demonstrated a 46% BCBM with median OS of 7.7 months [77]. When
CNS PFS at 1 year and median OS of 11.6 months compared to WBRT, no additional toxicity was
[73]. The trial enrolled 23 patients, of which 5 identified with the addition of veliparib [77].
had breast cancer (22%). Another study designed In an effort to expand the portfolio of second-
to investigate sorafenib with concurrent WBRT ary chemoprevention after SRS, a phase II trial of
for BCBM is actively recruiting patients sunitinib was completed in 14 patients (21% with
(NCT01724606) [74]. A novel PET tracer (FLT: BCBM) with 1–3 brain metastases [78].
3′deoxy-3′-fluorothymidine) will be used to Sunitinib, a TKI with anti-VEGF activity, was
improve response assessment. employed as an alternative to consolidation with
Lapatinib has also been investigated in con- WBRT after SRS. Study authors demonstrated
junction with WBRT with promising results in 6-month CNS PFS of 43% [78]. Another second-
early phase trials [75, 76]. In a phase II trial eval- ary chemoprevention trial is underway investigat-
uating lapatinib with WBRT in patients with ing T-DMI alone versus T-DMI plus metronomic
brain metastases from breast cancer and NSCLC, temozolomide after SRS in HER2-positive
patients with BCBM fared significantly better BCBM (NCT03190967).
than those with NSCLC. Breast cancer patients Immuno-oncology (IO) represents an area of
accounted for 22% of trial subjects and demon- active BCBM research, with many trials examin-
strated an 11.8-month median OS and 5-month ing the combination of immunotherapy and radi-
time to progression (TTP), compared to 4.2- ation. Although studies in melanoma and NSCLC
month median OS and 2.9-month TTP in patients have demonstrated CNS responses to IO agents
with NSCLC [75]. The difference in median OS, as monotherapy, breast cancer studies have so far
but not TTP, was statistically significant between failed to show improved outcomes [79]. Current
cohorts. An ongoing phase II trial is currently studies are investigating pembrolizumab or
investigating concurrent RT (WBRT or SRS) atezolizumab given concurrently with SRS in
with or without lapatinib in HER2-positive BCBM (NCT03449238 and NCT03483012)
BCBM (NCT01622868). (Table 15.3).
Table 15.3 Selected completed trials of systemic therapy with concurrent radiation therapy
Drug Phase RT Result
Bevacizumab I (65% 30 Gy WBRT OS 13.3 mos, TTP 7.1 mos [80]
BC)
Capecitabine + sunitinib II WBRT (dose 42% of pts removed from study due to toxicity, PFS
unspecified) 4.7 mos, OS 10 mos [81]
Cisplatin + vinorelbine II 30 Gy WBRT CNS RR 76%, PFS 3.7 mos, OS 6.5 mos [72]
Lapatinib I 37.5 Gy WBRT CNS RR 79% [76]
Lapatinib II (26% 30 Gy WBRT CNS RR 71%, OS 11.8 mos, TTP 5 mos [75]
BC)
Sorafenib I (22% SRS OS 11.6 mos, CNS progression 10 months [73]
BC)
Sunitinib II (21% SRS CNS PFS 6-mos after SRS 43% [78]
BC)
Veliparib I (31% 30 Gy or 37.5 Gy OS 7.7 mos [77]
BC) WBRT
BC breast cancer, RR response rate, OS median overall survival, PFS progression-free survival, TTP time to
progression
Complications of BCBM Therapy current with radiation, with a novel focus on

immunotherapy. The efficacy of immunotherapy
In addition to well-reported side effects from with radiation is not clear. Systemic monotherapy
individual systemic therapies, additional compli- is also being evaluated in in the prevention, first-
cations from therapy include radiation necrosis line, and recurrent settings. Promising therapeu-
and cognitive impairment. WBRT is notably tics include CDK 4/6 inhibitors, HER2-targeted
associated with well-reported neurocognitive therapy, cytotoxic therapy, TKIs, PARP inhibi-
side effects and fatigue. In a randomized, tors, and estrogen modulator therapy. Novel
placebo-controlled, double-blind trial of meman- mechanisms to deliver drugs to the brain through
tine started within 3 days of radiation initiation, the restrictive BBB/BTB are being pursued
Brown et al. found that the memantine cohort including liposomal drug delivery, pegylation,
showed less decline in executive function, mem- and use of nanoparticles [43]. In addition to
ory impairment, and processing speeds and better discrete, clinical endpoints, quality of life mea-
overall cognitive function [82]. sures are being incorporated into trial designs.
Another growing area of research is the rela- As more patients undergo genetic tissue typ-
tionship between systemic therapy and the inci- ing of their tumors, precision medicine will con-
dence of radiation necrosis [83]. Cerebral tinue to move to the forefront in many cancers,
radiation necrosis may present in an asymptom- including breast cancer. Precision medicine may
atic patient as simply an imaging finding or as a pave the way for a more tumor agnostic approach,
serious, potentially life-threatening complication in that the genomic features of the tumor impact
that can occur at any time up to years after radia- treatment decision-making more than the tumor
tion treatment. The diagnosis of radiation necro- type itself. Such an approach is being actively
sis can be difficult due to the complexity of investigated in the ASCO TAPUR (Targeted
distinguishing between tumor progression and Agent and Profiling Utilization Registry Study)
radiation necrosis on imaging. Biopsy remains and NCI MATCH (Molecular Analysis for
the gold standard but is not a feasible option in Therapy Choice) trials [89, 90]. However, both
many patients. In a retrospective review of 12 the TAPUR trial and MATCH trial exclude
BCBM patients, the incidence of radiation necro- patients with active, symptomatic brain metasta-
sis was 50% in those patients treated with T-DMI ses [89, 90].
simultaneously with SRS compared to 28.6% if Traditional clinical trials generally excluded
T-DMI was given sequentially with SRS [84]. In patients with active CNS disease. A 2017 meta-
a smaller retrospective review, four of seven analysis which reviewed phase I, II, and I/II tri-
patients treated with SRS prior to T-DMI devel- als for MBC concluded that only 29% allowed
oped clinically significant radiation necrosis patients with CNS metastases [91]. The 2018
[85]. Bevacizumab remains the most evidence- guidelines by the Response Assessment in
based option to treat radiation necrosis due to Neuro-Oncology Brain Metastases (RANO-BM)
VEGF dysregulation in radiation necrosis; how- working group stipulate clear recommendations
ever, bevacizumab cannot be used in patients for inclusion and exclusion of patients with
with acute cerebral hemorrhage [86]. brain metastases in an effort to safely advance
Bevacizumab can be a safe and effective therapy research for this challenging and underrepre-
for radiation necrosis that reduces steroid require- sented patient population [92]. The RANO-BM
ments [87, 88]. working group postulates clinical trial designs
based on the understanding of the potential CNS
activity of an experimental drug to recommend
Future Directions exclusion or inclusion of patients with brain
metastases into a particular clinical trial [92].
Many active trials have been specifically designed With updated guidelines for inclusion of patients
to advance the care of patients with BCBM. A with BCBM and numerous novel therapeutics to
number are investigating systemic therapy con- investigate, progress is anticipated in the treat-
ment of this challenging patient population. https://www.nccn.org/professionals/physician_gls/

This will require well-designed clinical trials pdf/nscl.pdf.
12. Cagney DN, Martin AM, Catalano PJ. Implications of
and collaboration between specialties across screening for brain metastases in patients with breast
multiple institutions. cancer and non–small cell lung cancer. JAMA Oncol.
2018;4(7):1001–3.
13. Morikawa A, Wang R, Patil S, Diab A, Yang J, Hudis
C, et al. Characteristics and prognostic factors for
References patients with HER2-overexpressing breast cancer and
brain metastases in the era of HER2-targeted therapy:
1. Martin AM, Cagney DN, Catalano PJ. Brain metas- an argument for earlier detection. Clin Breast Cancer.
tases in newly diagnosed breast cancer: a population- 2018;18(5):353–61.
based study. JAMA Oncol. 2017;3(8):1069–77. 14. Okines A, Irfan T, Khabra K, Smith I, O’Brien M,
2. Morikawa A, Jhaveri K, Seidman AD. Clinical tri- Parton M, et al. Development and responses of
als for breast cancer with brain metastases: chal- brain metastases during treatment with trastuzumab
lenges and new directions. Curr Breast Cancer Rep. emtansine (T-DM1) for HER2 positive advanced

2013;5:293–301. breast cancer: a single institution experience. Breast
3. Nieder C, Spanne O, Mehta MP, Grosu AL, Geinitz J. 2018;24(3):253–9.
H. Presentation, patterns of care, and survival in 15. NCCN. Central nervous system cancers Version
patients with brain metastases: what has changed in 1.2018. 2018. Available from: https://www.nccn.org/
the last 20 years? Cancer. 2010;117:2505. professionals/physician_gls/pdf/cns.pdf.
4. Graesslin O, Abdulkarim BS, Coutant C, Huguet F, 16. Bachelot T, Romieu G, Campone M, Dieras V, Cropet
Gabos Z, Hsu L, et al. Nomogram to predict subse- C, Dalenc F, et al. Lapatinib plus capecitabine in
quent brain metastasis in patients with metastatic patients with previously untreated brain metasta-
breast cancer. J Clin Oncol. 2010;28(12):2032–7. ses from HER2-positive metastatic breast cancer
5. Olson EM, Najita JS, Sohl J, Arnaout A, Burstein (LANDSCAPE): a single-group phase 2 study. Lancet
HJ, Winer EP, et al. Clinical outcomes and treatment Oncol. 2013;14(1):64–71.
practice patterns of patients with HER2-positive 17. Davies MA, Saiag P, Robert C, Grob J-J, Flaherty
metastatic breast cancer in the post-trastuzumab era. KT, Arance A. Dabrafenib plus trametinib in patients
Breast. 2013;22:525. with BRAFV600-mutant melanoma brain metastases
6. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau (COMBI-MB): a multicentre, multicohort, open-
FD, Lai P, Sawaya RE. Incidence proportions of brain label, phase 2 trial. Lancet Oncol. 2017;18(7):863–73.
metastases in patients diagnosed (1973 to 2001) in 18. Reungwetwattana T, Nakagawa K, Cho BC, Cobo
the metropolitan Detroit cancer surveillance system. M, Cho EK, Bertolini A, et al. CNS response to
J Clin Oncol. 2004;22(14):2865–72. osimertinib versus standard epidermal growth factor
7. Ramakrishna N, Temin S, Chandarlapaty S, Crews receptor tyrosine kinase inhibitors in patients with
JR, Davidson NE, Esteva FJ, et al. Recommendations untreated EGFR-mutated advanced non–small-cell
on disease management for patients with advanced lung cancer. J Clin Oncol. https://doi.org/10.1200/
human epidermal growth factor receptor 2–positive JCO.2018.78.3118.
breast cancer and brain metastases: American Society 19. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS,
of Clinical Oncology clinical practice guideline. J Moschos SJ, et al. Combined nivolumab and ipilim-
Clin Oncol. 2014;32(19):2100–9. umab in melanoma metastatic to the brain. N Engl J
8. Gaspar LE, Gore EM, Bradley GM, Germano I, Med. 2018;379(8):722–30.
Ghafoori P, Henderson MA, Lutz ST, McDermott 20. Kim M, Kizilbash S, Laramy J, Gampa G, Parrish
MW, Patchell RA, Patel SH, Robins HI, Vassil AD, K, Sarkaria J, et al. Barriers to effective drug treat-
Wippold FJ. ACR appropriateness criteria: pre- ment for brain metastases: a multifactorial problem
irradiation evaluation and management of brain metas- in the delivery of precision medicine. Pharm Res.
tases. American College of Radiology. 2011. http:// 2018;35(9):177.
www.acr.org/~/media/ACR/Documents/AppCriteria/ 21. Lassman AB, Abrey LE, Shah GD, Panageas KS,
Oncology/PreIrradiationEvaluationBrainMetastases. Begemann M, Malkin MG, et al. Systemic high-dose
pdf. intravenous methotrexate for central nervous system
9. NCCN. Melanoma Version 3.2018. 2018 [updated 12 metastases. J Neuro-Oncol. 2006;78(3):255–60.
July 2018]. Available from: https://www.nccn.org/ 22. Franciosi V, Cocconi G, Michiara M, Di Costanzo F,
professionals/physician_gls/pdf/melanoma.pdf. Fosser V, Tonato M, et al. Front-line chemotherapy
10. NCCN. Small cell lung cancer Version 1.2019. 2018 with cisplatin and etoposide for patients with brain
[updated 10 October 2018]. Available from: https:// metastases from breast carcinoma, nonsmall cell lung
www.nccn.org/professionals/physician_gls/pdf/sclc. carcinoma, or malignant melanoma: a prospective
pdf. study. Cancer. 1999;85(7):1599–605.
11. NCCN. Non-small cell lung cancer Version 1.2019. 23. Cocconi G, Lottici R, Bisagni G, Bacchi M, Tonato
2018 [updated 19 October 2018]. Available from: M, Passalacqua R, et al. Combination therapy with
platinum and etoposide of brain metastases from breast 35. Siena S, Crino L, Danova M, Del Prete S, Cascinu S,
carcinoma. Cancer Investig. 1990;8(3–4):327–34. Salvagni S, et al. Dose-dense temozolomide regimen
24. Fabi A, Vidiri A, Ferretti G, Felici A, Papaldo P, for the treatment of brain metastases from melanoma,
Carlini P, et al. Dramatic regression of multiple brain breast cancer, or lung cancer not amenable to surgery
metastases from breast cancer with capecitabine: or radiosurgery: a multicenter phase II study. Ann
another arrow at the bow? Cancer Investig. Oncol. 2010;21(3):655–61.
2006;24(4):466–8. 36.
Christodoulou C, Bafaloukos D, Kosmidis P,
25. Rivera E, Meyers C, Groves M, Valero V, Francis D, Samantas E, Bamias A, Papakostas P, et al. Phase
Arun B, et al. Phase I study of capecitabine in combi- II study of temozolomide in heavily pretreated can-
nation with temozolomide in the treatment of patients cer patients with brain metastases. Ann Oncol.
with brain metastases from breast carcinoma. Cancer. 2001;12(2):249–54.
2006;107(6):1348–54. 37.
Christodoulou C, Bafaloukos D, Linardou H,
26. Sutherland S, Ashley S, Miles D, Chan S, Wardley A, Aravantinos G, Bamias A, Carina M, et al.
Davidson N, et al. Treatment of HER2-positive meta- Temozolomide (TMZ) combined with cisplatin
static breast cancer with lapatinib and capecitabine in (CDDP) in patients with brain metastases from
the lapatinib expanded access programme, including solid tumors: a Hellenic Cooperative Oncology
efficacy in brain metastases--the UK experience. Br J Group (HeCOG) Phase II study. J Neuro-Oncol.
Cancer. 2010;102(6):995–1002. 2005;71(1):61–5.
27. Wang ML, Yung WK, Royce ME, Schomer DF,
38. Caraglia M, Addeo R, Costanzo R, Montella L, Faiola
Theriault RL. Capecitabine for 5-fluorouracil- V, Marra M, et al. Phase II study of temozolomide
resistant brain metastases from breast cancer. Am J plus pegylated liposomal doxorubicin in the treat-
Clin Oncol. 2001;24(4):421–4. ment of brain metastases from solid tumours. Cancer
28. Ekenel M, Hormigo AM, Peak S, Deangelis LM,
Chemother Pharmacol. 2006;57(1):34–9.
Abrey LE. Capecitabine therapy of central nervous 39. Thomas E, Gomez H, Li R. Ixabepilone plus

system metastases from breast cancer. J Neuro-Oncol. capecitabine for metastatic breast cancer progress-
2007;85(2):223–7. ing after anthracycline and taxane treatment. J Clin
29. Morikawa A, Peereboom DM, Thorsheim HR, Samala Oncol. 2007;25(33):5210–7.
R, Balyan R, Murphy CG, et al. Capecitabine and 40. Perez EA, Lerzo G, Pivot X, Thomas E, Vahdat L,
lapatinib uptake in surgically resected brain metasta- Bosserman L. Efficacy and safety of Ixabepilone
ses from metastatic breast cancer patients: a prospec- (BMS-247550) in a phase II study of patients with
tive study. Neuro-Oncology. 2015;17(2):289–95. advanced breast cancer resistant to an anthracy-
30. Lin NU, Dieras V, Paul D, Lossignol D, Christodoulou cline, a taxane, and capecitabine. J Clin Oncol.
C, Stemmler HJ, et al. Multicenter phase II study 2007;25(23):3407–14.
of lapatinib in patients with brain metastases from 41. Peereboom DM, Murphy C, Ahluwalia MS, Conlin
HER2-positive breast cancer. Clin Cancer Res. A, Eichler A, Poznak CV, et al. Phase II trial of patu-
2009;15(4):1452–9. pilone in patients with brain metastases from breast
31. Lin NU, Gelman RS, Younger WJ, Sohl J, Freedman cancer. Neuro-Oncology. 2014;16(4):579–83.
RA, Sorensen AG, Bullitt E, Harris GJ, Morganstern 42. Freedman RA, Bullitt E, Sun L, Gelman R, Harris
D, Schneider BP, Krop IE, Winer EP. Phase II trial of G, Ligibel JA, et al. A phase II study of sagopi-
carboplatin (C) and bevacizumab (BEV) in patients lone (ZK 219477; ZK-EPO) in patients with breast
(pts) with breast cancer brain metastases (BCBM) cancer and brain metastases. Clin Breast Cancer.
(ASCO Annual Meeting abstract). J Clin Oncol. 2011;11(6):376–83.
2013;31(supplement):513. 43. Shah N, Mohammad AS, Saralkar P, Sprowls SA,
32. Lu Y-S, Chen W-W, Lin C-H, Tseng L-M, Yeh
Vickers SD, John D, et al. Investigational chemo-
D-C, Wu P-F. Bevacizumab, etoposide, and cis- therapy and novel pharmacokinetic mechanisms
platin (BEEP) in brain metastases of breast cancer for the treatment of breast cancer brain metastases.
progressing from radiotherapy: results of the first Pharmacol Res. 2018;132:47–68.
stage of a multicenter phase II study. J Clin Oncol. 44. Perez EA, Awada A, O’Shaughnessy J, Rugo HS,
2012;30:1079. Twelves C, Im S-A. Etirinotecan pegol (NKTR-102)
33. Abrey LE, Olson JD, Raizer JJ, Mack M, Rodavitch versus treatment of physician’s choice in women with
A, Boutros DY, et al. A phase II trial of temozolomide advanced breast cancer previously treated with an
for patients with recurrent or progressive brain metas- anthracycline, a taxane, and capecitabine (BEACON):
tases. J Neuro-Oncol. 2001;53(3):259–65. a randomised, open-label, multicentre, phase 3 trial.
34. Trudeau ME, Crump M, Charpentier D, Yelle L, Lancet Oncol. 2015;16(15):1556–68.
Bordeleau L, Matthews S, et al. Temozolomide 45. Cortés J, Paez D, García JMP, Tormo SB, Parraga
in metastatic breast cancer (MBC): a phase II KA, Borrego MR, et al. Abstract CT154: Multicenter
trial of the National Cancer Institute of Canada - open-label, phase II trial, to evaluate the efficacy and
Clinical Trials Group (NCIC-CTG). Ann Oncol. safety of liposomal irinotecan (nal-IRI) for progress-
2006;17(6):952–6. ing brain metastases in patients with HER2-negative
breast cancer (The Phenomenal Study). Cancer Res. for patients with HER2-positive breast cancer brain
2018;78(13 Supplement):CT154. metastases. J Neuro-Oncol. 2011;105(3):613–20.
46. Kumthekar P, Tang S-C, Brenner AJ, Kesari S,
57. Lin NU, Carey LA, Liu MC, Younger J, Come SE,
Piccioni DE, Anders CK, et al. ANG1005, a novel Ewend M, et al. Phase II trial of lapatinib for brain
brain-penetrant taxane derivative, for the treatment metastases in patients with human epidermal growth
of recurrent brain metastases and leptomeningeal factor receptor 2-positive breast cancer. J Clin Oncol.
carcinomatosis from breast cancer. J Clin Oncol. 2008;26(12):1993–9.
2016;34(15_suppl):2004. 58. Freedman RA, Gelman RS, Wefel JS, Melisko ME,
47. Dickler MN, Tolaney SM, Rugo HS, Cortés J,
Hess KR, Connolly RM, et al. Translational breast
Diéras V, Patt D, et al. MONARCH 1, a phase II cancer research consortium (TBCRC) 022: a phase II
study of abemaciclib, a CDK4 and CDK6 inhibitor, trial of neratinib for patients with human epidermal
as a single agent, in patients with refractory HR+/ growth factor receptor 2-positive breast cancer and
HER2− metastatic breast cancer. Clin Cancer Res. brain metastases. J Clin Oncol. 2016;34(9):945–52.
2017;23(17):5218–24. 59. Brufsky AM, Mayer M, Rugo HS, Kaufman PA,

48. Hortobagyi GN, Stemmer SM, Burris HA, Yap Y-S, Tan-Chiu E, Tripathy D, et al. Central nervous sys-
Sonke GS, Paluch-Shimon S, et al. Ribociclib as first- tem metastases in patients with HER2-positive
line therapy for HR-positive, advanced breast cancer. metastatic breast cancer: incidence, treatment, and
N Engl J Med. 2016;375(18):1738–48. survival in patients from registHER. Clin Cancer Res.
49. Finn RS, Martin M, Rugo HS, Jones S, Im S-A, 2011;17(14):4834–43.
Gelmon K, et al. Palbociclib and Letrozole 60. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton
in advanced breast cancer. N Engl J Med. V, Bajamonde A, et al. Use of chemotherapy plus a
2016;375(20):1925–36. monoclonal antibody against HER2 for metastatic
50. Patnaik A, Rosen LS, Tolaney SM, Tolcher AW,
breast cancer that overexpresses HER2. N Engl J
Goldman JW, Gandhi L, et al. Efficacy and safety Med. 2001;344(11):783–92.
of abemaciclib, an inhibitor of CDK4 and CDK6, 61. Morikawa A, Pentsova E, Kemeny M, Patil S, Li
for patients with breast cancer, non–small cell lung BT, Tang K, et al. Phase I study of intermittent
cancer, and other solid tumors. Cancer Discov. high-dose lapatinib alternating with capecitabine
2016;6(7):740–53. for HER2- positive breast cancer with cen-
51. Patel AG, De Lorenzo SB, Flatten KS, Poirier
tral nervous system metastases. J Clin Oncol.
GG, Kaufmann SH. Failure of iniparib to inhibit 2018;36(15_suppl):e14016.
poly(ADP-ribose) polymerase in vitro. Clin Cancer 62. Pivot X, Manikhas A, Zurawski B, Chmielowska

Res. 2012;18(6):1655–62. E, Karaszewska B, Allerton R. CEREBEL
52. Anders C, Deal AM, Abramson V, Liu MC, Storniolo (EGF111438): a phase III, randomized, open-label
AM, Carpenter JT, et al. TBCRC 018: phase II study study of lapatinib plus capecitabine versus trastu-
of iniparib in combination with irinotecan to treat pro- zumab plus capecitabine in patients with human epi-
gressive triple negative breast cancer brain metasta- dermal growth factor receptor 2–positive metastatic
ses. Breast Cancer Res Treat. 2014;146(3):557–66. breast cancer. J Clin Oncol. 2015;33(14):1564–73.
53. Brosnan EM, Anders CK. Understanding patterns
63. Freedman R, Gelman R, Melisko M. TBCRC 022:
of brain metastasis in breast cancer and design- phase II trial of neratinib + capecitabine for patients
ing rational therapeutic strategies. Ann Trans Med. (Pts) with human epidermal growth factor receptor 2
2018;6(9):163. (HER2+) breast cancer brain metastases (BCBM). J
54. Baselga J, Cortés J, DeLaurentiis M, Dent S, Diéras Clin Oncol. 2017;35(15):1005.
V, Harbeck N, et al. SANDPIPER: phase III study 64. Awada A, Colomer R, Inoue K. Neratinib plus pacli-
of the PI3-kinase (PI3K) inhibitor taselisib (GDC- taxel vs trastuzumab plus paclitaxel in previously
0032) plus fulvestrant in patients (pts) with estro- untreated metastatic ERBB2-positive breast can-
gen receptor (ER)-positive, HER2-negative locally cer: the NEfERT-T randomized clinical trial. JAMA
advanced or metastatic breast cancer (BC) enriched Oncol. 2016;2(12):1557–64.
for pts with PIK3CA-mutant tumors. J Clin Oncol. 65. Fellner S, Bauer B, Miller DS, Schaffrik M, Fankhänel
2017;35(15_suppl):TPS1119. M, Spruss T, et al. Transport of paclitaxel (Taxol)
55. Cortes J, Dieras V, Ro J, Barriere J, Bachelot T, Hurvitz across the blood-brain barrier in vitro and in vivo. J
S, et al. Afatinib alone or afatinib plus vinorelbine Clin Invest. 2002;110(9):1309–18.
versus investigator’s choice of treatment for HER2- 66.
Murthy R, Borges VF, Conlin A, Chaves J,
positive breast cancer with progressive brain metasta- Chamberlain M, Gray T. Tucatinib with capecitabine
ses after trastuzumab, lapatinib, or both (LUX-Breast and trastuzumab in advanced HER2-positive meta-
3): a randomised, open-label, multicentre, phase 2 static breast cancer with and without brain metastases:
trial. Lancet Oncol. 2015;16(16):1700–10. a non-randomised, open-label, phase 1b study. Lancet
56. Lin NU, Eierman W, Greil R, Campone M, Kaufman Oncol. 2018;19(7):880–8.
B, Steplewski K, et al. Randomized phase II study of 67. Borges VF, Ferrario C, Aucoin N, et al. Tucatinib
lapatinib plus capecitabine or lapatinib plus topotecan combined with ado-trastuzumab emtansine in
advanced erbb2/her2-positive metastatic breast metastases: results of a phase 1 study. J Neuro-Oncol.

cancer: a phase 1b clinical trial. JAMA Oncol. 2015;122(2):409–17.
2018;4(9):1214–20. 78. Ahluwalia MS, Chao ST, Parsons MW, Suh JH, Wang
68. Terrell-Hall TB, Nounou MI, El-Amrawy F, Griffith D, Mikkelsen T, et al. Phase II trial of sunitinib as
JIG, Lockman PR. Trastuzumab distribution in an in- adjuvant therapy after stereotactic radiosurgery in
vivo and in-vitro model of brain metastases of breast patients with 1–3 newly diagnosed brain metastases. J
cancer. Oncotarget. 2017;8(48):83734–44. Neuro-Oncol. 2015;124(3):485–91.
69. Ricciardi GRR, Russo A, Franchina T, Schifano
79. Kotecki N, Lefranc F, Devriendt D, Awada A. Therapy
S, Mastroeni G, Santacaterina A, et al. Efficacy of of breast cancer brain metastases: challenges, emerg-
T-DM1 for leptomeningeal and brain metastases in a ing treatments and perspectives. Ther Adv Med
HER2 positive metastatic breast cancer patient: new Oncol. 2018;10:1758835918780312.
directions for systemic therapy - a case report and lit- 80. Lévy C, Allouache D, Lacroix J, Dugué AE, Supiot
erature review. BMC Cancer. 2018;18(1):97. S, Campone M, et al. REBECA: a phase I study of
70. Krop IE, Lin NU, Blackwell K, Guardino E, Huober bevacizumab and whole-brain radiation therapy for
J, Lu M, et al. Trastuzumab emtansine (T-DM1) ver- the treatment of brain metastasis from solid tumours.
sus lapatinib plus capecitabine in patients with HER2- Ann Oncol. 2014;25(12):2351–6.
positive metastatic breast cancer and central nervous 81. Niravath P, Tham Y, Wang T, Rodriguez A, Foreman
system metastases: a retrospective, exploratory analy- C, Hilsenbeck S, et al. A phase II trial of capecitabine
sis in EMILIA. Ann Oncol. 2015;26(1):113–9. concomitantly with whole-brain radiotherapy fol-
71. Lin NU, Lai C, Lacasia A, Stein A, Yoo B, Fung A, lowed by capecitabine and sunitinib for brain metas-
et al. An open-label, single-arm, phase II study of tases from breast cancer. Oncologist. 2015;20(1):13
pertuzumab with high-dose trastuzumab for treatment 82. Brown PD, Pugh S, Laack NN, Wefel JS, Khuntia
of central nervous system (CNS) progression post- D, Meyers C, et al. Memantine for the preven-
radiotherapy in patients (pts) with HER2-positive tion of cognitive dysfunction in patients receiving
metastatic breast cancer (MBC): PATRICIA. J Clin whole-brain radiotherapy: a randomized, double-
Oncol. 2016;34(15_suppl):TPS633. blind, placebo-controlled trial. Neuro-Oncology.
72. Cassier PA, Ray-Coquard I, Sunyach MP, Lancry
2013;15(10):1429–37.
L, Guastalla JP, Ferlay C, et al. A phase 2 trial of 83. Chao ST, Ahluwalia M, Barnett G, Stevens GHJ,

whole-brain radiotherapy combined with intravenous Murphy ES, Stockham AL, et al. Challenges with the
chemotherapy in patients with brain metastases from diagnosis and treatment of cerebral radiation necrosis.
breast cancer. Cancer. 2008;113(9):2532–8. Int J Radiat Oncol Biol Phys. 2013;87(3):449–57.
73. Arneson K, Mondschein J, Stavas M, Cmelak A, Attia 84. Geraud A, Xu H, Beuzeboc P, Kirova Y. Preliminary
A, Horn L, et al. A phase I trial of concurrent sorafenib experience of the concurrent use of radiosur-
and stereotactic radiosurgery for patients with brain gery and T-DMI for brain metastases in HER2-
metastases. J Neuro-Oncol. 2017;133(2):435–42. positive metastatic breast cancer. J Neuro-Oncol.
74. Morikawa A, Jhaveri KL, Patil S, Chen M, McDonnell 2017;131:69–72.
E, Neville DA, et al. A phase I trial of sorafenib with 85. Carlson JA, Nooruddin Z, Rusthoven C, Elias A,

whole brain radiotherapy (WBRT) in breast can- Borges VF, Diamond JR, et al. Trastuzumab emtan-
cer patients with brain metastases and a correlative sine and stereotactic radiosurgery: an unexpected
study of FLT-PET brain imaging to evaluate treat- increase in clinically significant brain edema. Neuro-
ment response after WBRT sorafenib. J Clin Oncol. Oncology. 2014;16(7):1006–9.
2014;32(15_suppl):TPS2103. 86. Gonzalez J, Kumar AJ, Conrad CA. Effect of bevaci-
75. Christodoulou C, Kalogera-Fountzila A, Karavasilis zumab on radiation necrosis of the brain. Int J Radiat
V, Kouvatseas G, Papandreou CN, Samantas E, et al. Oncol Biol Phys. 2007;67(2):323–6.
Lapatinib with whole brain radiotherapy in patients 87. Boothe D, Young R, Yamada Y, Prager A, Chan T,
with brain metastases from breast and non-small Beal K. Bevacizumab as a treatment for radiation
cell lung cancer: a phase II study of the Hellenic necrosis of brain metastases post stereotactic radio-
Cooperative Oncology Group (HeCOG). J Neuro- surgery. Neuro-Oncology. 2013;15(9):1257–63.
Oncol. 2017;134(2):443–51. 88. Delishaj D, Ursino S, Pasqualetti F, Cristaudo A,

76. Lin NU, Freedman RA, Ramakrishna N, Younger J, Cosottini M, Fabrini MG, et al. Bevacizumab for the
Storniolo AM, Bellon JR, et al. A phase I study of treatment of radiation-induced cerebral necrosis: a
lapatinib with whole brain radiotherapyin patients systematic review of the literature. J Clin Med Res.
with human epidermal growth factor receptor 2017;9(4):273–80.
2(HER2)-positive breast cancer brain metastases. 89. NCI. NCI-MATCH trial (molecular analysis for

Breast Cancer Res Treat. 2013;142:405–14. therapy choice). 2018 [updated August 21, 2018].
77. Mehta MP, Wang D, Wang F, Kleinberg L, Brade Available from: https://www.cancer.gov/about-
A, Robins HI, et al. Veliparib in combination with cancer/treatment/clinical-trials/nci-supported/
whole brain radiation therapy in patients with brain nci-match.
90. ASCO. ASCO TAPUR: targeted agent and profil-

92. Camidge DR, Lee EQ, Lin NU, Margolin K,

ing utilization registry study. 2018. Available from: Ahluwalia MS, Bendszus M. Clinical trial design
https://www.tapur.org/. for systemic agents in patients with brain metasta-
91. Costa R, Gill N, Rademaker A, Carneiro B, Chae Y, ses from solid tumours: a guideline by the Response
Kumthekar P, et al. Systematic analysis of early phase Assessment in Neuro-Oncology Brain Metastases
clinical studies for patients with breast cancer: inclu- working group. Lancet Oncol. 2018;19(1):30–2.
sion of patients with brain metastasis. Cancer Treat
Rev. 2017;55:10–5.
Systemic Therapy for Brain
Metastases: Melanoma
16
Sarah Weiss and Harriet Kluger
Introduction therapies on melanoma brain metastases was

understudied because few clinical trials allowed
Melanoma is the third most common cancer to for inclusion of these patients.
metastasize to the brain after lung cancer and Several recent clinical trials have demon-
breast cancer, but it has the highest propensity to strated activity of contemporary immune check-
spread to the brain [1]. Brain metastases occur in point inhibitors and BRAF and MEK inhibitors in
over 30% of patients with metastatic melanoma untreated melanoma brain metastases. This data
[2], and the incidence is even higher in autopsy has impacted current brain metastasis treatment
series [3, 4]. Risk factors for development of mel- paradigms, as local therapy may not be necessary
anoma brain metastases include primary melano- upfront in select patients. Even patients who
mas greater than 4 mm in size, location on the achieve disease control with systemic therapy for
scalp, and nodular histology [5]. Once brain extracranial metastases may relapse with CNS
metastases have developed, factors such as older metastases. It will be important to develop effec-
age, elevated lactate dehydrogenase, more than tive second-line therapies for these patients.
three brain metastases, poor performance status, Moreover, combinations of local and systemic
and neurologic symptoms are predictive of worse therapies have proven to be effective, but the
outcomes [3, 6]. Patients with leptomeningeal optimal sequencing of treatments remains
disease and those who develop brain metastases unclear, and there may be higher risk for chronic
after receiving systemic therapy for extracranial neurologic toxicity. Research is underway to
metastases also do poorly [3]. Historically, the understand how to sequence and combine sys-
median overall survival (OS) from the time of temic and local therapies for maximal activity
diagnosis of melanoma brain metastases was less with minimal toxicity and how to select patients
than 6 months [7]. Treatment approaches for mel- most likely to benefit from each treatment. Other
anoma brain metastases typically focused on important challenges to be addressed include
local therapies such as whole brain radiotherapy accurately assessing radiographic response to
(WBRT), stereotactic radiosurgery, or surgical treatment, managing radiation necrosis, and
resection. Until recently, the impact of systemic treating leptomeningeal disease, for which no
effective therapy currently exists. In this chapter,
we discuss systemic therapy approaches for the
S. Weiss · H. Kluger (*) management of melanoma brain metastases, cur-
Medicine (Medical Oncology), Yale University, rent therapeutic challenges, and new approaches
New Haven, CT, USA being investigated.
e-mail: harriet.kluger@yale.edu

https://doi.org/10.1007/978-3-030-42958-4_16
236 S. Weiss and H. Kluger
Historic Systemic Therapy Targeted Therapy

Approaches
BRAF V600 mutations are the most common tar-
There were essentially no effective systemic ther- getable driver mutations identified in melanoma.
apies for metastatic melanoma prior to the advent BRAF is a serine/threonine protein kinase that
of contemporary immunotherapy and targeted activates the MAP kinase/ERK signaling path-
therapies. Historically, chemotherapy including way and when mutated becomes constitutively
temozolomide, fotemustine, and several other activated, increasing melanoma cell proliferation
combination chemotherapies was studied in the and survival [15]. In small series, melanoma
setting of brain metastases without evidence of brain metastases have an estimated 44–55% rate
substantial activity [8–10]. of BRAF V600 mutations; BRAF status in mela-
Temozolomide is an oral alkylating agent noma brain metastases has been shown to be gen-
that was found to have equivalent efficacy to erally concordant with that of matched
intravenous dacarbazine in a phase 3 trial of extracranial metastases [16, 17]. Additional
patients with advanced extracranial melanoma molecular profiling of small cohorts of mela-
[11]. Due to the good CNS penetration of temo- noma brain metastases have shown a 23% rate of
zolomide, a phase 2 study enrolled 155 advanced NRAS mutations [18] as well as a significantly
melanoma patients with asymptomatic brain higher degree of PI3K/AKT pathway activation,
metastases not requiring immediate radiation PTEN loss, and increased expression of proteins
therapy. Intracranial disease benefit was such as STAT3 or SOCS1, compared to other
achieved in 36% of patients who were chemo- metastatic sites [19–22].
therapy-naïve, including 1 complete response Vemurafenib, one of the earliest mutant BRAF
(CR) (1%), 7 partial responses (PR) (6%), and inhibitors, was FDA approved in 2011 for patients
34 with stable disease (SD) (29%). Patients who with BRAF V600E mutant melanoma and was the
received prior chemotherapy had much lower first to be studied specifically in melanoma brain
intracranial response rates (1 PR (3%) and 6 SD metastases. In one pilot study, 24 patients with
(18%)). Responses did not last for meaningful BRAF V600-positive melanoma with unresect-
periods of time. Median progression-free surable, previously treated, symptomatic brain
vival (PFS) in the brain was only 1.2 months for metastases were treated with vemurafenib 960 mg
chemotherapy-naïve patients and 1 month who twice a day. Median treatment duration was only
had already received chemotherapy [12]. OS 3.8 months (range 0.1–11.3) limited mostly by
was also poor at 3.5 months and 2.2 months, disease progression, but 3/19 (16%) patients had a
respectively [11]. PR, demonstrating some intracranial activity in
Other studies using chemotherapy had equally previously treated patients; this set the foundation
disappointing results. For example, a phase 2 for further investigation of targeted approaches
study of fotemustine included a small cohort of for melanoma brain metastases [23].
patients with melanoma brain metastases and ini- A phase 2 study (BREAK-MB) used dab-
tially reported a 25% intracranial response rate rafenib, a BRAF inhibitor FDA approved in 2013,
[13]. However, a follow-up phase 3 trial of fote- to treat patients with V600E or V600K BRAF-
mustine versus fotemustine with whole brain mutant melanoma with at least one asymptomatic
radiation therapy in patients with melanoma brain metastasis (Cohort A included patients
brain metastases showed no significant differ- with no prior therapy, while Cohort B was com-
ences in intracranial response rates (7% vs 10%), prised of patients with progression of brain
which were much lower than that reported in the metastases after prior local therapy). The intra-
earlier phase 2 study. OS was brief, with no sig- cranial response rate for patients with V600E
nificant difference between the two groups (86 vs mutations was 39% (29/74) in Cohort A and 31%
105 days) [14]. (20/65) in Cohort B. Far fewer patients with
16 Systemic Therapy for Brain Metastases: Melanoma 237
V600K mutations were enrolled as this mutation Use of other available BRAF/MEK inhibitors
is much less frequent than V600E, but intracra- such as vemurafenib and cobimetinib is being
nial response rates were also lower than those studied after radiosurgery to brain metastases in
seen for V600E-mutant patients: 7% (1/15) in patients with BRAF-mutant melanoma
Cohort A and 22% (4/18) in Cohort B [24]. This (NCT03430947). Other targeted therapies being
study further confirmed the safety and activity of investigated include inhibitors of JAK2
BRAF inhibitors for melanoma patients with (NCT01904123), MEK (NCT03332589), and
treated or untreated brain metastases. PI3K (NCT02452294).
After combined BRAF and MEK inhibition
was found to be superior to BRAF inhibition
alone in advanced melanoma [25], the phase 2 Immune Checkpoint Inhibitors
COMBI-MB trial studied dabrafenib 150 mg
orally twice daily with trametinib 2 mg orally The CTLA-4 inhibitor ipilimumab was the first
daily in treatment-naïve melanoma patients with immune checkpoint inhibitor that was FDA
brain metastases. Patients were divided into four approved for the treatment of advanced mela-
cohorts: Cohort A (asymptomatic, V600E mutant, noma in 2011, followed by the PD-1 inhibitors
no prior local therapy, ECOG performance status pembrolizumab and nivolumab in 2014, and
0–1), Cohort B (asymptomatic, V600E mutant, combination regimen of ipilimumab and
prior local therapy, ECOG performance status nivolumab in 2015. Immunotherapy has the
0–1), Cohort C (asymptomatic, V600E/D/K/R potential to induce durable responses in patients
mutant with or without prior local therapy, ECOG with advanced melanoma. Historically, most
performance status 0–1), and Cohort D (symp- clinical trials studying these agents excluded
tomatic, V600E/D/K/R mutant with or without patients with brain metastases so their impact on
prior local therapy, ECOG performance status the CNS was initially unclear.
0–2). Intracranial response rates were 58%, 56%, Initial data on checkpoint inhibitors in mela-
44%, and 59% in Cohorts A, B, C, and D, respec- noma brain metastases came from a retrospective
tively [26]. Despite the encouraging response analysis of a phase 2 study of ipilimumab in
rates, the duration of response was brief. For patients with advanced melanoma. Twelve out of
example, median PFS in Cohort A was less than the 115 patients enrolled in the trial had stable
6 months, with a 1-year PFS rate of 19%. This is brain metastases at baseline. Of the 12 patients, 2
much lower than the PFS of over 11 months and had a PR and 3 had SD with treatment, with 3 of
a 3-year follow-up PFS rate of 22% for dab- these patients surviving beyond 4 years, suggest-
rafenib and trametinib in advanced melanoma ing activity and potential for durable responses in
patients without brain metastases [25, 27]. These the brain [30]. A subsequent phase 2 trial studied
results suggest that resistance to BRAF/MEK ipilimumab 10 mg/kg once every 3 weeks for
inhibition may occur more quickly in the brain four doses followed by maintenance every
compared to extracranial sites. Additional studies 12 weeks in brain metastasis patients, including
are necessary, but this may be related to insuffi- those who received prior WBRT or stereotactic
cient drug delivery or to development of distinct radiosurgery, as long as at least one untreated tar-
resistance pathways [28]. For example, in vitro get lesion was present. The majority of patients
studies have demonstrated that the addition of had received prior systemic therapy for mela-
cerebrospinal fluid to melanoma cell lines noma, including interferon, interleukin-2, or che-
reduced the cell death response to BRAF inhibi- motherapy. Two cohorts were studied: patients
tion, which was restored with addition of a PI3K with asymptomatic brain metastases not requir-
inhibitor [29]. ing corticosteroids (Cohort A) or symptomatic
There are multiple active clinical trials using patients on a stable dose of steroids (Cohort B).
targeted approaches for melanoma brain metastases. CNS objective response rates were 16% and 5%
in Cohorts A and B, respectively. PFS was less or leptomeningeal disease (Cohort C). At
than 2 months for both cohorts and median OS 17-month follow-up, intracranial responses were
was 7 months and 3.7 months, respectively. This seen in 46% (16/35), 20% (5/20), and 6% (1/16)
trial did not demonstrate unexpected adverse of patients in Cohorts A, B, and C, respectively.
events in the CNS and suggested modest activity Intracranial complete responses were seen in 6
for ipilimumab, particularly for small, asymp- (17%) and 3 (12%) patients in Cohorts A and B,
tomatic brain metastases in heavily pre-treated respectively, and none in Cohort C [34]. Adverse
patients [31]. events were similar to those seen in prior trials of
Anti-PD-1 therapy with or without ipilim- ipilimumab and nivolumab in melanoma patients
umab has since become the backbone for front- without CNS disease, and there were no unex-
line melanoma therapy and has recently been pected neurologic toxicities.
investigated in several clinical trials specific to Another phase 2 trial by Tawbi et al. evaluated
melanoma brain metastases. Pembrolizumab the safety and efficacy of combined ipilimumab
10 mg/kg every 2 weeks was studied in a phase 2 and nivolumab in patients with asymptomatic
trial of 23 melanoma patients with asymptomatic melanoma brain metastases measuring 5–30 mm,
brain metastases measuring 5–20 mm. At without prior local therapy. At 14-month follow-
24-month follow-up, the brain metastasis up of 94 patients, the intracranial response rate
response rate was 26%, and at follow-up the was 57% (CR 26%, PR 30%, SD 2%) and was
2-year OS rate was 48%. Intracranial and extra- concordant with the extracranial response rate of
cranial responses were concordant [32, 33]. 56%. Grade 3 or 4 adverse events occurred in
Although neurologic adverse events occurred in 55% of patients which is again consistent with
65% of patients, almost all instances were grade expected rates and is similar to Long et al.
1 or 2. The most common neurologic adverse Adverse events specific to the CNS occurred in
events included gait disturbance (22%) and head- 36% (34/94) of patients, and only 7% (7/94) were
ache (17%). Three patients developed seizures grade 3 or 4. The most common CNS adverse
which were controlled with antiepileptics, and event was headache which occurred in 21 patients
four patients developed neurologic symptoms (almost all grade 1 or 2). Only four patients had
related to perilesional edema. Radiation necrosis cerebral edema, three had intracranial hemor-
occurred in seven patients (30%), which was rhage, and two had seizures [35].
higher than expected. Based on these results, a These studies demonstrate the intracranial
trial of pembrolizumab in combination with bev- safety and efficacy of immune checkpoint inhibi-
acizumab is underway (NCT02681549) in tors for treatment of untreated melanoma brain
patients with untreated melanoma brain metasta- metastases. Historically stereotactic radiosurgery
ses to determine if bevacizumab can mitigate per- or resection has served as the initial therapy of
ilesional edema and radiation necrosis while also melanoma brain metastases. New data suggest
enhancing T-cell migration and thus antitumor that it is safe to start with upfront immunotherapy
immune responses. in select patients with close monitoring of
Combined treatment with ipilimumab and intracranial disease. However, stereotactic radio-
nivolumab has also been studied in two clinical surgery remains an important component, partic-
trials. The first was a phase 2 randomized study ularly for large lesions or lesions in neurologically
by Long et al. assessing the safety and efficacy of sensitive sites, and its use in combination with
ipilimumab plus nivolumab (Cohort A) compared systemic therapies is discussed below. Data dem-
to nivolumab monotherapy (Cohort B) in patients onstrate a higher rate of intracranial response
with asymptomatic melanoma brain metastases with combined ipilimumab with nivolumab com-
measuring 5–40 mm without prior local therapy. pared to anti-PD-1 therapy alone. Also, available
Nivolumab was also administered to a third evidence suggests that the durability of response
cohort of patients who had progression in the is longer with immunotherapy compared to tar-
brain after prior therapy, neurologic symptoms, geted therapy.
As new systemic therapy approaches become Combination of Immune Therapy

available and/or demonstrate efficacy for with SRS
extracranial disease, their efficacy is also likely to
be tested in the setting of melanoma brain metas- There is limited clinical trial experience regard-
tases. Clinical trials are now often including ing the combination of immunotherapy with
patients with small, asymptomatic, untreated SRS. The bulk of our current knowledge is from
brain metastases. Combination studies are also retrospective analyses, which are often single
underway that focus specifically on enrolling institution in nature. In one example, Knisely
patients with melanoma brain metastases— et al. demonstrated prolonged overall survival
examples include the previously mentioned trial with ipilimumab and SRS in a series of patients
of pembrolizumab and bevacizumab and another at Yale University [41]. In another series, Silk
investigating the use of atezolizumab with beva- et al. evaluated outcomes of melanoma patients
cizumab (NCT03175432). with brain metastasis treated at the University
of Michigan with radiation, some of whom also
received ipilimumab, and survival was simi-
Systemic Therapy Combined larly improved in the ipilimumab-treated group
with Radiation [42]. Additional studies have been published
regarding PD-1 inhibitors, demonstrating
Prior to the advent of stereotactic radiosurgery enhanced activity in combination with SRS
(SRS), patients with brain metastases were [43, 44].
treated with whole brain radiation therapy A phase 1 trial of ipilimumab in combination
(WBRT). The most commonly used WBRT regi- with SRS or WBRT in melanoma brain metasta-
men is 3 Gy daily over 10 days, for a total of sis patients (NCT01703507) has been completed
30 Gy, although various alternative strategies [45]. Patients were assigned to one of two treat-
have been utilized that spare sensitive parts of ment groups (SRS vs WBRT) based on tumor
the brain [36, 37]. This modality is only mini- burden, and the dose of ipilimumab was escalated
mally effective, and the median survival in mela- from 3 mg/kg in the first cohort in each treatment
noma patients treated with WBRT has ranged group to 10 mg/kg. Expected toxicities of ipilim-
from 2 to 4 months [38]. In recent years, SRS is umab were observed—although one patient
being used with increasing frequency. The deliv- experienced a grade 3 neurotoxicity, most were
ery of 22 Gy in a single fraction has proven to be relatively minor. Results regarding efficacy are
highly effective for melanoma, a disease which still pending.
is only modestly sensitive to radiation [39]. The Another clinical trial of pembrolizumab with
incorporation of SRS into standard treatment SRS is currently accruing patients at Emory
paradigms over a decade ago increased median University (NCT02858869). Patients with meta-
overall survival to 8 months, which has fortu- static melanoma or non-small cell lung cancer to
nately improved with the advent of immunother- the brain are being treated with standard pembro-
apy and targeted agents [40]. Initially SRS was lizumab dosing and escalating doses of radiation
reserved for patients with a small number of (6, 9, and 18–21 Gy). The primary endpoint is to
metastases (up to 4), but with improved technol- determine the safety of the combination. A simi-
ogy, it is now feasible to treat a larger number of lar trial assessing SRS with nivolumab is being
lesions [41]. Many institutions now forgo WBRT, conducted at Johns Hopkins University
given limited efficacy and potential neurocogni- (NCT02716948). A study evaluating the combi-
tive toxicities, and patients are often treated nation of SRS, ipilimumab, and nivolumab is
with a combination of systemic therapy and being planned in Australia (NCT03340129),
SRS. However, WBRT is still widely used for building on the high intracranial response rate
diffuse intracranial and/or leptomeningeal observed with this systemic therapy regimen in
disease. prior trials [34, 35].
equencing and Timing of Immune

S Challenges of Systemic Therapy
Therapy and Radiation Therapy for Melanoma Brain Metastasis
The timing of SRS relative to immune therapy esponse Criteria for Brain

R
appears to affect outcome. For example, in an Metastasis
institutional series published by Qian et al.,
administration of CTLA-4 or PD-1 inhibitors Accurately assessing responses to systemic ther-
within 4 weeks of SRS resulted in improved apy can pose a challenge. In general, radiographic
response of melanoma brain metastasis, com- responses to systemic therapy in extra-cerebral
pared to waiting longer to initiate immunother- sites are determined by RECIST criteria, which
apy [46]. Further, preclinical studies have utilize the sum of the largest single dimension of
suggested that concomitant therapy may be target lesions over 1 cm in size. The minimal size
superior to stepwise treatment [47]. In theory, criterion was determined as double the distance
intermittent radiation could increase tumor infil- between slices on CT. Brain lesions are typically
tration by T cells and ultimately improve antitu- imaged by high-resolution MRI with slices that
mor activity by avoiding exhaustion of are 1–2.5 mm apart, allowing for reliable assess-
intra-tumoral T cells. As murine models of mel- ment of differences in smaller lesions; response
anoma brain metastases improve, preclinical criteria for brain metastasis were recently stan-
studies may be able to address the optimal dardized. The international Response Assessment
sequence and timing of combined modality in Neuro-Oncology (RANO) Working Group
therapy. developed response criteria for brain metastases
(RANO-BM) [49]. Given the potential for initial
tumor inflammation in patients on immunother-
Targeted Therapy in Combination apy, the RANO group also developed the iRANO
with Radiation Therapy criteria [50]. Other groups have used modified
RECIST criteria for brain metastasis trials, allow-
Case reports and case series have described ing for lesions ≥5 mm to be used as target
institutional experiences of the combination of lesions—overall, differences between these mod-
BRAF inhibitors with radiation therapy, sum- ified RECIST criteria and the RANO-BM are
marized by Anker et al. [48]. An ongoing clini- minimal, and both allow inclusion of patients
cal trial is studying the activity of dabrafenib with smaller lesions into clinical trials [51]. All
in combination with SRS (NCT01721603). of these criteria require further prospective vali-
The primary endpoint of this study is 6-month dation and standardization.
distant brain metastasis-free survival. A trial of
cobimetinib and vemurafenib after SRS is
ongoing (NCT03430947). Toxicities associ- Radiation Necrosis
ated with the combination of BRAF/MEK
inhibitors in combination with radiation can be Radiation necrosis is becoming an increasingly
severe, leading to the establishment of treat- frequent challenge with the increasing utility of
ment guidelines [48]. Specifically, for brain SRS. Radiation necrosis is a delayed complica-
metastases, the Eastern Cooperative Oncology tion of SRS, occurring months to years after
Group recommends holding BRAF and/or treatment. It appears to occur more frequently in
MEK inhibitors for at least 3 days before and the brain than in other organs. Some studies have
after WBRT and at least 1 day before and after suggested increased incidence in patients treated
SRS. with combination SRS and immune therapy,
either concurrently or sequentially, compared one third of the patients in this series had disease
with the combination of SRS and other forms of progression that was too rapid for any further
systemic therapy [52]. A recent clinical trial of intervention—median survival in this group was
pembrolizumab in patients with untreated brain only 2.9 weeks.
metastases reported radiation necrosis in over Intrathecal drug administration is an interest-
30% of patients [33]. ing alternative approach for patients with LMD,
Radiation necrosis can sometimes be manas it overcomes concerns about limited drug pen-
aged by observation alone in patients with etration into the cerebrospinal fluid. Glitza et al.
asymptomatic lesions that do not grow or regress. recently reported results for over 100 melanoma
However, surgical intervention such as resection patients with LMD who were treated with intra-
or laser interstitial thermocoagulation therapy is thecal interleukin-2 over two decades [63]. The
required in over half of cases for symptom con- 43 cases in this series treated between 2006 and
trol [53–56]. Glioblastoma patients similarly 2014 showed 1-, 2-, and 5-year survival rates of
develop radiation necrosis, which sometimes 36%, 26%, and 13%, respectively, from the start
responds to bevacizumab [57]. This practice is of treatment. However, intrathecal IL-2 adminis-
being adopted in brain metastasis patients as tration is often poorly tolerated due to increased
well, although the true efficacy of bevacizumab intracranial pressure, indicating an urgent need
in either treating or preventing radiation necrosis for newer approaches for LMD patients. Limited
is unknown [58–60]. ongoing clinical trials for melanoma LMD
include NCT03025256 (combined systemic and
intrathecal nivolumab) and NCT02939300 (ipili-
Leptomeningeal Disease mumab and nivolumab, both systemic).
Additional approaches are sorely needed.
Despite the progress in treating parenchymal
brain metastasis, leptomeningeal disease (LMD)
remains a major challenge. The median survival Conclusions
for LMD from melanoma is still dismal [61].
Patients with LMD have almost universally been Improved systemic therapies for metastatic mela-
excluded from clinical trials, and the few small noma have resulted in superior outcomes for
trials that have been conducted for this disease patients with melanoma brain metastases.
population have not yielded promising results. Increased enrollment of patients with brain
Standard treatment includes supportive care, metastases in clinical trials has enhanced our
clinical trials when available or whole brain radi- understanding of the efficacy of systemic thera-
ation therapy for symptom palliation. Given the pies in the CNS. Brain metastasis response rates
lack of prospective randomized data for treating to immune therapies are similar to response rates
these patients, a number of groups have pub- in other advanced (stage M1C) melanoma
lished their institutional experience suggesting patients, while responses to targeted therapies in
that contemporary targeted or immune therapies the brain appear to be shorter than in extracranial
may improve survival with LMD. For example, sites. Further investigation is necessary to deter-
one retrospective analysis of 39 metastatic mela- mine whether this reflects limited drug penetra-
noma patients with LMD reported a median sur- tion into the CNS or other local factors. SRS
vival of 21 weeks in 21 patients treated with remains an important component of treatment of
targeted/immune therapy and radiotherapy, com- brain metastases, as response rates to all current
pared to 4.3 weeks in patients treated with radia- regimens are <60%, and large lesions or lesions
tion therapy alone [62]. Of note, approximately in neurologically sensitive sites such as the brain
stem, speech area or motor strip may require 10. Richard MA, Grob JJ, Zarrour H, Basseres N, Bizzari
local intervention. The combination of SRS with JP, Gerard B, et al. Combined treatment with dacar-
bazine, cisplatin, fotemustine and tamoxifen in
immune therapy may improve outcomes but may metastatic malignant melanoma. Melanoma Res.
also be associated with increased risk of radiation 1998;8:170–4.
necrosis, a phenomenon that is rarely seen out- 11. Middleton MR, Grob JJ, Aaronson N, Fierlbeck G,
side the brain. Patients with leptomeningeal dis- Tilgen W, Seiter S, et al. Randomized phase III study
of temozolomide versus dacarbazine in the treatment
ease have very limited options with poor of patients with advanced metastatic malignant mela-
outcomes, highlighting the need for further clini- noma. J Clin Oncol. 2000;18:158–66.
cal trials in this population. In conclusion, while 12. Agarwala SS, Kirkwood JM, Gore M, Dreno B,

outcomes of patients with melanoma brain Thatcher N, Czarnetski B, et al. Temozolomide for
the treatment of brain metastases associated with
metastasis have significantly improved, addi- metastatic melanoma: a phase II study. J Clin Oncol.
tional preclinical and clinical studies are required 2004;22:2101–7.
to increase survival while minimizing CNS 13. Jacquillat C, Khayat D, Banzet P, Weil M, Fumoleau
toxicity. P, Avril MF, et al. Final report of the French multi-
center phase II study of the nitrosourea fotemustine
in 153 evaluable patients with disseminated malignant
melanoma including patients with cerebral metasta-
References ses. Cancer. 1990;66:1873–8.
14. Mornex F, Thomas L, Mohr P, Hauschild A, Delaunay
1. Nayak L, Lee EQ, Wen PY. Epidemiology of brain MM, Lesimple T, et al. Randomised phase III trial of
metastases. Curr Oncol Rep. 2012;14:48–54. fotemustine versus fotemustine plus whole brain irra-
2. Zhang D, Wang Z, Shang D, Yu J, Yuan S. Incidence diation in cerebral metastases of melanoma. Cancer
and prognosis of brain metastases in cutaneous mela- Radiother. 2003;7:1–8.
noma patients: a population-based study. Melanoma 15. Ascierto PA, Kirkwood JM, Grob JJ, Simeone E,

Res. 2018;29:77. Grimaldi AM, Maio M, et al. The role of BRAF V600
3. Davies MA, Liu P, McIntyre S, Kim KB, Papadopoulos mutation in melanoma. J Transl Med. 2012;10:85.
N, Hwu WJ, et al. Prognostic factors for survival in 16. Capper D, Berghoff AS, Magerle M, Ilhan A, Wohrer
melanoma patients with brain metastases. Cancer. A, Hackl M, et al. Immunohistochemical testing of
2011;117:1687–96. BRAF V600E status in 1,120 tumor tissue samples
4. Budman DR, Camacho E, Wittes RE. The current of patients with brain metastases. Acta Neuropathol.
causes of death in patients with malignant melanoma. 2012;123:223–33.
Eur J Cancer. 1978;14:327–30. 17. Chen G, Chakravarti N, Aardalen K, Lazar AJ, Tetzlaff
5. Gardner LJ, Ward M, Andtbacka RHI, Boucher KM, MT, Wubbenhorst B, et al. Molecular profiling of
Bowen GM, Bowles TL, et al. Risk factors for devel- patient-matched brain and extracranial melanoma
opment of melanoma brain metastasis and disease metastases implicates the PI3K pathway as a thera-
progression: a single-center retrospective analysis. peutic target. Clin Cancer Res. 2014;20:5537–46.
Melanoma Res. 2017;27:477–84. 18. Colombino M, Capone M, Lissia A, Cossu A, Rubino
6. Tio M, Wang X, Carlino MS, Shivalingam B, Fogarty C, De Giorgi V, et al. BRAF/NRAS mutation frequen-
GB, Guminski AD, et al. Survival and prognostic cies among primary tumors and metastases in patients
factors for patients with melanoma brain metastases with melanoma. J Clin Oncol. 2012;30:2522–9.
in the era of modern systemic therapy. Pigment Cell 19. Davies MA, Stemke-Hale K, Lin E, Tellez C, Deng
Melanoma Res. 2018;31:509–15. W, Gopal YN, et al. Integrated molecular and clinical
7. Fife KM, Colman MH, Stevens GN, Firth IC, Moon analysis of AKT activation in metastatic melanoma.
D, Shannon KF, et al. Determinants of outcome in Clin Cancer Res. 2009;15:7538–46.
melanoma patients with cerebral metastases. J Clin 20. Niessner H, Forschner A, Klumpp B, Honegger JB,
Oncol. 2004;22:1293–300. Witte M, Bornemann A, et al. Targeting hyperactiva-
8. Kaba SE, Kyritsis AP, Hess K, Yung WK, Mercier R, tion of the AKT survival pathway to overcome ther-
Dakhil S, et al. TPDC-FuHu chemotherapy for the apy resistance of melanoma brain metastases. Cancer
treatment of recurrent metastatic brain tumors. J Clin Med. 2013;2:76–85.
Oncol. 1997;15:1063–70. 21. Xie TX, Huang FJ, Aldape KD, Kang SH, Liu M,
9. Franciosi V, Cocconi G, Michiara M, Di Costanzo F, Gershenwald JE, et al. Activation of stat3 in human
Fosser V, Tonato M, et al. Front-line chemotherapy melanoma promotes brain metastasis. Cancer Res.
with cisplatin and etoposide for patients with brain 2006;66:3188–96.
metastases from breast carcinoma, nonsmall cell lung 22. Huang FJ, Steeg PS, Price JE, Chiu WT, Chou PC,
carcinoma, or malignant melanoma: a prospective Xie K, et al. Molecular basis for the critical role of
study. Cancer. 1999;85:1599–605. suppressor of cytokine signaling-1 in melanoma brain
metastasis. Cancer Res. 2008;68:9634–42.
23. Dummer R, Goldinger SM, Turtschi CP, Eggmann NB, 35. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS,
Michielin O, Mitchell L, et al. Vemurafenib in patients Moschos SJ, et al. Combined nivolumab and ipilim-
with BRAF(V600) mutation-positive melanoma with umab in melanoma metastatic to the brain. N Engl J
symptomatic brain metastases: final results of an open- Med. 2018;379:722–30.
label pilot study. Eur. J. Cancer. 2014;50:611–21. 36. Tsao MN, Xu W, Wong RK, Lloyd N, Laperriere N,
Sahgal A, et al. Whole brain radiotherapy for the treat-
Ascierto PA, Chapman PB, et al. Dabrafenib in ment of newly diagnosed multiple brain metastases.
patients with Val600Glu or Val600Lys BRAF-mutant Cochrane Database Syst Rev. 2018;1:CD003869.
melanoma metastatic to the brain (BREAK-MB): a 37. Fan XW, Wang JQ, Wu JL, Wang HB, Wu

multicentre, open-label, phase 2 trial. Lancet Oncol. KL. Simultaneously avoiding the hippocampus
2012;13:1087–95. and hypothalamic-pituitary axis during whole
25. Long GV, Flaherty KT, Stroyakovskiy D, Gogas H, brain radiotherapy: a planning study. Med Dosim.
Levchenko E, de Braud F, et al. Dabrafenib plus tra- 2018;44(2):130–5.
metinib versus dabrafenib monotherapy in patients 38. Flanigan JC, Jilaveanu LB, Faries M, Sznol M,

with metastatic BRAF V600E/K-mutant melanoma: Ariyan S, Yu JB, et al. Melanoma brain metastases:
long-term survival and safety analysis of a phase 3 is it time to reassess the bias? Curr Probl Cancer.
study. Ann Oncol. 2017;28:1631–9. 2011;35:200–10.
26. Davies MA, Saiag P, Robert C, Grob JJ, Flaherty KT, 39. Redmond AJ, Diluna ML, Hebert R, Moliterno JA,
Arance A, et al. Dabrafenib plus trametinib in patients Desai R, Knisely JP, et al. Gamma knife surgery for
with BRAF(V600)-mutant melanoma brain metasta- the treatment of melanoma metastases: the effect of
ses (COMBI-MB): a multicentre, multicohort, open- intratumoral hemorrhage on survival. J Neurosurg.
label, phase 2 trial. Lancet Oncol. 2017;18:863–73. 2008;109(Suppl):99–105.
27. Robert C, Karaszewska B, Schachter J, Rutkowski P, 40. DiLuna ML, King JT Jr, Knisely JP, Chiang

Mackiewicz A, Stroiakovski D, et al. Improved over- VL. Prognostic factors for survival after stereotactic
all survival in melanoma with combined dabrafenib radiosurgery vary with the number of cerebral metas-
and trametinib. N Engl J Med. 2014;372:30–9. tases. Cancer. 2007;109:135–45.
28. Glitza Oliva IC, Schvartsman G, Tawbi H. Advances 41. Raldow AC, Chiang VL, Knisely JP, Yu JB. Survival
in the systemic treatment of melanoma brain metasta- and intracranial control of patients with 5 or more
ses. Ann Oncol. 2018;29:1509–20. brain metastases treated with gamma knife stereotac-
29. Seifert H, Hirata E, Gore M, Khabra K, Messiou C, tic radiosurgery. Am J Clin Oncol. 2013;36:486–90.
Larkin J, et al. Extrinsic factors can mediate resis- 42. Silk AW, Bassetti MF, West BT, Tsien CI, Lao

tance to BRAF inhibition in central nervous system CD. Ipilimumab and radiation therapy for melanoma
melanoma metastases. Pigment Cell Melanoma Res. brain metastases. Cancer Med. 2013;2:899–906.
2016;29:92–100. 43. Anderson ES, Postow MA, Wolchok JD, Young

30. Weber JS, Amin A, Minor D, Siegel J, Berman D, RJ, Ballangrud A, Chan TA, et al. Melanoma brain
O'Day SJ. Safety and clinical activity of ipilimumab in metastases treated with stereotactic radiosurgery and
melanoma patients with brain metastases: retrospec- concurrent pembrolizumab display marked regres-
tive analysis of data from a phase 2 trial. Melanoma sion; efficacy and safety of combined treatment. J
Res. 2011;21:530–4. Immunother Cancer. 2017;5:76.
31. Margolin K, Ernstoff MS, Hamid O, Lawrence D, 44. Liniker E, Menzies AM, Kong BY, Cooper A,

McDermott D, Puzanov I, et al. Ipilimumab in patients Ramanujam S, Lo S, et al. Activity and safety of
with melanoma and brain metastases: an open-label, radiotherapy with anti-PD-1 drug therapy in patients
phase 2 trial. Lancet Oncol. 2012;13:459–65. with metastatic melanoma. Oncoimmunology.
32. Goldberg SB, Gettinger SN, Mahajan A, Chiang
2016;5:e1214788.
AC, Herbst RS, Sznol M, et al. Pembrolizumab for 45. Williams NL, Wuthrick EJ, Kim H, Palmer JD, Garg
patients with melanoma or non-small-cell lung can- S, Eldredge-Hindy H, et al. Phase 1 study of ipilim-
cer and untreated brain metastases: early analysis of umab combined with whole brain radiation therapy or
a non-randomised, open-label, phase 2 trial. Lancet radiosurgery for melanoma patients with brain metas-
Oncol. 2016;17:976–83. tases. Int J Radiat Oncol Biol Phys. 2017;99:22–30.
33. Kluger HM, Chiang V, Mahajan A, Zito CR, Sznol 46. Qian JM, Yu JB, Kluger HM, Chiang VL. Timing
M, Tran T, et al. Long-term survival of patients and type of immune checkpoint therapy affect the
with melanoma with active brain metastases treated early radiographic response of melanoma brain
with pembrolizumab on a phase II trial. J Clin metastases to stereotactic radiosurgery. Cancer.
Oncol. 2018;37(1):52–60. https://doi.org/10.1200/ 2016;122:3051–8.
JCO.18.00204. 47.
Dewan MZ, Galloway AE, Kawashima N,
34. Long GV, Atkinson V, Lo S, Sandhu S, Guminski AD, Dewyngaert JK, Babb JS, Formenti SC, et al.
Brown MP, et al. Combination nivolumab and ipilim- Fractionated but not single-dose radiotherapy induces
umab or nivolumab alone in melanoma brain metas- an immune-mediated abscopal effect when com-
tases: a multicentre randomised phase 2 study. Lancet bined with anti-CTLA-4 antibody. Clin Cancer Res.
Oncol. 2018;19:672–81. 2009;15:5379–88.
48. Anker CJ, Grossmann KF, Atkins MB, Suneja G, 56. Lu AY, Turban JL, Damisah EC, Li J, Alomari AK, Eid
Tarhini AA, Kirkwood JM. Avoiding severe toxic- T, et al. Novel biomarker identification using metabo-
ity from combined BRAF inhibitor and radiation lomic profiling to differentiate radiation necrosis and
treatment: consensus guidelines from the Eastern recurrent tumor following gamma knife radiosurgery.
Cooperative Oncology Group (ECOG). Int J Radiat J Neurosurg. 2017;127:388–96.
Oncol Biol Phys. 2016;95:632–46. 57. Tye K, Engelhard HH, Slavin KV, Nicholas MK,

49. Lin NU, Lee EQ, Aoyama H, Barani IJ, Barboriak Chmura SJ, Kwok Y, et al. An analysis of radiation
DP, Baumert BG, et al. Response assessment criteria necrosis of the central nervous system treated with
for brain metastases: proposal from the RANO group. bevacizumab. J Neuro-Oncol. 2014;117:321–7.
Lancet Oncol. 2015;16:e270–8. 58. Remon J, Le Pechoux C, Caramella C, Dhermain F,
50. Okada H, Weller M, Huang R, Finocchiaro G,
Louvel G, Soria JC, et al. Brain radionecrosis treated
Gilbert MR, Wick W, et al. Immunotherapy with bevacizumab in a patient with resected squa-
response assessment in neuro-oncology: a report mous cell carcinoma of the lung. J Thorac Oncol.
of the RANO working group. Lancet Oncol. 2017;12:e1–3.
2015;16:e534–e42. 59. Delishaj D, Ursino S, Pasqualetti F, Pesaresi I,

51. Qian JM, Mahajan A, Yu JB, Tsiouris AJ, Goldberg Desideri I, Cosottini M, et al. The effectiveness of
SB, Kluger HM, et al. Comparing available criteria bevacizumab in radionecrosis after radiosurgery of a
for measuring brain metastasis response to immuno- single brain metastasis. Rare Tumors. 2015;7:6018.
therapy. J Neuro-Oncol. 2017;132:479–85. 60. Boothe D, Young R, Yamada Y, Prager A, Chan T,
52. Colaco RJ, Martin P, Kluger HM, Yu JB, Chiang Beal K. Bevacizumab as a treatment for radiation
VL. Does immunotherapy increase the rate of radia- necrosis of brain metastases post stereotactic radio-
tion necrosis after radiosurgical treatment of brain surgery. Neuro-Oncology. 2013;15:1257–63.
metastases? J Neurosurg. 2016;125:17–23. 61. Cohen JV, Tawbi H, Margolin KA, Amravadi R,

53. Rahmathulla G, Recinos PF, Valerio JE, Chao S,
Bosenberg M, Brastianos PK, et al. Melanoma central
Barnett GH. Laser interstitial thermal therapy for nervous system metastases: current approaches, chal-
focal cerebral radiation necrosis: a case report lenges, and opportunities. Pigment Cell Melanoma
and literature review. Stereotact Funct Neurosurg. Res. 2016;29:627–42.
2012;90:192–200. 62. Geukes Foppen MH, Brandsma D, Blank CU, van
54. Alomari A, Rauch PJ, Orsaria M, Minja FJ, Chiang Thienen JV, Haanen JB, Boogerd W. Targeted treat-
VL, Vortmeyer AO. Radiologic and histologic con- ment and immunotherapy in leptomeningeal metasta-
sequences of radiosurgery for brain tumors. J Neuro- ses from melanoma. Ann Oncol. 2016;27:1138–42.
Oncol. 2014;117:33–42. 63. Glitza IC, Rohlfs M, Guha-Thakurta N, Bassett RL
55. Nath SK, Sheridan AD, Rauch PJ, Yu JB, Minja FJ, Jr, Bernatchez C, Diab A, et al. Retrospective review
Vortmeyer AO, et al. Significance of histology in of metastatic melanoma patients with leptomeningeal
determining management of lesions regrowing after disease treated with intrathecal interleukin-2. ESMO
radiosurgery. J Neuro-Oncol. 2014;117:303–10. Open. 2018;3:e000283.
Systemic Therapy for Brain
Metastases in Other Primary
17
Cancers (Genitourinary,
Gastrointestinal, Gynecology,
Head/Neck)
Karishma M. Parikh and Rajiv S. Magge
Introduction With our rapidly growing understanding of

genetics and molecular phenotypes, novel sys-
Brain metastasis (BM) is most common in lung temic options are now available. Immunotherapy,
and breast cancers and melanomas; BM in these such as checkpoint inhibitors, has shown exciting
malignancies tends to make up 75% of all brain efficacy in specific malignancies. Interestingly,
metastases [1, 2]. Metastatic disease from renal some agents with minimal CNS penetration may
cell carcinoma and colorectal carcinoma makes still play a role in both preventing and treating
up a large portion of the remainder. BM [6].
In most cases, there is no standardized treat-
ment for BM; the treatment approach for BM
tends to be customized for each patient and often Gastrointestinal Cancers
involves multiple modalities of treatment [3, 4].
Patients with isolated, symptomatic, and acces- Colorectal
sible lesions are usually offered surgical resec-
tion, while stereotactic radiosurgery (SRS) and Gastrointestinal malignancies are some of the
whole brain radiation therapy (WBRT) are used most common in the United States, with colorec-
in the setting of multiple lesions. tal cancer (CRC) being the most frequent type
Systemic therapy for BM has been challeng- [8]. Colorectal cancer (CRC) is the third most
ing due to limited CNS penetration of traditional common cancer worldwide [9] and third most
chemotherapy—if drugs are only effective with common cause of cancer death in the United
non-CNS disease, the brain may become a sanc- States [8], with improving survival due to better
tuary site for metastases [5–7]. Efficacy of sys- screening and available treatment options [10].
temic therapy is dependent on both the intrinsic About 5% of CRC patients have an underlying
sensitivity of the primary malignancy and the genetic disorder that predisposes them to the
ability of the agent to pass the blood-brain barrier development of colon cancer, such as familial
(BBB), which can have heterogeneous break- adenomatous polyposis (FAP) and Lynch syn-
down within lesions [6]. drome (hereditary nonpolyposis colorectal can-
cer (HNPCC)) [11–13]. They are both associated
with the development of brain tumors—typically,
K. M. Parikh · R. S. Magge (*) patients with FAP can develop medulloblasto-
Department of Neurology, NewYork-Presbyterian/ mas, while those with Lynch syndrome have an
Weill Cornell Medicine, New York, NY, USA increased risk of glioma [14–16].
e-mail: ram9116@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_17
246 K. M. Parikh and R. S. Magge
About 20% of patients diagnosed with CRC

have metastatic disease at baseline; 25% of
patients have systemic metastases during the dis-
ease course, most commonly to the liver, lung,
and lymphatic tissue. BM occurs in approxi-
mately 1–4% of patients [17], often concomitant
with other metastases [18]. Cerebellar lesions are
seen in up to 33–55% of the patients [19]. Patients
often present with headache, hemiparesis, dizzi-
ness, ataxia, and/or seizures [20].
Curative treatment for localized CRC is usu-
ally surgical resection. The mainstay for meta-
static CRC (mCRC) treatment is
generally cytotoxic chemotherapy, including iri-
notecan or oxaliplatin, combined with 5-FU and
leucovorin or capecitabine [21]. These drugs are
generally considered ineffective for BMs as they
have limited CNS penetration [22].
Outcomes have improved with the addition of
targeted therapy with monoclonal antibodies
(mAbs) against vascular endothelial growth factor
(VEGF) and epidermal growth factor receptor
Fig. 17.1 T1-weighted post-contrast MRI demonstrating
(EGFR). EGFR mAbs cetuximab and panitu- brain metastases from colon adenocarcinoma. Post-
mumab have been approved for use in patients contrast T1 MRI image of brain metastases in a 54-year-
with RAS wild-type tumors. For RAS-mutant dis- old patient with metastatic colon adenocarcinoma
ease, the VEGF mAb bevacizumab, anti-VEGF
receptor 2 (VEGFR2) mAb ramucirumab, recom- with neurological symptoms at the time of pre-
binant fusion protein ziv-aflibercept, and multiki- sentation [37, 38]; some report that BM has been
nase inhibitor regorafenib have been approved associated with poorer prognosis than BM from
[21]. RAS mutations have been associated with other solid tumors [39]. No clear distribution pat-
increased incidence of brain and lung metastases tern for brain lesions has been found, with
[23–27]. There are no established associations patients presenting with both supratentorial and
between BM and alterations in PIK3CA, BRAF, infratentorial metastases [36]. Spread to the brain
EGFR, and CXCR4 or changes in tumor markers may occur via Batson’s vertebral venous plexus,
CA19.9 and CEA [23–25, 28–33] (Fig. 17.1). which allows vascular communication between
the brain and esophagus. It was traditionally
thought that brain metastases were more com-
Esophageal mon with adenocarcinoma histology; one study
[40] found adenocarcinoma as the primary histol-
Esophageal cancer is the eighth most common ogy for all BM in their series. However, a later
cancer worldwide and the sixth leading cause of 2017 study by Welch et al. observed no statisti-
cancer death due to its aggressiveness [34]. cally significant differences between the adeno-
Approximately 40% of patients with esophageal carcinoma and squamous cell carcinoma groups,
cancer present with metastatic disease at diagno- supporting no difference in neurotropism [36].
sis, usually to the lymph nodes, liver, peritoneum, On average, median survival for these patients
lung, and adrenal glands [8]. with BM was 3.8–10.5 months [37, 41].
BM is rare, with incidence ranging from 0% to Due to the aggressive nature of esophageal
6% [35, 36], with 75–80% of patients presenting cancer, most patients receive trimodality therapy
17 Systemic Therapy for Brain Metastases in Other Primary Cancers (Genitourinary, Gastrointestinal… 247
with chemotherapy (carboplatin, paclitaxel, and gastric cancer was associated with VEGF expres-
5-FU) alongside radiation therapy and surgical sion, and based on a preclinical model proposed
resection when indicated [42]. Carboplatin has that reducing VEGF expression may decrease
incomplete CNS penetration, but there is no evi- metastatic capacity, by using metformin to reduce
dence yet of clear efficacy with esophageal BM. VEGF expression and blocking epithelial to mes-
Abu et al. retrospectively looked at 142 cases enchymal transformation.
of esophageal cancer over a 10-year period and
found that HER2 overexpression correlated with
postoperative BM [43]. Similarly, Preusser et al. Hepatocellular Carcinoma
described 21 patients with esophageal BM with
good concordance of HER2 and EGFR expres- Hepatocellular carcinoma is the most common
sion between the primary tumor and brain metas- primary hepatic malignancy worldwide with
tasis; however, unlike the earlier study, HER2 increasing prevalence [52]. The dominant risk fac-
positivity did not seem to increase risk of BM tor for HCC in North America is hepatitis C-related
[44]. Ongoing investigation is required to deter- cirrhosis, while in Africa and Asia, HCC incidence
mine the role of HER2 [41]. is associated with hepatitis B infection [53].
HCC has a low affinity for the CNS, with stud-
ies indicating an incidence of around 1% (range
Gastric of 0.2–2.2%) [54]. Reports mostly describe a
solitary intracranial metastasis to the parietal or
Gastric cancer is the second most common cause frontal lobes [55]. In addition to hepatic encepha-
of cancer-related death worldwide [45]. lopathy, patients may develop intracranial hyper-
Adenocarcinoma is the most common subtype, tension and focal neurologic symptoms [56]. Of
and surgical resection is quite effective for early significant concern is that a reported 70% of
stage cancer. Postoperative chemoradiation is HCC BM is associated with intracerebral hemor-
often considered for patients with at least stage rhage [55]. In one study, patients with a single
IB disease. There are multiple regimens utilized BM, Child-Pugh grade A, had the best prognosis
worldwide based on heterogeneous populations; with median overall survival of 27 weeks [57].
in the United States, docetaxel tends to be the As is the case with other primary tumors,
drug of choice added to a regimen with cisplatin WBRT, SRS, and surgical resection are com-
and 5-fluorouracil [16, 46]. monly used, with improvements in survival noted
With the rare incidence of less than 1% (range: with combination of both surgery and RT [57,
0.16–0.69%), there are only a few reported stud- 58]. Sorafenib, an oral multi-kinase inhibitor that
ies characterizing gastric BM [47]. In one study, induces tumor stasis and inhibits tumor angio-
York et al. described 0.7% of their patients to genesis, has been theorized to reduce intracranial
have BM with all of them having concomitant disease, but its use has been limited by concern
systemic metastatic disease, with a median sur- that it may increase risk of intracranial hemor-
vival of 2.4 months [48]. The usual clinical pre- rhage [59, 60]. Targeted agents often only pro-
sentation tends to be headache, muscular vide partial inhibition of a signaling pathway, so
weakness, and visual difficulties [20]. combinatory regimens may be necessary [61–
Lemke et al. reported that surgical resection 65]. Unfortunately, there is no evidence yet to
provided the best chance of improved survival in support systemic therapy for HCC BM.
these patients [49]. In a study by Kasakura et al.,
11 of 2322 Japanese patients with gastric cancer
had BM (0.47%); patients that received both sur- Pancreatic Cancer
gical resection and WBRT lived longer
than patients who had surgery or WBRT alone Pancreatic cancer is one of the most lethal cancers
[50]. Jun et al. [51] found that BM from advanced with a 5-year survival of <5% [9, 66, 67], in large
part due to its diagnosis at an advanced disease hol exposure and separately with a distinct patho-
stage [68]. The incidence of BM in pancreatic can- physiology to infection by the human
cer remains poorly understood and is extremely papillomavirus (HPV). Overall, HNSCC is an
rare (0.33–0.57%) [39, 69]. No clear brain regional aggressive epithelial malignancy associated with
preference for metastases has been described [20]. lymph node metastasis and immunosuppression
Gemcitabine monotherapy is often used as the [75, 76].
first-line treatment for resected disease and has BM for patients with head and neck malignan-
only limited BBB penetration. Other regimens cies is extremely rare [1, 77]; there are primarily
include FOLFIRINOX (fluorouracil, folinic acid only case reports in setting of an untreated pri-
(leucovorin), irinotecan, and oxaliplatin) and gem- mary cancer.
citabine plus nanoparticle albumin-bound pacli- The treatment for early stage HNSCC is usu-
taxel (nab-paclitaxel) and similarly do not have ally surgery or radiotherapy; localized advanced
any clear efficacy in the brain [70]. Lemke et al. HNSCC often requires combination regimens,
described extended survival after surgical resec- such as surgery followed by postoperative radia-
tion of solitary brain metastases in two post-pan- tion and/or chemoradiation, including with cis-
createctomy patients [49]. platin or sequential induction chemotherapy.
Immunotherapy agents such as pembroli-
zumab or nivolumab, humanized monoclonal
Gallbladder Cancer antibodies targeting PD-1, have been approved
by the FDA for platinum-refractory recurrent/
Gallbladder cancer is rare but rapidly fatal with metastatic HNSCC. There is no data to support
about 5000 cases diagnosed annually in the United their efficacy in patients with BM; however, some
States. Most cancer is primarily adenocarcinoma brain lesions from primary melanoma and lung
and most commonly linked to chronic gallbladder cancer do respond to systemic immunotherapy,
inflammation due to gallstone, gallbladder polyps, indicating potential utility in the future. Several
and chronic infection. Initial symptoms can be clinical trials evaluating other immune check-
nonspecific, contributing to late diagnosis and sub- point inhibitors are ongoing including for HPV-
sequent treatment difficulty [71]. associated HNSCC, which tends to be more
BM from primary gallbladder carcinoma is immunogenic and responsive [78].
extremely rare, with an incidence of <0.5% [72]. The rare case of brain metastasis from head
Surgical resection offers the best chance of cure in and neck malignancy is usually treated with
patients with localized gallbladder cancer. There is SRS. One study by Patel et al. reported similar
no definitive standard regimen for adjuvant or pal- outcomes to other cancers with SRS for BM from
liative chemotherapy for gallbladder cancer, but head and neck carcinomas, without the neurotox-
gemcitabine has been used in most adjuvant/palli- icity seen with WBRT [79].
ative regimens [73]. As noted above, gemcitabine
has only limited BBB permeability. Few cases
have been reported, and subsequently no clear effi- Paraganglioma (Carotid Body Tumor)
cacy of systemic therapy has been established.
Paragangliomas of the head and neck are rare
vascular neuroendocrine tumors derived from the
Head and Neck paraganglia tissues originating from the neural
crest, comprising 0.6% of head and neck tumors
Squamous Cell Carcinoma [80]. Up to 40% of paragangliomas are heredi-
tary, and there are well-known tumor syndromes
Head and neck squamous cell carcinoma associated with the same, including multiple
(HNSCC) is the sixth leading cancer worldwide endocrine neoplasia type 2 (MEN 2), von Hippel-
[74], found to be linked to both tobacco and alco- Lindau (VHL), and neurofibromatosis (NF-1).
Paragangliomas can also occur in familial forms, 90]. The most common presentation is unfortu-
which tend to present at a younger age and at nately intracerebral hemorrhage; metastatic
multiple sites compared to sporadic paraganglio- choriocarcinoma should be on the differential for
mas. Paragangliomas are associated with PGL1 any woman of reproductive age with a new hem-
genes, with a mutation of the SDHB protein orrhagic brain lesion [90]. Most CNS metastases
being involved in head and neck paragangliomas are discovered alongside concurrent lung involve-
[81]. Carotid paraganglioma or carotid body ment—a chest CT should be routinely performed
tumors (CBTs) represent 60–70% of paragangli- as part of the workup [88].
omas of the head and neck [82]. Preferred method The use of systemic chemotherapy for meta-
of treatment and management of CBT involves static choriocarcinoma is a well-accepted prac-
surgical excision, often times difficult due to size tice, particularly with EMA-CO (etoposide,
and vascular involvement of these tumors, with a methotrexate, and actinomycin-D, alternating
high risk of cranial nerve damage and resultant weekly with cyclophosphamide and vincristine).
neurological dysfunction [83]. Methotrexate dosing in this regimen is lower than
Paragangliomas may arise at the skull base commonly used for other types of CNS malig-
with local invasion and involvement of cranial nancy but higher than established for other meta-
nerves, but brain parenchymal metastasis is very static sites in choriocarcinoma. Patients with high
uncommon. Wang et al. [84] described a 53-year- burden of CNS disease may receive low-dose
old woman with right limb weakness associated etoposide and cisplatin before the EMA-CO regi-
with dizziness and vomiting who presented with men, even though they have limited CNS pene-
intracranial metastasis from a carotid body para- tration [91]; in one study employing EMA-CO
ganglioma, one of few cases in the literature. She along with EMA-EP by Savage et al., 85% of the
recovered without reported neurological deficits 27 patients with BM had an overall cure [91],
following surgical resection of the brain tumor. with previous earlier and smaller studies having
cure rates ranging from 35% to 100% [89, 90,
92–95]. Additionally, utilizing the presence of
GYN multidrug-resistant-associated protein 1 (MRP1)
in the uterus and choroid plexus epithelium, one
Choriocarcinoma study using triple-knockout and double-knockout
mice for MRP1 was able to enhance the delivery
Choriocarcinoma is the most aggressive type of of etoposide tenfold through the BBB with lack
gestational trophoblastic disease (GTD), arising of the MRP1 protein; the utility and optimization
from the placental trophoblastic tissue after fer- of drug delivery through the CSF in this regard
tilization. It usually occurs after a molar preg- are being investigated [96, 97].
nancy but may present prior to a full-term/ectopic There is no established management with sur-
pregnancies or abortion [85]. Worldwide inci- gery, radiotherapy, or intrathecal chemotherapy
dence is variable mostly because of differences in for BM in choriocarcinoma; the benefit of WBRT
reporting and diagnostic criteria. The incidence and chemotherapy is not clearly established.
in the United States is 2–7 per 100,000 pregnan-
cies [86]. Choriocarcinoma is quite aggressive,
notably with rapid vascular invasion and diffuse Ovarian/Fallopian Tube Cancer
systemic metastases.
CNS metastases are seen in up to 40% of cho- Ovarian cancer is the second most common
riocarcinoma patients [87]. Choriocarcinoma gynecological cancer after endometrial cancer
BM often causes elevated intracranial pressure and a leading cause of mortality in women [98].
contributing to subsequent headache, vision Globally, it is the seventh most common cancer
changes, nausea, vomiting, tinnitus, hemiparesis, and the eighth most common cause of cancer
and seizures, especially with cortical lesions [88– death among women [99, 100].
BM is a rare and late manifestation of ovar- therapy if the disease is more limited; however, in
ian cancer, with an incidence ranging from most cases, patients will receive a platinum-
0.3% to 1.2% [101]. According to a review by based agent like cisplatin with bevacizumab, an
Pakneshan et al., most patients with brain anti-VEGF monoclonal antibody that acts as a
metastases present with sensory/motor distur- tumor angiogenesis inhibitor.
bances, ataxia, seizures, and altered conscious- There is no established chemotherapy regi-
ness; the cerebellum was the common site of men for patients with BM. There is no standard-
parenchymal metastasis. Most patients diag- ized treatment for BM from cervical cancer, and
nosed with BM have stage III or stage IV can- most patients will undergo resection, SRS, and/or
cer when diagnosed, but BM can occur both in WBRT [109].
the setting of disseminated or isolated disease;
30–44% of patients in one study were reported
to have isolated CNS relapse [100, 101]. Endometrial
Patients with concurrent extracranial disease
had a median overall survival of 9 months com- Endometrial carcinoma is the most common
pared to 21 months in patients with isolated gynecological malignancy in the United States
CNS metastases [102]. and the fourth most common malignancy in
Patients with high-grade (stage 3 or more) dis- women, with an overall increasing incidence due
ease usually receive adjuvant systemic chemo- to improved survival as well as higher rates of
therapy consisting of a platinum (carboplatin or obesity [74, 110, 111].
cisplatin) and taxane (paclitaxel or docetaxel). BM is exceedingly rare with primary endome-
BBB penetration of these drugs, especially tax- trial cancers, with incidence of 0.3–0.9% of
anes, is limited, and thus significant CNS activity patients based on case reports [112, 113]. The
may not be expected. cancer may first metastasize to the lungs and sub-
Most cases of ovarian BM are treated with sequently disseminate to the CNS hematoge-
surgical resection and/or SRS. In one study by nously; papillary serous, clear cell, and poorly
Niu et al., Gamma Knife radiotherapy and surgi- differentiated histologic subtypes carry higher
cal excision contributed to extended survival risk of BM [112, 114]. Cybulska et al. described
[103]. Another study by Kwon et al. found sig- 23 of the total 3052 patients who developed BM
nificantly prolonged survival after surgical resec- in the setting of low-grade endometrial carci-
tion for single or symptomatic BM [104]. There noma [113]. No specific neurological symptoms
are no novel systemic treatments identified for or predilection of certain areas in the brain have
the use of BM in ovarian cancer. been described.
Total hysterectomy with bilateral salpingo-
oophorectomy is the primary treatment for
Cervical patients with endometrial cancer followed by
XRT or chemotherapy. The combination of
Cervical cancer is the third most common cancer cisplatin, doxorubicin, and paclitaxel is the
in women and the fourth major cause of mortality most active regimen for advanced or recurrent
in women worldwide. endometrial cancer [115]. In one retrospective
The development of BM is rare in uterine cer- study from Gien et al., 8 of 1295 women
vical cancer, with an incidence of 0.4–1.2% [105, developed BM; brain involvement was diag-
106]. Most brain metastases tend to be supraten- nosed an average of 2 months following com-
torial without a propensity for a specific lobe pletion of primary tumor treatment [114],
[107]. Median survival is 2.3–8 months [105, indicating likely poor efficacy of the chemo-
106, 108]. therapy regimen for brain disease. No effec-
Metastatic cervical cancer can be managed tive systemic treatment for endometrial BM
with the use of recurrent surgery or radiation has been fully identified [116].
Genitourinary Cancers There have been ten FDA-approved systemic

therapies in the last decade for the treatment of
Prostate mRCC, including agents that target vascular
endothelial growth factor receptor (VEGFR),
Prostate cancer is the second most common can- mammalian target of rapamycin (mTOR), and
cer in men after lung cancer and the third leading immune checkpoint inhibitors [138]. Of the
cause of malignancy in the United States [117]. approved VEGFR tyrosine kinase inhibitors
BM is extremely uncommon with an inci- (TKI), most have not shown promising responses
dence of 0.2–2.0% [118–122]—most are soli- for BM. One such drug, sunitinib, did not show
tary metastases that are usually supratentorial. significant efficacy in patients with BM in RCC
BMs are present with concurrent osseous, lymph [139]; similar results were seen with cabozan-
node, liver, and/or lung metastases [122, 123]. tinib, which has additional targeting activity
Additionally, prostate cancer more commonly against c-MET, AXL, and RET [140].
spreads to the calvarium and dura, which can be Immunotherapy with checkpoint inhibitors has
seen in 2–8% of patients. Calvarial metastases changed the face of RCC treatment; nivolumab
are usually asymptomatic but can exert mass has had objective positive responses for BM
effect on venous structures leading to increased [138, 141]. A case report by Rothermundt et al.
ICP or venous infarcts [124–126]; dural metas- described a patient being treated with pembroli-
tases can mimic meningiomas or hematomas zumab, another PD-1 inhibitor, which with ste-
[122, 127]. roids contributed to regression of BM from renal
Patients with clinically localized disease may cell carcinoma [142].
be treated with a combination of radiation, andro-
gen deprivation therapy, or bilateral orchiectomy
and radical prostatectomy [128, 129]. Treatment Bladder
of metastatic disease is focused on androgen
deprivation therapy for castration-sensitive dis- Bladder cancer accounts for approximately 4.5%
ease; castration-resistant disease warrants andro- of malignancies in the United States and is the
gen inhibition with agents like abiraterone or fifth leading cause of cancer and one of the most
enzalutamide alongside chemotherapy including common malignancies seen in men [117]. Bladder
docetaxel, cabazitaxel, and mitoxantrone [130]. cancer usually spreads through local invasion as
As brain lesions are relatively uncommon, no well as hematogenous dissemination, usually
systemic treatment has shown definitive efficacy. involving the liver, lung, and bone [143, 144].
BM in bladder cancer is rare, with incidence
reported as 0–7% [145]. In a study by Mahmoud-
Renal Cell Carcinoma Ahmed et al., headache and motor weakness
were the most common symptoms seen with
Renal cell carcinoma (RCC) accounts for about bladder cancer BM [146].
1% of systemic cancers [131, 132]; clear cell car- Localized, non-muscle-invasive tumor can be
cinoma is the most common histological subtype responsive to transurethral resection and intra-
(70–80%). Frequently seen sites of distant metas- vesical delivery of chemotherapy and immuno-
tasis include the lungs, lymph nodes, liver, and logic therapy. For invasive tumors, adjuvant
bone. chemotherapy with cystectomy is often required.
BM occurs in about 3.5–17% of patients The most common chemotherapy regimen is
[133]. RCC brain metastases have a tendency to MVAC (methotrexate, vinblastine, doxorubicin,
be hemorrhagic [134], leading to complications and cisplatin); none of these drugs cross the
of intracerebral hemorrhage and extensive peritu- blood-brain barrier, potentially allowing for brain
moral edema [132, 135, 136]. Often BM can metastasis development late in the disease and/or
present as a solitary lesion [137]. CNS relapse [146, 147].
Testicular with a poor prognosis, particularly because of the

incomplete understanding of the various molecu-
Testicular cancer is the most common solid lar mechanisms and genetic phenotypes of these
malignancy diagnosed in men less than 40 years metastases, making the creation of standardized
of age but is one of the most curable solid tumors therapies extremely difficult. Additionally, many
in the United States [117]. More than 95% of tes- systemic therapies do not cross the BBB. Although
ticular cancer is testicular germ-cell tumor brain metastases are rare in gastrointestinal, head/
(TGCT), classified as seminomas and nonsemi- neck, genitourinary, and gynecological malignan-
nomatous germ-cell tumors. Seminomas usually cies, they remain a significant challenge. Further
are not aggressive and present with localized dis- research into targeted therapies, immunotherapy,
ease, while nonseminomatous germ-cell tumors and better surgical and radiation techniques may
carry a worse prognosis and can metastasize. improve patient outcomes in the future.
TGCT are usually chemotherapy responsive,
with up to 70–80% survival rates [148].
BM of TGCT is relatively uncommon, with an References
incidence ~1% (range of 0.4–4%) [149, 150], and
usually carries a poor prognosis. The International 1. Nussbaum ES, Djalilian HR, Cho KH, Hall
WA. Brain metastases. Histology, multiplicity, sur-
Germ Cell Cancer Collaborative Group reported gery, and survival. Cancer. 1996;78(8):1781–8.
incidence of BM at the initial diagnosis of all 2. DeAngelis LM, Posner JB, editors. Neurologic com-
advanced TGCT to be 10–15% and 1–2% of all plications of cancer. New York: Oxford University
TGCT cases in general [151]. Press; 2009.
3. Owonikoko TK, Arbiser J, Zelnak A, et al. Current
For stage I seminomas, orchiectomy is usually approaches to the treatment of metastatic brain
curative; for stage II, it is dependent on lymph tumours. Nat Rev Clin Oncol. 2014;11(4):203–22.
node involvement, so it can include radiotherapy 4. Sita TL, Petras KG, Wafford QE, Berendsen MA,
or cisplatin-based chemotherapy. The usual regi- Kruser TJ. Radiotherapy for cranial and brain metas-
tases from prostate cancer: a systematic review. J
men for nonseminomatous germ-cell tumors and Neuro-Oncol. 2017;133(3):531–8.
more aggressive seminomas involves triple com- 5. Maher EA, Mietz J, Arteaga CL, DePinho RA, Mohla
bination of bleomycin, etoposide, and cisplatin S. Brain metastasis: opportunities in basic and trans-
(BEP). Ginsberg et al. described that cisplatin lational research. Cancer Res. 2009;69(15):6015–20.
6. Lin X, DeAngelis LM. Treatment of brain metasta-
and etoposide can cross the BBB such as in the ses. J Clin Oncol. 2015;33(30):3475–84.
presence of CNS tumors or previous radiation to 7. Parrish KE, Sarkaria JN, Elmquist WF. Improving
that area, so presumably high-dose chemother- drug delivery to primary and metastatic brain
apy could work and improve outcomes in these tumors: strategies to overcome the blood-brain bar-
rier. Clin Pharmacol Ther. 2015;97(4):336–46.
patients [152]. In general, due to the penetration 8. Siegel RL, Miller KD, Jemal A. Cancer statistics,
of the chemotherapy in the BBB, BM from TGCT 2015. CA Cancer J Clin. 2015;65(1):5–29.
generally is sensitive and can have survival rates 9. Ferlay J, Soerjomataram I, Dikshit R, et al. Cancer
of 45% but much poorer prognosis of 12% if BM incidence and mortality worldwide: sources, meth-
ods and major patterns in GLOBOCAN 2012. Int J
and/or relapses present during or after chemo- Cancer. 2015;136(5):E359–86.
therapy [153]. 10. Jemal A, Simard EP, Dorell C, et al. Annual Report
to the Nation on the Status of Cancer, 1975-
2009, featuring the burden and trends in human
papillomavirus(HPV)-associated cancers and HPV
Summary vaccination coverage levels. J Natl Cancer Inst.
2013;105(3):175–201.
In conclusion, brain metastases continue to be 11. Burt RW, DiSario JA, Cannon-Albright L. Genetics
much more common than primary brain of colon cancer: impact of inheritance on colon can-
cer risk. Annu Rev Med. 1995;46:371–9.
tumors and are a significant cause of morbidity 12. Ponz de Leon M, Sassatelli R, Benatti P, Roncucci
and mortality in cancer patients. Although out- L. Identification of hereditary nonpolyposis colorec-
comes are improving, they are often associated tal cancer in the general population. The 6-year
experience of a population-based registry. Cancer. 27. Magni E, Santoro L, Ravenda PS, et al. Brain
1993;71(11):3493–501. metastases from colorectal cancer: main clini-
13. Lynch HT, Smyrk TC, Watson P, et al. Genetics, natu- cal factors conditioning outcome. Int J Color Dis.
ral history, tumor spectrum, and pathology of hered- 2014;29(2):201–8.
itary nonpolyposis colorectal cancer: an updated 28. Mongan JP, Fadul CE, Cole BF, et al. Brain metas-
review. Gastroenterology. 1993;104(5):1535–49. tases from colorectal cancer: risk factors, incidence,
14. Attard TM, Giglio P, Koppula S, Snyder C, Lynch and the possible role of chemokines. Clin Colorectal
HT. Brain tumors in individuals with familial Cancer. 2009;8(2):100–5.
adenomatous polyposis: a cancer registry expe- 29. Tran B, Kopetz S, Tie J, et al. Impact of BRAF
rience and pooled case report analysis. Cancer. mutation and microsatellite instability on the pat-
2007;109(4):761–6. tern of metastatic spread and prognosis in metastatic
15. Hamilton SR, Liu B, Parsons RE, et al. The molec- colorectal cancer. Cancer. 2011;117(20):4623–32.
ular basis of Turcot’s syndrome. N Engl J Med. 30. Scartozzi M, Bearzi I, Berardi R, Mandolesi A,
1995;332(13):839–47. Fabris G, Cascinu S. Epidermal growth factor recep-
16. Magge R, Diamond EL. Neurological complica- tor (EGFR) status in primary colorectal tumors does
tions of gastrointestinal cancer. In: Schiff D, not correlate with EGFR expression in related meta-
Arrillaga I, Wen PY, editors. Cancer neurology in static sites: implications for treatment with EGFR-
clinical practice: neurological complications of targeted monoclonal antibodies. J Clin Oncol.
cancer and its treatment. Zurich: Springer; 2018. 2004;22(23):4772–8.
p. 471–84. 31. De Maglio G, Casagrande M, Guardascione M, et al.
17. Michl M, Thurmaier J, Schubert-Fritschle G, et al. MGMT promoter methylation status in brain metas-
Brain metastasis in colorectal cancer patients: tases from colorectal cancer and corresponding pri-
survival and analysis of prognostic factors. Clin mary tumors. Future Oncol. 2015;11(8):1201–9.
Colorectal Cancer. 2015;14(4):281–90. 32. Onodera H, Nagayama S, Tachibana T, Fujimoto A,
18. Amichetti M, Lay G, Dessi M, et al. Results of Imamura M. Brain metastasis from colorectal can-
whole brain radiation therapy in patients with brain cer. Int J Color Dis. 2005;20(1):57–61.
metastases from colorectal carcinoma. Tumori. 33. Christensen TD, Spindler KL, Palshof JA, Nielsen
2005;91(2):163–7. DL. Systematic review: brain metastases from
19. Wronski M, Arbit E. Resection of brain metastases colorectal cancer--incidence and patient characteris-
from colorectal carcinoma in 73 patients. Cancer. tics. BMC Cancer. 2016;16:260.
1999;85(8):1677–85. 34. Pennathur A, Gibson MK, Jobe BA, Luketich
20. Esmaeilzadeh M, Majlesara A, Faridar A, et al. Brain JD. Oesophageal carcinoma. Lancet.
metastasis from gastrointestinal cancers: a system- 2013;381(9864):400–12.
atic review. Int J Clin Pract. 2014;68(7):890–9. 35. Gabrielsen TO, Eldevik OP, Orringer MB, Marshall
21. Martini G, Troiani T, Cardone C, et al. Present BL. Esophageal carcinoma metastatic to the brain:
and future of metastatic colorectal cancer treat- clinical value and cost-effectiveness of routine
ment: a review of new candidate targets. World J enhanced head CT before esophagectomy. AJNR
Gastroenterol. 2017;23(26):4675–88. Am J Neuroradiol. 1995;16(9):1915–21.
22. Tokoro T, Okuno K, Hida JC, et al. Prognostic fac- 36. Welch G, Ross HJ, Patel NP, et al. Incidence of brain
tors for patients with advanced colorectal cancer metastasis from esophageal cancer. Dis Esophagus.
and symptomatic brain metastases. Clin Colorectal 2017;30(9):1–6.
Cancer. 2014;13(4):226–31. 37. Weinberg JS, Suki D, Hanbali F, Cohen ZR, Lenzi
23. Tanriverdi O, Kaytan-Saglam E, Ulger S, et al. The R, Sawaya R. Metastasis of esophageal carcinoma to
clinical and pathological features of 133 colorectal the brain. Cancer. 2003;98(9):1925–33.
cancer patients with brain metastasis: a multicenter 38. Wadhwa R, Taketa T, Correa AM, et al. Incidence of
retrospective analysis of the Gastrointestinal Tumors brain metastases after trimodality therapy in patients
Working Committee of the Turkish Oncology Group with esophageal or gastroesophageal cancer: impli-
(TOG). Med Oncol. 2014;31(9):152. cations for screening and surveillance. Oncology.
24. Yaeger R, Cowell E, Chou JF, et al. RAS mutations 2013;85(4):204–7.
affect pattern of metastatic spread and increase pro- 39. Go PH, Klaassen Z, Meadows MC, Chamberlain
pensity for brain metastasis in colorectal cancer. RS. Gastrointestinal cancer and brain metas-
Cancer. 2015;121(8):1195–203. tasis: a rare and ominous sign. Cancer.
25. Tie J, Lipton L, Desai J, et al. KRAS mutation is 2011;117(16):3630–40.
associated with lung metastasis in patients with 40. Smith RS, Miller RC. Incidence of brain metasta-
curatively resected colorectal cancer. Clin Cancer sis in patients with esophageal carcinoma. World J
Res. 2011;17(5):1122–30. Gastroenterol. 2011;17(19):2407–10.
26. Kemeny NE, Chou JF, Capanu M, et al. KRAS 41. Ogawa K, Toita T, Sueyama H, et al. Brain metas-
mutation influences recurrence patterns in patients tases from esophageal carcinoma: natural his-
undergoing hepatic resection of colorectal metasta- tory, prognostic factors, and outcome. Cancer.
ses. Cancer. 2014;120(24):3965–71. 2002;94(3):759–64.
42. Sjoquist KM, Burmeister BH, Smithers BM, et al. nial and spinal metastatic groups. J Clin Neurosci.
Survival after neoadjuvant chemotherapy or chemo- 2011;18(3):364–8.
radiotherapy for resectable oesophageal carci- 57. Lim S, Lee S, Lim JY, et al. Hepatocellular carci-
noma: an updated meta-analysis. Lancet Oncol. noma specific graded prognostic assessment can
2011;12(7):681–92. predict outcomes for patients with brain metasta-
43. Abu Hejleh T, Deyoung BR, Engelman E, et al. ses from hepatocellular carcinoma. J Neuro-Oncol.
Relationship between HER-2 overexpression and 2014;120(1):199–207.
brain metastasis in esophageal cancer patients. 58. Han MS, Moon KS, Lee KH, et al. Brain metastasis
World J Gastrointest Oncol. 2012;4(5):103–8. from hepatocellular carcinoma: the role of surgery as
44. Preusser M, Berghoff AS, Ilhan-Mutlu A, et al. Brain a prognostic factor. BMC Cancer. 2013;13:567.
metastases of gastro-oesophageal cancer: evaluation 59. Kato H, Yoshida H, Taniguch H, et al. Cyberknife
of molecules with relevance for targeted therapies. treatment for advanced or terminal stage hepa-
Anticancer Res. 2013;33(3):1065–71. tocellular carcinoma. World J Gastroenterol.
45. Mathers CD, Shibuya K, Boschi-Pinto C, Lopez AD, 2015;21(46):13101–12.
Murray CJ. Global and regional estimates of can- 60. Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib in
cer mortality and incidence by site: I. Application advanced hepatocellular carcinoma. N Engl J Med.
of regional cancer survival model to estimate can- 2008;359(4):378–90.
cer mortality distribution by site. BMC Cancer. 61. Llovet JM, Villanueva A, Lachenmayer A, Finn
2002;2:36. RS. Advances in targeted therapies for hepatocel-
46. Van Cutsem E, Moiseyenko VM, Tjulandin S, et al. lular carcinoma in the genomic era. Nat Rev Clin
Phase III study of docetaxel and cisplatin plus fluo- Oncol. 2015;12(7):408–24.
rouracil compared with cisplatin and fluorouracil 62. Hammoud GM, Ibdah JA. Are we getting closer to
as first-line therapy for advanced gastric cancer: understanding intratumor heterogeneity in hepa-
a report of the V325 Study Group. J Clin Oncol. tocellular carcinoma? Hepatobiliary Surg Nutr.
2006;24(31):4991–7. 2016;5(2):188–90.
47. Kim M. Intracranial involvement by metastatic 63. Ohri N, Kaubisch A, Garg M, Guha C. Targeted
advanced gastric carcinoma. J Neuro-Oncol. therapy for hepatocellular carcinoma. Semin Radiat
1999;43(1):59–62. Oncol. 2016;26(4):338–43.
48. York JE, Stringer J, Ajani JA, Wildrick DM, 64. Tannock IF, Hickman JA. Limits to personalized can-
Gokaslan ZL. Gastric cancer and metastasis to the cer medicine. N Engl J Med. 2016;375(13):1289–94.
brain. Ann Surg Oncol. 1999;6(8):771–6. 65. Wang S, Wang A, Lin J, et al. Brain metastases
49. Lemke J, Barth TF, Juchems M, Kapapa T, Henne- from hepatocellular carcinoma: recent advances and
Bruns D, Kornmann M. Long-term survival follow- future avenues. Oncotarget. 2017;8(15):25814–29.
ing resection of brain metastases from pancreatic 66. Hidalgo M, Cascinu S, Kleeff J, et al. Addressing
cancer. Anticancer Res. 2011;31(12):4599–603. the challenges of pancreatic cancer: future direc-
50. Kasakura Y, Fujii M, Mochizuki F, Suzuki T, tions for improving outcomes. Pancreatology.
Takahashi T. Clinicopathological study of brain 2015;15(1):8–18.
metastasis in gastric cancer patients. Surg Today. 67. Ilic M, Ilic I. Epidemiology of pancreatic cancer.
2000;30(6):485–90. World J Gastroenterol. 2016;22(44):9694–705.
51. Jun KH, Lee JE, Kim SH, et al. Clinicopathological 68. Lemke J, Scheele J, Kapapa T, Wirtz CR, Henne-
significance of N-cadherin and VEGF in advanced Bruns D, Kornmann M. Brain metastasis in pancre-
gastric cancer brain metastasis and the effects atic cancer. Int J Mol Sci. 2013;14(2):4163–73.
of metformin in preclinical models. Oncol Rep. 69. Park KS, Kim M, Park SH, Lee KW. Nervous system
2015;34(4):2047–53. involvement by pancreatic cancer. J Neuro-Oncol.
52. Bosch FX, Ribes J, Diaz M, Cleries R. Primary 2003;63(3):313–6.
liver cancer: worldwide incidence and trends. 70. Kamisawa T, Wood LD, Itoi T, Takaori K. Pancreatic
Gastroenterology. 2004;127(5 Suppl 1):S5–s16. cancer. Lancet. 2016;388(10039):73–85.
53. Gaddikeri S, McNeeley MF, Wang CL, et al. 71. Sharma A, Sharma KL, Gupta A, Yadav A, Kumar
Hepatocellular carcinoma in the noncirrhotic liver. A. Gallbladder cancer epidemiology, pathogen-
AJR Am J Roentgenol. 2014;203(1):W34–47. esis and molecular genetics: recent update. World J
54. Menis J, Fontanella C, Follador A, Fasola G, Aprile Gastroenterol. 2017;23(22):3978–98.
G. Brain metastases from gastrointestinal tumours: 72. Takano S, Yoshii Y, Owada T, Shirai S, Nose
tailoring the approach to maximize the outcome. Crit T. Central nervous system metastasis from gall-
Rev Oncol Hematol. 2013;85(1):32–44. bladder carcinoma--case report. Neurol Med Chir.
55. Jiang XB, Ke C, Zhang GH, et al. Brain metastases 1991;31(12):782–6.
from hepatocellular carcinoma: clinical features and 73. Daines WP, Rajagopalan V, Grossbard ML, Kozuch
prognostic factors. BMC Cancer. 2012;12:49. P. Gallbladder and biliary tract carcinoma: a compre-
56. Hsiao SY, Chen SF, Chang CC, et al. Central nervous hensive update, Part 2. Oncology (Williston Park).
system involvement in hepatocellular carcinoma: 2004;18(8):1049–59; discussion 1060, 1065–1046,
clinical characteristics and comparison of intracra- 1068.
74. Siegel R, Naishadham D, Jemal A. Cancer statistics, plasia: an update on incidence, management and out-
2012. CA Cancer J Clin. 2012;62(1):10–29. come. Gynecol Oncol. 2015;137(1):73–6.
75. Haddad RI, Shin DM. Recent advances in head and 92. Yordan EL Jr, Schlaerth J, Gaddis O, Morrow
neck cancer. N Engl J Med. 2008;359(11):1143–54. CP. Radiation therapy in the management of gesta-
76. Shah JP. Patterns of cervical lymph node metastasis tional choriocarcinoma metastatic to the central ner-
from squamous carcinomas of the upper aerodiges- vous system. Obstet Gynecol. 1987;69(4):627–30.
tive tract. Am J Surg. 1990;160(4):405–9. 93. Newlands ES, Holden L, Seckl MJ, McNeish I,
77. Kotwall C, Sako K, Razack MS, Rao U, Bakamjian Strickland S, Rustin GJ. Management of brain metas-
V, Shedd DP. Metastatic patterns in squamous tases in patients with high-risk gestational tropho-
cell cancer of the head and neck. Am J Surg. blastic tumors. J Reprod Med. 2002;47(6):465–71.
1987;154(4):439–42. 94. Ghaemmaghami F, Behtash N, Memarpour N,
78. Xie X, O’Neill W, Pan Q. Immunotherapy for head Soleimani K, Hanjani P, Hashemi FA. Evaluation
and neck cancer: the future of treatment? Expert and management of brain metastatic patients with
Opin Biol Ther. 2017;17(6):701–8. high-risk gestational trophoblastic tumors. Int J
79. Patel RA, Bell JB, Kim T, et al. Stereotactic radio- Gynecol Cancer. 2004;14(5):966–71.
surgery for brain metastases from primary head and 95. Soper JT, Spillman M, Sampson JH, Kirkpatrick
neck carcinomas: a retrospective analysis. J Neuro- JP, Wolf JK, Clarke-Pearson DL. High-risk ges-
Oncol. 2017;134(1):197–203. tational trophoblastic neoplasia with brain metas-
80. Batsakis JG. Chemodectomas of the head and neck. tases: individualized multidisciplinary therapy in
In: Batsakis JG, editor. Tumors of the head and neck. the management of four patients. Gynecol Oncol.
Clinical and pathological considerations. 2nd ed. 2007;104(3):691–4.
Baltimore: Williams and Wilkins; 1976. p. 280–8. 96. Borst P, Evers R, Kool M, Wijnholds J. A family of
81. Lips C, Lentjes E, Hoppener J, Luijt R, Moll drug transporters: the multidrug resistance-associated
F. Familial paragangliomas. Hered Cancer Clin proteins. J Natl Cancer Inst. 2000;92(16):1295–302.
Pract. 2006;4(4):169–76. 97. Wijnholds J, de Lange EC, Scheffer GL, et al.
82. Lee JH, Barich F, Karnell LH, et al. National Cancer Multidrug resistance protein 1 protects the cho-
Data Base report on malignant paragangliomas of roid plexus epithelium and contributes to the
the head and neck. Cancer. 2002;94(3):730–7. blood-cerebrospinal fluid barrier. J Clin Invest.
83. Moore MG, Netterville JL, Mendenhall WM, 2000;105(3):279–85.
Isaacson B, Nussenbaum B. Head and neck paragan- 98. Monaco E 3rd, Kondziolka D, Mongia S, Niranjan
gliomas: an update on evaluation and management. A, Flickinger JC, Lunsford LD. Management of
Otolaryngology. 2016;154(4):597–605. brain metastases from ovarian and endometrial
84. Wang X, Zhu X, Chen J, Liu Y, Mao Q. Metastatic carcinoma with stereotactic radiosurgery. Cancer.
brain carotid body paraganglioma with endocrine 2008;113(9):2610–4.
activity: a case report and literature review. Br J 99. Ferlay J, Soerjomataram I, Ervik M, et al.
Neurosurg. 2017;33:1–3. GLOBOCAN 2012 v1.0, Cancer incidence and mor-
85. Guo J, Zhong C, Liu Q, et al. Intracranial choriocar- tality worldwide: IARC Cancerbase no. 11 [inter-
cinoma occurrence in males: two cases and a review net]. Lyon: International Agency for Research on
of the literature. Oncol Lett. 2013;6(5):1329–32. Cancer; 2013.
86. Altieri A, Franceschi S, Ferlay J, Smith J, La 100. Ly KI, Mrugala MM. Neurological complications
Vecchia C. Epidemiology and aetiology of ges- of female reproductive tract cancers. In: Schiff D,
tational trophoblastic diseases. Lancet Oncol. Arrillaga I, Wen PY, editors. Cancer neurology in
2003;4(11):670–8. clinical practice: neurological complications of
87. Graf AH, Buchberger W, Langmayr H, Schmid cancer and its treatment. Zurich: Springer; 2018.
KW. Site preference of metastatic tumours of the p. 497–514.
brain. Virchows Arch A Pathol Anat Histopathol. 101. Pakneshan S, Safarpour D, Tavassoli F, Jabbari
1988;412(5):493–8. B. Brain metastasis from ovarian cancer: a system-
88. Berkowitz RS, Goldstein DP. Current manage- atic review. J Neuro-Oncol. 2014;119(1):1–6.
ment of gestational trophoblastic diseases. Gynecol 102. Cormio G, Maneo A, Colamaria A, Loverro G,
Oncol. 2009;112(3):654–62. Lissoni A, Selvaggi L. Surgical resection of
89. Neubauer NL, Latif N, Kalakota K, et al. Brain solitary brain metastasis from ovarian carci-
metastasis in gestational trophoblastic neoplasia: an noma: an analysis of 22 cases. Gynecol Oncol.
update. J Reprod Med. 2012;57(7–8):288–92. 2003;89(1):116–9.
90. Cagayan MS, Lu-Lasala LR. Management of ges- 103. Niu X, Rajanbabu A, Delisle M, et al. Brain metas-
tational trophoblastic neoplasia with metastasis tases in women with epithelial ovarian cancer: mul-
to the central nervous system: a 12-year review at timodal treatment including surgery or gamma-knife
the Philippine General Hospital. J Reprod Med. radiation is associated with prolonged survival. J
2006;51(10):785–92. Obstet Gynaecol Can. 2013;35(9):816–22.
91. Savage P, Kelpanides I, Tuthill M, Short D, Seckl 104. Kwon JW, Yoon JH, Lim MC, et al. Treatment results
MJ. Brain metastases in gestational trophoblast neo- and prognostic factors of brain metastases from
ovarian cancer: a single institutional experience of 56 122. Tremont-Lukats IW, Bobustuc G, Lagos GK, Lolas
patients. Int J Gynecol Cancer. 2018;28(8):1631–8. K, Kyritsis AP, Puduvalli VK. Brain metastasis from
105. Chura JC, Shukla K, Argenta PA. Brain metasta- prostate carcinoma: the M. D. Anderson Cancer
sis from cervical carcinoma. Int J Gynecol Cancer. Center experience. Cancer. 2003;98(2):363–8.
2007;17(1):141–6. 123. Hatzoglou V, Patel GV, Morris MJ, et al. Brain
106. Cormio G, Pellegrino A, Landoni F, et al. Brain metastases from prostate cancer: an 11-year analysis
metastases from cervical carcinoma. Tumori. in the MRI era with emphasis on imaging charac-
1996;82(4):394–6. teristics, incidence, and prognosis. J Neuroimaging.
107. Fetcko K, Gondim DD, Bonnin JM, Dey M. Cervical 2014;24(2):161–6.
cancer metastasis to the brain: a case report and 124. Bubendorf L, Schopfer A, Wagner U, et al. Metastatic
review of literature. Surg Neurol Int. 2017;8:181. patterns of prostate cancer: an autopsy study of
108. Mahmoud-Ahmed AS, Suh JH, Barnett GH, Webster 1,589 patients. Hum Pathol. 2000;31(5):578–83.
KD, Kennedy AW. Tumor distribution and survival 125. Long MA, Husband JE. Features of unusual
in six patients with brain metastases from cervical metastases from prostate cancer. Br J Radiol.
carcinoma. Gynecol Oncol. 2001;81(2):196–200. 1999;72(862):933–41.
109. Matsunaga S, Shuto T, Sato M. Gamma knife sur- 126. Raizer JJ, DeAngelis LM. Cerebral sinus thrombosis
gery for metastatic brain tumors from gynecologic diagnosed by MRI and MR venography in cancer
cancer. World Neurosurg. 2016;89:455–63. patients. Neurology. 2000;54(6):1222–6.
110. Onstad MA, Schmandt RE, Lu KH. Addressing the 127. Kleinschmidt-DeMasters BK. Dural metastases.
role of obesity in endometrial cancer risk, prevention, A retrospective surgical and autopsy series. Arch
and treatment. J Clin Oncol. 2016;34(35):4225–30. Pathol Lab Med. 2001;125(7):880–7.
111. Lortet-Tieulent J, Ferlay J, Bray F, Jemal 128. Hoffman RM, Koyama T, Fan KH, et al. Mortality
A. International patterns and trends in endometrial after radical prostatectomy or external beam radio-
cancer incidence, 1978-2013. J Natl Cancer Inst. therapy for localized prostate cancer. J Natl Cancer
2018;110(4):354–61. Inst. 2013;105(10):711–8.
112. Cormio G, Lissoni A, Losa G, Zanetta G, Pellegrino 129. Nepple KG, Stephenson AJ, Kallogjeri D, et al.
A, Mangioni C. Brain metastases from endometrial Mortality after prostate cancer treatment with radi-
carcinoma. Gynecol Oncol. 1996;61(1):40–3. cal prostatectomy, external-beam radiation therapy,
113. Cybulska P, Stasenko M, Alter R, et al. Brain metas- or brachytherapy in men without comorbidity. Eur
tases in patients with low-grade endometrial carci- Urol. 2013;64(3):372–8.
noma. Gynecol Oncol Rep. 2018;26:87–90. 130. Basch E, Loblaw DA, Oliver TK, Carducci M,
114. Gien LT, Kwon JS, D’Souza DP, et al. Brain metas- Chen RC, Frame JN, et al. Systemic therapy in men
tases from endometrial carcinoma: a retrospective with metastatic castration-resistant prostate cancer:
study. Gynecol Oncol. 2004;93(2):524–8. American Society of Clinical Oncology and Cancer
115. Sorosky JI. Endometrial cancer. Obstet Gynecol. Care Ontario clinical practice guideline. J Clin
2012;120(2 Pt 1):383–97. Oncol. 2014;32(30):3436–48.
116. Uccella S, Morris JM, Multinu F, et al. Primary 131. Kim YH, Kim JW, Chung HT, Paek SH, Kim DG,
brain metastases of endometrial cancer: a report of Jung HW. Brain metastasis from renal cell carci-
18 cases and review of the literature. Gynecol Oncol. noma. Prog Neurol Surg. 2012;25:163–75.
2016;142(1):70–5. 132. Sheehan JP, Sun MH, Kondziolka D, Flickinger
117. Howlader N, Noone AM, Krapcho M, Garshell J, J, Lunsford LD. Radiosurgery in patients with
Miller D, Altekruse SF, et al. SEER cancer statis- renal cell carcinoma metastasis to the brain: long-
tics review, 1975–2012. Bethesda: National Cancer term outcomes and prognostic factors influenc-
Institute. http://seer.cancer.gov/csr/1975_2012/, ing survival and local tumor control. J Neurosurg.
based on November 2014 SEER data submission, 2003;98(2):342–9.
posted to the SEER web site, April 2015. Accessed 133. Bennani O, Derrey S, Langlois O, et al. Brain metas-
June 2015. tasis from renal cell carcinoma. Neuro-Chirurgie.
118. Catane R, Kaufman J, West C, Merrin C, Tsukada Y, 2014;60(1–2):12–6.
Murphy GP. Brain metastasis from prostatic carci- 134. Posner JB. Neurologic complications of cancer.
noma. Cancer. 1976;38(6):2583–7. Philadelphia: FA Davis; 1995.
119. Chung TS, Thannikkary C. Carcinoma of the 135. Choi WH, Koh YC, Song SW, Roh HG, Lim
prostate with brain metastasis. J Surg Oncol. SD. Extremely delayed brain metastasis from
1986;33(2):103–5. renal cell carcinoma. Brain Tumor Res Treat.
120. Lynes WL, Bostwick DG, Freiha FS, Stamey 2013;1(2):99–102.
TA. Parenchymal brain metastases from adenocarci- 136. Muacevic A, Siebels M, Tonn JC, Wowra
noma of prostate. Urology. 1986;28(4):280–7. B. Treatment of brain metastases in renal cell carci-
121. McCutcheon IE, Eng DY, Logothetis CJ. Brain noma: radiotherapy, radiosurgery, or surgery? World
metastasis from prostate carcinoma: antemortem J Urol. 2005;23(3):180–4.
recognition and outcome after treatment. Cancer. 137. Saitoh H. Distant metastasis of renal adenocarci-
1999;86(11):2301–11. noma. Cancer. 1981;48(6):1487–91.
138. Ramalingam S, George DJ, Harrison MR. How we 145. Anderson TS, Regine WF, Kryscio R, Patchell
treat brain metastases in metastatic renal cell carci- RA. Neurologic complications of bladder carcinoma:
noma. Clin Adv Hematol Oncol. 2018;16(2):110–4. a review of 359 cases. Cancer. 2003;97(9):2267–72.
139. Chevreau C, Ravaud A, Escudier B, et al. A phase 146. Mahmoud-Ahmed AS, Suh JH, Kupelian PA,
II trial of sunitinib in patients with renal cell can- et al. Brain metastases from bladder carcinoma:
cer and untreated brain metastases. Clin Genitourin presentation, treatment and survival. J Urol.
Cancer. 2014;12(1):50–4. 2002;167(6):2419–22.
140. Motzer RJ, Hutson TE, Glen H, et al. Lenvatinib, 147. Dhote R, Beuzeboc P, Thiounn N, et al. High inci-
everolimus, and the combination in patients with dence of brain metastases in patients treated with an
metastatic renal cell carcinoma: a randomised, M-VAC regimen for advanced bladder cancer. Eur
phase 2, open-label, multicentre trial. Lancet Oncol. Urol. 1998;33(4):392–5.
2015;16(15):1473–82. 148. Sheinfeld J, Bajorin D. Management of the post-
141. Escudier BJ, Chabaud S, Borchiellini D, et al. chemotherapy residual mass. Urol Clin North Am.
Efficacy and safety of nivolumab in patients with 1993;20(1):133–43.
metastatic renal cell carcinoma (mRCC) and brain 149. Raina V, Singh SP, Kamble N, et al. Brain metastasis
metastases: preliminary results from the GETUG- as the site of relapse in germ cell tumor of testis.
AFU 26 (Nivoren) study [ASCO abstract 4563]. J Cancer. 1993;72(7):2182–5.
Clin Oncol. 2017;35(15 suppl):4563. 150. Fossa SD, Bokemeyer C, Gerl A, et al. Treatment out-
142. Rothermundt C, Hader C, Gillessen S. Successful come of patients with brain metastases from malig-
treatment with an anti-PD-1 antibody for progress- nant germ cell tumors. Cancer. 1999;85(4):988–97.
ing brain metastases in renal cell cancer. Ann Oncol. 151. Nonomura N, Nagahara A, Oka D, et al. Brain
2016;27(3):544–5. metastases from testicular germ cell tumors: a retro-
143. Kishi K, Hirota T, Matsumoto K, Kakizoe T, Murase spective analysis. Int J Urol. 2009;16(11):887–93.
T, Fujita J. Carcinoma of the bladder: a clinical and 152. Ginsberg S, Kirshner J, Reich S, et al. Systemic
pathological analysis of 87 autopsy cases. J Urol. chemotherapy for a primary germ cell tumor of the
1981;125(1):36–9. brain: a pharmacokinetic study. Cancer Treat Rep.
144. Schaefer O, Lohrmann C, Harder J, Veelken H, 1981;65(5–6):477–83.
Langer M. Treatment of renal cell carcinoma- 153. Taylor J. Neurological complications of genito-
associated dermatomyositis with renal arterial urinary cancer. In: Schiff D, Arrillaga I, Wen PY,
embolization and percutaneous radiofrequency editors. Cancer neurology in clinical practice: neu-
heat ablation. J Vasc Interv Radiol. 2004;15(1 Pt rological complications of cancer and its treatment.
1):97–9. Zurich: Springer; 2018. p. 485–96.
Management of Solid Tumor CNS
Metastases in Children
18
Whitney E. Parker, Shahiba Q. Ogilvie,
Lily McLaughlin, and Mark M. Souweidane
Introduction and Epidemiology ney tumors [including Wilms tumors and clear

cell sarcomas of the kidney (CCSK), the latter
Essential differences between adult and pediat- of which have been found to have a 5–11% inci-
ric solid tumors suggest that the two ought to, in dence of CNS metastases], and germ cell tumors
many regards, be considered distinct pathological (with a particularly high rate of CNS metasta-
entities. This distinction is particularly profound ses in choriocarcinoma, up to 43%), reflecting
in the case of central nervous system (CNS) an increased representation of undifferentiated
metastases. While adult solid tumor brain metas- tumor types [3, 16–18]. Additionally, rare lung
tases occur in approximately 20–40% of primary tumors in children have been reported to have an
tumor cases, the frequency of solid tumor brain increased incidence of CNS metastases, includ-
metastases reported in children is only 1–10%, ing pleuropulmonary blastoma (PPB), with up to
or 6–13% reported at autopsy [1–14]. In adults, 25% incidence, and alveolar soft part sarcoma,
CNS metastases occur most frequently in cases with 15–29% incidence [6, 19, 20].
of lung, breast, and gastrointestinal primary Pediatric solid tumors can enter the CNS
tumors and melanoma [1, 15]. In contrast, the space via one of two mechanisms—direct exten-
most common solid primary tumor types to pression, as with sinonasal tumors, or hematogenous
ent with CNS metastases in the pediatric popula- metastatic spread which necessitates penetrating
tion are sarcomas (including soft tissue, Ewing, the blood-brain barrier (BBB). Treatments for
and osteosarcoma), melanomas (up to 18% prev- CNS metastases tend to vary based upon primary
alence), retinoblastomas, neuroblastomas, kid- tumor type, extent of intra- and extracranial dis-
ease, and goals of care. Decreased prevalence of
CNS metastases in pediatric versus adult tumors
W. E. Parker · M. M. Souweidane (*)
Department of Neurosurgery, suggests a difference in the BBB—the pediatric
New York-Presbyterian Hospital, BBB may be less permeable to tumor cells or,
Weill Cornell Medicine, New York, NY, USA more likely, have increased permeability to sys-
Department of Neurosurgery, temic therapies used to treat the primary tumor
Memorial Sloan-Kettering Cancer Center, or extracranial metastatic spread. Additionally,
New York, NY, USA
since tumor-instigated myeloid precursor cells
e-mail: mmsouwei@med.cornell.edu
are believed to play a role in metastasis, the
S. Q. Ogilvie · L. McLaughlin
increased tendency of children to receive mye-
Department of Neurosurgery,
Memorial Sloan-Kettering Cancer Center, loablative therapy for high-risk primary tumors,
New York, NY, USA particularly in the neuroblastoma population,

https://doi.org/10.1007/978-3-030-42958-4_18
260 W. E. Parker et al.
may contribute to a reduction in CNS spread [21– Although the CNS is a common site of extra-
23]. Similarly to adult cases, however, pediatric medullary spread in pediatric leukemia, it is
solid CNS metastases are generally associated rarely seeded by solid tumors in children [6, 25].
with a very poor prognosis, with survival times Because the occurrence of pediatric solid tumor
of typically less than 1 year after diagnosis [17]. CNS metastases is so infrequent, surveillance
imaging in children diagnosed with a primary
solid tumor is not routinely performed. Thus,
Tumor Characteristics most CNS metastases are diagnosed from imag-
and Pathophysiology ing in the setting of presenting symptoms such
as headache, nausea, vomiting, seizures, apha-
The majority of pediatric solid tumor brain sia, visual field deficits, focal motor or sensory
metastases are solitary (approximately 60–90% deficits, cranial neuropathies, ataxia, and altered
of cases), in contrast to adult cases where mul- mental status. These symptoms reflect the loca-
tiple CNS metastases are common. Pediatric tion and size of the tumor, extent of edema, pres-
brain metastases are supratentorial in 85–100% ence of intratumoral hemorrhage, and occurrence
of cases in recently published series, in contrast of obstructive or communicating hydrocephalus
to primary pediatric brain tumors which present [3–7, 9–11, 17]. Pediatric solid tumor CNS metas-
with an infratentorial predominance [3, 4, 6, 10]. tases are rarely the sole or initial metastases and,
Solid brain metastases tend to most commonly when they occur, are often a late disease finding.
be located in the cerebral hemispheres (less fre- Multiple retrospective studies have suggested
quently in the cerebellum and basal ganglia), pre- that there may be a direct correlation between
senting at the gray matter-white matter junction, the occurrence of pulmonary metastases and
as in adults, or in border zones between major brain metastases, across several different pri-
cerebral vascular territories, suggesting an arterial mary tumor types, with up to 70% of cases hav-
delivery mechanism [17, 24]. Interestingly, in our ing a known pulmonary metastasis at the time
surgical experience at Memorial Sloan Kettering of brain metastatic diagnosis [2–4, 6, 9, 13, 26].
Cancer Center (MSKCC), we have found numer- Mechanistically, it is plausible that tumor cells
ous brain metastases at the pial interface, such as at shed into the pulmonary circulation from a lung
the depth of a sulcus, suggesting a possible venous metastasis have a direct route to the brain via
or cerebrospinal fluid (CSF) mechanism of tumor the left atrium, with a subsequent direct arterial
cell seeding (Fig. 18.1, unpublished observations). conduit to brain circulation; this is supported
Fig. 18.1 T1-weighted
MRI demonstrating a
left frontal
neuroblastoma
metastasis at the pial
interface. Pre- (left) and
post-contrast (right) T1
MRI images of a
neuroblastoma
metastasis in a
7-year-old male patient
illustrate the presence of
tumor along the pial
margin, a pattern
commonly manifested in
our cohort
18 Management of Solid Tumor CNS Metastases in Children 261
by the presence of parenchymal metasta- hemorrhage, the type of primary tumor (spe-
ses in major cerebral arterial border zones [2, cifically, whether it is radiosensitive or radio-
24]. Additionally, Kramer and colleagues at resistant), location, and neurologic symptoms.
MSKCC found in their review of neuroblas- Surgical options for brain metastases can include
toma cases with bone marrow involvement an resection, debulking (as with lesions extending
association of lumbar punctures (LP) performed into eloquent areas or deep brain structures),
near the time of primary disease diagnosis with CSF diversion with shunting or endoscopic third
the development of CNS metastases, suggesting ventriculostomy (ETV), or implantation of an
a possible direct hematogenous to CSF seeding intraventricular reservoir for therapeutic delivery.
mechanism [8]. Long-term use of ventricular access reservoirs
Cumulatively across all histological subtypes, has been found to be safe—a recent study from
solid brain metastases in the pediatric population our center reported a 4% rate of acute and rela-
occur at a median age between 11 and 13 years tively minor complications, including catheter
and at a median interval of 8–16 months fol- migration and pericatheter cyst formation [28]. As
lowing the diagnosis of the primary tumor some of these patients may develop hydrocepha-
(Table 18.1) [2, 3, 6]. It has been suggested in lus and require conversion of the intraventricular
multiple prior studies that the incidence of pedi- reservoir to a shunt, a programmable shunt can
atric solid tumor CNS metastases is increas- be implanted for both therapeutic CSF diversion
ing [3, 6, 8]. However, the largest case series and drug delivery (by increasing shunt resistance
reported to date by Suki and colleagues from to the highest setting and thus effectively turning
MD Anderson Cancer Center found that the pro- it “off” during the time of drug infusion).
portion of patients with primary solid tumors
developing CNS metastases remained relatively
low at 1.4%, which was consistent with previ- Radiation
ously reported values from earlier studies [3].
Since pediatric CNS metastases are so rare and Radiation treatment may serve as a monotherapy
case studies have been limited to small cohorts, or supplement surgical resection and/or systemic
it has not yet been determined whether patients therapy; however, this is avoided if possible in
with this diagnosis have experienced an overall children under 3 years of age, due to the likeli-
improvement in survival over time. hood of disruption of normal neurocognitive
function during this critical period of brain devel-
opment and the possibility of developing latent
Treatment Options radiation-induced tumors such as meningiomas,
gliomas, or sarcomas [29]. Whole brain radiation
Largely limited by small cohort sizes, evidence therapy (WBRT) remains the most common radi-
for the efficacy of different treatment regimens ation treatment, delivered in fractionated doses,
for pediatric solid tumor metastases remains often totaling 10–50 Gy [2].
sparse. Treatment options generally include Stereotactic radiosurgery (SRS), however,
surgical resection, radiation, chemotherapy, or is increasingly used and commonly preferred
primary tumor-specific immunotherapy. Brain at our institution even for multiple metastases,
edema can usually be managed with steroids dur- offering the option of effective focal treatment
ing treatment. while minimizing side effects, particularly in
this vulnerable population. Recent studies have
suggested that there is little to no survival benefit
Surgical Considerations of WBRT over SRS, and in fact, that SRS alone
may improve survival in select patient popula-
Surgical treatment depends upon multiple con- tions under 50 years of age and with less than
siderations, including tumor size, presence of four brain metastases [2, 30, 31]. This may be
Table 18.1 Case series of solid tumor CNS metastases in the pediatric population
262
Patient age at Incidence of Interval: primary Survival after Survival after

Study Number of CNS CNS diagnosis to CNS primary CNS
reference Institution Primary cancer type Years patients diagnosis metastases metastasis diagnosis metastasis
Suki et al. [3] MD Anderson Mixed 1990– 54 11.37 years, 1.4% 17 months, median 29 months, 9 months,
2012 median median median
Wiens and Indiana Mixed, germ cell 1981– 26 10.6 years, 2.2%, 18 months, median 34.8 months, 12.5 months,
Hattab [4] University tumor predominance 2011 median excluding median median
germ cell
primary
tumors
Gobel et al. Ludwig Germ cell tumors 1982– 9 NR 1.1% NR NR NR
[32] Maximilians 2009
University
(Munich,
Germany)
Stefanowicz Medical Mixed 1992– 10 13.8 years, 2% 8 months, median NR NR
et al. [6] University of 2010 median
Gdansk
(Poland)
Hauser Semmelweis Mixed 1989– 14 NR 3.4% 11.4 months, mean 21.9 months, 10.4 months,
et al. [7] University 2002 mean mean
(Budapest,
Hungary)
Kebudi Istanbul Mixed 1989– 16 10.5 years, 1.45% 20 months, median NR 2 months,
et al. [5] University 2002 median median
(Turkey)
Spunt St. Jude Germ cell tumors 1962– 16 NR 7.8% NR NR NR
et al. [11] 2002
Paulino University of Sarcoma, 1965– 30 14 years, 4.9% 5 months, median NR 4 months,
et al. [10] Iowa neuroblastoma, 2000 median median
Wilms tumor
Postovsky Rambam Sarcoma 1990– 18 17.4 years, 4.3% NR NR 5.03 months,
et al. [33] Medical Center 2001 mean mean (death or
(Haifa, Israel) last follow-up)
W. E. Parker et al.
Kramer MSKCC Neuroblastoma 1980– 11 NR 6.3%, varying 12.2 months, median NR 6.7 months,
et al. [8] 1999 by study median
protocol
Parasuraman St. Jude Ewing sarcoma (ES), 1962– 21 NR ES: 3.3%, ES: 22 months, median; NR 2.7 months,
et al. [34] rhabdomyosarcoma 1998 RMS: 2.4% RMS: 12 months, median median
(RMS)
Lowis et al. UK Children’s Wilms tumor 1980– 7 NR 0.6% 19 months, median NR NR
[35] Cancer Study 1995
Group
Bouffet Centre Leon Mixed 1987– 12 9 years, 7.4% 15 months, median NR 3 months,
et al. [9] Berard (Lyon, 1995 median median
France)
Rodriguez- St. Jude Melanoma 1962– 8 NR 18% 20 months, median NR 5 months,
Galindo 1995 median
et al. [16]
Tasdemiroglu University of Mixed 1982– 12 NR 7.8%, 4.5% 327 days, mean NR 5.2 months,
and Patchell Kentucky 1994 parenchymal mean
[12]
Weyl-Ben Rambam Mixed 1986– 6 NR 9.8% 13 months, median NR 9.8 months,
Arush et al. Medical Center 1990 median
[36] (Haifa, Israel)
18 Management of Solid Tumor CNS Metastases in Children
Marina St. Jude Osteosarcoma 1962– 13 NR 5.1% 3 months, median NR 16 months,

et al. [27] 1989 median
Baram MD Anderson Osteosarcoma 1980– 5 NR 5.7% 12 months, median NR NR
et al. [37] 1986
Graus MSKCC Mixed 1973– 31 (including NR 22.3%, with Median ranges NR NR
et al. [13] 1982 postmortem 13% diagnosed 8.5 months
diagnosis) at autopsy (rhabdomyosarcoma) to
22 months
(osteosarcoma)
Vannucci and MSKCC Mixed 1951– 13 (including NR 6% 23 months, median NR 1 month
Baten [14] 1972 postmortem (31.5 days),
diagnosis) median
Characteristics of patients reported in case series of pediatric solid tumor metastases are summarized, including primary study center, dates, tumor subtype, number of patients
included, average patient age, incidence of CNS metastasis, time interval between diagnosis of primary tumor and diagnosis of CNS metastasis, survival time after primary
diagnosis, and survival time after CNS diagnosis. Averages are reported in each study, as either means or medians, as specified
263
due in part to the fact that SRS allows delivery of survival, and overall survival while minimizing
higher focal doses of radiation, rather than frac- side effects, particularly in medulloblastoma
tionated or hypofractionated doses, overcoming patients receiving craniospinal irradiation (CSI)
the radioresistance of certain primary cancer sub- [46–50]. Given the reduction in the risk of neu-
types such as melanomas and sarcomas [2, 38, rocognitive deficits associated with proton ther-
39]. Importantly, SRS may be associated with apy, it is a particularly appealing option in the
fewer neurocognitive side effects than WBRT. In susceptible pediatric population. With a favor-
a Phase 3 randomized control trial comparing able risk-benefit profile of proton versus photon
the outcomes of SRS alone to those of SRS plus beam therapy, proton therapy appears promising
WBRT, Chang and colleagues demonstrated for the treatment of pediatric CNS metastases as
decreased deficits in learning and memory in the well and will likely become more popular as a
group treated with SRS alone [40]. treatment option as specialized proton centers
To date, the efficacy of SRS for metastases in become more widespread.
the pediatric population has not been reported
outside of case reports; however, this technique
has been evaluated and found to be likely effec- Multimodal Treatment
tive in cases of pediatric arteriovenous malforma-
tions (AVMs) and primary brain tumors, such as Although pediatric solid tumor CNS metastases
juvenile pilocytic astrocytomas (JPAs), recurrent generally confer a grim prognosis, with survival
ependymomas, and pineocytomas [2, 41–43]. in case studies described in months (Table 18.1),
The development of frameless, optically guided there are rare reports of long-term survivors,
stereotactic systems has helped to overcome usually with patients who have received aggres-
many of the difficulties of SRS in the pediatric sive multimodal therapy incorporating surgical
population, such as intolerance of the headframe resection, radiation (often focal combined with
and the risk that movement could result in off- craniospinal), chemotherapy, immunotherapy,
target effects, making this now a much more pal- and/or stem cell transplantation. Osawa and
atable treatment option [42]. colleagues reported two cases of rhabdomyo-
sarcoma achieving disease freedom at 8 and
10 months, respectively (whereas most rhab-
Proton Therapy domyosarcoma CNS cases succumb in under
1 year), through a combination of surgical resec-
Though not yet widely described for use in pedi- tion, radiation to the tumor bed, ifosfamide/
atric CNS metastases, proton therapy has been carboplatin/etoposide (ICE) chemotherapy, and
shown to be effective in both pediatric primary additional CSI and allogenic stem cell transplan-
brain tumors (including astrocytic, embryonal, tation in one of the patients [51]. Hauser and col-
and ependymal tumors) and adult CNS metas- leagues also reported a case of long-term survival
tases [44–46]. Characteristics of proton delivery of 44.8 months after CNS diagnosis in a patient
optimize the risk-benefit profile, particularly for undergoing surgery and receiving radiation, high-
the pediatric population. Compared to photon dose chemotherapy, and stem cell transplantation
therapy, protons can deposit more precisely at a [7]. Notably, this was the patient in their reported
desired depth in the oncologic target, reducing cohort of 14 cases who received the most aggres-
entry and exit doses and thus sparing surrounding sive treatment regimen. Additionally, a few cases
normal tissues and enabling treatment of targets of long-term survivors with CNS osteosarcoma
adjacent to critical structures [45]. Additionally, metastases treated with multimodal therapy have
comparisons of proton therapy-treated primary been reported to survive beyond 5 years (this dis-
pediatric CNS malignancies to historical photon ease is otherwise associated with a 6-month sur-
beam-treated cohorts have shown non-inferiority vival) [4, 27, 52]. Rare long-term survival with
or superiority in local control, progression-free multimodal therapy has also been reported in
cases of CNS metastases from germ cell tumors, Predicting the Occurrence

hepatoblastoma, melanoma, Wilms tumor, clear of Pediatric CNS Metastases
cell sarcoma of the kidney, and neuroblastoma
[4, 11, 16, 53–56]. Croog and colleagues from While rare in incidence, pediatric solid tumor
our center demonstrated a survival advantage for metastases tend to be associated more frequently
CSI and intraventricular radio-immunotherapy with certain primary subtypes and metastatic dis-
in neuroblastoma patients with CNS relapse, ease characteristics. Certain rare primary tumors
postulating that neuroblastoma cells disseminate such as PPB, CCSK, and alveolar soft part sar-
through CSF along the neuraxis, necessitating coma have been reported to have a higher inci-
full craniospinal radiation [55]. Specifically, they dence of CNS metastasis, approximately 25%,
advocate for simultaneous radiation of cranial 5–11%, and 15–29%, respectively [18–20].
and spinal fields to avoid potential reseeding Additionally, although cases of choriocarcinoma
and for treatment with either intrathecal or intra- comprised a small subset of their cohort, Suki and
ventricular delivery of therapeutics or systemic colleagues found that these were associated with
delivery of BBB-penetrating compounds such a 43% incidence of brain metastases, suggesting
as irinotecan or temozolomide. With the advent that this germ cell subtype may have a particu-
of increasingly effective biological therapies for lar predilection for the CNS [3]. Furthermore, as
metastatic disease (such as the combination of pulmonary metastases often predate or co-occur
nivolumab and ipilimumab checkpoint inhibi- with CNS metastasis (as described previously),
tors), multimodal treatment options must always they appear to be a risk factor for CNS disease.
be considered [57]. It remains to be determined whether there
may be a role for CNS screening imaging in
diagnosed pediatric solid tumor cases. This con-
Consideration of Prophylaxis sideration remains controversial, as previous
studies have found a high rate of false positives
In contrast to the adult tumor and pediatric liquid with computerized tomography (CT) imaging
tumor populations, pediatric solid tumors rarely to survey for brain metastases in the melanoma
develop CNS metastases, and thus, prophylaxis population [59]. Risk factors such as primary
is generally not considered and is of unknown tumor type (as those listed above are more neu-
efficacy. Interestingly, in a comparison of Ewing rotropic) and presence of pulmonary or other vis-
sarcoma patients who received CNS prophylaxis ceral metastases should be taken into account in
(n = 92, WBRT and a single dose of intrathecal determining whether CNS screening is warranted
methotrexate) to those who did not (n = 62), Trigg [16]. Although CNS metastasis has generally
and colleagues found no significant difference been considered a late-stage finding, continuing
in the incidence of developing CNS metastases development of novel biological treatments, che-
between the cohorts, suggesting prophylaxis may motherapies, and radiation regimens yields hope
not be effective in preventing CNS spread, at for combatting metastatic disease. With this,
least in this primary tumor type [58]. However, CNS screening should be performed for primary
specific risk factors, such as a breach in the BBB, malignancy subtypes that have an effective sys-
may render CNS prophylaxis warranted in cer- temic therapy, like melanoma [57].
tain cases. As mentioned, Kramer and colleagues
found that prior LP in a neuroblastoma popula-
tion was significantly associated with the devel- Conclusions
opment of CNS metastases; it may be beneficial
to prophylactically treat such cases (undergoing CNS metastases in pediatric solid tumors
LP for primary tumors known to have high risk remain a relatively rare and late-stage occur-
of hematogenous spread and CNS seeding) with rence. This may reflect, in part, the early use
intrathecal chemotherapy [8]. of myeloablative therapy for pediatric primary
tumors, depleting the myeloid precursor pool, or 10. Paulino AC, Nguyen TX, Barker JL Jr. Brain metas-
tasis in children with sarcoma, neuroblastoma,
an increased permeability of the pediatric BBB, and Wilms’ tumor. Int J Radiat Oncol Biol Phys.
thus facilitating delivery of systemic therapy 2003;57(1):177–83.
into the brain. However, certain primary tumor 11. Spunt SL, Walsh MF, Krasin MJ, Helton KJ, Billups
subtypes, as well as the presence of pulmonary CA, Cain AM, et al. Brain metastases of malignant
germ cell tumors in children and adolescents. Cancer.
metastases, are associated with increased inci- 2004;101(3):620–6.
dence of CNS pathology in the pediatric pop- 12. Tasdemiroglu E, Patchell RA. Cerebral metasta-

ulation. Previous studies have suggested that ses in childhood malignancies. Acta Neurochir.
aggressive multimodal therapy may confer a 1997;139(3):182–7.
13. Graus F, Walker RW, Allen JC. Brain metastases in
survival advantage. Taken together, we propose children. J Pediatr. 1983;103(4):558–61.
that careful screening of select cases with risk 14. Vannucci RC, Baten M. Cerebral metastatic disease in
factors for CNS metastasis, particularly in cer- childhood. Neurology. 1974;24(10):981–5.
tain tumor subtypes and those with effective 15. Klos KJ, O’Neill BP. Brain metastases. Neurologist.
2004;10(1):31–46.
therapies available for metastatic disease, may 16. Rodriguez-Galindo C, Pappo AS, Kaste SC, Rao BN,
enable better outcomes. Cain A, Jenkins JJ, et al. Brain metastases in children
with melanoma. Cancer. 1997;79(12):2440–5.
17. Kumar R, Liu APY, Orr BA, Northcott PA, Robinson
GW. Advances in the classification of pediatric
References brain tumors through DNA methylation profiling:
from research tool to frontline diagnostic. Cancer.
1. Nayak L, Lee EQ, Wen PY. Epidemiology of brain 2018;124(21):4168–80.
metastases. Curr Oncol Rep. 2012;14(1):48–54. 18. Seibel NL, Li S, Breslow NE, Beckwith JB, Green
2. Sreeraman Kumar R, Rotondo RL. Pediatric brain DM, Haase GM, et al. Effect of duration of treat-
metastases. In: Mahajan A, Paulino A, editors. ment on treatment outcome for patients with
Radiation oncology for pediatric CNS tumors. Cham: clear-cell sarcoma of the kidney: a report from the
Springer International Publishing; 2018. p. 393–410. National Wilms’ Tumor Study Group. J Clin Oncol.
3. Suki D, Khoury Abdulla R, Ding M, Khatua S, Sawaya 2004;22(3):468–73.
R. Brain metastases in patients diagnosed with a solid 19. Priest JR, Magnuson J, Williams GM, Abromowitch
primary cancer during childhood: experience from M, Byrd R, Sprinz P, et al. Cerebral metastasis
a single referral cancer center. J Neurosurg Pediatr. and other central nervous system complications of
2014;14(4):372–85. pleuropulmonary blastoma. Pediatr Blood Cancer.
4. Wiens AL, Hattab EM. The pathological spectrum of 2007;49(3):266–73.
solid CNS metastases in the pediatric population. J 20. Wang CH, Lee N, Lee LS. Successful treatment for
Neurosurg Pediatr. 2014;14(2):129–35. solitary brain metastasis from alveolar soft part sar-
5. Kebudi R, Ayan I, Gorgun O, Agaoglu FY, Vural S, coma. J Neuro-Oncol. 1995;25(2):161–6.
Darendeliler E. Brain metastasis in pediatric extra- 21. Gao D, Mittal V. The role of bone-marrow-derived
cranial solid tumors: survey and literature review. J cells in tumor growth, metastasis initiation and pro-
Neuro-Oncol. 2005;71(1):43–8. gression. Trends Mol Med. 2009;15(8):333–43.
6. Stefanowicz J, Izycka-Swieszewska E, Szurowska E, 22. Matthay KK, Villablanca JG, Seeger RC, Stram DO,
Bien E, Szarszewski A, Liberek A, et al. Brain metas- Harris RE, Ramsay NK, et al. Treatment of high-risk
tases in paediatric patients: characteristics of a patient neuroblastoma with intensive chemotherapy, radio-
series and review of the literature. Folia Neuropathol. therapy, autologous bone marrow transplantation, and
2011;49(4):271–81. 13-cis-retinoic acid. Children’s Cancer Group. N Engl
7. Hauser P, Jakab Z, Lang O, Kondas O, Muller J, J Med. 1999;341(16):1165–73.
Schuler D, et al. Incidence and survival of central 23. Berthold F, Boos J, Burdach S, Erttmann R, Henze
nervous system involvement in childhood malignan- G, Hermann J, et al. Myeloablative megatherapy with
cies: Hungarian experience. J Pediatr Hematol Oncol. autologous stem-cell rescue versus oral maintenance
2005;27(3):125–8. chemotherapy as consolidation treatment in patients
8. Kramer K, Kushner B, Heller G, Cheung with high-risk neuroblastoma: a randomised con-
NK. Neuroblastoma metastatic to the central ner- trolled trial. Lancet Oncol. 2005;6(9):649–58.
vous system. The Memorial Sloan Kettering Cancer 24. Sul J, Posner JB. Brain metastases: epidemiology

Center experience and a literature review. Cancer. and pathophysiology. Cancer Treat Res. 2007;136:
2001;91(8):1510–9. 1–21.
9. Bouffet E, Doumi N, Thiesse P, Mottolese C, Jouvet 25. Laningham FH, Kun LE, Reddick WE, Ogg RJ,

A, Lacroze M, et al. Brain metastases in children with Morris EB, Pui CH. Childhood central nervous sys-
solid tumors. Cancer. 1997;79(2):403–10. tem leukemia: historical perspectives, current therapy,
and acute neurological sequelae. Neuroradiology. 39. Selek U, Chang EL, Hassenbusch SJ 3rd, Shiu AS,
2007;49(11):873–88. Lang FF, Allen P, et al. Stereotactic radiosurgical
26. Deutsch M, Wollman MR. Radiotherapy for metas- treatment in 103 patients for 153 cerebral mela-
tases to the mandible in children. J Oral Maxillofac noma metastases. Int J Radiat Oncol Biol Phys.
Surg. 2002;60(3):269–71. 2004;59(4):1097–106.
27. Marina NM, Pratt CB, Shema SJ, Brooks T, Rao B, 40. Chang EL, Wefel JS, Hess KR, Allen PK, Lang

Meyer WH. Brain metastases in osteosarcoma. Report FF, Kornguth DG, et al. Neurocognition in patients
of a long-term survivor and review of the St. Jude with brain metastases treated with radiosurgery or
Children’s Research Hospital experience. Cancer. radiosurgery plus whole-brain irradiation: a ran-
1993;71(11):3656–60. domised controlled trial. Lancet Oncol. 2009;10(11):
28. Kramer K, Smith M, Souweidane MM. Safety profile 1037–44.
of long-term intraventricular access devices in pediat- 41. Saran FH, Baumert BG, Khoo VS, Adams EJ, Garre
ric patients receiving radioimmunotherapy for central ML, Warrington AP, et al. Stereotactically guided
nervous system malignancies. Pediatr Blood Cancer. conformal radiotherapy for progressive low-grade
2014;61(9):1590–2. gliomas of childhood. Int J Radiat Oncol Biol Phys.
29. Aherne NJ, Murphy BM. Radiation-induced gliomas. 2002;53(1):43–51.
Crit Rev Oncog. 2018;23(1–2):113–8. 42. Keshavarzi S, Meltzer H, Ben-Haim S, Newman CB,
30. Soon YY, Tham IW, Lim KH, Koh WY, Lu
Lawson JD, Levy ML, et al. Initial clinical experience
JJ. Surgery or radiosurgery plus whole brain radio- with frameless optically guided stereotactic radiosur-
therapy versus surgery or radiosurgery alone for brain gery/radiotherapy in pediatric patients. Childs Nerv
metastases. Cochrane Database Syst Rev. 2014;(3): Syst. 2009;25(7):837–44.
CD009454. 43. Kano H, Yang HC, Kondziolka D, Niranjan A, Arai
31. Sahgal A, Aoyama H, Kocher M, Neupane B, Collette Y, Flickinger JC, et al. Stereotactic radiosurgery
S, Tago M, et al. Phase 3 trials of stereotactic radio- for pediatric recurrent intracranial ependymomas. J
surgery with or without whole-brain radiation ther- Neurosurg Pediatr. 2010;6(5):417–23.
apy for 1 to 4 brain metastases: individual patient 44. Gondi V, Yock TI, Mehta MP. Proton therapy for pae-
data meta-analysis. Int J Radiat Oncol Biol Phys. diatric CNS tumours - improving treatment-related
2015;91(4):710–7. outcomes. Nat Rev Neurol. 2016;12(6):334–45.
32. Gobel U, von Kries R, Teske C, Schneider DT,
45. Kirkpatrick JP, Laack NN, Halasz LM, Minniti G,
Beyerlein A, Bernbeck B, et al. Brain metastases Chan MD. Proton therapy for brain metastases: a
during follow-up of children and adolescents with question of value. Int J Radiat Oncol Biol Phys.
extracranial malignant germ cell tumors: risk adapted 2018;101(4):830–2.
management decision tree analysis based on data of 46. Sreeraman R, Indelicato DJ. Proton therapy for the
the MAHO/MAKEI-registry. Pediatr Blood Cancer. treatment of children with CNS malignancies. CNS
2013;60(2):217–23. Oncol. 2014;3(2):149–58.
33. Postovsky S, Ash S, Ramu IN, Yaniv Y, Zaizov
47. Miralbell R, Lomax A, Russo M. Potential role of
R, Futerman B, et al. Central nervous system proton therapy in the treatment of pediatric medul-
involvement in children with sarcoma. Oncology. loblastoma/primitive neuro-ectodermal tumors: spi-
2003;65(2):118–24. nal theca irradiation. Int J Radiat Oncol Biol Phys.
34. Parasuraman S, Langston J, Rao BN, Poquette CA, 1997;38(4):805–11.
Jenkins JJ, Merchant T, et al. Brain metastases in 48. St Clair WH, Adams JA, Bues M, Fullerton BC,

pediatric Ewing sarcoma and rhabdomyosarcoma: the La Shell S, Kooy HM, et al. Advantage of protons
St. Jude Children’s Research Hospital experience. J compared to conventional X-ray or IMRT in the
Pediatr Hematol Oncol. 1999;21(5):370–7. treatment of a pediatric patient with medulloblas-
35. Lowis SP, Foot A, Gerrard MP, Charles A, Imeson J, toma. Int J Radiat Oncol Biol Phys. 2004;58(3):
Middleton H, et al. Central nervous system metastasis 727–34.
in Wilms’ tumor: a review of three consecutive United 49. Yuh GE, Loredo LN, Yonemoto LT, Bush DA,

Kingdom trials. Cancer. 1998;83(9):2023–9. Shahnazi K, Preston W, et al. Reducing toxicity
36. Weyl-Ben Arush M, Stein M, Perez-Nachum M,
from craniospinal irradiation: using proton beams to
Dale J, Babilsky H, Zelnik N, et al. Neurologic treat medulloblastoma in young children. Cancer J.
complications in pediatric solid tumors. Oncology. 2004;10(6):386–90.
1995;52(2):89–92. 50. Jimenez RB, Sethi R, Depauw N, Pulsifer MB,

37. Baram TZ, van Tassel P, Jaffe NA. Brain metastases in Adams J, McBride SM, et al. Proton radiation ther-
osteosarcoma: incidence, clinical and neuroradiologi- apy for pediatric medulloblastoma and supratento-
cal findings and management options. J Neurooncol. rial primitive neuroectodermal tumors: outcomes
1988;6(1):47–52. for very young children treated with upfront chemo-
38. Brown PD, Brown CA, Pollock BE, Gorman DA, therapy. Int J Radiat Oncol Biol Phys. 2013;87(1):
Foote RL. Stereotactic radiosurgery for patients with 120–6.
“radioresistant” brain metastases. Neurosurgery. 51. Osawa S, Kumabe T, Saito R, Sonoda Y, Niizuma
2008;62(Suppl 2):790–801. H, Watanabe M, et al. Infratentorial brain metas-
tases of pediatric non-epithelial malignant tumors: in high-risk neuroblastoma. Int J Radiat Oncol Biol
three case reports. Brain Tumor Pathol. 2011;28(2): Phys. 2010;78(3):849–54.
167–74. 56. Kellie SJ, Hayes FA, Bowman L, Kovnar EH,

52.
Wexler LH, DeLaney TF, Saris S, Horowitz Langston J, Jenkins JJ 3rd, et al. Primary extracranial
ME. Long-term survival after central nervous sys- neuroblastoma with central nervous system metasta-
tem relapse in a patient with osteosarcoma. Cancer. ses characterization by clinicopathologic findings and
1993;72(4):1203–8. neuroimaging. Cancer. 1991;68(9):1999–2006.
53.
Robertson PL, Muraszko KM, Axtell RA. 57. Wolchok JD, Kluger H, Callahan MK, Postow MA,
Hepatoblastoma metastatic to brain: prolonged sur- Rizvi NA, Lesokhin AM, et al. Nivolumab plus ipi-
vival after multiple surgical resections of a solitary limumab in advanced melanoma. N Engl J Med.
brain lesion. J Pediatr Hematol Oncol. 1997;19(2): 2013;369(2):122–33.
168–71. 58. Trigg ME, Makuch R, Glaubiger D. Actuarial risk
54. Radulescu VC, Gerrard M, Moertel C, Grundy PE, of isolated CNS involvement in Ewing’s sarcoma
Mathias L, Feusner J, et al. Treatment of recur- following prophylactic cranial irradiation and intra-
rent clear cell sarcoma of the kidney with brain thecal methotrexate. Int J Radiat Oncol Biol Phys.
metastasis. Pediatr Blood Cancer. 2008;50(2): 1985;11(4):699–702.
246–9. 59. Buzaid AC, Tinoco L, Ross MI, Legha SS, Benjamin
55. Croog VJ, Kramer K, Cheung NK, Kushner BH,
RS. Role of computed tomography in the staging of
Modak S, Souweidane MM, et al. Whole neuraxis patients with local-regional metastases of melanoma.
irradiation to address central nervous system relapse J Clin Oncol. 1995;13(8):2104–8.
Part III
Radiation Therapy of CNS Metastasis
Basic Radiobiology and Radiation
Physics Primer
19
Emily S. Lebow, Marc R. Bussière,
and Helen A. Shih
Radiobiology translocations) that are propagated in future cel-

lular divisions.
The DNA-Damage Response Radiation-induced DNA damage, particularly
double-strand breaks, may be so deleterious to
Radiation therapy treats cancer through damag- cellular function that cell death results. One form
ing DNA, exerting its most potent effects through of cell death following radiation therapy is
double-strand DNA breaks which are challeng- mitotic catastrophe, which occurs as a cell
ing to accurately repair. Cells have a complex set attempts to divide in the presence of significant
of mechanisms to detect and repair DNA damage chromosomal aberration. Cells may also activate
collectively called the DNA-damage response a highly organized cell death in response to
[1–3]. Some types of DNA damage are more irreparable DNA damage, termed apoptosis.
readily repairable, including damage to individ- During apoptosis, cellular content including
ual DNA bases or single-strand DNA breaks. In DNA is divided into membranous apoptotic bod-
these cases, the opposite, intact strand of DNA ies which are digested by phagocytes to prevent
serves as a template. This allows for a high-fidel- leakage of damaging cellular proteins. A third
ity repair through processes including base exci- theory of cell death following ionizing radiation
sion repair, nucleotide excision repair, and is the bystander effect, in which cancerous cells
mismatch repair [4]. In contrast, homologous are killed due to irradiation of neighboring cells.
recombination and non-homologous end joining Irradiated cells may release danger signals or
are both utilized to repair double-strand breaks. other cytotoxic molecules inducing cell death of
However, these mechanisms are error-prone with nearby cells that were not directly irradiated.
potential for mutations (including deletions and Mitotic catastrophe, apoptosis, and the bystander
effect may induce cell death weeks to months
following irradiation [1].
E. S. Lebow
Inherited mutations in the DNA-damage
Department of Medicine, Memorial Sloan Kettering
Cancer Center, New York, NY, USA response are associated with a profound cancer
predisposition, demonstrating the extraordinary
M. R. Bussière
Department of Radiation Oncology, Massachusetts role of DNA repair on organismal function
General Hospital, Boston, MA, USA (Table 19.1). Xeroderma pigmentosum (XP) is
H. A. Shih (*) caused by deficient nucleotide excision repair.
Central Nervous System & Eye Services, Department XP is associated with marked ultraviolet r adiation
of Radiation Oncology, Massachusetts General sensitivity, such as to sunlight. Individuals with
Hospital, Boston, MA, USA
XP have a significant cancer risk in sun-exposed
e-mail: hshih@mgh.harvard.edu

https://doi.org/10.1007/978-3-030-42958-4_19
272 E. S. Lebow et al.
Table 19.1 Cancer syndromes secondary to inherited defects in DNA repair

Genetic
etiology Functional deficiency Cancer phenotype
Ataxia telangiectasia ATM Signal transduction Lymphoma, breast
BRCA1 and BRCA2, DNA associated BRCA1, BRCA2 Homologous Breast, ovarian
recombination
Hereditary non-polyposis colorectal MSH2, MSH3, Mismatch repair Colon, gastric, gynecologic
cancer (Lynch syndrome) MSH6 deficiency
Li-Fraumeni syndrome p53 Signal transduction Sarcoma, breast, brain
Xeroderma pigmentosum XPA-XPG Nucleotide excision Melanoma and non-
repair melanoma skin cancer
areas, including a 10,000-fold increase in risk for apoptosis in the setting of radiation-induced
non-melanoma skin cancer, 2000-fold increased DNA damage [14]. While there are likely a
risk of melanoma skin cancer, and a 3000-fold complex set of mechanisms contributing to a
increased risk in intraoral cancers, most com- radioresistant phenotype, evidence suggests
monly on the tip of the tongue and the dorsal that DNA damage repair is an important com-
tongue [5, 6]. Inherited mutations in BRCA1 or ponent to this phenomenon.
BRCA2 result in impaired homologous recombi-
nation and confer an increased risk of breast,
uterus, ovarian, fallopian tube, prostate, and pan- Tumor Hypoxia
creatic cancers [7, 8]. Hereditary non-polyposis
colorectal cancer (Lynch syndrome) is caused by Tumor hypoxia is a common feature of solid
deficient mismatch repair and confers a predispo- tumors. Rapidly proliferating tumors frequently
sition to colorectal, gastric, endometrial, and outgrow their neovascular supply, which tends to be
ovarian cancers [9]. Inherited mutations in p53 chaotic and poorly developed. The result is a state
and ATM, both critical components of signal of diffusion-limited chronic hypoxia in which some
transduction, cell cycle regulation, and promot- cells are too far away from the vasculature to be
ing conditions for repair of DNA and other cel- adequately oxygenated. Hypoxia may also develop
lular damage, also result in hereditary cancer acutely, as blood vessels are temporarily obstructed
predispositions [10–12]. or have variable blood flow, resulting in perfusion-
It is hypothesized that robust DNA repair limited hypoxia [15]. Tumor hypoxia is associated
capacity also confers resistance to the effects with radioresistance, requiring up to three times
of ionizing radiation, providing a potential the dose of radiation to achieve the equivalent cell-
explanation for why some cancer subpopula- killing effect. Hypoxic radioresistance is due to
tions are more radioresistant. High-grade glio- inefficient production of reactive oxygen species,
mas are aggressive tumors that have a a dominant mechanism of radiation-induced DNA
propensity to recur even with adjuvant radia- damage. Furthermore, in response to a hypoxic
tion therapy [13]. Recurrence is hypothesized microenvironment, cells activate hypoxia-induc-
to be secondary to subpopulations of glioma ible transcription factors to promote anaerobic
stem cells in which cell cycle checkpoints are metabolism, invasion, and angiogenesis. These
readily activated following radiation-induced changes result in a more aggressive and radioresis-
DNA damage, halting division, creating an tant phenotype [16]. Hypoxia within solid tumors
effective DNA repair, and conferring a net rela- is heterogeneous, with some areas of the tumor
tive resistance to radiation therapy. In contrast, well oxygenated (e.g., cells near the blood ves-
glioma cells with less robust pathways sup- sels) and others hypoxic. Hypoxic niches within
porting DNA repair are more likely to undergo solid tumors support radioresistant cancer stem
19 Basic Radiobiology and Radiation Physics Primer 273
cells that can repopulate the tumor following irra- Accelerated Repopulation
diation, accounting for local recurrence after treat-
ment [17]. Accelerated repopulation is an important source
When the total dose of radiation is divided into of treatment failure following radiation therapy.
many smaller doses delivered over days or weeks During the course of fractionated radiation ther-
(termed fractionation), oxygenation may be apy, surviving tumor cells may proliferate,
improved, thereby increasing the efficacy of radi- replacing tumor cells already killed by radiation
ation therapy. Reoxygenation between fractions therapy. The rate of tumor cell repopulation
of radiation therapy may occur quickly via reper- accelerates during the course of radiation ther-
fusion through vessels that were temporarily apy, with growth occurring increasingly quickly
closed, reoxygenating the tumor within minutes. after the start of radiation therapy. Accelerated
Reoxygenation may also occur over days or repopulation has been observed in several types
weeks as tumor cells die secondary to effects of of malignancy, including head and neck squa-
radiation therapy, thereby shrinking the size of the mous cell carcinoma, cervical cancer, and blad-
tumor and decreasing the distance between blood der cancer [18]. Because of accelerated
vessels and surviving tumor cells. These mecha- repopulation, prolonging overall treatment time
nisms of reoxygenation provide a fundamental results in a larger number of tumor cells that need
rationale for fractionating radiation delivery. to be eradicated to achieve local control. This
necessitates a higher total dose of radiation ther-
apy. One method to counter the effects of repopu-
Cell Cycle and Redistribution lation is accelerated radiotherapy, where radiation
is delivered over fewer days giving tumor cells
The cell cycle is comprised of a series of phases as less time to repopulate. However, this strategy
it goes through its process of growth, replication, also reduces the opportunity for normal tissue
and division to produce two daughter cells. repair from radiation injury, thereby increasing
Radiosensitivity varies depending on the phase of risk of toxicity of treatment. It also leaves less
the cell cycle. Cells during the phases of mitosis opportunity for tumor cell reoxygenation and
and G2 (the gap immediately prior to mitosis) are transition through the phases of the cell cycle,
more radiosensitive. In contrast, cells during S both of which increase radiosensitivity.
phase (DNA duplication) are more radioresistant.
Tumor cells grow in an asynchronous manner,
present at various phases of the cell cycle at any Radiosensitizers
point in time. The first few fractions of a course of
radiation therapy will preferentially kill cells in the Radiosensitizing drugs decrease the proportion
radiosensitive phases (mitosis and G2) while kill- of cells that survive after radiation therapy
ing fewer cells in radioresistant phases (S phase). (Fig. 19.1). There are multiple mechanisms by
However, surviving cells will continue to transi- which classes of systemic therapies increase the
tion through the cell cycle and will redistribute to efficacy of radiation therapy. Many classes of
more sensitive phases of the cell cycle during sub- systemic therapies may do so through more than
sequent fractions of a course of radiation therapy. one mechanism (Tables 19.2 and 19.3).
Cells in the S phase during the first fraction of Antimetabolites increase radiation-associated
radiation therapy may transition to the G2 or M DNA damage by incorporating itself into the
phase for a subsequent fraction, thereby becoming cell’s DNA and by inhibiting DNA repair pro-
more sensitive to the effects of ionizing radiation. cesses, thereby making the cell more susceptible
As a result, fractionated radiation therapy is con- to DNA damage [19]. The antimetabolite
sidered to be more effective than delivery of a 5-fluorouracil blocks the synthesis of the pyrimi-
single, large dose in some circumstances [1]. dine thymidine (a nucleoside required for DNA
1.0 Table 19.3 Mechanisms of drug-radiation interactions

Mechanism Selected examples
Increased radiation- Antimetabolites,
Radioresistant associated DNA damage alkylators, platinum
Surviving fraction
and/or impaired DNA repair agents, topoisomerase

inhibitors
0.1
Redistribution of cells to Antimetabolites,
more radiosensitive phases topoisomerase inhibitors
of the cell cycle
Radiosensitive Cytostatic agents that reduce Molecularly targeted
accelerated repopulation agents, most classical
after radiation therapy chemotherapy
Tumor microvasculature VEGF-targeted agents
2 4 6 8 10 12 normalization that reduces
Radiation therapy dose Gray (j/kg) hypoxia-associated
radioresistance
Fig. 19.1 Radiosensitization decreases the proportion of Release of immunogenic Immune checkpoint
surviving cells for any given dose of radiation therapy tumor antigens, induction of inhibitors
resulting in a leftward shift of the curve. Radioresistance, pro-inflammatory signaling
in contrast, results in a greater proportion of surviving pathways
cells for any given dose of radiation, resulting in a right-
ward shift of the curve. Many factors are known to impact
sensitivity or resistance of cells to radiation therapy, undergo cell death. The addition of 5-fluorouracil
including radiosensitizing drugs, hypoxia in the tumor to radiation therapy has been demonstrated to
microenvironment, cell cycle phase, and DNA repair improve survival in squamous cell carcinoma of
capacity
the head and neck compared to radiation therapy
alone [20].
Table 19.2 Classes of systemic therapy Temozolomide is an alkylating agent with
Selected examples greater efficacy in the subset of glioblastoma
Antimetabolites 5-Fluorouracil patients with epigenetic silencing of O-6-
Gemcitabine methylguanine-DNA methyltransferase
Pemetrexed (MGMT), a gene critical in DNA repair. A ran-
Alkylators Cyclophosphamide domized controlled trial demonstrated that the
Temozolomide
addition of temozolomide to radiation therapy
Platinum agents Cisplatin
Carboplatin
preferentially improved survival among glio-
Microtubule stabilizers Vincristine blastoma patients with MGMT silencing versus
Docetaxel patients without MGMT silencing (21.7 months
Topoisomerase inhibitors Etoposide versus 15.3 months, p = 0.007). The MGMT
Doxorubicin protein removes temozolomide-induced methyl
Molecularly targeted agents Bevacizumab (VEGF) groups from the DNA base guanine to prevent
Trastuzumab
errors during DNA transcription and replica-
(HER2-neu)
Gefitinib (EGFR) tion. MGMT promoter methylation causes
Immunotherapeutic checkpoint Pembrolizumab MGMT silencing, and those tumors harboring
inhibitors (PD-1) MGMT promoter hypermethylation are thus
Nivolumab (PD-1) more sensitive to the effects of concurrent radi-
Ipilimumab (CTLA-4) ation therapy and temozolomide [21, 22].
There are several additional mechanisms of
replication) through inhibition of thymidylate radiosensitizing drugs. Any therapy that slows or
synthase. Without thymidine, DNA replication or halts proliferation of cancer cells will mitigate
repair from radiation-associated damage is accelerated repopulation of tumors associated
impaired and preferentially affects cancer cells with radiation therapy [19]. This includes many
that are actively dividing and thus preferentially classes of systemic therapy, including m
olecularly
targeted agents. VEGF-targeted systemic ther- fibroblasts, vascular endothelial cells, and macro-
apy normalizes chaotic, tortuous, and dilated phages. Ionizing radiation induces pro-fibrosis
tumor neovasculature. This process of vascular signaling and growth factor cascades (such as
normalization improves tumor oxygenation, TGFβ) that result in progressive deposition of
thereby reducing hypoxic radioresistance [23]. extracellular matrix and collagen. Fibrosis is
Interaction between radiation therapy and associated with organ loss of function (such as
enhancement of the immune system is an area of bowel malabsorption) as well as other symptoms
active investigation, with some successful such as pain, neuropathy, decreased strength, and
approaches either enhancing tumor cell recogni- reduced joint range of motion. Higher radiation
tion by increased release of tumor-associated doses decrease the latency between irradiation
antigens and other approaches enhancing pro- and the onset of late toxicity [1].
inflammatory signals [24]. Radiation exposure also harbors a low but real
risk of developing a secondary primary cancer, a
distinct type of late adverse effect. Ionizing radia-
Normal Tissue Side Effects tion can cause genomic instability or the acquisi-
tion of a mutator phenotype through mutations in
The goal of radiation therapy is to deliver highly genes critical for efficient and high-fidelity DNA
conformal radiation that maximizes dose deliv- replication and repair. As a result, cells will more
ered to the tumor while minimizing toxicity to readily acquire additional mutations that are
normal tissues. Modern advances in radiation oncogenic and may result in transformation to a
therapy have resulted in significant strides toward malignancy [26]. The risk of secondary malig-
achieving this objective. Nonetheless, a number nancy is inversely related to the age at treatment,
of early and late radiation-related toxicities are with patients treated in childhood at the highest
still commonly observed in patients treated with risk of secondary malignancy. For example, chil-
radiation therapy. dren treated with cranial irradiation have between
Early radiation effects occur during or within an 8.1 and 52.3 times higher risk of developing
weeks of radiation therapy in highly proliferative subsequent central nervous system malignancies
tissues and are likely to be reversible. Proliferative compared to the general populations [27]. There
tissues have a precise balance between cell loss are other risk-modifying factors as well, includ-
and cell production. Radiation impairs cell divi- ing the site of irradiation, the size of the radiation
sion while accelerating the rate of cell loss, dis- field, radiation dose, systemic therapy, and envi-
rupting this equilibrium. Examples of common ronmental exposures, such as cigarette smoking
early toxicities include mucositis of the upper [28].
and lower gastrointestinal tracts, bone marrow
hypoplasia, and hair loss. With cessation of radia-
tion treatment, the balance between cell replica- Radiation Physics
tion and cell loss is eventually restored. Tissues
heal and return to essentially normal function Ionizing Radiation
with few or no long-term sequelae of treatment.
Late radiation effects, in contrast, occur within Therapeutic radiation therapy takes advantage
months or years of radiation therapy and tend to of the portion of the electromagnetic spectrum
be permanent and progressive. Historically late with sufficient energy capable of creating ion-
radiation effects were thought to occur secondary izations. Radiowaves and microwaves represent
to functional deficiency caused by depletion of low energy, whereas visible light represents
organ parenchymal cells, called the target-cell mid-energy range of the electromagnetic spec-
hypothesis [25]. However, more recent under- trum. X-rays are in the high-energy range of the
standings of late toxicity incorporate complex electromagnetic spectrum and are capable of
interactions between organ parenchymal cells, ejecting orbital electrons and thereby creating
highly reactive oxygen species which can dam- thereby leaving behind an ion (an atom with a net
age DNA. This form of electromagnetic radia- positive electric charge, from which the desig-
tion – high-energy X-rays that are capable of nated ionizing radiation originates). The photon,
ionizing atoms – is utilized in therapeutic radia- now decreased in energy, scatters and continues
tion. The dose of radiation used to treat cancer is to travel through the material along an altered
measured in Gray, representing the energy pathway. The ejected electron gains energy from
deposited by ionizing radiation per unit mass of its collision with the photon. Cells are mostly
the material [29]. comprised of water, and therefore photons are
Linear energy transfer (LET) describes the most likely to interact with water molecules and
amount of energy that is transferred from an produce reactive oxygen species, most impor-
energy source to another material, measured by tantly the hydroxyl radical. Reactive oxygen spe-
the amount of ionizations created. LET depends cies are highly electron-affinic and damage
on the radiation source. Photons and protons pro- nearby DNA, producing DNA base damage, sin-
duce few ionizations and are low LET, whereas gle-strand breaks, and double-strand breaks. As
neutrons and carbon are examples of radiation previously discussed, the double-strand break is
with high-density ionizations and are high most difficult to repair and deleterious to cellular
LET. In general, high-LET radiation is more bio- function.
logically effective (produces more DNA damage) There are two primary sources of photons uti-
per dose of radiation than low-LET radiation. lized for radiation therapy: radioactive decay and
This concept is quantified by relative biological linear acceleration. Radioactive decay is the pro-
effectiveness (RBE). As LET increases, RBE cess by which elements with unstable nuclei emit
also increases in a nonlinear manner until reach- energy, thereby increasing their nuclear stability.
ing a maximum (at approximately 100 keV/μm) Cobalt-60 has been a key source of therapeutic
after which point RBE decreases with increasing radiation for many decades. Cobalt-60 undergoes
LET. Photons produced by cobalt-60 are consid- beta decay to produce high-energy photons called
ered to have a low LET and are an established gamma rays, in the process becoming Nickel-60.
reference with an RBE of 1. Protons have an Linear accelerators are now the most common
RBE of 1.1. Thus, for a given dose of radiation, source of high-energy photons. Linear accelera-
protons have a 10% greater biological effect than tors use microwaves to accelerate electrons that
photons. In fact, this is an oversimplification as collide with a heavy metal target to produce high-
the RBE of protons and all other particle beam energy X-rays, interchangeably called photons.
profiles is variable, but the estimation of 1.1 is the This process is called bremsstrahlung, “braking
currently accepted convention until its effects are radiation” or “decelerating radiation,” since the
better understood. However, the net effect of photon is produced by the braking or decelera-
radiation includes the fractionation (how many tion of a high-energy electron as it collides with a
sessions the radiation course is delivered in and heavy metal nucleus. The kinetic energy of the
the dose given with each treatment) and the electron is converted into radiation in the form of
intrinsic tumor radiosensitivity [30]. a high-energy X-ray. The only difference between
the X-ray and gamma ray is the means of produc-
tion, with gamma rays produced from radioactive
Photon Radiation isotope decay and X-rays produced from electron
collisions. All photon beams, whether gamma
Photons are the most common source of radiation rays produced from radioactive decay or X-rays
therapy utilized for treatment of cancer. In the produced from a man-machine, must be precisely
setting of radiation therapy, photons interact with regulated for clinical use to deliver a desired dose
matter predominantly via the Compton effect. As to a desired target [29].
a high-energy photon travels though matter, it As photons move through the patient, energy
eventually collides with an orbital electron. The is deposited in a characteristic pattern called a
collision ejects an electron from its atomic orbit, depth-dose curve (Fig. 19.2). Within the first few
Normal tissue
Normal tissue proximal Tumor distal to the
to the tumor tumor
100
Photon 6 MV
Proton peaks 149, 158,
Relative dose (%)
75 178 mEV
Proton Spread Out Bragg
Peak (SOBP)
50
Photon
25 exit
dose
0
0 4 8 12 16
Depth in tissue (cm)
Fig. 19.2 Comparison of relative depth-dose distribution proton dose deposition. Protons also deposit energy dur-
of photons versus protons. The blue line demonstrates the ing the entire pathway, but travel only a finite distance and
photon’s deposition of energy as a function of depth. After rapidly increase in dose deposition near its end of range,
entering the patient, photon dose deposition increases and reaching a maximum dose transfer at the Bragg peak fol-
peaks and then steadily decreases. Because photons lowed by an extremely rapid drop-off in energy. Multiple
deposit energy during its entire pathway, tissue both prox- proton beams are summed together to form the spread-out
imal to tumor and distal is exposed to relatively high radi- Bragg peak (SOBP) to provide coverage to the entire
ation dose. The red lines demonstrate three examples of depth of the tumor
millimeters of entering the patient, photon dose levels using magnetic fields [11]. Protons have a
deposition increases and peaks and then steadily distinct pattern of energy deposition as they
decreases in energy deposition until exiting the travel through tissue. Protons enter tissue with a
body. Because photons deposit energy during high energy but relative low dose deposition.
their entire pathway through the body, tissue both Protons decelerate quickly and eventually stop.
proximal to the treatment target and distal to the Just before reaching its end of range, the proton
target inevitably receives some radiation dose. beam transfers the great bolus of dose known as
The shape of the depth-dose curve is dependent the Bragg peak and then abruptly stops. By mod-
on the photon energy. Higher-energy photons ulating the energy of protons, a spread-out Bragg
more effectively spare the skin but have less peak (SOBP) can be generated to treat a range of
attenuation of dose deposition after passing depths. Relative to photons, protons offer the
through the target, increasing the dose deposition advantage of decreased dose to both proximal
distal to the target. Despite advances in radiation and distal normal tissues within a given beam,
delivery techniques that allow high-precision tar- offering a theoretic advantage of reduced toxic-
geting of the intended volume, dose deposition ity including risk of secondary malignancy. Dose
distal to the target is a physical limitation of the reduction to normal adjacent tumor also allows
photon beam. This dose to nontarget tissue confor dose escalation to the tumor, potentially
tributes to radiation-associated toxicities [31]. improving disease control [31].
The dosimetric superiority of proton therapy
relative to photon therapy is well-established,
Proton Radiation but whether there is also a clinical benefit is only
widely recognized in a few conditions and other-
Proton radiation consists of hydrogen atoms wise largely remains an area of active inquiry for
(composed of one proton and one electron) most types of benign and malignant diagnoses.
stripped of their electron. Most common proton Widely recognized uses of proton therapy include
accelerators are cyclotrons or synchrotrons many childhood malignancies such as medullo-
which accelerate protons to therapeutic energy blastoma, ependymoma, craniopharyngioma, and
rhabdomyosarcoma. Because many of these pedi- 5. Bradford PT, Goldstein AM, Tamura D, Khan SG,
atric diseases are highly curable, these children are Ueda T, Boyle J, et al. Cancer and neurologic degen-
eration in xeroderma pigmentosum: long term follow-
particularly susceptible to late effects of radiation up characterises the role of DNA repair. J Med Genet.
therapy that may develop over many decades such 2011;48(3):168–76.
as secondary malignancy, endocrinopathy, and 6. Mahindra P, DiGiovanna JJ, Tamura D, Brahim JS,
cognitive dysfunction. Several studies suggest a Hornyak TJ, Stern JB, et al. Skin cancers, blindness,
and anterior tongue mass in African brothers. J Am
decreased risk of secondary malignancy among Acad Dermatol. 2008;59(5):881–6.
children treated with proton radiation compared 7. Antoniou A, Pharoah PD, Narod S, Risch HA, Eyfjord
with photon radiation [32, 33]. Proton therapy is JE, Hopper JL, et al. Average risks of breast and ovar-
also commonly utilized for ocular melanoma, ian cancer associated with BRCA1 or BRCA2 muta-
tions detected in case series unselected for family
skull base malignancies, and sinonasal malignan- history: a combined analysis of 22 studies. Am J Hum
cies. In contrast, the uncertain benefit of proton Genet. 2003;72(5):1117–30.
therapy for other disease sites remains an area of 8. Thompson D, Easton DF, Breast Cancer Linkage
active investigation, including for prostate cancer Consortium. Cancer incidence in BRCA1 mutation
carriers. J Natl Cancer Inst. 2002;94(18):1358–65.
(NCT01617161, NCT01352429) and breast cancer 9. Terdiman JP. Colorectal cancer at a young age.
(NCT02603341). Gastroenterology. 2005;128(4):1067–76.
The primary disadvantage of proton radiation 10. Hwang SJ, Lozano G, Amos CI, Strong LC. Germline
therapy is the high cost and complexity of con- p53 mutations in a cohort with childhood sarcoma:
sex differences in cancer risk. Am J Hum Genet.
struction and operating a proton facility, which 2003;72(4):975–83.
have historically limited the availability of proton 11. Agarwalla PK, Royce TJ, Koch MJ, Daartz J, Loeffler
therapy. As technology has advanced, more cost- JS. Application of current radiation delivery sys-
effective and user-friendly facilities have been tems and radiobiology. Neurologic surgery. 4th ed.
Philadelphia: Elsevier; 2018.
designed; approximately 30 proton centers are 12. Pawlik TM, Keyomarsi K. Role of cell cycle in medi-
operational in the United States alone in 2018, ating sensitivity to radiotherapy. Int J Radiat Oncol
and this number is actively increasing. The his- Biol Phys. 2004;59(4):928–42.
torically limited availability of proton therapy 13. Corso CD, Bindra RS, Mehta MP. The role of radia-
tion in treating glioblastoma: here to stay. J Neuro-
has led to several important attempts to best Oncol. 2017;134(3):479–85.
ration this resource. Considerations include the 14. Bao S, Wu Q, McLendon RE, Hao Y, Shi Q,

degree of dosimetric and clinical benefit of pro- Hjelmeland AB, et al. Glioma stem cells promote
ton therapy (versus the next best option, typically radioresistance by preferential activation of the DNA
damage response. Nature. 2006;444(7120):756–60.
photon therapy), patient age (prioritization for 15. Hong BJ, Kim J, Jeong H, Bok S, Kim YE, Ahn
pediatric patients), resource equity (no penalty GO. Tumor hypoxia and reoxygenation: the
for lifestyle-associated cancers), advancement of yin and yang for radiotherapy. Radiat Oncol J.
medical research (treat patients on research pro- 2016;34(4):239–49.
16.
Brown JM, Wilson WR. Exploiting tumour
tocols), and the financial reality of operational hypoxia in cancer treatment. Nat Rev Cancer.
cost and thus ability to pay [34, 35]. 2004;4(6):437–47.
17. Peitzsch C, Perrin R, Hill RP, Dubrovska A, Kurth
I. Hypoxia as a biomarker for radioresistant cancer
stem cells. Int J Radiat Biol. 2014;90(8):636–52.
References 18. Kim JJ, Tannock IF. Repopulation of cancer cells dur-
ing therapy: an important cause of treatment failure.
1. Joiner M, van der Kogel A. Basic clinical radiobiol- Nat Rev Cancer. 2005;5(7):516–25.
ogy. 4th ed. London: Hodder Arnold; 2009. p. vi, 375 p. 19. Gunderson LL, Tepper JE, Bogart JA. Clinical

2. Jackson SP, Bartek J. The DNA-damage radiation oncology. 3rd ed. Philadelphia: Elsevier
response in human biology and disease. Nature. Saunders; 2012. p. xxiii, 1638 p.
2009;461(7267):1071–8. 20. Budach W, Hehr T, Budach V, Belka C, Dietz K. A
3. Roos WP, Thomas AD, Kaina B. DNA damage and meta-analysis of hyperfractionated and accelerated
the balance between survival and death in cancer biol- radiotherapy and combined chemotherapy and radio-
ogy. Nat Rev Cancer. 2016;16(1):20–33. therapy regimens in unresected locally advanced
4. Wang H, Mu X, He H, Zhang XD. Cancer radiosensi- squamous cell carcinoma of the head and neck. BMC
tizers. Trends Pharmacol Sci. 2018;39(1):24–48. Cancer. 2006;6:28.
21. Stupp R, Mason WP, van den Bent MJ, Weller M, the potential risk of second malignancies. Cancer.
Fisher B, Taphoorn MJ, et al. Radiotherapy plus con- 2016;122(12):1809–21.
comitant and adjuvant temozolomide for glioblas- 29. Saw CB, Celi JC, Saiful HM. Therapeutic radia-

toma. N Engl J Med. 2005;352(10):987–96. tion physics primer. Hematol Oncol Clin North Am.
22. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de 2006;20(1):25–43.
Tribolet N, Weller M, et al. MGMT gene silencing 30. Jones B, McMahon SJ, Prise KM. The radiobiol-

and benefit from temozolomide in glioblastoma. N ogy of proton therapy: challenges and opportunities
Engl J Med. 2005;352(10):997–1003. around relative biological effectiveness. Clin Oncol
23. Batchelor TT, Sorensen AG, di Tomaso E, Zhang WT, (R Coll Radiol). 2018;30(5):285–92.
Duda DG, Cohen KS, et al. AZD2171, a pan-VEGF 31. Mohan R, Grosshans D. Proton therapy – present and
receptor tyrosine kinase inhibitor, normalizes tumor future. Adv Drug Deliv Rev. 2017;109:26–44.
vasculature and alleviates edema in glioblastoma 32. Chung CS, Yock TI, Nelson K, Xu Y, Keating NL,
patients. Cancer Cell. 2007;11(1):83–95. Tarbell NJ. Incidence of second malignancies among
24. De Ruysscher D. Combination of radiotherapy
patients treated with proton versus photon radiation.
and immune treatment: first clinical data. Cancer Int J Radiat Oncol Biol Phys. 2013;87(1):46–52.
Radiother. 2018;22(6–7):564–6. 33. Sethi RV, Shih HA, Yeap BY, Mouw KW, Petersen
25. Bentzen SM. Preventing or reducing late side effects R, Kim DY, et al. Second nonocular tumors among
of radiation therapy: radiobiology meets molecular survivors of retinoblastoma treated with contem-
pathology. Nat Rev Cancer. 2006;6(9):702–13. porary photon and proton radiotherapy. Cancer.
26.
Allan JM, Travis LB. Mechanisms of ther- 2014;120(1):126–33.
apy-related carcinogenesis. Nat Rev Cancer. 34. Jagsi R, DeLaney TF, Donelan K, Tarbell NJ. Real-
2005;5(12):943–55. time rationing of scarce resources: the Northeast
27. Bowers DC, Nathan PC, Constine L, Woodman C, Proton Therapy Center experience. J Clin Oncol.
Bhatia S, Keller K, et al. Subsequent neoplasms of the 2004;22(11):2246–50.
CNS among survivors of childhood cancer: a system- 35. Bekelman JE, Asch DA, Tochner Z, Friedberg J,

atic review. Lancet Oncol. 2013;14(8):e321–8. Vaughn DJ, Rash E, et al. Principles and reality of
28. Kamran SC, Berrington de Gonzalez A, Ng A, Haas- proton therapy treatment allocation. Int J Radiat
Kogan D, Viswanathan AN. Therapeutic radiation and Oncol Biol Phys. 2014;89(3):499–508.
Role of Whole-Brain Radiotherapy
20
Connor Lynch, Jeffrey P. Gross, and Vinai Gondi
Introduction salvage therapies. Subsequent trials demonstrat-

ing cognitive preservation using neuroprotective
Whole-brain radiotherapy (WBRT) has been strategies of prophylactic memantine and hippo-
integral to the management of brain metastases campal avoidance have led to efforts seeking to
for several decades. Early studies demonstrated redefine the role of WBRT, especially since prior
the efficacy of WBRT in relieving neurologic trials comparing cognitive outcomes between
symptoms related to intracranial disease and focal therapy and WBRT did not include these
improving survival for patients with brain metas- neuroprotective strategies and no longer apply in
tases. However, concerns over cognitive side the modern WBRT era.
effects with conventional WBRT and improve- In recent years, multiple attempts have been
ments in local treatment techniques have led to a made to optimize the efficacy of WBRT. The
shifting dynamic in how and when WBRT is used most common dose prescription for WBRT is
[1]. As a result, focal therapies involving stereo- 30 Gy in 10 fractions, though other dosing regi-
tactic radiosurgery (SRS) with or without surgi- mens have been studied without proven superior-
cal resection have been increasingly used as an ity. The use of systemic agents during and
alternative to conventional WBRT in patients following WBRT has also been studied exten-
with limited brain metastases at a cost of sively. Although enthusiasm for radiosensitizers
increased risk of distant brain relapse and use of was sparked by studies of motexafin gadolinium
showing benefits in non-small cell lung cancer,
other radiosensitizers have failed to show added
C. Lynch value. The use of targeted agents and immune
Northwestern University Feinberg School of checkpoint inhibitors with WBRT remain areas
Medicine, Chicago, IL, USA
of active study.
J. P. Gross Radiation-related toxicity secondary to con-
Department of Radiation Oncology, Northwestern
University, Robert H. Lurie Comprehensive Cancer ventional WBRT manifests as early, early-
Center, Chicago, IL, USA delayed, and late delayed forms, with the last one
V. Gondi (*) being the most permanent. This toxicity ranges
Brain and Spine Tumor Center, Department of from mild cognitive impairment to rarely demen-
Radiation Oncology, Northwestern Medicine Cancer tia and can be a concern for patients and clini-
Center Warrenville, Warrenville, IL, USA cians alike. However, practice-changing clinical
Northwestern Medicine Chicago Proton Center, trials have demonstrated that prophylactic
Warrenville, IL, USA memantine, combined with minimal radiation
e-mail: vinai.gondi@nm.org

https://doi.org/10.1007/978-3-030-42958-4_20
282 C. Lynch et al.
dose to the hippocampal neural stem cell com- receiving either accelerated fractionation (30 Gy
partment (hippocampal avoidance), prevents at 3 Gy daily) or accelerated hyperfractionation
cognitive toxicity in patients undergoing (54.4 Gy at 1.6 Gy BID). The authors found no
WBRT. This chapter traces the course of the difference in survival or toxicity between the two
research that established the use of WBRT and groups [4]. Rades and colleagues retrospectively
discusses the evolving role and delivery of WBRT compared 30 Gy in 10 fractions to either 40 Gy in
in contemporaneous management of brain metas- 20 fractions or 45 Gy in 15 fractions. The alterna-
tases. In order to improve care for patients requir- tive dose-escalated schedules did not signifi-
ing WBRT, knowledge of the optimal candidates cantly improve survival or local control [5].
for WBRT and techniques for safer delivery of Neider and colleagues demonstrated a 25% com-
WBRT are important. plete and 39% partial radiographic response at
The efficacy of WBRT was noted as early as 3 months after WBRT with 30 Gy in 10 fractions.
1954 when Chao et al. published a case series of Radiographic response was associated with
38 patients with symptomatic brain metastases improved survival across multiple cancer histolo-
treated with two opposed lateral x-ray fields gies [6, 7]. Likewise, tumor shrinkage in those
targeting the whole brain. Chao started with with favorable response following WBRT was
doses of 0.5 Gray (Gy) per fraction and eventu- associated with preserved neurocognitive func-
ally increased to 4 Gy per fraction to deliver tion relative to those with poor response in both
total dose up to 35 or 40 Gy. Of these patients, mini mental status exam and specific tests of
63% experienced improvement of a variety of executive function and fine-motor skills [8, 9].
symptoms related to tumor shrinkage in the These seminal studies established WBRT as the
brain. Incontinence, aphasias, and hemiplegia standard of care for management of brain metasta-
improved or resolved in many of these patients. ses and support 30 Gy in 10 fractions as the most
At least one returned to work and another was standard regimen. More recently, concerns over the
able to play the piano again [2]. While limited in neurocognitive sequelae of WBRT have prompted
many respects, this foundational study was the a reevaluation of the technique. Recognizing the
largest series to date demonstrating the pallia- connection between memory formation and the
tive benefit of WBRT and prompted further production of neural progenitor cells (NPCs) in the
study to define the role of WBRT. In 1980, subgranular zone (SGZ) of the hippocampal den-
Borgelt et al. published the results of two phase tate gyrus, a technique was devised that would
III trials—Radiation Therapy Oncology Group avoid this highly radiosensitive region [10]. Termed
(RTOG) 6901 and 7361—demonstrating symp- hippocampal avoidance (HA), results from a sin-
tomatic improvement in 43–64% of patients gle-arm phase II trial and subsequently a random-
with brain metastases at 2 weeks following ized phase III trial combining this strategy with the
WBRT, and noted a threefold increase in median neuroprotective agent memantine (NRG-Oncology
survival time compared to standard supportive CC001) showed significant prevention of cognitive
care (3–6 months vs 1–2 months). These studies toxicity and better preservation of patient-reported
evaluated five different dose schedules ranging quality of life (QoL) [11, 12]. Prior research
from hypofractionated regimens (e.g., 10 Gy in comparing WBRT to focal therapy modalities,
one fraction or 12 Gy in two fractions) to more particularly SRS, does not account for these neu-
conventional schedules of 20–40 Gy in 5 to 20 roprotective strategies, which is crucial to bear in
fractions. They did not identify significant dif- mind when considering the differences in cognitive
ferences in outcomes between the different dose toxicity between WBRT and SRS presented below.
schedules [3]. Though SRS is considered to have a more favor-
More contemporary studies have confirmed able side effect profile, future trials are being
these findings as well. RTOG 9104 assessed designed to reevaluate this in light of these trials’
1-year survival and acute toxicity in patients practice-changing findings.
20 Role of Whole-Brain Radiotherapy 283
Conventional Whole-Brain but should be at least 1 cm, even under optimal

Radiotherapy conditions. A special problem arises anteriorly
because sparing of the ocular lenses and lacrimal
Approach glands may require blocking with <5-mm mar-
gins at the cribriform plate.
Conventional WBRT is administered through The anterior border of the field should be
parallel-opposed lateral portals. The inferior field approximately 3 cm posterior to the ipsilateral
border should be inferior to the cribriform plate, eyelid for the diverging beam to exclude the con-
the middle cranial fossa, and the foramen mag- tralateral lens. However, this results in only
num, all of which should be distinguishable on approximately 40% of the prescribed dose to the
simulation or portal localization radiographs posterior eye. A better alternative is to angle the
(Fig. 20.1). The safety margin depends on pen- beam approximately 3 degrees or more (100- or
umbra width, head fixation, and anatomic factors, 80-cm source-to-axis distance midline, but also
field size dependent) against the frontal plane so
that the anterior beam border traverses posterior
to the lenses (approximately 2 cm posterior to
eyelid markers). Placing a radiopaque marker on
both lateral canthi and aligning the markers per-
mits individualization in terms of the couch angle.
This arrangement provides full dose to the poste-
rior eyes. However, the eyelid-to-lens and -retina
topography is individually more constant than the
canthus, and lateral beam eye shielding is better
individualized with the aid of computed tomogra-
phy (CT) or magnetic resonance imaging (MRI)
Fig. 20.1 Lateral portal of conventional whole-brain scans. When in doubt about tumor coverage or
radiotherapy (WBRT) treatment. Conventional WBRT is
lens sparing for tumors in a subfrontal or middle
administered through parallel-opposed lateral portals. The
inferior field border should be inferior to the cribriform cranial fossa location, CT-based contouring and
plate, the middle cranial fossa, and the foramen magnum planning should be considered (Fig. 20.2a).
a
Conventional whole-brain
radiotherapy
30 Gy 30 Gy
b
Hippocampal avoidance
during whole-brain
radiotherapy
30 Gy
30 Gy
8 Gy 8 Gy
Fig. 20.2 Comparison of treatment plans between WBRT achieves several-fold reduction in radiation
(a) conventional whole-brain radiotherapy (WBRT) dose to hippocampi (yellow). (Adapted with permis-
and (b) hippocampal avoidant WBRT. Hippocampal sion from Brown et al. [12])
avoidance using intensity-modulated radiotherapy during
284 C. Lynch et al.
cute, Early-Delayed, and Late-

A assessed the impact of adjuvant WBRT after
Delayed Complications SRS, deterioration of immediate memory,
delayed memory, processing speed, and execu-
Toxicity following conventional WBRT may be tive function were associated with conventional
categorized as acute, early-delayed, or late- WBRT [15]. Radiation necrosis is another late
delayed depending on the time of presentation. complication of WBRT. Necrosis may result in
Acute effects of radiation manifest during the mass-effect-related symptoms that make these
course of treatment or shortly after completion. lesions difficult to distinguish from tumor recur-
Common complications include those associated rence. These lesions can require surgical inter-
with increased intracranial pressure such as vention if unresponsive to corticosteroids.
headache, fatigue, nausea, and dizziness. These Radiation-related leukoencephalopathy is seen
side effects may be due to interruption of the in rare cases and results in severe dementia and
blood-brain barrier and the development of cere- cortical atrophy. Higher per-fraction doses (in
bral edema immediately following radiation excess of 3.5 Gy) have been associated with
exposure. These symptoms generally respond greater risk of radiation-related leukoencepha-
well to corticosteroids [13]. Patients may also lopathy [16]. The capacity of neuroprotective
acutely experience mild, self-limited dermatitis, strategies including prophylactic memantine and
and hair loss. Early-delayed toxicity appears hippocampal avoidance during modern WBRT
weeks to months following treatment and is to prevent radiation-related leukoencephalopa-
thought to arise due to transient demyelination. thy remains unclear.
It manifests as weakness, headache, and fatigue
[13]. Additional non-neurological side effects
include serous otitis media, dry sinuses, and lac- The Evolving Role of Conventional
rimal gland dysfunction. Lhermitte’s sign may WBRT
be present in some of these patients, identified as
the sensation of a shock spreading down the Omission of WBRT
neck and upper limbs with flexion of the neck. In poor performance status patients with limited
Radiation somnolence syndrome is a rare early- survival, there is better understanding regarding
delayed complication of central nervous system the benefit of WBRT versus modern best support-
(CNS) radiation characterized by extreme som- ive care. The Quality of Life after Treatment for
nolence accompanied by anorexia, apathy, and Brain Metastases (QUARTZ) trial was designed
headache. The syndrome is commonly associ- in part to address this question, randomizing 538
ated with prophylactic cranial irradiation in patients with non-small cell lung cancer (NSCLC)
pediatric patients with acute lymphocytic leuke- to either WBRT with optimal supportive care or
mia, but has been described in adult patients supportive care alone. Eligible patients had brain
undergoing radiation therapy for primary CNS metastases that were not amenable to stereotactic
tumors. Management and prevention involve radiosurgery or resection. Using quality-adjusted
administering corticosteroids during radiation life-years as the primary outcome measure, the
treatment [14]. trial found that omitting WBRT resulted in a loss
Late-delayed toxicities appear beginning at of 4.7 days (in terms of QALYs). Overall survival
6 months after radiation but can present many time was also diminished by less than a week for
years later. They are often the most debilitating those receiving supportive care alone when com-
and the least likely to improve with time. pared to those receiving WBRT [17]. While this
Permanent neurocognitive dysfunction follow- study is commonly used to dismiss the use of
ing conventional WBRT ranges from mild WBRT in the palliative setting, it is important to
impairment in most cases to severe dementia in avoid overgeneralizing the results. First, clini-
rare cases (<5%) [13]. For instance, in the previ- cians were encouraged to recruit patients into the
ously mentioned NCCTG N107C/CEC.3, which trial if they had doubts regarding the benefit of
WBRT. The median survival on this study was Cancer (EORTC 22852-26,001) [21] demon-
8–9 weeks, highlighting an extremely unfavor- strated that adding WBRT does indeed improve
able cohort of patients in both groups. intracranial disease control, resulting in a signifi-
Symptomatic benefit from tumor regression may cant reduction in the absolute risk of new brain
take 3–6 months; therefore it is not surprising metastases by between 18% and 22% at 1 year
that there was no difference in quality of life for and by 15% at 2 years. Recurrence rates at local
patients undergoing WBRT on this study. sites were also reduced. Notably, the EORTC and
Furthermore, subgroup analysis did demonstrate MD Anderson studies found reduced quality of
a significant survival benefit to WBRT for patients life and reduced Hopkins Verbal Learning Test-
younger than 60, with a non-significant trend in Revised (HVLT-R) scores, respectively, with the
favor of WBRT observed for patients of better addition of WBRT. In addition, contrary to expec-
prognosis as measured by recursive partitioning tations, the MD Anderson trial identified a sur-
analysis (RPA) and disease-specific generalized vival difference in patients managed with SRS
prognostic assessment (ds-GPA) scores. Finally, alone, who experienced a median survival time of
relative to this study, brain metastases may be 15.2 months compared to 5.7 months in patients
associated with a better median survival with the receiving combination therapy. The differences
emergence of immunotherapy or other systemic in survival could have in part contributed to the
therapies or with other types of cancer. Thus, differences in neurocognition and quality of life
while the QUARTZ trial demonstrated that observed between the arms [20]. To address these
NSCLC patients with poor prognosis might not conflicting data on overall survival, Sahgal et al.
benefit from WBRT, those with a better prognosis (2013) analyzed individual patient data from
or a better performance status may experience these trials and identified patient age as a signifi-
survival and/or quality-of-life improvements cant effect modifier. After stratifying by age, they
with WBRT. To aid in decision making, Sperduto found that SRS alone was associated with favor-
and colleagues have developed prognostic sys- able survival outcomes in patients younger than
tems to provide survival time estimates for 50 years old, although the majority of this differ-
patients with brain metastases [18]. However, for ence was driven by the MD Anderson trial. For
patients who develop brain metastases in the set- patients older than 50, there was no difference
ting of systemic progression and are not planned between SRS alone and SRS with WBRT. This
for further systemic therapy due to poor perfor- meta-analysis also identified higher rates of sal-
mance status and/or limited prognosis, the vage treatment in the SRS alone arm, highlight-
QUARTZ trial provides a rationale for omission ing the need for regular imaging follow-up with
of WBRT to manage the brain metastases. SRS alone [22].
Alliance trial N0574 assessed the impact of
Conventional WBRT Following adjuvant WBRT after SRS on quality of life,
Stereotactic Radiosurgery functional independence, and radiation-related
Stereotactic radiosurgery (SRS) offers the ability cognitive dysfunction at 3 months using a battery
to deliver targeted, high doses of radiation to dis- of standardized cognitive tests to assess learning,
crete foci of metastatic disease within the brain. memory, fine motor control, verbal fluency, pro-
The hypothesis that SRS followed by adjuvant cessing speed, and executive function. Cognitive
WBRT for patients with limited brain metastases deterioration—defined as decline greater than 1
could achieve superior intracranial control and standard deviation (SD) below baseline in any of
survival has been tested in multiple phase-III ran- these cognitive domains at 3 months—was more
domized controlled trials (RCTs). Prospective frequent with SRS and adjuvant WBRT com-
studies conducted by the Japanese Radiation pared with SRS alone (91.7% vs. 63.5%,
Oncology Study Group (JROSG 99–1) [19], MD P < 0.001). Specifically, patients receiving com-
Anderson Cancer Center [20], and the European bined therapy were more likely to experience
Organization for Research and Treatment of impairments in immediate memory, delayed
286 C. Lynch et al.
memory, and verbal fluency than those receiving improved local and distant control. Specifically,
SRS alone. Quality of life was significantly better adjuvant SRS led to a 20% decrement in surgi-
with SRS alone and functional independence was cal bed control at 12 months compared to
the same between arms. Overall survival was not WBRT (60% compared to 80%, P = 0.00068).
different between groups despite the improved While this improved intracranial control was
intracranial control of combined therapy [23]. not associated with an increase in overall sur-
vival, this trial lacked a comparison of the rates
Conventional WBRT Following Surgical of neurologic cause of death [15]. With respect
Resection to cognitive deterioration, adjuvant WBRT per-
Upfront surgery for large or symptomatic brain formed significantly worse than adjuvant SRS,
metastases is associated with survival benefits. with an overall rate of cognitive deterioration of
However, multiple studies have demonstrated 85% versus 52%, respectively, at 6 months
that the rate of local recurrence following MRI- (P = 0.0003). Within specific cognitive domains,
confirmed gross total resection of brain metasta- patients in the WBRT arm had significantly
ses without adjuvant therapy is around 50% [21, higher rates of deterioration in immediate
24, 25]. Two large RCTs, a multi-center study recall, delayed recall, processing speed, and
published by Patchell et al. in 1998 and EORTC executive function.
22952-26001, have investigated the use of sur- Taken together, the evidence supports the use
gery with adjuvant WBRT versus surgery alone. of postoperative radiotherapy following surgical
Both studies demonstrated a statistically signifi- resection for brain metastasis. Both WBRT and
cant improvement in local control, reduction in SRS remain effective treatment options but have
the incidence of distant brain metastases, and some limitations [26]. Neuroprotective strategies
reduced incidence of neurologic death with the to prevent cognitive toxicity from WBRT are dis-
addition of adjuvant WBRT [21, 24]. However, cussed below. The inferior surgical bed control of
these studies did not find a significant difference SRS remains an area of concern. An Alliance
in survival for adjuvant WBRT over observation phase III trial of fractionated versus single-
following surgery, though they were not powered fraction radiosurgery to improve local control
to do so. following surgical resection will seek to address
Stereotactic radiosurgery has been shown to this issue.
improve local control following surgical resec-
tion while minimizing the potential for neurocog- Prophylactic Cranial Irradiation
nitive toxicity. A phase III trial of postoperative WBRT may be used prophylactically (i.e., before
SRS compared to observation (MD Anderson disease is radiologically detectable) in select
Cancer Center 2009-0381) demonstrated patients with small cell lung cancer (SCLC), who
improved surgical bed control with SRS com- demonstrate up to 80% risk of developing brain
pared to observation (12- month surgical bed metastases 2 years after diagnosis [27]. As such,
relapse rate: 28% with SRS vs. 57% with obser- WBRT is considered the standard of care for
vation, p = 0.015). While there was no survival patients with limited-stage (LS) SCLC that has
advantage to adjuvant SRS, there was a trend responded to chemotherapy, given the potential
toward reduced neurologic death with SRS but for prolonged survival. The seminal meta-
this did not reach statistical significance analysis by Aupérin et al. (1999) demonstrated a
(p = 0.13). significant increase in overall survival (pooled
A phase III trial from a collaboration between relative risk of death 0.84, P = 0.01) and
Alliance and the Canadian Cancer Trials Group significantly reduced the incidence of brain
(N107C/CEC.3) compared surgery with adju- metastases (0.46, P < 0.001) in patients with a
vant WBRT to surgery with adjuvant SRS and complete response (CR) to chemotherapy [28].
examined both overall survival and cognitive These results were reinforced by a 2001 system-
side effects. WBRT was again associated with atic review by Meert et al., which also showed
decreased incidence of brain metastasis (HR of However, these trials of PCI in NSCLC were
0.48, 95% CI of 0.39–0.60) and improved overall conducted prior to the emergence of immune
survival (HR of death 0.82, 95% CI 0.71–0.96) in checkpoint inhibitors, now considered the stan-
patients with LS SCLC and CR [29]. dard of care for most locally advanced and meta-
Prophylactic WBRT in patients with extensive static NSCLC patients. Thus, in the modern era
stage (ES) SCLC is more controversial. A 2007 of NSCLC management, the role of PCI remains
EORTC trial seemed to demonstrate improved uncertain. The use of neuroprotective strategies
survival for patients with ES SCLC and any posi- such as hippocampal avoidance during PCI to
tive response to chemotherapy [30]. This study, prevent cognitive toxicity also remains an area of
however, did not include brain imaging as a part ongoing investigation through the current NRG
of its inclusion criteria, raising the possibility that Oncology CC003 trial.
some patients had asymptomatic brain metasta-
ses upon enrollment (making cranial irradiation
for these patients therapeutic rather than prophy- Modern WBRT
lactic). A later phase III Japanese trial of 224 ES
SCLC patients addressed this concern by exclud- Preceding and concurrent with trials establish-
ing patients with brain lesions visible on MRI ing the neurocognitive toxicity of conventional
prior to enrollment. This study showed no benefit WBRT, several investigations have been pur-
to overall survival (survival HR 1.27; P = 0.094) sued to identify approaches to deliver WBRT
with prophylactic cranial irradiation (PCI) versus more safely. These approaches have included
observation and was halted for futility [31]. In both pharmacologic and technologic strategies
light of this most recent trial, prophylactic WBRT and have led to practice-changing findings that
for ES SCLC is controversial, and a planned have ushered in the era of modern WBRT inclu-
SWOG phase III trial MAVERICK seeks to sive of prophylactic memantine and hippocam-
address this question. SCLC patients in this study pal avoidance.
will be randomized to PCI with hippocampal
avoidance versus MR surveillance; the primary
endpoint is overall survival. MDA Receptor Antagonists
N
Given its success in limited stage SCLC, (Memantine)
WBRT has also been studied extensively in non-
small cell lung cancer (NSCLC). While no study N-methyl-D-aspartate (NMDA) receptors are
has demonstrated an advantage to overall sur- ionotropic glutamate receptors that mediate syn-
vival, two phase III trials have demonstrated sig- aptic plasticity and memory in the brain, particu-
nificantly reduced incidence of brain metastases larly in the neurons of the hippocampus.
[32, 33]. An additional phase III study by De Overstimulation of these receptors following
Ruysscher et al. in 2018 confirmed a reduced inci- insults to the brain by ischemia, trauma, or radia-
dence of brain metastases with PCI versus obser- tion can lead to apoptosis and necrosis via a phe-
vation (7% vs 27.2%, P = 0.001), albeit with a nomenon known as excitotoxicity. Preclinical
reduced quality of life with PCI at 3 months post- studies have demonstrated that blockade of these
treatment and a non-significant trend toward QoL receptors by the noncompetitive NMDA antago-
benefit to observation at 2, 3, and 4 years [34]. nist memantine protects against NMDA-receptor-
RTOG 0214 was a phase III trial that randomized mediated neurotoxicity [35, 36]. Animal studies
stage III NSCLC to PCI or observation but did not have also demonstrated that giving memantine
complete target accrual to detect an overall sur- ahead of radiation can preserve long-term poten-
vival benefit. However, unplanned analyses of tiation—a process involved in synaptic plastic-
longer-term results revealed an overall survival ity—in rodents [37, 38]. Phase II clinical studies
benefit of PCI in stage III NSCLC patients who have demonstrated the effectiveness of meman-
did not undergo upfront surgical resection. tine in managing vascular dementia [39, 40]. The
288 C. Lynch et al.
apparent neuroprotective effects of memantine the HVLT-R Delayed Recognition) and processing
generated interest in its use in for managing speed (Trail-Making Test A) at 24 weeks [41].
radiation-related neurotoxicity. However, when cognitive toxicity was assessed as
A phase III trial (RTOG 0641) was designed to a composite endpoint, defined as a decline in the
assess the neuroprotective effects memantine in reliable change index on the HVLT-R, Trail-
patients treated with WBRT. Patients were ran- Making Test, or Controlled Oral Word Association
domized to WBRT (37.5 Gy in 15 fractions) with tests, the use of memantine during WBRT led to a
either memantine or placebo. The dose of meman- 22% relative reduction in risk of cognitive toxicity.
tine was escalated over the course of treatment These results, combined with the favorable safety
beginning with 5 mg QD in week 1 of treatment profile of memantine, have made the drug appro-
and rising to 10 mg BID for weeks 4 through 24. priate for use in clinical practice to mitigate the
The full regimen is detailed in Table 20.1. Because cognitive toxicity of WBRT, particularly in con-
memantine is primarily cleared renally, exceptions junction with hippocampal avoidance as detailed
were made for patients for patients with low cre- below. It is not known at this time what the optimal
atinine clearance. Those with clearance below dosing schedule and duration of memantine is to
30 mL/min received 5 mg BID and those with attenuate radiation-induced neurotoxicity, and fur-
clearance less than 5 mL/min were taken off the ther trials may help guide future management
drug. The primary endpoint was whether meman- recommendations.
tine preserved memory, as assessed by the HVLT-R
Delayed Recall at 24 weeks. Although patients
treated with memantine were found to experience Hippocampal Avoidance
less cognitive decline than control patients, this
difference was not statistically significant (0 com- The hippocampus plays a critical role in the for-
pared to −0.9, P = 0.059), possibly due to the high mation of episodic and spatial memory. Its ability
rate of attrition in the trial. Among the positive to do so stems in part from the production of new
findings in the trial were a significantly longer neurons by neural progenitor cells (NPCs) within
time to cognitive deterioration in the memantine the subgranular zone (SGZ) of the hippocampal
arm (HR 0.78, P = 0.01) and significantly less dentate gyrus. Animal studies have demonstrated
deterioration in delayed recognition (measured by that these NPCs are highly sensitive to radiation
and that radioablation of these cells results in
Table 20.1 Memantine dosing in RTOG 0614 deficits in hippocampus-dependent learning and
Extended memory tasks [42]. Given this interaction with
Twice daily dosinga release dosingb radiation, it is unsurprising that memory deficits
Morning Evening are commonly reported in WBRT patients. One
Week(s) dose (mg) dose (mg) Daily dose (mg) recent study (NCCTG N107C/CEC.3) examining
1 5 – 7
WBRT versus SRS following surgical resection
2 5 5 14
found deterioration of immediate and delayed
3 10 5 21
4–24 10 10 21 memory in 49% and 62% of patients, respec-
a
A dosage reduction to 5 mg orally twice daily is recom-
tively. This was significantly more than in patients
mended in patients with severe renal impairment [creati- treated with focal radiotherapy via SRS [15].
nine clearance (CrCl), 5–29 milliliters/minute (mL/min)]. Clinical studies have also demonstrated a clear
No dosage adjustment is needed in patients with mild dose-response relationship between hippocampal
(CrCl greater than 50–80 mL/min) or moderate (CrCl
30–49 mL/min) renal impairment
radiation exposure and memory deterioration,
b
A dosage reduction to 14 milligrams (mg) orally daily is with a study by Gondi et al. (2013) demonstrat-
recommended in patients with severe renal impairment ing an association between the delivery of 7.3 Gy
(creatinine clearance (CrCl), 5–29 milliliters/minute (mL/ to 40% of the bilateral hippocampi (in the equiva-
min)). No dosage adjustment is needed in patients with
mild (CrCl greater than 50–80 mL/min) or moderate
lent of 2 Gy fractions) and long-term deteriora-
(CrCl 30–49 mL/min) renal impairment tion in list-learning delayed verbal recall as
measured by the Weschler Memory Scale-III (p = 0.67), indicating that the cognitive benefit of
Word Lists test [43]. Given this association and hippocampal avoidance does not differ by age.
given the relatively rare rate of metastasis to the Importantly, the addition of hippocampal
hippocampi, a technique was devised using avoidance to WBRT+memantine preserved
intensity-modulated radiotherapy (IMRT) to patient-reported symptom burden, as assessed by
limit the dose delivered to the hippocampus the M.D. Anderson Symptom Inventory Brain
(Fig. 20.2b) [10]. Tumor Module (MDASI-BT). Patients on the
A phase II study, RTOG 0933, was designed HA-WBRT+memantine arm experienced less
to evaluate the benefits of this hippocampal symptom interference and less cognitive symp-
avoidance strategy. The study found that com- toms at 6 months (estimate = −1.02, p = 0.008
pared with historical controls, patients treated and estimate = −0.63, p = 0.011, respectively)
with hippocampal avoidance (HA) WBRT expe- compared to the WBRT+memantine arm.
rienced significantly less deterioration in delayed Cognitive symptom differences were driven pri-
memory as measured by the HLVT-R Delayed marily by two items: problems with remember-
Recall. Consistent with previous observations, ing things and difficulty speaking. At 6 months,
4.5% of patients experienced progression in the patients on the HA-WBRT+memantine arm had
hippocampal avoidance region [44]. less difficulty remembering things (mean 0.16 vs.
A phase III trial, NRG Oncology-CC001, was 1.29, p = 0.013) and less difficulty speaking
conducted to validate these findings in patients (mean − 0.20 vs. 0.45, p = 0.049) as compared to
treated with memantine and WBRT with or with- the WBRT+memantine arm. Greater improve-
out HA. The study recruited and randomized 518 ment in fatigue at 6 months was reported in the
adult patients with brain metastases between July HA-WBRT+memantine arm as compared to the
2016 and March 2018. The primary endpoint was WBRT+memantine arm (mean 0.93 vs. −0.16,
cognitive toxicity, defined as a decline in the reli- p = 0.036).
able change index on the HVLT-R, Trail-Making Analyses with longer follow-up (median fol-
Test, or Controlled Oral Word Association tests. low-up of 12.1 months) additionally demon-
There was no difference in grade 3 or higher tox- strated better preservation of overall symptom
icity between the treatment arms. The median burden (p < 0.0001) at 6 months on the
follow-up for alive patients was 7.8 months. HA-WBRT+memantine arm compared to the
There was no difference between arms in terms WBRT+memantine arm, while continuing to
of baseline cognitive function, overall survival show similar benefits in cognitive function and
(HR = 1.13, 95% CI: 0.89–1.44, p = 0.31), or patient-reported quality of life with hippocampal
intracranial progression (HR 1.12, 95% CI 0.90– avoidance.
1.39, p = 0.33). The summation of these findings remains con-
The addition of hippocampal avoidance to sistent with cognition-specific hypothesis of hip-
WBRT+memantine significantly prevented cog- pocampal avoidance but also underscore the
nitive toxicity (Fig. 20.2b) with an adjusted haz- palliative intent of brain metastasis management
ard ratio of 0.74, or a 26% relative reduction in and the capacity of HA-WBRT to provide opti-
risk of cognitive toxicity with the addition of hip- mal intracranial control to limit neurologic symp-
pocampal avoidance to memantine [12, 26]. The tom burden.
difference was first seen at 4 months and main-
tained throughout the follow-up period, and was
attributable to improvements in executive func- Future Directions
tion at 4 months (p = 0.01) and learning
(p = 0.049) and memory (p = 0.02) at 6 months. All of the trials observing higher cognitive toxic-
While age also predicted for prevention of cogni- ity in patients receiving WBRT were conducted
tive function failure, test for interaction between in the conventional era of WBRT without the
treatment arm and age was non-significant inclusion of neuroprotective strategies including
290 C. Lynch et al.
memantine and hippocampal avoidance, which metastases. However, as mentioned above, these
have demonstrated significant cognitive toxicity studies were largely conducted prior to the publi-
prevention. In the modern era of brain metastasis cation of large brain metastasis trials testing
management, the role of WBRT with neuropro- pharmacologic and technologic neuroprotec-
tective strategies remains under investigation. tive strategies during WBRT and leading to the
Given the increased requirement for imaging safer delivery of WBRT. Thus, the appropriate
follow-up and the higher rate of salvage therapies management of patients with multiple brain
associated with SRS alone, modern WBRT may metastases remains unclear.
be appropriate for patients who do not wish to To address this question in the newly diag-
undergo extensive surveillance or subsequent sal- nosed setting of multiple brain metastases, mul-
vage therapy. Generally, however, SRS with tiple trials have been launched. Originally, a trial
omission of WBRT can be considered standard of comparing SRS to conventional WBRT for
care for patients whose survival is anticipated to patients with greater than five brain metastases
extend multiple years, as the capacity of meman- was initiated by the North American Gamma
tine and hippocampal avoidance to prevent the Knife Consortium. Although this trial was of
rare occurrence of radiation-related leukoenceph- interest, it was limited in its scope to only one of
alopathy in long-term survivors of WBRT the several radiosurgical platforms and limited in
remains unclear. its statistical power (39 patients planned to be
It is worth noting too that SRS is being investi- accrued per treatment arm) and the trial closed
gated for use in five or more brain metastases, long before reaching the total target accrual.
with one prospective observational study demon- More recently, the Canadian Clinical Trials
strating that survival in patients receiving SRS Group (CCTG) launched a cooperative-group
alone for 5–10 brain metastases was not inferior phase III trial of SRS versus conventional WBRT
to that seen in patients receiving SRS alone for in 5–15 brain metastases with co-primary end-
two to four brain metastases [45]. Currently, four points of overall survival and neurocognitive
RCTs are either planned or actively accruing progression-free survival. Given the practice-
patients to directly compare SRS versus WBRT changing evidence from NRG CC001, this trial
for four or more brain metastases (up to as many has subsequently been amended to compare SRS
as 20 in one study) [46]. Absent conclusive evi- versus modern WBRT with hippocampal avoid-
dence for non-inferiority of SRS alone to WBRT ance and memantine and has also been endorsed
for patients with more than four brain metastases, by NRG Oncology and Alliance. The question of
modern WBRT with hippocampal avoidance and whether SRS or modern WBRT with hippocam-
memantine remains a standard of care for these pal avoidance and memantine is the optimal
patients. modality in patients with 5–15 brain metastases
is significant from a societal and medical
resources standpoint since the charges related to
odern WBRT for Newly Diagnosed
M SRS and IMRT for HA-WBRT can be consider-
Brain Metastases ably higher than those of conventional
WBRT. However, examining therapy-associated
Radiosurgery, for as many as 15 brain metasta- costs is particularly complex in patients with mul-
ses, has been found to be safe, notably in a series tiple brain metastases, because such patients are
of 360 patients from Japan [45]. The feasibility likely to undergo additional salvage p rocedures
and safety of multiple-brain metastasis SRS, as for new brain metastases. Therefore, the addi-
well as studies demonstrating inferior cognitive tional costs of salvage are also important to incor-
outcomes following upfront WBRT relative to porate into economic comparisons, especially
upfront SRS for one to four brain metastases, when SRS is anticipated to result in higher intra-
have led several institutions to consider SRS cranial relapse rate and need for salvage therapies
alone for patients with more than four brain [20, 21, 23, 47].
Brain Metastasis Velocity brain metastasis velocity exceeding four brain

metastases/year is being developed through NRG
Brain metastasis velocity (BMV) is a useful mea- Oncology with anticipated activation in 2020.
sure for predicting outcomes in patients with The primary objective of this trial is to determine
brain metastases who experience distant brain if the addition of HA-WBRT with memantine to
relapse following their first SRS treatment. It is salvage radiosurgery effectively prevents neuro-
defined as the cumulative number of brain metas- logic death in this high-risk patient population.
tases developed since upfront SRS divided by the
number of years following SRS. For example, a
patient who develops two brain metastases mall Cell Lung Cancer Brain
S
6 months after upfront SRS would have a BMV Metastases
of 2/0.5 = 4. Developed by Farris et al. (2017),
BMV was found to be significantly associated Intracranial failure is a frequent problem in
with overall survival, neurologic death, and rates patients with small cell lung cancer (SCLC).
of salvage WBRT in a cohort of 737 patients [48]. SCLC accounts for approximately 15% of all
This remained true when the same analysis was cases of lung cancer, tends to disseminate earlier
applied to a validation set featuring an additional in the course of its natural history than non-small
2092 patients across multiple institutions [49]. cell lung cancer and is more clinically aggressive.
Farris et al. (2017) stratified patients into low As a result, approximately 10–20% of SCLC
(<4), intermediate (4–13), and high (>13 BMV) patients present with brain metastases at the time
categories, finding that patients with high BMV of initial diagnosis, and an additional 40–50%
experienced a cumulative incidence of neuro- will develop brain metastases some time during
logic death roughly twice that of low-BMV the course of their disease. In addition, brain
patients. Neurologic death was defined by the metastases have an impact on the quality and
authors as death with progressive neurologic length of survival. Prophylactic cranial irradia-
decline, regardless of extracranial disease status tion (PCI) has historically been used as a strategy
[48]. The significant association of BMV with to reduce the incidence of brain metastases in
neurologic death, thus defined, makes it a useful SCLC; however, the National Comprehensive
marker for predicting intracranial control, as does Cancer Network (NCCN) guidelines recommend
the association between BMV at first distant caution regarding PCI delivery in older patients
brain relapse and BMV at second distant brain and PCI is omitted in up to 40–50% of patients,
relapse. The prognostic value of BMV has since primarily due to concerns over cognitive toxicity
been validated in two additional published series [52, 53]. NRG CC003 is an ongoing phase III
[50, 51]. trial testing whether the cognitive toxicity of PCI
This predictive ability is of interest for its can be prevented with hippocampal avoidance
potential utility in triaging patients at risk for during PCI for SCLC patients.
poor intracranial control to optimal intracranial Due to the high propensity for micro-
control offered by SRS plus WBRT. With contin- metastatic seeding of the brain, WBRT remains
ued refinement, BMV could be used to identify standard of care for patients with SCLC brain
and treat patients who would benefit from the metastases. Studies demonstrating cognitive tox-
superior intracranial control offered by icity from conventional WBRT have led to
WBRT. This in turn could reduce both neurologic questions as to whether upfront SRS followed by
death and, more generally, the neurological close imaging surveillance for patients with
sequelae of a high burden of brain metastatic dis- SCLC brain metastases is an acceptable alterna-
ease in this patient population. A phase III trial tive. Importantly, SCLC patients have been
(NRG BN009) of salvage SRS with or without excluded from the landmark randomized trials
modern WBRT with hippocampal avoidance and testing SRS for brain metastases [20, 21, 23, 47].
memantine for recurrent brain metastases with Historic objections to the use of SRS in SCLC
292 C. Lynch et al.
have included the concern for diffuse interval gle-institution retrospective analysis that
CNS progression, which could potentially result identified more significant leukoencephalopathy
in diminished overall survival. in patients treated with PCI using HAWBRT than
However, there is growing evidence to sug- conventional WBRT [58] and a multi-institution
gest that SRS alone may be safe and appropriate phase II trial of early HA-PCI that saw similar
for some patients with SCLC brain metastases. neurocognitive outcomes as PCI using conven-
A multi-institutional retrospective analysis of tional WBRT techniques [59].
293 patients treated with SRS for SCLC brain
metastases observed the risk of radiation necro-
sis to be <5% [54], comparable to outcomes fol- Alternating Electric Field Therapy
lowing SRS for brain metastases from other
histologies. Serizawa et al. (2002) [55] com- Alternating electric fields—commonly called
pared the outcomes of SCLC (N = 34) and tumor treating fields or TTFields—have been
NSCLC (N = 211) patients with brain metasta- increasingly used as part of management for glio-
ses treated with SRS alone and found compara- blastoma. The low-intensity, intermediate fre-
ble rates of overall survival, central nerve system quency fields are applied via an adhesive cap
control, and neurologic mortality in SCLC and consisting of an array of transducers and serve to
NSCLC patients. Yomo and Hayashi (2015) interrupt cell replication by two principal mecha-
[56] reported on 70 SCLC patients treated with nisms. First, TTFields interact with the strong elec-
SRS (including 46 without prior PCI or WBRT), tric dipole moments of the microtubules forming
with a median overall survival of 7.8 months the mitotic spindle, disrupting spindle formation
and encouraging 1-year and 2-year neurologic and stalling mitosis. Second, the fields have been
mortality free survival of 94% and 84%, respec- shown to destroy cells nearing the end of cytokine-
tively. A recent analysis of the National Cancer sis, rupturing the cell membrane and generating
Database compared upfront WBRT with upfront membrane blebs that resemble the products of
SRS for SCLC patients with brain metastases apoptosis. These results were observed in vitro in
and reported favorable overall survival with both glioma and melanoma cell lines [60].
SRS both overall and after propensity-score A phase III trial comparing TTFields with
matching [57]. temozolomide to temozolomide alone in patients
Although retrospective analyses are subject to with glioblastoma (GBM) has demonstrated a
cofounding from selection bias, they do suggest survival advantage to adding TTFields. Patients
that some patients may be safely and effectively treated with TTFields were exposed to low-inten-
managed with a strategy of SRS alone. Overall, sity, 200-kHz alternating electric fields for at
there is growing equipoise regarding the role of least 18 h per day via a portable device. These
SRS versus WBRT in the management of SCLC patients had an overall median survival of
brain metastases, and prospective randomized 20.9 months compared to 16 months in those
data are urgently needed to address this knowl- treated with temozolomide alone (P < 0.001).
edge gap especially given practice-changing evi- Systemic adverse events occurred at about the
dence demonstrating the cognitive preservation same rate in each arm with the most common
benefits of hippocampal avoidance and meman- side effect of treatment being mild-moderate skin
tine as neuroprotective strategies during irritation of the scalp in 52% of patients in the
WBRT. NRG Oncology is currently developing a TTFields arm [61]. Prompted by this success in
phase III trial of SRS versus modern WBRT with the management of primary CNS malignancy
hippocampal avoidance and memantine for 10 or and by preclinical data showing effectiveness in
fewer brain metastases from small cell lung cannon-CNS malignancies, a phase III trial is cur-
cer with a primary endpoint of cognitive toxicity. rently underway to evaluate the use of TTFields
There is data from Switzerland that is raising in conjunction with radiosurgery for patients
questions about HA-WBRT for PCI that this pro- with 1–10 NSCLC metastases. The METIS trial
posed trial may help examine, specifically a sin- (ClinicalTrials.gov identifier: NCT02831959)
will evaluate as its primary outcome the time to TMZ. A phase III study from Antonadou et al.
intracranial progression and, as a secondary out- (2002) demonstrated a significantly higher radio-
come, track cognitive function in patients receiv- graphic response rate with combined therapy ver-
ing this novel therapy. sus WBRT alone (53.4 vs. 33.3%, P = 0.039). The
difference in response rate was even more dra-
matic in patients <60 years of age and with a
Concomitant Systemic Agents Karnofsky performance score of 90–100 (70.6 vs.
32.4%, P = 0.003 in the latter group). The study
Since the 1980s, a variety of systemic therapies found no difference in neurological response or
have been investigated for use in conjunction with median survival, however [66].
WBRT for patients with brain metastases. The A later phase III trial in NSCLC patients from
imidazoles such as metronidazole and misonida- Sperduto et al. (2013) investigated WBRT and
zole were among the first agents to be tested in this SRS with or without TMZ or erlotinib. The
context. Neither agent added any survival benefit authors again found no significant survival
over WBRT alone [62]. More recently, a trial of advantage with the addition of temozolomide.
sodium glycididazole did demonstrate improved They also found no difference in time to progres-
intracranial control and longer progression- free sion [67]. A 2016 meta-analysis of seven trials
survival, but did not find a benefit to overall sur- (including those discussed above) comparing
vival [63]. One area of success has been with the radiotherapy with TMZ to radiotherapy alone
use of motexafin gadolinium (MGd), a redox mod- found no advantage in survival to adding TMZ,
ulating agent that catalyzes the oxidation of vari- despite a significant increase in response rate on
ous intracellular metabolites, increasing the combination therapy. Patients treated with TMZ
toxicity of reactive oxygen species and limiting were more likely to experience grade 3 to 4 nau-
the cell’s ability to repair itself. While one phase sea and grade 3 to 4 thrombocytopenia [68]. As
III trial of MGd in patients with brain metastases TMZ has not shown a survival benefit and is
demonstrated no overall benefit in survival time or accompanied by an increase in toxicity, it is not
time to neurologic progression overall, a subset of recommended for use in clinical practice for
patients with NSCLC did experience a benefit in patients with brain metastases.
time to neurologic progression [64]. The phase III Inhibitors of the epidermal growth factor
trial that followed compared WBRT with or with- receptor (EGFR) tyrosine kinase are under inves-
out MGd in NSCLC patients and demonstrated tigation for adjuvant use with WBRT. Erlotinib is
that patients initiating WBRT within 28 days of one such (EGFR) inhibitor with known radiosen-
brain metastasis diagnosis experienced a signifi- sitizing properties. Recent trials of this agent
cant improvement in time to neurologic progres- highlight the importance of patient selection with
sion with the addition of MGd. This effect was respect to pathway-specific mutations. While a
identified on geographic subgroup analysis when preliminary trial from Welsh et al. (2013) sug-
it was found that patients in North America (where gested a benefit for adjuvant erlotinib with WBRT
investigators were more likely to initiate WBRT in lung cancer patients, subsequent trials have
earlier) had a significantly longer time to neuro- contradicted this [69]. The above-mentioned
logic progression than the overall cohort [65]. study from Sperduto et al. (2013) showed that
Based on these data, MGd was deemed an appro- adding erlotinib provided no benefit to overall
priate adjunct therapy in NSCLC patients, pro- survival or time to progression [67]. Another
vided that WBRT is initiated promptly, but has not study from Lee et al. (2014) again found that add-
been widely accepted. ing erlotinib had no effect on neurological
Temozolomide (TMZ) is a DNA alkylating progression-free survival or overall survival [70].
chemotherapeutic agent that is notable for its high Notably, however, over 50% of the patients in
blood-brain barrier (BBB) penetrance. This prop- Welsh et al. with known tumor EGFR mutation
erty has prompted a number of trials evaluating status possessed EGFR mutations. Patients in that
the efficacy and toxicity of WBRT with adjuvant study with EGFR-mutated tumors had a median
294 C. Lynch et al.
survival time of 19.1 months compared to erlotinib. There was also trend toward a survival
9.3 months in those with wild-type EGFR tumors. advantage to WBRT over erlotinib, though this
With a total sample of only 17 patients, however, was not significant [74]. Given the intracranial
this difference was not significant (P = 0.534). In activity of osimertinib as a newer generation
contrast, Sperduto et al. did not assess EGFR EGFR-targeting agent [75, 76], and its establish-
mutation status and in the study from Lee et al. ment as first-line therapy for EGFR-mutated non-
only 1 of the 35 patients with known tumor EGFR small cell lung cancer, treatment with osimertinib
mutation status possessed a mutation. The and omission of upfront radiotherapy for small
SATURN trial investigating the use of erlotinib in asymptomatic brain metastases is increasingly
patients with advanced NSCLC demonstrated that being utilized, although further study is needed.
although erlotinib provided a benefit to NSCLC The use of immune checkpoint inhibitors in
patients generally, those with EGFR mutations conjunction with brain radiotherapy is a matter of
derived the greatest benefit from it [71]. It seems active and ongoing study. A retrospective analy-
likely then that erlotinib is most effective in intra- sis of patients with melanoma brain metastases
cranial metastases in which a mutated EGFR receiving ipilimumab (an anti-CTLA-4 mono-
drives cancer growth and proliferation. As such, clonal antibody) with either SRS or WBRT did
further study is warranted in this patient subpopu- not demonstrate a survival advantage for com-
lation. Other EGFR inhibitors such as gefitinib bined WBRT-ipilimumab therapy compared
and icotinib have also been investigated as adju- with historical controls treated with WBRT and
vant therapy with WBRT with similarly mixed bortezomib. The authors did find an advantage to
results. A study of icotinib versus WBRT with SRS and ipilimumab versus SRS alone [77].
chemotherapy in patients with EGFR-mutant There are, however, currently no published RCTs
NSCLC demonstrated superior intracranial pro- investigating the use of WBRT with immunother-
gression-free survival in the icotinib arm (10 vs apy compared to WBRT alone or immunotherapy
4.8 months, P = 0.014) [72]. It is worth noting, alone. Future trials of immune checkpoint inhibi-
however, that a phase II study of NSCLC patients tors and other targeted agents should consider not
has shown a survival advantage to icotinib with just efficacy, but toxicity as well. In particular,
WBRT compared to WBRT alone, with a particu- with significantly improved survivorship with the
lar advantage for patients with EGFR-mutated use of immune checkpoint inhibitors for mela-
tumors [73]. This suggests a possible benefit to noma and non-small cell lung cancer brain metas-
combination therapy rather than icotinib alone. tases, cognitive side effects become a significant
RTOG 1119 is an ongoing study assessing the component of both brain metastatic disease and
treatment of HER2-positive breast cancer patients associated therapies, and new treatments should
with WBRT plus adjuvant trastuzumab and lapa- be evaluated for impact on these symptoms.
tinib. Lapatinib is a dual EGFR and HER2 inhibi-
tor that, unlike trastuzumab, can cross the BBB
and has shown preclinical promise. Until more Optimal Patient Selection: Summary
persuasive clinical evidence emerges, however,
the benefit of combining WBRT and targeted Recent research has helped identify which
therapies for the purpose of improving intracra- patients may stand to benefit the most from
nial control remains unclear. WBRT versus or in conjunction with other defini-
A retrospective analysis of NSCLC patients in tive treatment modalities for brain metastases.
whom radiation therapy was deferred provides WBRT should be considered as primary treat-
further reason for clinicians to exercise caution ment for patients with good performance status
before omitting radiation treatment. The study and with systemic therapy options for managing
found that patients who received upfront SRS for extracranial disease when metastatic lesions
brain metastases had a significantly longer within the brain are not amenable to surgical
median survival time than those receiving upfront resection or SRS. This can occur when a metasta-
sis is too large for SRS and is located in an unre- even more apparent, and the appropriate usage of
sectable region, when the burden of metastatic modern WBRT will be further refined.
disease is too extensive for other techniques (≥ 5
metastases), in the case where there is diffuse dis-
ease (extensive dural, pachymeningeal, or lepto- References
meningeal metastases), or in cases where there is
a possibility for microscopic metastatic disease 1. Brown PD, Ahluwalia MS, Khan OH, Asher AL,
(particularly for limited-stage SCLC responsive Wefel JS, Gondi V. Whole-brain radiotherapy for
to chemotherapy or high brain metastasis veloc- brain metastases: evolution or revolution? J Clin
Oncol. 2018;36(5):483–91.
ity after upfront SRS). These patients, at high risk 2. Chao J-H, Phillips R, Nickson JJ. Roentgen-ray ther-
for developing or experiencing progression of apy of cerebral metastases. Cancer. 1954;7(4):682–9.
multiple brain metastases that may cause neuro- 3. Borgelt B, et al. The palliation of brain metastases:
logic and cognitive impairment, may benefit the final results of the first two studies by the radiation
therapy oncology group. Int J Radiat Oncol Biol Phys.
most from modern WBRT which can alleviate 1980;6(1):1–9.
these symptoms and improve survival. 4. Murray KJ, et al. A randomized phase III study of
Neurocognitive protective strategies of hippo- accelerated hyperfractionation versus standard in
campal avoidance and memantine can effectively patients with unresected brain metastases: a report of
the radiation therapy oncology group (RTOG) 9104.
prevent WBRT-associated cognitive toxicity. Int J Radiat Oncol Biol Phys. 1997;39(3).
5. Rades D, Haatanen T, Schild SE, Dunst J. Dose escala-
tion beyond 30 grays in 10 fractions for patients with mul-
Summary tiple brain metastases. Cancer. 2007;110(6):1345–50.
6. Nieder C, Berberich W, Schnabel K. Tumor-related
prognostic factors for remission of brain metasta-
WBRT remains a valuable asset in the manage- ses after radiotherapy. Int J Radiat Oncol Biol Phys.
ment of brain metastases. Several trials of SRS 1997;39(1):25–30.
versus conventional WBRT demonstrated infe- 7. Stea B, Suh JH, Boyd AP, Cagnoni PJ, Shaw E. Whole-
brain radiotherapy with or without efaproxiral for the
rior cognitive outcomes of WBRT in the setting treatment of brain metastases: determinants of response
of one to four brain metastases. These findings and its prognostic value for subsequent survival. Int J
have led to a declining use of WBRT and rapidly Radiat Oncol Biol Phys. 2006;64(4):1023–30.
rising use of SRS alone. However, practice- 8. Regine WF, Scott C, Murray K, Curran
W. Neurocognitive outcome in brain metastases
changing evidence demonstrating preservation of patients treated with accelerated-fractionation vs.
cognitive function with hippocampal avoidance accelerated-hyperfractionated radiotherapy: an anal-
and memantine has ushered in the modern era of ysis from radiation therapy oncology group study
WBRT. Importantly, prior trials demonstrating 91–04. Int J Radiat Oncol Biol Phys. 2001;51(3).
9. Li J, Bentzen SM, Renschler M, Mehta MP. Regression
inferior cognitive outcomes with conventional after whole-brain radiation therapy for brain metasta-
WBRT did not include these neuroprotective ses correlates with survival and improved neurocogni-
strategies and thus have limited relevance in the tive function. J Clin Oncol. 2007;25(10).
modern management of brain metastases. 10. Gondi V, et al. Hippocampal-sparing whole brain

radiotherapy: a “how-to” technique, utilizing heli-
Several phase III trials are currently accruing or cal tomotherapy and LINAC-based intensity modu-
under development to better define the role of lated radiotherapy. Int J Radiat Oncol Biol Phys.
modern WBRT either in lieu of SRS for newly 2010;78(4).
diagnosed 5–15 brain metastases or small cell lung 1
1. Gondi V, et al. Preservation of neurocognitive function
(NCF) with conformal avoidance of the hippocampus
cancer brain metastases or adjunctive to SRS for during whole-brain radiotherapy (HA-WBRT) for
recurrent brain metastases with high brain metas- brain metastases: preliminary results of phase III trial
tasis velocity. In addition, as improvements in sys- NRG oncology CC001. Int J Radiat Oncol Biol Phys.
temic therapy continue to prolong survival in brain 2018;102(5).
12. Brown PD, Gondi V, Pugh S, Tome WA, Wefel JS,
metastasis patients, the impact of optimizing brain Armstrong TS, et al. Hippocampal avoidance dur-
metastasis control and minimizing associated neu- ing whole-brain radiotherapy plus memantine for
rologic and quality-of-life sequelae will become patients with brain metastases: phase III trial NRG
296 C. Lynch et al.
oncology. J Clin Oncol. 2020;38(10):1019–29. changing pattern with lengthening survival. Cancer.
https://doi.org/10.1200/JCO.19.02767. 1979;44(5).
13. Nolan CP, DeAngelis LM. Neurologic complications 28. Aupérin A, et al. Prophylactic cranial irradiation

of chemotherapy and radiation therapy. Continuum. for patients with small-cell lung cancer in complete
2015;21(2). remission. N Engl J Med. 1999;341.
14. Woodford K. Somnolence syndrome after cranial radi- 29. Meert A-P, et al. Prophylactic cranial irradiation in
ation: a literature review. Radiograp. 2007;54(3):30. small cell lung cancer: a systematic review of the lit-
15. Brown PD, et al. Postoperative stereotactic radio-
erature with meta-analysis. BMC Cancer. 2001;1(5).
surgery compared with whole brain radiotherapy for 30. Slotman B, et al. Prophylactic cranial irradiation

resected metastatic brain disease (NCCTG N107C/ in extensive small-cell lung cancer. N Engl J Med.
CEC·3): a multicentre, randomised, controlled, phase 2007;357(7).
3 trial. Lancet Oncol. 2017;18(8). 31. Takahashi T, et al. Prophylactic cranial irradia-

16. DeAngelis LM, Delattre JY, Posner JB. Radiation- tion versus observation in patients with extensive-
induced dementia in patients cured of brain metasta- disease small-cell lung cancer: a multicentre,
ses. Neurology. 1989;39(6):789–96. randomised, open-label, phase 3 trial. Lancet Oncol.
17. Mulvenna P, et al. Dexamethasone and supportive 2017;18(5):663–71.
care with or without whole brain radiotherapy in 32. Gore EM, et al. Phase III comparison of prophylactic
treating patients with non-small cell lung cancer with cranial irradiation versus observation in patients with
brain metastases unsuitable for resection or stereotac- locally advanced non–small-cell lung cancer: primary
tic radiotherapy (QUARTZ): results from a phase 3, analysis of radiation therapy oncology group study
non-inferiority. Lancet. 2016;388(10055):2004–14. RTOG 0214. J Clin Oncol. 2011;29(3).
18. Sperduto PW, et al. Summary report on the graded 33. Li N, et al. Randomized phase III trial of prophylactic
prognostic assessment: an accurate and facile cranial irradiation versus observation in patients with
diagnosis-specific tool to estimate survival for patients fully resected stage IIIA–N2 nonsmall-cell lung can-
with brain metastases. J Clin Oncol. 2012;30(4). cer and high risk of cerebral metastases after adjuvant
19. Aoyama H, Tago M, Shirato H. Stereotactic radiochemotherapy. Ann Oncol. 2014;26(3).
surgery with or without whole-brain radiotherapy 34. De Ruysscher D, et al. Prophylactic cranial irradia-
for brain metastases secondary analysis of the tion versus observation in radically treated stage iii
JROSG 99-1 randomized clinical trial. JAMA Oncol. non–small-cell lung cancer: a randomized phase III
2015;1(4):457–64. NVALT-11/DLCRG-02 study. J Clin Oncol. 2018.
20. Chang EL, et al. Neurocognition in patients with brain 35. Chen HS, et al. Open-channel block of N-methyl-D-
metastases treated with radiosurgery or radiosurgery aspartate (NMDA) responses by memantine: thera-
plus whole-brain irradiation: a randomised controlled peutic advantage against NMDA receptor-mediated
trial. Lancet Oncol. 2009;10(11). neurotoxicity. J Neurosci. 1992;12(11):4427–36.
21. Kocher M, et al. Adjuvant whole-brain radiotherapy 36. Chen HS, Lipton SA. Mechanism of memantine

versus observation after radiosurgery or surgical block of NMDA-activated channels in rat retinal
resection of one to three cerebral metastases: results ganglion cells: uncompetitive antagonism. J Physiol.
of the EORTC 22952-26001 study. J Clin Oncol. 1997;499(1).
2011;29(2):134–41. 37. Wu PH, et al. Radiation induces acute alterations in
22. Sahgal A, et al. Phase 3 trials of stereotactic radio- neuronal function. PLoS One. 2012;7(5).
surgery with or without whole-brain radiation ther- 38. Zhang D, et al. Radiation induces age-dependent

apy for 1 to 4 brain metastases: individual patient deficits in cortical synaptic plasticity. J Neurooncol.
data meta-analysis. Int J Radiat Oncol Biol Phys. 2018.
2015;91(4):710–7. 39. Orgogozo J-M, Rigaud A-S, Stöffler A, Möbius

23. Brown PD, et al. Effect of radiosurgery alone vs radio- H-J, Forette F. Efficacy and safety of Memantine in
surgery with whole brain radiation therapy on cogni- patients with mild to moderate vascular dementia: a
tive function in patients with 1 to 3 brain metastases: randomized, placebo-controlled trial (MMM 300).
a randomized clinical trial. JAMA. 2016;316(4). Stroke. 2002;33(7):1834–9.
24. Patchell RA, et al. Postoperative radiotherapy in the 40. Wilcock G, Möbius HJ, Stöffler A. A double-blind,
treatment of single metastases to the brain a random- placebo-controlled multicentre study of memantine in
ized trial. JAMA. 1998;280(17). mild to moderate vascular dementia (MMM500). Int
25. Mahajan A, et al. Post-operative stereotactic radiosur- Clin Psychopharmacol. 2002;17(6):297–305.
gery versus observation for completely resected brain 41. Brown PD, et al. Memantine for the prevention of
metastases: a single-centre, randomised, controlled, cognitive dysfunction in patients receiving whole-
phase 3 trial. Lancet Oncol. 2017;18(8). brain radiotherapy: a randomized, double-blind,
26. Gondi V, Mehta MP. Control versus cognition: the placebo-controlled trial. Neuro Oncol. 2013;15(10).
changing paradigm of adjuvant therapy for resected 42. Deng W, Aimone JB, Gage FH. New neurons and new
brain metastasis. Neuro Oncol. 2018;20(1):2–3. memories: how does adult hippocampal neurogenesis
27. Nugent JL, et al. CNS metastases in small cell
affect learning and memory? Nat Rev Neurosci.
bronchogenic carcinoma. Increasing frequency and 2010;11(5):339–50.
43.
Gondi V, Hermann BP, Mehta MP, Tomé 57. Tyler R, et al. Radiosurgery alone is associated with
WA. Hippocampal dosimetry predicts neurocogni- favorable outcomes for brain metastases from small-
tive function impairment after fractionated stereo- cell lung cancer. Lung Cancer. 2018;120.
tactic radiotherapy for benign or low-grade adult 58. Mayinger M, Kraft J, Lohaus N, Weller M, Schanne
brain tumors. Int J Radiat Oncol Biol Phys. February D, Heitmann J, et al. Leukoencephalopathy after pro-
2013;85(2):348–54. phylactic whole-brain irradiation with or without hip-
44. Gondi V, et al. Preservation of memory with confor- pocampal sparing: a longitudinal magnetic resonance
mal avoidance of the hippocampal neural stem-cell imaging analysis. Eur J Cancer. 2020;124:194–203.
compartment during whole-brain radiotherapy for https://doi.org/10.1016/j.ejca.2019.11.008. Epub 2019
brain metastases (RTOG 0933): a phase II multi- Dec 6.
institutional trial. J Clin Oncol. 2014;32(34):3810–6. 59. Vees H, Caparrotti F, Eboulet EI, Xyrafas A, Fuhrer
45. Yamamoto M, et al. Stereotactic radiosurgery for
A, Meier U, et al. Impact of Early Prophylactic
patients with multiple brain metastases (JLGK0901): Cranial Irradiation With Hippocampal Avoidance on
a multi-institutional prospective observational study. Neurocognitive Function in Patients With Limited
Lancet Oncol. 2014;15(4):387–95. Disease Small Cell Lung Cancer. A Multicenter
46. Soike MH, et al. Does stereotactic radiosurgery
Phase 2 Trial (SAKK 15/12). Int J Radiat Oncol Biol
have a role in the management of patients present- Phys. 2020 Mar 4. pii: S0360-3016(20)30255-8.
ing with 4 or more brain metastases? Neurosurgery. https://doi.org/10.1016/j.ijrobp.2020.02.029. [Epub
2019;84(3). ahead of print].
47. Aoyama H, et al. Stereotactic radiosurgery plus
60. Kirson ED, et al. Disruption of Cancer cell repli-
whole-brain radiation therapy vs stereotactic radio- cation by alternating electric fields. Cancer Res.
surgery alone for treatment of brain metastases: a ran- 2004;64(9):3288–95.
domized controlled trial. JAMA. 2006;295(21). 61. Stupp R, et al. Effect of tumor-treating fields plus
48. Farris M, et al. Brain metastasis velocity: a novel prog- maintenance temozolomide vs maintenance temozolo-
nostic metric predictive of overall survival and free- mide alone on survival in patients with glioblastoma:
dom from whole-brain radiation therapy after distant a randomized clinical trial. JAMA. 2017;318(23).
brain failure following upfront radiosurgery alone. Int 62. Komarnicky LT, Phillips TL, Martz K, Asbell S,

J Radiat Oncol Biol Phys. 2017;98(1):131–41. Isaacson S, Urtasun R. A randomized phase III pro-
49. Mctyre E, et al. Multi-institutional validation of brain tocol for the evaluation of misonidazole combined
metastasis velocity, a recently defined predictor of with radiation in the treatment of patients with brain
outcomes following stereotactic radiosurgery. Int J metastases (RTOG-7916). Int J Radiat Oncol Biol
Radiat Oncol Biol Phys. 2017;99(S2):E93. Phys. 1991;20(1).
50. Yamamoto M, et al. Validity of a recently proposed 63. Zeng YC, et al. Radiation-enhancing effect of sodium
prognostic grading index, brain metastasis velocity, glycididazole in patients suffering from non-small
for patients with brain metastasis undergoing mul- cell lung cancer with multiple brain metastases: a ran-
tiple radiosurgical procedures. Int J Radiat Oncol Biol domized, placebo-controlled study. Cancer Radiother.
Phys. 2019;103(3). 2016;20(3).
51. Fritz C, et al. Repeated courses of radiosurgery for new 64. Mehta MP, et al. Survival and neurologic outcomes
brain metastases to defer whole brain radiotherapy: fea- in a randomized trial of motexafin gadolinium and
sibility and outcome with validation of the new prognos- whole-brain radiation therapy in brain metastases. J
tic metric brain metastasis velocity. Front Oncol. 2018;8. Clin Oncol. 2003;21(13).
52. Giuliani M, et al. Utilization of prophylactic cranial 65. Mehta MP, et al. Motexafin gadolinium combined
irradiation in patients with limited stage small cell with prompt whole brain radiotherapy prolongs time
lung carcinoma. Cancer. 2010;116(24). to neurologic progression in non–small-cell lung can-
53. Lok B, et al. Factors influencing the utilization of pro- cer patients with brain metastases: results of a phase
phylactic cranial irradiation in patients with limited-stage III trial. Int J Radiat Oncol Biol Phys. 2009;73(4).
small cell lung cancer. Adv Radiat Oncol. 2017;2(4). 66. Antonadou D, et al. Whole brain radiotherapy alone
54. Cifarelli C, et al. Role of gamma knife radiosurgery in or in combination with temozolomide for brain metas-
small cell lung cancer: a multi-institutional retrospec- tases. A phase III study. Int J Radiat Oncol Biol Phys.
tive study of the international radiosurgery research 2002;2(1).
foundation (IRRF). Neurosurgery. 2019; epub ahead 67. Sperduto PW, et al. A phase 3 trial of whole brain
of print. radiation therapy and stereotactic radiosurgery alone
55. Serizawa T, et al. Gamma knife radiosurgery for meta- versus WBRT and SRS with temozolomide or erlo-
static brain tumors from lung cancer: a comparison tinib for non-small cell lung cancer and 1 to 3 brain
between small cell and non-small cell carcinoma. J metastases: radiation therapy oncology group 0320.
Neurosurg. 2002;97(5). Int J Radiat Oncol Biol Phys. 2013;85(5):1312–8.
56. Yomo S, et al. Is stereotactic radiosurgery a rational 68. Zhao Q, et al. Brain radiotherapy plus concurrent
treatment option for brain metastases from small cell temozolomide versus radiotherapy alone for patients
lung cancer? A retrospective analysis of 70 consecu- with brain metastases: a meta-analysis. PLoS One.
tive patients. BMC Cancer. 2015;15(95):95. 2016;11(3).
298 C. Lynch et al.
69. Welsh JW, et al. Phase II trial of erlotinib plus concur- 74. Magnuson WJ, Yeung JT, Guillod PD, Gettinger SN,
rent whole-brain radiation therapy for patients with Yu JB, Chiang VL. Impact of deferring radiation ther-
brain metastases from non–small-cell lung cancer. J apy in patients with epidermal growth factor recep-
Clin Oncol. 2013;31(7). tor–mutant non-small cell lung cancer who develop
70. Lee SM, et al. Randomized trial of erlotinib plus brain metastases. Int J Radiat Oncol Biol Phys.
whole-brain radiotherapy for NSCLC patients 2016;95(2):673–9.
with multiple brain metastases. J Natl Cancer Inst. 75.
Goss G, Tsai CM, Shepherd FA, Ahn MJ,
2014;106(7). Bazhenova L, et al. CNS response to osimertinib
71. Neal JW. The SATURN trial: the value of mainte- in patients with T790M-positive advanced NSCLC:
nance erlotinib in patients with non-small-cell lung pooled data from two phase II trials. Ann Oncol.
cancer. Future Oncol. 2010;6(10). 2018;29(3):687–93.
72. Yang J-J, et al. Icotinib versus whole-brain irradiation 76. Wu YL, Ahn MJ, Garassino MC, Han JY, et al. CNS
in patients with EGFR -mutant non-small-cell lung efficacy of Osimertinib in patients with T790M-
cancer and multiple brain metastases (BRAIN): a positive advanced non-small-cell lung cancer: data
multicentre, phase 3, open-label, parallel, randomised from a randomized phase III trial (AURA3). J Clin
controlled trial. Lancet Respir Med. 2017;5(9). Oncol. 2018;36(26):2702–9.
73. Fan Y, et al. A phase II study of icotinib and whole- 77. Silk AW, Bassetti MF, West BT, Tsien CI, Lao

brain radiotherapy in Chinese patients with brain CD. Ipilimumab and radiation therapy for melanoma
metastases from non-small cell lung cancer. Cancer brain metastases. Cancer Med. 2013;2(6).
Chemother Pharmacol. 2015;76(3):517–23.
Stereotactic Radiosurgery
Technology
21
Diana A. R. Julie and Jonathan P.S. Knisely
Introduction able intracranial lesions [1, 3–6, 11, 14–16].

Leksell and his team experimented with dental x-ray
Stereotactic radiosurgery (SRS) aims to non- tubes, as well as with proton therapy, before ulti-
invasively deliver ablative radiation doses to intra- mately settling upon Cobalt-60 (Co-60) sources for
cranial lesions, with minimal dose to nearby administering SRS. The first Gamma Knife (GK)
healthy tissues, traditionally in a single treatment. unit employed 179 Co-60 sources convergently
The fundamental principles of SRS include: (1) focused upon a stereotactically targeted lesion. The
stereotactic target localization; (2) high dose, pre- first GK treatment occurred in 1967 for the manage-
cise radiation delivery; (3) steep dose fall off, mini- ment of a craniopharyngioma [13–16].
mizing dose to surrounding tissues; and (4) For many decades, the Leksell GK was the
acquisition of volumetric imaging for treatment gold standard in SRS. However, since the 1950s,
planning [1–11]. Hypofractionated SRS, in which in parallel with advances in imaging and comput-
the radiation dose is delivered in two to five frac- ing, SRS technology has rapidly evolved, and
tions, has also been employed [9, 11]. today, numerous platforms are available for intra-
Hypofractionation confers several benefits, includ- cranial SRS [1, 3–5, 7, 9, 11, 15]. The major sys-
ing decreased toxicity risk, ability to treat larger tems to be reviewed in this chapter include the
lesions, and potential for safer re-irradiation [2, 7, GK, manufactured by Elekta (Elekta AB,
8, 11, 12]. Since its development, SRS has become Stockholm, Sweden), linear accelerator (linac)
a key treatment option for a variety of benign and SRS systems, and the CyberKnife (CK), manu-
malignant intracranial lesions, as an alternative or factured by Accuray (Accuray Inc., Sunnyvale,
adjunct to surgery [1, 2, 4, 5, 7, 10–13]. Lesions CA) [3, 8, 9, 12]. Proton therapy is also emerging
appropriate for SRS include benign and malignant as an SRS treatment option [1, 17–19].
tumors, arteriovenous malformations (AVMs), and
some functional disorders [1, 3, 5, 7, 10–13]. The
use of SRS and its applications continue to expand. General SRS Concepts
SRS was devised by the Swedish neurosurgeon
Lars Leksell in the 1950s as a treatment for inoper- Beam Shaping
D. A. R. Julie (*) · J. P.S. Knisely Collimators are devices used to shape the radia-
Department of Radiation Oncology, NewYork- tion beam so that its edges are sharply defined.
Presbyterian/Weill Cornell Medicine, Conical collimators are typically composed of
New York, NY, USA lead or tungsten and come in a variety of aperture
e-mail: daj9066@nyp.org

https://doi.org/10.1007/978-3-030-42958-4_21
300 D. A. R. Julie and J. P.S. Knisely
diameters. They are divergently milled to mini- iteratively adjust input specifications for the algo-
mize the penumbra, providing a sharp beam edge rithm and also manually tweak the final plan [13].
and rapid dose fall off beyond the target [16, 20]. Inverse planning carries some advantages over
Their use can yield small, sharply focused beams, forward planning. Since trial-and-error is not
as small as 0.4–0.5 cm. Multi-leaf collimators required, plans can be generated more efficiently.
(MLCs) are composed of many (up to160) paired Additionally, plans of better quality should be
tungsten leaves, each with a width of 0.5–1 cm; generated, as the optimization algorithm can eval-
micro-MLCs (mMLCs), with width 0.2–0.5 cm, uate and compare thousands of plans [20].
can also be used. A computer controls the posi-
tion of each independent leaf at every point in
time. Compared to circular collimators, MLCs Immobilization and Image Guidance
and mMLCs improve conformity to irregularly
shaped target volumes and homogeneity of dose Given the high doses per fraction delivered with
distribution throughout the target volume [1, 3, intracranial SRS and the close proximity to sensi-
6–8, 13, 16, 20–22]. The CyberKnife may be tive OARs, accurate patient positioning is vital to
equipped with MLCs that function much like ensuring efficacy and safety. End-to-end submil-
those on other linacs or may be equipped with an limeter accuracy and precision are desired. The
‘iris’ collimator that employs two banks of six methods used to position and immobilize patients
tungsten blocks that provide a dodecagonal beam for SRS can be broadly categorized into two
whose size can be changed at will. Circular coni- groups: frame-based and frameless systems [4, 8,
cal collimators in a range of sizes may also be 12, 15, 21–26].
employed. When SRS was first introduced, rigid head-
For proton beam therapy, the generation of frames applied to the patient’s skull and subse-
focused beams requires custom apertures and quently attached to the treatment table were
range compensators. Passively scattered beams always employed, and SRS continues to be deliv-
and scanned pencil beam treatments have slightly ered with headframes for both GK and linac plat-
different equipment requirements to generate a forms, depending upon facility capabilities and
spread-out Bragg peak beam at the depth of the expertise. Invasive frames consist of a rigid head-
target tissue. Only two to four beams are used for frame, attached to the outer table of the skull
proton radiosurgery because of the dosimetric using three to four metal pins, and serve two pur-
characteristics of protons. poses, immobilizing the patient and providing
stereotactic coordinates for target localization [4,
5, 7, 12, 15, 21, 24, 27, 28]. Studies of headframe
Treatment Planning systems have demonstrated submillimeter
intrafractional translational and rotational motion
All modern SRS platforms are capable of both [12, 28]. There are several shortcomings of the
forward and inverse radiation planning. With for- invasive headframe, including patient inconve-
ward planning, the planner completes a repetitive nience and discomfort, risks associated with a
trial-and-error process, sequentially changing (minimally) invasive procedure, and regimented
delivery parameters, such as beam shape, angle, workflow where all treatment-related activities
or weight, to improve the plan. This process con- must be completed in 1 day. Errors can occur
tinues until an acceptable plan is generated. In with frame-based SRS secondary to frame slip-
contrast, inverse planning is a more automated page or deformation, and can also be introduced
approach. First, the target and organs at risk from inaccuracy in imaging or image co-
(OARs) are outlined, and treatment goals are set registration [7, 12, 21, 27, 28].
with regard to these structures. Next, an optimiza- To overcome these drawbacks of invasive
tion algorithm generates a plan that best meets the headframes, non-invasive immobilization sys-
predefined goals [13, 20, 21, 23]. The planner can tems have been developed. Today, both GK and
21 Stereotactic Radiosurgery Technology 301
linac can deliver frameless SRS, while the CK is to confirm patient positioning and target location
a dedicated frameless platform. Frameless SRS relative to the treatment plan. An alternate image-
improves patient comfort and convenience, guidance system integrates stereoscopic x-rays in
avoids the potential complications of frame fixa- the treatment room, such that two orthogonal
tion, and facilitates hypofractionated treatment. kilovoltage (kV) x-rays are taken. Digitally
Frameless SRS is also more convenient for pro- reconstructed radiographs (DRRs) are generated,
viders, as planning can occur over several days. which are x-ray images synthesized from the
This additional time can be especially important planning CT. Bony anatomy is compared between
for optimizing complex plans. Given these advan- the x-rays and DRRs, and patient position can be
tages, there has recently been a transition to pre- verified or corrected [3, 4, 7, 12, 21, 23, 27]. This
dominantly frameless systems [4, 7, 10, 12, 16, method can also be used to acquire images during
21, 27, 28]. treatment delivery, monitoring for patient motion.
Many non-invasive SRS immobilization sys- For intrafraction monitoring, optical tracking can
tems have been introduced. Some systems use a also be used. Reflective markers can be placed on
headframe attached to the treatment table, com- the mask, detected by wall-mounted infrared
bined with a mouthpiece with a vacuum-fixed cameras, and compared to fixed reflectors on the
bite-block. The individualized bite-block is suc- couch head cradle. A threshold level is set for
tioned to the patient’s hard palate and maxillary allowable patient deviation from the initial posi-
teeth, and loss of vacuum is a surrogate for tion, and treatment delivery is automatically
motion. Vacuum bite-block systems require stopped if this is exceeded [12, 21, 27]. It is com-
patient compliance and adequate dentition. mon for proton facilities performing cranial treat-
Perhaps the simplest frameless system in use is ments to implant several tiny stainless steel BBs
the thermoplastic mask. With this setup a custom into the outer table of the skull under local anes-
thermoplastic mask is created, with a mouthpiece thesia to serve as a rigid fiducial system that can
and supplemental reinforcing strips across the be employed to confirm accuracy of patient posi-
forehead, below the nose and across the chin. tioning at the time of treatment.
Frameless SRS is not without shortcomings, Invasive headframes were long considered the
including most importantly inaccuracies in most accurate immobilization and localization
patient positioning and target localization. Since systems. However, with experienced users, mod-
frameless systems do not rigidly immobilize, ern frameless systems combined with image
there may be errors due to patient motion as well guidance can achieve submillimeter accuracy,
as setup errors with subsequent immobilizations; and setup accuracies routinely approach those
machine characteristics may also fluctuate achieved with frame-based treatments [4, 7, 12,
slightly from day to day [12, 21, 27, 28]. 16, 21, 27, 29–34].
In order to replicate the reliability of frame-
based SRS with frameless systems, sophisticated
image guidance is crucial for initial patient posi- Workflow
tioning, as well as real-time or near real-time
monitoring of intrafraction motion. Image- The general workflow required to administer
guidance systems commonly employed for SRS SRS is similar regardless of the system used.
include onboard CBCT (cone beam computed Figure 21.1 provides an overview of frame-based
tomography) and ceiling- and floor-mounted and frameless SRS treatment workflow. When
x-rays [3, 9, 12, 21, 27, 28]. The term CBCT delivering frame-based SRS, the first step on the
derives from the conical x-ray beam employed, day of treatment is attachment of the headframe.
unlike the fan beam used in diagnostic CT (com- The rigid lightweight frame is affixed with metal
puted tomography) [3]. CBCT is now integrated screws penetrating the skin and outer table of the
into nearly all modern SRS systems and gener- skull at three or four points. Next, volumetric
ates high-quality 3D images, which can be used imaging is acquired to ascertain target and patient
a
Lesion and Image-
Diagnostic Frameless Volumetric Diagnostic Treatment Treatment Motion
OAR Plan QA guided
imaging immobilization imaging imaging planning delivery monitoring
delineation positioning
b
Lesion and
Diagnostic Volumetric Diagnostic Treatment Positioning Treatment
Headframe OAR Plan QA
imaging imaging imaging planning verification delivery
delineation
Fig. 21.1 General workflow for (a) frame-based SRS and (b) frameless SRS
position relative to the stereotactic frame. A ste- Gamma Knife

reotactic MRI brain, or other required diagnostic
imaging, is obtained, as necessary. This can be eam Properties, Arrangement,
B
performed prior to the treatment day, or with the and Shaping
frame in place, but the frame limits the imaging
that can be obtained. Diagnostic imaging is then Gamma rays are a form of ionizating radiation
rigidly co-registered with the volumetric imaging produced from the radioactive decay of a Co-60
for treatment planning. Target volumes and nucleus. Co-60 sources have a half-life of approx-
OARs are contoured on the co-registered data imately 5 years and produce gamma radiation of
sets, and the treatment planning software (TPS) 1.17 and 1.33 MeV (mean 1.25 MeV). The total
is used to generate the plan [1, 10, 15, 21, 23–25]. activity of new GK Co-60 sources is approxi-
The plan must be approved by the treating physi- mately 6000 Ci, with a dose rate of 3.3–3.6 Gy/
cians, and quality assurance (QA) checks must be min. Over time, the Co-60 sources decay and
performed to ensure that the plan generated in treatment times increase, such that the source
virtual reality parallels what is delivered to the must be replaced approximately every 5 years [1,
patient. Once the plan is ready to be delivered, 2, 15, 16, 24].
the patient lies supine and the headframe is again A GK uses either 192 or 201 highly collimated
affixed to the treatment table. Patient positioning Co-60 sources arranged hemispherically or coni-
is verified, and treatment is delivered [1, 10, 15, cally around an isocenter, representing the point
21, 24, 27]. where the beams converge. These sources are able
When delivering frameless SRS, the first step to deliver radiation to a specific target volume in a
is creation of any of a variety of SRS-quality highly conformal manner, with sharp dose fall off
immobilization devices (i.e., thermoplastic within a few millimeters, and minimal dose to
mask). With the patient immobilized, a non- surrounding tissue [1, 2, 4–7, 15, 16, 24, 26].
contrast CT scan of the head is acquired, which is Shaping of each Co-60 beam, or ‘shot,’ is accom-
necessary for target localization and treatment plished by a secondary circular collimator system
planning. A stereotactic MRI brain, or other consisting of collimators of different diameters (4,
required diagnostic imaging, is obtained prior to 8, and 16 mm or 4, 18, 14, and 18 mm). These
treatment planning and rigidly co-registered with collimators can be combined or selectively
the CT scan. Target and OAR delineation, treat- blocked to allow conformal treatment of lesions
ment planning, plan approval, and QA proceed as of complex shape [1, 5, 7, 10, 15, 16, 24].
with frame-based treatment. For treatment deliv-
ery, the patient lies supine and is immobilized.
Image guidance is used to confirm patient posi- Treatment Planning and Delivery
tioning in 3D space relative to the planning CT
scan. With frameless SRS, patient motion during With GK SRS, dose is prescribed to an isodose
treatment is monitored using real-time or near surface between 30% and 90%, most often 50–
real-time imaging [1, 10, 13, 15, 21–25, 27]. 80%. This increases the maximum dose and plan
inhomogeneity, with a high central dose and steep Modern Model

fall off [2, 4, 8, 15, 20, 24]. GK treatment plan-
ning employs Leksell Gamma Plan, a dedicated The most modern GK system is the Icon (Elekta
TPS capable of single or multiple isocenter treat- AB, Stockholm, Sweden). It consists of 192
ments, and forward or inverse planning. Simple Co-60 sources in a conical arrangement of 8 sec-
spherical lesions can be treated using a single iso- tors, with 24 sources per sector. It also contains
center, while complex volumes are created by several collimator sizes (4, 8, and 16 mm). Each
combining dose to multiple isocenters. A plan is source can be placed in front of a specific collima-
generated by ‘shot packing,’ combining individ- tor or can be blocked [1, 5, 7, 15, 24]. Intermediate
ual shots of radiation of varying sizes to create a sizes can be achieved by sequential combination
dose distribution corresponding to the desired of collimators. For the Icon system, all source,
volume. For treatment delivery, the isocenter is collimator, and couch motions are automated,
stereotactically aligned with the focal point of the representing a key improvement over prior mod-
GK sources, and radiation is administered one els. The entire collimation system is embedded
isocenter at a time [1, 5, 10, 15, 16, 20, 24]. within the unit as a 12–cm-thick tungsten collima-
tor array, with no secondary collimator helmet.
This results in a 300% increase in the inner diam-
eter of the unit compared to prior models, allow-
Immobilization and Image Guidance ing for treatment of multiple or peripheral lesions,
with minimal collision risk. Compared to prior
Traditionally, GK SRS is delivered with a rigid models, the majority of sources are closer to the
headframe, with accuracy, documented as isocenter in the Icon, increasing the dose rate [5,
<0.3 mm. More recently, frameless systems have 7, 15, 24, 27]. Increased automation and this
been developed for GK SRS, with target localiza- higher dose rate combine to reduced treatment
tion accomplished via an onboard CBCT, specifi- times with the Icon. The Icon can deliver treat-
cally integrated into the GK machine. The accuracy ment with improved conformity, dose fall off,
of the GK CBCT imaging system has been inves- sparing of adjacent OARs, and sparing of normal
tigated, with excellent results, comparable to brain tissue compared to prior GK models [5, 7,
frame-based treatment. Intrafraction motion can 24, 35–38]. This machine can be used to treat
be monitored with optical tracking [1, 15, 24, 27]. intracranial lesions or vertebral lesions up to and
including the C3 level. The Icon can employ fra-
meless immobilization and is equipped with on-
board CBCT and optical tracking for intrafraction
Cost motion [1, 4, 5, 15, 24]. With the Icon, studies
have shown a total error of below 1 mm [1, 2, 9,
The initial cost of a GK unit is approximately 15, 24, 39].
US$3.2 million, with a total start-up cost of
US$3–5 million. Shielding requirements are
lower for GK relative to linac systems, and there- Linac
fore, less expensive. Source replacement every
5 years costs US$0.5–1 million. The break-even Linacs were first developed in the 1950s and have
patient volume for GK in the United States has been used since the 1960s for the overwhelming
been estimated at 86 patients annually, which may majority of patients treated with conventional
limit its use for smaller practices. Relatively large radiotherapy (RT) [13, 21, 22]. Leksell and his
annual volumes of over 200 patients (with reim- colleagues did not pursue linac SRS because
bursement levels at current rates in the United existing linacs had a variety of shortcomings that
States) are required for GK to be cost effective made them unreliable, including low beam
relative to linac SRS [16–18, 20, 21, 23, 24]. energy, low output, limited range of motion of the
gantry and couch, and inaccuracies in patient Treatment Planning and Delivery

positioning and dose delivery. In the 1990s, many
of these limitations were overcome and linac In linac SRS, the treatment is often prescribed to
SRS systems were developed [1, 3, 6, 7, 14, 21, the 80–90% isodose surface. There are dedicated
22]. Key technological developments included TPS for linac SRS, which can create single and
smaller machines, higher beam energy and out- multiple isocenter plans, with planar and non-
put, improved collimation, flattening filter-free coplanar fields, and forward or inverse planning.
(FFF) beams, greater and more precise automated In the earliest linac SRS plans, a single isocenter
gantry and couch mobility, improved target was used, creating a roughly spherical dose dis-
localization and image guidance systems, and tribution. Non-spherical targets were treated
dedicated TPS [1, 5, 7, 21, 22]. with the smallest sphere of dose encompassing
the entire target. With multiple isocenter plans,
by combining dose distributions, more complex
eam Properties, Arrangement,
B treatment volumes can be delivered [8, 20–22].
and Shaping There are several different methods available
for treatment planning and delivery with linac
In a linac, an electron beam is accelerated SRS, including static fields, dynamic conformal
toward a heavy metal alloy, and the interaction arcs, intensity-modulated radiotherapy (IMRT),
of the electron beam with the metal produces and volumetric modulated arc therapy (VMAT).
x-rays, which can be focused upon a target. For In static field planning, fields of fixed position
linac SRS, numerous high-energy x-ray beams are used, and shaped with collimators, equiva-
are sequentially focused on the intracranial tar- lent to 3D conformal technique in conventional
get [1, 2, 5, 13, 21, 22]. The energies commonly RT. With dynamic conformal arcs, the field
used for linac SRS are 6 and 10 MV, associated rotates in an arc around the target, and MLCs or
with constant dose rates of 3–6 Gy/min over the mMLCs alter field shape to match the shape of
lifetime of the machine [16, 21, 22]. The gantry the target at each position [16, 20–22]. With
can be rotated around an isocenter located IMRT planning, the time that the beam spends in
within the patient, and the couch itself can any location and the leaf configuration can be
also be rotated, so that the numerous non- modulated during radiation delivery, along with
coplanar beams are focused upon the lesion, the photon fluence [1, 3, 12, 13, 16, 20–22].
allowing for conformal dose around a target IMRT allows for increased dose to areas of
volume, with minimal dose to surrounding tis- tumor and decreased dose to normal tissues, in a
sues [8, 13, 21, 22]. more complex and precise manner. VMAT
The linac radiation beam is shaped by conical allows for delivery of intensity-modulated radia-
collimators or by MLCs and mMLCs [1, 2, 5, 7, tion along an arc around the isocenter, rather
13, 16, 20–22]. In conventional RT, the beam than from a series of static positions, and can be
passes through a flattening filter (FF), which delivered from n on-coplanar arcs, improving the
homogenizes the beam as low energy x-rays are dose fall off outside the target. As the gantry
filtered out. Treating without an FF significantly rotates along its arc, dose rate, MLC shape,
increases the dose rate. For the high fractional intensity, and gantry speed can all be indepen-
doses delivered with SRS, flattening filter-free dently controlled by the TPS running the linac.
(FFF) treatment allows for significant decreases IMRT and VMAT planning allows for highly
in treatment time. In addition to increased patient conformal complex dose distribution plans, with
comfort, the reduced treatment time potentially sharp dose fall off, as well as shortened treat-
decreases patient intrafraction motion. All major ment time [12, 20–22].
linac SRS platforms now offer a FFF treatment With advances in diagnostic techniques and
mode [1, 12, 21]. oncologic therapies, patients with intracranial
metastatic disease are living longer. This, com- used for conventional RT. Because linacs are
bined with improved SRS capabilities, has more complex than other SRS platforms, they
resulted in patients with increasing numbers of require more intensive maintenance and QA [10,
brain metastases receiving SRS. Consequently, 16, 22].
vendors have been developing TPS specifically
for planning treatment of multiple brain metasta-
ses. Elements (Brainlab) and HyperArc (Varian) Modern Model
are examples of such TPS [21].
An example of a modern linac for SRS treat-
ment is the Novalis TX (Brainlab AG), a linac
Immobilization and Image Guidance able to deliver radiation at 2 energies, with beams
shaped by an mMLC consisting of 120 2.5 mm
Initially linac SRS used a rigid headframe, but leaves. The Novalis TX system can be used for
in modern systems, frameless treatment can be intra- or extracranial treatments, and for SRS,
delivered, in combination with image guidance SBRT or conventional RT. It can deliver SRS
[12, 16, 21, 22, 27]. Linac image guidance can treatment via static fields, dynamic arcs, IMRT,
be accomplished with a CBCT mounted perpen- or VMAT. Novalis TX can perform frame-based
dicularly on the linac gantry, using orthogonal and frameless SRS, as necessary. For image
kV x-rays, optical tracking, or a combination of guidance, the Novalis TX machine includes dedi-
these tools [3, 5, 12, 21, 22, 27]. Modern linac cated stereotactic kV x-ray imaging equipment
couches have 6 degrees of freedom (6DOF), mounted in the floor and above the linac and a
allowing for motion along the three primary CBCT mounted on the linac gantry [1, 7, 21,
Cartesian axes, as well as three rotational direc- 22]. Treatment of multiple metastases simultane-
tions (pitch, roll, yaw). Such flexibility of couch ously with a non-coplanar VMAT or 3D forward
motion was crucial in the development of fra- planned approach on a linac permits brain metas-
meless SRS, as it allowed for the most precise tasis radiosurgical treatment session times to be
patient positioning based on image guidance [7, reduced below a half an hour, though consider-
12, 21–23, 27]. Overall, modern linacs are able able time is required to perform the necessary
to deliver SRS with the same accuracy and pre- planning steps, of course.
cision as GK [39].
CyberKnife
Cost
The CK (Accuray Inc., Sunnyvale, CA), devel-
Dedicated linac SRS platforms are less expensive oped by neurosurgeon John Adler in the late
than GK SRS, and modifying an existing linac to 1990s, is a lightweight 6 MV linac mounted on
perform SRS is the most cost-effective means of a robotic arm. The goal in designing the CK was
establishing an SRS program. In the early 2000s, to create an SRS tool, which could deliver high
the cost of implementing a new linac SRS system doses of RT conformally to target lesions, with
was approximately US$2.5–3.2 million. Updated sharp dose fall off, without requiring an invasive
estimates of cost are difficult to obtain, as this headframe or being limited to intracranial sites
information is not made public by vendors but is [2–7, 10, 13, 21, 23, 25]. It was first used to treat
likely US$3–4 million. A TPS may cost an addi- patients in 1994 at Stanford University, and at
tional several hundred thousand dollars. It has that time was called the Neurotron 1000. The
been estimated that 122 SRS patients must be CK system gained FDA approval for intracra-
treated annually to break even with a dedicated nial use in 1999 and for full-body use in 2001
linac SRS system, though the system can also be [5, 6, 13, 23].

B inverse planning [1, 6, 7, 9, 21, 23, 25]. Non-
and Shaping isocentric plans are particularly useful for con-
formally and homogeneously treating complex
The CK produces an unflattened 6 MV beam target volumes. Since treatment time is a signifi-
from a lightweight linac with dose rates of cant concern with CK SRS, an attempt is made to
3–6 Gy/min, constant over the lifetime of the reduce the number of nodes, while maintaining
machine [4, 6, 12, 21, 23, 25]. A major innova- plan quality. Based upon the treatment plan, the
tion of the CK system is the robotic arm, which robotic arm travels along a predetermined path
moves the linac about the patient with 6 DOF, from node to node, delivering radiation in a step-
compared to conventional linac gantries, which and-shoot manner [9, 21, 23, 25].
can rotate only in one plane. It directs the radia-
tion from numerous angles and positions above
(but not below) the patient, such that hundreds of Immobilization and Image Guidance
small, circular non-coplanar beams converge
upon the target [1, 2, 5–7, 10, 13, 21, 23, 25]. The CK is an exclusively frameless SRS system
Each beam of treatment is defined by a ‘node,’ that employs real-time image guidance to assure
consisting of a robotic arm location, a beam accuracy. In CK, image guidance for initial
direction, and a field size. Usually, 23–133 unique patient setup, prior to treatment delivery from
beams are employed in a single CK treatment, each node, and for movement correction during
but there are over 1000 possible beam directions. treatment consists of the orthogonal kV x-ray
This flexibility is a distinct advantage of CK, method described [1, 4, 6–10, 12–14, 21, 23,
allowing for treatment planning with excellent 25]. For CK, this system is referred to as 6D
homogeneity and conformity, even with large and skull tracking. If any patient motion or target
complex targets [6, 9, 12, 21, 23, 25]. misalignment occurs, radiation delivery is auto-
The CK system can use fixed circular colli- matically halted, the robot readjusts, and treat-
mators, creating a beam diameter of 5–60 mm. ment resumes [6, 7, 14, 21, 23, 25]. Different
The more advanced IRIS variable collimator publications recommend different frequencies of
employs 12 secondary tungsten-copper alloy cir- image guidance during treatment, ranging from
cular collimators and automatically changes every 1–5 minutes [12].
aperture size, decreasing treatment time [6, 7,
12, 21, 23, 25]. The newest CK collimation sys-
tem, the InCise, consists of 2 banks of 26 leaves, Cost
and brings MLC beam shaping to CK treatment.
The InCise MLC is able to further reduce treat- CK systems require a substantial initial invest-
ment time and improve conformity, especially ment, in addition to considerable maintenance
for complex targets [23]. The InCise system can expenditures. The cost of a new CK system is
also be employed for multitarget radiosurgical approximately US$3.5 million [16, 25]. Since
treatments to decrease treatment delivery radiation beams can be directed from such a
time [40]. variety of locations and directions, all walls,
including the ceiling, must be primary radiation
barriers reinforced with concrete, so a dedicated
Treatment Planning and Delivery or upgraded vault is required. Alternatively, if
CK is to be used in a conventional linac room,
CK treatments are prescribed to the 50–80% iso- the beam angles employed must be limited [5].
dose surface [21, 23, 25]. Given their unique fea- Hardware and software for the CK system are
tures, CK systems use a dedicated TPS, Multiplan, estimated to cost US$225,000–450,000 per
or the more recent Precision, capable of isocen- year. Approximately 109 treatments annually
tric and non-isocentric treatment, with forward or are necessary to break even with the CK system,
but these numbers can be reached with the addi- dosimetric features make proton therapy an
tion of extracranial treatments to intracranial attractive candidate for SRS [17–19]. For thera-
SRS treatments [16, 25]. peutic use, protons are accelerated using a
cyclotron or synchrotron and directed to the
gantry, maintaining their energy via bending
Modern Model magnets. Proton beam energies used for RT are
usually 150–250 MeV. Apertures or collimators
The modern CK system, the CK M6, can can be used to shape the beam, and a range of
deliver dose rates of up to 1000 MU/min, in as compensators can be used to finely control the
many as 1600 beam directions. CK M6 includes distal edge of the field [17–19].
a couch with 6DOF, such that patients can be set
up efficiently and reliably. The CK M6 also
includes fixed collimators, the IRIS variable col- Treatment Planning
limator system, and the InCise MLC system.
With these updated features, the CK M6 can Since the Bragg peak is very narrow, a single proton
deliver intracranial treatments in 15–60 minutes, beam is not practical therapeutically. For treatment
depending upon the complexity of the case; aver- of targets wider than the Bragg peak, two planning
age treatment times are approximately 30 min. methods have been established: passive scattering
With the CK M6, employing initial and near real- and scanning beam. Passive scattering involves
time imaging, treatment accuracy has been scattering and flattening the beam, combining
reported as submillimeter, comparable to that beams of different energies such that the target is
delivered with headframe systems [9, 23]. covered. Passive scattering technique is comparable
to conventional 3D conformal radiation planning,
making highly conformal complex plans challeng-
Protons ing to generate [1, 17–19]. However, studies have
demonstrated that with experienced users, passive
B scattering proton plans can be generated, which are
and Shaping dosimetrically superior to photon SRS plans.
Scanning beam planning involves using changes in
Proton therapy confers several advantages over beam energy to vary depth of dose deposition, as the
photon RT. Uniquely, protons stop their propa- beam scans across the width of the target. As the
gation through tissue at the point of maximum proton nears the gantry, an electromagnetic field is
energy deposition in the tissue. There is a slow used to modulate the direction and energy of the
increase in dose deposition with depth, followed beam such that the proton deposits energy in a spe-
by a steep increase toward the end of the range, cific plane or voxel. Scanning beam technique can
at the Bragg peak. There are minimal ioniza- be employed such that each field covers the target,
tions beyond the Bragg peak, sparing tissue single-field uniform doses, or multiple fields with
beyond the target. This confers dosimetric intensity modulation and inverse planning can be
advantages to proton therapy, especially in the used. Since this planning technique is similar to
context of SRS, where sparing of normal tissue photon IMRT, it is sometimes called intensity-
is paramount [1, 14, 17–19]. With reduced angle modulated proton therapy (IMPT). IMPT is useful
of scatter, proton therapy also attains a steep lat- for conferring increased conformity with irregularly
eral dose fall off, minimizing dose spill laterally shaped targets or those near critical structures. One
[1, 14, 19, 41]. Protons in the energy range used shortcoming of this planning approach is that, sec-
therapeutically have radiobiological effective- ondary to its great conformity, it is very sensitive to
ness (RBE) of 1.1. Therefore, the ability of promotion or setup error [17–19].
tons to induce cellular damage is comparable to Proton SRS planning carries unique dosimetric
that of photons, but perhaps 10% higher. These challenges. There is some uncertainty in the con-
version of CT Hounsfield units to proton stopping Cost

power, in the range of each beam, and in the depth
of the Bragg peak [17–19]. Additionally, most pro- Because protons have substantially higher mass
ton TPS apply a constant RBE of 1.1 for convert- than electrons, cyclotrons or synchrotrons capa-
ing proton dose to equivalent photon dose, but the ble of accelerating protons are of considerable
RBE likely changes along the proton range. The size and require high energy input. The cost of
RBE may also change based upon histology, dose, establishing a proton facility is at least an order
and fractionation, factors not accounted for in pro- of magnitude higher than for any photon SRS
ton TPS [17–19, 42, 43]. One approach for over- platform. Historically, building a proton therapy
coming these uncertainties is to add a margin to medical facility could cost as much as US$120–
the lesion when creating the treatment volume. 200 million. With more modern technology and
While this will increase the likelihood of target increasingly compact systems, it may be possible
coverage, it mitigates some of the benefits of pro- to establish a single room proton facility for
ton therapy in sparing surrounding tissues. Another approximately US$30 million. A multi-room
method is known as the “smearing technique,” in facility could cost over US$100 million.
which compensator dimensions are changed Additionally, dedicated maintenance, QA, and
within the range of uncertainty to ensure that the trained staff such as physicists and engineers
target lesion is covered adequately [18, 19]. confer high operational costs [14, 18, 19].
With regard to treatment efficiency, any
change in proton energy requires approximately
2 seconds. In IMPT plans, energy could be Efficacy and Appropriate Indications
changed as many as 50 or 60 times, as layers of
treatment are delivered. In a three- or four-beam Delivery of therapeutic radiation doses to CNS
IMPT proton plan, 5 or 6 minutes would be lesions is often challenging, secondary to dose
required simply for energy change, in addition to constraints of adjacent OARs. Given the dosimet-
actual radiation delivery time [19]. ric advantages of proton therapy, treatment of CNS
lesions may be an appropriate application for pro-
ton therapy. Evidence regarding feasibility, effi-
Immobilization and Image Guidance cacy, and toxicity of proton SRS is limited and is
largely derived from retrospective single institu-
Given the highly conformal nature of proton ther- tion series. Given limited availability and high cost
apy, the sharp dose fall off, and the sensitivity of of proton SRS, coupled with a lack of robust, pro-
protons to changes in tissue shape or density, spective, randomized efficacy and toxicity data,
immobilization and localization are even more guidelines regarding which patients are appropri-
critical than in conventional SRS. The immobili- ate for such treatment are lacking [17–19]. There
zation and patient positioning methods used for continues to be great controversy regarding appro-
other SRS modalities can be used for proton priate use of proton therapy [17–19, 44, 45]. In
SRS. Dedicated frame-based and frameless general, proton SRS may be considered for
immobilization systems have also been devel- patients who are young, have benign intracranial
oped specifically for intracranial proton RT, simi- lesions, have malignant lesions with long expected
lar to those for other platforms. As with other survival, have lesions adjacent to sensitive OARs,
SRS modalities, a variety of models and tech- require re-irradiation, and finally for those partici-
niques exist. Patients sometimes undergo addi- pating in clinical trials. The feasibility, efficacy,
tional steps to ensure appropriate target and safety of proton SRS have been demonstrated
localization, such as placement of fiducial mark- for several intracranial lesions, including vestibu-
ers within the outer table of the skull. On-board lar schwannoma, AVM, pituitary adenoma, and
imaging systems compatible with proton meningioma [17–19]. There is great need for more
machines, such as x-ray, fluoroscopy, or CBCT, compelling evidence from randomized controlled
are also well-developed [17–19]. trials regarding the benefits of proton SRS.
Table 21.1 Key features of different SRS platforms

GK Icon LINAC CK M6 Proton SRS
Beam source Gamma ray X-ray X-ray Proton
Beam energy 1.25 MeV Co-60 6 or 10 MV photons 6 MV photons 150–250 MeV
Beam 192 fixed Planar and non-coplanar Non-coplanar beams Planar and
arrangement converging beams beams non-coplanar
beams
Arcs
Beam shaping Circular Circular collimators Circular collimators Collimators
collimators
MLCs MLCs Apertures
Compensators
Treatment Shot-packing Static fields Shot-packing Passive scattering
planning
IMRT Scanning beam
Dynamic arcs (IMPT)
VMAT
Dosimetry High conformity High conformity High conformity High conformity
Low homogeneity High homogeneity High homogeneity High homogeneity
Decreased Greater low-dose spillage Greater low-dose spillage Least low-dose
low-dose spillage spillage
Immobilization Invasive Invasive Headframe Frameless Invasive
Headframe Headframe
Frameless Frameless Frameless
Image guidance CBCT Planar x-rays Planar x-rays Planar x-rays
Optical CBCT CBCT
monitoring
Optical monitoring
Machine Dedicated SRS, SBRT, and SRS and SBRT only, Limited
availability intracranial SRS conventional RT, intracranial intracranial and availability
unit and extracranial sites extracranial sites
Table 21.1 provides a summary of the key fea- target volumes, target locations, target num-
tures of the different SRS platforms. bers, beam arrangements, planning techniques,
outcomes, etc. [7]. This further complicates
comparisons of SRS technologies, and adds to
Comparisons the controversy.
While each SRS system has distinct advan-
Much debate has been generated regarding the tages and disadvantages, it is important to stress
benefits and shortcomings of the various SRS that the overall clinical efficacy of all modern
modalities and, unfortunately, existing litera- platforms is felt to be equivalent. Each system
ture comparing different SRS systems is should be able to generate and deliver acceptably
largely retrospective with low sample sizes [7, accurate and precise high-quality SRS plans [4,
12, 16]. These studies are often multi-institu- 14, 16, 21, 46]. As technological advances lead to
tional, with a different machine at each facility, improved plan quality, the dosimetric differences
as it would be impractical and costly for one between platforms are likely to decrease further.
center to have two SRS platforms. There may It warrants emphasis that user expertise is likely
be significant differences in patient popula- the most vital determinant of plan quality [8, 12,
tions, staff expertise, and treatment planning 16]. Large, multi-institutional, prospective trials
across institutions [3]. SRS planning is very would likely be required to definitively determine
complex, and different studies have evaluated whether any SRS platform confers dosimetric or
different machine models, dose prescriptions, clinical advantages.
Clinical Outcomes Dosimetry
Since GK SRS has been in use the longest, the Treatment of single brain metastases is relatively
published literature regarding clinical outcomes straightforward, and a high-quality plan can be
is most robust for GK. Despite significant dif- generated regardless of modality used. As physi-
ferences in the SRS platforms with regard to cians are treating patients with increasing num-
how radiation is prescribed and planned, and the bers of sometimes complex targets, this places
resultant dose distributions, studies to date have additional dosimetric demands on the SRS sys-
not found significant differences in clinical out- tem. Several studies have compared dosimetric
comes for patients with brain metastases features of the different SRS platforms in the
depending upon treatment modality [10, 47–49]. management of multiple brain metastases [7, 39,
The use of SRS for patients with a limited num- 46, 47, 52–54]. With regard to conformity in mul-
ber of intracranial metastases (1–4) has been tiple brain metastasis plans, the data generally
well established [46, 48–50]. As oncologic indicate superiority of GK compared to linac
treatments improve and patients with metastatic SRS. Further, linac may perform better than CK
disease live longer, SRS is increasingly being [39, 46, 54]. GK treatments are often prescribed
used in the management of multiple brain to the 50–80% isodose surface, leading to
metastases [41, 46, 48]. More recently, increased maximum dose, dose inhomogeneity,
Yamamoto et al. demonstrated that SRS man- and dose fall off [8, 15, 20, 24, 46, 47, 55].
agement of patients with 5–10 brain metastases Prescription to a higher isodose surface, such as
resulted in non-inferior overall survival relative with linac and CK, allows more homogeneous
to those with 1–4 metastases [51]. Currently, dose distributions [16, 20–22, 46, 55]. The stron-
there are no published clinical trial data estab- gest determinants of dose fall off are target vol-
lishing the role of SRS in patients with multiple ume and number, but treatment planning
(>4) brain metastases, though there are ongoing techniques can also influence dose fall off.
phase III trials exploring the role of SRS in this Comparative studies either demonstrate the sharp-
patient population [48]. That said, it may well est dose fall off with GK or report similar dose fall
be that volume of disease treated in the brain is off across the platforms [7, 39, 46, 52–55].
more important than the number of metastases. Since a goal of SRS is to reduce radiation
Numerous prospective and retrospective studies exposure to surrounding tissues, dose to normal
have been published regarding disease outcomes brain is an important factor to consider. In addi-
with SRS for multiple brain metastases [2, 7, 9, tion, V10 or V12, the volume of nor-
47, 48]. In these studies, a wide range of local mal brain receiving 10 or 12 Gy, are important
control (LC) rates is found, likely due to hetero- validated predictors of symptomatic radionecro-
geneity in patient diagnosis, performance status, sis for single-fraction SRS [4, 7, 8, 46, 47, 55].
and burden of systemic or intracranial disease In management of multiple brain metastases,
[48]. However, in general, excellent outcomes there is literature to support improved V12 with
are conferred by SRS, with LC rates at 1 year GK relative to linac and CK [46, 50, 52, 53, 55,
ranging from 69.5% to 97% [2, 7, 9, 47–49]. 56]. However, these findings have not been uni-
Given their reduced life expectancy, patients versally observed, and some studies have not
with brain metastases have not been managed demonstrated any difference in normal brain
extensively with proton SRS. In fact, to our V12 across the SRS platforms [54]. The greatest
knowledge, there is only one publication regard- concern with treating multiple intracranial
ing proton SRS for brain metastases. Atkins lesions is the integral dose to normal brain tis-
et al. retrospectively evaluated 370 patients with sues [53]. It is noteworthy that there are no pub-
815 brain metastases. They reported 1 year LC lished data establishing any neurocognitive
of 91.5% [41]. consequences of this low-dose spillage and any
benefits of normal tissue sparing are theoretical malignancies in patients treated for conditions
[41, 46, 53]. The published literature most com- for which long survivals are expected (Paddick I,
monly demonstrates sparing of normal brain tis- Personal communication, 2019).
sue from low-dose radiation, measured as V3–6, Treatment of multiple brain metastases via
with GK, compared to the other platforms [46– linac confers significantly reduced treatment
48, 53, 54, 56]. At least one study has further time compared to GK and CK [39, 46, 48, 50,
demonstrated improved V4 with linac compared 54]. Single isocenter VMAT plans, especially,
to CK [46]. Of interest, in the sole brain metasta- can increase treatment efficiency, potentially at
sis proton SRS study published to date, the the cost of increased low-dose spillage [46].
authors generated linac plans for 10 representa- Reduced treatment time increases patient com-
tive patients and reported improved V4 and V10 fort and reduces the likelihood of intrafraction
with proton SRS [41]. Other dosimetric studies motion and the need for real-time imaging
have confirmed decreased integral brain dose [48]. Treatment time with GK can be as much
with proton SRS compared to photons [41, as three to five times longer than with VMAT
57–59]. linac SRS in multiple lesion plans [39]. In
Extracranial radiation doses for various pho- patients with multiple brain metastases in
ton radiosurgery platforms have been measured whom minimizing treatment time is an impor-
and compared, and the lowest radiation doses tant consideration, linac SRS may be the
outside the brain are seen with the Gamma Knife modality of choice.
platform. Successively higher doses are delivered Table 21.2 provides a summary of the major
with linac radiosurgery and CyberKnife radiosur- advantages and disadvantages of the different
gery, which may confer higher risks of secondary SRS modalities.
Table 21.2 Advantages and disadvantages of different SRS platforms

Advantages Disadvantages
GK Icon Most robust, long-term efficacy, and Dedicated intracranial or high cervical SRS
safety data machine
Sharp dose fall off, minimal low-dose Co-60 decay causing variable dose rate over time
spillage and necessitating eventual source replacement
Simple system, minimal maintenance Longer treatment time
and QA
Least shielding required
Best suited for very small lesions near
critical OARs
LINAC Widespread machine availability More maintenance and QA
Easiest and cheapest to establish SRS Higher low-dose spillage
program
Machine versatility (conventional RT,
SBRT, SRS)
Extracranial treatments
Most efficient treatment delivery
Best suited for large lesions and
efficient treatments
CK M6 Flexibility in beam arrangement More maintenance and QA
Extracranial targets Most shielding required
SRS and SBRT only
Higher low-dose spillage
Longer treatment time
Protons Best dose fall off distal to target and Limited machine availability
laterally High cost
Least low-dose spill, best normal tissue Least efficacy and safety data
sparing Uncertainties in dosimetry
Conclusion 8. Pinkham MB, Whitfield GA, Brada M. New devel-

opments in intracranial stereotactic radiother-
apy for metastases. Clin Oncol (R Coll Radiol).
The critical features of several modern SRS plat- 2015;27(5):316–23.
forms, GK, linac-based, CK, and proton have 9. Hara W, Tran P, Li G, et al. Cyberknife for brain
been reviewed and critically compared with metastases of malignant melanoma and renal cell car-
cinoma. Neurosurgery. 2009;64(2):A26–32.
regard to clinical efficacy and plan quality. 10. Hoffelt CS. Gamma Knife vs. CyberKnife. Oncol

Overall, the choice of ideal SRS platform for Issues. 2006;21(5):18–20.
each individual case is a complex one, and myr- 11. Kurshnirsky M, Patil V, Schulder M. The history of
iad machine, patient, and target factors can be stereotactic radiosurgery. In: Chin LS, Regine WF, edi-
tors. Principles and practice of stereotactic radiosur-
considered, including patient performance status gery. 2nd ed. New York: Springer Science+Business
and prognosis, radiation source, dose rate, colli- Media; 2015. p. 3–10. Print.
mator type, beam arrangement, method of immo- 12. Farha G, Schlesinger D, Sarfehnia A, Sahgal A,

bilization, image guidance, TPS, treatment time, Ruschin M. Current state of the art in intracranial
stereotactic radiosurgery technology. In: Haffty B,
lesion size and shape, and distances to OARs. Sharad G, editors. Precision radiation oncology. New
Of course, the experience of the entire treat- Brunswick: Rutgers University Press; 2018. Print.
ment team, including radiation oncologist, neuro- 13. Niranjan A, Lunsford LD. Radiosurgery: where

surgeon, dosimetrists, physicists, and therapists, we were, are, and may be in the third millennium.
Neurosurgery. 2000;46(3):531–43.
must also be considered. It may be the case that 14. Lasak JM, Gorecki P. The history of stereotactic

patients with more complex and challenging radiosurgery and radiotherapy. Otolaryngol Clin N
treatments should be evaluated at regional cen- Am. 2009;42(4):593–9.
ters of excellence, given that treatments can be 15. Prasad D. Gamma knife® stereotactic radiosurgery
and hypo-fractionated stereotactic radiotherapy. In:
provided in one or only a few treatments, mini- Chang EL, Brown PD, Lo SS, Sahgal A, Suh JH, edi-
mizing inconvenience and potentially optimizing tors. Adult CNS radiation oncology. Cham: Springer
long-term outcomes. International Publishing AG; 2018. p. 665–85. Print.
16.
Stieber VW, Bourland JD, Tome WA, Mehta
MP. Gentlemen (and ladies), choose your weapons:
gamma knife vs. linear accelerator radiosurgery.
References Technol Cancer Res Treat. 2003;2(2):79–86.
17. Shirvani SM, Chang JY. Charged particles in stereo-
1. De Salles AAF, Gorgulho AA, Pereira JLB, tactic radiosurgery. In: Chin LS, Regine WF, editors.
McLaughlin N. Intracranial stereotactic radiosur- Principles and practice of stereotactic radiosurgery.
gery concepts and techniques. Neurosurg Clin N Am. 2nd ed. New York: Springer Science+Business
2013;24(4):491–8. Media; 2015. p. 135–46. Print.
2. Lippitz B, Lindquist C, Paddick I, Peterson D, 18. Yerramilli D, Bussiere MR, Loeffler JS, Shih

O’Neill K, Beaney R. Stereotactic radiosurgery in the HA. Proton beam therapy (for CNS Tumors). In:
treatment of brain metastases: the current evidence. Chang EL, Brown PD, Lo SS, Sahgal A, Suh JH, edi-
Cancer Treat Rev. 2014;40(1):48–59. tors. Adult CNS radiation oncology. Cham: Springer
3. Andrews DW, Bednarz G, Evans JJ, Downes B. A International Publishing AG; 2018. p. 709–22. Print.
review of 3 current radiosurgery systems. Surg 19. Mohan R, Grosshans D. Proton therapy – present and
Neurol. 2006;66(6):559–64. future. Adv Drug Deliv Rev. 2017;109:26–44.
4. Wowra B, Muacevic A, Tonn J-C. CyberKnife radio- 20. Shepard DM, Yu C, Murphy MJ, Bussiere M, Bova
surgery for brain metastases. In: Kim DG, Lunsford FJ. Treatment planning for stereotactic radiosurgery.
LD, editors. Current and future management of brain In: Chin LS, Regine WF, editors. Principles and prac-
metastases. Basel: Krager; 2012. p. 201–9. Print. tice of stereotactic radiosurgery. 2nd ed. New York:
5. Levivier M, Gevaert T, Negretti L. Gamma Knife, Springer Science+Business Media; 2015. p. 73–94.
cyberknife, tomotherapy: gadgets or useful tools? Print.
Curr Opin Neurol. 2011;24(6):616–25. 21.
Thomas EM, Popple RA, Bredel M, Fiveash
6. Kurup G. CyberKnife: a new paradigm in radiother- JB. Linac-based stereotactic radiosurgery and hypo-
apy. J Med Phys. 2010;35(2):63–4. fractionated stereotactic radiotherapy. In: Chang
7. Sahgal A, Ma L, Chang E, et al. Advances in technol- EL, Brown PD, Lo SS, Sahgal A, Suh JH, editors.
ogy for intracranial stereotactic radiosurgery. Technol Adult CNS radiation oncology. Cham: Springer
Cancer Res Treat. 2009;8(4):271–80. International Publishing AG; 2018. p. 639–63. Print.
22.
Rahman M, Murad GJA, Bova FJ, Friedman 35. Lindquist C, Paddick I. The leksell Gamma Knife
WA. LINAC: past, present, and future of radiosurgery. perfexion and comparisons with its predecessors.
In: Chin LS, Regine WF, editors. Principles and prac- Neurosurgery. 2007;61:130–40.
tice of stereotactic radiosurgery. 2nd ed. New York: 36. Petti P, Larson D, Kunwar S. Comparison of treatment
Springer Science+Business Media; 2015. p. 121–34. planning and delivery parameters for the Model C and
Print. Perfexion Gamma Knife units. In: Proceedings of the
23. Ding C, Saw CB, Timmerman RD. Cyberknife ste- 14th International meeting of the Leksell Gamma
reotactic radiosurgery and radiation therapy treatment Knife Society; 2008. p. 264.
planning system. Med Dosim. 2018;43(2):129–40. 37. Régis J, Tamura M, Guillot C, et al. Radiosurgery with
24. Niranjan A, Bowden G, Flickinger JC, Lunsford
the world’s first fully robotized Leksell Gamma Knife
LD. Gamma knife radiosurgery. In: Chin LS, PerfeXion in clinical use: a 200-patient prospective,
Regine WF, editors. Principles and practice of ste- randomized, controlled comparison with the Gamma
reotactic radiosurgery. 2nd ed. New York: Springer Knife 4C. Neurosurgery. 2009;64:346–55.
Science+Business Media; 2015. p. 111–9. Print. 38. Niranjan A, Novotny J Jr, Bhatnagar J, et al. Efficiency
25. Fasola CE, Wang L, Adler JR, et al. CyberKnife. In: and dose planning comparisons between the perfexion
Chin LS, Regine WF, editors. Principles and prac- and 4C leksell Gamma Knife units. Stereotact Funct
tice of stereotactic radiosurgery. 2nd ed. New York: Neurosurg. 2009;87:191–8.
Springer Science+Business Media; 2015. p. 147–61. 39. Liu H, Andrews DW, Evans JJ, et al. Plan quality
Print. and treatment efficiency for radiosurgery to multiple
26. Vesper J, Bolke B, Wille C, et al. Current concepts brain metastases: non-coplanar RapidArc vs. Gamma
in stereotactic radiosurgery - a neurosurgical and Knife. Front Oncol. 2016;6:26.
radiooncological point of view. Eur J Med Res. 40. Jang SY, Lalonde R, Ozhasoglu C, Burton S, Heron
2009;14(3):93–101. D, Huq MS. Dosimetric comparison between cone/
27. Tse VCK, Yashar SK, Adler JR. Techniques of stereo- Iris-based and InCise MLC-based CyberKnife plans
tactic localization. In: Chin LS, Regine WF, editors. for single and multiple brain metastases. J Appl Clin
Principles and practice of stereotactic radiosurgery. Med Phys. 2016;17(5):184–199. https://doi.org/
2nd ed. New York: Springer Science+Business 10.1120/jacmp.v17i5.6260
Media; 2015. p. 121–34. Print. 41. Atkins KM, Pashtan IM, Bussiere MR, et al. Proton
28. Ramakrishna N, Rosca F, Friesen S, Tezcanli E,
stereotactic radiosurgery for brain metastases: a
Zygmanszki P, Hacker F. A clinical comparison of single-institution analysis of 370 patients. Int J Radiat
patient setup and intra-fraction motion using frame- Oncol Biol Phys. 2018;101(4):820–9.
based radiosurgery versus a frameless image-guided 42. Carabe A, Moteabbed M, Depauw N, Schuemann J,
radiosurgery system for intracranial lesions. Radiother Paganetti H. Range uncertainty in proton therapy due
Oncol. 2009;95(1):109–15. to variable biological effectiveness. Phys Med Biol.
29. Alheit H, Dornfeld S, Dawel M. Patient position
2012;57(5):1159–72.
reproducibility in fractionated Stereotactically guided 43. Paganetti H. Nuclear interactions in proton therapy:
conformal radiotherapy using the BrainLab mask sys- dose and relative biological effect distributions origi-
tem. Strahlenther Onkol. 2001;177:264–8. nating from primary and secondary particles. Phys
30. Willner J, Flentje M, Bratengeier K. CT simulation Med Biol. 2002;47(5):747–64.
in stereotactic brain radiotherapy — analysis of iso- 44. Lodge M, Pijls-Johannesma M, Stirk L, Munro

center reproducibility with mask fixation. Radiother AJ, De Ruysscher D, Jefferson T. A systematic lit-
Oncol. 1997;45:83–8. erature review of the clinical and cost-effectiveness
31. Meeks SL, Bova FJ, Wagner TH. Image localization of hadron therapy in cancer. Radiother Oncol.
for frameless stereotactic radiotherapy. Int J Radiat 2007;83(2):110–22.
Oncol Biol Phys. 2000;46:1291–9. 45. Jones B. The case for particle therapy. Br J Radiol.
32. Lamba MS, Plotkin O, Weik J, et al. Evaluation of the 2006;79(937):24–31.
Novalis Exactrac system for frameless image-guided 46. Ma L, Nichol A, Hossain S, et al. Variable dose inter-
radiosurgery in the head. Int J Radiat Oncol Biol play effects across radiosurgical apparatus in treating
Phys. 2006;66:S244. multiple brain metastases. Int J Comput Assist Radiol
33.
Chen JC, Rahimian J, Girvigian MR, et al. Surg. 2014;9(6):1079–86.
Contemporary methods of radiosurgery treatment 47. Ma L, Sahgal A, Descovich M, et al. Equivalence in
with the Novalis linear accelerator system. Neurosurg dose fall-off for isocentric and nonisocentric intra-
Focus. 2007;23:E4. cranial treatment modalities and its impact on dose
34. Ma J, Chang Z, Wang Z, et al. ExacTrac X-ray 6 fractionation schemes. Int J Radiat Oncol Biol Phys.
degree-of-freedom image-guidance for intracranial 2010;76(3):943–8.
noninvasive stereotactic radiotherapy: comparison 48. Sahgal A, Ruschin M, Ma L, Verbakel W, Larson D,
with kilo-voltage cone-beam CT. Radiother Oncol. Brown PD. Stereotactic radiosurgery alone for mul-
2009;93:602–8.
tiple brain metastases? a review of clinical and techni- gery for multiple cranial metastases. Neurosurgery.
cal issues. Neuro-Oncology. 2017;19(2):112–5. 2014;75(4):409–18.
49. Barani IJ, Larson D, Berger MS. Future directions 55. Sio TT, Jang S, Lee SW, Curran B, Pyakuryal AP,
in treatment of brain metastases. Surg Neurol Int. Sternick ES. Comparing gamma knife and cyberknife
2013;4(4):S220–30. in patients with brain metastases. J Appl Clin Med
50. Zhang I, Antone J, Li J, et al. Hippocampal-sparing Phys. 2014;15(1):4095.
and target volume coverage in treating 3 to 10 brain 56. McDonald D, Schuler J, Takacs I, Peng J, Jenrette J,
metastases: a comparison of Gamma Knife, single- Vanek K. Comparison of radiation dose spillage from
isocenter VMAT, CyberKnife, and TomoTherapy the Gamma Knife Perfexion with that from volu-
stereotactic radiosurgery. Pract Radiat Oncol. metric modulated arc radiosurgery during treatment
2017;7:183–9. of multiple brain metastases in a single fraction. J
51. Yamamoto M, Kawabe T, Barfod BE. How many Neurosurg. 2014;121:51–9.
metastases can be treated with radiosurgery? Prog 57. Baumert BG, Lomax AJ, Miltchev V, et al. A compari-
Neurol Surg. 2012;25:261–72. son of dose distributions of proton and photon beams
52. Ma L, Petti P, Wang B, et al. Apparatus dependence in stereotactic conformal radiotherapy of brain lesions.
of normal brain tissue dose in stereotactic radiosur- Int J Radiat Oncol Biol Phys. 2001;49(5):1439–49.
gery for multiple brain metastases. J Neurosurg. 58. Baumert BG, Norton IA, Lomax AJ, et al. Dose con-
2011;114(6):1580–4. formation of intensity-modulated stereotactic photon
53. Hossain S, Keeling V, Hildebrand K, et al. Normal beams, proton beams, and intensity-modulated proton
brain sparing with increasing number of beams and beams for intracranial lesions. Int J Radiat Oncol Biol
isocenters in volumetric-modulated arc beam radio- Phys. 2004;60(4):1314–24.
surgery of multiple brain metastases. Technol Cancer 59. Bolsi A, Fogliata A, Cozzi L. Radiotherapy of small
Res Treat. 2016;15(6):L766–71. intracranial tumours with different advanced tech-
54. Thomas EM, Popple RA, Wu X, et al. Comparison niques using photon and proton beams: a treatment
of plan quality and delivery time between volumetric planning study. Radiother Oncol. 2003;68(1):1–14.
arc therapy (RapidArc) and Gamma Knife radiosur-
Stereotactic Radiosurgery:
Indications and Outcomes in
22
Central Nervous System and
Skull Base Metastases
Henry Jeison Ruiz-Garcia, Daniel M. Trifiletti,
and Jason P. Sheehan
Abbreviations NCCN National Comprehensive Cancer

Network
ASTRO American Society for Radiation NSCLC Non-small cell lung cancer
Oncology OS Overall survival
CDFS Cognitive Deterioration Free Survival PTV Planning target volume
CP Cognitive preservation QoL Quality of life
CSF Cerebrospinal fluid RCT Randomized controlled clinical trial
CT Computed tomography RN Radiation necrosis
CTV Clinical target volume RPA Recursive partitioning analysis
DS-GPA Diagnosis-specific Graded Prognostic RTOG Radiation Therapy Oncology Group
Assessment SRS Stereotactic radiosurgery
EORTC European Organization for Research UPMC University of Pittsburgh Medical
and Treatment of Cancer Center
FI Functional independence WBRT Whole-brain radiotherapy
GPA Graded Prognostic Assessment
GTV Gross tumor volume
HSRS Hypofractionated stereotactic radio-
surgery
Introduction
KPS Karnofsky Performance Status
LC Local control
Brain metastases account for the majority of
LINAC Linear accelerator
intracranial brain tumors, most frequently origi-
LMD Leptomeningeal disease
nating from cutaneous melanoma and carcino-
MRI Magnetic resonance imaging
mas of the lung, kidney, and breast. Brain
metastases appear in 20–40% of cancer patients,
even in the setting of controlled extracranial dis-
H. J. Ruiz-Garcia · D. M. Trifiletti ease, leading to 200,000 newly diagnosed cases
Department of Radiation Oncology, Mayo Clinic, per year in the United States [1].
Jacksonville, FL, USA
The prognosis of patients with brain metastases
J. P. Sheehan (*) has evolved over time, with survivals historically
Neurosurgery and Radiation Oncology, Department
ranging from 1 to 2 months for untreated patients to
of Neurosurgery, University of Virginia,
Charlottesville, VA, USA up to 27 months and beyond with multi-modal ther-
e-mail: jsheehan@virginia.edu apy. Whole-brain radiotherapy (WBRT) has been

https://doi.org/10.1007/978-3-030-42958-4_22
316 H. J. Ruiz-Garcia et al.
In this chapter, we will discuss the rationale

for patient selection in SRS, SRS in the post-
operative and preoperative setting, SRS for
previously irradiated patients, and SRS near
critical intracranial structures.
Stereotactic Radiosurgery
During SRS, a large dose of highly conformal

radiation is delivered in one to five fractions at
the targeted lesion. This is possible due to the
creation of a sharp dose fall-off at the margin of
the tumor that allows for the sparing of sur-
rounding normal tissue. Since the Swedish neu-
rosurgeon Lars Leksell described the
stereotactic utilization of therapeutic irradia-
tion in 1951 in the paper entitled “The
Stereotaxic Method and Radiosurgery of the
Fig. 22.1 Typical appearance of a T1-weighted post- Brain” [6], newer systems have been launched
contrast axial MRI image from a patient with a metastatic allowing improved sparing of normal brain tis-
brain lesion from primary lung cancer located on the left sue. Currently, linear accelerator (LINAC)-
postcentral gyrus with associated edema based SRS, Cyberknife®, and Gamma Knife®
technologies allow treating patients with “fra-
utilized for more than 60 years and has shown a meless” SRS with safety and reliability afforded
benefit in the treatment of neurologic symptoms by real- time patient tracking during
and intracranial tumor control. However, in more irradiation.
recent years, WBRT has been shown to increase the SRS has emerged as one of the most effective
risk of iatrogenic neurocognitive deficits and worsen treatments for the management of brain metasta-
quality of life (QoL) relative to stereotactic radio- ses. SRS has similar survival outcomes and is
surgery (SRS) [2]. Advancements in imaging tech- associated with less neurocognitive side effects,
nology have allowed for early (presymptomatic) as compared to WBRT [2, 7]. Furthermore, it is
identification of brain metastatic lesions in cancer often delivered in a single ambulatory session
patients (Fig. 22.1). As a result, SRS has become a and does not interrupt or delay systemic
dominant therapeutic option in the management of therapies.
selected patients with one to four metastases and
even in patients harboring 10 or more lesions [3].
Contemporary management of patients with Prognostic Scoring Systems
brain metastases typically involves a multimodal- and Patient Selection
ity regimen, including some combination of sur-
gery, WBRT, SRS, glucocorticoids, and/or Patients with brain metastases were generally
systemic therapy. Each patient should be evalu- classified as a single group until 1997, when a
ated in a personalized manner, and ideally, every paradigm shift occurred after the publication of
patient eligible for treatment should also be con- the Radiation Therapy Oncology Group
sidered for radiosurgery, weighing the risks and (RTOG) recursive partitioning analysis (RPA)
benefits [4, 5]. [8]. The RPA identified patient clinical factors
22 Stereotactic Radiosurgery: Indications and Outcomes in Central Nervous System and Skull Base… 317
that influence survival and prognosis, allowing if neurologic symptoms refractory to steroid
for improved clinical decision making. Later, management, as surgical decompression is the
specific biological tumor features were fastest manner to improve neurological function
included in the Graded Prognostic Assessment [20, 21].
(GPA) and diagnosis-specific GPA (DS-GPA)
scoring systems [9, 10], incorporating more
disease-specific parameters and even molecu- ostoperative Irradiation: SRS or
P
lar profiles into the prognostic systems. WBRT?
Consequently, clinicians have more tools than
ever to provide patients with optimized and Even as the studies from Patchell et al. [14] and
personalized therapy. Kocher et al. [15] positioned postoperative
WBRT as the standard of care in oligometastatic
patients, concerns were raised over the detrimen-
Stereotactic Radiosurgery tal effects of WBRT on quality of life (QoL)
for the Management of Patients domains such as fatigue and cognitive impair-
with One to Four Brain Metastases ment [2, 19, 22, 23].
As a result of the most recent advances in
Role of Surgical Resection SRS, radiosurgery has challenged the historical
use of WBRT. Postoperative SRS to the surgical
Phase III randomized clinical trials (RCT) have cavity following the resection of brain metasta-
established that surgery improves the survival of ses has established itself as a reasonable stan-
oligometastatic intracranial disease [11–13]. dard of care, owing to data from phase III RCT
Patchell et al. described the benefit of adding sur- [24]. The parallel development of hypofraction-
gery to WBRT in patients with solitary brain ated postoperative SRS and preoperative SRS
metastasis, by randomizing patients into “surgery could potentially both minimize symptomatic
+ WBRT” versus “biopsy only + WBRT”; sur- radiation-induced injury and improve local
gery improved local control, preservation of tumor control [25–27].
functional status, and most importantly, overall
survival (OS) [12]. To determine if surgery alone ostoperative Resection Cavity SRS
P
without WBRT was sufficient for patients pre- Apart from the neurotoxicity associated with
senting with solitary brain metastasis, Patchell WBRT, postoperative WBRT can delay systemic
et al. conducted a subsequent phase III RCT and therapy, especially if the patient needs to recover
found that surgery with WBRT was superior to from acute side effects.
surgery alone in terms of intracranial tumor con- Although numerous retrospective studies
trol (local and distal failure) and decreasing neu- reported local control rates from 70% to 90%
rologic death; however, there was no significant with SRS to the postoperative resection cavity
difference with regard to OS [14]. Very similar [28], Brennan et al. from Memorial Sloan
findings in oligometastatic patients presenting Kettering Cancer Center published the first pro-
with one to three lesions were reported more than spective trial and detailed local control, distant
a decade later by Kocher et al. as part of the failure, and overall survival for patients with lim-
EORTC 22952-26001 study (Table 22.1) [15]. ited number of metastases. Delivering a median
Thus, patients with oligometastatic disease margin dose of 18 Gy (15–22 Gy), approximately
should routinely receive neurosurgical evaluation 85% local control was reported during a median
for potential resection. This is especially impor- follow-up of 12 months [29].
tant in patients with large tumors (generally Two recent phase III RCTs further validated
>3 cm), particularly if it is causing edema and/or the role of adjuvant postoperative SRS after
Table 22.1 Randomized controlled clinical trials evaluating different treatment combinations for patients carrying limited brain metastases
318
Phase III randomized controlled trials evaluating the role of surgery, SRS, and WBRT through the time
Primary end Tumor control Functional Radiation
Study Randomization Criteria point Local control Distal control Survival outcomes necrosis
Evaluating the addition of surgery to WBRT
Patchell et al. (1990) WBRT + (n = 25) 1 lesion NR 52% 20% 40 w Sx > Bx NR
[12] Surgery
WBRT + (n = 23) No RT 20% 13% (p = 0.52) 15 w (p < 0.01) (p < 0.005)
Biopsy (p < 0.02)
Vetch et al. (1993) WBRT + (n = 32) 1 lesion Overall survival NR NR 10 m Sx + WBRT> NR
[11] Surgery
WBRT (n = 31) 6 m (p < 0.04)∗ WBRT (p < 0.06)
Mintz et al. (1996) WBRT + (n = 41) 1 lesion Overall survival NR NR 6.3 m NS NR
[13] Surgery
WBRT (n = 43) 5.6 m (p = 0.24)
Evaluating the addition of WBRT to surgery
Patchell et al. (1998) Surgery + (n = 49) 1 lesion Local control 90% 86% NS NS NR
[14] WBRT
Surgery (n = 46) 54% 37% (p < 0.01
(p < 0.01)
Kocher et al. (2011)a Surgery + (n = 81) 1–3 OS with FI 59% 42% 10.7 NS NR
[15] WBRT lesions
Surgery (n = 79) 27% 23% 10.9 (p = 0.89)
(p < 0.001) (p < 0.008)
Evaluating the addition of SRS to WBRT
Kondziolka et al. WBRT + SRS (n = 13) 2–4 Local control 92% 34 m∗∗ 11 m NR 0%
(1999) [16] lesions
WBRT (n = 14) <2.5 cm 0% 5 m 7.5 m (p < 0.22) 0%
(p < 0.001) (p < 0.002)
Andrews et al. (2004) WBRT + SRS (n = 164) 1–3 Overall survival 82% NR 6.5 m WBRT + SRS >
[17] lesions
WBRT (n = 167) <4 cm 71% NR 4.9 m WBRT (p = 0.03)
(p = 0.01) (p = 0.04)∗∗∗
H. J. Ruiz-Garcia et al.
Evaluating the addition of WBRT to SRS
Aoyama et al. (2006) SRS + WBRT (n = 65) 1–4 Overall survival 96.9% 67.6% 7.5 m 33.9% 4.6%
[18] lesions
SRS (n = 67) each 91.0% 49.2% 8 m (p = 0.42) 26.9% (p = 0.53) 1.5%
<3 cm (p = 0.02) (p < 0.003)
Kocher et al. (2011)b SRS + WBRT (n = 99) 1–3 OS with FI 81% 67% 10.7 NS 13%
[15] lesions
SRS (n = 100) 69% 52% (p < 0.02) 10.9 (p = 0.89) 8%
(p = 0.04)
Chang et al. (2009) SRS + WBRT (n = 28) 1–3 Cognitive 100% 73% 63% 52%∗∗∗∗
[19] lesions outcomes
SRS (n = 30) 67% 45% (p = 0.02) 21% (p < 0.003) 24%
(p = 0.012)
Brown et al. (2016) SRS + WBRT (n = 102) 1–3 Cognitive 90% 92.3% 7.4 m SRS > WBRT+SRS 2.9%
[2] lesions outcomes
SRS (n = 111) <3 cm 73% 69.9% 10.4 m (p = 0.92) For CP and QoL 4.5%
(p < 0.003) (p < 0.001) (p = 0.72)
Abbreviations: WBRT whole-brain radiation therapy, SRS stereotactic radiosurgery, Sx surgery, NR not reported, NS not significant, LC local control, OS overall survival,
FI functional independence, CP cognitive preservation
∗No differences for patients with active extracranial disease. ∗∗Time to any brain failure. ∗∗∗Survival time for patients with single metastasis (months). ∗∗∗∗HVLT-R total recall
mean probability to decline
a & b are part of the same RCT (EORTC 22952-26001)
22 Stereotactic Radiosurgery: Indications and Outcomes in Central Nervous System and Skull Base…
319
surgical resection of a limited number of metas- contributed to a statistically significant reduction

tases. Mahajan et al. [30] randomized 132 in local failure at 12 months (16% vs 3%), with
patients with one to three lesions to receive sur- no significant increase in toxicity. The use of
gery and SRS or surgery alone, with respective margin expansions is heavily dependent on radio-
local tumor control rates of 72% and 42%, sup- surgical platform and technique; extrapolation
porting the use of SRS in the postoperative set- between centers should be done with caution.
ting. Brown et al. reported results of NCCTG Soliman et al. [37] published the Contouring
(N107C/CEC3) [31], a cooperative group phase Consensus Guidelines for Postoperative
III RCT comparing surgery + SRS versus sur- Completely Resected Cavity Stereotactic
gery + WBRT in 194 patients with resected Radiosurgery for Brain Metastases in 2017,
single metastatic brain lesions. Cognitive dete- where SRS experts contoured 10 postoperative
rioration at 6 months was less frequent with resection cavities of brain metastasis patients
SRS than with WBRT. As no differences were with lesions located in either supratentorial or
found in overall survival during a median infratentorial regions. Overall, the absolute kappa
11.1 months follow-up, SRS was recommended agreement for clinical target volume (CTV) was
over WBRT as a less toxic alternative in these high in each of the cases (mean sensitivity 0.75,
patients (Table 22.2). mean specificity 0.98). The findings led to the
Larger tumor size/volume has been reported following recommendations on CTV contouring:
as an unfavorable risk factor for local control (1) CTV should include the entire contrast-
[32–34]. Brennan et al. had reported that tumor enhancing surgical cavity using the T1-weighted
diameter >3 cm as well as superficial dural/pial gadolinium-enhanced axial MRI scan, excluding
invasion were associated with increased local any vasogenic edema determined by MRI; (2)
failure [29]. On the other hand, lesions <3 cm, CTV should include the entire surgical tract seen
deep lesions, and non-small cell lung cancer on postoperative CT or MRI; (3) if the tumor was
(NSCLC) histology were associated with in contact with the dura preoperatively, CTV
improved local control in the same study. In gen- should include a 5- to 10-mm margin along the
eral, tumor recurrence at the surgical site was bone flap beyond the initial region of preopera-
associated with increased volume of the surgical tive tumor contact; (4) if the tumor was not in
cavity or the lack of a 1–3 mm margin. contact with the dura, CTV should include a mar-
The Radiation Therapy Oncology Group gin of 1–5 mm along the bone flap; and (5) if the
95-08 trial established the initial SRS margin tumor was in contact with a venous sinus preop-
dose recommendations in recurrent brain metas- eratively, CTV should include a margin of
tases and gliomas based on tumor diameter. 1–5 mm along the sinus. Clinical judgment is still
However, it is now clear that dose prescription for required on a case-by-case basis until these rec-
SRS to a resection bed will depend on the postop- ommendations are fully validated by clinical out-
erative resection cavity volume on postoperative comes and patterns of recurrence [37].
imaging, as well as tumor location, previous irra- Another important factor for postoperative
diation, and prescription isodose. SRS is the resection cavity volume dynamic [24].
The role of the margin expansion in target Iorio-Morin et al. [38] recommended 3 weeks
delineation was initially studied by the Stanford after resection as ideal timing to deliver SRS,
group. Soltys et al. [35] found improved local after they found longer surgery-to-SRS delay to
control in treatment plans with a lower confor- be associated with local recurrence on a multi-
mality index—a measure of the compactness of variate analysis. This agrees with Patel et al. [39]
the high-dose radiation given during SRS relative who recommend against delaying SRS after sur-
to the target volume. Choi et al. [36] later pro- gery. After prospectively reviewing 79 cases, the
spectively studied the role of target margin on authors found that there was a 28% increase in
tumor control of resection cavities treated by the postoperative cavity volume with a median
SRS, finding that the addition of 2 mm margins time of surgery-to-SRS of 20 days and that, the
Table 22.2 Randomized clinical trials evaluating postoperative stereotactic radiosurgery to resection cavities in patients with brain metastases
Postoperative stereotactic radiosurgery to resection cavity in brain metastasis
Primary end Tumor control Functional Radiation
Study Randomization Criteria point Local control Distal control Survival outcomes necrosis
Brennan et al. Surgery + SRS (n = 49) 1–2 lesions LC at 85% 44% 14.7 m NR 17.50%
(2014)
Phase II [29] > 18 yo 12 m 50% (p = 0.08)a
(MSKCC)
PTV = cavity + 2 mm
Mahajan et al. Surgery + SRS (n = 64) 1–3 lesions LC 72% 42% 17 m NR 0%
(2017)
Phase III [30] Surgery + Obs (n = 68) >3 yo 43% 33% (p = 0.35) 18 m
(p = 0.015) (p = 0.24)
PTV = cavity + 1 mm
Brown et al. Surgery + SRS (n = 98) 1 lesion, >18 yo OS and 61% 64.70% 12.2 CDFS: 3.7 m 1%
(2017)
Phase III [31] Surgery + (n = 96) <5 cm CDFS 81% 89.2% 11.6 3 m (p < 0.001) 0%
WBRT (p < 0.0007) (p < 0.0005) (p = 0.7)
PTV = cavity + 2 mm
Abbreviations: WBRT whole-brain radiation therapy, SRS stereotactic radiosurgery, NR not reported, LC local control, OS overall survival, CDFS cognitive deterioration free
survival, PTV planning target volume, Obs observation
a
Based on competing risk analysis including patients who completed postsurgical SRS and those who did not receive SRS (n = 40 and n = 10, respectively)
22 Stereotactic Radiosurgery: Indications and Outcomes in Central Nervous System and Skull Base…
321
smaller the cavity, the higher the probability of SRS over single-fraction SRS with regard to
postoperative cavity volume enlargement. The efficacy and toxicity.
ideal interval between surgical resection and
delivery of SRS was conjectured to be 2–3 weeks, Preoperative SRS
as it allows for recovery after surgery and limits A novel potential strategy to approach some of
risk of local recurrence. Ultimately, though, it is the drawbacks associated with postoperative SRS
clear that resection cavity sizes fluctuate after is the use of preoperative SRS. Advantages
surgery. As such, it is imperative that planning include lack of need for margin addition to the
MRIs be performed as close as possible to the gross tumor volume (GTV; GTV = PTV or plan-
actual time of radiation delivery. Platforms that ning target volume), no delay in treatment deliv-
involve MRI acquisition on the day of radiation ery, and the decreased risk of potential seeding of
treatment may thus have an inherent advantage in viable malignant cells into the CSF during sur-
accuracy. gery. Given that preoperative SRS treats a non-
violated brain metastatic lesion, the borders will
Hypofractionation and Postoperative be well defined for target delineation; this could
Resection Cavity SRS explain the decreased risk of radiation necrosis
Single fraction SRS may have increased risk of reported with this technique [25, 43].
toxicity in patients who have been previously Asher et al. [44] published the first study
irradiated, have lesions larger than 3 cm in diam- regarding local efficacy and safety of preoperative
eter, produce more than 1 cm of midline shift, SRS for patients with one to three metastases
and/or abut critical organs-at-risk [40, 41]. where at least one of them was scheduled for sur-
According to the RTOG 90-05, recurrent previ- gical resection. A dose reduction strategy was
ously irradiated lesions of 3.1–4.0 cm receiving used under the principle that intact brain metasta-
15 Gy, as the maximum tolerated dose, present a ses would maintain their blood supply and oxy-
risk of unacceptable neurological toxicity up to genation and consequently a lower dose would be
16 times that of lesions <2 cm [42]. necessary to reach the same biological effect;
Hypofractionated SRS is being increasingly 80% of the standard dose according to RTOG
used as it allows for dose escalation while limit- 95-08 was delivered 48 hours before surgery and
ing the risk profile, taking advantage of the no margins were applied for delineation
improved repair of normal brain tissue. Eaton (GTV = PTV). Overall survival at 6 and 12 months
et al. reported on local control and the incidence was 77.8% and 60% and local control at 6, 12, and
and severity of radiation necrosis (RN) among 24 months was 97.8%, 85.6%, and 71.8%, respec-
patients treated with single fraction SRS or hypo- tively. There were no reports of leptomeningeal
fractionated SRS (HSRS) for postoperative disease (LMD) during the 12-month follow-up.
resection cavities ≥3 cm in diameter. Seventy-six A subsequent study from the same group com-
patients with a median follow-up of 11 months pared postoperative WBRT with preoperative
were included. No significant differences in local SRS. There were no differences in OS or LC, and
control were found, but single-fraction SRS was interestingly no advantage with regard to LMD
associated with higher risk of radiation necrosis with WBRT [45].
on multivariate analysis (HR: 3.81; 95% CI 1.04– Two potential drawbacks could arise with the
13.93, p = 0.043). use of preoperative SRS. The first is the possibil-
Although several other retrospective studies ity of incomplete resection of the metastatic
support the utilization of hypofractionated lesion after a lower and less ideal preoperative
SRS in the postsurgical setting for brain metas- radiosurgical dose. The second and major draw-
tases [25, 26], there is still a lack of RCT data back is the lack of pathological confirmation of
supporting the superiority of hypofractionated the lesion. Although there are no robust data, the
reported rate of false positive lesions ranges from SRS support this premise [20, 21]. Regardless,
2% to 11%. it is clear that surgical resection, unlike SRS,
can provide immediate intracranial decompres-
sion and pathologic confirmation.
Radiosurgery as Definitive Treatment The National Comprehensive Cancer Network
(NCCN) recommends that surgery is followed
SRS Versus Surgery by either WBRT or SRS for patients with one
Currently, there are no clinical trials available to three lesions and limited systemic disease.
comparing SRS and surgery. In 1996, Bindal The choice between surgery and SRS depends
et al. [46] from MD Anderson reported on this on several factors such as size and location; a
comparison. They prospectively followed 31 small, deep lesion should be treated with SRS
patients with lesions <3 cm who underwent SRS at an experienced institution [49]. Surgery also
between 1991 and 1994 and matched them to 62 can lead to almost immediate symptom relief as
patients from a pool of retrospective cases that well as rapid discontinuation of glucocorticoid
had only received surgery. Median SRS dose was therapy.
20 Gy (12–22 Gy) and WBRT was given equally
in both groups. They found improved overall sur- RS with or Without WBRT
S
vival and local control with surgery. The authors Two RCT comparing WBRT with WBRT + SRS
suggested that SRS should be limited to surgi- reported suboptimal local control with WBRT
cally inaccessible lesions or patients with signifi- alone in patients with limited metastases [16, 17].
cant medical comorbidities. Four recent randomized studies evaluated SRS
Muacevic et al. [47] reported the results from versus WBRT + SRS in patients with up to three
a phase III RCT that was stopped prematurely to four metastases [2, 15, 18, 19] and reported the
given poor accrual. In the final analysis based on following conclusions: (1) adjuvant WBRT
64 patients with a single lesion <3 cm and ran- improves local and distal control; (2) adjuvant
domized into surgery + WBRT or SRS alone, the WBRT increases the risk of neurotoxicity, with
authors found similar OS (median, 9.5 vs. consequent neurocognitive and quality-of-life
10.8 months, p = 0.8), LC (82% vs. 96%, decline; and (3) adjuvant WBRT does not
p = 0.06), and neurological death rates (29% vs. improve survival over SRS alone (Table 22.1).
11%, p = 0.3). Although higher rates of distal This last conclusion has been challenged by ret-
recurrence were observed with SRS, this differ- rospective studies. Wang et al. [50] who analyzed
ence was not seen after salvage therapy. 15 years of experience from Columbia University
A phase III RCT comparing surgery and SRS Medical Center and a new secondary analysis of
(both with adjuvant WBRT) was reported by the JROSG 99-1RCT published by Aoyama et al.
Ross et al. [48]. Although there was a trend favor- [51] have suggested that WBRT + SRS may
ing SRS regarding OS (6.2 vs 2.8 months) and improve OS in select patients with favorable
median failure free survival (3.1 vs. 1.7 months), prognoses. A secondary analysis of EORTC
the number of patients (n = 21) was too small to 22952-26001 did not find any survival advantage
obtain any robust conclusions. for WBRT relative to SRS in patients with lim-
In general, either treatment should not ited systemic disease or favorable GPA scores
exclude the other. We have already discussed [52]. The National Comprehensive Cancer
the benefit of postoperative resection cavity Network (NCCN) recommends SRS plus WBRT
SRS, and there is a growing body of knowledge (Level 1 evidence) or SRS alone (Level 2B evi-
on ways to balance the risks and benefits of dence) for patients with a single brain metastasis,
these two approaches. Recent retrospective limited systemic disease, and good performance
series showing the benefit of adding surgery to status.
The American Society for Radiation Oncology SRS for patients with five or more lesions; how-
(ASTRO) [53] released a list of definitive recom- ever, further prospective data are needed to vali-
mendations as part of the Choosing Wisely cam- date other aspects of this treatment; recursive
paign and recommended against routinely adding partitioning analyses could be useful to identify
adjuvant WBRT to SRS for patients with limited the groups of patients that can benefit the most
brain metastases. The impact of WBRT on QoL from SRS. Several such studies are currently
and cognition should be taken into consideration, underway.
especially as salvage SRS or WBRT is always an
option for dealing with future recurrences with-
out worsening toxicity. Stereotactic Radiosurgery
in the Reirradiation Setting
Stereotactic Radiosurgery Radiation necrosis is a known potential compli-

for the Management of Patients cation of SRS and can be difficult to distinguish
with More Than Four Brain clinically and/or radiographically from tumor
Metastases recurrence. For intact brain metastases treated
with SRS, rates of radiographic radiation necro-
Patients with a higher number of brain metastasis (RN) could reach up to 24%, while in the
ses should be managed with WBRT or SRS as postoperative resection cavity setting RN rates
primary treatment, unless at least one of the range from 1.5% to 18% [24]. If there is a high
indications for surgery is present. While select index of suspicion for recurrence, resection or
patients with poor prognosis are offered WBRT stereotactic biopsy should be considered.
[15], SRS is indicated for patients with good If recurrence is pathologically confirmed, SRS
performance status and low overall tumor vol- could be delivered as a salvage treatment in this
ume [49]. context after previous WBRT. In the setting of
The group from University of Pittsburgh resection for tumor recurrence after previous
Medical Center (UPMC) [54] published out- SRS, adjuvant therapy should be individualized,
comes of SRS for patients with four or more although observation after gross total resection is
metastatic brain lesions. They found that cumu- a reasonable approach. Repeat SRS can be
lative tumor treatment volume was the most offered, and other options include resection with
important prognostic factor for survival, sup- intraoperative brachytherapy, detailed elsewhere
porting the use of the total volume of brain in this book, and laser interstitial thermal therapy
metastases rather than the number of lesions for (LITT) to cauterize the tissue.
treatment decision making. In their analysis, For patients who have previously been
patients with a total treatment volume <7 cc and treated with SRS, the NCCN guidelines [49]
<7 brain metastases benefited the most from sin- recommend repeat SRS if there was a durable
gle SRS [55]. response longer than 6 months as long as imag-
Yamamoto et al. [3] published the results of a ing supports active tumoral lesion and not
non-inferiority trial in 2014 finding no differ- necrosis (2B recommendation). That said,
ences in survival or treatment-related adverse imaging in the recurrent, post-treatment setting
events between the group of patients treated is often a mixed picture, and thus clinician best
with SRS for 5–10 brain metastases and the judgment must prevail. Because of the possi-
group with 2–4 lesions (largest tumor <10 mL in bility of pseudoprogression in patients with
volume and <3 cm in longest diameter, total metastatic disease, it is often prudent to moni-
cumulative volume ≤15 mL, KPS ≥70, SRS tor suspicious post-radiosurgical abnormalities
only treatment). This study supports the use of unless they become symptomatic.
RS for Brain Metastases Involving

S In a study of radiosurgery in 161 patients har-
Eloquent or Critical Structures boring 189 metastases in the brainstem, 52% of
had received whole brain radiotherapy (WBRT)
Radiating eloquent regions of the brain requires a prior to SRS. These results suggest that SRS can
careful analysis of risk and benefit in order to pre- be safely administered after WBRT, even in elo-
vent damage to adjacent tissues that serve impor- quent or critical brain locations [58]. However,
tant neurologic functions (Fig. 22.2). after this report, we conducted an international
Sensorimotor, language, visual cortex, hypothal- cooperative study to define response and toxicity
amus, thalamus, brainstem, cerebellar nuclei, in brainstem metastases and found an increased
optic pathways, and regions immediately adja- risk of injury when SRS is administered shortly
cent to these structures are generally considered after WBRT [40]. This could be due to sublethal
organs at risk of symptomatic radiation injury. damage from WBRT decreasing with time,
Two retrospective series evaluating SRS for allowing for recovery and lower radiation-
metastases located in eloquent areas (primary induced injury risk with subsequent SRS. It is
motor, somatosensory, speech, and visual cortex; evident that previous intracranial therapies, spe-
basal ganglia; thalamus; and brainstem) indicated cifically radiation, should be considered during
that it is safe and effective [56, 57]. Hsu et al. treatment decision making.
reported no differences in the overall survival when Taken together, it is possible for an experienced
compared to the cohort harboring non-
eloquent team to perform stereotactic radiosurgery to brain
lesions receiving a higher median prescription dose. metastases located within or near critical struc-
New neurological deficits were transient and rates tures. In the presence of an intact tumor capsule,
of radiation necrosis were as expected for SRS. the target would consist solely of tumor cells (i.e.,
non-neural tissue), and therefore accurate delinea-
tion and accurate conformal delivery should rarely
result in clinical toxicity. Furthermore, given the
dismal prognosis of patients carrying metastatic
brain lesions, it is possible that the survival is not
long enough for late complications such as radia-
tion necrosis to present.
Hypofractionation in SRS is advantageous for
larger lesions, allowing maintenance of therapeu-
tic dose while decreasing the risk of radionecro-
sis. Hypofractionated SRS delivery for lesions
located in critical structures is a topic of ongoing
prospective clinical research.
Conclusion
Stereotactic radiosurgery has proven safety and

efficacy for the management of brain metastatic
Fig. 22.2 T1-weighted post-contrast axial MRI image
from a patient presenting a metastatic brain lesion from a lesions in the definitive and adjuvant setting.
soft tissue sarcoma primary located on the right postcen- The total volume of brain metastases, rather
tral gyrus. Given the tumor volume, SRS was delivered to than the number of lesions, seems to be more
a dose of 20 Gy. There have been no complications, and important to clinical decision making. With the
local control was maintained in the last follow-up at
12 months appropriate clinical and biological factors taken
into consideration, SRS is a powerful therapeu- 12. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ,

tic tool that can improve the quality of life of Maruyama Y, Kryscio RJ, et al. A randomized trial of
surgery in the treatment of single metastases to the
our patients. Prospective data are needed to fur- brain. N Engl J Med. 1990;322(8):494–500.
ther validate the superiority of novel SRS 13. Mintz AH, Kestle J, Rathbone MP, Gaspar L,

approaches. Hugenholtz H, Fisher B, et al. A randomized trial
to assess the efficacy of surgery in addition to radio-
therapy in patients with a single cerebral metastasis.
Cancer. 1996;78(7):1470–6.
References 14. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ,
Mohiuddin M, Kryscio RJ, et al. Postoperative
1. Dagogo-Jack I, Carter SL, Brastianos PK. Brain radiotherapy in the treatment of single metas-
metastasis: clinical implications of branched evolu- tases to the brain: a randomized trial. JAMA.
tion. Trends Cancer. 2016;2(7):332–7. 1998;280(17):1485–9.
2. Brown PD, Jaeckle K, Ballman KV, Farace E, Cerhan 15. Kocher M, Soffietti R, Abacioglu U, Villa S, Fauchon
JH, Anderson SK, et al. Effect of radiosurgery alone F, Baumert BG, et al. Adjuvant whole-brain radiother-
vs radiosurgery with whole brain radiation therapy apy versus observation after radiosurgery or surgical
on cognitive function in patients with 1 to 3 brain resection of one to three cerebral metastases: results
metastases: a randomized clinical trial. JAMA. of the EORTC 22952-26001 study. J Clin Oncol Off J
2016;316(4):401–9. Am Soc Clin Oncol. 2011;29(2):134–41.
3. Yamamoto M, Serizawa T, Shuto T, Akabane A, 16. Kondziolka D, Patel A, Lunsford LD, Kassam A,
Higuchi Y, Kawagishi J, et al. Stereotactic radio- Flickinger JC. Stereotactic radiosurgery plus whole
surgery for patients with multiple brain metastases brain radiotherapy versus radiotherapy alone for
(JLGK0901): a multi-institutional prospective obser- patients with multiple brain metastases. Int J Radiat
vational study. Lancet Oncol. 2014;15(4):387–95. Oncol Biol Phys. 1999;45(2):427–34.
4. Flickinger JC, Lunsford LD, Somaza S, Kondziolka 17. Andrews DW, Scott CB, Sperduto PW, Flanders AE,
D. Radiosurgery: its role in brain metastasis manage- Gaspar LE, Schell MC, et al. Whole brain radiation
ment. Neurosurg Clin N Am. 1996;7(3):497–504. therapy with or without stereotactic radiosurgery
5. Gerosa M, Nicolato A, Foroni R, Zanotti B, Tomazzoli boost for patients with one to three brain metastases:
L, Miscusi M, et al. Gamma knife radiosurgery for phase III results of the RTOG 9508 randomised trial.
brain metastases: a primary therapeutic option. J Lancet. 2004;363(9422):1665–72.
Neurosurg. 2002;97(5):515–24. 18. Aoyama H, Shirato H, Tago M, Nakagawa K, Toyoda
6. Leksell L. The stereotaxic method and radiosurgery of T, Hatano K, et al. Stereotactic radiosurgery plus
the brain. Acta Chir Scand. 1951;102(4):316–9. whole-brain radiation therapy vs stereotactic radiosur-
7. Skeie BS, Eide GE, Flatebo M, Heggdal JI, Larsen gery alone for treatment of brain metastases: a random-
E, Bragstad S, et al. Quality of life is maintained ized controlled trial. JAMA. 2006;295(21):2483–91.
using Gamma Knife radiosurgery: a prospective study 19. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF,
of a brain metastases patient cohort. J Neurosurg. Kornguth DG, et al. Neurocognition in patients with
2017;126(3):708–25. brain metastases treated with radiosurgery or radio-
8. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, surgery plus whole-brain irradiation: a randomised
Wasserman T, et al. Recursive partitioning analysis controlled trial. Lancet Oncol. 2009;10(11):1037–44.
(RPA) of prognostic factors in three radiation therapy 20. Prabhu RS, Press RH, Patel KR, Boselli DM,
oncology group (RTOG) brain metastases trials. Int J Symanowski JT, Lankford SP, et al. Single-fraction
Radiat Oncol Biol Phys. 1997;37(4):745–51. stereotactic radiosurgery (SRS) alone versus surgical
9. Sperduto PW, Kased N, Roberge D, Xu Z, Shanley resection and SRS for large brain metastases: a multi-
R, Luo X, et al. Summary report on the graded prog- institutional analysis. Int J Radiat Oncol Biol Phys.
nostic assessment: an accurate and facile diagnosis- 2017;99(2):459–67.
specific tool to estimate survival for patients with 21. Quigley MR, Bello N, Jho D, Fuhrer R, Karlovits S,
brain metastases. J Clin Oncol Off J Am Soc Clin Buchinsky FJ. Estimating the additive benefit of sur-
Oncol. 2012;30(4):419–25. gical excision to stereotactic radiosurgery in the man-
10. Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran agement of metastatic brain disease. Neurosurgery.
W. A new prognostic index and comparison to three 2015;76(6):707–12; discussion 12-3
other indices for patients with brain metastases: an 22. Pulenzas N, Khan L, Tsao M, Zhang L, Lechner B,
analysis of 1,960 patients in the RTOG database. Int J Thavarajah N, et al. Fatigue scores in patients with
Radiat Oncol Biol Phys. 2008;70(2):510–4. brain metastases receiving whole brain radiotherapy.
11.
Vecht CJ, Haaxma-Reiche H, Noordijk EM, Support Care Cancer. 2014;22(7):1757–63.
Padberg GW, Voormolen JH, Hoekstra FH, et al. 23. Aoyama H, Tago M, Kato N, Toyoda T, Kenjyo M,
Treatment of single brain metastasis: radiotherapy Hirota S, et al. Neurocognitive function of patients
alone or combined with neurosurgery? Ann Neurol. with brain metastasis who received either whole
1993;33(6):583–90. brain radiotherapy plus stereotactic radiosurgery or
radiosurgery alone. Int J Radiat Oncol Biol Phys. prospective evaluation of target margin on tumor con-
2007;68(5):1388–95. trol. Int J Radiat Oncol Biol Phys. 2012;84(2):336–42.
24. Marchan EM, Peterson J, Sio TT, Chaichana KL,
37. Soliman H, Ruschin M, Angelov L, Brown PD,

Harrell AC, Ruiz-Garcia H, et al. Postoperative cavity Chiang VLS, Kirkpatrick JP, et al. Consensus
stereotactic radiosurgery for brain metastases. Front contouring guidelines for postoperative com-
Oncol. 2018;8:342. pletely resected cavity stereotactic radiosurgery
25. Keller A, Dore M, Cebula H, Thillays F, Proust F, for brain metastases. Int J Radiat Oncol Biol Phys.
Darie I, et al. Hypofractionated stereotactic radiation 2018;100(2):436–42.
therapy to the resection bed for intracranial metastases. 38. Iorio-Morin C, Masson-Cote L, Ezahr Y, Blanchard
Int J Radiat Oncol Biol Phys. 2017;99(5):1179–89. J, Ebacher A, Mathieu D. Early Gamma Knife stereo-
26. Steinmann D, Maertens B, Janssen S, Werner M,
tactic radiosurgery to the tumor bed of resected brain
Fruhauf J, Nakamura M, et al. Hypofractionated ste- metastasis for improved local control. J Neurosurg.
reotactic radiotherapy (hfSRT) after tumour resection 2014;121:69–74.
of a single brain metastasis: report of a single-Centre 39. Patel RA, Lock D, Helenowski IB, Chandler JP,

individualized treatment approach. J Cancer Res Clin Sachdev S, Tate MC, et al. Postsurgical cavity evolu-
Oncol. 2012;138(9):1523–9. tion after brain metastasis resection: how soon should
27. Patel KR, Burri SH, Asher AL, Crocker IR, Fraser postoperative radiosurgery follow? World Neurosurg.
RW, Zhang C, et al. Comparing preoperative with 2018;110:e310–e4.
postoperative stereotactic radiosurgery for Resectable 40. Trifiletti DM, Lee CC, Kano H, Cohen J, Janopaul-
brain metastases: a multi-institutional analysis. Naylor J, Alonso-Basanta M, et al. Stereotactic radio-
Neurosurgery. 2016;79(2):279–85. surgery for brainstem metastases: an international
28. Roberge D, Souhami L. Tumor bed radiosurgery
cooperative study to define response and toxicity. Int J
following resection of brain metastases: a review. Radiat Oncol Biol Phys. 2016;96(2):280–8.
Technol Cancer Res Treat. 2010;9(6):597–602. 41. Trifiletti DM, Lee CC, Shah N, Patel NV, Chen SC,
29. Brennan C, Yang TJ, Hilden P, Zhang Z, Chan K, Sheehan JP. How does brainstem involvement affect
Yamada Y, et al. A phase 2 trial of stereotactic radio- prognosis in patients with limited brain metasta-
surgery boost after surgical resection for brain metas- ses? Results of a matched-cohort analysis. World
tases. Int J Radiat Oncol Biol Phys. 2014;88(1):130–6. Neurosurg. 2016;88:563–8.
30. Mahajan A, Ahmed S, McAleer MF, Weinberg JS, Li 42. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R,
J, Brown P, et al. Post-operative stereotactic radiosur- Loeffler J, et al. Single dose radiosurgical treat-
gery versus observation for completely resected brain ment of recurrent previously irradiated primary
metastases: a single-Centre, randomised, controlled, brain tumors and brain metastases: final report of
phase 3 trial. Lancet Oncol. 2017;18(8):1040–8. RTOG protocol 90-05. Int J Radiat Oncol Biol Phys.
31. Brown PD, Ballman KV, Cerhan JH, Anderson SK, 2000;47(2):291–8.
Carrero XW, Whitton AC, et al. Postoperative ste- 43. Minniti G, Clarke E, Lanzetta G, Osti MF, Trasimeni
reotactic radiosurgery compared with whole brain G, Bozzao A, et al. Stereotactic radiosurgery for brain
radiotherapy for resected metastatic brain dis- metastases: analysis of outcome and risk of brain
ease (NCCTG N107C/CEC.3): a multicentre, ran- radionecrosis. Radiat Oncol. 2011;6:48.
domised, controlled, phase 3 trial. Lancet Oncol. 44. Asher AL, Burri SH, Wiggins WF, Kelly RP, Boltes
2017;18(8):1049–60. MO, Mehrlich M, et al. A new treatment paradigm:
32. Jensen CA, Chan MD, McCoy TP, Bourland JD,
neoadjuvant radiosurgery before surgical resec-
deGuzman AF, Ellis TL, et al. Cavity-directed radio- tion of brain metastases with analysis of local
surgery as adjuvant therapy after resection of a brain tumor recurrence. Int J Radiat Oncol Biol Phys.
metastasis. J Neurosurg. 2011;114(6):1585–91. 2014;88(4):899–906.
33. Hartford AC, Paravati AJ, Spire WJ, Li Z, Jarvis LA, 45. Patel KR, Burri SH, Boselli D, Symanowski JT, Asher
Fadul CE, et al. Postoperative stereotactic radiosur- AL, Sumrall A, et al. Comparing pre-operative ste-
gery without whole-brain radiation therapy for brain reotactic radiosurgery (SRS) to post-operative whole
metastases: potential role of preoperative tumor size. brain radiation therapy (WBRT) for resectable brain
Int J Radiat Oncol Biol Phys. 2013;85(3):650–5. metastases: a multi-institutional analysis. J Neuro-
34. Jagannathan J, Yen CP, Ray DK, Schlesinger D,
Oncol. 2017;131(3):611–8.
Oskouian RJ, Pouratian N, et al. Gamma Knife radio- 46.
Bindal AK, Bindal RK, Hess KR, Shiu A,
surgery to the surgical cavity following resection of Hassenbusch SJ, Shi WM, et al. Surgery versus
brain metastases. J Neurosurg. 2009;111(3):431–8. radiosurgery in the treatment of brain metastasis. J
35. Soltys SG, Adler JR, Lipani JD, Jackson PS, Choi CY, Neurosurg. 1996;84(5):748–54.
Puataweepong P, et al. Stereotactic radiosurgery of 47. Muacevic A, Wowra B, Siefert A, Tonn JC, Steiger
the postoperative resection cavity for brain metasta- HJ, Kreth FW. Microsurgery plus whole brain irra-
ses. Int J Radiat Oncol Biol Phys. 2008;70(1):187–93. diation versus Gamma Knife surgery alone for treat-
36. Choi CY, Chang SD, Gibbs IC, Adler JR, Harsh GR, ment of single metastases to the brain: a randomized
Lieberson RE, et al. Stereotactic radiosurgery of the controlled multicentre phase III trial. J Neuro-Oncol.
postoperative resection cavity for brain metastases: 2008;87(3):299–307.
48. Roos DE, Smith JG, Stephens SW. Radiosurgery

53. ASTRO. ASTRO releases second list of five radia-
versus surgery, both with adjuvant whole brain radio- tion oncology treatments to question, as part
therapy, for solitary brain metastases: a randomised of national Choosing Wisely® campaign 2014.
controlled trial. Clin Oncol. 2011;23(9):646–51. Available from: http://www.choosingwisely.org/
49.
National Comprehensive Cancer Network astro-releases-second-list/.
Clinical Practice Guidelines in Oncology (NCCN 54.
Bhatnagar AK, Flickinger JC, Kondziolka D,
Guidelines®) - Central Nervous System Cancers Lunsford LD. Stereotactic radiosurgery for four or
2018 updated March 20, 2018. Available from: more intracranial metastases. Int J Radiat Oncol Biol
https://www.nccn.org/professionals/physician_gls/ Phys. 2006;64(3):898–903.
pdf/cns.pdf. 55.
Bhatnagar AK, Kondziolka D, Lunsford LD,
50. Wang TJ, Saad S, Qureshi YH, Jani A, Isaacson SR, Flickinger JC. Recursive partitioning analysis of
Sisti MB, et al. Outcomes of Gamma Knife radiosur- prognostic factors for patients with four or more intra-
gery, bi-modality & tri-modality treatment regimens cranial metastases treated with radiosurgery. Technol
for patients with one or multiple brain metastases: Cancer Res Treat. 2007;6(3):153–60.
the Columbia University medical center experience. 56. Dea N, Borduas M, Kenny B, Fortin D, Mathieu

J Neuro-Oncol. 2015;122(2):399–408. D. Safety and efficacy of Gamma Knife surgery for
51. Aoyama H, Tago M, Shirato H. Japanese radiation brain metastases in eloquent locations. J Neurosurg.
oncology study group I. stereotactic radiosurgery with 2010;113:79–83.
or without whole-brain radiotherapy for brain metas- 57. Hsu F, Nichol A, Ma R, Kouhestani P, Toyota B,
tases: secondary analysis of the JROSG 99-1 random- McKenzie M. Stereotactic radiosurgery for metasta-
ized clinical trial. JAMA Oncol. 2015;1(4):457–64. ses in eloquent central brain locations. Can J Neurol
52. Churilla TM, Handorf E, Collette S, Collette L,
Sci. 2015;42(5):333–7.
Dong Y, Aizer AA, et al. Whole brain radiotherapy 58. Trifiletti DM, Lee CC, Winardi W, Patel NV, Yen CP,
after stereotactic radiosurgery or surgical resection Larner JM, et al. Brainstem metastases treated with
among patients with one to three brain metastases and stereotactic radiosurgery: safety, efficacy, and dose
favorable prognoses: a secondary analysis of EORTC response. J Neuro-Oncol. 2015;125(2):385–92.
22952-26001. Ann Oncol. 2017;28(10):2588–94.
Hypofractionated Stereotactic
Radiosurgery (HF-SRS)
23
in the Treatment of Brain
Metastases
Jordan A. Torok, Scott R. Floyd, Peter E. Fecci,

and John P. Kirkpatrick
Introduction Oncology Group (RTOG) 9005, a dose escalation

study of SF-SRS delivered to either recurrent brain
Single-fraction stereotactic radiosurgery (SF-SRS) metastases following whole-brain radiotherapy
is recognized as a primary treatment for brain (WBRT) or recurrent gliomas after partial brain
metastases based on its ability to deliver a high irradiation, dose limits of 24, 18, and 15 Gy were
dose of radiation to a target, killing all viable established for lesions <2, 2–3, and 3–4 cm in
tumor cells, while minimizing damage to sur- maximum dimension, respectively. Paradoxically,
rounding normal tissue. Initially, delivery of high as tumor diameter and volume increase, the
dose of radiation with sufficient accuracy and pre- number of tumor cells increases dramatically, so
cision necessitated that a stereotactic headframe administering lower doses to a much greater num-
be fixed to the patient’s skull. Thus, radiosurgical ber of tumor cells limits the success of SRS for
treatments were most often delivered in a single larger tumors. This limitation has definite clinical
fraction, as it was difficult and uncomfortable to consequences; a study from the Cleveland Clinic
remove and reattach the headframe on successive found that the dose limit imposed by increasing
days. While single-dose treatment of small lesions size resulted in much lower rates of local control
is effective at killing small brain metastases and for brain metastases larger than 2 cm diameter [1].
sparing normal tissue injury, the ability to safely The location of a brain metastasis often
deliver a sterilizing dose in a single fraction to imposes an additional constraint on SF-SRS. In
larger tumors is lost because normal brain tissue SF-SRS, a limiting maximum dose of 8–10 Gy is
cannot be adequately spared. In Radiation Therapy typically employed as single fractions of 12 Gy
or more to the anterior visual pathways carry a
risk of radiation-induced optic neuropathy [2, 3].
J. A. Torok · S. R. Floyd Indeed, patients with metastases within 5 mm of
Department of Radiation Oncology, Duke Cancer the optic nerves and chiasm were not even eligi-
ble for several RTOG trials involving SF-SRS [4,
P. E. Fecci 5]. Similarly, metastases either abutting or within
Center for CNS Metastasis, Department of
Neurosurgery, Duke Cancer Institute, the brainstem impose a substantial limitation on
Durham, NC, USA the use of SF-SRS, as it is desirable not to exceed
J. P. Kirkpatrick (*) the 13–14 Gy maximum dose to this organ in a
Center for CNS Metastasis, Departments of Radiation single fraction [6, 7]. The “implicit bias” imposed
Oncology and Neurosurgery, Duke Cancer Institute, by volume and critical organ dose constraints is
Durham, NC, USA less well appreciated. In particular, when con-
e-mail: john.kirkpatrick@duke.edu

https://doi.org/10.1007/978-3-030-42958-4_23
330 J. A. Torok et al.
touring the dural attachments of intact or resected underlying HF-SRS and presents clinical data on
metastases, it is important to include the entire HF-SRS in the treatment of brain metastases.
involved dural surface in the target volume [8].
However, this often results in an irregularly
shaped target where the maximum diameter is odeling Tumor Kill and Normal
M
much larger than that of the intraparenchymal Tissue Damage in Radiosurgery [9]
tumor or resection cavity. Consequently, when
intending to employ a single-fraction technique, The relationship between radiation dose and
there is a tendency to draw a structure that does tumor cell survival may be represented by the lin-
not exceed the geometric limits that would permit ear quadratic model, at least below 10 Gy per
single-fraction SRS. fraction [9]. In the linear-quadratic model, a plot
In contrast to SF-SRS, “conventional” radia- of surviving cell fraction versus single radiation
tion therapy minimizes damage by utilizing mul- doses shows that the log of the surviving cell
tiple small fractions, typically administered to fraction (SCF) can be represented by a two-
large target volumes consisting of the tumor and parameter model, where the first parameter, α, is
the surrounding tissue at risk for tumor involve- the initial, linear portion of the plot—for exam-
ment. Normal tissue repair between fractions per- ple, where the SCF is linearly proportional to
mits one to administer high total doses of dose (D, units Gy) on a log-linear graph with a
radiation and still obtain acceptable toxicity in slope of −α, that is, SCF = exp[−αD]. The sec-
the surrounding normal tissue. In conventionally ond parameter, β, describes the portion of the
fractionated regimens, a course of radiation ther- curve, where SCF is proportional to the square of
apy delivered to a brain tumor typically spans dose. A radiation survival curve thus “bends” at
weeks. In addition, older linear accelerator-based moderate doses, and SCF depends on both dose
technology employed relocatable immobilization and the square of dose, that is, SCF = exp[−αD –
devices that resulted in larger day-to-day varia- βD2]. In this linear-quadratic (LQ) model, the
tions of patient positioning than the fixed head- response to radiation is often characterized by the
frames. Over the past decade, the implementation α/β ratio, which tends to be on the order of
of high-resolution image guidance, integrated 2–3 Gy for brain tissue and 10 Gy for rapidly pro-
into the radiosurgery system and coupled to a liferating tumors, such as brain metastases.
robotically controlled couch, permitted correc- Admittedly, this is a somewhat simplistic view
tion of translational and rotational errors in and the response to radiation is also influenced by
patient position at the time of treatment. These many other factors including the microenviron-
technological advances allow for tighter, or ment (e.g., oxygen concentration); the capacity
smaller, volume expansion to account for set-up of cells to repair, repopulate, and redistribute in
error or patient motion during the administration the cell cycle; and the immunologic milieu.
of radiation therapy. The concept of using multiple small daily
The introduction of image-guided radiosur- fractions of radiation to minimize normal tissue
gery systems offers highly accurate, precise, and toxicity is well-supported by preclinical data and
reproducible patient positioning and target local- clinical experience [9]. Using the linear-quadratic
ization, facilitating multi-fraction treatments model, one can calculate a biologically effective
with radiosurgical quality. The use of a hypofrac- dose (BED) for a particular α/β ratio (units Gy),
tionated (nominally two to five sessions) stereo- total dose (D), and dose/fraction (d, Gy):
tactic radiosurgery (HF-SRS) may provide an
improved balance of tumor control and normal BEDa / b = D éë1 + d (a / b ) ùû

tissue toxicity over single-fraction radiosurgery
(SRS), particularly in larger tumors and those For a low α/β tissue, BED will increase much
located next to or within critical structures. This more rapidly with increasing dose per fraction
chapter discusses the radiobiologic rationale than the BED for a high α/β tissue. Consequently,
23 Hypofractionated Stereotactic Radiosurgery (HF-SRS) in the Treatment of Brain Metastases 331
one could potentially exploit the difference in α/β a

24
ratio between tumor and normal tissue, by frac-
22
tionating the dose and, thereby, improving the 100 Gy2 120 Gy2
20
therapeutic ratio. For example, consider the case 160 Gy2 200 Gy2
18
of a tumor in a normal tissue with α/β ratios of 10
Dose/Fraction (Gy)
16
and 2 Gy, respectively. For 16 Gy delivered in a 14
single fraction, the BED will be 41.6 Gy10 and 12
144 Gy2 for tumor and normal tissue, respec- 10
tively. However, for a course of eight fractions 8
delivered at 5.08 Gy/fraction, BED for the nor- 6

4
mal tissue remains at 144 Gy2 but the BED for the
2
tumor is 61.3 Gy10, an increase of 47%. 1 2 3 4 5 6 7 8 9 10
Alternatively, treating in five fractions at 5.4 Gy/ Number of Fractions
fraction yields the same BED for the tumor b
(41.6 Gy10), though the BED for the normal tis- 24
sue is reduced by 31% (99.9 Gy2 versus 144Gy2). 22

50 Gy10 65 Gy10
On-line BED calculators can be used to guide 20
80 Gy10 100 Gy10
radiation dose prescriptions. Representative BED 18
Dose/Fraction (Gy)
16
isoeffect plots are presented in Fig. 23.1.
14
The calculated BED2 (associated with normal
12
tissue toxicity) and BED10 (modeling control of 10
rapidly proliferating tumors) for typical single- 8
fraction SRS and HF-SRS regimens encountered 6
in the treatment of brain metastases are shown in 4
Fig. 23.2. Note the improved balance of lower 2
1 2 3 4 5 6 7 8 9 10
BED2 and higher BED10, which should yield
Number of Fractions
decreased toxicity and improved tumor control
for the hypofractionated versus single-fraction Fig. 23.1 Biologically effective dose (BED) isoeffect
schemes. In particular, consider the case of a sin- plots for dose/fraction and number of fractions adminis-
gle fraction of 15 Gy—the maximum “safe” dose tered for (a) α/β = 2 Gy and (b) α/β = 10 Gy calculated
using the linear-quadratic (LQ) mode [9]. (From
for a 3–4 cm diameter lesion established in RTOG Kirkpatrick et al. [10]. Reprinted with permission from
9005. In this case, BED2 and BED10 are 127.5 Gy2 Duke University Press)
and 37.5 Gy10, respectively. When utilizing three
8 Gy fractions, BED2 decreases to 120 Gy2 while
BED10 increases to 43.2 Gy10. While this clearly of this model [14–16]. Others assert that radiobi-
favors the HF-SRS regimen over SF-SRS, other ologic mechanisms, such as profound vascular
patient, tumor, and treatment factors must be damage [17, 18] and antigen expression, apart
considered when selecting the appropriate dose from classic DNA damage, are evoked above a
regimen for an individual patient, including per- threshold dose of 8–12 Gy, and that the high lev-
formance status, medical comorbidities, psycho- els of tumor control observed in radiosurgery
social issues, logistics, histology, and the timing reflect this “new radiobiology” and enhanced
and nature of surgery, and systemic treatments. dose-response [16, 19–21] (Fig. 23.3). If one
The shape of the dose-response curve above assumes that there is an exaggerated tumor
10 Gy is controversial [11–13]. Some argue that response above some threshold dose, it would
the linear-quadratic model provides an adequate seem appropriate to design treatment plans and
representation of the dose-response relationship select dose regimens such that the dose encom-
at high doses, and that observed clinical out- passing the metastasis always exceeds this
comes are entirely consistent with the predictions threshold dose [10]. Conversely, the plan should
70
20Gyx1
60
9Gyx3
BED10 (Gy10)
18Gyx1
50 6Gyx5
8Gyx3
16Gyx1
40 5Gyx5
15Gyx1
14Gyx1
30 Single-Fraction SRS
12Gyx1
Hypofractionated SRS
20
50 100 150 200 250
BED2 (Gy2)
Fig. 23.2 BED2 and BED10 (biologically effective doses more damage to normal tissue. The response of rapidly pro-
calculated for α/β ratios of 2 and 10 Gy, respectively) for liferating tissue, such as brain metastases, to radiation is
clinically employed single-fraction SRS and hypofraction- better represented by BED10 with a higher BED10 suggest-
ated SRS. Note that BED2 is associated with response of ing improved local control. (From Kirkpatrick et al. [10].
normal tissue to radiation, and increasing BED2 results in Reprinted with permission from Duke University Press)
Fig. 23.3 Speculative 1.E+00

surviving cell fraction
(SCF) versus single-
dose irradiation response 1.E-01
curves for the linear-
Surviving Cell Fraction
In vitro
quadratic (LQ) model, Observations
in vitro cell cultures and 1.E-02
in vivo tumors with SCF
determined by the LQ
product of direct cell kill 1.E-03 Model
and indirect vascular
In vivo
damage. (From
Cytotoxic x
Kirkpatrick et al. [10]. 1.E-04
Vascular Damage
Reprinted with
Effect?
permission from Duke
University Press) 1.E-05
0 5 10 15 20
Dose (Gy)
be designed such that the dose in the surrounding nisms for metastases and normal tissues (which
normal tissue rarely goes above the threshold. It may differ) would not only aid rational plan
is also important to recall that there is no funda- design but could open new avenues for increasing
mental reason that the threshold for a dose effect the therapeutic ratio.
should be the same for a brain metastasis as the The above issue of dose-response does not
surrounding normal brain parenchyma. In any include the other critical element in assessing
case, an improved understanding of the in vivo toxicity—the volume of normal tissue irradiated.
dose-response curves and underlying mecha- As discussed by Marks et al. in the Quantitative
Analyses of Normal Tissue Effects in the Clinic tionated SRS, treatment may be delivered once
(QUANTEC) series of papers [22–24], normal daily on consecutive days or as infrequently as
tissue complications increase as the volume of twice/week. In this case, the issue still remains on
tissue receiving some minimum dose increases, the optimum timing that permits adequate repair
and this behavior is observed in a wide variety of of normal tissues while minimizing the adverse
tissues. For example, the volume of brain tissue impact of tumor cell repopulation.
receiving 12 Gy or more (V12Gy) in radiosurgery Finally, intriguing evidence is emerging that irra-
appears to be correlated with the risk of radione- diation of tumors may also release epitopes that
crosis, particularly when this volume exceeds stimulate the immune system, improve local con-
10–15 ml. Note, however, that this limitation trol, and, perhaps more importantly, decrease the
appears overly restrictive, as nearly all single- appearance of new, distant disease in the brain and
fraction radiosurgery plans would exceed this body [26]. While the high dose per fraction observed
limit when even moderately sized lesions were in single-fraction SRS may be quite effective at
treated to accepted doses [6]. While the linear- damaging the vasculature and enhancing local con-
quadratic model is useful in comparing BED trol, the resulting impaired perfusion could limit
delivered via different dose fractionation regi- transport of antigens and immune cells, inhibiting
mens, the most relevant method for doing so the global immunomodulatory effect of radiation
remains unclear [25]. Recognizing these limita- [27]. Thus, it has been suggested that a hypofrac-
tions, the fundamental principles of stereotactic tionated regimen could still generate antigens with-
radiosurgery—highly conformal treatment plans, out impairing transport, and that this treatment
small margins around the target, accurate target strategy would produce a more robust immune
localization, minimal position deviation, exqui- response [26, 28]. Such an approach might have
site attention to detail, and unwavering quality even greater impact when combined with one or
assurance—should aid in minimizing the irradi- more of the immunomodulating drugs (discussed
ated volume and should always be employed. elsewhere in this book) that are profoundly chang-
To identify and select the optimal dose regi- ing clinical practice, though a great deal remains to
men that maximizes tumor kill and minimizes be understood about this complex relationship.
normal tissue damage, one should also consider
time. Decreasing the time between fractions and
the total length of the treatment course should Clinical Outcomes
decrease tumor cell repopulation and, thus,
enhance the efficacy of the regimen. In particular, Brain metastases occur in approximately 20–40%
this should be more beneficial in the more rapidly of patients with advanced cancer and have become
growing malignant tumors (e.g., metastases) than more prevalent over the past decade, given
in the indolent tumors (e.g., benign schwanno- improved systemic therapies for certain cancers,
mas and meningiomas). If insufficient recovery such as trastuzumab for Her2-amplified breast
time between treatments is allowed, the normal cancer, targeted tyrosine kinase inhibitors for
brain parenchyma would experience incomplete EGFR-mutated and ALK-rearranged non-small
repair and would exhibit more pronounced late cell lung cancer, and immunotherapy for mela-
effects. While an interval of at least 8 h between noma. Given these systemic therapy advances,
fractions has generally been considered sufficient patients not only live longer with more time to
to permit repair of normal tissues, the QUANTEC develop brain metastases, but these patients with
analysis of daily versus twice-daily brain treat- treated brain metastases also live longer, thus
ments called this into question. The analysis by leading to the rising prevalence of brain metas-
Lawrence et al. [23] suggested that hyperfrac- tases. Given the longer survival of many cancer
tionated treatment was associated with an patients with brain metastases, the local control
increased risk of radionecrosis compared to once of each treated lesion and potential long-term
daily treatment at equivalent BED. In hypofrac- toxicity are increasingly important clinical con-
siderations that may influence not only survival, lated to the highest dose levels of 27–30 Gy in
but neurocognitive function and quality of life. three fractions and 31–35 Gy in five fractions
There have been numerous retrospective [59]. One-year local control was 69%, and no
studies published about the outcomes of intact grade 3 toxicities were reported.
and resected larger brain metastases following Many of the above studies included patients with
HF-SRS [29–71]. To date, there is no prospective intact brain metastases. In patients who undergo
randomized evidence demonstrating either supe- initial surgical resection, the need to adequately
rior efficacy or reduced toxicity between SF-SRS cover the surgical cavity often results in large and
and HF-SRS. However, multiple retrospective irregular target volumes. For SF-SRS, this necessi-
series lend support for this approach and will be tates a lower dose to limit significant toxicity. Two
reviewed further in this section. Single institu- recent prospective studies of postoperative SF-SRS
tion experience from Seoul National University lend insight into the challenges of this situation.
compared their results with SF-SRS (n = 58) to Investigators at the MD Anderson Cancer Center
HF-SRS (n = 40), where the latter typically utilized randomized patients after surgical resection for
a regimen of 6 Gy × 6. In both groups, 1-year local brain metastases to either observation or SF-SRS
progression-free survival was approximately 70%, (with a maximum allowed resection cavity diame-
while toxicity was significantly less for HF-SRS ter of 4 cm). SRS dose was based on the SRS target
[48]. Minniti et al. first reported their experience volume: 16 Gy for ≤10 cc, 14 Gy for 10.1–15 cc,
of HF-SRS (9–12 Gy × 3) in patients with one and 12 Gy for >15 cc. The addition of SRS signifi-
to three brain metastases, describing 1-year local cantly reduced local recurrence, with a 1-year free-
control and radionecrosis rates of 88% and 7%, dom from local recurrence of 72% with SF-SRS
respectively [54]. These authors subsequently [52]. Tumor size even after resection was an inde-
compared their institutional experience of SF-SRS pendent predictor for local recurrence; however,
versus HF-SRS (9 Gy × 3) for brain metastases suggesting lower dose SF-SRS provided subopti-
measuring >2 cm. With 289 patients, the 1-year mal local control. In our experience, resection of
cumulative local control for SF-SRS and HF-SRS intact brain metastases usually results in a planning
were 77% and 91%, respectively, while the rate target volume for postoperative SRS greater than
of radionecrosis with HF-SRS and SF-SRS were 3 cm in maximum diameter and, consequently, our
20% and 8%, respectively [57]. These differ- practice is to hypofractionate such patients.
ences remained significant after propensity score NCCTG N107C/CEC.3 was a multi-
adjustment. institutional randomized phase III trial compar-
At Dana Farber/Harvard Cancer Center, ing the efficacy and toxicity of postoperative
HF-SRS (~90% received 5 Gy × 5) was generally WBRT versus SF-SRS (with a maximum allowed
utilized for tumors of size >3 cm, cases with a resection cavity diameter of 5 cm). SF-SRS dose
high V12Gy, or for those in close proximity to a was based on cavity size in a similar manner to
critical structure. In 70 patients treated with the above study. The 1-year surgical bed local
HF-SRS, the authors reported a 1-year LC of control was 60% with SF-SRS, which was infe-
56% with symptomatic radiation-induced treat- rior to that of WBRT at 81% [30]. While cogni-
ment changes occurring in only 4% [63]. In one tive endpoints were improved with SF-SRS and
of the few prospective studies investigating survival was similar to the WBRT arm, the
HF-SRS, Murai et al. performed a dose escala- disappointing surgical bed control warrants fur-
tion study utilizing three- and five-fraction regi- ther study to improve efficacy.
mens. Patients with tumors ≥2.5 cm were In evaluating the optimal surgical cavity target
included—tumors in the 2.5–4 cm range were volume, investigators at Stanford reviewed their
treated with three fractions while those with experience with HF-SRS with a 2 mm margin to
tumors ≥4 cm were treated with five fractions. A a cohort primarily treated with SF-SRS with no
total of 54 patients with 61 large brain metastases expansion. Compared to the SF-SRS cohort, the
were included, with the dose safely being esca- 12-month cumulative incidence of local failure
for the HF-SRS group was only 3% (compared to cohort that did not receive surgery, where 1-year
16%), while toxicity rates trended in favor of local failure was 28% with HF-SRS alone.
HF-SRS [31]. Minniti et al. reported their experi- Collectively, these studies suggest reduced
ence of HF-SRS after resection for melanoma toxicity rates and potentially increased local con-
brain metastases and found a 1-year local failure trol with the use of HF-SRS in situations where
rate of 12% [72]. This compared favorably to a high-dose SF-SRS is not feasible. Further valida-
tion with prospective controlled trials will be
Table 23.1 Summary of studies of single-fraction, staged, and hypofractionated SRS in the treatment of intact and
resected brain metastases
Local
Median Median Median lesion control
follow-up, SRS marginal dose/ diametera 1-year post Rate of
Study Year n months approach Fx × Fx # (range), cm SRS RN
SRS to intact brain
metastases
Angelov 2017 54 NR Staged 15Gy × 2 2.7 (1.7–3.9) NR 11.1%
et al. [29]
Dohm 2017 54 7.7 Staged 14.5Gy × 2 2.8 (1.2–4.9) 88.7% 10.3%
et al. [32]
Feuvret 201424 20 SF-SRS 20Gy × 1 4.5 (3.2–6.0) vs 60.4% 0%
et al. [37] vs vs vs vs 3.8 (3.3–4.7) vs vs
12 11 HF-SRS 11Gy × 3 100% 0%
Han et al. [40] 2012 80 13.8 SF-SRS 13.8Gy × 1 4.0 (mean) 84.6% 38.8%
Hasegawa 2017 56 6.0 HF-SRS 13Gy × 2 3.4 80.0% NR
et al. [42]
Higuchi 2009 43 10 Staged 10Gy × 3 3.2 (mean) 75.9% 2.0%
et al. [43]
Inoue 2014 78 6.2 HF-SRS 6.2Gy × 5 2.9 98.4% 2.0%
et al. [44]
Jeong 2015 37 10 HF-SRS 11.7Gy × 3 3.2 (2.6–4.6) 86.7% 15.8%
et al. [46]
Kim et al. [47] 2016 36 13.4 HF-SRS 8Gy × 3 3.3 (mean) NR 2.7%
Minniti 2016 151 NR SF-SRS vs 18Gy × 1 All >2 cm 77% 27.7%
et al. [57] vs HF-SRS vs vs vs
138 9Gy × 3 90% 14.4%
Minniti 2017 60 13 HF-SRS 9Gy × 3 2.5 (1.6–4.3) 72% 8%
et al. [72]
Mucaevic 2008 31 21 SF-SRS 21Gy × 1 2 96.8% NR
et al. [58]
Murai 2014 51 NR SF-SRS vs 8-14Gy × 1 >2.5 76% NR
et al. [59] HF-SRS vs vs
5-11Gy × 3 59%
Navarria 2016 51 NR HF-SRS 9Gy × 3 2.9 (2.1–5) 100% 5.9%
et al. [60] vs vs vs
8Gy × 4 91% 5.9%
Prabhu 2017 60 10.3 SF-SRS 18Gy × 1 2.2 63.3% 17.2%
et al. [62]
Wegner 2015 36 NR HF-SRS 24Gy in 2–5Fx 3.1 (2.7–5.4) 63.0% NR
et al. [67]
Yomo 2014 58 14 HF-SRS 5.5Gy × 5 3.2 (2.7–4.7) 98.4% 2.0%
et al. [69]
Zimmerman 2015 52 NR SF-SRS 15Gy × 1 3.5 (3.0–5.8) 80.2% 7.0%
et al. [71]
(continued)
Local
Median Median Median lesion control
follow-up, SRS marginal dose/ diametera 1-year post Rate of
Study Year n months approach Fx × Fx # (range), cm SRS RN
SRS to resection cavity
Brown 2017 39 NR SF-SRS 15Gy × 1 60% <3 cm, 40% 34% 10.3%
et al. [30] >3 cm
Choi 2012 97 10 HF-SRS 8Gy × 3 2.5 (0.5–5.0) 90.7% 5%
et al. [31]
Doré 2017 95 17 HF-SRS 7.7Gy × 3 2.8 (1.2–5.0) 84% 20.6%
et al. [73]
Ling 2015 99 12.2 HF-SRS 7.3Gy × 3 >3.0 71.8% NR
et al. [51]
Mahajan 2017 17 11.1 SF-SRS 12Gy × 1 2.6 (1.2–3.8) 72% 0%
et al. [52]
Minniti 2013 101 16 HF-SRS 9Gy × 3 3.2 (2.9–4.1) 93% 9.0%
et al. [56]
Minniti 2017 60 13 HS-SRS 9Gy × 3 2.8 (1.7–4.8) 88% 13%
et al. [72]
Pessina 2016 69 12.5 HF-SRS 10Gy × 3 3.8 (2.0–7.3) 100% 0%
et al. [61]
Prabhu 2017 93 10.3 SF-SRS 15Gy × 1 2.6 80.8% 33.5%
et al. [62]
Vogel 2015 30 9.5 HF-SRS 7.8Gy × 5 3.8 (2.8–6.7) 68.5% 10%
et al. [65]
Zimmerman 2015 33 NR SF-SRS 15Gy × 1 4.0 (3.0–6.8) 79% 3.0%
et al. [71]
When only tumor or resection cavity volume reported, diameter calculated from (1.91 × volume)1/3
a
Abbreviations: Fx fraction(s), Gy Gray, n total number of patients, NR not reported, RN radionecrosis
critical to further optimize the HF-SRS platform life, local control, appearance of distant metasta-
(Table 23.1). ses in the brain, and overall survival. Patients are
treated on 3 consecutive days to doses of 24, 27,
30, or 33 Gy (8–11 Gy/fraction) using a 3 + 3
Ongoing Clinical Trials dose escalation scheme. Preliminary results have
been presented in abstract form.
Clinical trials of hypofractionated SRS for brain A retrospective study of patients with four
metastases registered at ClinicalTrials.gov are or more brain metastases treated with single-
presented in Table 23.2. To illustrate the oppor- isocenter, multi-target (SIMT) SRS showed that
tunities to better define the role of HF-SRS in the patients with higher V12Gy exhibited poorer over-
treatment of brain lesions, it is worthwhile to all survival than those with lower V12Gy [74]; an
discuss several of these trials. For example, the increased number of brain metastases was not
Stanford study (NCT00928226) asks the ques- associated with diminished survival. As these
tion, “What is the maximum tolerated dose patients were primarily treated with single-
(MTD) of HF-SRS for large brain metastases fraction SIMT SRS, the authors speculated
treated using a 3 fraction regimen?” Eligible that hypofractionated SRS might have reduced
patients have one to four brain metastases, one of radiation-related toxicity and improved outcome.
which is 4.2–33.5 cm3 (equivalent to a uniform This notion informed the subsequent prospective
sphere 2–4 cm in diameter), intact and unresect- trial at Duke University of SIMT SRS in patients
able. The primary outcome is MTD with second- with 4–10 brain metastases (NCT02886572). In
ary measures of acute and late toxicity, quality of that study, patients are initially planned for sin-
Table 23.2 Clinical trials of hypofractionated stereotactic radiosurgery (SRS) or stereotactic radiotherapy (SRT) in
patients with brain metastases
Principal ClinicalTrials.
Study Institution investigator gov identifier Primary outcome
Phase I/II study of fractionated Stanford S. Soltys NCT00928226 Determine MTD of SRS given
stereotactic radiosurgery to University in 3 fractions for brain
treat large brain metastases metastases 4.2–14.1 cm3 and
14.2–33.5 cm3
SRS to brain metastases Heidelberg J. Debus NCT03285932 Evaluate safety/efficacy of
resection cavity vs. post-op University post-op SRS compared to
WBRT (ESTRON) WBRT
SRS or hypofractionated Oncology Inst. G. Pesce NCT03561896 Rate of local recurrence in
stereotactic radiotherapy of Southern post-op SRS or HF-SRT
(HF-SRT) to brain metastases Switzerland
resection cavity
Fractionated stereotactic Moffitt Cancer S. Sahebjam NCT02187822 Determine MTD of TPI 287
radiotherapy (FSRT) in Center given concurrently with FSRT
treatment of brain metastases to treat brain metastases
Hypofractionated stereotactic Emory B. Eaton NCT01705548 Determine MTD of 5-fraction
radiosurgery in treating University SRS for brain metastases,
patients with large brain 3–6 cm diameter
metastasis
Hypofractionated stereotactic Istituto Clinico M. Tedeschi NCT02576522 Rate of local recurrence in
VMAT to the resection cavity Humanitas post-op HF-SRS for a single
from a single, large brain large brain metastasis
metastasis
Perfexion brain metastasis Princess C. Chung NCT00805103 Determine MTD of HF-SRS
(HF-SRT) Margaret for recurrent brain metastases
Hospital (at least 1 > 2 cm diameter)
post WBRT
Single-isocenter multi-target Duke Cancer G. Kim NCT02886572 Overall survival (SF-SRS vs
SRS for patients with 4–10 Institute HF-SRS for
brain metastases V12Gy < vs. > 20 ml)
Fractionated stereotactic National Taiwan C.-C. Wang NCT02672995 Determine MTD of 3-fraction
radiosurgery with concurrent University SRS + bevacizumab for brain
bevacizumab for brain Hospital metastases, 1.5–3.5 cm
metastases: a phase I dose diameter
escalation trial
Frameless fractionated MD Anderson A. Garg NCT02798029 Local control based on
stereotactic radiation therapy imaging for each lesion (up to
(FSRT) for brain mets 5 cm diameter) after 3–5
fraction SRS
Fractionated stereotactic University of D. Heron NCT02054689 MTD for 3-fraction SRS for
radiosurgery for large brain Pittsburgh brain metastases, 3–5 cm
metastases diameter
Hypofractionated stereotactic Institut de P. Royer NCT02913534 Overall survival of patients
radiation therapy of brain Cancérologie de with 1–3 brain metastases
metastases: evaluation of Lorraine treated with HF-FSRT
WBRT
gle-fraction SIMT SRS using the dose-volume is replanned and treated with a marginal pre-
constraints imposed by RTOG 9005, that is, scribed dose of 25 Gy administered in five 5-Gy
24 Gy marginal dose for targets <2 cm maximum fractions over consecutive days. Accrual to this
dimension and 18 Gy for targets 2–3 cm maxi- protocol was completed in August 2019 and the
mum diameter. However, if either the planned results will be reported in 2020.
V12Gy to normal brain parenchyma exceeds 20 ml Given the need to optimize treatment strate-
or any lesion involves the brainstem, the patient gies and to identify appropriate patient popula-
tions for HF-SRS in the treatment of large brain 2. Mayo C, Martel MK, Marks LB, Flickinger J, Nam J,
Kirkpatrick J. Radiation dose-volume effects of optic
lesions, all eligible patients should consider nerves and chiasm. Int J Radiat Oncol Biol Phys.
enrollment on a clinical trial. 2010;76(3 Suppl):S28–35.
3. Milano MT, Grimm J, Soltys SG, Yorke E, Moiseenko
V, Tome WA, et al. Single- and multi-fraction stereo-
tactic radiosurgery dose tolerances of the optic path-
Conclusion ways. Int J Radiat Oncol Biol Phys. 2018. https://doi.
org/10.1016/j.ijrobp.2018.01.053.
Compared to single-fraction SRS, hypofractionated 4. Andrews DW, Scott CB, Sperduto PW, Flanders AE,
SRS appears to offer a superior balance of efficacy Gaspar LE, Schell MC, et al. Whole brain radiation
therapy with or without stereotactic radiosurgery
and toxicity in patients with large brain metastases boost for patients with one to three brain metastases:
and resection cavities, as well as in lesions located phase III results of the RTOG 9508 randomised trial.
close to critical normal organs. This potential ben- Lancet. 2004;363(9422):1665–72.
efit comes at the expense of an increased number of 5. Shaw E, Scott C, Souhami L, Dinapoli R, Kline R,
Loeffler J, et al. Single dose radiosurgical treatment
treatment fractions. Fundamental studies and clini- of recurrent previously irradiated primary brain tumors
cal trials are required to identify the most appropri- and brain metastases: final report of RTOG protocol
ate applications for HF-SRS based on tumor and 90-05. Int J Radiat Oncol Biol Phys. 2000;47(2):291–8.
patient characteristics. It would be worthwhile to 6. Kirkpatrick JP, Marks LB, Mayo CS, Lawrence YR,
Bhandare N, Ryu S. Estimating normal tissue toxicity
establish the optimal scheme for total dose and in radiosurgery of the CNS: application and limitations
dose/fraction in HF-SRS as a function of tumor of QUANTEC. J Radiosurg SBRT. 2011;1:95–102.
histology, diameter, and location. Moreover, given 7. Mayo C, Yorke E, Merchant TE. Radiation associ-
the rapid progress in targeted and immunomodu- ated brainstem injury. Int J Radiat Oncol Biol Phys.
2010;76(3 Suppl):S36–41.
lating treatments for primary and metastatic malig- 8. Soliman H, Ruschin M, Angelov L, Brown PD,
nancies, these outcomes with HF-SRS need to be Chiang VLS, Kirkpatrick JP, et al. Consensus contour-
evaluated in the setting of concurrent and adjuvant ing guidelines for postoperative completely resected
systemic therapies. Establishing the benefits and cavity stereotactic radiosurgery for brain metastases.
Int J Radiat Oncol Biol Phys. 2018;100(2):436–42.
toxicities of combined HF-SRS and immunomod- 9. Hall EJ, Giaccia AJ. Radiobiology for the radiolo-
ulating therapy is particularly important, given the gist. 7th ed. Philadelphia: Lippincott Williams and
imputed role of HF-SRS in stimulating the immune Wilkins; 2012.
response. Because of the known toxicity of high- 10. Kirkpatrick JP, Soltys SG, Lo SS, Beal K, Shrieve
DC, Brown PD. The radiosurgery fractionation quan-
dose, single-fraction SRS when treating large dary: single fraction or hypofractionation? Neuro
lesions, it would be difficult to perform a random- Oncol. 2017;19(suppl_2):ii38–49.
ized trial of single-fraction SRS versus HF-SRS 11. Guerrero M, Li XA. Extending the linear-quadratic
with equipoise. Consequently, optimizing HF-SRS model for large fraction doses pertinent to stereotactic
radiotherapy. Phys Med Biol. 2004;49(20):4825–35.
demands well-constructed and well-analyzed clini- 12. Hanin LG, Zaider M. Cell-survival probability at

cal trials with sufficient size and length of follow- large doses: an alternative to the linear-quadratic
up to determine the safety and efficacy of HF-SRS model. Phys Med Biol. 2010;55(16):4687–702.
for the treatment of large and/or critically located 13. Park C, Papiez L, Zhang S, Story M, Timmerman
RD. Universal survival curve and single fraction
lesions, either intact or surgically resected. equivalent dose: useful tools in understanding potency
of ablative radiotherapy. Int J Radiat Oncol Biol Phys.
Conflict of Interest The authors report no conflicts of 2008;70(3):847–52.
interest. 14. Brown JM, Carlson DJ, Brenner DJ. The tumor

radiobiology of SRS and SBRT: are more than
the 5 Rs involved? Int J Radiat Oncol Biol Phys.
2014;88(2):254–62.
References 15. Brown JM, Carlson DJ, Brenner DJ. Dose escalation,
not “new biology,” can account for the efficacy of ste-
1. Vogelbaum MA, Angelov L, Lee SY, Li L, Barnett reotactic body radiation therapy with non-small cell
GH, Suh JH. Local control of brain metastases by ste- lung cancer. In reply to Rao et al. Int J Radiat Oncol
reotactic radiosurgery in relation to dose to the tumor Biol Phys. 2014;89(3):693–4.
margin. J Neurosurg. 2006;104(6):907–12.
16. Kirkpatrick JP, Brenner DJ, Orton CG. Point/

radiotherapy for resected metastatic brain dis-
Counterpoint. The linear-quadratic model is inappro- ease (NCCTG N107C/CEC.3): a multicentre, ran-
priate to model high dose per fraction effects in radio- domised, controlled, phase 3 trial. Lancet Oncol.
surgery. Med Phys. 2009;36(8):3381–4. 2017;18(8):1049–60.
17.
Fuks Z, Kolesnick R. Engaging the vascular 31. Choi CY, Chang SD, Gibbs IC, Adler JR, Harsh GR,
component of the tumor response. Cancer Cell. Lieberson RE, et al. Stereotactic radiosurgery of the
2005;8(2):89–91. postoperative resection cavity for brain metastases:
18. Garcia-Barros M, Paris F, Cordon-Cardo C, Lyden D, prospective evaluation of target margin on tumor con-
Rafii S, Haimovitz-Friedman A, et al. Tumor response trol. Int J Radiat Oncol Biol Phys. 2012;84(2):336–42.
to radiotherapy regulated by endothelial cell apopto- 32. Dohm A, McTyre ER, Okoukoni C, Henson A,

sis. Science. 2003;300(5622):1155–9. Cramer CK, LeCompte MC, et al. Staged stereotactic
19. Kirkpatrick JP, Meyer JJ, Marks LB. The linear-
radiosurgery for large brain metastases: local control
quadratic model is inappropriate to model high dose and clinical outcomes of a one-two punch technique.
per fraction effects in radiosurgery. Semin Radiat Neurosurgery. 2018;83(1):114–21.
Oncol. 2008;18(4):240–3. 33. Dohm AE, Hughes R, Wheless W, Lecompte M,

20. Song CW, Lee YJ, Griffin RJ, Park I, Koonce NA, Lanier C, Ruiz J, et al. Surgical resection and post-
Hui S, et al. Indirect tumor cell death after high- operative radiosurgery versus staged radiosurgery
dose hypofractionated irradiation: implications for for large brain metastases. J Neuro-Oncol. 2018;
stereotactic body radiation therapy and stereotac- 140:749.
tic radiation surgery. Int J Radiat Oncol Biol Phys. 34. Eaton BR, Gebhardt B, Prabhu R, Shu HK, Curran
2015;93(1):166–72. WJ Jr, Crocker I. Hypofractionated radiosurgery for
21. Sperduto PW, Song CW, Kirkpatrick JP, Glatstein
intact or resected brain metastases: defining the opti-
E. A hypothesis: indirect cell death in the radiosurgery mal dose and fractionation. Radiat Oncol. 2013;8:135.
era. Int J Radiat Oncol Biol Phys. 2015;91(1):11–3. 35. Ernst-Stecken A, Ganslandt O, Lambrecht U, Sauer
22. Bentzen SM, Constine LS, Deasy JO, Eisbruch A, R, Grabenbauer G. Phase II trial of hypofractionated
Jackson A, Marks LB, et al. Quantitative analyses of stereotactic radiotherapy for brain metastases: results
normal tissue effects in the clinic (QUANTEC): an and toxicity. Radiother Oncol. 2006;81(1):18–24.
introduction to the scientific issues. Int J Radiat Oncol 36. Fahrig A, Ganslandt O, Lambrecht U, Grabenbauer G,
Biol Phys. 2010;76(3 Suppl):S3–9. Kleinert G, Sauer R, et al. Hypofractionated stereo-
23. Lawrence YR, Li XA, el Naqa I, Hahn CA, Marks LB, tactic radiotherapy for brain metastases – results from
Merchant TE, et al. Radiation dose-volume effects in three different dose concepts. Strahlenther Onkol.
the brain. Int J Radiat Oncol Biol Phys. 2010;76(3 2007;183(11):625–30.
Suppl):S20–7. 37. Feuvret L, Vinchon S, Martin V, Lamproglou I, Halley
24. Marks LB, Yorke ED, Jackson A, Ten Haken RK, A, Calugaru V, et al. Stereotactic radiotherapy for
Constine LS, Eisbruch A, et al. Use of normal tissue large solitary brain metastases. Cancer Radiother.
complication probability models in the clinic. Int J 2014;18(2):97–106.
Radiat Oncol Biol Phys. 2010;76(3 Suppl):S10–9. 38. Fokas E, Henzel M, Surber G, Kleinert G, Hamm K,
25. Blonigen BJ, Steinmetz RD, Levin L, Lamba MA, Engenhart-Cabillic R. Stereotactic radiosurgery and
Warnick RE, Breneman JC. Irradiated volume as a fractionated stereotactic radiotherapy: comparison of
predictor of brain radionecrosis after linear accelera- efficacy and toxicity in 260 patients with brain metas-
tor stereotactic radiosurgery. Int J Radiat Oncol Biol tases. J Neuro-Oncol. 2012;109(1):91–8.
Phys. 2010;77(4):996–1001. 39. Follwell MJ, Khu KJ, Cheng L, Xu W, Mikulis DJ,
26. Demaria S, Golden EB, Formenti SC. Role of local Millar BA, et al. Volume specific response criteria for
radiation therapy in cancer immunotherapy. JAMA brain metastases following salvage stereotactic radio-
Oncol. 2015;1(9):1325–32. surgery and associated predictors of response. Acta
27. Park HJ, Griffin RJ, Hui S, Levitt SH, Song
Oncol. 2012;51(5):629–35.
CW. Radiation-induced vascular damage in tumors: 40. Han JH, Kim DG, Chung HT, Paek SH, Park CK,
implications of vascular damage in ablative hypofrac- Jung HW. Radiosurgery for large brain metastases. Int
tionated radiotherapy (SBRT and SRS). Radiat Res. J Radiat Oncol Biol Phys. 2012;83(1):113–20.
2012;177(3):311–27. 41. Han JH, Kim DG, Kim CY, Chung HT, Jung

28. Demaria S, Formenti SC. Radiation as an immuno- HW. Stereotactic radiosurgery for large brain metas-
logical adjuvant: current evidence on dose and frac- tases. Prog Neurol Surg. 2012;25:248–60.
tionation. Front Oncol. 2012;2:153. 42. Hasegawa T, Kato T, Yamamoto T, Iizuka H,

29. Angelov L, Mohammadi AM, Bennett EE, Abbassy Nishikawa T, Ito H, et al. Multisession gamma knife
M, Elson P, Chao ST, et al. Impact of 2-staged stereo- surgery for large brain metastases. J Neuro-Oncol.
tactic radiosurgery for treatment of brain metastases ≥ 2017;131(3):517–24.
2 cm. J Neurosurg. 2018;129(2):366–82. 43. Higuchi Y, Serizawa T, Nagano O, Matsuda S, Ono
30. Brown PD, Ballman KV, Cerhan JH, Anderson SK, J, Sato M, et al. Three-staged stereotactic radio-
Carrero XW, Whitton AC, et al. Postoperative ste- therapy without whole brain irradiation for large
reotactic radiosurgery compared with whole brain
metastatic brain tumors. Int J Radiat Oncol Biol Phys. 56. Minniti G, Esposito V, Clarke E, Scaringi C, Lanzetta
2009;74(5):1543–8. G, Salvati M, et al. Multidose stereotactic radiosur-
44. Inoue HK, Sato H, Seto K, Torikai K, Suzuki Y,
gery (9 Gy × 3) of the postoperative resection cavity
Saitoh J, et al. Five-fraction CyberKnife radiotherapy for treatment of large brain metastases. Int J Radiat
for large brain metastases in critical areas: impact on Oncol Biol Phys. 2013;86(4):623–9.
the surrounding brain volumes circumscribed with a 57. Minniti G, Scaringi C, Paolini S, Lanzetta G, Romano
single dose equivalent of 14 Gy (V14) to avoid radia- A, Cicone F, et al. Single-fraction versus multifrac-
tion necrosis. J Radiat Res. 2014;55(2):334–42. tion (3 × 9 Gy) stereotactic radiosurgery for large (>2
45. Inoue HK, Sato H, Suzuki Y, Saitoh J, Noda SE, Seto cm) brain metastases: a comparative analysis of local
K, et al. Optimal hypofractionated conformal radio- control and risk of radiation-induced brain necrosis.
therapy for large brain metastases in patients with Int J Radiat Oncol Biol Phys. 2016;95(4):1142–8.
high risk factors: a single-institutional prospective 58. Muacevic A, Wowra B, Siefert A, Tonn JC, Steiger
study. Radiat Oncol. 2014;9:231. HJ, Kreth FW. Microsurgery plus whole brain irra-
46. Jeong WJ, Park JH, Lee EJ, Kim JH, Kim CJ, Cho diation versus Gamma Knife surgery alone for treat-
YH. Efficacy and safety of fractionated stereotac- ment of single metastases to the brain: a randomized
tic radiosurgery for large brain metastases. J Korean controlled multicentre phase III trial. J Neuro-Oncol.
Neurosurg Soc. 2015;58(3):217–24. 2008;87(3):299–307.
47. Kim JW, Park HR, Lee JM, Kim JW, Chung HT, 59. Murai T, Ogino H, Manabe Y, Iwabuchi M, Okumura
Kim DG, et al. Fractionated stereotactic gamma T, Matsushita Y, et al. Fractionated stereotactic radio-
knife radiosurgery for large brain metastases: a therapy using CyberKnife for the treatment of large
retrospective, Single Center Study. PLoS One. brain metastases: a dose escalation study. Clin Oncol
2016;11(9):e0163304. (R Coll Radiol). 2014;26(3):151–8.
48. Kim YJ, Cho KH, Kim JY, Lim YK, Min HS, Lee 60. Navarria P, Pessina F, Cozzi L, Ascolese AM, De Rose
SH, et al. Single-dose versus fractionated stereotactic F, Fogliata A, et al. Hypo-fractionated stereotactic radio-
radiotherapy for brain metastases. Int J Radiat Oncol therapy alone using volumetric modulated arc therapy
Biol Phys. 2011;81(2):483–9. for patients with single, large brain metastases unsuit-
49. Kwon AK, Dibiase SJ, Wang B, Hughes SL, Milcarek able for surgical resection. Radiat Oncol. 2016;11:76.
B, Zhu Y. Hypofractionated stereotactic radiother- 61. Pessina F, Navarria P, Cozzi L, Ascolese AM, Maggi
apy for the treatment of brain metastases. Cancer. G, Riva M, et al. Outcome evaluation of oligometa-
2009;115(4):890–8. static patients treated with surgical resection followed
50. Lim TK, Kim WK, Yoo CJ, Kim EY, Kim MJ, Yee by hypofractionated stereotactic radiosurgery (HSRS)
GT. Fractionated stereotactic radiosurgery for brain on the tumor bed, for single, large brain metastases.
metastases using the Novalis Tx(R) system. J Korean PLoS One. 2016;11(6):e0157869.
Neurosurg Soc. 2018;61(4):525–9. 62. Prabhu RS, Press RH, Patel KR, Boselli DM,

51. Ling DC, Vargo JA, Wegner RE, Flickinger JC,
Symanowski JT, Lankford SP, et al. Single-fraction
Burton SA, Engh J, et al. Postoperative stereotactic stereotactic radiosurgery (SRS) alone versus surgical
radiosurgery to the resection cavity for large brain resection and SRS for large brain metastases: a multi-
metastases: clinical outcomes, predictors of intra- institutional analysis. Int J Radiat Oncol Biol Phys.
cranial failure, and implications for optimal patient 2017;99(2):459–67.
selection. Neurosurgery. 2015;76(2):150–6; discus- 63. Rajakesari S, Arvold ND, Jimenez RB, Christianson
sion 6–7; quiz 7. LW, Horvath MC, Claus EB, et al. Local control after
52. Mahajan A, Ahmed S, McAleer MF, Weinberg JS, Li fractionated stereotactic radiation therapy for brain
J, Brown P, et al. Post-operative stereotactic radiosur- metastases. J Neuro-Oncol. 2014;120(2):339–46.
gery versus observation for completely resected brain 64. Serizawa T, Higuchi Y, Yamamoto M, Matsunaga S,
metastases: a single-centre, randomised, controlled, Nagano O, Sato Y, et al. Comparison of treatment
phase 3 trial. Lancet Oncol. 2017;18:1040. results between 3- and 2-stage Gamma Knife radiosur-
53. Manning MA, Cardinale RM, Benedict SH, Kavanagh gery for large brain metastases: a retrospective multi-
BD, Zwicker RD, Amir C, et al. Hypofractionated institutional study. J Neurosurg. 2018;131:227–37.
stereotactic radiotherapy as an alternative to radiosur- 65. Vogel J, Ojerholm E, Hollander A, Briola C, Mooij R,
gery for the treatment of patients with brain metasta- Bieda M, et al. Intracranial control after Cyberknife
ses. Int J Radiat Oncol Biol Phys. 2000;47(3):603–8. radiosurgery to the resection bed for large brain
54. Minniti G, D’Angelillo RM, Scaringi C, Trodella LE, metastases. Radiat Oncol. 2015;10:221.
Clarke E, Matteucci P, et al. Fractionated stereotac- 66. Wang CC, Floyd SR, Chang CH, Warnke PC, Chio
tic radiosurgery for patients with brain metastases. J CC, Kasper EM, et al. Cyberknife hypofractionated
Neuro-Oncol. 2014;117(2):295–301. stereotactic radiosurgery (HSRS) of resection cavity
55. Minniti G, Esposito V, Clarke E, Scaringi C, Bozzao after excision of large cerebral metastasis: efficacy
A, Falco T, et al. Fractionated stereotactic radiosur- and safety of an 800 cGy × 3 daily fractions regimen.
gery for patients with skull base metastases from sys- J Neuro-Oncol. 2012;106(3):601–10.
temic cancer involving the anterior visual pathway. 67. Wegner RE, Leeman JE, Kabolizadeh P, Rwigema JC,
Radiat Oncol. 2014;9:110. Mintz AH, Burton SA, et al. Fractionated stereotac-
tic radiosurgery for large brain metastases. Am J Clin large brain metastases with stereotactic radiosurgery.
Oncol. 2015;38(2):135–9. Technol Cancer Res Treat. 2016;15(1):186–95.
68. Yomo S, Hayashi M. A minimally invasive treatment 72. Minniti G, Paolini S, D’Andrea G, Lanzetta G, Cicone
option for large metastatic brain tumors: long-term F, Confaloni V, et al. Outcomes of postoperative ste-
results of two-session Gamma Knife stereotactic reotactic radiosurgery to the resection cavity versus
radiosurgery. Radiat Oncol. 2014;9:132. stereotactic radiosurgery alone for melanoma brain
69. Yomo S, Hayashi M, Nicholson C. A prospective metastases. J Neuro-Oncol. 2017;132(3):455–62.
pilot study of two-session Gamma Knife surgery 73. Doré M, Martin S, Delpon G, Clément K, Campion
for large metastatic brain tumors. J Neuro-Oncol. L, Thillays F. Stereotactic radiotherapy following sur-
2012;109(1):159–65. gery for brain metastasis: predictive factors for local
70. Zhong J, Ferris MJ, Switchenko J, Press RH,
control and radionecrosis. Cancer Radiother. 2017
Buchwald Z, Olson JJ, et al. Postoperative stereotac- Feb;21(1):4–9.
tic radiosurgery for resected brain metastases: a com- 74. Limon D, McSherry F, Herndon J, Sampson J, Fecci
parison of outcomes for large resection cavities. Pract P, Adamson J, et al. Single fraction stereotactic radio-
Radiat Oncol. 2017;7(6):e419–e25. surgery for multiple brain metastases. Adv Radiat
71. Zimmerman AL, Murphy ES, Suh JH, Vogelbaum Oncol. 2017;2(4):555–63.
MA, Barnett GH, Angelov L, et al. Treatment of
Challenges and Controversies
in Stereotactic Radiosurgery
24
Jugal K. Shah and Douglas Kondziolka
Introduction 2 weeks of headache, 2 days of worsening

ataxia, and two episodes of nausea and vomit-
Patients diagnosed with cancer have a 20–40% ing. She presented to the emergency depart-
incidence of brain metastases, and 70% of those ment and neurological exam was notable for no
patients have multiple brain metastases [1]. For a focal deficits. Brain magnetic resonance imaging
decade, treatment has evolved to include more focal (MRI) revealed bilateral cerebellar hemisphere
approaches, such as radiosurgery and resection, and tumors: a right superior cerebellar tumor mea-
less use of whole-brain radiation therapy (WBRT) suring 2 cm with mild surrounding edema and a
[2]. Clinicians strive to reduce undesirable effects, left mid lateral cerebellar tumor measuring 3 cm
including induced cognitive deficits, hair loss, nau- with surrounding edema and mild 4th ventricular
sea, and fatigue, and to improve functional survival. effacement, but no hydrocephalus (Fig. 24.1a).
Stereotactic radiosurgery (SRS) was developed
to deliver powerful and precise energy to targeted
tumors [3]. While the treatment of one to four brain Current Treatment Paradigm
metastases with SRS alone without WBRT has
reached the level of Class 1 evidence, controversy The specific challenges in this case are both the
persists in the case of larger, multiple, and radio- size and location of the lesions. Larger brain metas-
resistant tumors. Management of radiation effects tasis (≥3 cm in diameter) present a treatment chal-
can be challenging both in terms of diagnosis and lenge for stereotactic radiosurgery, as the doses
therapy. Here we present five cases of uses of SRS required may increase the risk of acute and late side
in challenging and controversial ways. effects and dose to the surrounding tissue, and risk
of radiation necrosis [4]. A dose too low may be
ineffective. Metastasis to the cerebellum has been
Case 1: Large Metastases reported as a negative prognostic factor for survival
[5, 6]. In addition, the risk of acute decompensation
We first present the case of a 62-year-old woman from obstructive hydrocephalus with compression
with locally advanced colorectal cancer diag- of the 4th ventricle must be weighed when manag-
nosed 7 years previously who presented with ing posterior fossa masses [7, 8]. Treatment options
include (1) resection followed by postoperative ste-
J. K. Shah · D. Kondziolka (*) reotactic radiosurgery/radiotherapy, (2) stereotactic
Department of Neurosurgery, New York University radiosurgery only, or (3) stereotactic radiosurgery
Langone Health, New York, NY, USA followed by resection.
e-mail: Douglas.Kondziolka@nyumc.org

https://doi.org/10.1007/978-3-030-42958-4_24
344 J. K. Shah and D. Kondziolka
Fig. 24.1 Large bilateral cerebellar hemisphere colorec- effacement and surrounding edema. (b) One-month fol-
tal cancer metastases treated with stereotactic radiosur- low-up post-contrast MRI demonstrating major regres-
gery monotherapy. (a) Pre-treatment post-contrast (above) sion of both lesions. (c) Five-month post-contrast MRI
and T2 (below) MRI, demonstrating 4th ventricular demonstrating further regression
24 Challenges and Controversies in Stereotactic Radiosurgery 345
Resection followed by radiotherapy has been a regression of tumors (Fig. 24.1b). Her balance
mainstay of treatment for single brain metastasis improved. Five-month follow-up demonstrated
[9]. For colorectal cancer specifically, Wronski further regression of the tumors (Fig. 24.1c). At
et al. studied a series of 73 patients who under- her most recent follow-up, 22 months after radio-
went brain metastasis resection and found that surgery showed no regrowth and no new tumors.
resection may increase survival of these patients
[5]. In addition, for metastasis to the cerebellum,
Rajendra et al. concluded that surgical treatment Case 2: Melanoma
appeared beneficial, provided the absence of
postoperative complications [8, 10]. We present the case of a 68-year-old man who
More recently, evidence for stereotactic radio- originally presented with abdominal pain and was
surgery monotherapy for cerebellar metastases is found to have lung, adrenal, and gastric masses
increasing. Hill et al. studied 100 patients with found to be S100 positive and HMB45 positive
155 cerebellar metastases and found that SRS is melanoma metastases on biopsy. Staging MRI
generally safe and effective [11]. In this study, demonstrated six intracranial metastases—right
only 10% of patients had undergone pre-SRS and left cerebellar, right temporal, right occipi-
resection, with resection prior to SRS associ- tal, and right frontal (Fig. 24.2a). He had no focal
ated with increased long-term risk for subsequent deficits on examination.
hydrocephalus.
Alternatively, a new treatment paradigm for the
management of cerebral metastases has emerged. Current Treatment Paradigm
Preoperative neoadjuvant SRS (NaSRS) has been
described in two studies, the rationale for which Half of melanoma patients will develop brain
is clearer target delineation and the theoreti- metastases during their treatment course, with
cal reduction of intraoperative dissemination of prior reports noting a poor median overall survival
tumor cells. Asher et al. reported on 47 consecu- of 4–5 months, but recent data is more encourag-
tive patients treated with NaSRS with a median ing [14]. For this patient, the diagnosis-specific
dose of 14 Gy to 80% isodose followed by surgi- Graded Prognostic Assessment (DS-GPA) pre-
cal resection a median of 1 day later and found dicted a survival of 4.7 months [15]. Traditionally,
local control of 97.8%, 85.6%, and 71.8% at 6, treatment options included SRS, resection, or
12, and 24 months, respectively [12]. Patel et al. a combination thereof. Although resection fol-
reported 12 patients treated with 16 Gy followed lowed by SRS or SRS alone has been a mainstay
by resection a median of 1 day later, with poste- of management of brain metastases, this treat-
rior fossa tumors comprising 75% of their cohort ment strategy is optimized for local control of
[13]. Local control rates at 6 and 12 months were few metastases. Addition of WBRT for preven-
81.8% and 49.1%, respectively. tion of distant intracranial metastasis results in
cognitive deficits that are unacceptable to many
patients and may not be effective in melanoma.
Case Outcome An alternative treatment option that has been
emerging is the use of targeted (BRAF mutation)
After reviewing her options, the patient wished to therapy or immunotherapy. Immune checkpoint
avoid open surgery and elected for treatment with inhibitors have generated promising results in
primary SRS. The tumor margin dose was 17 Gy some patients, with food and drug administration
at the 50% isodose line to the larger left cerebel- (FDA) approval of ipilimumab followed by pem-
lar tumor and 18 Gy to the 50% isodose line for brolizumab and nivolumab in 2014. However,
the right cerebellar tumor. Her dexamethasone immune-related adverse events are a known side
dose was tapered over the course of 1 month. effect due to the mechanism of releasing inhi-
One-month follow-up demonstrated major bition on T cell proliferation and activity. The
Fig. 24.2 Multiple melanoma brain metastases treated occipital, right temporal, and right cerebellar (left to
with first-line stereotactic radiosurgery and systemic right). (b) Sixteen months post-treatment post-contrast T1
immunotherapy. (a) Pre-treatment post-contrast T1 MRI MRI demonstrating regressed appearance of tumors
demonstrating the three most significant lesions: right
value may be in patients with very small tumors, combination immunotherapy with ipilimumab
but even in these we see treatment failures. and nivolumab. He developed hypopituitarism
Specifically in the case of ipilimumab, incidence and started hormone replacement therapy. Four
of immune hypophysitis has been reported to be months after his initial GKRS, the treated tumors
as high as 17%, with frequent hormonal deficien- regressed, but a new punctate metastasis in the
cies at diagnosis [16]. Increased doses of immune right postcentral gyrus was found on imaging.
checkpoint inhibitors are also associated with He underwent GKRS to that lesion. New intra-
an increased risk of hypophysitis, further limit- cranial metastases were subsequently found and
ing the clinical potential of immune checkpoint treated with GKRS an additional two times, for a
inhibitor monotherapy [17]. total of four procedures, in addition to a specific
dosing regimen of ipilimumab and nivolumab.
Twenty-one months after initial Gamma Knife
Case Outcome and Discussion radiosurgery, the patient remains without neu-
rological deficits. Recent imaging shows a
After staging, the patient underwent Gamma regressed appearance of all tumors (Fig. 24.2b).
Knife radiosurgery (GKRS) for his six intracra- A recent unpublished study at our institution
nial metastases. Afterwards, he was started on revealed among 123 multiple melanoma patients
combining first-line SRS with immunotherapy of preventing local and distal intracranial relapse
had a median OS of 17.5 mo (31% 3-year OS). in addition to treating undetected lesions, the
Furthermore, among BRAF mutated patients, presence of which increases in probability with
median OS was 31.0 mo (47% 3-year OS) in the increasing number of detected lesions. WBRT
setting of combined immune checkpoint inhibi- has significant side effects, including hair loss,
tion and BRAF targeted therapy. Combination fatigue, nausea, and most significantly, neuro-
SRS with immunotherapy is part of a trend of cognitive decline [3]. Nevertheless, it remains an
improving survival in patients with multiple important therapy for some patients.
brain metastases. The maximum number of lesions suit-
able for SRS alone has been controversial but
increasing steadily. Early studies focused on
Case 3: Multiple Metastases comparing the addition of SRS to WBRT com-
pared to WBRT alone, which demonstrated
We present a 50-year-old woman with ER/ improved local disease control in the WBRT
PR/Her2+ metastatic breast cancer diagnosed plus SRS group compared with the WBRT
4 years prior to first brain metastasis. She under- alone group (92% vs 0% at 1 year) and non-
went lumpectomy, mastectomies, radiation, statistically significant trends toward improved
and chemotherapy with tamoxifen, letrozole, survival in WBRT plus SRS group (11 months
trastuzumab, and vinorelbine. Her initial brain vs 7.5 months), (p = 0.22) [19].
MRI demonstrated miliary metastases, and she A subsequent randomized control trial com-
underwent whole-brain radiation therapy with pared SRS alone to WBRT plus SRS in patients
an excellent early response on serial scans for with one to four brain metastases and found no
6 months, at which time recurrence of 26 tumors difference in overall survival [20]. However, the
was noted (Fig. 24.3). Her only neurologic defi- SRS alone group had increased rates of distal
cit was a left facial weakness, which was the metastases. Comparable overall survival in SRS
result of a left parotid metastasis resection. She alone versus SRS plus WBRT was recapitulated
remained normal in cognition and continued to in two more studies by different groups [21, 22].
work and drive. Since then, SRS, while withholding WBRT in
patients with one to four metastases with con-
trolled systemic and primary disease, has become
Current Treatment Paradigm the preferred treatment [18].
The efficacy of GKRS in up to 10 metastases
Whole-brain radiation therapy (WBRT) has been was studied in a prospective multi-institutional
the primary treatment for multiple intracranial trial, Japanese Leksell Gamma Knife (JLGK)
metastases for decades [18], with the rationale 0901 [23]. Included were 1194 patients with
Fig. 24.3 Multiple breast cancer metastasis treated with multiple rounds of stereotactic radiosurgery. Pre-treatment
post-contrast T1 MRI demonstrating multiple small lesions throughout different axial slices
newly diagnosed 1–10 brain metastases, KPS vival were melanoma primary, active systemic
>70, and absence of cerebrospinal fluid (CSF)dis- disease, and higher RPA class.
semination, with the primary endpoint of overall In summary, no randomized controlled trials
median survival. The purpose was to study non- exist for greater than four metastases, and current
inferiority of treatment of 5–10 brain metastases outcome data indicates that SRS may be used in
compared to 2–4 metastases as measured by sur- patients with >10 brain metastases, specifically
vival. In this study, overall median survival after those with controlled primary cancer, absence of
SRS was 12 months for all patients included. The systemic disease, and good KPS. It can be impor-
median survival was 13.9 months in patients with tant as initial targeted brain tumor care prior to
a single lesion, and a similar 10.8 months in those systemic cancer therapy for which longer surviv-
with 2–4 lesions and 10.8 months in those with als may be expected.
5–10 lesions. Local recurrence rates were similar
throughout; however, risk of new lesion occur-
rence was significantly lower in single lesion Case Outcome
group compared with other groups (2–4 and 5–10
lesions) (p < 0.0001). The patient first underwent GKRS for 10
A follow-up study to the JLGK trial found tumors. A conformal radiosurgery plan was cre-
total tumor volume to correlate with overall sur- ated to give these tumors a margin dose of 16
vival, with longer survival time in patients with Gy, keeping the surrounding brain at a low dose.
≤15 cm3 total tumor volume than those with One week later, 16 more tumors underwent
volume >15 cm3 (p < 0.0001). Other significant GKRS. All tumors regressed initially; how-
prognostic factors for poor survival were MRI ever, 10 months later 16 new small tumors were
evidence of CSF dissemination and KPS <70 detected on imaging. She underwent GKRS for
[24]. Thus, a new paradigm of prognostication by these tumors, followed by two more sessions
overall tumor volume rather than number of brain for multiple new small tumors. Over the next
metastases has emerged. 3 years, she continued to have multiple new
Use of SRS in greater than 10 metastases tumors which were treated with GKRS. She
was specifically studied in a retrospective trial entered a clinical trial and stabilized for
by Kim et al. [25]. This small study included 18 months, but then additional radiosurgery was
26 patients with 10 or more intracranial metas- required. Over 100 tumors were targeted over 6
tases who underwent GKRS. All patients had years following WBRT.
KPS >70, mean age was 55, and mean cumula-
tive tumor volume was 10.3 cm3. Overall median
survival was 34 weeks, with 79.5% local control Case 4: Radiation Injury
rate, and control of all lesions at 54% 6 months
post-GKRS. Synchronous onset in non-small cell We present the case of a 75-year-old woman
lung cancer, high KPS (≥80), and controlled pri- diagnosed with endometrial adenocarcinoma,
mary disease were found to be favorable prog- who underwent total abdominal hysterectomy/
nostic factors in their analysis. bilateral salpingo-oophorectomy and lymph node
However, data for SRS alone for multiple dissection. She underwent four cycles of carbo-
metastases has not been completely favor- platin and 5040 Gy of adjuvant local radiation.
able. Grandhi et al. retrospectively reviewed Two months later, she presented with imbal-
patients with 10 or more metastases managed ance, widened gait, nausea/vomiting, malaise,
with Gamma Knife surgery and found that fac- and hyponatremia. Imaging revealed a 2 cm left
tors associated with poor survival included medial frontal tumor (Fig. 24.4a). Given the sur-
greater than 14 metastases (which decreased rounding edema, the lesion was considered to be
median survival from 6 months to 3 months) a metastasis, and the patient elected to proceed
[26]. Other factors correlating with poor sur- with SRS and declined surgery or a biopsy. She
Fig. 24.4 Left medial frontal endometrial cancer metas- of the tumor and more edema. (c) Ten-month follow-up
tasis treated with first-line SRS. (a) Pre-treatment post- MRI demonstrating regression of lesion. (d) Sixteen-
contrast T1 (left) and T2 (right) MRI demonstrating lesion month follow-up demonstrating further regression on
with surrounding edema. (b) Three-month follow-up MRI post-contrast T1 (left) and minimal edema on FLAIR
demonstrating further peripheral increased enhancement (right)
underwent GKRS. The dose plan consisted of Case Outcome

10 isocenters using 8- and 4-mm collimators.
The tumor margin dose was 17.5 Gy at the 50% She was initially observed and there was neuro-
isodose line. She was placed on a 3-week dexa- logical improvement with respect to cognition.
methasone taper starting at 4 mg TID. One-month follow-up MRI demonstrated slight
Three-month follow-up imaging after GKRS improvement without intervention. She contin-
demonstrated further peripheral increased ued to be observed and continued to experience
enhancement of the tumor and more edema improvement clinically and radiographically,
(Fig. 24.4b). No other new tumors were seen. The with a salvage plan of high-dose steroids or beva-
maximum tumor measurements increased from cizumab therapy should her imaging or clinical
20, to 30, to 35 mm on serial scans. Although status worsen. Ten-month follow-up imaging after
radiation-associated expansion was likely, these initial GKRS demonstrated significant decrease
imaging findings raised the concern for tumor in the size and surrounding edema of the lesion
growth, or reconsideration of the diagnosis which (Fig. 24.4c). At last follow-up, 20 months after
presumed was metastatic, given its appearance GKRS, the imaging remained stable (Fig. 24.4d).
and the active cancer history. A separate pathol-
ogy such as a malignant glioma was also possible.
Case 5: Tumor Bed Radiosurgery
Current Treatment Paradigm A 47-year-old man was diagnosed with mela-

noma and treated with 1 month of interferon.
Treatment options at this point included beva- A serial positron emission tomography (PET)
cizumab therapy, biopsy, biopsy and laser inter- revealed a lesion in the brain and subsequent MRI
stitial tumor therapy, and continued observation demonstrated it to be a 21 × 17 × 22 mm left pari-
with use of corticosteroids. Surgical resection etal cortex with surrounding edema (Fig. 24.5a).
was felt to be hazardous in this location. The patient elected to proceed with resection and
Fig. 24.5 Left occipital melanoma metastasis treated onstrating gross total resection. (c) Last follow-up
with resection followed by GKRS. (a) Pre-treatment post- 59 months after initial imaging, demonstrating no residual
contrast T1 (left) and T2 (right) MRI demonstrating lesion tumor on T1 post-contrast (left) and minimal evidence of
and surrounding edema. (b) Postoperative imaging dem- parenchymal alteration on T2 (right)
he underwent an uncomplicated gross total resec- Current Treatment Paradigm

tion (Fig. 24.5b). Two weeks later, he underwent and Outcome
GKRS with a tumor bed margin dose of 16 Gy.
Subsequent serial imaging every 3 months ini- As in Case 2 above, mean survival after diagno-
tially demonstrated a regressed appearance of the sis of melanoma brain metastases is 4–5 months.
tumor and he did not require systemic melanoma This patient exceeded this outcome and follow-
treatment. At last follow-up, 59 months after ini- up imaging continues to demonstrate no residual.
tial adjuvant GKRS, no recurrence and no new
tumors were visible (Fig. 24.5c). The patient has
References
no neurological complaints. Local tumor man-
agement and avoidance of whole-brain radiation
1. Gavrilovic IT, Posner JB. Brain metastases: epi-
therapy have fostered continued normal neuro- demiology and pathophysiology. J Neuro-Oncol.
logical function and full-time employment. 2005;75(1):5–14.
2. Kurtz JM, et al. The palliation of brain metastases in 16. Albarel F, et al. Long-term follow-up of ipilimumab-
a favorable patient population: a randomized clinical induced hypophysitis, a common adverse event
trial by the Radiation Therapy Oncology Group. Int J of the anti-CTLA-4 antibody in melanoma. Eur J
Radiat Oncol Biol Phys. 1981;7(7):891–5. Endocrinol. 2015;172(2):195–204.
3. Tallet AV, et al. Neurocognitive function impairment 17. Faje AJP. Immunotherapy and hypophysitis: clini-

after whole brain radiotherapy for brain metastases: cal presentation, treatment, and biologic insights.
actual assessment. Radiat Oncol. 2012;7:77. Pituitary. 2016;19(1):82–92.
4. Masucci GL. Hypofractionated radiation therapy for 18. Hatiboglu MA, et al. Treatment of high numbers of
large brain metastases. Front Oncol. 2018;8:379. brain metastases with Gamma Knife radiosurgery: a
5. Wronski M, Arbit E. Resection of brain metastases review. Acta Neurochir. 2016;158(4):625–34.
from colorectal carcinoma in 73 patients. Cancer. 19. Kondziolka D, et al. Stereotactic radiosurgery plus
1999;85(8):1677–85. whole brain radiotherapy versus radiotherapy alone
6. Chaichana KL, et al. Factors associated with survival for patients with multiple brain metastases. Int J
and recurrence for patients undergoing surgery of cer- Radiat Oncol Biol Phys. 1999;45(2):427–34.
ebellar metastases. Neurol Res. 2014;36(1):13–25. 20. Aoyama H, et al. Stereotactic radiosurgery plus whole-
7. Fadul C, Misulis KE, Wiley RG. Cerebellar metas- brain radiation therapy vs stereotactic radiosurgery
tases: diagnostic and management considerations. J alone for treatment of brain metastases: a randomized
Clin Oncol. 1987;5(7):1107–15. controlled trial. JAMA. 2006;295(21):2483–91.
8. Ghods AJ, Munoz L, Byrne R. Surgical treatment of 21. Chang EL, et al. Neurocognition in patients with brain
cerebellar metastases. Surg Neurol Int. 2011;2:159. metastases treated with radiosurgery or radiosurgery
9. Patchell RA, et al. A randomized trial of surgery in the plus whole-brain irradiation: a randomised controlled
treatment of single metastases to the brain. N Engl J trial. Lancet Oncol. 2009;10(11):1037–44.
Med. 1990;322(8):494–500. 22. Kocher M, et al. Adjuvant whole-brain radiotherapy
10. Rajendra T, et al. Results of surgical treatment for cere- versus observation after radiosurgery or surgical
bral metastases. J Clin Neurosci. 2003;10(2):190–4. resection of one to three cerebral metastases: results
11. Hill C, Trifiletti DM, Romano KD, Showalter TN, of the EORTC 22952-26001 study. J Clin Oncol.
Sheehan JP. Stereotactic radiosurgery for cerebellar 2011;29(2):134–41.
metastases and the risk of obstructive hydrocephalus. 23. Yamamoto M, et al. Stereotactic radiosurgery for

Appl Rad Oncol. 2017;6(1):17–23. patients with multiple brain metastases (JLGK0901):
12. Asher AL, et al. A new treatment paradigm: neoad- a multi-institutional prospective observational study.
juvant radiosurgery before surgical resection of brain Lancet Oncol. 2014;15(4):387–95.
metastases with analysis of local tumor recurrence. 24. Serizawa T, et al. Analysis of 2000 cases treated with
Int J Radiat Oncol Biol Phys. 2014;88(4):899–906. gamma knife surgery: validating eligibility criteria for
13. Patel AR, et al. Neoadjuvant stereotactic radiosur- a prospective multi-institutional study of stereotactic
gery before surgical resection of cerebral metastases. radiosurgery alone for treatment of patients with 1-10
World Neurosurg. 2018;120:e480. brain metastases (JLGK0901) in Japan. J Radiosurg
14. Davies MA, et al. Prognostic factors for survival in SBRT. 2012;2(1):19–27.
melanoma patients with brain metastases. Cancer. 25. Kim CH, et al. Gamma knife radiosurgery for ten
2011;117(8):1687–96. or more brain metastases. J Korean Neurosurg Soc.
15. Sperduto PW, et al. Summary report on the graded 2008;44(6):358–63.
prognostic assessment: an accurate and facile 26. Grandhi R, et al. Stereotactic radiosurgery using the
diagnosis-specific tool to estimate survival for patients Leksell Gamma Knife Perfexion unit in the manage-
with brain metastases. J Clin Oncol. 2012;30(4): ment of patients with 10 or more brain metastases. J
419–25. Neurosurg. 2012;117(2):237–45.
Synergy of Immunotherapy
and Radiosurgery
25
Andrew G. Brandmaier, Rohan Ramakrishna,
and Silvia C. Formenti
T Cells in Immunity emerge in the host to respond against certain anti-

gens and ignore others [2]. During development,
The immune system encompasses a broad spec- each T cell clone is individually educated to rec-
trum of cells from the hematopoietic lineage. ognize a peptide: major histocompatibility com-
Each cell type contributes specialized functions plex (MHC), but emerging clones that respond
which together perform key steps in host immu- to self-antigens are directed to undergo apop-
nity: sensing danger stimuli, secreting cytokines tosis. Despite its broad applicability, SNS does
that recruit and activate effector cells, display- not completely explain key phenomena such as
ing peptide fragments, detecting antigens, and tumor immunity. The “Danger Model” is a newer
engulfing and lysing targets. A guiding hallmark theory developed by Matzinger, and it proposes
of immune activity was defined by immunolo- that immune activation is controlled by the con-
gists Burnet and Medawar, who proposed the text of innate immune signals in the tissue envi-
Self-Nonself Model (SNS), which posited that ronment that are generated by perturbation of
immune cells cooperate to recognize and attack homeostasis. Viewing immune responses through
foreign (non-self) antigens [1]. This framework the lens of “danger” has provided a logical way to
rationalized how microbes and infected or trans- interpret fundamental principles of tumor immu-
formed cells harbor aberrant “Nonself” protein nology [3]. These include the requirement of
antigens, which are recognized and subjected to co-stimulation signaling to activate naïve T cells
immune attack, whereas regular normal “Self” and the regulatory role of checkpoint molecules,
host tissues are spared from immune targeting. which attenuate T cell activation and prolifera-
SNS helps explain how B and T cell repertoires tion. In the tumor microenvironment, homeo-
static regulatory processes and suppressive
signals maintain a balance that suppresses T cell
function. These mechanisms create a high thresh-
A. G. Brandmaier · S. C. Formenti (*) old for the immune system to activate effective
Department of Radiation Oncology, NewYork- anti-tumor responses.
Presbyterian Hospital, Weill Cornell Medicine, αβ T cells play a central role in the adaptive
New York, NY, USA
immune system and are also the key constituent
e-mail: formenti@med.cornell.edu
of anti-tumor immunity. Individual T cell clones
R. Ramakrishna
express a unique T cell receptor (TCR) dimer on
Department of Neurological Surgery,
NewYork-Presbyterian Hospital, Weill Cornell the plasma membrane. The receptors contain an
Medicine, New York, NY, USA immunoglobulin-like subunit with a unique

https://doi.org/10.1007/978-3-030-42958-4_25
356 A. G. Brandmaier et al.
variable sequence at the surface, which scans and enters a state of reduced proliferation and
MHC:peptide complexes expressed on adjacent diminished IL-2 secretion [6]. The Danger Model
antigen-presenting cells (APCs). MHC class I predicts that immune cells are activated by alarm
molecules are expressed on nearly all cell types, signals from pathogens or distressed cells. Toll-
and these complexes are recognized by CD8+ T like receptors (TLRs) represent a prominent fam-
cells. MHC class II molecules are primarily pre- ily of innate danger sensors expressed by APCs,
sented by professional APCs, which include den- and activation through these receptors induces
dritic cells (DCs) and macrophages, and they are their maturation. When mature, professional
recognized by CD4+ T helper cells. When a TCR APCs, such as dendritic cells, increase surface
is engaged by an MHC:peptide complex with expression of B7-1 and B7-2; these molecules
sufficient affinity to bind, activation signals are co-stimulate naïve T cells by binding CD28 [7].
transmitted from the TCR through downstream TCR stimulation coupled with CD28 co-
signaling cascades that activate T cell effector stimulation activates a naïve T cell to adopt its
functions and clonal proliferation. Class I anti- mature, effector status; its TCR subunits reorga-
gens stimulate CD8+ T cells to produce TNF-α nize at the plasma membrane to respond to future
and IFN-γ and to secrete cytotoxic granules that antigen encounters at a lower threshold. The cell
release perforin and granzyme, which cause lysis also expresses CD25 to enable rapid proliferation
of the targeted cell [4]. Class II antigens stimulate in response to IL-2. With these changes, the acti-
CD4+ T cells, which mediate helper activities vated T cell can clonally expand and effectively
including release of supportive cytokines and attack antigenic targets in the periphery. It also
expression of CD40 ligand, which binds CD40 spawns effector and memory daughter cells to
on adjacent APCs and promotes their activation. expand the reach and longevity of cells recogniz-
Type I helper cells (Th1) can promote anti-tumor ing the antigen in question.
activity when activated to secrete IFN-γ, which is Mature DCs perform key functions necessary
a strong paracrine signal that promotes surround- for T cell priming. They phagocytose distressed
ing cells to present class I and class II MHC com- or dead cells, process and present antigens for T
plexes. For tumor cells, this increases recognition cell recognition, and release stimulatory cyto-
by cytotoxic CD8+ T cells [5]. Additionally, Th1 kines [8]. In tumors, DCs are activated if they
cells release IL-2, which is a growth factor that encounter danger-associated molecular patterns
promotes survival and proliferation of surround- (DAMPS) which bind their TLRs. In particular, a
ing T cells. Collectively, CD8+ and CD4+ T cells subpopulation of BATF-3-dependent CD103+
directly attack tumor cells and produce immuno- DCs have been found to efficiently engulf and
stimulatory signals that promote anti-tumor process tumor cells and vesicles and transport
activity of other immune populations. In the fol- this cargo to tumor draining lymph nodes [9].
lowing sections, we will describe some of the Upon arrival, the DCs present MHC class I and II
mechanisms cancers utilize to escape immune peptide antigens from the tumor and prime anti-
recognition and rejection. tumor T cells. The presence of sufficient DAMPs
in the tumor microenvironment is critical for this
initial step in generating an anti-tumor response.
APC Activation of T Cells Multiple innate regulatory signals are utilized
by the immune system to prevent overactive or
T cell clones emerging from the thymus are in a redundant T cell responses and maintain homeo-
naïve phenotypic state prior to antigen encounter. stasis. This includes checkpoint signals, which
The context of a T cell’s initial recognition of are transmitted through an array of receptors to
cognate antigen is critical for its long-term fate in control the duration and amplitude of T cell activ-
the immune system. When a naïve T cell’s TCR ity. The two most prominent checkpoint targets
binds MHC:peptide antigen without co-with a proven efficacy in immunotherapy are
stimulation, the cell is induced to undergo anergy CTLA-4 and PD-1. T cells upregulate surface
25 Synergy of Immunotherapy and Radiosurgery 357
expression of CTLA-4 following stimulation of tumors, microglia function can be subverted to a

their TCR. This provides a negative signaling tolerogenic phenotype, similar to M2 macro-
axis, wherein the B7-1 and B7-2 costimulatory phages. Peripheral macrophages and monocytes
molecules can transmit regulatory signals by are often recruited to brain tumors and can func-
binding CTLA-4. In mouse models, a germline tion together with altered microglia to release
knockout of CTLA-4 leads to fatal autoimmunity tumor-promoting cytokines and growth factors.
associated with generalized T cell activation, In doing so, they help vascularize the tumor tis-
illustrating the regulatory power of this check- sue and promote tumor cell growth and invasion
point molecule in suppressing T cells [10]. A sec- [14].
ond checkpoint pathway is mediated by the The blood-brain barrier stringently regulates
receptor, PD-1. T cells upregulate PD-1 expres- passage of substrates and cells into the brain from
sion following activation, and ligand binding the vasculature. It is comprised of tight junctions
regulates tissue inflammation, which protects between endothelial cells and support from astro-
against autoimmunity. Its ligands, PD-L1 and cytes and pericytes. This tight barrier hinders
PD-L2, are expressed on tumor cells and regula- immune cell trafficking, and therefore the brain
tory immune cells. When PD-1 is bound, the T has been sometimes characterized as an “immune
cell downregulates kinases involved in activation privileged” site due to the limited cross-talk
and acquires an “exhausted” phenotype with lim- between its tissue epitopes and the APCs and lym-
ited function and potentially apoptosis [11]. Mice phocytes of the immune system, though this inter-
with a genetic knockout of PD-1 demonstrate pretation has been challenged. Three sites of
tissue-specific autoimmunity, though less severe immune cell access into the brain have been pro-
than CTLA-4 [12]. Overall, the immune check- posed: choroid plexus, leptomeningeal vessels,
point molecules maintain homeostasis by damp- and parenchymal vessels [15]. Metastatic tumors
ening immune activation. They also are utilized exhibit heterogenous vascular regions with selec-
in the tumor microenvironment to create a barrier tive disruption of the blood-brain barrier, which is
to anti-tumor immunity. due, in part, to their suppression of molecular sig-
naling pathways of CNS endothelial cells [16].
Nevertheless, modeling of the “blood tumor bar-
Immunity in the Brain rier” has found that chemotherapeutic agents are
significantly excluded from brain tumors relative
The central nervous system (CNS) has a unique to non-CNS tissues [17]. Tumor-directed radiation
landscape compared to other tissue types with its has been found to disrupt the blood-brain barrier. It
own resident immune cells, a distinct lymph is not clear how or whether more extensive altera-
drainage pathway, and restricted vascular perme- tions of the blood-brain barrier in tumors would
ability maintained by the blood-brain barrier. impact the systemic anti-tumor immune response.
Microglia reside exclusively within the CNS, and The origins and extent of tumor immunosur-
they perform similar functions to macrophages, veillance in the CNS are not fully defined.
including processing and presentation of antigens Preclinical evidence has shown that APCs are
and expression of MHC class II complexes. At present in the brain parenchyma and ultimately
baseline, microglial cells maintain immune drain into cervical nodes of the neck, where they
homeostasis; they also stimulate and remove var- can present tumor antigens to circulating T cells
ious neighboring cells for maintenance of the to generate a systemic immune response.
microenvironment. Innate immune signals can Recently, discovery of draining lymphatics
activate microglia and turn on their antigen-pre- along the dura has provided more insight into
senting and immune-priming functions. Their this pathway. Intraparenchymal cerebrospinal
persistent activation has been associated with fluid (CSF) carrying cells and antigens from the
destructive inflammation and neurodegenerative brain tissue flow out to the subarachnoid reser-
diseases [13]. In metastatic and primary brain voirs of CSF [18]. Enriched by these substrates,
Fig. 25.1 Histologic
evaluation has revealed Superior sagittal sinus
the presence of
lymphatic vessels in the
meninges of the brain.
They line the dural
sinuses and serve as an Inferior sagittal sinus
interface with
cerebrospinal fluid
carrying cells and
soluble particles from
the brain parenchyma. Confluence of sinuses
The brain lymphatics are
a channel for immune
Transverse sinus
cells and fluids to drain
to the deep cervical
nodes where they can
interact with the
peripheral immune
system
Internal jugular vein

Deep cervical lymph nodes
the CSF diffuses into lymphatic vessels, which Tumor-associated antigens (TAAs) are proteins
run in parallel along the dura. The lymph fluid associated with cell function that may be recog-
follows this path along the sagittal sinus, which nized by T cells when expressed at aberrant lev-
ultimately reaches deep cervical lymph nodes to els. In melanoma, several substrates involved
interface with the peripheral immune system with pigment synthesis are TAAs, such as
(Fig. 25.1). Overall, the brain has a unique MART-1 and GP100. In breast cancer, HER2/
immune microenvironment. Its resident immune neu is a TAA. Germline antigens are proteins
cell population, blood-brain barrier, and distinct normally restricted to the gonads but ectopically
lymph drainage channels add to the complexity expressed by tumor cells. MAGE-A and
of strategically targeting metastatic CNS tumors NY-ESO-1 are well-characterized germline
with immunotherapy. antigens expressed by various cancers. Tumor-
specific antigens, also known as “neoantigens,”
are proteins expressed from nonsynonymous
Tumor Immunosurveillance gene mutations occurring in cancer cells that
result in novel peptide epitopes recognized as
Tumor immunosurveillance is a model of the foreign by lymphocytes. Innovations in bioin-
dynamic interaction between the immune sys- formatics are creating new applications to apply
tem and emerging cancers. It postulates that whole genome sequencing and mass spectrom-
most neoplastic cells are eliminated before they etry data from tumor samples to predict the
proliferate to form tumors. Newly transformed presence of neoantigens and identify corre-
cells possess genetic or cellular aberrations that sponding reactive lymphocytes from the patient
are presented in antigen complexes and recog- [20]. Tumors that contain a high mutational
nized by circulating T cells. Schreiber et al. load, such as in the setting of defective mismatch
defined three main categories of tumor antigens: repair genes, have shown an increased response
tumor-associated antigens, cancer germline to immunotherapy. This may be due to an
antigens, and tumor-specific antigens [19]. increased abundance of neoantigens susceptible
to T cell attack. The ability to analyze tumors class II complexes and CD86 costimulatory mol-
and predict antigenic targets may lead to new ecules [21]. TGF-β promotes expansion of Tregs
opportunities in immunotherapy. and induces differentiation of naïve CD4+ T cells
to Foxp3+ Tregs. It also induces apoptosis of acti-
vated CD8+ T cells, attenuates DC activation, and
Immune Suppression by Tumors directs macrophages toward a suppressor pheno-
type [27]. In summary, the tumor microenviron-
In addition to immunoediting, tumors also acti- ment maintains specific populations of cells and
vate regulatory processes that suppress host anti- produces a profile of cytokines that are potently
tumor immunity. Histologically, the tumor immunosuppressive, establishing a significant
microenvironment contains supporting and regu- barrier to effective anti-tumor immune responses.
latory stromal cells dispersed among the primary
cancer cells. They include fibroblasts, myeloid
cells, and tumor-associated vascular endothe- Radiation Therapy in Cancer
lium. In cancer, these populations converge to
create an immunosuppressive network resem- Radiation biology dogma has traditionally
bling an unhealed wound [21]. They condition attributed anti-tumor effects of radiotherapy to
the microenvironment by secreting growth fac- cytocidal DNA damage. Measurements of
tors and chemokines including vascular endothe- tumor cell sensitivity to radiation, such as sur-
lial growth factor (VEGF), chemokine ligand 2 vival curves generated from clonogenic assays,
(CCL2), and granulocyte-macrophage colony- have traditionally provided a means to model
stimulating factor (GM-CSF), which attract therapeutic efficacy of various dose and frac-
myeloid cells from the periphery that differenti- tionation approaches [28]. This approach inter-
ate into myeloid-derived suppressor cells prets radiotherapy through the lens of cell kill.
(MDSCs) and macrophages; these cells potently However, more modern data has revealed that
suppress APCs and T cells within the tumor [22]. radiation also has a substantial effect on the
Clinically, high levels of tumor infiltrating poly- tumor microenvironment that influences sys-
morphonuclear MDSCs have been associated temic processes. In vivo mouse studies have
with disease progression and worse prognosis in shown that radiation treatment can activate
cancer patients, which illustrates how local anti-tumor immune responses and synergize
immunosuppression favors tumor persistence with immunotherapeutic agents [29]. Radiation
and growth [23]. They produce reactive oxygen releases cell death substrates that activate
species which affect CD8+ T cells by reducing innate immune receptors that promote T cell
levels of the TCR zeta chain and BCL-2, which priming [30]. Furthermore, the production of
increases their proclivity to undergo apoptosis double-stranded DNA breaks and formation of
[24]. MDSC metabolism also suppresses immune micronuclei turn on the type I interferon path-
function, by depleting arginine in the tumor way [31, 32]. These phenomena, which will be
microenvironment, disrupting the function of the elaborated in greater detail later, are established
TCR complex and limiting proliferation of acti- mechanisms by which radiation stimulates
vated T cells [25]. MDSCs also express the tumor immunity, and they substantiate the ben-
enzyme IDO, which catabolizes tryptophan to eficial role of radiotherapy as an adjuvant when
kynurenines. Low tryptophan concentration sen- combined with immunotherapy.
sitizes T cells to apoptosis, and kynurenines
induce Treg cell differentiation [26]. Tolerogenic
DCs also synthesize IDO and metabolize trypto- Immunogenic Cell Death
phan. MDSCs, tumor macrophages, and Tregs all
produce IL-10 and transforming growth factor The contribution of radiation to anti-tumor
(TGF)-β. IL-10 attenuates DC activation and immunity is partly due to how the malignant cells
reduces macrophage expression of both MHC die and the associated signals that are released
into the microenvironment. Zitvogel and Kroemer repeated in immunodeficient mice, bolstering the
reported that various cell death pathways can causal relationship between adaptive immuniza-
produce DAMPS, which are danger signals that tion and protection against tumor growth. Golden
activate innate immune receptors and ultimately et al. evaluated levels of ICD biomarkers pro-
trigger adaptive T cell activation against antigens duced in tumor cell cultures and found that tumor
from the dying cells. This type of cell death is cell radiation results in release of ATP and
categorized as “immunogenic cell death” (ICD) HMGB1 and promotes externalization of plasma
[33]. Strategic induction of ICD is an emerging membrane calreticulin, all in a dose-dependent
therapeutic strategy to elicit activation of the fashion [38]. These studies have shown that radi-
immune system within the tumor. Three impor- ation of tumor cells induces bona fide ICD with
tant DAMPs have been conventionally associated production of the hallmark DAMPs.
with cells undergoing ICD [34]:
1. Calreticulin, an endoplasmic reticulum pro- adiation Upregulates MHC

R
tein, translocates to the extracellular surface and IFN-β
of the plasma membrane. External exposure
of calreticulin corresponds to endoplasmic Radiation also promotes tumor MHC:peptide
reticulum stress and the molecule signals antigen presentation. Reits et al. showed that
CD91 on DCs and macrophages, leading to radiation of human melanoma cultures increased
phagocytosis of the dying cell [35]. the level of tumor MHC class I molecules in a
2. HMGB2 is a chromatin-binding factor that is dose-dependent fashion. Radiation was also
released from the cell. It signals TLR4 on DCs shown to upregulate MHC expression on normal
leading to maturation. Mature DCs upregulate host tissues in vivo [39]. An orthotopic murine
costimulatory molecules such as CD80, effi- glioma model demonstrated that whole-brain
ciently phagocytose dead cells, and cross- radiation upregulated MHC-I expression on
present exogenous antigens [36]. GL261 tumor cells, which improved the efficacy
3. ATP is secreted by the dying cells, which
of concomitant vaccination [40]. Radiation also
recruits professional APCs and stimulates broadens the antigen peptide pool by activating
IL-1β production by DCs, thus promoting mammalian target of rapamycin (mTOR), which
antigen cross-presentation. promotes processing of proteins into peptide
fragments and increases synthesis of new pro-
Altogether, ICD facilitates anti-tumor immu- teins. Moreover, radiation of different types of
nity by producing an array of DAMPs that pro- human tumor cells demonstrably increased the
mote tumor infiltration and activation of APCs, production of cancer-testis antigens, including
engulfment of dead and dying tumor cells, and MAGE-A1 and NY-ESO-1, which lead to activa-
effective cross-presentation and priming of tion of corresponding T cells reactive against
tumor-specific T cells. these epitopes. These findings taken together
The discovery that some cell death pathways show that radiation promotes MHC display with
promote adaptive immunity has led to evaluation a diverse ensemble of peptide antigens.
of various anti-neoplastic therapies for their Radiation of tumors also stimulates an innate
immunogenicity. Among various chemotherapy immune pathway that leads to type I interferon
classes, anthracyclines, cyclophosphamide, and production. Specifically, production of double
oxaliplatin have been demonstrated to induce stranded DNA (dsDNA) breaks followed by cell
ICD in vitro and in vivo [37]. Classic tumor vac- mitosis generates micronuclei that contain chro-
cination/re-challenge assays have also shown that mosome fragments. The cGAS molecule senses
radiation induces ICD. Mice injected with irradi- these dsDNA fragments and activates down-
ated cells fail to grow tumors following a second stream STING, which ultimately leads to tran-
challenge injection. Importantly, this finding was scription of type I interferon [41]. Production of
not recapitulated when the experiment was IFN-β stimulates maturation of DCs with
increased expression of costimulatory mole- increase HIF1-α expression, TGF-β production,

cules and efficient cross-presentation of anti- and activation and release of chemokines that
gens to T cells, which enhances priming of recruit Tregs, MDSCs, and macrophages. These
adaptive immunity. Combinatorial therapy with phenomena have prompted research into regi-
radiation and checkpoint blockade relies on mens combining radiation with immunomodula-
IFN-β activation of Batf-3-dependent DCs to tory drugs to “release the brakes” from these
cross-prime CD8+ T cells and generate effective regulatory signals. TGF-β is a prominent target
anti-tumor responses [32]. An in vivo mouse for this objective; it diminishes cross-priming by
model of breast cancer utilizing combination APCs, reduces activation of CD8+ T cells, and
anti-CTLA-4 and tumor radiotherapy showed increases the prevalence of Tregs. A preclinical
that doses greater than 12–15 Gy per fraction model with 4T1 breast cancer evaluated tumor
attenuated anti-tumor immune responses. radiation and TGF-β blockade, which showed
Mechanistically, higher doses of radiation increased activation of anti-tumor T cells,
induce expression of the nuclease, Trex1, which decreased tumor growth and metastases, and
degrades cytosolic dsDNA and thereby removes improved survival [44]. This approach was incor-
the immune signal for activation of the cGAS- porated in a clinical trial for metastatic breast
STING pathway [31]. This model demonstrated cancer: patients received three fractions of 7.5 Gy
the significance of radiation dose and fraction- to one lesion and either low- or high-dose anti-
ation for immunotherapy applications. TGF-β antibody; receipt of the high dose of
The immune-activating effects of radiation immunotherapy boosted memory CD8+ T cells
have provided a basis for models combining and was associated with improved overall sur-
tumor radiotherapy with immune targeting vival [45]. Chemokine receptor 2 (CCR) is also a
drugs. One of the first preclinical models testing relevant target for combination therapy. Radiation
this concept utilized Flt-3 ligand, a growth factor signals through cGas-STING to increase intratu-
for DCs, together with radiation to treat mice mor levels of chemokines that bind CCR2 and
challenged with Lewis lung carcinoma. The attract MDSCs to the tumor microenvironment.
cohorts that received monotherapy of either Notably, tumor-challenged mice treated with
agent alone showed limited survival because of radiation and CCR2 blockade demonstrated
lung metastases. However, a combination of enhanced CD8+ T cell-mediated tumor rejection
radiation with Flt-3 ligand reduced the number versus cohorts receiving radiation alone [46].
of pulmonary metastases and improved overall Overall, radiation has both stimulatory and sup-
survival [42]. Subsequently, Demaria and pressive effects on the immune system. Strategic
Formenti showed a bona fide abscopal effect molecular targeting of potent regulatory path-
with a combination of radiation and immuno- ways together with radiotherapy can successfully
therapy. In mice with bilateral flank tumors of elicit anti-tumor immunity.
mammary carcinoma (67NR), radiation of one
tumor and Flt-3 ligand treatment reduced the
growth of the contralateral tumor [43]. This Immune Checkpoint Inhibitors
effect was abrogated in athymic mice lacking αβ
T cells, highlighting a synergy of the two thera- Clinical trials have demonstrated the efficacy of
pies for the adaptive immune response. checkpoint inhibitors for several tumor types and
thus established immunotherapy as a mainstream
modality in oncology. New applications continue
I mmune Regulation Induced by to emerge, and at present, most are focused on
Radiation metastatic or locally advanced disease. Allison
and colleagues originally elucidated the T cell
Radiation also activates homeostatic mechanisms regulatory molecule, CTLA-4, and demon-
of the immune system that play an important role strated that antibody blockade (anti-CLTA-4)
in suppressing immune attack. Irradiated tumors unleashed anti-tumor immunity. Mice that were
challenged with characteristically immunogenic needed, and radiotherapy is being investigated for
tumors showed pronounced rejection of the this purpose.
tumors after receipt of anti-CTLA-4 antibody Recent trials have assessed whether combined
[47]. An in vivo study with melanoma demon- checkpoint inhibition may synergistically
strated that anti-CTLA-4 therapy contributed to enhance clinical anti-tumor responses. Check-
tumor immunity by amplifying effector T cell mate 067 was a phase III clinical trial evaluating
function and minimizing Treg cell activity [48]. monotherapy checkpoint inhibition versus a
Notably, a subsequent study using anti-CTLA-4 combination of ipilimumab and nivolumab
to treat the poorly immunogenic melanoma, administration for patients with metastatic mela-
B16-BL6, showed minimal ability to inhibit noma [54]. The cohort receiving combined ther-
tumor growth. It was only when mice received apy had a longer progression-free survival (PFS)
anti-CTLA-4 therapy combined with a vaccina- and higher objective response rate compared to
tion injection of irradiated B16-BL6 cells modi- the cohort receiving ipilimumab alone, albeit at
fied to express GM-CSF that elimination of tumor the price of increased toxicity. Notably, most of
could be achieved in vivo [49]. These results high- the trials utilizing immunotherapy for advanced
lighted that most tumors may require multiple stage cancer have excluded patients with brain
sources of immunogenic stimuli for a therapeutic metastases. However, Margolin and colleagues
response. The preclinical work characterizing reported a phase II study of dual checkpoint inhi-
anti-CTLA-4 ultimately translated to clinical bition with nivolumab and ipilimumab for
applications with ipilimumab. In the first major patients with melanoma brain metastases. Com-
phase III trial with a checkpoint inhibitor, the drug bined therapy resulted in a high response rate of
showed improved overall survival for metastatic 56%. Complete response was seen in 26% of
melanoma, which set the stage for further devel- patients [55]. These impressive results provide a
opment of checkpoint inhibitors in oncology [50]. foundation for exploring checkpoint inhibition
The PD-1 signaling axis is the second T cell for different types of brain metastases.
checkpoint pathway that has been successfully
incorporated for tumor immunotherapy. Several
established human tumors such as lung, ovary, Combination of Checkpoint
colon, and melanoma increase expression of Inhibitors with Radiation
PD-L1 to suppress T cell activity in their microen-
vironment [51]. Immune cells recruited by tumors, Radiation of tumors associated with off-target
including MDSCs, can also express PD-L1 [52]. responses (abscopal effect) has been described in
When surface PD-1 is engaged by the ligand, T a small number of case reports dating back several
cells adopt an exhausted phenotype and display decades. This includes patients with a wide vari-
diminished activity. Anti-PD-1 antibodies block ety of tumor types such as melanoma, renal cell
this signal and help revive tumor infiltrating T carcinoma, and lymphoma [56–58]. The impact
cells, thus facilitating adaptive anti- tumor of radiation on systemic tumor responses may be
responses. PD-1 checkpoint inhibitors have dem- related to anti-tumor immunity. As previously
onstrated success and are approved for use in an described, radiation induces stimulatory immune
increasing number of malignancies, including danger signals that create an in situ vaccine effect
advanced stage melanoma, non-small cell lung in the tumor microenvironment, which helps
cancer (NSCLC) urothelial carcinoma, Hodgkin’s prime adaptive T cell responses. Potential synergy
disease, and head and neck squamous cell cancer, of these effects with checkpoint inhibition has
as well as microsatellite instability-high cancers been extensively explored in preclinical studies
[53]. Though cohorts of cancer patients receiving [59]. Formenti and Demaria showed that mice
checkpoint inhibition have improved clinical out- challenged with 4T1 breast carcinoma derived
comes as a group, most patients do not achieve a minimal benefit from treatment with radiation or
significant response to treatment. New approaches anti-CTLA-4 monotherapy. Yet, combined treat-
to increase the proportion of responders are ment with both agents significantly reduced the
number of lung metastases in recipients and CTLA-4 blockade decreased Tregs. Thus, dual
improved survival [60]. This approach has also checkpoint blockade increased the ratio of CD8/
demonstrated efficacy in an orthotopic glioma Treg cells [62]. Rudqviist et al. also found that
mouse model: combinatorial therapy with anti- CTLA-4 blockade and radiation therapy
PD-1 and a single fraction of 10Gy to the tumor for tumor-challenged mice synergized to expand
resulted in a significant improvement in survival the TCR repertoire within tumor- infiltrating
over either treatment alone [61]. Minn et al. lymphocytes (TIL). Their evaluation identified an
showed that dual checkpoint therapy with anti- increased diversity and number of CDR3 motifs
PD-1 and anti-CTLA-4 in addition to tumor radia- among the population of receptors [63]. The evi-
tion provided complementary, non-redundant dence from these and several other preclinical
immune activation signals. The anti-tumor TCR models have provided a compelling rationale to
repertoire was expanded by radiation. PD-L1 explore combinatorial strategies with radiation
blockade revived exhausted CD8+ T cells, and and checkpoint inhibitors (Fig. 25.2).
8Gy × 3
Anti-PD-1
Anti-CTLA-4
TCR Treg
CD25
Activated CTLA-4
DC Anti-CTLA-4
Tumor CD-80 CD-28
cell T cell
IFN-γ HLA:peptide TCR
PD-1
CD8+ Anti-CTLA-4
T cell Anti-PD-1
CD4+
T cell Tumor
IL-2 PD-L1
PD-1 cell
PD-1
Anti-PD-1
Fig. 25.2 Immunotherapy with anti-CTLA-4 and anti- the anti-tumor T cell repertoire. Clinical trials are explor-
PD-1 monoclonal antibodies activates non-redundant ing paradigms for combining immunotherapy with tumor
mechanisms that promote clonal expansion of T cells and radiation to synergistically activate and expand anti-tumor
revive exhausted effector cells. Tumor radiation enhances T cells that mediate systemic tumor rejection
MHC antigen presentation and increases the diversity of
Results from preclinical data have influenced a hypofractionated course, though no standard
new oncology trials for patients incorporating prescription has emerged. One core question is
synchronous immunotherapy and radiation. Most the comparative efficacy of different doses per
of the findings are limited to small cohort studies fraction of radiotherapy. In preclinical work
or anecdotal case reports. For example, a mela- with B16 melanoma, a single fraction of 20Gy
noma patient who reportedly progressed after activated anti-tumor CD8+ T cells in mice,
receipt of ipilimumab received palliative radia- whereas this response was not seen in a com-
tion in three fractions for a spinal metastasis. parison cohort treated with 5 Gy × 4 fractions
Within 3 months, distant hilar metastases, and [69]. On the other hand, Vanpouille-Box
splenic lesions responded, representing nearly a treated mice bearing two subcutaneous TSA
complete disease regression [64]. Also a phase II breast carcinomas with anti-CTLA-4 and vari-
study treating Merkel cell carcinoma with pem- ous radiation regimens directed only to one
brolizumab reported that two patients who tumor. Cohorts that received 8 Gy × 3 demon-
received palliative radiation following disease strated abscopal tumor response (measured in
progression had subsequent off-target tumor the non-irradiated tumor) and increased sur-
response [65]. Formenti and colleagues recently vival compared to those that received a single
reported the results of a trial for patients with fraction of 20 Gy. In this model, the abscopal
NSCLC who, after failing chemotherapy, went response from radiation diminished as doses
on to receive radiation therapy to a single metas- were escalated above 12 Gy per fraction [31].
tasis and concurrent ipilimumab. Notably, two This trend paralleled dose-dependent induction
previous prospective randomized trials of of Trex-1, an exonuclease that digests cyto-
CTLA-4 blockade with chemotherapy failed to plasmic dsDNA and removes the substrate for
demonstrate significant activity in advanced cGAS/Sting, which attenuates induction of
NSCLC [66, 67]. However, in Formenti’s trial type I interferon.
combining ipilimumab with focal radiotherapy, With no consensus dose established for immu-
31% of the patients achieved disease control, and notherapy applications, clinical trials are utilizing
18% demonstrated an objective response [68]. a variety of radiation prescriptions. Chmura et al.
One patient who achieved a complete response conducted a phase I clinical trial treating meta-
after originally presenting with synchronous lung static solid tumors with pembrolizumab and
cancer and brain metastases was found to have a SBRT doses from 30 to 50 Gy. They reported a
clonal expansion of T cells recognizing a muta- favorable toxicity profile, but the objective
tion within his tumor. This result demonstrated response was only 13.2%, which was similar to
translational success of radiotherapy in inducing the outcome of pembrolizumab alone in an
neo-antigens and converting the tumor into an in- unselected cohort of patients with metastatic dis-
situ vaccine. As ongoing combinatorial trials ease. The median PFS was 3.1 months [70]. In
continue to mature, more sophisticated conclu- comparison, the Netherlands Cancer Institute
sions can be reached regarding the efficacy of reported preliminary phase II results from NSCLC
combining tumor radiation with immune check- patients, who were randomized to pembrolizumab
point inhibitors. alone versus pembrolizumab with a sub-ablative
radiation dose of 8 Gy × 3. The pembrolizumab
alone cohort achieve a 19% response rate, while
ptimal Radiation Parameters
O the cohort receiving combination therapy had a
for Immunotherapy 41% objective response. Also, the median PFS
was 1.8 versus 6.4 months, respectively [71].
The optimal dose and fractionation of radio- These preliminary findings suggest that a dose/
therapy in combinatorial regimens with check- fraction effect may govern the immune activat-
point inhibitors are yet to be determined. ing potential of radiotherapy. Further investiga-
Several cases reported in the literature utilized tion is needed to validate this phenomenon and,
if confirmed, determine whether this is due to brain metastases, a retrospective analysis showed
Trex-1 induction or other signals. that patients receiving anti-PD-L1 and anti-
Modern clinical trials have not yet reported CTLA-4 therapy followed by stereotactic radio-
high-level data for combinatorial regimens with surgery within 4 weeks of checkpoint blockade
checkpoint inhibitors and radiation of brain demonstrated a greater median reduction in lesion
metastases. Standard whole-brain radiation pre- volume compared to patients with a longer sepa-
scriptions include 30 Gy in 10 fractions and ration of treatments. However, this result could be
20 Gy in 5 fractions as palliative options for attributable to patient selection since progression
extensive disease. Stereotactic radiosurgery through ipilumimab may correspond to more
(SRS) using a single-fraction ablative dose has aggressive metastatic disease [75]. An unplanned
demonstrated excellent local control for patients analysis of the Pacific Trial for NSCLC found that
with a limited number and size of brain metasta- patients who received durvalumab (anti-PD-1)
ses. SRS also has superior preservation of long- after responding to platinum-based chemo-radia-
term cognition compared to whole-brain tion had improved PFS. The finding was particu-
radiation. Furthermore, Knisely and colleagues larly significant if checkpoint blockade was
reported findings that bolstered the prospect of administered within 2 weeks from completion of
combination SRS and checkpoint inhibition. In a chemoradiation [76]. Also, Chiang and colleagues
retrospective analysis of cases of melanoma reported retrospective data of melanoma patients
brain metastases, they showed that the cohort of with brain metastases who were treated with SRS
patients who received ipilimumab in addition to and immune checkpoint inhibition. Administration
SRS had an overall survival of 21.4 months ver- of immune checkpoint therapy within 4 weeks of
sus 4.9 months for patients who received SRS SRS resulted in greater reduction in tumor size
alone [72], a significant difference even if the compared with patients who received treatment
retrospective nature of the study likely reflects that was not concurrent [77]. Going forward,
patient selection. Additionally, hypofractionated results from clinical trials that are currently under-
regimens may have comparable efficacy to SRS way will provide a clearer understanding of the
for larger brain tumors >2 cm. A meta-analysis significance of dose/fractionation and sequencing
of 24 trials showed similar 1-year local control to the overall success of therapy.
for patients receiving SRS versus multi-fraction
RT. The most common multi-fractionation regi-
men utilized was 27 Gy in three fractions [73], a Lymph Nodes as an Organ at Risk (OAR)
prescription that aligns well with the preclinical
data from Vanpouille-Box modeling optimal Utilization of radiotherapy for tumor immune
immunogenic doses to induce tumor production activation will elevate the importance of lympho-
of type I interferon. cytes and lymph nodes as organs at risk for treat-
In addition to dose and fractionation, the opti- ment planning. Functional lymph nodes provide
mal sequencing of radiation and immunotherapy an interface for T cells and APCs draining from
continues to be evaluated. Preclinical work com- tumors to interact and receive priming signals for
paring different sequences showed that upfront activation and proliferation. Marciscano and col-
checkpoint blockade with anti-CTLA-4 followed leagues examined the impact of radiation target
by radiotherapy achieved the greatest tumor treat- fields that included tumor-draining lymph nodes
ment efficacy. The study concluded that early in a preclinical model. Mice were challenged
depletion of Tregs facilitated immune priming of with flank tumors and treated with checkpoint
CD8+ T cells when tumors were irradiated [74]. blockade and a single fraction of 12 Gy that
Limited results from currently available trials either included or omitted the regional draining
suggest that overlapping or close sequencing of lymph nodes. The cohort that received radiation
checkpoint blockade with radiotherapy is likely to with a field encompassing their draining lymph
be the most effective approach. For melanoma nodes had a diminished tumor infiltrating lym-
phocytes population and worse survival compared because the brain has a distinct immune profile.
to the cohort where draining lymph nodes were Furthermore, many clinical trials with check-
avoided [78]. A second area of consideration is point inhibitors have excluded such patients.
the impact of fractionated radiation on lympho- Additional data regarding optimal dose, timing,
cytes in the peripheral blood. Ford and colleagues and targeting with radiation is rapidly emerging.
modeled the radiation dose to the circulating pool This data should be incorporated into new clini-
of lymphocytes. In their calculation, a single frac- cal trials for brain metastases to ultimately
tion of 2 Gy would deliver 0.5 Gy to 5% of circu- develop the most effective combinations of ste-
lating cells. Notably, a 30-fraction course would reotactic radiation and immunotherapy.
result in ≥0.5 Gy to 99% of circulating blood
cells. These studies support a strategy of lymph
node sparing and the utilization of hypofraction- References
ated courses of radiation to best protect the host T
cell pool if attempting to stimulate an anti-tumor 1. Ribatti D. Peter Brian Medawar and the discovery
of acquired immunological tolerance. Immunol Lett.
immune response. For radiation of the brain, an 2015;167(2):63–6.
additional consideration could be the anatomic 2. Medzhitov R, Janeway CA Jr. Decoding the patterns
avoidance of the previously described lymphoid of self and nonself by the innate immune system.
drainage network that traces along the sinuses to Science. 2002;296(5566):298–300.
3. Matzinger P. The danger model: a renewed sense of
the cervical nodes. Louvea and colleagues showed self. Science. 2002;296(5566):301–5.
that ablation of meningeal lymphatics reduces T 4. Zhang N, Bevan MJ. CD8(+) T cells: foot soldiers of
cells and inflammatory responses in the brain in a the immune system. Immunity. 2011;35(2):161–8.
model of multiple sclerosis [79]. It has not been 5. Borst J, Ahrends T, Babala N, Melief CJM,
Kastenmuller W. CD4(+) T cell help in cancer immu-
determined how treatment such as whole-brain nology and immunotherapy. Nat Rev Immunol.
radiation impacts the integrity of the brain lym- 2018;18:635.
phatic channels and, furthermore, what impact 6. Appleman LJ, Boussiotis VA. T cell anergy and
this has on anti-tumor T cell responses within the costimulation. Immunol Rev. 2003;192:161–80.
7. Lanzavecchia A, Sallusto F. Dynamics of T lympho-
brain. Future preclinical studies may be needed to cyte responses: intermediates, effectors, and memory
explore how brain radiation specifically impacts cells. Science. 2000;290(5489):92–7.
all of these variables. 8. Akira S, Takeda K, Kaisho T. Toll-like receptors: criti-
cal proteins linking innate and acquired immunity.
Nat Immunol. 2001;2(8):675–80.
9. Sanchez-Paulete AR, Teijeira A, Cueto FJ, Garasa S,
Summary Perez-Gracia JL, Sanchez-Arraez A, et al. Antigen
cross-presentation and T-cell cross-priming in can-
Immunotherapy is transforming the practice of cer immunology and immunotherapy. Ann Oncol.
2017;28(suppl_12):xii74.
oncology and rapidly integrating into main- 10. Waterhouse P, Penninger JM, Timms E, Wakeham A,
stream treatment paradigms. The utility of radio- Shahinian A, Lee KP, et al. Lymphoproliferative dis-
therapy as an adjuvant with immunotherapy is orders with early lethality in mice deficient in Ctla-4.
well established by preclinical data showing how Science. 1995;270(5238):985–8.
11. Zou W, Chen L. Inhibitory B7-family molecules in
tumor radiation releases danger signals that may the tumour microenvironment. Nat Rev Immunol.
convert the irradiated tumor into an in situ vac- 2008;8(6):467–77.
cine. The rarity of abscopal effects confirms the 12. Keir ME, Butte MJ, Freeman GJ, Sharpe AH. PD-1
evidence of the robust immunosuppressive and its ligands in tolerance and immunity. Annu Rev
Immunol. 2008;26:677–704.
microenvironment of established tumors. 13. Li Q, Barres BA. Microglia and macrophages in

Tipping the balance by adding immunomodula- brain homeostasis and disease. Nat Rev Immunol.
tors to local radiotherapy, such as checkpoint 2018;18(4):225–42.
inhibitors, can create a synergistic effect that 14. Wu SY, Watabe K. The roles of microglia/macro-
phages in tumor progression of brain cancer and
promotes therapeutic anti-tumor T cell responses. metastatic disease. Front Biosci (Landmark Ed).
Brain metastases present a unique challenge 2017;22:1805–29.
15. Wilson EH, Weninger W, Hunter CA. Trafficking of 30. Golden EB, Apetoh L. Radiotherapy and immunogenic
immune cells in the central nervous system. J Clin cell death. Semin Radiat Oncol. 2015;25(1):11–7.
Invest. 2010;120(5):1368–79. 31. Vanpouille-Box C, Alard A, Aryankalayil MJ, Sarfraz
16. Tiwary S, Morales JE, Kwiatkowski SC, Lang FF, Y, Diamond JM, Schneider RJ, et al. DNA exonucle-
Rao G, McCarty JH. Metastatic brain tumors disrupt ase Trex1 regulates radiotherapy-induced tumour
the blood-brain barrier and alter lipid metabolism by immunogenicity. Nat Commun. 2017;8:15618.
inhibiting expression of the endothelial cell fatty acid 32. Deng L, Liang H, Xu M, Yang X, Burnette B, Arina A,
transporter Mfsd2a. Sci Rep. 2018;8(1):8267. et al. STING-dependent cytosolic DNA sensing pro-
17. Lockman PR, Mittapalli RK, Taskar KS, Rudraraju V, motes radiation-induced type I interferon-dependent
Gril B, Bohn KA, et al. Heterogeneous blood-tumor antitumor immunity in immunogenic tumors.
barrier permeability determines drug efficacy in Immunity. 2014;41(5):843–52.
experimental brain metastases of breast cancer. Clin 33. Kepp O, Senovilla L, Vitale I, Vacchelli E, Adjemian S,
Cancer Res. 2010;16(23):5664–78. Agostinis P, et al. Consensus guidelines for the detec-
18. Da Mesquita S, Fu Z, Kipnis J. The meningeal lym- tion of immunogenic cell death. Oncoimmunology.
phatic system: a new player in neurophysiology. 2014;3. United States:e955691.
Neuron. 2018;100(2):375–88. 34. Gebremeskel S, Johnston B. Concepts and mecha-
19. Gubin MM, Artyomov MN, Mardis ER, Schreiber nisms underlying chemotherapy induced immu-
RD. Tumor neoantigens: building a framework for nogenic cell death: impact on clinical studies and
personalized cancer immunotherapy. J Clin Invest. considerations for combined therapies. Oncotarget.
2015;125(9):3413–21. 2015;6(39):41600–19.
20. Pasetto A, Gros A, Robbins PF, Deniger DC,
35. Wiersma VR, Michalak M, Abdullah TM, Bremer
Prickett TD, Matus-Nicodemos R, et al. Tumor- and E, Eggleton P. Mechanisms of translocation of ER
neoantigen- reactive T-cell receptors can be identi- chaperones to the cell surface and immunomodula-
fied based on their frequency in fresh tumor. Cancer tory roles in cancer and autoimmunity. Front Oncol.
Immunol Res. 2016;4(9):734–43. 2015;5:7.
21. Joyce JA, Fearon DT. T cell exclusion, immune
36. Pathak SK, Skold AE, Mohanram V, Persson C,

privilege, and the tumor microenvironment. Science. Johansson U, Spetz AL. Activated apoptotic cells
2015;348(6230):74–80. induce dendritic cell maturation via engagement of
22. Kumar V, Patel S, Tcyganov E, Gabrilovich DI. The Toll-like receptor 4 (TLR4), dendritic cell-specific
nature of myeloid-derived suppressor cells in intercellular adhesion molecule 3 (ICAM-3)-grabbing
the tumor microenvironment. Trends Immunol. nonintegrin (DC-SIGN), and beta2 integrins. J Biol
2016;37(3):208–20. Chem. 2012;287(17):13731–42.
23. Weide B, Martens A, Zelba H, Stutz C, Derhovanessian 37.
Zitvogel L, Galluzzi L, Smyth MJ, Kroemer
E, Di Giacomo AM, et al. Myeloid-derived suppres- G. Mechanism of action of conventional and targeted
sor cells predict survival of patients with advanced anticancer therapies: reinstating immunosurveillance.
melanoma: comparison with regulatory T cells and Immunity. 2013;39(1):74–88.
NY-ESO-1- or melan-A-specific T cells. Clin Cancer 38. Golden EB, Frances D, Pellicciotta I, Demaria S,

Res. 2014;20(6):1601–9. Helen Barcellos-Hoff M, Formenti SC. Radiation
24. Ezernitchi AV, Vaknin I, Cohen-Daniel L, Levy
fosters dose-dependent and chemotherapy-induced
O, Manaster E, Halabi A, et al. TCR zeta down- immunogenic cell death. Oncoimmunology.
regulation under chronic inflammation is mediated 2014;3:e28518.
by myeloid suppressor cells differentially distrib- 39. Reits EA, Hodge JW, Herberts CA, Groothuis TA,
uted between various lymphatic organs. J Immunol. Chakraborty M, Wansley EK, et al. Radiation modu-
2006;177(7):4763–72. lates the peptide repertoire, enhances MHC class I
25. Rodriguez PC, Ochoa AC. Arginine regulation by
expression, and induces successful antitumor immu-
myeloid derived suppressor cells and tolerance in notherapy. J Exp Med. 2006;203(5):1259–71.
cancer: mechanisms and therapeutic perspectives. 40. Newcomb EW, Demaria S, Lukyanov Y, Shao Y,

Immunol Rev. 2008;222:180–91. Schnee T, Kawashima N, et al. The combination of
26. Marvel D, Gabrilovich DI. Myeloid-derived suppres- ionizing radiation and peripheral vaccination pro-
sor cells in the tumor microenvironment: expect the duces long-term survival of mice bearing estab-
unexpected. J Clin Invest. 2015;125(9):3356–64. lished invasive GL261 gliomas. Clin Cancer Res.
27.
Travis MA, Sheppard D. TGF-beta activation 2006;12(15):4730–7.
and function in immunity. Annu Rev Immunol. 41. Harding SM, Benci JL, Irianto J, Discher DE, Minn
2014;32:51–82. AJ, Greenberg RA. Mitotic progression following
28. Buch K, Peters T, Nawroth T, Sanger M, Schmidberger DNA damage enables pattern recognition within
H, Langguth P. Determination of cell survival after micronuclei. Nature. 2017;548(7668):466–70.
irradiation via clonogenic assay versus multiple MTT 42. Chakravarty PK, Alfieri A, Thomas EK, Beri V,

assay – a comparative study. Radiat Oncol. 2012;7:1. Tanaka KE, Vikram B, et al. Flt3-ligand administra-
29. Demaria S, Coleman CN, Formenti SC. Radiotherapy: tion after radiation therapy prolongs survival in a
changing the game in immunotherapy. Trends Cancer. murine model of metastatic lung cancer. Cancer Res.
2016;2(6):286–94. 1999;59(24):6028–32.
43. Demaria S, Ng B, Devitt ML, Babb JS, Kawashima N, after extracranial stereotactic radiotherapy in meta-
Liebes L, et al. Ionizing radiation inhibition of distant static renal cell carcinoma. Acta Oncol. 2006;45.
untreated tumors (abscopal effect) is immune medi- England:493–7.
ated. Int J Radiat Oncol Biol Phys. 2004;58(3):862–70. 58. Robin HI, AuBuchon J, Varanasi VR, Weinstein

44. Vanpouille-Box C, Diamond JM, Pilones KA, Zavadil AB. The abscopal effect: demonstration in lympho-
J, Babb JS, Formenti SC, et al. TGFbeta is a master matous involvement of kidneys. Med Pediatr Oncol.
regulator of radiation therapy-induced antitumor 1981;9(5):473–6.
immunity. Cancer Res. 2015;75(11):2232–42. 59. Demaria S, Golden EB, Formenti SC. Role of local
45. Formenti SC, Lee P, Adams S, Goldberg JD, Li X, Xie radiation therapy in cancer immunotherapy. JAMA
MW, et al. Focal irradiation and systemic TGFbeta Oncol. 2015;1(9):1325–32.
blockade in metastatic breast cancer. Clin Cancer Res. 60. Demaria S, Kawashima N, Yang AM, Devitt ML,
2018;24(11):2493–504. Babb JS, Allison JP, et al. Immune-mediated inhibi-
46. Liang H, Deng L, Hou Y, Meng X, Huang X, Rao E, tion of metastases after treatment with local radiation
et al. Host STING-dependent MDSC mobilization and CTLA-4 blockade in a mouse model of breast
drives extrinsic radiation resistance. Nat Commun. cancer. Clin Cancer Res. 2005;11(2 Pt 1):728–34.
2017;8(1):1736. 61. Zeng J, See AP, Phallen J, Jackson CM, Belcaid Z,
47. Leach DR, Krummel MF, Allison JP. Enhancement of Ruzevick J, et al. Anti-PD-1 blockade and stereotac-
antitumor immunity by CTLA-4 blockade. Science. tic radiation produce long-term survival in mice with
1996;271(5256):1734–6. intracranial gliomas. Int J Radiat Oncol Biol Phys.
48. Peggs KS, Quezada SA, Chambers CA, Korman AJ, 2013;86(2):343–9.
Allison JP. Blockade of CTLA-4 on both effector and 62. Twyman-Saint Victor C, Rech AJ, Maity A, Rengan
regulatory T cell compartments contributes to the R, Pauken KE, Stelekati E, et al. Radiation and
antitumor activity of anti-CTLA-4 antibodies. J Exp dual checkpoint blockade activate non-redun-
Med. 2009;206(8):1717–25. dant immune mechanisms in cancer. Nature.
49. van Elsas A, Hurwitz AA, Allison JP. Combination 2015;520(7547):373–7.
immunotherapy of B16 melanoma using anti- 63. Rudqvist NP, Pilones KA, Lhuillier C, Wennerberg
cytotoxic T lymphocyte-associated antigen 4 (CTLA- E, Sidhom JW, Emerson RO, et al. Radiotherapy
4) and granulocyte/macrophage colony-stimulating and CTLA-4 blockade shape the TCR repertoire
factor (GM-CSF)-producing vaccines induces rejec- of tumor-infiltrating T cells. Cancer Immunol Res.
tion of subcutaneous and metastatic tumors accom- 2018;6(2):139–50.
panied by autoimmune depigmentation. J Exp Med. 64. Postow MA, Callahan MK, Barker CA, Yamada Y,
1999;190(3):355–66. Yuan J, Kitano S, et al. Immunologic correlates of the
50. Hodi FS, O’Day SJ, McDermott DF, Weber RW,
abscopal effect in a patient with melanoma. N Engl J
Sosman JA, Haanen JB, et al. Improved survival with Med. 2012;366(10):925–31.
ipilimumab in patients with metastatic melanoma. N 65. Xu MJ, Wu S, Daud AI, Yu SS, Yom SS. In-field and
Engl J Med. 2010;363(8):711–23. abscopal response after short-course radiation therapy
51. Dong H, Strome SE, Salomao DR, Tamura H, Hirano in patients with metastatic Merkel cell carcinoma pro-
F, Flies DB, et al. Tumor-associated B7-H1 promotes gressing on PD-1 checkpoint blockade: a case series.
T-cell apoptosis: a potential mechanism of immune J Immunother Cancer. 2018;6(1):43.
evasion. Nat Med. 2002;8(8):793–800. 66. Govindan R, Szczesna A, Ahn MJ, Schneider CP,
52. Ballbach M, Dannert A, Singh A, Siegmund DM, Gonzalez Mella PF, Barlesi F, et al. Phase III trial of
Handgretinger R, Piali L, et al. Expression of check- ipilimumab combined with paclitaxel and carboplatin
point molecules on myeloid-derived suppressor cells. in advanced squamous non-small-cell lung cancer. J
Immunol Lett. 2017;192:1–6. Clin Oncol. 2017;35(30):3449–57.
53.
Tang J, Shalabi A, Hubbard-Lucey 67. Reck M, Luft A, Szczesna A, Havel L, Kim SW,
VM. Comprehensive analysis of the clinical immuno- Akerley W, et al. Phase III randomized trial of
oncology landscape. Ann Oncol. 2018;29(1):84–91. ipilimumab plus etoposide and platinum versus
54.
Wolchok JD, Chiarion-Sileni V, Gonzalez R, placebo plus etoposide and platinum in exten-
Rutkowski P, Grob JJ, Cowey CL, et al. Overall sive-stage small- cell lung cancer. J Clin Oncol.
survival with combined nivolumab and ipi- 2016;34(31):3740–8.
limumab in advanced melanoma. N Engl J Med. 68. Formenti SC, Rudqvist NP, Golden E, Cooper B,

2017;377(14):1345–56. Wennerberg E, Lhuillier C, et al. Radiotherapy
55. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS, induces responses of lung cancer to CTLA-4 block-
Moschos SJ, et al. Combined nivolumab and ipilim- ade. Nat Med. 2018;24(12):1845–51.
umab in melanoma metastatic to the brain. N Engl J 69. Lee Y, Auh SL, Wang Y, Burnette B, Meng Y, Beckett
Med. 2018;379(8):722–30. M, et al. Therapeutic effects of ablative radiation on
56. Kingsley DP. An interesting case of possible absco- local tumor require CD8+ T cells: changing strategies
pal effect in malignant melanoma. Br J Radiol. for cancer treatment. Blood. 2009;114(3):589–95.
1975;48(574):863–6. 70. Luke JJ, Lemons JM, Karrison TG, Pitroda SP,

57. Wersall PJ, Blomgren H, Pisa P, Lax I, Kalkner KM, Melotek JM, Zha Y, et al. Safety and clinical activ-
Svedman C. Regression of non-irradiated metastases ity of pembrolizumab and multisite stereotactic body
radiotherapy in patients with advanced solid tumors. J 75. Qian JM, Yu JB, Kluger HM, Chiang VL. Timing
Clin Oncol. 2018;36(16):1611–8. and type of immune checkpoint therapy affect the
71. Theelen W, Peulen H, Lalezari F, de Vries J, De
early radiographic response of melanoma brain
Langen J, Aerts J, et al., editors. Randomized phase metastases to stereotactic radiosurgery. Cancer.
II study of pembrolizumab after stereotactic body 2016;122(19):3051–8.
radiotherapy (SBRT) versus pembrolizumab alone in 76. Wang Y, Deng W, Li N, Neri S, Sharma A, Jiang W,
patients with advanced non-small cell lung cancer: the et al. Combining immunotherapy and radiotherapy for
PEMBRO-RT study. ASCO Annual Meeting; 2018; cancer treatment: current challenges and future direc-
Chicago. tions. Front Pharmacol. 2018;9:185.
72. Knisely JPS, Yu JB, Flanigan J, Sznol M, Kluger 77. Qian JM, Yu JB, Kluger HM, Chiang VL. Timing
HM, Chiang VLS. Radiosurgery for melanoma and type of immune checkpoint therapy affects
brain metastases in the ipilimumab era and early radiographic response of melanoma brain
the possibility of longer survival. J Neurosurg. metastases to stereotactic radiosurgery. Cancer.
2012;117(2):227–33. 2016;122(19):3051–8.
73. Lehrer EJ, Peterson JL, Zaorsky NG, Brown PD,
78. Marciscano AE, Ghasemzadeh A, Nirschl TR,

Sahgal A, Chiang VL, et al. Single versus multifrac- Theodros D, Kochel CM, Francica BJ, et al. Elective
tion stereotactic radiosurgery for large brain metasta- nodal irradiation attenuates the combinatorial efficacy
ses: an international meta-analysis of 24 trials. Int J of stereotactic radiation therapy and immunotherapy.
Radiat Oncol Biol Phys. 2019;103(3):618–30. Clin Cancer Res. 2018;24(20):5058–71.
74.
Young KH, Baird JR, Savage T, Cottam B, 79. Louveau A, Herz J, Alme MN, Salvador AF,

Friedman D, Bambina S, et al. Optimizing tim- Dong MQ, Viar KE, et al. CNS lymphatic drain-
ing of immunotherapy improves control of tumors age and neuroinflammation are regulated by
by hypofractionated radiation therapy. PLoS One. meningeal lymphatic vasculature. Nat Neurosci.
2016;11(6):e0157164. 2018;21(10):1380–91.
Salvage Irradiation for Patients
with Recurrent Brain Metastases
26
Christian Iorio-Morin, Laurence Masson-Côté,
and David Mathieu
Introduction sequent Disease-Specific Graded Prognostic

Assessment (DS-GPA) [5] led to a paradigm shift
Brain metastases (BMs) are the most common in BM management. These tools highlighted the
tumors of the central nervous system. Their exact wide range of outcomes observed in BM, such as
incidence is difficult to assess, since they are not a median overall survival of 4 years from the
part of the data collected by national registries diagnosis of BM in ALK-rearranged NSCLC [6],
such as the Surveillance, Epidemiology, and End compared to 3 months in SCLC with the lowest
Results (SEER) [1] and the Central Brain Tumor DS-GPA score [3]. By allowing reliable outcome
Registry of the United States (CBTRUS) [2]. In prediction before the start of therapy, these tools
prospective cohorts, 10–50% of patients with enabled clinicians to identify patients who might
cancer have been reported to be diagnosed with benefit from aggressive treatment despite a diag-
BM before their death [3]. Given that, in 2020, nosis of BM. In turn, aggressive treatment of BM
1,806,590 new cases of cancer are projected to also led to the realization that, once treated,
occur in the United States alone [1], the annual patients tend to die of their systemic, rather than
number of new BM is considerable. their neurologic disease [7–9]. As survival
Historically, a BM diagnosis was an indica- improved and active and aggressive treatment of
tion for palliative care and the prognosis was BM became mainstream, recurring BM became
typically <6 months. The development and itera- an additional challenge in the management of
tive refinement of prognostic models such as the cancer patients.
original recursive partitioning analysis (RPA)
of prognostic factors in three Radiation Therapy
Oncology Group (RTOG) trials [4] and the sub- Recurrent Brain Metastases
Depending on the treatment modality and goal

C. Iorio-Morin · D. Mathieu (*) of therapy, recurrence of BM can be assessed as
Division of Neurosurgery, Department of Surgery, either a global process encompassing all BM of
Université de Sherbrooke, Centre de Recherche du
Centre Hospitalier Universitaire de Sherbrooke, a given patient (i.e., the cerebral disease
Sherbrooke, QC, Canada response) or at a metastasis-specific level (i.e.,
e-mail: david.mathieu@usherbrooke.ca the local response). Studies of systemic thera-
L. Masson-Côté pies and whole-brain radiation therapy (WBRT)
Nuclear Medicine and Radiobiology, Centre will usually report outcomes in terms of disease
Hospitalier Universitaire de Sherbrooke, response, whereas studies of local therapies such
Sherbrooke, QC, Canada

https://doi.org/10.1007/978-3-030-42958-4_26
372 C. Iorio-Morin et al.
Table 26.1 RANO-BM CNS disease response criteria for brain metastases
Complete
response Partial response Stable disease Progressive diseasea
Target lesions None ≥30% decrease in the Between the partial ≥20% increase in the sum of the
sum of the longest response and longest diameters of all target lesions
diameters of all target progressive disease relative to the smallest measurement
lesions relative to the criteria obtained during follow-up after
baseline treatment
Nontarget None Stable or improved Stable or improved Unequivocal progression
lesions
New lesion None None None Present
Corticosteroids None Stable or decreased Stable or decreased Not applicable
Clinical status Stable or Stable or improved Stable or improved Worse
improved
Adapted from Lin et al. [10]
a
All listed criteria are required for the CNS disease to be considered in complete response, partial response or stable. If
any criteria listed under the progressive disease column are met, the CNS disease is considered to be progressing
as surgical resections, stereotactic radiosurgery spectroscopy, or FLT, FET, MET, or FDG-PET

(SRS), and laser interstitial thermal therapy [14, 15], although no approach has been proven
(LITT) will focus on the local response of the sufficient to reliably distinguish radiation necro-
treated lesion. This distinction is important when sis from true progression in all patients [10].
interpreting the literature on recurrent In light of these challenges, unless a surgical
BM. Furthermore, the definition of response and resection is performed and pathology is avail-
progression varies considerably across clinical able, progressive BMs are usually diagnosed by
trials, making meta-analyses challenging to pro- the clinical judgment of a multidisciplinary team.
duce. The most recent response assessment crite-
ria for BM have been proposed by the Response
Assessment in Neuro- Oncology Brain herapeutic Options for Recurrent
T
Metastases (RANO-BM) working group [10] Brain Metastases
and are presented in Table 26.1. These criteria
are built on the older RECIST [11], Macdonald Management of recurrent brain metastases
[12] and WHO [13] criteria and define progres- depends on the pattern of recurrence, the previ-
sive disease as any patients in whom target ously employed modalities, as well as patient fac-
lesions are progressing, nontarget lesions are tors, such as his or her functional status, systemic
progressing, new lesions are appearing, or clini- treatment options, and personal preference.
cal status is worsening. A progressive lesion is Patient factors are used to estimate the useful-
defined as a ≥20% increase in the longest diam- ness of pursuing further treatments in the setting
eter relative to the smallest measurement of recurrent, incurable disease. For most patients
obtained during follow-up after treatment [10]. with a poor functional status with recurrent brain
An exception exists for lesions treated by SRS or disease, palliative care should be considered.
immunotherapy, because transient increase in When patients have a good functional status,
lesion size or edema can often be observed in more aggressive BM management can be consid-
lesions which will eventually respond. These ered because the short-term impact of treatment
treatment effects should not be mistaken for on QOL might be outweighed by a better long-
recurrence [10]. In addition, radiation necrosis term prognosis.
can be seen as a complication of SRS-treated In the National Comprehensive Cancer
lesions and does not constitute recurrence. A Network (NCCN) guidelines [16], recurrence is
presumed diagnosis of radiation necrosis can be conceptually addressed as either local (i.e., the
supported by advanced imaging modalities progression of a previously known and treated
such as perfusion MRI, magnetic resonance lesion) or distant (i.e., the appearance of a new
26 Salvage Irradiation for Patients with Recurrent Brain Metastases 373
lesion during follow-up of a patient with other SRS has been shown to be as good as the
known BM). The CNS disease is further defined response after a first SRS [20], suggesting that
as limited, extensive, or leptomeningeal. The failure might be a random event not necessarily
threshold between limited and extensive disease related to intrinsic tumor characteristics. This
is not specified. Philosophically, this stratifica- peculiarity is relevant to the management of BM,
tion is used to distinguish patients in whom SRS because a previous failure might not be predic-
would be “equally effective and offers significant tive of a future failure. For BM patients where
cognitive protection compared with WBRT” the maintenance of short-term quality of life is
(limited disease), from those in whom SRS is not critical, the option of repeating SRS and sparing
thought to be advantageous or feasible (extensive the patient a surgery is appealing.
disease) [16]. Randomized controlled trials cur- We will now discuss various re-irradiation
rently support the advantage of primary SRS in paradigms with a special focus on their impact on
up to four metastases [17–19] although some survival and functional outcomes.
centers have reported results for many more
lesions [8].
For patients with recurrent previously irradi- SRS After SRS
ated BM and for whom active treatment is pur-
sued, therapeutic options include surgical Eight series have reported the use of repeat SRS
resection, laser interstitial thermal therapy on the same lesion (Table 26.2). Because of the
(LITT), systemic chemotherapy, and repeat irra- retrospective nature of these series, heterogeneous
diation with either SRS or WBRT. We will now inclusion criteria and the lack of standardized
discuss the various combinations of repeat irra- definition of radiation necrosis and tumor control,
diation in the setting of recurrent BM. aggregation of data, and meta-analysis are not
feasible. Reported 1-year local control ranged
from 61% to 83% and median survival after the
Rationale for Repeat Irradiation second SRS ranged from 8 months to more than
2 years. Some series were heavily biased toward
The majority of patients with recurrent BM will melanoma [28] or SCLC [24], while others were
have undergone previous SRS or WBRT, as sole more representative of standard BM histologies
primary treatment modality or as adjuvant ther- [22]. We recently analyzed our personal series,
apy following surgical resection. Distant recur- which consisted of 75 recurrent lesions in 56
rences are new lesions that have not previously patients. We used the standardized RANO criteria
been exposed to radiation, although the sur- to define outcome. Patients were treated using a
rounding brain parenchyma might have been. median dose of 20 Gy (range 14–24) for the first
Salvage irradiation in this setting is therefore SRS and 18 Gy (range 12–20) for the second.
thought to have the same efficacy on each indi- Actuarial local control at 1, 2, and 5 years was
vidual lesion as if it was a primary treatment, 68%, 54%, and 54%, respectively, and median
albeit with an increased risk of adverse radiation survival was 14 months (Journal of Neuro-
effects on the surrounding tissue. For locally Oncology, https://doi.org/10.1007/s11060-019-
recurrent lesions, however, the rationale for 03323-8). Factors associated with failure of the
repeat irradiation is different. The mechanism of repeat SRS were an absence of initial response
action of SRS is not completely understood and observed after the first SRS, a lower KPS, a lower
lesion response is different than that seen after maximal dose, and having an uncontrolled pri-
WBRT. Inherently radioresistant histologies, mary cancer at the time of the second SRS. Other
such as melanoma, renal cell carcinoma, and sar- authors associated a volume >4 cc with poor local
coma BM, which have a significantly higher control [24]. Radiation necrosis occurred in
recurrence rate following WBRT have been 5–30% of patients across all series. Risk factors
shown to respond to SRS. Moreover, for vestibu- identified for radiation necrosis included a treat-
lar schwannomas, the response after a second ment volume >7 cc [22] and the volume of lesion
Table 26.2 Series reporting SRS after SRS

Median
survival from % of
Median Median One-year last radiation % of patients radiation
first SRS second SRS local treatment with symptom induced
References n dose (Gy) dose (Gy) control (months) improvement effects
Iorio-Morin et al. [21] 56 20 (14–24) 18 (12–20) 68% 14 18% 5% RN
Moreau et al. [22] 30 18 (12–20) 18 (12–20) 68% 14.2 NR 10%

McKay et al. [23] 32 20 (12–24) 20 (14–22) 79% >24 NR 30%
Koffer et al. [24] 22 18 (17–20) 15.5 61% 8.7 NR 16.7%
(10–20)
Minniti et al. [25] 43 NR 3 × 7–8 38–78% 10 NR 19%
Trifiletti et al. [26] 24 20 18 NR 12.2 NR 9%
Jayachandran et al. [27] 19 22 (16–24) 17.3 83% 26 NR 21%
(14.5–24)
Terakedis et al. [28] 37 20 (15–24) 20 (14–24) 81% 8.3 NR 16%
receiving 40 Gy [23]. Previous WBRT was asso- were treated with WBRT (37.5 Gy in 15 frac-
ciated with an increased risk of radiation necrosis tions) with or without SRS (15–24 Gy at the mar-
in one series [22] while two others did not show a gin). This trial showed improved KPS at 6 months
significant correlation [24]. and improved survival in RPA class 1 patients
Together, these studies demonstrate the feasi- with the combination therapy. Importantly it also
bility and safety of a repeat course of SRS for demonstrated that toxicity did not differ between
locally recurrent BM. Given the lack of standard- both groups [32]. This led to a subsequent study
ized response assessment and the concurrent of SRS with or without WBRT which further
administration of other treatments in most series supported the safety of combined irradiation
(including subsequent WBRT and systemic che- [33]. Multiple series have since been published
motherapy), the level of efficacy of the second assessing SRS as a salvage treatment (i.e., not as
SRS alone remains to be determined. However, a boost) for patients who previously underwent
in properly selected patients, repeat SRS can be WBRT (Table 26.3).
used to control locally recurrent BM and post- Survival in this setting ranged between 4 and
pone WBRT or surgery. 11.7 months from the SRS. As discussed, sur-
vival can be biased by heterogeneous inclusion
criteria and practice across studies. The largest
SRS After WBRT study included 310 patients [37]. The median
survival in this series was 8.4 months overall,
For historical reasons, the combination of SRS and 12.0 versus 7.9 months in patients with a
and WBRT is the most studied double BM irra- single or multiple retreated BM. Favorable
diation paradigm. In the 1990s, multiple random- prognostic factors depended on the primary can-
ized controlled trials demonstrated the benefit of cer histology. For breast cancer, factors identi-
BM surgical resection on survival [29, 30] as well fied were an age <50, a smaller total target
as the role of WBRT to reduce recurrence [31], volume, and a longer interval between WBRT
establishing resection followed by adjuvant and SRS. For NSCLC, factors were a smaller
WBRT as the standard of care of the time. When number of BM, a KPS >60, and a controlled pri-
SRS for BM was introduced, it was proposed as mary. In melanoma, the only favorable prognos-
an alternative to surgical resection [32]. Patients tic factor was having a smaller total target
with 1–3 BM in the landmark RTOG 9508 trial volume [37].
Table 26.3 Series of salvage SRS after WBRT since 2000

Median
survival from % of
Median Median Median time last radiation % of patients radiation
WBRT dose SRS dose to local failure treatment with symptom induced
References n (Gy) (Gy) (months) (months) improvement effects
Huang et al. [34] 39 40 (30–50) 17 (12–25) 6.5 11.4 43% NR
Lucas et al. [9] 293 NR NR 14.8 4 NR NR
Kurtz et al. [35] 106 NR 21 (12–24) 6.2 11.7 NR NR
Hsu et al. [36] 78 30 (20–30) 24 (12–24) NR 11.2 NR NR
Caballero et al. 310 30 18 (7.5–22) NR 8.4 NR NR
[37] (19.8–60)
Kelly et al. [38] 76 NR 18 (16–20) 5.7 9.8 NR NR
Gwak et al. [39] 46 32.7 23 (10–36) 21 10 NR 4% of RN
(18–54.9)
Chao et al. [40] 111 37.5 23.6 8.4–15.3 9.9 NR 2%
(30–50) (9.6–25.4)
Noël et al. [41] 54 NR 17.2 >24 7.8 NR 0% RN
(11–22.9)
Local control was heterogeneously reported 20 Gy (range 4–30.6 Gy) for the second. Median
across studies. Median time to local failure survival in this series was 3.6 months (range
ranged from 5.7 months to more than 2 years, and 0.2–45) from the second treatment with 14%
control was shown to be improved in patients surviving less than a month after the second
with a favorable histology (NSCLC) [40], an WBRT. Prognostic factors associated with poor
interval between WBRT and SRS >14 months survival at the time of the second WBRT were an
[41] and a SRS dose >22 Gy [40]. In breast can- SCLC histology, the presence of extracranial
cer, overall cerebral disease control was affected metastases, a KPS <80, an interval between both
by HER2 status [38] and the systemic disease WBRT courses <9 months and an uncontrolled
status [38]. primary. These five factors were combined to cre-
As discussed, the safety of the combined irra- ate a reirradiation score in which each factor is
diation was prospectively demonstrated in multi- worth 1 point, and in this series, patients with 4–5
ple randomized controlled trials of the SRS boost points had a median survival of 2.2 months com-
paradigm [32, 33]. WBRT has recently fallen out pared to 3 months for patients with 3 points, and
of favor for most BM patients with limited brain 7.2 months for patients with 1–2 points [42]. This
disease. However, it is still commonly used for system allowed better prognostication than histol-
patients with disseminated BM, so the ability to ogy or RPA class alone, although it still requires
salvage new lesions arising after the end of external validation.
WBRT remains relevant as systemic treatments The relevance of WBRT after WBRT is
and overall survival improve. The studies in challenged by the lack of studies reporting
Table 26.3 confirm the relevance of this approach. patient-centered outcomes. With a significant
neurocognitive decline observed as soon as
3 months after the first course of WBRT [17],
WBRT After WBRT there is concern that any gain in terms of BM
control or survival provided by a second
Fourteen studies have described repeat WBRT course of WBRT could be offset by a worsened
after WBRT (Table 26.4). Reported median sur- quality of life resulting from poor cognitive
vival ranged from 2 to 6.9 months. The largest function. Improvement in symptoms after repeat
study included 205 patients from nine Canadian WBRT has been reported in 14% [48] to 80%
centers [42]. The median treatment dose was [47], highlighting inconsistent definitions,
20 Gy (range 12–48 Gy) for the first course and reporting standards, and follow-ups of these
Table 26.4 Series of WBRT after WBRT

Median survival
Median dose at Median dose at from second WBRT % of patients with
References n first WBRT (Gy) second WBRT (Gy) (months) symptom improvement
Logie et al. [42] 205 20 (12–48) 20 (4–30.6) 2.2–7.2 NR
Aktan et al. [43] 34 30 (25–30) 25 (20–30) 5.3 24%
Scharp et al. [44] 134 30 (30–40) 20 (20–30) 2.8 39%
Ozgen et al. [45] 28 30 (20–30) 25 (20–30) 3 39%
Akiba et al. [46] 31 30 (26–42) 30(30–40) 4 68%
Son et al. [47] 17 35 (28–40) 21.6 (14–30) 5.2 80%
Karam et al. [48] 37 NR NR 6.9 14%
Sadikov et al. [49] 72 30 (20–30) NR 4.1 31%
Abdel et al. [50] 15 30 (30–55) 30 (30–35) NR 60%
Wong et al. [51] 86 30 (1.5–50.6) 20 (8–30.6) 4 NR
Cooper et al. [52] 52 NR NR 5 42%
Hazuka et al. [53] 44 30 (30–36) 25 (6–36) 2 27%
Kurup et al. [54] 56 NR (18–30) 20 3.5 NR
Shehata et al. [55] 35 NR NR NR NR
terminally ill patients. In addition, the assessing cognitive functions, and maximizing local
ment of adverse radiation effects is unreliable control. WBRT remains an option for patients with
because most patients in these studies did not disseminated intracranial disease unresponsive to
undergo follow- up imaging. Given the safety systemic therapies, or with poor performance sta-
and probable efficacy of the other previously dis- tus who require palliative symptom control not
cussed paradigms, WBRT after WBRT should be achievable by best supportive care alone with ste-
reserved for symptom relief in patients with a roids. In the end, patient survival still mostly
very short expected survival not otherwise eligi- depends on the control of primary cancer. As
ble for SRS, and in whom this approach is such, when selecting the best management plan
believed, based on individual clinical judgment, for BM, clinicians should consider not only opti-
to be superior to best supportive care alone with mal local control but also target the quality of life
steroids. and symptoms control—issues poorly assessed
in current studies.
Conclusion
References
As systemic therapies for cancer are improving,
keeping cerebral metastatic disease under control 1. Siegel RL, Miller KD, Jemal A. Cancer statistics,
will become increasingly important. Salvage irra- 2020. CA Cancer J Clin. 2020;70(1):7–30.
2. Ostrom QT, Gittleman H, Xu J, Kromer C, Wolinsky
diation of previously irradiated tumors is a useful Y, Kruchko C, et al. CBTRUS Statistical Report:
strategy to achieve this goal. Multiple paradigms Primary brain and other central nervous system
have been studied, including SRS after SRS, SRS tumors diagnosed in the United States in 2009–2013.
after WBRT, WBRT after SRS, and WBRT after Neuro-oncology. 2016;18(suppl_5):v1–v75.
3. Arvold ND, Lee EQ, Mehta MP, Margolin K,
WBRT. All have been shown to be feasible and Alexander BM, Lin NU, et al. Updates in the man-
safe in properly selected patients. The choice of agement of brain metastases. Neuro-Oncology.
salvage strategy depends on the initially chosen 2016;18(8):1043–65.
management modalities. For patients with a good 4. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T,
Wasserman T, et al. Recursive partitioning analysis
functional status and a limited intracranial disease, (RPA) of prognostic factors in three Radiation Therapy
multiple courses of SRS, irrespective of a previous Oncology Group (RTOG) brain metastases trials. Int J
WBRT, are likely to provide the best results, spar- Radiat Oncol Biol Phys. 1997;37(4):745–51.
5. Sperduto PW, Kased N, Roberge D, Xu Z, Shanley 18. Soon YY, Tham IWK, Lim KH, Koh WY, Lu

R, Luo X, et al. Summary report on the graded prog- JJ. Surgery or radiosurgery plus whole brain radio-
nostic assessment: an accurate and facile diagnosis- therapy versus surgery or radiosurgery alone for
specific tool to estimate survival for patients with brain metastases. Cochrane Gynaecological, Neuro-
brain metastases. J Clin Oncol. 2012;30(4):419–25. oncology and Orphan Cancer Group, editor. Cochrane
6. Johung KL, Yeh N, Desai NB, Williams TM, Database Syst Rev. 2014;295(3):CD009454.
Lautenschlaeger T, Arvold ND, et al. Extended sur- 19. Brown PD, Jaeckle K, Ballman KV, Farace E, Cerhan
vival and prognostic factors for patients with ALK- JH, Anderson SK, et al. Effect of radiosurgery alone
rearranged non-small-cell lung cancer and brain vs radiosurgery with whole brain radiation therapy
metastasis. J Clin Oncol. 2016;34(2):123–9. on cognitive function in patients with 1 to 3 brain
7. Iorio-Morin C, Masson-Côté L, Ezahr Y, Blanchard metastases: a randomized clinical trial. JAMA.
J, Ebacher A, Mathieu D. Early Gamma Knife stereo- 2016;316(4):401–9.
tactic radiosurgery to the tumor bed of resected brain 20. Iorio-Morin C, Liscak R, Vladyka V, Kano H, Jacobs
metastasis for improved local control. J Neurosurg. RC, Lunsford LD, et al. Repeat stereotactic radiosur-
2014;121 Suppl 2:69–74. gery for progressive or recurrent vestibular schwan-
8. Yamamoto M, Serizawa T, Shuto T, Akabane A, nomas. Neurosurgery. 2018;19(suppl_5):v1.
Higuchi Y, Kawagishi J, et al. Stereotactic radio- 21.
Iorio-Morin C, Mercure-Cyr R, Figueiredo G,
surgery for patients with multiple brain metastases Touchette CJ, Masson-Côté L, Mathieu D. Repeat
(JLGK0901): a multi-institutional prospective obser- stereotactic radiosurgery for the management of
vational study. Lancet Oncol. 2014;15(4):387–95. locally recurrent brain metastases. J Neuro oncol.
9. Lucas JT, Colmer HG, White L, Fitzgerald N, Isom 2019;145(3):551–9. https://doi.org/10.1007/
S, Bourland JD, et al. Competing risk analysis of s11060-019-03323-8.
neurologic versus nonneurologic death in patients 22. Moreau J, Khalil T, Dupic G, Chautard E, Lemaire
undergoing radiosurgical salvage after whole-brain J-J, Magnier F, et al. Second course of stereotactic
radiation therapy failure: who actually dies of their radiosurgery for locally recurrent brain metastases:
brain metastases? Int J Radiat Oncol Biol Phys. safety and efficacy. Zhang Q, editor. PLoS ONE.
2015;92(5):1008–15. 2018;13(4):e0195608.
10. Lin NU, Lee EQ, Aoyama H, Barani IJ, Barboriak 23. McKay WH, McTyre ER, Okoukoni C, Alphonse-
DP, Baumert BG, et al. Response assessment criteria Sullivan NK, Ruiz J, Munley MT, et al. Repeat ste-
for brain metastases: proposal from the RANO group. reotactic radiosurgery as salvage therapy for locally
Lancet Oncol. 2015;16(6):e270–8. recurrent brain metastases previously treated with
11. Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, radiosurgery. J Neurosurg. 2017;127(1):148–56.
Sargent D, Ford R, et al. New response evaluation 24. Koffer P, Chan J, Rava P, Gorovets D, Ebner D,

criteria in solid tumours: revised RECIST guideline Savir G, et al. Repeat stereotactic radiosurgery for
(version 1.1). Eur J Cancer. 2009;45:228–47. locally recurrent brain metastases. World Neurosurg.
12. Macdonald DR, Cascino TL, Schold SC, Cairncross 2017;104:589–93.
JG. Response criteria for phase II studies of 25. Minniti G, Scaringi C, Paolini S, Clarke E, Cicone
supratentorial malignant glioma. J Clin Oncol. F, Esposito V, et al. Repeated stereotactic radiosur-
1990;8(7):1277–80. gery for patients with progressive brain metastases. J
13. Miller AB, Hoogstraten B, Staquet M, Winkler
Neuro-Oncol. 2016;126(1):91–7.
A. Reporting results of cancer treatment. Cancer. 26. Trifiletti DM, Patel NV, Sheehan JP. Repeated stereo-
1981;47(1):207–14. tactic radiosurgery for intracranial metastases after
14. Shah R, Vattoth S, Jacob R, Manzil FFP, O’Malley JP, local failure: the safety and efficacy of repeat radiosur-
Borghei P, et al. Radiation necrosis in the brain: imag- gery. Int J Radiat Oncol Biol Phys. 2015;93(3):E73.
ing features and differentiation from tumor recur- 27. Jayachandran P, Shultz D, Modlin L, Eyben Von

rence. Radiographics. 2012;32(5):1343–59. R, Gibbs IC, Chang S, et al. Repeat Stereotactic
15. Vellayappan B, Tan CL, Yong C, Khor LK, Koh WY, Radiosurgery (SRS) for brain metastases locally
Yeo TT, et al. Diagnosis and management of radia- recurrent following initial SRS. Int J Radiat Oncol
tion necrosis in patients with brain metastases. Front Biol Phys. 2014;90(1):S320.
Oncol. 2018;8:395. 28. Terakedis BE, Jensen RL, Boucher K, Shrieve

16. National Comprehensive Cancer Network. Central
DC. Tumor control and incidence of radiation necrosis
Nervous System Cancers [Internet]. 1st ed. NCCN. after reirradiation with stereotactic radiosurgery for
org. 2018 [cited 2018 Nov 1]. pp. 1–136. Available brain metastases. J Radiosurg SBRT. 2014;3(1):21–8.
from: https://www.nccn.org/professionals/physician_ 29. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ,

gls/pdf/cns_blocks.pdf. Maruyama Y, Kryscio RJ, et al. A randomized trial of
17. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, surgery in the treatment of single metastases to the
Kornguth DG, et al. Neurocognition in patients with brain. N Engl J Med. 1990;322(8):494–500.
brain metastases treated with radiosurgery or radio- 30. Vecht CJ, Haaxma-Reiche H, Noordijk EM, Padberg
surgery plus whole-brain irradiation: a randomised GW, Voormolen JH, Hoekstra FH, et al. Treatment
controlled trial. Lancet Oncol. 2009;10(11):1037–44. of single brain metastasis: radiotherapy alone or
combined with neurosurgery? Ann Neurol. Wiley- 41. Noël G, Proudhom MA, Valery CA, Cornu P,

Blackwell. 1993;33(6):583–90. Boisserie G, Hasboun D, et al. Radiosurgery for re-
31. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, irradiation of brain metastasis: results in 54 patients.
Mohiuddin M, Kryscio RJ, et al. Postoperative Radiother Oncol. 2001;60(1):61–7.
radiotherapy in the treatment of single metas- 42. Logie N, Jimenez RB, Pulenzas N, Linden K, Ciafone
tases to the brain: a randomized trial. JAMA. D, Ghosh S, et al. Estimating prognosis at the time
1998;280(17):1485–9. of repeat whole brain radiation therapy for multiple
32. Andrews DW, Scott CB, Sperduto PW, Flanders AE, brain metastases: the reirradiation score. Adv Radiat
Gaspar LE, Schell MC, et al. Whole brain radiation Oncol. 2017;2(3):381–90.
therapy with or without stereotactic radiosurgery 43. Aktan M, Koc M, Kanyilmaz G, Tezcan Y. Outcomes
boost for patients with one to three brain metastases: of reirradiation in the treatment of patients with mul-
phase III results of the RTOG 9508 randomised trial. tiple brain metastases of solid tumors: a retrospective
Lancet. 2004;363(9422):1665–72. analysis. Ann Transl Med. 2015;3(21):325.
33. Aoyama H, Shirato H, Tago M, Nakagawa K,
44. Scharp M, Hauswald H, Bischof M, Debus J, Combs
Toyoda T, Hatano K, et al. Stereotactic radio- SE. Re-irradiation in the treatment of patients with
surgery plus whole-brain radiation therapy vs cerebral metastases of solid tumors: retrospective
stereotactic radiosurgery alone for treatment analysis. Radiat Oncol. BioMed Central. 2014;9(1):4.
of brain metastases: a randomized controlled 45. Ozgen Z, Atasoy BM, Kefeli AU, Seker A, Dane F,
trial. JAMA. American Medical Association. Abacioglu U. The benefit of whole brain reirradia-
2006;295(21):2483–91. tion in patients with multiple brain metastases. Radiat
34. Huang Z, Sun B, Shen G, Cha L, Meng X, Wang Oncol. BioMed Central. 2013;8(1):186.
J, et al. Brain metastasis reirradiation in patients 46. Akiba T, Kunieda E, Kogawa A, Komatsu T, Tamai Y,
with advanced breast cancer. J Radiat Res. Ohizumi Y. Re-irradiation for metastatic brain tumors
2017;58(1):142–8. with whole-brain radiotherapy. Jpn J Clin Oncol.
35. Kurtz G, Zadeh G, Gingras-Hill G, Millar B-A,
2012;42(4):264–9.
Laperriere NJ, Bernstein M, et al. Salvage radiosur- 47. Son CH, Jimenez R, Niemierko A, Loeffler JS, Oh
gery for brain metastases: prognostic factors to con- KS, Shih HA. Outcomes after whole brain reirradia-
sider in patient selection. Int J Radiat Oncol Biol tion in patients with brain metastases. Int J Radiat
Phys. 2014;88(1):137–42. Oncol Biol Phys. 2012;82(2):e167–72.
36. Hsu F, Kouhestani P, Nguyen S, Cheung A, McKenzie 48. Karam I, Nichol A, Woods R, Tyldesley S. Population-
M, Ma R, et al. Population-based outcomes of boost based outcomes after whole brain radiotherapy and re-
versus salvage radiosurgery for brain metastases irradiation in patients with metastatic breast cancer in
after whole brain radiotherapy. Radiother Oncol. the trastuzumab era. Radiat Oncol. BioMed Central.
2013;108(1):128–31. 2011;6(1):181.
37. Caballero JA, Sneed PK, Lamborn KR, Ma L,
49. Sadikov E, Bezjak A, Yi Q-L, Wells W, Dawson L,
Denduluri S, Nakamura JL, et al. Prognostic fac- Millar B-A, et al. Value of whole brain re-irradiation
tors for survival in patients treated with stereotactic for brain metastases – single centre experience. Clin
radiosurgery for recurrent brain metastases after prior Oncol (R Coll Radiol). 2007;19(7):532–8.
whole brain radiotherapy. Int J Radiat Oncol Biol 50. Abdel-Wahab MM, Wolfson AH, Raub W, Landy H,
Phys. 2012;83(1):303–9. Feun L, Sridhar K, et al. The role of hyperfractionated
38. Kelly PJ, Lin NU, Claus EB, Quant EC, Weiss
re-irradiation in metastatic brain disease: a single insti-
SE, Alexander BM. Salvage stereotactic radiosur- tutional trial. Am J Clin Oncol. 1997;20(2):158–60.
gery for breast cancer brain metastases: outcomes 51. Wong WW, Schild SE, Sawyer TE, Shaw EG. Analysis
and prognostic factors. Cancer. Wiley-Blackwell. of outcome in patients reirradiated for brain metasta-
2012;118(8):2014–20. ses. Int J Radiat Oncol Biol Phys. 1996;34(3):585–90.
39. Gwak H-S, Yoo HJ, Youn S-M, Lee DH, Kim MS, 52. Cooper JS, Steinfeld AD, Lerch IA. Cerebral metas-
Rhee CH. Radiosurgery for recurrent brain metasta- tases: value of reirradiation in selected patients.
ses after whole-brain radiotherapy: factors affecting Radiology. 1990;174(3 Pt 1):883–5.
radiation-induced neurological dysfunction. J Korean 53. Hazuka MB, Kinzie JJ. Brain metastases: results and
Neurosurg Soc. Korean Neurosurgical Society;. effects of re-irradiation. Int J Radiat Oncol Biol Phys.
2009;45(5):275–83. 1988;15(2):433–7.
40. Chao ST, Barnett GH, Vogelbaum MA, Angelov
54. Kurup P, Reddy S, Hendrickson FR. Results of

L, Weil RJ, Neyman G, et al. Salvage stereotac- re-irradiation for cerebral metastases. Cancer.
tic radiosurgery effectively treats recurrences 1980;46(12):2587–9.
from whole-brain radiation therapy. Cancer. Wiley 55. Shehata WM, Hendrickson FR, Hindo WA. Rapid
Subscription Services, Inc., A Wiley Company. fractionation technique and re-treatment of cerebral
2008;113(8):2198–204. metastases by irradiation. Cancer. 1974;34(2):257–61.
Applications of Stereotactic
Radiosurgery for Brain Metastases
27
Akshay V. Save, Dominique M. O. Higgins,
Mark D. Mayeda, and Tony J. C. Wang
Introduction treatments with masks have become feasible

without sacrificing accuracy. In this chapter, we
Radiation therapy has been a mainstay of man- outline seven particularly interesting applica-
aging brain metastases for decades. In the past, tions in which SRS was used to increase the like-
clinicians have relied on whole brain radiother- lihood of achieving local control in patients with
apy (WBRT) and in some instances partial brain a variety of brain metastatic lesions.
radiation therapy to deliver conformal radiation
dose distributions using MRI and CT-guidance
to plan treatment. The dose limitations of vari- Case 1: Skull-Base Metastasis
ous critical structures of the brain and treatment-
induced neurocognitive side effects made History
intracranial treatment challenging. However, the
development and popularization of stereotactic This is a 43-year-old woman with history of
radiosurgery (SRS), including Gamma Knife stage IIIA cT3N2M0 right-sided invasive ductal
stereotactic radiosurgery (GKRS) and linear breast carcinoma with ER+/PR+/HER2+ status
accelerator (LINAC)-based SRS, has had a pro- who underwent neoadjuvant chemotherapy,
found impact on the field of neuro-oncology. As mastectomy, adjuvant irradiation of the right
the name suggests, SRS typically uses a method chest wall and internal mammary lymph nodes,
of immobilization to maintain a particular patient and adjuvant herceptin and anastrazole. Over a
position for the duration of treatment. While this year after this initial treatment, she developed
method of immobilization has traditionally been multiple brain metastases, for which she
an invasive, rigid frame placed onto a patient’s received memantine and hippocampal-sparing
head, with newer technical advances, frameless WBRT. Several months later, she represented
with new metastatic brain lesions which were
treated with GKRS. Four months after complet-
A. V. Save · D. M. O. Higgins ing her treatment, she was found to have interval
Columbia University Irving Medical Center, enlargement of a right cerebellar lesion in the
New York, NY, USA posterior fossa adjacent to the skull base. Her
M. D. Mayeda · T. J. C. Wang (*) case was discussed at a multidisciplinary con-
Department of Radiation Oncology, ference and the consensus was that she should
Columbia University Irving Medical Center, receive repeat GKRS (Fig. 27.1).
New York, NY, USA
e-mail: tjw2117@cumc.columbia.edu

https://doi.org/10.1007/978-3-030-42958-4_27
380 A. V. Save et al.
Fig. 27.1 Gamma Knife stereotactic radiosurgery treatment plan for breast cancer metastatic to the skull base of the
posterior fossa
Radiosurgery for Skull Base Tumors shown to have an excellent 1-year local control
rate of 89% in a series of calvarial and skull
While metastases to the skull base are less com- base lesions [4]. Further, GKRS can be repeated
mon than metastases to the cerebral hemispheres, with minimal consequences in the case of dis-
they present a unique challenge for management. ease progression. Overall, SRS has become an
Retrospective case series have found that lung, attractive option for patients with inoperable
breast, and prostate cancers have the highest inci- skull base tumors or an overwhelming burden
dence of skull base metastasis [1, 2]. Though of disease.
prostate cancers rarely metastasize to the brain, a
large review of the French and English case lit-
erature found that 38% of reported skull base Dosage and Treatment
metastases were from primary prostate cancers Considerations
and 20% were from primary breast cancers [1].
In general, skull base tumors are challenging This patient received single-fraction GKRS of
to treat due to their proximity to critical neural 20 Gy to the lesion, which, although it is not
structures. Surgical resection can be limited or physically within the clivus or petrous bone,
even impossible due to the risk of damage to shares with tumors in those sites the issues of
the brainstem or cranial nerves [3]. This is fur- relative surgical inaccessibility and concerns
ther complicated in the case of metastasis due about tolerance of cranial nerves emerging from
to the frequency of multifocal disease. SRS the brainstem. Approximately 2 weeks after treat-
provides the ability to selectively deliver high ment, she developed nausea and vomiting, which
doses of radiation to tumor regions, with sharp resolved after treatment with dexamethasone.
dropoffs in dosage that minimizes the toxicity Repeat imaging 2 months after treatment showed
to surrounding tissue [3]. GKRS has been resolution of the treated lesion.
27 Applications of Stereotactic Radiosurgery for Brain Metastases 381
ase 2: Re-irradiation GKRS After

C e-irradiation After Whole Brain
R
WBRT Radiotherapy
History In the setting of diffusely distributed metastatic

disease to the brain, WBRT remains the main-
This is a 34-year-old woman who presented post- stay of treatment. One of the advantages of
partum with fevers, persistent abdominal and pel- WBRT is the ability to preemptively treat
vic pain, abdominal distension, and dark urine. microscopic preclinical lesions before they
Abdominal imaging revealed multiple hypodense become symptomatic or even radiographically
lesions in the liver with biopsy revealing poorly evident. However, despite this, many patients
differentiated carcinoma consistent with a pri- progress and develop new lesions. In that set-
mary ER-/PR 10%+/HER2+ breast cancer. She ting, salvage therapy options include SRS or
began treatment with paclitaxel, trastuzumab, repeat WBRT. In general, clinicians have
and pertuzumab, but began experiencing visual avoided repeat WBRT because of the concerns
field deficits. MRI at the time revealed numerous for violating normal brain organ constraints
lesions diffusely distributed throughout the brain. and also the increased risk of neurocognitive
Given the overwhelming burden of disease, she deterioration. In this setting, SRS has had
was treated with WBRT, 3750 cGy in 15 frac- promising results in several trials.
tions, instead of SRS. Over time, she developed The RTOG trial 90-05 set out to determine
new metastatic lesions and was referred for the maximum dose of focal radiation that could
GKRS to prevent progression of disease be delivered without causing significant CNS
(Fig. 27.2). toxicity in a large cohort of recurrent h igh-grade
Fig. 27.2 Gamma Knife stereotactic radiosurgery after prior WBRT for patient with multifocal metastases. All lesions
were treated to between 18 and 20 Gy in one fraction
gliomas and brain metastases. For tumors tus, the consensus from the panel was to recom-
≤20 mm, 21–30 mm, and 31–40 mm in maxi- mend GKRS without WBRT with close interval
mum diameter, the highest safe dose for previ- follow-up (Fig. 27.3).
ously irradiated brains were 24 Gy, 18 Gy, and
15 Gy, respectively [5]. The most common
severe CNS toxicities were irreversible cere- Treatment Considerations
bral edema requiring steroids and radiation for Multifocal Metastases
necrosis requiring craniotomy. Increased tumor
volume was associated with increased CNS The development of SRS has made a consider-
toxicity. able impact on the management of patients with
In a dedicated series for SRS to patients multiple brain metastases. Standard treatment for
with brain metastases that had failed WBRT, multiple metastases had traditionally been
response rates were 91%, with a 1-year local WBRT. However, high rates of neurocognitive
control rate of 74%. Re-irradiation was found deterioration and memory deficits have made
to be safe and effective with low rates of radia- WBRT a less attractive option, especially consid-
tion necrosis (6%). Chao et al. found that ering the accuracy and local control rates achiev-
patients who had sustained responses to initial able with modern SRS.
WBRT tended to have longer survival after In 2014, a large prospective observational
SRS treatment [6]. Based on these results, SRS study compared the overall survival of patients
after WBRT has been shown to provide high receiving SRS, stratified by the number of metas-
rates of local control without significant side tases. This multi-institutional study enrolled
effects [5–8]. 1194 patients over a 3-year period and provided
a median follow-up of 20.9 months. Fairly high-
functioning patients, with a KPS >70, with evi-
ase 3: Whole Brain Versus
C dence of newly diagnosed metastatic lesions and
Stereotactic Radiotherapy a cumulative tumor volume less than 15 mL,
for Multiple Metastases who were treated with SRS without WBRT, were
included in the study. Patients who had one met-
History astatic lesion were found to have better overall
survival (13.9 months) compared to patients
This is a 44-year-old woman with history of with more than one metastasis (10.8 months).
left-sided poorly differentiated invasive ductal However, importantly, the authors found no dif-
breast carcinoma with ER+/PR+/HER2- dis- ference in survival between patients that had five
ease, initial stage unknown. She underwent to ten metastatic brain lesions compared to those
bilateral mastectomies with chemotherapy and with two to four, with both groups having a
unilateral radiation to the left chest wall. Several median survival of 10.8 months [9]. Furthermore,
years after treatment, she presented with persis- univariate analysis of volumetric data found that
tent cough and chest pain. Comprehensive dimension-related factors such as a maximum
imaging studies at the time revealed pleural diameter of the largest lesion greater than 1.6 cm
nodules and several areas of dural enhancement. and a cumulative tumor volume greater than
Shortly after, she had an episode of right scalp, 1.9 mL were associated with worse overall sur-
face, and arm numbness followed by loss of vival. Though these factors were not statisti-
consciousness. Imaging revealed multifocal cally significant predictors in the multivariate
metastases in the frontal lobes, right parietal analysis, these findings together suggest that
lobe, right cerebellar hemisphere, left corona volumetric factors may be more important than
radiata, and right temporal lobe. Her case was the absolute number of metastases in clinical
discussed at multidisciplinary conference, and decision-making for patients with multiple
given her young age and good p erformance sta- metastases [10, 11].
Fig. 27.3 Gamma Knife stereotactic radiosurgery treatment plan for a patient with multifocal metastatic breast cancer.
All lesions were treated between 18 and 20 Gy each
ase 4: SRS for Neurocognitive-

C SRS, she was considered to be a good candidate
Sparing Treatment of Multifocal for continued SRS. Several months after her sec-
Disease ond course of GKRS, she progressed and devel-
oped widely disseminated lesions with significant
History changes to her baseline, including visual field
deficits, headaches, nausea, vomiting, and debili-
This is the case of a 43-year-old woman with a tating lethargy leaving her wheel chair bound. At
history of ER-/PR+/HER2-invasive ductal carci- this point she decided to pursue WBRT and was
noma complicated by multifocal, multisite considered for enrollment in an ongoing clinical
metastases for which she has undergone several trial for hippocampal-sparing WBRT with
radiotherapy regimens. She presented to her memantine (Fig. 27.4).
oncologist with worsening headaches, fatigue,
and confusion, with neuroimaging showing six
brain lesions distributed throughout the temporal Treatment Considerations
lobe, occipital lobe, and cerebellum with com- for Neurocognitive Sparing
pression of the fourth ventricle. She was initially Radiotherapy in Patients with Good
treated with GKRS, but over time she developed Performance Status
additional metastases. At the time, she wished to
avoid WBRT. Given her young age, good perfor- Though recent evidence suggests that SRS may
mance status, and effective response to prior provide similar rates of local control for up to ten
Fig. 27.4 Gamma Knife stereotactic radiosurgery plan for multifocal metastatic disease rather than WBRT for neuro-
cognitive protection. All four lesions were treated to 20 Gy
brain metastases, whole brain radiotherapy memory and motor dexterity, though it did not
(WBRT) remains the mainstay treatment for widely change the overall neurocognitive testing score
disseminated metastatic disease. The rationale for [15]. Memantine slowed the cognitive decline in
WBRT is to minimize symptom burden from the patients undergoing WBRT for metastatic dis-
current lesions and also to preemptively treat disease, though the study failed to meet statistical
tant sites of subclinical disease. Conventional frac- significance (p = 0.059) likely due to a relatively
tionation schedules for WBRT involve 20 Gy in 5 low rate of patient follow-up [16]. Additionally, a
fractions, 30 Gy in 10 fractions, and 37.5 Gy in 15 phase II multi-institutional trial studied the
fractions, which have similar overall survival and effects of hippocampal-sparing WBRT and found
preservation of neurological function probabilities significant preservation of memory and quality of
[12]. While two studies looked at accelerated life compared to previous studies [17].
schedules of 40 Gy in 20 twice-daily fractions and Preliminary results from an ongoing phase III
found better rates of local control with slower time trial suggest that hippocampal avoidance pro-
to progression, they were unable to find any benefit vides a cognitive benefit that is noticeable as
in terms of overall survival [12–14]. early as 3 months into treatment, most notably in
WBRT is effective in treating multifocal brain executive function and total recall and recogni-
metastatic disease; however, it has been associ- tion. However, there have been no statistically
ated with long-term neurocognitive and memory significant differences in terms of intracranial
deficits in a large percentage of patients. To mini- progression or overall survival. Despite the
mize these toxicities, neurocognitive agents such evolving role of SRS, whole brain radiotherapy
as donepezil and memantine have been studied. remains an important treatment option in the
Donepezil has been shown to prevent decline in management of multifocal brain metastases.
ase 5: Frameless SRS to Post-

C distinguished between scarring or residual neo-
resection Cavities plasm. Given the large size of her brain metasta-
sis, she was referred for post-operative SRS of
History the resection cavity for treatment of suspected
residual disease (Fig. 27.5).
This is a 64-year-old woman with stage IIIA
cT1aN2M0 lung adenocarcinoma with mutated
KRAS treated with three cycles of carboplatin/ tereotactic Radiosurgery for Post-
S
pemetrexed, left upper lobectomy with complete resection Cavities
thoracic lymph node dissection, and post-
operative radiation therapy to the chest. Almost Two landmark trials in the 1990s showed that
3 years later, she began experiencing daily severe WBRT after surgical resection of a single brain
holocephalic headaches that would wake her up metastasis offers better progression-free survival
from sleep in the morning. She underwent neuro- than either WBRT or surgical resection alone [18,
imaging, which revealed a large 3.9 × 2.7 × 3.8 cm 19]. As such, radiation to the post-operative
peripherally enhancing, centrally necrotic mass resection cavities has been a mainstay in the
in the right parietal lobe with substantial vaso- treatment of large single metastases that are ame-
genic edema causing 6 mm midline shift. This nable to surgery. However, as previously
mass was resected with pathology consistent described, WBRT carries a higher risk of neuro-
with a poorly differentiated metastatic carci- cognitive decline, which has prompted increased
noma. Her postoperative brain MRI showed min- research in the role of SRS for treatment. SRS to
imal surrounding enhancement that could not be the resection cavity in fewer than four metastases
Fig. 27.5 Gamma Knife radiosurgery plan for the post-operative resection cavity
results in longer time to progression compared to ment with GKRS. The ability to fractionate treat-
observation alone [20]. Additionally, a phase III ment provides a radiobiological advantage to local
study compared SRS and WBRT in patients after tumor control by exploiting the relative dysfunction
resection of metastatic disease [21]. Though of DNA-repair pathways in tumor cells compared
patients in the SRS cohort had worse local and to healthy cells. Additionally, the ability to better
distant control compared to patients with WBRT, manage constraints to critical organs such as the
there were no differences in overall survival brainstem, optic pathways, and cochlea through
between the treatment groups. Further, patients fractionation minimizes toxicity. While the concern
treated with SRS had less frequent decline in of frameless treatment is a decrease in treatment
cognition and a better quality of life compared to accuracy compared to treatment with a stereotactic
patients who received WBRT. A systematic frame, we found that the use of real-time monitor-
review and meta-analysis of the literature found ing with an infrared camera was able to successfully
that SRS may be associated with a higher rate of ensure that the patients remained in the stereotactic
leptomeningeal disease compared to WBRT; space. Whenever rare deviations are greater than a
however, both treatment options provide similar pre-set threshold, a new cone-beam CT is required
survival and disease control rates [22]. prior to re-starting treatment. LINAC mask-based
The largest clinical concern for using SRS over radiosurgical systems employing image guidance in
WBRT is the risk of recurrent disease at distant the treatment room provide the same ability to frac-
sites. Factors associated with distant brain failure tionate treatments and may have some advantages
are uncontrolled systemic disease, melanoma in terms of rapidity of treatment delivery to larger
lesions, and higher numbers of brain metastases target volumes and target volumes that are spatially
[23]. Interestingly, size of the pre-operative meta- separated from each other.
static lesion or post-operative resection cavity is
not associated with a differential response to SRS
versus WBRT. A comparison of patients with Dosage and Treatment
large tumors (>4 cm) compared to smaller tumors Considerations
(≤4 cm) found no statistically significant differ-
ence between 1-year rates of local control, radia- This patient was treated with 24 Gy in three frac-
tion necrosis, or overall survival [24]. Thus, tions of frameless GKRS. She experienced some
post-operative SRS to the resection cavity should fatigue in the weeks after treatment, but tolerated
be considered as a main treatment option for the procedure without significant issues.
patients with single metastases.
ase 6: Management of Brainstem

C
rameless GKRS for Fractionated
F Lesions
Treatment
History
The advent of frameless GKRS has made treatment
with GKRS more acceptable to patients because it This is a 49-year-old woman with a past smok-
obviates the need to place an invasive stereotactic ing history who presented to an outside hospital
Leksell G Frame. Placement of s tereotactic frames with complaints of worsening abdominal pain,
can be uncomfortable and can rarely result in infec- decreased appetite, bloating, dyspnea, and
tions or persistent pain at the insertion sites. Our nonproductive cough. She was found to have
institution published a case series on the first 100 multiple nodules in her lung fields, lesions in
consecutive patients that were treated with frame- the liver, moderate ascites in the abdomen and
less GKRS on the Gamma Knife Icon [25]. In our pelvis, and bone lesions, findings consistent
experience, using the frameless mask for treatment with diffusely metastatic disease. Bronchoscopy
resulted in improvements in our workflow and an and biopsy revealed lung adenocarcinoma.
increase in the number of patients eligible for treat- Comprehensive imaging revealed an 8 mm
lesion in the midbrain in the left cerebral superior local control in their cohort [26]. Based
peduncle, and three other lesions between 5 on a large multi-institutional series, estimated
and 7 mm in both frontal lobes, and she was rates of severe toxicity after brainstem SRS are
referred for consideration of SRS. 7.4%, with the most common reported adverse
effects being radiation necrosis, intra-tumoral
hemorrhage, and symptomatic peri-tumoral
se of SRS for Management
U edema [27]. Prior whole-brain irradiation has
of Brainstem Metastases been shown to increase the risk of severe toxicity
in patients undergoing SRS for brainstem metas-
SRS plays an important role in treating lesions in tases [27]. Tumor location further sub-localized
eloquent brain regions. Although the recom- within the brainstem has not been definitively
mended maximal dose tolerance for the brainstem shown to predict for toxicity [28]. Overall, SRS
is 12–12.5 Gy in a single fraction and radiation for brainstem lesions has been shown to be a safe
dose to the margins of radiosurgically treated and effective treatment with high rates of local
tumors routinely can exceed that amount. Despite control and low toxicity [26–31].
this concern, the sharp drop-off in radiation with
SRS at a sub-millimeter level still makes it an
attractive option for brainstem metastases. Dosage and Treatment
Though some studies have been insufficient pow- Considerations
ered to determine a statistically significant rela-
tionship between tumor margin dosage and local This brainstem lesion was treated to 15 Gy in 1
control rates, Trifiletti et al. showed that higher fraction (Fig. 27.6). The mean dose to the
radiation doses at the tumor margin provided brainstem was 1.9 ± 2.2 Gy. About 1% of the
Fig. 27.6 Gamma Knife stereotactic radiosurgery plan for brainstem metastatic lesion
Fig. 27.7 Dose-volume cm3

histogram for the 25
brainstem, outlined in
pink on MRI
20
Brainstem volume
15
10
10.0 12.0 Gy
0 0.317
0 10 20 30 40
Radiation dose
brainstem volume received a dose ≥10 Gy and RS for Multifocal Radio-Resistant

S
less than 0.5% received more than 12 Gy Disease
(Fig. 27.7).
Different tumor histologies often have varying
expected responses to radiation therapy.
ase 7: Multiple Melanoma
C Radiosensitive tumors include breast, prostate,
Metastases in Elderly Patient ovarian, and neuroendocrine carcinomas,
with Comorbidities whereas traditionally radio-resistant tumors
include renal cell carcinoma, sarcoma, and mela-
History noma [32, 33]. Of these radio-resistant tumor
histologies, melanoma is the most common to
This is an 88-year-old man with a past medical metastasize to the brain. Generally, radio-
history of diabetes mellitus, hypertension, and resistant tumors cannot be adequately treated
hyperlipidemia who presented to an outside hos- with whole brain radiotherapy and require SRS to
pital with several months of worsening imbal- deliver the high doses of radiation for successful
ance on his left side resulting in several bicycle local control. In a large retrospective study of
accidents. Initial imaging studies revealed a lung patients undergoing SRS, our institution found
mass and several brain lesions. Bronchoscopy no difference in local control or overall survival
with biopsy of the lung lesion revealed pathology in patients with radiosensitive versus radio-
consistent with melanoma. He also underwent resistant histologies, suggesting that upfront SRS
biopsy and resection of the main brain lesion. may be the optimal treatment strategy for brain
Subsequently, he developed additional lesions, metastases with radio-resistant histologies [33].
which were treated with GKRS. He progressed The introduction of anti-PD1 and anti-CTLA4
both intra-cranially and systemically despite immunotherapy has had a significant impact on
treatment with pembrolizumab and was started survival in patients with advanced melanoma.
on ipilimumab and nivolumab. Despite treat- However, initial studies on the effects of immune
ment, he developed worsening disease causing checkpoint inhibition in melanoma excluded
loss of mental acuity, confusion, as well as wors- patients with brain metastases. Some estimates
ening imbalance. On latest imaging, he was suggest that almost 50% of patients with meta-
found to have five new brain metastases lesions static melanoma will have brain metastases [34];
for which he was referred for SRS (Fig. 27.8). therefore, optimizing the treatment strategy for
Fig. 27.8 Gamma Knife radiosurgery plan for treatment of a large intracranial lesion consistent with metastatic mela-
noma. This lesion was treated to 30 Gy in five fractions
this patient cohort will have a significant clinical metastases. This would allow de-escalation of
impact. A recent multicenter phase II trial stud- immunotherapy to decrease systemic toxicity,
ied the effect of combined dual anti-PD1 and while maintaining high rates of intracranial con-
anti-
CTLA4 immunotherapy on melanoma trol. Preclinical and clinical data suggest that
patients with non-radiated brain metastases. The radiation therapy may enhance the effects of
authors reported that 57% of patients had an immunotherapy by increasing the extent of lym-
intracranial response with dual therapy [35], phocytic infiltration into diseased tissue or by
whereas reported results for monotherapy of the abscopal effect. The abscopal effect is a
immune checkpoint inhibitors have response form of activation of the adaptive immune sys-
rates of 20–24% [36, 37]. Additionally, 64% of tem whereby local tumor death releases tumor-
patients were progression-free at 6-month fol- specific antigens that initiate a systemic immune
low-up. Unfortunately, 55% of patients experi- response [38]. Retrospective studies have found
enced grade 3 or 4 toxicities including hepatic a survival benefit when combining radiation
and CNS toxicity, with one patient dying from therapy and immune checkpoint inhibition in
immune-related myocarditis. terms of lesion volume [39], regional control
Given this high level of toxicity experienced rates, time to progression, and overall survival
in patients undergoing dual immune-checkpoint [40, 41], Results from ongoing prospective clin-
inhibitor therapy, studies are looking at the syn- ical trials that specifically aim to understand this
ergistic effect of immunotherapy and radiother- synergism and the most effective scheduling of
apy in the management of melanoma brain treatments are pending; however, at this time,
combined radiation and immune therapies are J Neuro-Oncol. 2013;113(3):459–65. https://doi.

org/10.1007/s11060-013-1138-y.
an active and exciting area of research in the 9. Yamamoto M, Serizawa T, Shuto T, et al. Stereotactic
management of brain metastases. radiosurgery for patients with multiple brain metas-
tases (JLGK0901): a multi-institutional prospective
observational study. Lancet Oncol. 2014;15(4):387–
95. https://doi.org/10.1016/S1470-2045(14)70061-0.
Conclusion 10. Yamamoto M, Serizawa T, Higuchi Y, et al. A multi-
institutional prospective observational study of ste-
SRS is a versatile technique that should be used reotactic radiosurgery for patients with multiple brain
in the management of metastatic disease to the metastases (JLGK0901 study update): irradiation-
related complications and long-term maintenance of
brain. The ability to selectively deliver high lev- mini-mental state examination scores. Int J Radiat
els of concentrated radiation without irradiating Oncol Biol Phys. 2017;99(1):31–40. https://doi.
surrounding normal brain allows for excellent org/10.1016/j.ijrobp.2017.04.037.
rates of local control while minimizing toxicity. 11. Ali MA, Hirshman BR, Wilson B, et al. Survival
patterns of 5750 stereotactic radiosurgery–treated
Recent technical advances have made frameless patients with brain metastasis as a function of the
treatments with the Gamma Knife feasible and number of lesions. World Neurosurg. 2017;107:944–
successful, which will undoubtedly increase the 951.e1. https://doi.org/10.1016/j.wneu.2017.07.062.
applicability of SRS for a wide variety of indica- 12. Tsao MN, Xu W, Wong RK, et al. Whole brain radio-
therapy for the treatment of newly diagnosed mul-
tions, including fractionated treatments. tiple brain metastases. Cochrane Gynaecological,
Neuro-oncology and Orphan Cancer Group, ed.
Cochrane Database Syst Rev. 2018. https://doi.
org/10.1002/14651858.CD003869.pub4.
References 13. Davey P, Hoegler D, Ennis M, Smith J. A phase III
study of accelerated versus conventional hypofraction-
1. Laigle-Donadey F, Taillibert S, Martin-Duverneuil ated whole brain irradiation in patients of good per-
N, Hildebrand J, Delattre J-Y. Skull-base metasta- formance status with brain metastases not suitable for
ses. J Neuro-Oncol. 2005;75(1):63–9. https://doi. surgical excision. Radiother Oncol. 2008;88(2):173–
org/10.1007/s11060-004-8099-0. 6. https://doi.org/10.1016/j.radonc.2008.05.020.
2. Chamoun RB, DeMonte F. Management of skull base 14. Graham PH, Bucci J, Browne L. Randomized com-
metastases. Neurosurg Clin N Am. 2011;22(1):61–6. parison of whole brain radiotherapy, 20 Gy in four
https://doi.org/10.1016/j.nec.2010.08.005. daily fractions versus 40 Gy in 20 twice-daily frac-
3. Chin LS. Principles and practice of stereotactic radio- tions, for brain metastases. Int J Radiat Oncol Biol
surgery. New York: Springer; 2014. Phys. 2010;77(3):648–54. https://doi.org/10.1016/j.
4. Kotecha R, Angelov L, Barnett GH, et al. Calvarial ijrobp.2009.05.032.
and skull base metastases: expanding the clini- 15. Rapp SR, Case LD, Peiffer A, et al. Donepezil for
cal utility of Gamma Knife surgery. J Neurosurg. irradiated brain tumor survivors: a phase III ran-
2014;121:91–101. https://doi.org/10.3171/2014.7. domized placebo-controlled clinical trial. J Clin
GKS141272. Oncol. 2015;33(15):1653–9. https://doi.org/10.1200/
5. Shaw E, Scott C, Souhami L, et al. Single dose radio- JCO.2014.58.4508.
surgical treatment of recurrent previously irradiated 16. Brown PD, Pugh S, Laack NN, et al. Memantine for
primary brain tumors and brain metastases: final the prevention of cognitive dysfunction in patients
report of RTOG protocol 90-05. Int J Radiat Oncol receiving whole-brain radiotherapy: a randomized,
Biol Phys. 2000;47(2):291–8. https://doi.org/10.1016/ double-blind, placebo-controlled trial. Neuro Oncol.
S0360-3016(99)00507-6. 2013;15(10):1429–37. https://doi.org/10.1093/
6. Chao ST, Barnett GH, Vogelbaum MA, et al. Salvage neuonc/not114.
stereotactic radiosurgery effectively treats recur- 17. Gondi V, Pugh SL, Tome WA, et al. Preservation
rences from whole-brain radiation therapy. Cancer. of memory with conformal avoidance of the hippo-
2008;113(8):2198–204. https://doi.org/10.1002/ campal neural stem-cell compartment during whole-
cncr.23821. brain radiotherapy for brain metastases (RTOG
7. Maranzano E, Trippa F, Casale M, et al. Reirradiation 0933): a phase II multi-institutional trial. J Clin
of brain metastases with radiosurgery. Radiother Oncol. 2014;32(34):3810–6. https://doi.org/10.1200/
Oncol. 2012;102(2):192–7. https://doi.org/10.1016/j. JCO.2014.57.2909.
radonc.2011.07.018. 18. Patchell RA, Tibbs PA, Regine WF, et al. Postoperative
8. Yomo S, Hayashi M. The efficacy and limitations of radiotherapy in the treatment of single metastases to
stereotactic radiosurgery as a salvage treatment after the brain: a randomized trial. JAMA. 1998;280(17).
failed whole brain radiotherapy for brain metastases. https://doi.org/10.1001/jama.280.17.1485.
19. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized 2010;113(Special_Supplement):79–83. https://doi.
trial of surgery for single metastasis to the brain. N org/10.3171/2010.8.GKS10957.
Engl J Med. 1990;322(8):494–500. 31. Hsu F, Nichol A, Ma R, Kouhestani P, Toyota B,
20. Mahajan A, Ahmed S, McAleer MF, et al. Post-
McKenzie M. Stereotactic radiosurgery for metasta-
operative stereotactic radiosurgery versus observation ses in eloquent central brain locations. Can J Neurol
for completely resected brain metastases: a single- Sci. 2015;42(05):333–7. https://doi.org/10.1017/
centre, randomised, controlled, phase 3 trial. Lancet cjn.2015.55.
Oncol. 2017;18(8):1040–8. https://doi.org/10.1016/ 32. Laufer I, Rubin DG, Lis E, et al. The NOMS frame-
S1470-2045(17)30414-X. work: approach to the treatment of spinal metastatic
21.
Brown PD, Ballman KV, Cerhan JH, et al. tumors. Oncologist. 2013;18(6):744–51. https://doi.
Postoperative stereotactic radiosurgery compared org/10.1634/theoncologist.2012-0293.
with whole brain radiotherapy for resected metastatic 33. Yaeh A, Nanda T, Jani A, et al. Control of brain metas-
brain disease (NCCTG N107C/CEC·3): a multicen- tases from radioresistant tumors treated by stereotac-
tre, randomised, controlled, phase 3 trial. Lancet tic radiosurgery. J Neuro-Oncol. 2015;124(3):507–14.
Oncol. 2017;18(8):1049–60. https://doi.org/10.1016/ https://doi.org/10.1007/s11060-015-1871-5.
S1470-2045(17)30441-2. 34. Bedikian AY, Wei C, Detry M, et al. Predictive factors
22. Lamba N, Muskens IS, DiRisio AC, et al. Stereotactic for the development of brain metastasis in advanced
radiosurgery versus whole-brain radiotherapy after unresectable metastatic melanoma. Am J Clin
intracranial metastasis resection: a systematic review Oncol. 2011;34(6):603–10. https://doi.org/10.1097/
and meta-analysis. Radiat Oncol. 2017;12(1). https:// COC.0b013e3181f9456a.
doi.org/10.1186/s13014-017-0840-x. 35. Tawbi HA, Forsyth PA, Algazi A, et al. Combined
23. Ling DC, Vargo JA, Wegner RE, et al. Postoperative nivolumab and ipilimumab in melanoma metastatic to
stereotactic radiosurgery to the resection cavity for the brain. N Engl J Med. 2018;379(8):722–30. https://
large brain metastases: clinical outcomes, predictors doi.org/10.1056/NEJMoa1805453.
of intracranial failure, and implications for optimal 36. Goldberg SB, Gettinger SN, Mahajan A, et al.

patient selection. Neurosurgery. 2015;76(2):150–7. Pembrolizumab for patients with melanoma or non-
https://doi.org/10.1227/NEU.0000000000000584. small-cell lung cancer and untreated brain metastases:
24. Zhong J, Ferris MJ, Switchenko J, et al. Postoperative early analysis of a non-randomised, open-label, phase
stereotactic radiosurgery for resected brain metasta- 2 trial. Lancet Oncol. 2016;17(7):976–83. https://doi.
ses: a comparison of outcomes for large resection cav- org/10.1016/S1470-2045(16)30053-5.
ities. Pract Radiat Oncol. 2017;7(6):e419–25. https:// 37.
Margolin K, Ernstoff MS, Hamid O, et al.
doi.org/10.1016/j.prro.2017.04.016. Ipilimumab in patients with melanoma and brain
25. Vulpe H, Save AV, Xu Y, et al. Frameless stereotactic metastases: an open-label, phase 2 trial. Lancet
radiosurgery on the gamma knife ICON: early experi- Oncol. 2012;13(5):459–65. https://doi.org/10.1016/
ence from 100 patients. Neurosurgery. 2019. PMID S1470-2045(12)70090-6.
31375826. 38. Stokes WA, Binder DC, Jones BL, et al. Impact of
26. Trifiletti DM, Lee C-C, Winardi W, et al. Brainstem immunotherapy among patients with melanoma
metastases treated with stereotactic radiosurgery: brain metastases managed with radiotherapy. J
safety, efficacy, and dose response. J Neuro-Oncol. Neuroimmunol. 2017;313:118–22. https://doi.
2015;125(2):385–92. https://doi.org/10.1007/ org/10.1016/j.jneuroim.2017.10.006.
s11060-015-1927-6. 39. Qian JM, Yu JB, Kluger HM, Chiang VLS. Timing and
27. Trifiletti DM, Lee C-C, Kano H, et al. Stereotactic type of immune checkpoint therapy affect the early
radiosurgery for brainstem metastases: an interna- radiographic response of melanoma brain metastases
tional cooperative study to define response and toxic- to stereotactic radiosurgery: immunotherapy and SRS
ity. Int J Radiat Oncol Biol Phys. 2016;96(2):280–8. in brain metastases. Cancer. 2016;122(19):3051–8.
https://doi.org/10.1016/j.ijrobp.2016.06.009. https://doi.org/10.1002/cncr.30138.
28. Patel A, Mohammadi H, Dong T, et al. Brainstem 40. Kiess AP, Wolchok JD, Barker CA, et al. Stereotactic
metastases treated with Gamma Knife stereotactic radiosurgery for melanoma brain metastases in
radiosurgery: the Indiana University Health expe- patients receiving ipilimumab: safety profile and effi-
rience. CNS Oncol. 2018;7(1):15–23. https://doi. cacy of combined treatment. Int J Radiat Oncol Biol
org/10.2217/cns-2017-0029. Phys. 2015;92(2):368–75. https://doi.org/10.1016/j.
29. Murray L, Menard C, Zadeh G, et al. Radiosurgery ijrobp.2015.01.004.
for brainstem metastases with and without whole 41. Skrepnik T, Sundararajan S, Cui H, Stea B. Improved
brain radiotherapy: clinical series and literature time to disease progression in the brain in patients
review. J Radiat Oncol. 2017;6(1):21–30. https://doi. with melanoma brain metastases treated with con-
org/10.1007/s13566-016-0281-4. current delivery of radiosurgery and ipilimumab.
30. Dea N, Borduas M, Kenny B, Fortin D, Mathieu
Oncoimmunology. 2017;6(3):e1283461. https://doi.
D. Safety and efficacy of Gamma Knife surgery for org/10.1080/2162402X.2017.1283461.
brain metastases in eloquent locations. J Neurosurg.
Radiation Necrosis Following
the Radiosurgical Treatment
28
of Brain Metastases
Stephanie M. Robert and Veronica L. Chiang
Introduction Early-delayed (also known as pseudoprogres-

sion) is typically seen less than 12 weeks post-
Radiosurgery is a growing treatment strategy for radiation, and late effects occur months to years
numerous neurological pathologies, including after treatment. Most concerning of these effects
vascular lesions, brain tumors, trigeminal neural- are late AREs, which include leukoencephalopa-
gia, as well as functional procedures for the treat- thy and radiation necrosis. Unlike earlier effects,
ment of epilepsy, Parkinson’s disease, and late changes are typically irreversible and more
essential tremor. In the 1950s, Lars Leksell, a often symptomatic and progressive [2].
Swedish neurosurgeon, began using photon and Furthermore, because of the time course of devel-
proton beams directed into the brain, in an opment and appearance on imaging, differentiat-
attempt to treat neurological diseases. Over the ing radiation necrosis from tumor progression is
centuries, his protocol and methods has been an important and growing challenge.
refined and developed into what is now broadly
referred to as stereotactic radiosurgery (SRS) [1].
Although SRS is used to treat many different Development of Radiation Necrosis
neurological conditions, its use has grown expo-
nentially in the field of neuro-oncology as a result Although one of the most common side effects of
of improved cancer survival and increased sur- SRS, the true incidence of radiation necrosis
veillance imaging for intracranial metastasis. As remains unclear, largely due to the challenges that
more intracranial lesions are identified and exist in defining and diagnosing this pathology.
treated, radiation-induced side effects increase in The reported incidence of AREs ranges from 5%
tandem. The term adverse radiation effect (ARE) to 68%, depending on the imaging and clinical cri-
is a radiological definition used to describe these teria used. Furthermore, the duration of time
post-radiation changes identified by imaging patients are followed clinically, or with additional
modalities. AREs are further distinguished based neuroimaging, varies, and significant heterogene-
on length of time from exposure. Early/acute ity exists in the degree of symptomatology that
AREs occur within days of radiation exposure. manifests clinically. In addition, the incidence can
be highly variable depending on whether the num-
ber of cases of radiation necrosis is reported rela-
S. M. Robert · V. L. Chiang (*) tive to the number of lesions treated, the number of
Department of Neurosurgery, Yale University School patients treated, or the number of at risk patients
of Medicine, New Haven, CT, USA (i.e., only survivors) over a variety of time points
e-mail: Veronica.chiang@yale.edu

https://doi.org/10.1007/978-3-030-42958-4_28
394 S. M. Robert and V. L. Chiang
after radiosurgical treatment. Symptomatic cases increases the risk as well. This will be discussed
have been reported to occur less often with a less later in the chapter. Interestingly, Colaco et al., in
variable incidence ranging between 2% and 14% a single-institution retrospective review, found
[2, 3]. that while 37.5% of patients treated with systemic
Risk factors for the development of radiation immunotherapy developed radiation necrosis,
necrosis is an active area of investigation. Although 25.0% of patients receiving targeted therapy also
still poorly understood, a few independent factors developed radiation necrosis which was
have been identified. Treatment platform, significantly higher than the 16.9% rate in those
including LINAC and Gamma Knife, has not been receiving chemotherapy [12]. Of particular
shown to affect rates of radiation necrosis concern is the BRAF-inhibitor vemurafenib, a
development. The most consistently identified proven pre-clinical radiosensitizer. Patel et al.
factors found to be associated with development of reported that the rate of both radiographic and
radiation necrosis are increased dose of radiation, symptomatic radiation necrosis was significantly
larger volume of treated tissue – as measured by increased if SRS was administered concurrently
target volume or 12Gy (V12) and 10Gy (V10) with vemurafenib (radiographic – 22 vs. 11.1% at
volumes – and concurrent chemotherapy 1 year, p < 0.001; symptomatic 28.2 vs. 11.1%,
administration [4–7]. Furthermore, risk of p < 0.001) [13]. In comparison, second-generation
radiation necrosis also increases with repeated BRAF-inhibitor dabrafenib does not seem to have
SRS treatment. Sneed et al. showed that risk the same increased risks. Despite this, consensus
increases significantly with size and volume of guidelines from the Eastern Cooperative
lesion; however, a 20% 1-year risk of symptomatic Oncology Group recommend holding BRAFi
lesions was found with prior SRS to the same area. and/or MEK inhibitors for three days or more
In comparison, risk of prior whole brain before and after fractionated radiotherapy,
radiotherapy (WBRT) or concurrent WBRT are and one day or more before and after
4% and 8%, respectively [8]. radiosurgery [14].
The use of chemotherapy in conjunction with In addition, varied reports have also sug-
radiation to enhance tumor killing properties is gested that radiation necrosis rates are increased
often used in the treatment of cancers outside the in lung cancer patients with oncogenic driver
brain, such as neoadjuvant therapy for mutations (EGFR or ALK) or in those patients
gastrointestinal cancers, skull base tumors, and receiving tyrosine kinase inhibitors [15]. Kim
primary treatment of non-operative lung cancers, et al. (2017) retrospectively reviewed 1650
as well as many other malignancies. For treatment patients treated for 2843 brain metastases across
of brain metastases, combining chemotherapy all histologies and found that radiation necrosis
with whole brain radiation therapy (WBRT) developed in 8% of lesions overall [15, 16].
results in unacceptable rates of normal tissue Concurrent systemic therapy significantly
toxicity without improved survival [9] and increased the rate of radiation necrosis if admin-
therefore chemotherapy is traditionally put on istered with upfront SRS with WBRT (8.7%
hold while WBRT is being administered. With the compared with 3.7%, p = 0.04), and the specific
increasing use of SRS rather than WBRT, it is less agents most likely to be associated with radia-
clear how much separation is needed between tion necrosis were VEGFR tyrosine kinase
radiation treatment and chemotherapy. In inhibitors (TKIs) and EGFR TKIs (14.3% and
malignant glioma, SRS in combination with 15.6%, respectively, compared with 6% for non-
temozolomide increases the incidence of radiation TKIs). The differences were particularly notable
necrosis [10, 11]. Sneed et al. (2015) reported that when comparing cumulative incidences suggest-
the only chemotherapy agent to independently ing that the increased duration of survival in
increase the rate of radiation necrosis was patients receiving these agents probably also
capecitabine [8]. Systemic immunotherapy, contributed to their increased risk of develop-
frequently used in metastatic cancer, significantly ment of radiation necrosis.
28 Radiation Necrosis Following the Radiosurgical Treatment of Brain Metastases 395
Pathophysiology of Radiation coagulative necrosis. As a result of either isch-

Necrosis emia or intrinsic injury from radiation, oligo-
dendrocyte damage and demyelination also
Currently, no data is available on the pathophysiol- occur [18, 19]. Interestingly, these changes may
ogy of pseudoprogression. Radiation necrosis by explain the radiographic characteristics seen in
strict definition is the death of healthy tissue radiation necrosis that makes it difficult to
caused by radiation therapy; however, it is the distinguish from tumor progression.
downstream pathological side effect of this initial The resulting ischemia and cellular injury
tissue death that is now more loosely defined by likely induces the activation and release of microg-
this term. Histologically, the key changes found in lial, macrophage, and lymphocyte cytokines.
specimens containing radiation necrosis are Upregulation of pro-inflammatory cytokines, such
regions of coagulative necrosis surrounded by as IL-1 alpha, TNF-alpha, and IL-6, can initiate
demyelinated white matter containing vessels with chemokine networks such as the CXCL12/CXCR4
thickened, sclerosed and hyalinized walls, reactive axes, which then likely contributes to much of the
astrocytosis, and extensive macrophage infiltrates progressive nature demonstrated by these lesions
[3]. In addition, a process referred to as delayed [20]. Specifically, hypoxia-inducible factor-1
radiation-induced vasculitic leukoencephalopathy alpha (HIF-1α) is released by activated microglia,
(DRIVL) has been reported in radiation necrosis and may lead to upregulation of vascular endothe-
specimens. Diffuse infiltrates of both CD4+ and lial growth factor (VEGF). VEGF is known to be
CD8+ T lymphocytes are widely present in SRS- elevated in the areas of radiation necrosis [20], is
treated tissue. T cells are commonly found dif- detected as early as 4 weeks post-treatment, and
fusely scattered throughout the tissue; however, has been shown to increase over time in a mouse
Rauch et al. also demonstrate the presence of model of radiation necrosis [21]. VEGF overex-
transmural infiltration into small- and medium- pression is known to promote angiogenesis, result-
sized vessels, suggesting an immune-driven, active ing in the development of leaky blood vessels.
vasculitis present in SRS-treated tissue [17]. This effect contributes to the permeability of the
The underlying biology and pathophysiol- blood–brain barrier and the resulting vasogenic
ogy of radiation necrosis remains a widely edema seen with these lesions [22].
debated topic. Radiation necrosis has not only One of the most challenging aspects of radia-
been reported following the high dose radiation tion necrosis is to understand its delayed presenta-
treatment of both malignant glioma and brain tion and progressive nature. Realistically, given
metastases, but also following the radiosurgical that neither current model fully explains the histo-
treatment of arteriovenous malformations and logical changes seen in radiation necrosis tissue, it
other benign brain tumors. Because of this, the is likely that a combination of these mechanisms,
implicated mechanisms underlying the devel- as well as yet unknown mechanisms, underlies the
opment of radiation necrosis include radiation- true pathophysiology of this disease process.
induced neuronal/glial cell damage, vascular Furthermore, beyond vascular-induced changes,
injury, and immune-mediated changes. Two diffuse cellular infiltration, including T cells and
models have been proposed to explain the activated macrophages, is commonly seen in
development of radiation necrosis – primary SRS-treated tissue. These infiltrating macro-
injury to endothelial cells versus primary injury phages readily express pro- inflammatory cyto-
to glial cells (predominantly oligodendrocytes). kines such as tumor necrosis factor-alpha (TNFα)
Early studies of rodent and human tissue sup- and IL-6 [22], possibly supporting an important
port both vascular and glial injuries underlying immunological aspect of radiation necrosis.
the pathophysiology. Currently, it is thought Oligodendrocytes are exquisitely radiosensitive,
that endothelial injury is the initial insult, which resulting in demyelinating lesions seen in radiation
then leads to the development of intravascular necrosis [23]. Interestingly, one hypothesis that may
thrombosis and subsequent ischemia, causing explain the late-onset and progressive nature of the
disease is an underlying autoimmune etiology. As usually concurrent with radiographic regrowth of

oligodendrocytes are damaged, and as a result lyse the radiosurgically treated lesion. Just like with
open, exposure of the brain’s immune cells to tumor growth, the symptoms of radiation necrosis
released intracellular components, such as myelin can be highly variable but are typically either
basic protein (MBP), which can be detected in cere- related to increased intracranial pressure (head-
brospinal fluid months after radiation exposure, ache, confusion, and altered mental status), focal
may trigger ongoing demyelination and further neurological dysfunction (such as motor weakness,
inflammation [17]. Furthermore, albeit rare, a few sensory loss, dysphasia, gait imbalance), or sei-
cases of radiation necrosis remote from the site of zures. Further, as with tumor, symptomatology is
SRS treatment have been reported [23, 24]. The determined by the size, extent of edema, and loca-
involvement of the corpus callosum and periven- tion in the brain, rather than the underlying pathol-
tricular subependymal areas in these cases further ogy. It has been suggested that fatigue and cognitive
supports oligodendroglial injury from an autoim- dysfunction in areas such as memory and concen-
mune-type response after initial SRS exposure. tration are more likely related to radiation injury,
Lastly, the reporting of the vasculitic changes seen but this data has not been validated [25]. In the
in the entity called DRIVL, as discussed above, majority of patients, initiation of steroids and anti-
may also lend support to the autoimmune theory. convulsants results in good initial control of symp-
Further work is necessary to identify the antigen(s) toms, regardless of underlying pathology.
to which the T-cells are primarily reactive. Given that symptomatology often does not dif-
Regardless of the underlying pathophysiology, ferentiate tumor regrowth from radiation necrosis,
radiation necrosis, then, is defined as a high-dose it is important then to try to accurately differen-
radiation-induced, self-perpetuating, inflammatory, tiate radiation necrosis from tumor recurrence
and demyelinating process that surrounds a central by some reliable method to appropriately guide
core of coagulative necrosis. management and treatment decisions for patients.
Conventional, first-line imaging, which consists of
computerized tomography (CT) and magnetic res-
Diagnosing Radiation Necrosis onance imaging (MRI), often shows indistinguish-
able imaging changes for both tumor recurrence
The dilemma in the diagnosis of pseudoprogres- and radiation necrosis. One of the challenges in
sion and radiation necrosis is differentiating it using MRI to differentiate tumor versus radiation
from tumor recurrence after initial response to necrosis lies in the similarities in the pathophysi-
treatment with radiosurgery. As discussed, ology of these lesions. MRI analysis of intracra-
pseudoprogression and radiation necrosis are nial tumors relies heavily on the breakdown of the
thought to be similar entities, differentiated by the blood–brain barrier in regions with malignancy-
timing of their presentation. Both tumor and induced neovascularization, which appears as con-
radiation necrosis then can manifest similarly, trast enhancement on imaging. Radiation necrosis,
both clinically and radiographically; however, as an inflammatory process, similarly disrupts the
treatment options for the two diagnoses are very blood–brain barrier, therefore showing similar
different. What can also be very different about enhancement on contrast-enhanced MRI scans
the two diagnoses is their potential course. With [26]. Characteristically, gadolinium-enhanced
tumor recurrence, growth will be persistent and T1-weighted MRI shows a rim-enhancing lesion
always requires intervention, whereas many cases with central necrosis with mass effect, and FLAIR
of radiation necrosis can be self-resolving even if imaging will show significant perilesional edema
presenting with symptomatology, and therefore around the area previously targeted by SRS. As
may not always require intervention. previously described, central necrosis is seen in all
Interestingly, patients with radiation necrosis cases of radiation necrosis, but many tumors also
many develop significant changes on imaging, contain significant areas of necrosis, further con-
without developing concurrent symptomatology fusing interpretation on imaging. Lastly, because
[25]. If they do occur, development of symptoms is of its delayed onset, radiation necrosis typically
presents in a similar time period as does tumor this finding, a subsequent publication by Nath et al.
regrowth, therefore, timing of lesion development showed that only patients in whom pathology
is not a realiable indicator to differentiate between showed absolutely no tumor had increased survival
these pathologies. and that even in patients with 2% tumor in their
In the largest study of brain metastasis-related specimen subsequently went on to have tumor pro-
radiation necrosis by Sneed et al. [8], the authors gression at the previously treated site [28].
showed that of their 2200 metastatic lesions treated, Extensive research continues to focus on iden-
9.2% of regrowth was tumor recurrence, 5.4% was tifying new imaging modalities and protocols to
isolated radiation necrosis, and 1.4% was a combi- accurately diagnose radiation necrosis without
nation of tumor and necrosis. The finding that in a the need for invasive procedures to obtain tis-
certain percentage of cases these lesions likely con- sue for pathological evaluation. Advances in the
tain a component of both radiation necrosis and development of novel MRI sequences are leading
tumor tissue further complicates diagnosis. Further, the field currently, with new protocols and tech-
the percentage of lesions that may show mixed niques such as perfusion protocols and spectros-
pathology may vary across institutions depending copy. As yet, however, there is no gold standard
on the radiosurgical dosing across institutions imaging technique that provides sufficient accu-
depending on the radiosurgical dosing, surgical racy in predicting radiation necrosis versus tumor
aggressiveness, and a host of other factors. In a regrowth in the clinical setting.
study by Alomari et al., 36 consecutively identified Based on T1 gadolinium-enhanced MRI, use
lesions that regrew after radiosurgical treatment of lesional morphology has been well document
required surgical resection for either diagnosis or to frequently be insufficient for differentiation of
management of symptomatology [27]. Pathology the two diagnoses [29] although some lesions do
was consistent with tumor regrowth in 31% and show the characteristic “Swiss cheese,” “cut bell
radiation necrosis in 36%; for the remaining 33%, pepper,” or “soap bubble” changes (Fig. 28.1)
there was less than 2% of tumor within the speci- described by Kumar et al. in gliomas treated
mens. In determining the clinical significance of using high-dose fractionated radiation [23].
Soap-bubble Cut bell pepper

Swiss-cheese
Fig. 28.1 Lesional morphology of radiation necrosis on of radiation necrosis lesions. (These imaging characteris-
T1 gadolinium-enhanced MRI. The characteristic “Swiss tics were described by Kumar et al. [23] in gliomas treated
cheese,” “soap bubble,” and “cut bell pepper” appearance with radiosurgery)
Apparent diffusion coefficient (ADC) is an MRI Cho/NAA and Cho/Cr ratios, and lower NAA/Cr
sequence based on diffusion-weighted imaging ratios [23, 33]. However, there is a significant
that was initially hypothesized to differentiate variability in the ability of MR spectroscopy and
tumor versus radiation necrosis based on hyper- MR perfusion to accurately predict tumor versus
cellularity of tumor-bearing regions relative to radiation necrosis in the literature, and therefore
necrotic areas. In theory, tumor recurrence should their sensitivity and specificity are currently
cause restriction of water diffusion due to the debated [34, 35].
increased number of cells compared to radiation Lastly, nuclear medicine imaging is becoming
necrosis. However, recurrent tumors have vari- more widely used, and is specifically being
able cellularity, and many regrowths have signifi- investigated for use in identifying tumors and
cant necrotic cores, preventing the development metastatic lesions. Techniques using positron
of clear parameters for diagnosis [30]. Similarly, emission tomography (PET) in combination with
T1/2 mismatch, initially developed and proposed 2-deoxy-[18F]fluoro-D-glucose (FDG) allows
by Dequesada et al., has been ultimately shown measurement of cellular metabolism by detecting
to have a poor positive predictive value in distin- glucose uptake. Malignant lesions, with higher
guishing radiation necrosis from tumor recur- metabolic activity, show up as brighter lesions
rence [31, 32]. given their enhanced glucose uptake [36].
The most promising MRI-based technique Unfortunately, due to the high basal glucose
currently available takes advantages of the subtle metabolism in brain tissue, FDG-PET is not as
differences in pathophysiology of radiation useful as in extracranial lesions. PET scans using
necrosis and tumor regrowth, using MR spectros- radiolabeled amino acids have shown promise at
copy combined with MR perfusion. These several centers for differentiating radiation necro-
modalities measure metabolism and physiology, sis and tumor, specifically with the 11C methio-
and together are more likely to accurately predict nine-PET-labeled isotope. However, due to its
the pathology of the lesion. MR perfusion uses short half-life and need for an on-site cyclotron,
dynamic susceptibility-weighted contrast- its use remains relatively limited [37].
enhanced imaging to determine relative cerebral Given the lack of any specific imaging test to
blood volume (rCBV). As a measure of micro- reliably differentiate radiation necrosis from
vascular density, rCBV is typically decreased in tumor regrowth, our institution still relies heavily
radiation necrosis given the small-vessel injury on histological analysis of tissue as studies have
that underlies its development. In contrast, tumor shown accuracy to be >98% to distinguish radia-
regrowth promotes angiogenesis for cell nutrition tion necrosis from tumor recurrence [38–40]. In
and survival, therefore enhancing rCBV and those patients in whom surgical management is
increasing perfusion on MRI [26]. not possible, serial imaging with standard
MR spectroscopy evaluates the chemical T1-weighted gadolinium-enhanced MRI and
makeup of the brain tissue, and using the ratio of FLAIR sequence MRI over a 3- to 6-month
metabolites N-acetyl aspartate (NAA), choline period ultimately also allows for differentiation
(Cho), and creatine (Cr) measured, further insight of diagnosis with the majority of cases of radia-
into the pathology of lesions can be gathered. tion necrosis being self-limited.
NAA is a marker of neuronal function, and a
decrease in this metabolite indicates neuronal
injury. Increase in choline suggests increased cel- Surveillance for Radiation Necrosis
lular proliferation, as it is a component of the cel-
lular membrane. Creatine, a marker of energy As seen in the radiological study by Patel et al.
reserve, remains stable and is used as a control (2011), of 500 metastases after SRS treatment,
for comparison. To differentiate tumor versus the authors found that up to one-third of SRS-
radiation necrosis, Cho, NAA, and Cr ratios are treated lesions regrew after an initial favorable
compared. Tumor regrowth tends to have higher radiographic response [41]. Radiation necrosis
can develop as early as 2 months post-radiation, to follow up any concerning CT findings with
and as far out as 10 years after exposure, although an MRI of the brain.
imaging changes begin most commonly between
7 and 11 months post-SRS [7, 42–44].
In a recent study by Fujimoto et al. (2018), the Immunotherapy and Radiation
authors retrospectively reviewed imaging for all Necrosis
their radiosurgically treated brain metastases
patients to identify patients who had lesional Monoclonal antibody immunotherapies, specifi-
regrowth beyond 18 months after radiation. A cally PD-1 inhibitors alone or in combination
total of 13 patients with 19 problematic lesions with CTLA4 inhibitors, are being increasingly
were identified with a median follow-up of used in many types of cancer but most commonly
48.2 months and a median overall survival of currently for patients with small cell and non-
73 months. Of the 19 lesions, 12 were identified small cell lung cancer, melanoma, and renal cell
as radiation necrosis (either by demonstration of carcinoma. Brain metastases occur in all of these
spontaneous resolution on serial imaging or by cancer types and therefore, as SRS is also being
pathology) and 7 were tumor recurrence. The increasing used as a therapeutic modality for
radiation necrosis cases demonstrated first con- brain metastases, the interaction between immu-
cerning imaging changes at a median of notherapies and the development of radiation
33.2 months (range 18.5–63.2 months) and there- necrosis is an important new focus of research.
fore the latest case occurred at 5.3 years. In com- Studies have shown that radiation therapy
parison, the tumor recurrence cases occurred at a enhances the innate immune response within
median time of 23.6 months (range 19.8– tumors as well as within the immune system of
45.3 months), with the latest case occurring at the patient [46], and that combining SRS with
3.8 years [45]. immune checkpoint inhibitors may enhance anti-
This study highlights the need for continued tumor activity [47, 48].
imaging surveillance for both tumor recurrence These immune therapies, also known as
and radiation necrosis and the need for recog- checkpoint inhibitors, block inhibitory check-
nition that both diagnoses can still occur many points in the immune response, allowing the
years after radiosurgical treatment. This issue immune system to generate a more robust
will become increasingly important as patients response to malignant lesions, enhancing anti-
with brain metastases experience improving tumor activity [49]. The brain has traditionally
survival durations from newer systemic thera- been known as an “immunologically privileged
pies that also may have variable central ner- site” referring to its ability to limit the entrance
vous system penetration. Within our institution, of peripheral immune cells due to the presence
post-radiosurgical surveillance includes serial of the blood–brain barrier (BBB). Activated T
MR imaging at 6 weeks, and then 3, 6, 9, 12, cells however have been shown to be able to
18, and 24 months after radiosurgical treat- cross this barrier [50], and furthermore it has
ment as long as systemic control is maintained. been suggested that not only do brain metasta-
MR imaging is performed more frequently ses have a disrupted BBB, but SRS may further
(usually every 6 weeks) if there is recurrence disrupt the BBB, allowing enhanced crossing of
of systemic disease or if neurological symp- peripheral immune cells into the central nervous
toms arise and are associated with new changes system (CNS), an effect that in animal studies
on MRI. Beyond 2 years, yearly MRIs are has been found to last as long as a month post-
advisable. Given that the first change often treatment [51]. Recent data by Qian et al. [52]
seen on MRI is the development of perilesional has shown an improved therapeutic response
edema, it may also be reasonable to perform of brain metastases in melanoma patients
surveillance imaging using serial CT scans of when radiosurgery is delivered concurrently
the brain to minimize imaging costs and then with immunotherapy, supporting the idea of
increased T-cell penetrance and/or efficacy in cal variables in the care of each patient, may be
the brain following r adiosurgery [52]. required to determine the true interactive effect of
The negative consequences of combining SRS immunotherapy and radiosurgery. In our own cen-
and immunotherapy, however, remain unclear. ter, we see examples of radiation necrosis develop-
Given that one of the pathophysiological mecha- ing immediately after GKSRS + immunotherapy
nisms proposed in the development of radiation treatment, as well as delayed effects (Fig. 28.2).
necrosis was autoimmune reactivity, it is feasible Interestingly, radiation necrosis is seen to develop
that immunotherapy may accentuate the risk for immediately after immunotherapy (Fig. 28.2a), in a
development of radiation necrosis. During the era delayed fashion (Fig. 28.2b), and even years later,
when the anti-CTLA4 agent ipilimumab was seen in long-term survivors (Fig. 28.1c). Although
commonly used for melanoma, there was no data is suggesting significant interactions between
clear evidence that risk of radiation necrosis was immunotherapy and radiation necrosis, more
increased, although a study by Colaco et al. research is needed to provide insight into the mech-
(2016) showed that when all immunotherapies anisms underlying the intricate interaction between
were aggregated together, the majority of which immunotherapy and SRS in the treatment of malig-
included ipilimumab, the rates of radiation necro- nant lesions and development of radiation
sis were higher than the rates seen in patients necrosis.
treated with either targeted therapies or chemo-
therapies [12]. This finding however was not sup-
ported by a recent study by Diao et al. (2018) Treatment of Radiation Necrosis
who studied acute toxicity specifically in 91
patients treated concurrently with ipilimumab Interestingly, and unlike tumor regrowth, many
and radiosurgery [53]. radiation-induced lesions spontaneously resolve
Similarly discrepant results are being reported without intervention. Although no clinical or
for anti-PD-1 and combination immunotherapies imaging parameters have been shown to predict
used in conjunction with radiosurgery, uncover- the likely evolution of the lesion, the likelihood
ing an association between the development of of improvement of radiation necrosis increases
radiation necrosis and the use of concurrent with time, with up to 76% of lesions resolving
immunotherapy. A recent study by Martin et al. 18 months after initial diagnosis. However, up to
(2018) found a higher incidence of radiation 25% of these lesions may not improve with
necrosis in patients treated with combined medical treatment alone, necessitating surgical
immunotherapy and SRS. Their patient popula- intervention [42].
tion included those with lung cancer, melanoma, Management and treatment of radiation necro-
and renal cell cancer, and interestingly, the sis therefore varies depending on symptomatology.
increased risk of radiation necrosis was heavily If the patient’s clinical scenario and imaging seem
biased toward the melanoma population [54]. most consistent with radiation necrosis, then
Similar results were found by Kaider-Person asymptomatic, small, and non-progressive lesions
et al. (2017) in their 58-patient population with can be managed conservatively.
melanoma brain metastases [55]. In contrast, First-line medical therapy for symptomatic
Fang et al. did not find any increase in the rate of radiation necrosis is corticosteroids. Steroids are
radiation necrosis in their study of 137 mela- presumed to contribute anti-inflammatory effects,
noma patients treated with radiosurgery to 1094 stabilize the blood–brain barrier, and decrease
lesions especially when compared to rates of edema. Steroids, however, are not required unless
radiation necrosis seen in patients treated with symptoms arise, as steroids alone have not been
chemotherapy [56]. shown to change the course of radiation necrosis.
Given the conflicting data currently in the litera- If symptoms arise, however, then the lowest dose
ture, larger multicentered studies, particularly of steroids allowing management of symptoms
either standardized or stratified for the many clini- should be administered and recurrent attempts to
Pembrolizumab
a Oct 2014
April 2014 August 2014 Nov 2014 Feb 2015
6/2015 12/2015 1/2016
GK 12/2014 10/2015 1/2016
Fig. 28.2 Immunotherapy and the development of radia- apy. He showed good response for 8 months; however, he
tion necrosis in three patients with melanoma. The first then developed radiation necrosis in the GKSRS cavity
patient (a) had previously received Gamma Knife stereo- only. This lesion was then surgically resected. For the last
tactic radiosurgery (GKSRS) several times with improve- patient (c), GKSRS was provided to a single lesion, fol-
ment in her right temporal lesion. However, 1 month after lowed by immunotherapy, which provided good response
receiving pembrolizumab, she developed progressive for 2 years until the development of radiation necrosis in
radiation necrosis that required resection. In the second the previous area of GKSRS treated lesion. This phenom-
patient (b), the initial lesion was resected, with GKSRS to enon is seen in many long-term survivors of peripheral
the post-op cavity only, in combination with immunother- cancers
wean steroids should occur until the lesion or is preferred for readily accessible lesions in the
symptoms resolve [26]. Unfortunately, some brain, and for patients healthy enough to undergo
patients are unable to tolerate this medication due a more invasive surgical procedure. Furthermore,
to its side effects, and others continue to experi- surgical resection has been shown to provide
ence symptoms in spite of treatment. complete local control of the lesion [28].
Bevacizumab, a humanized monoclonal anti- For less surgically accessible lesions, a more
body, has shown efficacy in studies of central ner- recent technique called laser interstitial
vous system radiation necrosis. While it is unclear thermotherapy (LITT) has been developed to
how bevacizumab works to resolve radiation necro- address both the need for tissue diagnosis and the
sis, as a VEGF-inhibitor, it is a highly effective nontreatment of radiation necrosis. Since its
steroidal therapy for management of perilesional introduction, it has been increasingly used for
edema. In a randomized, placebo-controlled study patients with lesions amenable to standard
of 14 patients with radiation necrosis, all treated craniotomy because of its minimally invasive
patients showed improvement, both on imaging and technique. Using this procedure, through a small
symptomatically, while none of the patients in the scalp incision and a 5 mm twist drill hole made in
placebo group showed improvement [57]. Similarly, the skull, a biopsy can first be obtained to provide
Gonzalez et al. (2007) showed that treatment with a diagnosis, and then the lesion can be ablated
bevacizumab reduced the MRI fluid-attenuated using a diode laser introduced into the center of
inversion- recovery (FLAIR) abnormalities and the lesion. Light emitted from the laser is
T1-weighted post-Gd-contrast abnormalities in converted in the surrounding tissues into heat.
radiation necrosis. These findings suggest that ves- The progression of heat delivery is monitored
sel leakage and associated edema is decreased as a using continuous intra-operative MR gradient
result of this treatment strategy. Bevacizumab use echo imaging, and through the use of proprietary
also enabled a reduction in daily dexamethasone software, a real-time ablation map is created
dose required for these patients [58]. Notably, how- corresponding to the amount of time each
ever, this treatment strategy has significant side imaging voxel has been at an elevated temperature.
effects and again is not tolerated by all patients. One of the significant advantages of using
Aspirin, non-steroidal agents, anticoagula- LITT for regrowing lesions after SRS is the
tion, and vitamin E supplementation have all success it has shown in treating both radiation
been anecdotally reported to have efficacy in the necrosis and regrowing metastatic lesions.
treatment of radiation necrosis although consis- Multiple retrospective studies have demonstrated
tent results have not been replicable. Hyperbaric the efficacy of LITT regardless of diagnosis [59,
oxygen is a less commonly used treatment, in 60]. Furthermore, since it provides a minimally
large part because of the limitations of delivery. invasive alternative to an open surgical approach,
This therapy enhances angiogenesis in hypoxic it allows more patients to undergo treatment.
tissue and oxygen delivery. Although no large- Many cancer patients with metastatic lesions are
scale studies have been performed, smaller trials poor surgical candidates or have lesions located
have also shown improvement in imaging and in brain regions too deep to justify the morbidity
symptoms in patients undergoing hyperbaric that would result from open surgical resection. A
oxygen [26]. recent prospective, multi-center study (Laser
For lesions that do not respond to medical Ablation After Stereotactic Radiosurgery
therapies, or those in whom a tissue diagnosis is [LAASR]), looked at longer-term results of using
needed to rule out regrowing tumor, surgical LITT for regrowing SRS-treated lesions.
resection can be performed. In addition to Interestingly, they found that for biopsy-
providing definitive pathology, removal of the confirmed radiation necrosis lesions, local
lesion is frequently the most rapid mechanism to control at 6 months was 100%; however, for
relieve neurological symptoms by immediately regrowing tumor lesions, control was less at 74%.
reducing mass effect and edema, and allowing for These findings suggest that LITT is an effective
a quicker taper of steroid dosing. This approach option for treatment of both radiation necrosis
and tumor regrowth, with the caveat that further comes, and risk factors with emphasis on radiation
surveillance and treatment may be needed post- parameters and chemotherapy. Int J Radiat Oncol Biol
Phys. 2006;65(2):499–508.
LITT for recurrent metastatic lesions [20]. 5. Blonigen BJ, Steinmetz RD, Levin L, Lamba MA,
Warnick RE, Breneman JC. Irradiated volume as a
predictor of brain radionecrosis after linear accelera-
tor stereotactic radiosurgery. Int J Radiat Oncol Biol
Conclusion Phys. 2010;77(4):996–1001.
6. Korytko T, Radivoyevitch T, Colussi V, Wessels BW,
Radiation necrosis is a growing problem due in Pillai K, Maciunas RJ, et al. 12 Gy gamma knife
large part to the success of SRS in treating a vast radiosurgical volume is a predictor for radiation
necrosis in non-AVM intracranial tumors. Int J Radiat
number of intracranial pathologies. Although Oncol Biol Phys. 2006;64(2):419–24.
SRS-induced radiation necrosis has been 7. Minniti G, Clarke E, Lanzetta G, Osti MF, Trasimeni
described following treatment for benign and G, Bozzao A, et al. Stereotactic radiosurgery for brain
non-tumor lesions, given its increasing use in metastases: analysis of outcome and risk of brain
radionecrosis. Radiat Oncol. 2011;6:48.
malignant primary and metastatic brain tumors, 8. Sneed PK, Mendez J, Vemer-van den Hoek JGM,
important questions remain unanswered for the Seymour ZA, Ma L, Molinaro AM, et al. Adverse
field of neuro-oncology. radiation effect after stereotactic radiosurgery for
Furthermore, newer evidence strongly sug- brain metastases: incidence, time course, and risk fac-
tors. J Neurosurg. 2015;123(2):373–86.
gests a significant immunologically mediated 9. Qin H, Pan F, Li J, Zhang X, Liang H, Ruan
component underlying the development of Z. Whole brain radiotherapy plus concurrent che-
radiation necrosis. Combining the current motherapy in non-small cell lung cancer patients
proposed models, immune activation likely with brain metastases: a meta-analysis. PLoS One.
2014;9(10):e111475.
occurs in response to the vascular and cellular 10. Brandes AA, Franceschi E, Tosoni A, Blatt V, Pession
injury created by radiation treatment. As such, as A, Tallini G, et al. MGMT promoter methylation sta-
both radiosurgery and immunotherapy/targeted tus can predict the incidence and outcome of pseudo-
therapies become more heavily utilized and progression after concomitant radiochemotherapy in
newly diagnosed glioblastoma patients. J Clin Oncol.
combined clinically, and as more metastatic 2008;26(13):2192–7.
brain lesions are identified and treated, the 11. Chamberlain MC, Glantz MJ, Chalmers L, Van Horn
incidence of radiation necrosis will likely A, Sloan AE. Early necrosis following concurrent
continue to grow. Fortunately, there are a number Temodar and radiotherapy in patients with glioblas-
toma. J Neuro-Oncol. 2007;82(1):81–3.
of effective treatment options for patients to treat 12. Colaco RJ, Martin P, Kluger HM, Yu JB, Chiang
radiation necrosis. However, a better VL. Does immunotherapy increase the rate of radia-
understanding of the pathophysiology of tion necrosis after radiosurgical treatment of brain
radiation necrosis is needed to guide treatment metastases? J Neurosurg. 2016;125(1):17–23.
13. Patel BG, Ahmed KA, Johnstone PAS, Yu H-HM,
decisions and patient management, both before Etame AB. Initial experience with combined BRAF
radiation necrosis develops, as well as how best and MEK inhibition with stereotactic radiosurgery for
to treat once it occurs. BRAF mutant melanoma brain metastases. Melanoma
Res. 2016;26(4):382–6.
14. Anker CJ, Grossmann KF, Atkins MB, Suneja G,
Tarhini AA, Kirkwood JM. Avoiding severe toxicity
from combined BRAF inhibitor and radiation treat-
References ment: consensus guidelines from the eastern coopera-
tive oncology group (ECOG). Int J Radiat Oncol Biol
1. Leksell L. Stereotactic radiosurgery. J Neurol Phys. 2016;95(2):632–46.
Neurosurg Psychiatry. 1983;46(9):797–803. 15. Miller JA, Balagamwala EH, Angelov L, Suh JH,
2. Brandsma D, Stalpers L, Taal W, Sminia P, van den Yang K, Tariq MB, et al. The impact of receptor status
Bent MJ. Clinical features, mechanisms, and man- on local control of breast metastases following spine
agement of pseudoprogression in malignant gliomas. stereotactic radiosurgery. Int J Radiat Oncol Biol
Lancet Oncol. 2008;9(5):453–61. Phys. 2016;96(2):E88.
3. Yoshii Y. Pathological review of late cerebral radione- 16. Kim JM, Miller JA, Kotecha R, Xiao R, Juloori A,
crosis. Brain Tumor Pathol. 2008;25(2):51–8. Ward MC, et al. The risk of radiation necrosis fol-
4. Ruben JD, Dally M, Bailey M, Smith R, McLean CA, lowing stereotactic radiosurgery with concurrent sys-
Fedele P. Cerebral radiation necrosis: incidence, out- temic therapies. J Neuro-Oncol. 2017;133(2):357–68.
17. Rauch PJ, Park HS, Knisely JPS, Chiang VL,

A radiographic-pathological study. Neurosurgery.
Vortmeyer AO. Delayed radiation-induced vasculitic 2008;63(5):898–903; discussion 904.
leukoencephalopathy. Int J Radiat Oncol Biol Phys. 32. Stockham AL, Tievsky AL, Koyfman SA, Reddy CA,
2012;83(1):369–75. Suh JH, Vogelbaum MA, et al. Conventional MRI
18. Rahmathulla G, Marko NF, Weil RJ. Cerebral radia- does not reliably distinguish radiation necrosis from
tion necrosis: a review of the pathobiology, diagno- tumor recurrence after stereotactic radiosurgery. J
sis and management considerations. J Clin Neurosci. Neuro-Oncol. 2012;109(1):149–58.
2013;20(4):485–502. 33. Zhang H, Ma L, Wang Q, Zheng X, Wu C, Xu B. Role
19. Chao ST, Ahluwalia MS, Barnett GH, Stevens GHJ, of magnetic resonance spectroscopy for the differen-
Murphy ES, Stockham AL, et al. Challenges with the tiation of recurrent glioma from radiation necrosis: a
diagnosis and treatment of cerebral radiation necrosis. systematic review and meta-analysis. Eur J Radiol.
Int J Radiat Oncol Biol Phys. 2013;87(3):449–57. 2014;83(12):2181–9.
20. Furuse M, Nonoguchi N, Kawabata S, Miyatake S-I, 34. van Dijken BRJ, van Laar PJ, Holtman GA, van der
Kuroiwa T. Delayed brain radiation necrosis: patho- Hoorn A. Diagnostic accuracy of magnetic reso-
logical review and new molecular targets for treat- nance imaging techniques for treatment response
ment. Med Mol Morphol. 2015;48(4):183–90. evaluation in patients with high-grade glioma, a
21. Perez-Torres CJ, Yuan L, Schmidt RE, Rich KM,
systematic review and meta-analysis. Eur Radiol.
Drzymala RE, Hallahan DE, et al. Specificity of vas- 2017;27(10):4129–44.
cular endothelial growth factor treatment for radiation 35. Chuang M-T, Liu Y-S, Tsai Y-S, Chen Y-C, Wang
necrosis. Radiother Oncol. 2015;117(2):382–5. C-K. Differentiating radiation-induced necro-
22. Kureshi SA, Hofman FM, Schneider JH, Chin
sis from recurrent brain tumor using MR perfu-
LS, Apuzzo ML, Hinton DR. Cytokine expression and spectroscopy: a meta-analysis. PLoS One.
sion in radiation- induced delayed cerebral injury. 2016;11(1):e0141438.
Neurosurgery. 1994;35(5):822–9; discussion 829. 36. Belohlávek O, Simonová G, Kantorová I, Novotný
23. Kumar AJ, Leeds NE, Fuller GN, Van Tassel P, Maor J, Liscák R. Brain metastases after stereotactic
MH, Sawaya RE, et al. Malignant gliomas: MR imag- radiosurgery using the Leksell gamma knife: can
ing spectrum of radiation therapy- and chemotherapy- FDG PET help to differentiate radionecrosis from
induced necrosis of the brain after treatment. tumour progression? Eur J Nucl Med Mol Imaging.
Radiology. 2000;217(2):377–84. 2003;30(1):96–100.
24. Tsuruda JS, Kortman KE, Bradley WG, Wheeler
37. Tomura N, Kokubun M, Saginoya T, Mizuno Y,

DC, Van Dalsem W, Bradley TP. Radiation effects Kikuchi Y. Differentiation between treatment-induced
on cerebral white matter: MR evaluation. AJR Am J necrosis and recurrent tumors in patients with
Roentgenol. 1987;149(1):165–71. metastatic brain tumors: comparison among
25. Giglio P, Gilbert MR. Cerebral radiation necrosis. 11C-methionine-PET, FDG-PET, MR permeability
Neurologist. 2003;9(4):180–8. imaging, and MRI-ADC-preliminary results. AJNR
26. Patel U, Patel A, Cobb C, Benkers T, Vermeulen
Am J Neuroradiol. 2017;38(8):1520–7.
S. The management of brain necrosis as a result of 38. Heper AO, Erden E, Savas A, Ceyhan K, Erden I,
SRS treatment for intra-cranial tumors. Transl Cancer Akyar S, et al. An analysis of stereotactic biopsy of
Res. 2014;3:373–82. brain tumors and nonneoplastic lesions: a prospective
27. Alomari A, Rauch PJ, Orsaria M, Minja FJ, Chiang clinicopathologic study. Surg Neurol. 2005;64(Suppl
VL, Vortmeyer AO. Radiologic and histologic con- 2):S82–8.
sequences of radiosurgery for brain tumors. J Neuro- 39. Kreth FW, Muacevic A, Medele R, Bise K, Meyer
Oncol. 2014;117(1):33–42. T, Reulen HJ. The risk of haemorrhage after image
28. Nath SK, Sheridan AD, Rauch PJ, Yu JB, Minja FJ, guided stereotactic biopsy of intra-axial brain
Vortmeyer AO, et al. Significance of histology in tumours--a prospective study. Acta Neurochir.
determining management of lesions regrowing after 2001;143(6):539–45; discussion 545.
radiosurgery. J Neuro-Oncol. 2014;117(2):303–10. 40. Narloch JL, Farber SH, Sammons S, McSherry F,
29. Dooms GC, Hecht S, Brant-Zawadzki M, Berthiaume Herndon JE, Hoang JK, et al. Biopsy of enlarging
Y, Norman D, Newton TH. Brain radiation lesions: lesions after stereotactic radiosurgery for brain metas-
MR imaging. Radiology. 1986;158(1):149–55. tases frequently reveals radiation necrosis. Neuro-
30. Detsky JS, Keith J, Conklin J, Symons S, Myrehaug Oncology. 2017;19(10):1391–7.
S, Sahgal A, et al. Differentiating radiation necrosis 41. Patel TR, McHugh BJ, Bi WL, Minja FJ, Knisely
from tumor progression in brain metastases treated JPS, Chiang VL. A comprehensive review of
with stereotactic radiotherapy: utility of intravoxel MR imaging changes following radiosurgery to
incoherent motion perfusion MRI and correlation with 500 brain metastases. AJNR Am J Neuroradiol.
histopathology. J Neuro-Oncol. 2017;134(2):433–41. 2011;32(10):1885–92.
31. Dequesada IM, Quisling RG, Yachnis A, Friedman 42. Chin LS, Ma L, DiBiase S. Radiation necrosis following
WA. Can standard magnetic resonance imaging gamma knife surgery: a case-controlled comparison of
reliably distinguish recurrent tumor from radiation treatment parameters and long-term clinical follow up. J
necrosis after radiosurgery for brain metastases? Neurosurg. 2001;94(6):899–904.
43. Kohutek ZA, Yamada Y, Chan TA, Brennan CW,

early radiographic response of melanoma brain
Tabar V, Gutin PH, et al. Long-term risk of radio- metastases to stereotactic radiosurgery. Cancer.
necrosis and imaging changes after stereotactic 2016;122(19):3051–8.
radiosurgery for brain metastases. J Neuro-Oncol. 53. Diao K, Bian SX, Routman DM, Yu C, Ye JC,

2015;125(1):149–56. Wagle NA, et al. Stereotactic radiosurgery and ipi-
44. Minniti G, Scaringi C, Paolini S, Lanzetta G, Romano limumab for patients with melanoma brain metasta-
A, Cicone F, et al. Single-fraction versus multifraction ses: clinical outcomes and toxicity. J Neuro-Oncol.
(3 × 9 Gy) stereotactic radiosurgery for large (>2 cm) 2018;139(2):421–9.
brain metastases: a comparative analysis of local control 54. Martin AM, Cagney DN, Catalano PJ, Alexander BM,
and risk of radiation-induced brain necrosis. Int J Radiat Redig AJ, Schoenfeld JD, et al. Immunotherapy and
Oncol Biol Phys. 2016;95(4):1142–8. symptomatic radiation necrosis in patients with brain
45. Fujimoto D, von Eyben R, Gibbs IC, Chang SD, Li G, metastases treated with stereotactic radiation. JAMA
Harsh GR, et al. Imaging changes over 18 months fol- Oncol. 2018;4(8):1123–4.
lowing stereotactic radiosurgery for brain metastases: 55. Kaidar-Person O, Zagar TM, Deal A, Moschos SJ,
both late radiation necrosis and tumor progression can Ewend MG, Sasaki-Adams D, et al. The incidence
occur. J Neuro-Oncol. 2018;136(1):207–12. of radiation necrosis following stereotactic radio-
46. Zhang T, Yu H, Ni C, Zhang T, Liu L, Lv Q, et al. therapy for melanoma brain metastases: the poten-
Hypofractionated stereotactic radiation therapy tial impact of immunotherapy. Anti-Cancer Drugs.
activates the peripheral immune response in oper- 2017;28(6):669–75.
able stage I non-small-cell lung cancer. Sci Rep. 56. Fang P, Jiang W, Allen P, Glitza I, Guha N, Hwu
2017;7(1):4866. P, et al. Radiation necrosis with stereotactic radio-
47. Herskind C, Wenz F, Giordano FA. Immunotherapy surgery combined with CTLA-4 blockade and
combined with large fractions of radiotherapy: stereo- PD-1 inhibition for treatment of intracranial dis-
tactic radiosurgery for brain metastases-implications ease in metastatic melanoma. J Neuro-Oncol.
for intraoperative radiotherapy after resection. Front 2017;133(3):595–602.
Oncol. 2017;7:147. 57. Levin VA, Bidaut L, Hou P, Kumar AJ, Wefel JS,
48. Knisely JPS, Yu JB, Flanigan J, Sznol M, Kluger Bekele BN, et al. Randomized double-blind placebo-
HM, Chiang VLS. Radiosurgery for melanoma brain controlled trial of bevacizumab therapy for radiation
metastases in the ipilimumab era and the possibility of necrosis of the central nervous system. Int J Radiat
longer survival. J Neurosurg. 2012;117(2):227–33. Oncol Biol Phys. 2011;79(5):1487–95.
49. Korman AJ, Peggs KS, Allison JP. Checkpoint
58. Gonzalez J, Kumar AJ, Conrad CA, Levin VA. Effect
blockade in cancer immunotherapy. Adv Immunol. of bevacizumab on radiation necrosis of the brain. Int
2006;90:297–339. J Radiat Oncol Biol Phys. 2007;67(2):323–6.
50. Prins RM, Vo DD, Khan-Farooqi H, Yang M-Y, Soto 59. Chaunzwa TL, Deng D, Leuthardt EC, Tatter SB,
H, Economou JS, et al. NK and CD4 cells collaborate Mohammadi AM, Barnett GH, et al. Laser ther-
to protect against melanoma tumor formation in the mal ablation for metastases failing radiosurgery: a
brain. J Immunol. 2006;177(12):8448–55. multicentered retrospective study. Neurosurgery.
51. Nakata H, Yoshimine T, Murasawa A, Kumura E, 2018;82(1):56–63.
Harada K, Ushio Y, et al. Early blood-brain barrier dis- 60. McKay WH, McTyre ER, Okoukoni C, Alphonse-
ruption after high-dose single-fraction irradiation in rats. Sullivan NK, Ruiz J, Munley MT, et al. Repeat ste-
Acta Neurochir. 1995;136(1–2):82–6; discussion 86. reotactic radiosurgery as salvage therapy for locally
52. Qian JM, Yu JB, Kluger HM, Chiang VLS. Timing recurrent brain metastases previously treated with
and type of immune checkpoint therapy affect the radiosurgery. J Neurosurg. 2017;127(1):148–56.
Neurocognitive Effects of Brain
Metastases and Their Treatment
29
Karine A. Al Feghali, Mariana E. Bradshaw,
Caroline Chung, and Jeffrey S. Wefel
Introduction ease course, and most therapies do not substan-

tially cross the blood–brain barrier.
Brain metastases are the most common intracra- Multidisciplinary management of brain metas-
nial tumors. The overall frequency of parenchy- tases consists of surgery, radiation therapy [ste-
mal brain metastases in cancer patients was found reotactic radiosurgery (SRS) or whole brain
to be as high as 15–17% on autopsy studies [1, 2]. radiation therapy (WBRT)], and/or systemic
The cancers that are the most prone to metasta- therapy. Unfortunately, patients are sometimes
size to the brain are lung, breast, melanoma, plagued with debilitating and life-altering neuro-
renal, and colorectal [3–6], with incidences as cognitive adverse effects due to off-target or on-
high as 60% in patients with small-cell lung can- target/off-tumor toxicity. Efforts to prolong
cer (SCLC) [7] and EGFR-mutated or ALK- survival can indeed come at the detriment of
rearranged non-small-cell lung cancer [8, 9]. The cognitive dysfunction and/or impairment of func-
emergence of new and more effective systemic tional independence. A decline in neurocognitive
therapies and the development of better radiation function (NCF) after WBRT in patients with
therapy techniques have significantly improved brain metastases has been shown to precede dete-
locoregional rates and overall survival in cancer rioration in the quality of life (QoL) by
patients. These advances have paradoxically led 9–153 days [12], although this association has
to an increased incidence of brain metastases [7, not always been demonstrated across all studies
10, 11], as these typically emerge later in the dis- [13]. It is sometimes challenging for the clinician
to isolate the effect of a specific treatment on cog-
K. A. Al Feghali · C. Chung nitive function, as cognitive decline is most often
Department of Radiation Oncology, The University multifactorial, resulting from the interplay of dif-
of Texas MD Anderson Cancer Center, ferent therapies and from the central nervous sys-
Houston, TX, USA tem (CNS) disease itself. Balancing the benefits
M. E. Bradshaw and toxicities of brain metastasis treatment is a
Section of Neuropsychology, Department of challenging task that medical, radiation, and sur-
Neuro-Oncology, The University of Texas MD
Anderson Cancer Center, Houston, TX, USA gical oncologists face daily.
Radiation therapy, chemotherapy, and surgery
J. S. Wefel (*)
Section of Neuropsychology, Department of can affect processing speed, attention, learning
Neuro-Oncology and Department of Radiation and memory, executive function, and motor dex-
Oncology, The University of Texas MD Anderson terity among other brain functions. These
Cancer Center, Houston, TX, USA changes can occur in the acute setting, such as
e-mail: jwefel@mdanderson.org

https://doi.org/10.1007/978-3-030-42958-4_29
408 K. A. Al Feghali et al.
with encephalopathy [14–16], and in the chronic The prevalence varies among studies, as it
setting, such as in neurotoxicity due to delayed depends on the sensitivity of the cognitive tests
white matter tract damage [17]. Neurocognitive employed, the measured endpoints (for example,
side effects from treatment are often partially or some studies use dementia as an endpoint versus
completely reversible, but can also be irrevers- milder deficits in other studies), and the patient
ible. Many factors, including age, comorbidi- population included. In a trial randomizing
ties [18], and lower baseline pretreatment patients with brain metastases to different radia-
cognitive capacity, sometimes called “cognitive tion fractionation schedules, 16% of patients had
reserve” [19, 20] can increase the risk of chronic dementia at baseline, and the average mini-
neurotoxicity. mental status examination (MMSE) score was
The pathophysiology of radiation-induced 26, which is considered at the lower quartile of
neurotoxicity is not well understood. However, normal for the United States population [34, 35].
animal studies have demonstrated that radiation In a pilot study of patients treated with SRS for
blocks neurogenesis in the dentate gyrus of the one to three brain metastases, two-thirds of
hippocampus [21]. Chemotherapeutic agents patients had impaired NCF at baseline, with
such as paclitaxel have also been shown to reduce impairments observed in measures of executive
neurogenesis in the hippocampus [22]. Advanced functioning, motor dexterity, and learning and
imaging techniques such as structural and func- memory [36]. Some evidence of neurocognitive
tional MRI, as well as animal studies in rodents, dysfunction can be found at baseline in up to
have shaped our understanding of some of the 90% of patients with brain metastases [30].
neurobiological mechanisms underlying chemo- Neurocognitive testing done at baseline can dif-
therapy and radiotherapy-induced adverse neuro- ferentiate the effect of the disease itself on
logical effects. The main culprits are believed to patients’ cognition from treatment side effects. It
be oxidative stress causing DNA single-strand or has also been suggested that a combination of
double-strand breaks, increased apoptosis, vascu- tumor prognostic variables and brain function
lar injury, damage to white matter tracts, and neu- assessments is better at predicting survival than
roinflammation [19, 23–26]. Imaging biomarkers tumor variables alone in patients with brain
are being investigated as surrogates for early metastasis and in those with leptomeningeal dis-
assessment of RT-induced neurotoxicity. In par- ease [30, 37, 38].
ticular, dynamic contrast-enhanced (DCE) mag- Tumor location is a factor that can affect the
netic resonance imaging (MRI) can be used to likelihood of developing neurocognitive impair-
detect early changes in vasculature and predict ment. Traditionally, patients with tumors in the
late neurocognitive dysfunction [27, 28]. frontal or temporal lobes were thought to have
more propensity to demonstrate cognitive dys-
function than patients with tumors in less “elo-
he Impact of Baseline Brain
T quent” brain regions [29]. More recent work is
Metastases Characteristics challenging this “localizationist” view of cogni-
on Neurocognitive Function tion; it describes whole-brain network distur-
bances in patients with brain tumors, indicating
NCF at baseline, i.e., before any treatment, is that cognitive deficits cannot be explained by
directly affected by tumor location [29], tumor tumor location alone, but are rather reflections of
burden [30–32], rate of growth [33], and extent of the brain’s intricate interconnecting neural net-
surrounding edema [31]. Paraneoplastic effects, works [39].
prior neurologic disease, and use of certain medi- Tumor burden, and more specifically the vol-
cations can also play a role in impaired NCF at ume of the lesions, rather than the number of
baseline. brain metastases, has been consistently associ-
Neurocognitive impairment at baseline has ated with worse NCF at baseline. In a large trial
been extensively documented in the literature. of patients with brain metastases from a variety
29 Neurocognitive Effects of Brain Metastases and Their Treatment 409
of primary cancers, there was a statistically sig- patients treated with surgery in addition to
nificant small-to-moderate degree of correlation WBRT maintained Karnofsky Performance
between the indicator lesion volume at baseline Scores (KPS) of ≥70 significantly longer than
and measures of memory, verbal fluency, fine those who were treated with WBRT alone [43].
motor control and executive function [30]. In a In contrast, no significant difference was found
prospective study of 97 patients with brain metas- in length of functional independence (again as
tases, patients with large tumor volume showed a measured by KPS) in patients who had surgery
trend toward worse verbal memory and informa- alone as compared to those who received surgery
tion processing speed, whereas the number of plus WBRT [44]. The impact of surgery on cog-
brain metastases was not correlated with any of nitive functioning in patients with malignant
the seven cognitive domains studied [32]. Larger gliomas has recently been described in several
total tumor volume of brain metastases (≥3 ver- publications. The incidence of cognitive decline
sus <3 cm3) and a larger extent of tumor edema at has varied across studies with rates ranging from
baseline were also shown to be associated with 20 to 60% [32, 45–48]. Domains commonly
worse MMSE in a Japanese trial [31]. reported as most vulnerable to surgical impact
In patients with gliomas, the faster the rate of include memory, executive function, processing
tumor growth, the more dramatic the cognitive speed, and attention. Further, language functions
changes [40–42], which is probably due to less are at increased risk when lesions and resections
chance for compensatory neuroplastic reorgani- involve the dominant hemisphere. While some
zation. Although not well established, this asso- studies have found an increased risk for surgi-
ciation may hold true in patients with brain cally associated cognitive decline in patients
metastases as well, in whom tumor “momentum” with dominant hemisphere tumors, other risk
may be an independent predictor of cognitive factors have not been routinely identified.
changes [33].
he Impact of Systemic Therapy

T
he Impact of Surgery
T on Neurocognitive Function
on Neurocognitive Function
Chemotherapy
As noted above, brain metastases may impact
cognition prior to any intervention. Surgical Chemotherapy-induced cognitive side effects
resection of brain metastases may result in the (coined “chemobrain” or “chemofog”) are fre-
improvement of these cognitive symptoms quent occurrences in cancer patients, varying in
through relief of mass effect. However, there is frequency from 15% to 80% [15, 19, 49]. Many
also the potential to damage healthy tissue, chemotherapeutic agents have been inculpated,
resulting in some long-term neurocognitive including methotrexate, BCNU, melphalan,
impairment; this might occur through either fludarabine, cytarabine, 5-fluorouracil, levami-
direct focal damage or disruption of larger dis- sole, cisplatin, and capecitabine [14, 50, 51].
tributed networks. While the use of surgery is Structural brain differences between patients
often a critical component of the treatment of who received chemotherapy and control patients
brain metastases, particularly for solitary lesions, have been demonstrated, as well as reductions in
there is little information regarding functional the volume of white and grey matter [52, 53]. A
outcomes after resection. When it has been Dutch study utilized diffusion tensor imaging
assessed, functional outcome after surgical (DTI) MRI in chemotherapy-exposed breast can-
resection of brain metastases has generally been cer survivors to report on changes in white matter
measured using broad ratings of performance microstructural integrity over time as measured
status as opposed to more precise neuropsycho- by fractional anisotropy, mean/axial/radial diffu-
logical assessment. Patchell et al. found that sivity and tractography. This study revealed that
these metrics of global and focal white matter therapeutic response intra-cranially and extra-
organization significantly deteriorated over time cranially, which had traditionally been solely
following treatment [54]. attributed to inadequate penetration of the blood–
It has been demonstrated that a disrupted brain barrier. Nowadays, immunotherapy and tar-
blood–brain barrier around brain tumors (as geted therapies play an increasingly important
opposed to healthy brain tissue) helps systemic role in the management of patients with brain
agents gain access [55, 56]. There is also grow- metastases. Tyrosine kinase inhibitors (TKIs),
ing literature on ways to circumvent the blood– such as some of the ALK-TKIs, EGFR-TKIs, and
brain barrier, by inhibiting transporters that HER2-TKIs, and other agents, such as BRAF
function in extruding drugs or toxins from the inhibitors, anti-PD-1, and anti-CTLA4, have
brain [57]. However, the literature on the impact been shown to have good activity against CNS
of chemotherapeutic agents on cognition in metastatic disease [69, 70]. The impact of these
patients with brain metastases specifically is novel therapeutics on NCF is still not elucidated,
scarce. Whether the addition of chemotherapy to as most phase 2 and 3 trials testing these new
radiation results in more cognitive dysfunction is agents have not yet incorporated formal neuro-
uncertain. Few clinical trials have examined the cognitive testing. Data on cognitive effects of
role of chemotherapy versus chemotherapy with immunotherapy or targeted therapy remain,
WBRT [58–61], but none reported on the end- therefore, very scarce. Although these options
point of cognitive dysfunction. have been hypothesized to come at a lower cost
Because of its good penetration in the brain to neurocognitive impairment [71], there is a lack
and its proven efficacy in glioblastoma multi- of high-level evidence to prove this assumption.
forme [62], temozolomide, an oral cytotoxic There were also growing concerns on the
alkylating agent, was tested in brain metastases safety of combining EGFR-TKIs with cranial
in few phase II studies and a phase III study, with irradiation in non-small cell lung cancer patients
or without WBRT or SRS. Its addition was asso- in terms of neurotoxicity. A systematic review on
ciated with good overall response rates, but no that topic concluded that although WBRT used
survival benefit [63–66]. While a phase II trial concurrently with TKI did not seem to increase
suggested that the addition of temozolomide to neurotoxicity, there was also a lack of high-
RT might lead to greater neurologic improvement quality evidence to support the use of these two
than RT alone [65], a phase III trial suggested therapies concurrently [72]. In fact, only one
increased toxicities that could lead to a possible study included in this review used a formal neu-
survival detriment [61]. However, none of these rocognitive battery of tests [73], and another one
trials performed a neurocognitive assessment for used MMSE, the EORTC QLQ-C30 cognitive
the patients included. function subscale and Trail Making Test Part B
(for executive function) [74]. In both these stud-
ies, the addition of TKI (erlotinib and gefitinib)
merging Role of Immunotherapy
E seemed to be well tolerated.
and Targeted Agents in Brain Checkpoint blockade directed against the pro-
Metastases: Uncertain Impact grammed death-1 (PD-1) pathway has been
on Neurocognition shown to improve cognitive performance in
murine models of Alzheimer’s disease [75].
Advances in genetic characterization of brain Whether this finding can be extrapolated to
metastases have paved the way to new therapeu- patients with brain metastases treated with anti-
tic avenues. Actionable mutations have been PD-1 remains to be proven. In a study with 36
identified in secondary brain lesions, which are patients with brain metastases secondary to
sometimes distinct from the mutations harbored NSCLC or melanoma treated with pembroli-
by primary cancer [67, 68]. This genetic hetero- zumab, one patient developed grade 3 cognitive
geneity might have resulted in the differential dysfunction, which could have resulted from per-
ilesional edema and not from the direct effect of assessed by the Folstein MMSE [13]. The limita-
the drug [76]. Some experts have hypothesized tions of MMSE will be discussed later in the
that there might be differences in individual sus- chapter, but this study was performed before the
ceptibility to cognitive impairments imparted by RTOG’s growing use of more elaborate neuro-
increased immune activation using some of these cognitive batteries. In a study by Shibamoto
agents [77]. Many questions in the field of immu- et al. including 101 patients treated with WBRT
notherapy and cognitive side effects remain (40 Gy in 20 fractions), there was a decrease in
unanswered and there needs to be more neuro- MMSE scores of ≥4 points in 7.4%, 11%, 20%,
cognitive assessment incorporated in trials test- 12%, 5.9% of assessable patients at 3, 6, 9, 12,
ing these agents. and 15 months, respectively. Brain atrophy was
observed in 30% of patients but was not corre-
lated with MMSE decrease [78]. In a study by
he Impact of Radiation Therapy
T Regine et al., accelerated fractionation (1.6 Gy
Targeting Brain Metastases twice a day to 54.4 Gy) was compared to stan-
on Neurocognitive Function dard WBRT fractionation (3 Gy daily to 30 Gy),
with no difference in NCF between the two regi-
Whole Brain Radiation Therapy mens as evaluated by MMSE [35]. Studies have
reported neurocognitive dysfunction in 11–85%
Whole Brain Radiation Therapy (WBRT) con- of patients treated with postoperative WBRT for
sists of two opposed-lateral treatment fields that brain metastases [79, 80]. These numbers vary
encompass the entire brain (Fig. 29.1). It has long depending on the assessment tool used and on
been considered the standard of care for patients the definition of neurocognitive deterioration in
with brain metastases, either postoperatively or the different trials. Given these alarming num-
as the sole treatment, especially for inoperable bers, the last decade has witnessed the decline of
patients or in the setting of multiple brain metas- WBRT in favor of the less ‘toxic’ and more tar-
tases. Its impact on neurocognition can, however, geted stereotactic radiosurgery. The next section
be quite dramatic, ranging from mild cognitive will offer an overview of the studies comparing
impairment to full-fledged dementia. the two techniques in terms of neurocognitive
In a secondary analysis of a trial including side effects. This decline in WBRT use was also
patients with multiple brain metastases receiving potentiated by the recent advancement in sys-
WBRT with or without thalidomide, both arms temic therapies [81] that have the potential to
experienced a steady neurocognitive decline as control microscopic disease, while surgery or
a b c
Fig. 29.1 Example of a whole brain radiation therapy Computed Tomography (CT) scans with isodose lines –
plan with a dose prescription of 30 Gray in ten fractions. Transverse (a), sagittal (b) and coronal (c) views
The three panels below represent cuts from the simulation
SRS is responsible for gross disease control. maintained in 92%, 91% and 89% of patients in
Even the role of WBRT in palliation (in patients groups A, B, and C, respectively [84].
who are not a candidate for surgical resection or Adding WBRT to SRS is associated with bet-
SRS) has been challenged in the recent British ter local control and distant intracranial control,
QUARTZ trial, as optimal supportive care but not with improved overall survival as com-
proved to be non-inferior [82]. pared to SRS alone [79, 85]. It is, therefore,
important to know whether this local control ben-
efit outweighs the possible neurocognitive side
Stereotactic Radiosurgery effects of adding WBRT. Table 29.1 outlines the
studies that compare neurocognitive side effects
Stereotactic radiosurgery is a form of local radio- from SRS to those from WBRT with or without
therapy that precisely delivers a single high dose SRS.
of radiation using multiple beams of high-energy In a randomized controlled trial of 58 patients
x-rays, gamma rays, or protons that converge on at MD Anderson Cancer Center, those who
a discrete treatment volume, while maximally received WBRT plus SRS showed a significantly
sparing/minimizing the irradiation to the adjacent greater decline in HVLT-R Total Recall at
normal brain parenchyma and other surrounding 4 months than patients treated with SRS alone
normal structures [83] (Fig. 29.2). (52% versus 24%, respectively), a difference
SRS alone, when used for the treatment of which persisted at the 6-month follow-up. These
brain metastases, is not associated with as many patients also had a greater drop in executive func-
neurocognitive side effects as WBRT. Also, the tioning as compared to patients randomly assigned
number of metastases treated with SRS seems to the SRS alone arm [86]. These findings suggest
not to correlate with the extent of the decline in that even though patients treated with SRS alone
NCF. In a Japanese study comparing outcomes of had higher rates of recurrences in the brain,
patients with 1 (group A), 2–4 (group B), and WBRT neurotoxicity (a decline in verbal learning
5–10 brain metastases (group C) treated with and memory) appeared to be worse than the cog-
SRS, the incidences of MMSE deterioration were nitive decline associated with recurrences, as long
low overall and were not significantly different as close surveillance with early diagnosis of
between the three groups. MMSE score was recurrent brain metastases was performed.
Fig. 29.2 Example of a hippocampal avoidance whole brain radiation therapy plan using intensity-modulated radiation
therapy
Table 29.1 Neurocognitive side effects of stereotactic radiosurgery (SRS) versus whole brain radiation therapy (WBRT) alone or with SRS
Definition of the Results
SRS Instrument for endpoint of cognitive Results with with
Study population dose WBRT dose cognitive testing deterioration Results with SRS SRS + WBRT WBRT
Brown 213 patients with 18– 30 Gy in 12 Learning and Decline >1 SD from 63.5% at 3 months 91.7% at 3 months –
et al. 2016 1–3 brain 22 Gy fx immediate memory baseline on at least 1
[85] metastases (no (HVLT-R Immediate cognitive test at 3 months
surgical resection) Recall)
Fine motor control
(Grooved Pegboard
Test)
Verbal fluency
(COWAT)
Processing speed (TMT
part A)
Executive function
(TMT part B)
Delayed memory
(HVLT-R delayed
recall)
Recognition (HVLT-R
recognition)
Chang 58 patients with 15– 30 Gy in 12 HVLT–R total recall Deterioration (5-point 24% at 4 months 52% at 4 months –
29 Neurocognitive Effects of Brain Metastases and Their Treatment
et al. 2009 1–3 brain 20 Gy fx drop compared with

[86] metastases (no baseline) in HVLT-R
surgical resection) total recall at 4 months
Aoyama 28 patients with 18– 30 Gy in 10 MMSE Median MMSE change Pretreatment MMSE: 27 Pretreatment –
et al. 2006 1–3 brain 25 Gy fx between pre- and Posttreatment MMSE: 28 MMSE: 28
[88] metastases (no posttreatment Posttreatment
surgical resection) MMSE: 27
Soffietti 341 patients with 25 Gy 30 Gy in 10 Cognitive functioning Differences between the 10.7-point difference in cognitive function at –
et al. 2012 1–3 brain fx as part of the EORTC two treatment arms for 12 months between the two treatment arms, in favor
[89] metastases (no QLQ-C30 (used as an all post-baseline time of SRS alone (statistically significant)
surgical resection) HRQOL measure) points
(continued)
413
414
Definition of the Results

SRS Instrument for endpoint of cognitive Results with with
Study population dose WBRT dose cognitive testing deterioration Results with SRS SRS + WBRT WBRT
Aoyama 92 patients with 18– 30 Gy in 10 MMSE 3-point drop in the 59.3% at 1 year, 51.9%, at 76.1%, at 1 year, –
et al. 2007 1–4 brain 25 Gy fx MMSE 2 years and 51.9% at 3 years 68.5%, at 2 years
[31] metastases (no The average duration until and 14.7% at 3 years
surgical resection) deterioration was 7.6 months The average duration
until deterioration
was 16.5 months
Brown 213 patients with 12– 30 Gy in 10 Learning and Decline >1 SD from 52% at 6 months 85% at
2017 [79] total/subtotal 20 Gy fx or immediate memory baseline on at least 1 6 months
resection of a 37·5 Gy in (HVLT-R Immediate cognitive test at 3 months
single BM 15 fx Recall)
Verbal fluency
(controlled Oral word
association test)
Processing speed (TMT
part A)
Executive function
(TMT part B)
Delayed memory
(HVLT-R delayed
recall)
Recognition (HVLT-R
recognition)
Kepka 59 patients with 15 Gy 30 Gy in 10 MMSE 3-point drop in the 4.5% at 6 months 11% at
2016 [80] total/subtotal fx MMSE 6 months
resection of a
single BM
Abbreviations: SRS stereotactic radiosurgery, WBRT whole brain radiation therapy, Gy gray, fx fractions, SD standard deviation, HVLT-R Hopkins Verbal Learning Test–Revised,
COWAT Controlled Oral Word Association Test, TMT Trail Making Test, MMSE Mini-Mental State Examination, EORTC QLQ-C30 EORTC Quality of Life Questionnaire C30,
HRQOL health-related quality of life
K. A. Al Feghali et al.
In the postoperative setting, a randomized trial Table 29.2 Clinical trial battery of neurocognitive tests
of adjuvant WBRT vs. SRS also demonstrated recommended for cognitive function assessment in
patients with brain metastases
better intracranial control with postoperative
WBRT, at the expense of a greater cognitive Cognitive Domain Test
Learning & Hopkins Verbal Learning
decline associated with WBRT that persisted at
Memory Test-Revised (HVLT-R) [165]
the 12-month follow-up visit [79]. Verbal fluency Controlled Oral Word
These randomized trials clearly demonstrate Association [166]
that WBRT compromises neurocognition more Executive Trail Making Test Part B [167,
than SRS, without yielding a survival benefit functioning 168]
Information Trail Making Test Part A [167]
[86–88]. It is thus reasonable to consider SRS
processing speed
first when a patient presents with a limited num-
ber of brain metastases and reserve WBRT as a
last resort after the failure of one or several more recent studies are using [79, 85, 86, 92].
courses of SRS and surgical salvage. While SRS Table 29.2 lists a brief, core battery of neurocog-
controls gross disease, systemic therapy might nitive tests that has been deemed appropriate for
also be needed to control microscopic disease in the clinical trial setting in patients with both CNS
the brain. A strategy of close follow-up and regu- and non-CNS cancers [92, 93]. While not an
lar high-quality neuroimaging to detect recur- exhaustive list of tests that would be appropriate
rences is preferred nowadays over more for use with this patient population, these mea-
aggressive treatment and is consistent with the sures have been found to have appropriate psy-
trend towards personalized treatment. It is, how- chometric properties and are sensitive to the
ever, dependent on the patient and medical effects of the tumor and anticancer treatment on
team’s willingness to adhere to a strict follow-up the domains of memory and learning, informa-
schedule. tion processing speed, and executive function.
It is difficult to quantitatively combine results The mechanisms underlying this differential
from these different studies, as they use different neurotoxicity of SRS versus WBRT have been
definitions of neurocognitive deterioration and investigated using MRI. It has been found that
time to assessment, as well as different assess- delayed white matter leukoencephalopathy is
ment methods, each with a different sensitivity. very common in patients treated with WBRT,
One study used self-reported measures, namely reported in up to 97% of patients [94, 95],
the European Organization for Research and whereas its incidence is much lower (1–3%) in
Treatment of Cancer (EORTC) Quality of Life patients treated with SRS [84, 94].
Questionnaire C30 [89]. The Japanese studies
[31, 88] used a mini-mental status examination to
compare cognitive outcomes between the differ- Neurotoxicity in the Setting
ent treatment modalities. Data from these studies of Prophylactic Cranial Irradiation
should be interpreted with caution, as self-
reporting and actual formal cognitive testing are Studying the effect of prophylactic cranial irra-
poorly correlated, and MMSE has been deemed diation (PCI) on cognitive function can help dis-
not sensitive enough in a brain tumor population entangle CNS toxicity inherent to the presence of
[90]. The Food and Drug Administration (FDA) CNS disease from the effect of radiation therapy
stated that objective assessment is preferred to itself.
subjective self-report in neuro-oncology, due to PCI is considered standard of care in SCLC
the challenges in assessing patient-centered out- based on an individual patient data meta-analysis
comes in individuals with malignant brain tumors in limited-stage SCLC [96], and a seminal trial in
[91]. A battery of standardized neuropsychologi- extended-stage SCLC [97], showing an overall
cal tests is now recommended in clinical trials survival benefit with the use of PCI. The role of
reporting on cognitive function, and this is what PCI in non-small-cell lung cancer (NSCLC) is
not as straightforward; no overall survival benefit mentioned above: RTOG 0214, and RTOG 0212
has been demonstrated in the different random- [107]. As compared to observation, PCI was
ized controlled trials randomizing patients with associated with a significant threefold higher risk
NSCLC to PCI versus no PCI [98–103], although of decline in self-reported cognitive functioning
a meta-analysis indicated a disease-free survival at 6 months and 12 months. PCI was also associ-
benefit in a subset of patients [104]. ated with a significant decline in HVLT-Total
Late neurological complications from PCI Recall and HVLT-Delayed Recall at 6 and
have only been formally studied in three trials 12 months. Interestingly, the decline in HVLT
[103, 105, 106]. In RTOG 0214, a trial that ran- and decline in self-reported cognitive functioning
domized 340 patients with stage III NSCLC to were not closely correlated [108].
PCI or no PCI, there was a trend toward a greater
decline in patient-reported cognitive functioning
in the PCI arm. There was no significant differ- Radiation-Induced Neuroxoticity:
ence in MMSE scores between the two arms, Timeline and Risk Factors
except at 3 months. The only significant differ-
ence in the NCF analysis was in the Hopkins Sheline’s report in the 1980s was the first to sub-
Verbal Learning Test (HVLT): patients who were divide radiation-induced brain injury into acute
treated with PCI had significantly greater deterio- (early, during radiation), subacute (up to 6 months
ration in learning and memory at 1 year as com- postradiation therapy), and late effects (chronic,
pared to patients in the observation arm. There more than 6 months postradiation therapy) [109,
were no significant differences in QoL between 110].
the patients who received PCI and those who did Acute encephalopathy, consisting of head-
not [105], unlike the sequential association ache, nausea, vomiting, and fever with onset dur-
between NCF decline and QoL deterioration ing treatment, occurs almost exclusively if a high
noted earlier in this chapter in the setting of dose per fraction is used, and not with the con-
WBRT for brain metastases. ventionally used dose of 3 Gray or less per frac-
The RTOG 0212 trial randomized 265 patients tion [111, 112]. This acute effect has been linked
with limited stage-SCLC and a complete response to edema formation secondary to blood–brain
after chemotherapy and thoracic RT to either barrier disruption, due to apoptosis of endothelial
standard-dose PCI (25 Gy in 10 fractions) or cells [113–116]. Corticosteroids can help in treat-
higher-dose PCI (36 Gy). The 36 Gy cohort was ing these symptoms.
secondarily randomized to receive PCI in either18 Subacute complications include somnolence
fractions of 2 Gy or twice daily in 24 fractions. syndrome, whose symptoms are transient and
Detailed neuropsychological test batteries were include excessive sleepiness, drowsiness, and
carried out on the study population. The baseline anorexia, and are mainly documented in children
assessments prior to PCI demonstrated abnor- receiving PCI for ALL [117, 118], or in adults
malities in multiple parameters including lan- receiving definitive doses of radiation therapy
guage, visual and spatial scanning, attention, (45–55 Gray) for primary brain tumors [119,
sequencing, and speed. Chronic neurotoxicity in 120]. Another subacute effect is impairment in
this study was defined as the deterioration in at verbal memory function 6–8 weeks after PCI
least one of the six cognitive domains without the completion as demonstrated by Welzel et al.
development of brain metastasis at 12 months. [121].
The incidence of chronic neurotoxicity was sig- Late or chronic effects are the most dreaded of
nificantly higher in patients treated with 36 Gy all radiation-induced injuries, as they are usually
compared with 25 Gy (85 and 89% versus 60%, irreversible. Molecular mechanisms underlying
respectively, p = 0.02) [106]. the development of these chronic effects are
A study by Gondi et al. pooled QoL and NCF inflammation [122, 123], hypoxia with vascular
results from the two RTOG randomized studies endothelial growth factor upregulation [124,
125], and neurogenesis inhibition [126]. This d ysfunction has been tested in the phase III trial,
cascade of events can lead to radiation-induced RTOG 0614. There was a trend toward less decline
demyelination and leukoencephalopathy that can in the primary endpoint of HVLT-R Delayed
occur months to years after irradiation [110], as Recall at 24 weeks with memantine as compared
well as radiation necrosis [127]. In long-term to placebo, but this result did not reach statistical
SCLC survivors, PCI has been shown to result in significance (p-value = 0.059), probably because
progressive ventricular dilatation or cerebral of significant drop out, resulting in statistical power
atrophy up to 8 years after therapy completion, as of only 35%. The patients on the memantine arm
well as a slow decline in NCF [128, 129]. had a significantly longer time to cognitive decline,
The incidence and severity of radiation- and better results in executive functioning and pro-
induced toxicities do not only depend on radia- cessing speed [137]. Donepezil, a reversible acetyl-
tion dose but also depend on some patient-related choline esterase inhibitor, has also been tested for
factors, such as age, chemotherapy and existing its ability to improve cognitive dysfunction in a
comorbidities [110]. In RTOG 0212, age phase III trial in 198 adult brain tumor survivors,
(>60 years) was the most significant predictor for and although it did not show significant improve-
the development of chronic neurotoxicity (p ment in the overall composite cognitive score (pri-
value = 0.005) [106]. Preexisting medical condi- mary endpoint), it showed significant benefit over
tions, such as hypertension, have also been shown placebo in some specific cognitive functions, such
to accelerate vascular radiation damage [18]. as memory, as well as motor speed and dexterity
Some patients may also have a genetic predis- [138]. One of the limitations of this study was the
position to develop more treatment-related neu- low dose of donepezil used (10 mg/day), given
rocognitive toxicities. Based on the premise that that studies on patients with moderate-to-severe
the APOE e4 allele confers an increased risk of Alzheimer’s disease showed significantly greater
Alzheimer’s disease, a retrospective analysis of cognitive benefits with higher doses of donepezil
RTOG 0614 evaluated the relationship between 23 mg/day than donepezil 10 mg/day [139].
APOE e4 carrier status and NCF after treatment Another strategy to avoid cognitive dysfunc-
with WBRT in patients with brain metastases. tion, and more specifically short-term memory
Carrying the APOE e4 allele was shown to be a loss, is hippocampal avoidance whole-brain radi-
risk factor for worse memory function after treat- ation therapy (HA-WBRT) (Fig. 29.3). It uses
ment with WBRT (with or without memantine) conformal radiation therapy to avoid neural stem
[130]. cells in the hippocampal dentate gyrus, which are
mitotically active and radiosensitive, and are
responsible for the formation of new memories
Strategies to Mitigate Neurotoxicity [126, 140, 141].
This technique was tested in the phase II
Some strategies have been tested to potentially cooperative trial RTOG 0933, which showed sig-
mitigate the neurocognitive complications of brain nificant memory preservation with hippocampal
irradiation [131]. One of them is the use of neuro- avoidance cranial irradiation, whereby relative
protective drugs, such as angiotensin-converting decline in Hopkins Verbal Learning Test-Revised
enzyme inhibitors [132], angiotensin type-1 recep- Delayed Recall at 4 months was 7% in the exper-
tor blockers [133], erythropoietin [134], and lith- imental arm, which was significantly lower than
ium [135, 136], all of which have been tested prespecified historical control of patients with
in vivo. Two of these potential neuroprotective brain metastases treated without hippocampal
drugs, memantine and donepezil, deserve special avoidance [142]. NRG-CC001 examined the
mention, as they have both been investigated in combined use of HA-WBRT + memantine ver-
phase III clinical trials. The effectiveness of sus WBRT + memantine in patients with brain
memantine, an N-Methyl-D-aspartate (NMDA) metastases. Recently reported results demon-
receptor antagonist, in preventing cognitive strated a delay in the time to neurocognitive
Fig. 29.3 Example of a stereotactic radiosurgery plan using Gamma Knife for a left parietal lesion, treated with a
prescription dose of 18 Gy to the 50% isodose line
decline in the HA-WBRT + memantine arm with deterioration, without necessarily improving sur-
no difference in OS or PFS [143]. Two ongoing vival, is also on the rise. The new studies outlined
trials, NRG-CC003 (ClinicalTrials.gov in this chapter are helping in shaping better risk-
Identifier: NCT01780675) and the Spanish versus-benefits evaluations for interventions tar-
PREMER trial (NCT02397733) [144], are cur- geted against brain metastases.
rently examining the role of hippocampal avoid- Although treatment-related cognitive deficits
ance in the setting of PCI for SCLC specifically. are being increasingly reported in clinical trials
Behavioral interventions have also been on cancer patients with brain metastases, their
attempted to mitigate treatment-related neuro- incidence and patterns are sometimes inconsis-
cognitive impairment. Cognitive rehabilitation tent between studies. This can be explained by
consists of clinic-based therapeutic programs different factors, such as the heterogeneity of the
designed to improve cognitive skills and func- patient population included, the wide range of
tional capacity [145]. While to our knowledge treatment modalities used, the different cognitive
these strategies have not been studied in patients tests employed, and the various statistical meth-
with brain metastases specifically, there is evi- ods used to measure and report neuropsychologi-
dence to suggest benefit in cancer patients, cal changes (some more sensitive than others)
including those with brain tumors [146–148]. [156, 157]. Some studies rely on self-reporting
Studies have reported improvement in executive instead of objective neurocognitive testing. Self-
function, working memory, processing speed, and reporting is problematic, as patients with cogni-
attention with the use of cognitive behavioral ther- tive impairments may not be fully aware of the
apy among other interventions focused on self- extent of their cognitive problems [158].
awareness, mindfulness, and meditation [149–153]. Moreover, testing at baseline is not always avail-
A home-based, computerized Lumosity program able but it is critical to assess the effects of anti-
has also been tested successfully in breast cancer cancer agents. Some studies have also struggled
survivors to improve executive function, process- with patient compliance when it comes to follow-
ing speed, and verbal fluency [154], while patients up neurocognitive testing [89], though feasibility
with primary brain tumors failed to comply with of repeated neurocognitive assessment in this
the intervention and did not demonstrate improve- patient population has been demonstrated [86].
ments in cognitive function [155]. It is, therefore, difficult to draw meaningful con-
clusions when pooling these studies together. To
make this task easier in the future, there is a need to
Limitations of the Existing make the tools used to evaluate cognition uniform
Literature and Future Directions across studies and to enforce a battery of validated
tools. In light of these challenges, the “International
In the development of new therapeutic agents, the Cognition and Cancer Task Force” was created and
endpoint of NCF is gaining increasing popularity, issued recommendations to harmonize studies of
and the search for drugs that delay neurocognitive cognitive function in cancer patients [92]. The core
clinical trial battery represented in Table 29.2 and 8. Zhang J, Yu J, Sun X, Meng X. Epidermal growth
endorsed by the ICCTF was also endorsed by the factor receptor tyrosine kinase inhibitors in the treat-
ment of central nerve system metastases from non-
RANO group [92, 159–163]. small cell lung cancer. Cancer Lett. 2014;351:6–12.
As brain function is multifaceted and subject 9. Rangachari D, Yamaguchi N, VanderLaan PA, et al.
to subtle changes over time, capturing it through Brain metastases in patients with EGFR-mutated or
formal and standardized neurocognitive testing is ALK-rearranged non-small-cell lung cancers. Lung
Cancer. 2015;88:108–11.
critical. The same battery of tests should be used 10. Law A, Karp DD, Dipetrillo T, Daly BT. Emergence
consistently across studies, as a sensitive mea- of increased cerebral metastasis after high-dose
sure of brain functioning [90]. Improving the preoperative radiotherapy with chemotherapy in
way the endpoint of NCF is reported can be criti- patients with locally advanced nonsmall cell lung
carcinoma. Cancer. 2001;92:160–4.
cal, as a study has demonstrated that cognitive 11. Cox JD, Scott CB, Byhardt RW, Emami B, Russell
deterioration can precede radiological evidence AH, Fu KK, Parliament MB, Komaki R, Gaspar
of progression by around 6 weeks in patients with LE. Addition of chemotherapy to radiation therapy
primary brain tumors [164]. Whether this finding alters failure patterns by cell type within non-small
cell carcinoma of lung (NSCCL): analysis of radia-
can be extrapolated to patients with brain metas- tion therapy oncology group (RTOG) trials. Int J
tases remains to be studied. Radiat Oncol Biol Phys. 1999;43:505–9.
As new therapies are currently revolutionizing 12. Li J, Bentzen SM, Li J, Renschler M, Mehta
cancer treatment, and allowing patients with MP. Relationship between neurocognitive function
and quality of life after whole-brain radiotherapy in
brain metastases to live longer, the concern for patients with brain metastasis. Int J Radiat Oncol.
delayed neurotoxicity is all the more real, and 2008;71:64–70.
strategies to delay or prevent this life-altering 13. Corn BW, Moughan J, Knisely JPS, et al. Prospective
morbidity all the more important. More efforts in evaluation of quality of life and neurocognitive
effects in patients with multiple brain metastases
the field of genetic characterization of brain receiving whole-brain radiotherapy with or without
metastasis should be deployed, which could per- thalidomide on Radiation Therapy Oncology Group
mit the identification of actionable mutations and (RTOG) trial 0118. Int J Radiat Oncol Biol Phys.
allow treatment personalization. In the future, 2008;71:71–8.
14. Perry A, Schmidt RE. Cancer therapy-associated
this approach should be favored over the “one- CNS neuropathology: an update and review of the
size-fits-all” strategy. literature. Acta Neuropathol. 2006;111:197–212.
15. Wefel JS, Saleeba AK, Buzdar AU, Meyers
CA. Acute and late onset cognitive dysfunction
associated with chemotherapy in women with breast
References cancer. Cancer. 2010;116:3348–56.
16. Choi SM, Lee SH, Yang YS, Kim BC, Kim MK,
1. Posner JB, Chernik NL. Intracranial metastases from Cho KH. 5-fluorouracil-induced leukoencephalopa-
systemic cancer. Adv Neurol. 1978;19:579–92. thy in patients with breast cancer. J Korean Med Sci.
2. Takakura K, Sano K, Hojo S, Hirano A. Metastatic 2001;16:328–34.
tumors of the central nervous system. Tokyo: Igaku- 17. Nagesh V, Tsien CI, Chenevert TL, Ross BD,
Shoin; 1983. Lawrence TS, Junick L, Cao Y. Radiation-induced
3. Barnholtz-Sloan JS, Sloan AE, Davis FG, Vigneau changes in normal-appearing white matter in
FD, Lai P, Sawaya RE. Incidence proportions of patients with cerebral tumors: a diffusion ten-
brain metastases in patients diagnosed (1973 to 2001) sor imaging study. Int J Radiat Oncol Biol Phys.
in the Metropolitan Detroit Cancer Surveillance 2008;70:1002–10.
System. J Clin Oncol. 2004;22:2865–72. 18. Hopewell JW, Wright EA. The nature of latent
4. Taillibert S, Le Rhun E. Epidemiology of brain cerebral irradiation damage and its modification by
metastases. Cancer Radiother. 2015;19:3–9. hypertension. Br J Radiol. 1970;43:161–7.
5. Nayak L, Lee EQ, Wen PY. Epidemiology of brain 19. Ahles TA, Saykin AJ, McDonald BC, Li Y,
metastases. Curr Oncol Rep. 2012;14:48–54. Furstenberg CT, Hanscom BS, Mulrooney TJ,
6. Nussbaum ES, Djalilian HR, Cho KH, Hall Schwartz GN, Kaufman PA. Longitudinal assess-
WA. Brain metastases. Histology, multiplicity, sur- ment of cognitive changes associated with adjuvant
gery, and survival. Cancer. 1996;78:1781–8. treatment for breast cancer: impact of age and cogni-
7. Komaki R, Cox JD, Whitson W. Risk of brain metas- tive reserve. J Clin Oncol. 2010;28:4434–40.
tasis from small cell carcinoma of the lung related 20. Stern Y. What is cognitive reserve? Theory and
to length of survival and prophylactic irradiation. research application of the reserve concept. J Int
Cancer Treat Rep. 1981;65:811–4. Neuropsychol Soc. 2002;8:448–60.
21. Madsen TM, Kristjansen PEG, Bolwig TG, 34. Murray KJ, Scott C, Zachariah B, Michalski
Wortwein G. Arrested neuronal proliferation and JM, Demas W, Vora NL, Whitton A, Movsas
impaired hippocampal function following fraction- B. Importance of the mini-mental status examina-
ated brain irradiation in the adult rat. Neuroscience. tion in the treatment of patients with brain metasta-
2003;119:635–42. ses: a report from the Radiation Therapy Oncology
22. Lee BE, Choi BY, Hong DK, Kim JH, Lee SH, Group protocol 91-04. Int J Radiat Oncol Biol Phys.
Kho AR, Kim H, Choi HC, Suh SW. The cancer 2000;48:59–64.
chemotherapeutic agent paclitaxel (Taxol) reduces 35. Regine WF, Scott C, Murray K, Curran
hippocampal neurogenesis via down-regulation of W. Neurocognitive outcome in brain metastases
vesicular zinc. Sci Rep. 2017;7:11667. patients treated with accelerated-fractionation
23. Han R, Yang YM, Dietrich J, Luebke A, Mayer- vs. accelerated-hyperfractionated radiotherapy:
Proschel M, Noble M. Systemic 5-fluorouracil an analysis from Radiation Therapy Oncology
treatment causes a syndrome of delayed myelin Group Study 91-04. Int J Radiat Oncol Biol Phys.
destruction in the central nervous system. J Biol. 2001;51:711–7.
2008;7:12. 36. Chang EL, Wefel JS, Maor MH, et al. A pilot study
24. Seigers R, Fardell JE. Neurobiological basis of of neurocognitive function in patients with one to
chemotherapy-induced cognitive impairment: a three new brain metastases initially treated with
review of rodent research. Neurosci Biobehav Rev. stereotactic radiosurgery alone. Neurosurgery.
2011;35:729–41. 2007;60:274–7.
25. Seigers R, Schagen SB, Van Tellingen O, Dietrich 37. Mehta MP, Shapiro WR, Glantz MJ, et al. Lead-in
J. Chemotherapy-related cognitive dysfunction: phase to randomized trial of motexafin gadolinium
current animal studies and future directions. Brain and whole-brain radiation for patients with brain
Imaging Behav. 2013;7:453–9. metastases: centralized assessment of magnetic reso-
26. Vichaya EG, Chiu GS, Krukowski K, Lacourt nance imaging, neurocognitive, and neurologic end
TE, Kavelaars A, Dantzer R, Heijnen CJ, Walker points. J Clin Oncol. 2002;20:3445–53.
AK. Mechanisms of chemotherapy-induced behav- 38. Sherman AM, Jaeckle K, Meyers CA. Pretreatment
ioral toxicities. Front Neurosci. 2015;9:1–17. cognitive performance predicts survival in
27. Farjam R, Pramanik P, Aryal MP, Srinivasan A, patients with leptomeningeal disease. Cancer.
Chapman CH, Tsien CI, Lawrence TS, Cao Y. A 2002;95:1311–6.
radiation-induced hippocampal vascular injury sur- 39. Derks J, Reijneveld JC, Douw L. Neural network
rogate marker predicts late neurocognitive dysfunc- alterations underlie cognitive deficits in brain tumor
tion. Int J Radiat Oncol Biol Phys. 2015;93:908–15. patients. Curr Opin Oncol. 2014;26:627–33.
28. Cao Y, Tsien CI, Sundgren PC, Nagesh V, Normolle 40. Wefel JS, Noll KR, Rao G, Cahill DP. Neurocognitive
D, Buchtel H, Junck L, Lawrence TS. Dynamic function varies by IDH1 genetic mutation status
contrast-enhanced magnetic resonance imaging in patients with malignant glioma prior to surgical
as a biomarker for prediction of radiation-induced resection. Neuro Oncol. 2016;18:1656–63.
neurocognitive dysfunction. Clin Cancer Res. 41. Klein M. Lesion momentum as explanation for pre-
2009;15:1747–54. operative neurocognitive function in patients with
29. Tucha O, Smely C, Preier M, Lange KW. Cognitive malignant glioma. Neuro Oncol. 2016;18:1595–6.
deficits before treatment among patients with brain 42. Noll KR, Sullaway C, Ziu M, Weinberg JS, Wefel
tumors. Neurosurgery. 2000;47:324. JS. Relationships between tumor grade and neuro-
30. Meyers CA, Smith JA, Bezjak A, et al. cognitive functioning in patients with glioma of the
Neurocognitive function and progression in patients left temporal lobe prior to surgical resection. Neuro
with brain metastases treated with whole-brain radia- Oncol. 2015;17:580–7.
tion and motexafin gadolinium: results of a random- 43. Patchell RA, Tibbs PA, Walsh JW, Dempsey
ized phase III trial. J Clin Oncol. 2004;22:157–65. RJ, Maruyama Y, Kryscio RJ, Markesbery WR,
31. Aoyama H, Tago M, Kato N, et al. Neurocognitive Macdonald JS, Young B. A randomized trial of
function of patients with brain metastasis who surgery in the treatment of single metastases to the
received either whole brain radiotherapy plus ste- brain. N Engl J Med. 1990;322:494–500.
reotactic radiosurgery or radiosurgery alone. Int J 44. Patchell RA, Tibbs PA, Regine WF, Dempsey RJ,
Radiat Oncol Biol Phys. 2007;68:1388–95. Mohiuddin M, Kryscio RJ, Markesbery WR, Foon
32. Habets EJJ, Dirven L, Wiggenraad RG, Verbeek-De KA, Young B. Postoperative radiotherapy in the
Kanter A, Lycklama À, Nijeholt GJ, Zwinkels H, treatment of single metastases to the brain: a ran-
Klein M, Taphoorn MJB. Neurocognitive function- domized trial. JAMA. 1998;280:1485–9.
ing and health-related quality of life in patients 45. Wu AS, Witgert ME, Lang FF, Xiao L, Bekele BN,
treated with stereotactic radiotherapy for brain Meyers CA, Ferson D, Wefel JS. Neurocognitive
metastases: a prospective study. Neuro Oncol. function before and after surgery for insular gliomas.
2016;18:435–44. J Neurosurg. 2011;115:1115–25.
33. Witgert ME, Meyers CA. Neurocognitive and qual- 46. Talacchi A, Santini B, Savazzi S, Gerosa M. Cog-
ity of life measures in patients with metastatic brain nitive effects of tumour and surgical treatment in
disease. Neurosurg Clin N Am. 2011;22:79–85. glioma patients. J Neurooncol. 2011;103:541–9.
47. Noll KR, Weinberg JS, Ziu M, Benveniste RJ, Suki bine combination in inoperable brain metastasis
D, Wefel JS. Neurocognitive changes associated of non-small-cell lung cancer: Groupe Francais de
with surgical resection of left and right temporal Pneumo-Cancerologie (GFPC) Protocol 95-1. Ann
lobe glioma. Neurosurgery. 2015;77:777–85. Oncol. 2001;12:59–67.
48. Hoffermann M, Bruckmann L, Mahdy Ali K, Zaar 61. Sperduto PW, Wang M, Robins HI, et al. A phase
K, Avian A, von Campe G. Pre- and postopera- 3 trial of whole brain radiation therapy and stereo-
tive neurocognitive deficits in brain tumor patients tactic radiosurgery alone versus WBRT and SRS
assessed by a computer based screening test. J Clin with temozolomide or erlotinib for non-small cell
Neurosci. 2017;36:31–6. lung cancer and 1 to 3 brain metastases: Radiation
49. Koppelmans V, Breteler MMB, Boogerd W, Sey- Therapy Oncology Group 0320. Int J Radiat Oncol
naeve C, Gundy C, Schagen SB. Neuropsycho- Biol Phys. 2013;85:1312–8.
logical performance in survivors of breast cancer 62. Stupp R, Mason WP, van den Bent MJ, et al.
more than 20 years after adjuvant chemotherapy. J Radiotherapy plus concomitant and adjuvant
Clin Oncol. 2012;30:1080–6. temozolomide for glioblastoma. N Engl J Med.
50. Videnovic A, Semenov I, Chua-Adajar R, et al. 2005;352:987–96.
Capecitabine-induced multifocal leukoencephalopa- 63. Agarwala SS, Kirkwood JM, Gore M, Dreno B,
thy: a report of five cases. Neurology. 2005;65:1792– Thatcher N, Czarnetski B, Atkins M, Buzaid A,
4; discussion 1685. Skarlos D, Rankin EM. Temozolomide for the
51. Wefel JS, Schagen SB. Chemotherapy-related cog- treatment of brain metastases associated with meta-
nitive dysfunction. Curr Neurol Neurosci Rep. static melanoma: a phase II study. J Clin Oncol.
2012;12:267–75. 2004;22:2101–7.
52. Ahles TA, Saykin AJ. Candidate mechanisms for 64. Abrey LE, Olson JD, Raizer JJ, Mack M, Rodavitch
chemotherapy-induced cognitive changes. Nat Rev A, Boutros DY, Malkin MG. A phase II trial of temo-
Cancer. 2007;7:192–201. zolomide for patients with recurrent or progressive
53. Koppelmans V, de Ruiter MB, van der Lijn F, brain metastases. J Neurooncol. 2001;53:259–65.
Boogerd W, Seynaeve C, van der Lugt A, Vrooman 65. Antonadou D, Paraskevaidis M, Sarris G, Coliarakis
H, Niessen WJ, Breteler MMB, Schagen SB. Global N, Economou I, Karageorgis P, Throuvalas N. Phase
and focal brain volume in long-term breast cancer II randomized trial of temozolomide and concurrent
survivors exposed to adjuvant chemotherapy. Breast radiotherapy in patients with brain metastases. J Clin
Cancer Res Treat. 2012;132:1099–106. Oncol. 2002;20:3644–50.
54. Koppelmans V, de Groot M, de Ruiter MB, Boogerd 66. Fiveash JB, Arafat WO, Naoum GE, Guthrie BL,
W, Seynaeve C, Vernooij MW, Niessen WJ, Schagen Sawrie SM, Spencer SA, Meredith RF, Markert JM,
SB, Breteler MMB. Global and focal white mat- Conry RM, Nabors BL. A phase 2 study of radiosur-
ter integrity in breast cancer survivors 20 years gery and temozolomide for patients with 1 to 4 brain
after adjuvant chemotherapy. Hum Brain Mapp. metastases. Adv Radiat Oncol. 2016;1:83–8.
2014;35:889–99. 67. Brastianos PK, Carter SL, Santagata S, et al.
55. Heimans JJ, Vermorken JB, Wolbers JG, Eeltink CM, Genomic characterization of brain metastases
Meijer OW, Taphoorn MJ, Beijnen JH. Paclitaxel reveals branched evolution and potential therapeutic
(Taxol) concentrations in brain tumor tissue. Ann targets. Cancer Discov. 2015;5:1164–77.
Oncol. 1994;5:951–3. 68. Saunus JM, Quinn MCJ, Patch A-M, et al. Integrated
56. Stewart DJ. A critique of the role of the blood–brain genomic and transcriptomic analysis of human brain
barrier in the chemotherapy of human brain tumors. metastases identifies alterations of potential clinical
J Neurooncol. 1994;20:121–39. significance. J Pathol. 2015;237:363–78.
57. Deeken JF, Loscher W. The blood–brain barrier and 69. Lazaro T, Brastianos PK. Immunotherapy and
cancer: transporters, treatment, and Trojan horses. targeted therapy in brain metastases : emerg-
Clin Cancer Res. 2007;13:1663–74. ing options in precision medicine. CNS Oncol.
58. Postmus PE, Haaxma-Reiche H, Gregor A, Groen 2017;6:139–51.
HJM, Lewinski T, Scolard T, Kirkpatrick A, Curran 70. Neagu MR, Gill CM, Batchelor TT, Brastianos
D, Sahmoud T, Giaccone G. Brain-only metastases PK. Genomic profiling of brain metastases: cur-
of small cell lung cancer; efficacy of whole brain rent knowledge and new frontiers. Chin Clin Oncol.
radiotherapy. An EORTC phase II study. Radiother 2015;4:22.
Oncol. 1998;46:29–32. 71. Tawbi HA, Boutros C, Kok D, Robert C. New era in
59. Mornex F, Thomas L, Mohr P, et al. A prospective the management of melanoma brain metastases. Am
randomized multicentre phase III trial of fotemus- Soc Clin Oncol Educ Book. 2018;38:741–50.
tine plus whole brain irradiation versus fotemustine 72. Hendriks LE, Schoenmaekers J, Zindler JD, Eekers
alone in cerebral metastases of malignant melanoma. DB, Hoeben A, De Ruysscher DK, Dingemans
Melanoma Res. 2003;13:97–103. AM. Safety of cranial radiotherapy concurrent with
60. Robinet G, Thomas P, Breton JL, et al. Results of a tyrosine kinase inhibitors in non-small cell lung can-
phase III study of early versus delayed whole brain cer patients: a systematic review. Cancer Treat Rev.
radiotherapy with concurrent cisplatin and vinorel- 2015;41:634–45.
73. Welsh JW, Komaki R, Amini A, et al. Phase II trial 83. Suh JH. Stereotactic radiosurgery for the man-
of erlotinib plus concurrent whole-brain radia- agement of brain metastases. N Engl J Med.
tion therapy for patients with brain metastases 2010;362:1119–27.
from non-small-cell lung cancer. J Clin Oncol. 84. Yamamoto M, Serizawa T, Higuchi Y, et al. A multi-
2013;31:895–902. institutional prospective observational study of ste-
74. Pesce GA, Klingbiel D, Ribi K, et al. Outcome, reotactic radiosurgery for patients with multiple brain
quality of life and cognitive function of patients metastases (JLGK0901 study update): irradiation-
with brain metastases from non-small cell lung can- related complications and long-term maintenance of
cer treated with whole brain radiotherapy combined mini-mental state examination scores. Int J Radiat
with gefitinib or temozolomide. A randomised Oncol Biol Phys. 2017;99:31–40.
phase II trial of the Swiss Group for Clinical 85. Brown PD, Jaeckle K, Ballman KV, et al. Effect of
Cancer Research (SAKK 70/03). Eur J Cancer. radiosurgery alone vs radiosurgery with whole brain
2012;48:377–84. radiation therapy on cognitive function in patients
75. Baruch K, Deczkowska A, Rosenzweig N, Tsitsou- with 1 to 3 brain metastases: a randomized clinical
Kampeli A, Sharif AM, Matcovitch-Natan O, trial. JAMA. 2016;316:401–9.
Kertser A, David E, Amit I, Schwartz M. PD-1 86. Chang EL, Wefel JS, Hess KR, et al. Neurocognition
immune checkpoint blockade reduces pathology and in patients with brain metastases treated with radio-
improves memory in mouse models of Alzheimer’s surgery or radiosurgery plus whole-brain irradia-
disease. Nat Med. 2016;22:135–7. tion: a randomised controlled trial. Lancet Oncol.
76. Goldberg SB, Gettinger SN, Mahajan A, et al. 2009;10:1037–44.
Pembrolizumab for patients with melanoma or non- 87. Kocher M, Soffietti R, Abacioglu U, et al. Adjuvant
small-cell lung cancer and untreated brain metasta- whole-brain radiotherapy versus observation after
ses: early analysis of a non-randomised, open-label, radiosurgery or surgical resection of one to three
phase 2 trial. Lancet Oncol. 2016;17:976–83. cerebral metastases: results of the EORTC 22952-
77. McGinnis GJ, Raber J. CNS side effects of immune 26001 study. J Clin Oncol. 2011;29:134–41.
checkpoint inhibitors: preclinical models, genet- 88. Aoyama H, Shirato H, Tago M, et al. Stereotactic
ics and multimodality therapy. Immunotherapy. radiosurgery plus whole-brain radiation therapy vs
2017;9:929–41. stereotactic radiosurgery alone for treatment of brain
78. Shibamoto Y, Baba F, Oda K, Hayashi S, Kokubo M, metastases: a randomized controlled trial. JAMA.
Ishihara S-I, Itoh Y, Ogino H, Koizumi M. Incidence 2006;295:2483–91.
of brain atrophy and decline in mini-mental state 89. Soffietti R, Kocher M, Abacioglu UM, et al. A
examination score after whole-brain radiotherapy in European Organisation for Research and Treatment
patients with brain metastases: a prospective study. of Cancer phase III trial of adjuvant whole-brain
Int J Radiat Oncol Biol Phys. 2008;72:1168–73. radiotherapy versus observation in patients with one
79. Brown PD, Ballman KV, Cerhan JH, et al. to three brain metastases from solid tumors after sur-
Postoperative stereotactic radiosurgery compared gical resection or radiosurgery: quality-of-life. J Clin
with whole brain radiotherapy for resected meta- Oncol. 2013;31:65–72.
static brain disease (NCCTG N107C/CEC·3): a 90. Meyers CA, Wefel JS. The use of the mini-mental
multicentre, randomised, controlled, phase 3 trial. state examination to assess cognitive functioning in
Lancet Oncol. 2017;18:1049–60. cancer trials: no ifs, ands, buts, or sensitivity. J Clin
80. Kepka L, Tyc-Szczepaniak D, Bujko K, Olszyna- Oncol. 2003;21:3557–8.
Serementa M, Michalski W, Sprawka A, Trabska- 91. Sul J, Kluetz P, Papadopopoulos E, Keegan P. Clinical
Kluch B, Komosinska K, Wasilewska-Tesluk E, outcome assessments in neuro-oncology: a regula-
Czeremszynska B. Stereotactic radiotherapy of tory perspective. Neurooncol Pract. 2016;3:4–9.
the tumor bed compared to whole brain radio- 92. Wefel JS, Vardy J, Ahles T, Schagen SB. International
therapy after surgery of single brain metastasis: cognition and Cancer Task Force recommendations
results from a randomized trial. Radiother Oncol. to harmonise studies of cognitive function in patients
2016;121:217–24. with cancer. Lancet Oncol. 2011;12:703–8.
81. Martinez P, Mak RH, Oxnard GR. Targeted ther- 93. Wefel JS, Cloughesy T, Zazzali JL, et al.
apy as an alternative to whole-brain radiotherapy Neurocognitive function in patients with recur-
in EGFR-mutant or ALK-positive non-small-cell rent glioblastoma treated with bevacizumab. Neuro
lung cancer with brain metastases. JAMA Oncol. Oncol. 2011;13:660–8.
2017;3:1274–5. 94. Monaco EA 3rd, Faraji AH, Berkowitz O, Parry PV,
82. Mulvenna P, Nankivell M, Barton R, et al. Hadelsberg U, Kano H, Niranjan A, Kondziolka D,
Dexamethasone and supportive care with or without Lunsford LD. Leukoencephalopathy after whole-
whole brain radiotherapy in treating patients with brain radiation therapy plus radiosurgery versus
non-small cell lung cancer with brain metastases radiosurgery alone for metastatic lung cancer.
unsuitable for resection or stereotactic radiotherapy Cancer. 2013;119:226–32.
(QUARTZ): results from a phase 3, non-inferiority, 95. Ebi J, Sato H, Nakajima M, Shishido F. Incidence
randomised trial. Lancet. 2016;388:2004–14. of leukoencephalopathy after whole-brain radiation
therapy for brain metastases. Int J Radiat Oncol Biol with limited-disease. Int J Radiat Oncol Biol Phys.
Phys. 2013;85:1212–7. 2011;81:77–84.
96. Auperin A, Arriagada R, Pignon JP, et al. Pro- 107. Le Pechoux C, Dunant A, Senan S, et al. Standard-
phylactic cranial irradiation for patients with dose versus higher-dose prophylactic cranial irradia-
small-cell lung cancer in complete remission. Pro- tion (PCI) in patients with limited-stage small-cell
phylactic Cranial Irradiation Overview Collabora- lung cancer in complete remission after chemother-
tive Group. N Engl J Med. 1999;341:476–84. apy and thoracic radiotherapy (PCI 99-01, EORTC
97. Slotman BJ, van Tinteren H, Praag JO, Knegjens JL, 22003-08004, RTOG 0212, and IFCT 99-01): : a.
El Sharouni SY, Hatton M, Keijser A, Faivre-Finn C, Lancet Oncol. 2009;10:467–74.
Senan S. Use of thoracic radiotherapy for extensive 108. Gondi V, Paulus R, Bruner DW, Meyers CA, Gore
stage small-cell lung cancer: a phase 3 randomised EM, Wolfson A, Werner-Wasik M, Sun AY, Choy H,
controlled trial. Lancet. 2015;385:36–42. Movsas B. Decline in tested and self-reported cogni-
98. Umsawasdi T, Valdivieso M, Chen TT, et al. Role tive functioning after prophylactic cranial irradiation
of elective brain irradiation during combined chemo- for lung cancer: pooled secondary analysis of radia-
radiotherapy for limited disease non-small cell lung tion therapy oncology group randomized trials 0212
cancer. J Neurooncol. 1984;2:253–9. and 0214. Int J Radiat Oncol. 2013;86:656–64.
99. Miller T, Crowley J, Mira J, Schwartz J, Hutchins L, 109. Sheline GE, Wara WM, Smith V. Therapeutic irradi-
Baker L. A randomized trial of chemotherapy and ation and brain injury. Int J Radiat Oncol Biol Phys.
radiotherapy for stage III non-small cell lung can- 1980;6:1215–28.
cere. Cancer Ther. 1998;1:229–36. 110. Giordano FA, Welzel G, Abo-Madyan Y, Wenz
100. Cox JD, Stanley K, Petrovich Z, Al E. Cranial irra- F. Potential toxicities of prophylactic cranial irradia-
diation in cancer of the lung of all cell types. JAMA. tion. Transl Lung Cancer Res. 2012;1:254–62.
1981;245:469–72. 111. Young DF, Posner JB, Chu F, Nisce L. Rapid-course
101. Gore EM, Bae K, Wong SJ, Sun A, Bonner JA, radiation therapy of cerebral metastases: results and
Schild SE, Gaspar LE, Bogart JA, Werner-Wasik complications. Cancer. 1974;34:1069–76.
M, Choy H. Phase III comparison of prophy- 112. Soussain C, Ricard D, Fike JR, Mazeron JJ, Psimaras
lactic cranial irradiation versus observation in D, Delattre JY. CNS complications of radiotherapy
patients with locally advanced non-small-cell lung and chemotherapy. Lancet. 2009;374:1639–51.
cancer: primary analysis of Radiation Therapy 113. Wong CS, Van der Kogel AJ. Mechanisms of radia-
Oncology Group study RTOG 0214. J Clin Oncol. tion injury to the central nervous system: implications
2011;29:272–8. for neuroprotection. Mol Interv. 2004;4:273–84.
102. Gore E, Paulus R, Wong S, Sun A, Videtic G, Dutta 114. Li Y-Q, Chen P, Jain V, Reilly RM, Wong CS. Early
S. Phase III comparison of prophylactic cranial irra- radiation-induced endothelial cell loss and blood-
diation versus observation in patients with locally spinal cord barrier breakdown in the rat spinal cord.
advanced non-small cell lung cancer-an updated Radiat Res. 2004;161:143–52.
analysis of RTOG 0214. Int J Radiat Oncol Biol 115. Yuan H, Gaber MW, Boyd K, Wilson CM, Kiani
Phys. 2012;84:S103. MF, Merchant TE. Effects of fractionated radiation
103. Li N, Zeng ZF, Wang SY, et al. Randomized phase III on the brain vasculature in a murine model: blood–
trial of prophylactic cranial irradiation versus obser- brain barrier permeability, astrocyte proliferation,
vation in patients with fully resected stage IIIA-N2 and ultrastructural changes. Int J Radiat Oncol Biol
nonsmall-cell lung cancer and high risk of cerebral Phys. 2006;66:860–6.
metastases after adjuvant chemotherapy. Ann Oncol. 116. Balentova S, Adamkov M. Molecular, cellular and
2015;26:504–9. functional effects of radiation-induced brain injury:
104. Al Feghali KA, Ballout RA, Khamis AM, Akl EA, a review. Int J Mol Sci. 2015;16:27796–815.
Geara FB. Prophylactic cranial irradiation in patients 117. Uzal D, Ozyar E, Hayran M, Zorlu F, Atahan L,
with non-small-cell lung cancer: a systematic review Yetkin S. Reduced incidence of the somnolence
and meta-analysis of randomized controlled trials. syndrome after prophylactic cranial irradiation
Front Oncol. 2018;8:115. in children with acute lymphoblastic leukemia.
105. Sun A, Bae K, Gore EM, et al. Phase III trial of pro- Radiother Oncol. 1998;48:29–32.
phylactic cranial irradiation compared with observa- 118. Littman P, Rosenstock J, Gale G, Krisch RE,
tion in patients with locally advanced non-small-cell Meadows A, Sather H, Coccia P, Decamargo B. The
lung cancer: neurocognitive and quality-of-life anal- somnolence syndrome in leukemic children follow-
ysis. J Clin Oncol. 2011;29:279–86. ing reduced daily dose fractions of cranial radiation.
106. Wolfson AH, Bae K, Komaki R, Meyers C, Movsas Int J Radiat Oncol Biol Phys. 1984;10:1851–3.
B, Le Pechoux C, Werner-Wasik M, Videtic GM, 119. Faithfull S, Brada M. Somnolence syndrome in
Garces YI, Choy H. Primary analysis of a phase adults following cranial irradiation for primary brain
II randomized trial Radiation Therapy Oncology tumours. Clin Oncol (R Coll Radiol). 1998;10:250–4.
Group (RTOG) 0212: impact of different total doses 120. Powell C, Guerrero D, Sardell S, Cumins S,

and schedules of prophylactic cranial irradiation on Wharram B, Traish D, Gonsalves A, Ashley S,
chronic neurotoxicity and quality of life for patients Brada M. Somnolence syndrome in patients
receiving radical radiotherapy for primary brain receptor antagonist, L-158,809, prevents or amelio-
tumours: a prospective study. Radiother Oncol. rates fractionated whole-brain irradiation-induced
2011;100:131–6. cognitive impairment. Int J Radiat Oncol Biol Phys.
121. Welzel G, Fleckenstein K, Schaefer J, Hermann B, 2009;73:499–505.
Kraus-Tiefenbacher U, Mai SK, Wenz F. Memory 134. Senzer N. Rationale for a phase III study of eryth-
function before and after whole brain radiotherapy ropoietin as a neurocognitive protectant in patients
in patients with and without brain metastases. Int J with lung cancer receiving prophylactic cranial irra-
Radiat Oncol Biol Phys. 2008;72:1311–8. diation. Semin Oncol. 2002;29:47–52.
122. Monje ML, Toda H, Palmer TD. Inflammatory 135. Malaterre J, McPherson CS, Denoyer D, et al.
blockade restores adult hippocampal neurogenesis. Enhanced lithium-induced brain recovery following
Science. 2003;302(80):1760–5. cranial irradiation is not impeded by inflammation.
123. Kyrkanides S, Moore AH, Olschowka JA, Stem Cells Transl Med. 2012;1:469–79.
Daeschner JC, Williams JP, Hansen JT, Kerry 136. Huo K, Sun Y, Li H, Du X, Wang X, Karlsson N, Zhu
O’Banion M. Cyclooxygenase-2 modulates brain C, Blomgren K. Lithium reduced neural progeni-
inflammation-related gene expression in central ner- tor apoptosis in the hippocampus and ameliorated
vous system radiation injury. Brain Res Mol Brain functional deficits after irradiation to the immature
Res. 2002;104:159–69. mouse brain. Mol Cell Neurosci. 2012;51:32–42.
124. Tsao MN, Li YQ, Lu G, Xu Y, Wong 137. Brown PD, Pugh S, Laack NN, et al. Memantine for
CS. Upregulation of vascular endothelial growth fac- the prevention of cognitive dysfunction in patients
tor is associated with radiation-induced blood-spinal receiving whole-brain radiotherapy: a randomized,
cord barrier breakdown. J Neuropathol Exp Neurol. double-blind, placebo-controlled trial. Neuro Oncol.
1999;58:1051–60. 2013;15:1429–37.
125. Li YQ, Ballinger JR, Nordal RA, Su ZF, Wong 138. Rapp SR, Case LD, Peiffer A, et al. Donepezil for
CS. Hypoxia in radiation-induced blood-spinal cord irradiated brain tumor survivors: a phase III random-
barrier breakdown. Cancer Res. 2001;61:3348–54. ized placebo-controlled clinical trial. J Clin Oncol.
126. Monje ML, Mizumatsu S, Fike JR, Palmer 2015;33:1653–9.
TD. Irradiation induces neural precursor-cell dys- 139. Salloway S, Mintzer J, Cummings JL, Geldmacher
function. Nat Med. 2002;8:955–62. D, Sun Y, Yardley J, Mackell J. Subgroup analy-
127. Furuse M, Nonoguchi N, Kawabata S, Miyatake S-I, sis of US and non-US patients in a global study of
Kuroiwa T. Delayed brain radiation necrosis: patho- high-dose donepezil (23 mg) in moderate and severe
logical review and new molecular targets for treat- Alzheimer’s disease. Am J Alzheimers Dis Other
ment. Med Mol Morphol. 2015;48:183–90. Demen. 2012;27:421–32.
128. Johnson BE, Becker B, Goff WB, Petronas N, 140. Eriksson PS, Perfilieva E, Björk-Eriksson T,
Krehbiel MA, Makuch RW, McKenna G, Glatstein Alborn AM, Nordborg C, Peterson DA, Gage
E, Ihde DC. Neurologic, neuropsychologic, and FH. Neurogenesis in the adult human hippocampus.
computed cranial tomography scan abnormalities in Nat Med. 1998;4:1313–7.
2- to 10-year survivors of small-cell lung cancer. J 141. Gondi V, Tom WA, Mehta MP. Why avoid the hippo-
Clin Oncol. 1985;3:1659–67. campus? A comprehensive review. Radiother Oncol.
129. Johnson BE, Patronas N, Hayes W, et al. Neurologic, 2010;97:370–6.
computed cranial tomographic, and magnetic reso- 142. Gondi V, Pugh SL, Tome WA, et al. Preservation
nance imaging abnormalities in patients with small- of memory with conformal avoidance of the hippo-
cell lung cancer: further follow-up of 6- to 13-year campal neural stem-cell compartment during whole-
survivors. J Clin Oncol. 1990;8:48–56. brain radiotherapy for brain metastases (RTOG
130. Wefel JS, Deshmukh S, Brown PD, et al. Impact of 0933): a phase II multi-institutional trial. J Clin
apolipoprotein E (APOE) genotype on neurocogni- Oncol. 2014;32:3810–6.
tive function (NCF) in patients with brain metastasis 143. Gondi V, Deshmukh S, Brown P, et al Preservation
(BM): an analysis of NRG Oncology’s RTOG 0614. of neurocognitive function with conformal avoid-
J Clin Oncol. 2018;36:2065. ance of the hippocampus during whole-brain
131. Day J, Zienius K, Gehring K, Grosshans D, Taphoorn radiotherapy (HA-WBRT) for brain metastases:
M, Grant R, Li J, Brown PD. Interventions for pre- preliminary results of phase III trial NRG Oncology
venting and ameliorating cognitive deficits in adults CC001. Presented at the 2018 annual meeting of the
treated with cranial irradiation. Cochrane Database American Society of Radiation Oncology (ASTRO),
Syst Rev. 2014; https://doi.org/10.1002/14651858. San Antonio, TX, October 23, 2018.
CD011335.pub2. 144. Rodriguez de Dios N, Counago F, Lopez JL, et al.
132. Jenrow KA, Brown SL, Liu J, Kolozsvary A, Treatment design and rationale for a randomized trial
Lapanowski K, Kim JH. Ramipril mitigates of prophylactic cranial irradiation with or without
radiation-induced impairment of neurogenesis in the hippocampal avoidance for SCLC: PREMER trial
rat dentate gyrus. Radiat Oncol. 2010;5:6. on behalf of the Oncologic Group for the Study of
133. Robbins ME, Payne V, Tommasi E, Diz DI, Hsu FC, Lung Cancer/Spanish Radiation Oncology Group-
Brown WR, Wheeler KT, Olson J, Zhao W. The AT1 Radiation. Clin Lung Cancer. 2018;19:e693–7.
145. Wefel JS, Kesler SR, Noll KR, Schagen SB. Clinical and education day of the Society for Neuro-Oncology,
characteristics, pathophysiology, and manage- San Antonio, TX, November 21, 2015.
ment of noncentral nervous system cancer-related 156. Ouimet LA, Stewart A, Collins B, Schindler D,
cognitive impairment in adults. CA Cancer J Clin. Bielajew C. Measuring neuropsychological change
2015;65:123–38. following breast cancer treatment: an analysis
146. Maschio M, Dinapoli L, Fabi A, Giannarelli D, of statistical models. J Clin Exp Neuropsychol.
Cantelmi T. Cognitive rehabilitation training in 2009;31:73–89.
patients with brain tumor-related epilepsy and 157. Soon YY, Tham IWK, Lim KH, Koh WY, Lu
cognitive deficits: a pilot study. J Neurooncol. JJ. Surgery or radiosurgery plus whole brain radio-
2015;125:419–26. therapy versus surgery or radiosurgery alone for
147. Han EY, Chun MH, Kim BR, Kim HJ. Functional brain metastases. Cochrane Database Syst Rev.
improvement after 4-week rehabilitation therapy and 2014, 2014;(3):CD009454.
effects of attention deficit in brain tumor patients: 158. Meyers CA, Brown PD. Role and relevance of neu-
comparison with subacute stroke patients. Ann rocognitive assessment in clinical trials of patients
Rehabil Med. 2015;39:560–9. with CNS tumors. J Clin Oncol. 2006;24:1305–9.
148. Zucchella C, Capone A, Codella V, De Nunzio 159. Reardon DA, Galanis E, DeGroot JF, et al. Clinical
AM, Vecchione C, Sandrini G, Pace A, Pierelli trial end points for high-grade glioma: the evolving
F, Bartolo M. Cognitive rehabilitation for early landscape. Neuro Oncol. 2011;13:353–61.
post-surgery inpatients affected by primary brain 160. van den Bent MJ, Wefel JS, Schiff D, et al. Response
tumor: a randomized, controlled trial. J Neurooncol. assessment in neuro-oncology (a report of the RANO
2013;114:93–100. group): assessment of outcome in trials of diffuse
149. Cherrier MM, Anderson K, David D, Higano CS, low-grade gliomas. Lancet Oncol. 2011;12:583–93.
Gray H, Church A, Willis SL. A randomized trial of 161. Blakeley JO, Coons SJ, Corboy JR, Kline Leidy N,
cognitive rehabilitation in cancer survivors. Life Sci. Mendoza TR, Wefel JS. Clinical outcome assess-
2013;93:617–22. ment in malignant glioma trials: measuring signs,
150. Ercoli LM, Castellon SA, Hunter AM, Kwan L, symptoms, and functional limitations. Neuro Oncol.
Kahn-Mills BA, Cernin PA, Leuchter AF, Ganz 2016;18 Suppl 2:ii13–20.
PA. Assessment of the feasibility of a rehabilitation 162. Alexander BM, Brown PD, Ahluwalia MS, et al.
intervention program for breast cancer survivors Clinical trial design for local therapies for brain
with cognitive complaints. Brain Imaging Behav. metastases: a guideline by the Response Assessment
2013;7:543–53. in Neuro-Oncology Brain Metastases working
151. Ferguson RJ, Ahles TA, Saykin AJ, McDonald BC, group. Lancet Oncol. 2018;19:e33–42.
Furstenberg CT, Cole BF, Mott LA. Cognitive- 163. Camidge DR, Lee EQ, Lin NU, et al. Clinical
behavioral management of chemotherapy-related trial design for systemic agents in patients with
cognitive change. Psychooncology. 2007;16:772–7. brain metastases from solid tumours: a guideline
152. Ferguson RJ, McDonald BC, Rocque MA, by the Response Assessment in Neuro-Oncology
Furstenberg CT, Horrigan S, Ahles TA, Saykin Brain Metastases working group. Lancet Oncol.
AJ. Development of CBT for chemotherapy-related 2018;19:e20–32.
cognitive change: results of a waitlist control trial. 164. Meyers CA, Hess KR. Multifaceted end points
Psychooncology. 2012;21:176–86. in brain tumor clinical trials: cognitive deterio-
153. McDougall GJ, Becker H, Acee TW, Vaughan PW, ration precedes MRI progression. Neuro Oncol.
Delville CL. Symptom management of affective and 2003;5:89–95.
cognitive disturbance with a group of cancer survi- 165. Brandt J, Benedict R. Hopkins Verbal Learning
vors. Arch Psychiatr Nurs. 2011;25:24–35. Test professional manual – revised: Psychological
154. Kesler S, Hadi Hosseini SM, Heckler C, Janelsins M, Assessment Resources, Inc; Lutz, Florida: 1991.
Palesh O, Mustian K, Morrow G. Cognitive training 166. Benton AL, Hamscher KD. Multilingual aphasia
for improving executive function in chemotherapy- examination. Iowa City: AJA Associates; 1989.
treated breast cancer survivors. Clin Breast Cancer. 167. Army Individual Test Battery. Manual of directions
2013;13:299–306. and scoring. Washington, DC: War Department,
155. Wefel J, Bradshaw M, Sullaway C, Gilbert M, Adjutant General’s Office; 1944.
Armstrong T (2015) A brain-plasticity based com- 168. Arbuthnott K, Frank J. Trail making test, part B as
puterized intervention to treat attention and memory a measure of executive control: validation using a
problems in adult brain tumor (BT) survivors. Poster set-switching paradigm. J Clin Exp Neuropsychol.
session presented at the 20th annual scientific meeting 2000;22:518–28.
Part IV
Surgical Treatment of Brain Metastasis
The Role of Surgery
in the Management of Brain
30
Metastases
Jeffrey I. Traylor, Aditya Srivatsan, Rajan Patel,

Rohan Ramakrishna, and Ganesh Rao
Background and Overview the overall outcome and response to treatment,

management of these patients may be divided
Brain metastases are a devastating cause of can- into four categories: systemic radiation, whole-
cer morbidity, comprising more than half of all brain radiation therapy (WBRT), stereotactic
intracranial malignancies. In the United States, radiosurgery (SRS), and surgical resection. It
brain metastases occur in more than 100,000 peo- also should be noted that newer targeted and
ple each year, affecting between 10% and 30% of immunotherapies have demonstrated efficacy for
all patients with systemic malignancy [1]. brain metastases as well [4–6]. Trials in the last
Additionally, 2% of all cancer patients as well as two decades have illuminated primary prognostic
12.5% of patients with systemic disease will have factors that contribute to an overall management
brain metastases discovered at diagnosis [2]. plan, including baseline neurologic status, often
Within this patient population, the most common measured by the Karnofsky Performance Scale
underlying tumors are lung cancer (28%), mela- (KPS), combined performance status, and tumor
noma (21%), and renal carcinoma (19%) with histology measured by the Graded Prognostic
numerous other primary tumors contributing to Assessment (GPA), age (<65 years), and control
the annual incidence [3]. of the primary malignancy [7]. Based on these
The development of brain metastases repre- and other variables, a combined therapeutic regi-
sents a terminal stage in the management of the men may optimize patient outcomes.
primary tumor with a prior median survival of 5 Surgical resection has held a defined role in
months from diagnosis [2]. However, there have the management of brain metastases since the
been numerous advancements in the treatment of early 1990s. Since that time, numerous advances
this disease process and median overall survival have been made in neurosurgical techniques with
rates are now demonstrating a significant additional trials investigating the efficacy of this
increase. Though a host of factors contribute to treatment option. A seminal study was reported
in 1990, when Patchell et al. compared resection
followed by WBRT to biopsy and WBRT in
J. I. Traylor · A. Srivatsan · R. Patel · G. Rao (*)
Department of Neurosurgery, The University of Texas patients with a single brain metastasis. In addi-
M.D. Anderson Cancer Center, Houston, TX, USA tion to demonstrating improved survival in the
e-mail: grao@mdanderson.org combined treatment cohort, the authors reported
R. Ramakrishna an increased rate of local tumor control in the
Department of Neurological Surgery, group undergoing resection [8]. The next major
NewYork-Presbyterian Hospital, Weill Cornell
trial was completed 3 years later with a slightly
Medicine, New York, NY, USA

https://doi.org/10.1007/978-3-030-42958-4_30
430 J. I. Traylor et al.
larger cohort of 63 patients with single brain Presurgical Evaluation

metastases showing similar findings. The authors
reported increased survival in the cohort receiv- Cerebral metastases have numerous clinical man-
ing combined resection and WBRT but also ifestations and should be suspected in patients
established that surgery was most beneficial for with underlying malignancy with neurologic
patients with stable extracranial disease [9]. symptoms or behavioral changes. In these
Conversely, Mintz et al. reported no survival ben- patients, both mass effect and cerebral edema
efit for surgical resection of brain metastases in a play a role in symptomatology. The tumor can
1996 trial [10]. However, these findings have compress adjacent structures resulting in corre-
since been disputed as a higher proportion of the sponding neurologic deficits. Additionally, the
cohort had significant extracranial disease and cumulative mass of these lesions can further dis-
lower neurologic baseline than the subjects of place cerebral tissue and increase intracranial
preceding trials. pressure. Should these masses interfere with nor-
Studies have investigated the role of surgery in mal cortical depolarization, new-onset epilepsy
multiple and recurrent brain metastases. Wrónski can occur which is often a presenting symptom.
et al. reported no outcome difference in patients Vasogenic edema resulting from tumor disrup-
with single and multiple cerebral metastases, and tion of the blood–brain barrier, allowing protein-
Iwadate and colleagues published comparable aceous fluid into the extracellular space, is
findings [11, 12]. Similarly, studies have revealed another mechanism by which metastatic lesions
a survival benefit for reoperation in patients with can contribute to increased intracranial pressure
tumor recurrence, further demonstrating the (Fig. 30.1). Yet another rare manifestation of
applicability and efficacy of surgical resection in intracranial spread of underlying malignancy is
this patient population [13–15]. More recent stroke, as the tumor can compress cerebral vascu-
research has investigated the benefit of novel sur- lature or hemorrhage. However, even within this
gical techniques, incorporating stereotactic guid- high-risk patient demographic, the rates of brain
ance as a means of more precise margination, metastases among those with new-onset neuro-
advanced imaging sequences capable of delineat- logic symptoms are low, highlighting the impor-
ing critical white matter structures, and the utility tance of additional imaging and evaluation for
of surgery under awake conditions for tumor underlying disease before establishing a concrete
resections in eloquent cortical regions to mini- diagnosis [18]. More oncologists are recognizing
mize postoperative neurologic deficits [16]. the elevated risk of specific cancers to metasta-
Surgical resection for cerebral metastasis has size to the brain (e.g., lung and melanoma) and
become a mainstay for lesions that are accessible are frequently screening patients for brain metas-
and of sufficient size to warrant removal. More tases at initial presentation. Further, prophylactic
recently, it has been established there may be a cranial irradiation may be indicated with certain
role for tissue sampling of brain metastases in an cancers that have an especially high risk of pre-
effort to genomically characterize these lesions senting with brain metastases (e.g., small-cell
as they can often be distinct from the primary lung cancer).
tumor with brain-specific targetable mutations
(this is not standard of care, however) [17].
Depending on the proximity of the tumor mass to Prognosis
eloquent cortical and subcortical structures, sur-
gery carries a risk of worsening pre-existing neu- The prognosis of a patient with brain metasta-
rologic deficits in addition to risks of hemorrhage ses has a significant influence on the decision
and infection associated with any operative pro- for operative intervention. Evaluation of a neu-
cedure. Even with these risks, neurosurgery for rologic baseline is another important factor in
patients with the appropriate indications has been determining whether surgery will improve the
associated with shorter hospitalizations and over- quality of life. To date, clinical trials have
all improved neurologic outcomes. demonstrated better outcomes for patients with
30 The Role of Surgery in the Management of Brain Metastases 431
a b
Fig. 30.1 (a) Axial T1-weighted postcontrast MRI of a temporal lobe metastasis. (b) Axial FLAIR MRI demonstrating
significant perilesional edema
stable extracranial disease and good neurologi- Preoperative Risk Assessment

cal function (usually defined as KPS ≥ 70).
Other factors that contribute to operative man- A preoperative risk assessment is crucial for
agement include surgical accessibility of the determining surgical candidacy as the risks of
metastasis and lesion-induced deficit, with the undergoing a craniotomy must be weighed against
potential for dramatic improvement in patient the potential benefits of relieved tumor burden.
quality of life. Though Sawaya et al. first dem- Assessment of cardiovascular health to prevent
onstrated the feasibility of resection of tumors intraoperative blood pressure fluctuations and
in more critical cortical structures (e.g., the myocardial ischemia as well as evaluation of
“eloquent” brain), deep-seated tumors in prox- baseline respiratory function is essential to good
imity to corticospinal or corticobulbar tracts surgical outcomes and efforts should be made to
are less amenable to resection with a high risk optimize the function of these organ systems.
of worsened neurologic dysfunction [19]. Hyperglycemia has also been reported to increase
Likewise, patients with poor prognoses and perioperative morbidity after neurosurgical proce-
low KPS are less suitable for surgery and are dures and must be adequately controlled in
conventionally managed with radiation, advance of surgery and kept between 100 mg/dL
although the specific modality may vary and 150 mg/dLperioperatively[21, 22].
depending on the size and number of lesions Reductions in hepatic and renal function also con-
[20]. It is important, however, to distinguish tribute significantly to surgical morbidity and
poor performance status that results from must be properly evaluated. Clinical decision-
symptomatic brain metastases versus poor per- making tools can provide insight into the feasibil-
formance status from systemic disease. The ity of surgical management for patients with
former may benefit from surgical excision underlying organ disease [23]. Renal function
while the latter may not. also has implications on fluid and electrolyte
homeostasis and thus modulates the approach to intracranial pressure may require admission and
maintenance of blood volume and osmolarity. more aggressive management including the
This is especially significant for neurosurgical administration of mannitol or hypertonic saline.
procedures where changes in plasma tonicity
greatly influence cerebral volume. Additionally,
the presence of peritumoral edema must be recog- Surgery Versus Radiation
nized and is generally managed with glucocorti-
coids to reduce the severity of neurological deficit The decision to pursue radiation or surgery in the
or headache (Fig. 30.2). Evidence of significant management of brain metastases often depends
a b
c d
Fig. 30.2 (a) Axial T1-weighted postcontrast MRI of a of dexamethasone, the lesion is considerably less edema-
right cerebellar metastasis from lung cancer. (b) Axial tous on axial T1-weighted postcontrast MRI and (d) axial
FLAIR MRI demonstrating significant perilesional edema FLAIR image showing less mass effect on the fourth
with mass effect on the fourth ventricle. (c) After 10 days ventricle
upon numerous factors including but not limited for LITT has existed for decades, only recently
to tumor size, number, accessibility, and patient has this technique been applied to the manage-
condition. As mentioned above, sometimes sur- ment of neurosurgical disorders, including brain
gery can be utilized for brain metastasis sampling metastases [28, 29]. However, because this ther-
in an effort to identify unique, targetable muta- apy is relatively new, it is primarily utilized for
tions in patients with multiple lesions. In general, the ablation of deep-seated tumors that are less
surgery is first-line therapy for single brain accessible to standard resection techniques or in
metastases of sufficient size to warrant resection, the management of recalcitrant radiation necro-
which may be 2 cm or greater. However, for sis. It is typically not used in the management of
lesions smaller in size or in locations not amena- newly diagnosed brain metastases.
ble to resection such as in close proximity to elo-
quent cortex, other approaches are utilized.
Stereotactic radiosurgery (SRS) is another tech- Intraoperative Management
nique for treatment that is less invasive than con-
ventional resection techniques. Though this Stereotactic Navigation
therapy has emerged as an alternative to resection
for cerebral metastases, particularly for small, Stereotactic navigation is a minimally invasive
deep-seated lesions, there is debate as to whether method for precisely locating targets of interest
or not SRS is superior for single lesions less than based on a three-dimensional (3D) coordinate
3.5 cm [23]. Multiple retrospective cohort studies system. Though the concept for this technique
have reported similar survival outcomes between has existed for over a century, advances in imag-
the two while a study by Bindal and colleagues ing sequences and delivery systems have made
reported significantly longer survival times and stereotaxy an integral part of numerous
reduced incidence of death from neurologic neurosurgical procedures requiring anatomical

causes with surgical resection [13]. A prospective precision. Initially, this technology was frame-
phase III trial comparing these modalities con- based requiring application of a stereotactic
cluded SRS to have similar local tumor control to frame to the patient’s head. In recent years, fra-
combined resection and WBRT, but with signifi- meless systems have been developed with
cantly less distant tumor control, though the increasing sophistication allowing the surgeon to
study was prematurely discontinued due to inad- register the volumetric MRI scan to the patient’s
equate patient accrual [24]. head at the beginning of surgery. During the
WBRT is often used as an adjunctive therapy operation, the 3D coordinates of the navigation
with surgery or SRS, though it remains the treat- probes and registered instruments will be super-
ment of choice for patients with numerous meta- imposed on preoperative imaging on the naviga-
static lesions. Altogether, WBRT has become less tion suite screen in multiple planes. Throughout
utilized as a single therapy as it may not improve the operation, the surgeon will be able to refer-
overall survival [25]. Additionally, studies have ence the precise location of instruments to ana-
demonstrated that WBRT advances cognitive tomical landmarks in real time in order to
decline in this patient population which also con- minimize collateral damage to adjacent structures
tributes to further selective implementation [26, and the associated morbidity. Additionally, the
27]. In the context of this combined evidence, ability of modern stereotactic delivery systems to
WBRT often serves as an alternative for poor sur- register compatible instruments has made it a
gical candidates and those with high intracranial core component of the LITT procedure; the pre-
tumor burden not amenable to more localized cision offered by stereotaxy allows for real-time
techniques. guidance of the ablation probe to the intracranial
Within the scope of operative therapy for brain target.
metastases, there is a newly described, minimally Technological advances in the last two decades
invasive surgical technique: laser interstitial ther- have made surgical resection a safer and more
mal ablation therapy (LITT). Though the concept reliable option for the management of cerebral
metastases. These advances have spurred new Diffusion Tensor Imaging (DTI)
research into the incorporation of these technol-
ogies into cerebral tumor resection and thus Diffusion tensor imaging (DTI) is another MRI
augmented the role of the neurosurgeon in the sequence that detects small movements of water
management of these patients. Advanced mag- molecules and is particularly useful for delineat-
netic resonance imaging (MRI) sequences, real- ing critical white matter tracts, such as the corti-
time intraoperative imaging guidance, and novel cospinal tract and arcuate fasciculus, due to their
cortical mapping techniques have all contrib- unique directionality; water molecules are more
uted to the modern landscape of neurosurgical displaced along the directionality of the axon
oncology. fiber than in the perpendicular direction [35].
Intraoperatively, DTI has been used to reduce
postoperative neurologic morbidity by minimiz-
Functional MRI (fMRI) ing damage to these critical white matter tracts.
By clearly visualizing these white matter bun-
The use of imaging for intracranial tumors is dles, surgeons can avoid their transgression
paramount to diagnosis, preoperative procedural intraoperatively [36]. Unlike fMRI, DTI does
planning, and postoperative surveillance. not rely on task-oriented feedback to display
Increasingly available for preoperative assess- axon bundles. Instead, this unique MR sequence
ment, functional MRI (fMRI) measures meta- delineates critical subcortical white matter struc-
bolic cerebral activity. The basis for fMRI is the tures that can be superimposed on a neuronavi-
quantification of changes in blood flow corre- gation model for real-time stereotactic feedback
sponding to neural depolarization based on the (Fig. 30.3). Though DTI mapping of white mat-
variable magnetic properties of hemoglobin ter structures has been shown to improve the
depending on oxygen concentration, called extent of resection and minimize postoperative
blood oxygenation level-dependent (BOLD) neurologic deficits in patients with high-grade
imaging[30]. With conventional fMRI, the gliomas, there is much less utility and research
patient will perform tasks during a preoperative into this modality for patients with brain metas-
fMRI in order to map the corresponding cortex. tases [37]. Because primary brain tumors, par-
These images are then closely referenced to the ticularly high-grade ones, have a propensity to
patient intraoperatively to avoid damaging cere- infiltrate neighboring white matter structures,
bral loci critical to speech as well as motor plan- DTI is often used to assess the extent of invasion
ning and execution. Subsequent stereotactic and guide subsequent resection. On the other
navigation can then be superimposed on these hand, intracranial metastases are better circum-
images with real-time guidance for optimized scribed and less likely to infiltrate adjacent neu-
procedural execution, based on these demarcated ral tissue, making DTI less valuable in this
cortical and subcortical boundaries [31]. Though patient population. Nevertheless, DTI remains a
fMRI has advantages in being noninvasive, it is tool at the neurosurgeon’s disposal for delineat-
not as accurate as other techniques and thus can- ing nearby white matter structures at risk of
not be solely relied upon for mapping cortical damage during operative resection.
structures. Specifically, fMRI has a reported sen-
sitivity and specificity of 61.7% and 93.7%,
respectively, for motor mapping [32]. Although ranscranial Magnetic Stimulation
T
some studies have reported predictive value for (TMS)
language “lateralization” with fMRI, others have
found significant discordance between fMRI and Unlike fMRI and DTI, transcranial magnetic
the gold standard Direct Cortical Stimulation stimulation (TMS) is a method for cortical and
(DCS) for cortical speech area mapping with subcortical mapping that does not rely on
sensitivities <60% [33, 34]. imaging data. Instead, TMS provides mapping
a b
Fig. 30.3 (a)Sagittal T1-weighted postcontrast MRI tance between the tract and the resection cavity is reduced.
demonstrating the proximity of a metastatic lesion to the The tract demonstrates continuity
descending corticospinal tract. (b) Post-resection, the dis-
information to neuronavigational models by begins the operation through the completion of

correlating magnetic stimulation of cortical resection [40]. While language mapping
foci with motor activation. Thus, cortical map- requires the patient to be awake, traditionally
ping information is based on the response to through the asleep–awake–asleep methodol-
TMS at various extracranial loci. Though TMS ogy, motor mapping can be performed either in
has existed for decades, it was not until it was the awake or sleep setting. If asleep, phase
combined with neuronavigation software suites reversal between motor and sensory cortex is
(nTMS) that studies began to investigate its first assessed, followed by direct cortical stim-
efficacy. Subsequent articles have reported sig- ulation of the brain to identify the motor gyrus.
nificant accuracy in mapping the motor cortex Continuous somatosensory and motor-evoked
in patients with primary brain tumors when potentials can be run throughout the tumor
compared to direct cortical stimulation (DCS) resection to be certain of intact cortical-sub-
of these regions intraoperatively, thus making cortical connectivity and function. If the
it a noninvasive and precise cortical mapping patient is awake, a similar mapping procedure
option for surgery on the eloquent cortex [38]. can be performed, with the added advantage of
Later studies in patients with brain metastases assessing active neurologic function. Brain
have echoed these results expanding the poten- mapping can thus serve to detect motor, sen-
tial role of this mapping technique [39]. sory, and language deficits in early stages, pre-
venting more significant neurologic morbidity.
Since the emergence of “awake craniotomies”
Electrocorticography for tumor resection, a number of articles have
reported success in minimizing postoperative
Brain mapping during surgery under awake or neurologic deficits [40]. For the resection of
asleep conditions offers the benefit of direct brain metastases specifically, a 2018 system-
and immediate feedback on critical neurologic atic review of published data on the subject
parameters such as language or motor func- concluded that this technique can optimize out-
tion. This continuous observation allows the comes for patients with lesions in eloquent cor-
surgeon to monitor neurologic function as she tical regions [41].
Method of Resection tively reviewing the outcome data of 1033

patients who received surgical resection of single
Surgical resection is a mainstay of treatment for brain metastases. However, they concluded that
selected patients with brain metastases. A num- an en bloc approach was as safe as piecemeal
ber of studies have examined the efficacy of vari- resection with similar complication rates. This
ous resection techniques and their respective data further solidifies the preference for the en
associated complications. Current evidence sug- bloc technique in carefully selected patients.
gests that 46% of resected, nonirradiated lesions Leptomeningeal disease (LMD) following the
eventually recur, highlighting the significant resection of cerebral metastases has become the
impact of the method of resection on the local subject of concern due to its aggressive nature.
recurrence rate [42]. Of these, en bloc and piece- This devastating complication of intracranial
meal resection represent the two categorized malignancy results from contiguous spread of
techniques for operative therapy of these lesions. microscopic tumor components or seeding of the
Local disease control and minimization of CSF. LMD has been associated with brain tumor
recurrence, coupled with optimized postoperative progression and more recently by the technique
outcomes, are the goals of resection in these employed during tumor resection. In 2008, Suki
patients. Thus, en bloc resection is the preferred et al. reported an increased risk of leptomenin-
approach for resection as it reduces the risk of geal disease in a cohort of 260 patients who
residual disease and local recurrence. Conversely, underwent piecemeal resection of cerebral metas-
the piecemeal approach does not afford the same tases in the posterior fossa [44]. Ahn and col-
level of control of the tumor mass, though it is leagues reported similar findings—leptomeningeal
commonly utilized for lesions in otherwise inac- seeding was significantly more likely to result
cessible locations. Additionally, piecemeal resec- from a piecemeal approach, particularly for
tion may be necessary depending on tumor lesions adjacent to CSF pathways [45].
characteristics unique to the patient. If the tumor In response to high rates of tumor recurrence
is friable or infiltrating eloquent regions, the in this patient population, Yoo et al. described a
required piecemeal resection may be unavoidable. novel resection technique to supplement conven-
Regardless of approach, a high-quality postopera- tional en bloc and piecemeal approaches [46].
tive MRI is required to assess for residual disease This technique for microscopic total resection
and to plan potential radiation therapy. involved utilization of an ultrasonic aspirator for
Many studies have investigated the extent to removal of fine infiltrating tumor cells along the
which resection techniques afford local disease white matter margin following gross resection.
control and optimal patient outcomes in the post- When comparing this technique to gross total
operative period. Of these, a 2010 study by Patel resections, the novel technique was associated
et al. found that the type of resection and tumor with significantly fewer local recurrences sug-
volume were the two primary variables predictive gesting the need for additional therapeutic meth-
of local tumor recurrence in a cohort of 570 ods for ensuring local control.
patients with a solitary brain metastasis [3]. They
reported specifically that piecemeal resection
was associated with significantly higher local Postoperative Management
recurrence rates compared to an en bloc tech-
nique, though larger tumors were found to have Good surgical candidates will often leave the hos-
higher rates of local recurrence regardless of the pital within a few days of surgery to either home or
type of resection. Additional evidence to support acute rehabilitation. One of the benefits of surgery
en bloc resection comes from a later 2015 study is the ability to rapidly wean steroid medications,
by Patel and colleagues at MD Anderson Cancer which are often given to minimize symptoms
Center [43]. The authors originally wanted to related to vasogenic edema. Once the tumor has
investigate the risk of potential complications been removed, steroids can be rapidly tapered,
associated with en bloc resection by retrospec- minimizing the complications of prolonged s teroid
therapy. Patients should be quickly encouraged to intracranial tumor burden or contraindications to

ambulate, given their propensity for deep venous surgery, it does result in superior brain control
thrombosis following surgery. globally at the expense of cognitive deterioration
In the context of brain metastases, progression of [48]. However, it does not clearly improve sur-
the primary malignancy often contributes to overall vival when added to surgery or SRS.
survival. However, it is important to devise a strat- SRS is another radiotherapy technique often
egy for both local and distant brain control both to used as an adjunctive therapy to surgical resection
aid overall survival and improve the quality of life that works by precisely intersecting multiple radi-
from neurologic complications. This often includes ation beams at a 3D target. The precision of SRS
stereotactic radiosurgery, targeted medical therapy has made it particularly useful for treating small,
(i.e., EGFR inhibitors for EGFR mutant non-small- well-demarcated intracranial lesions in otherwise
cell lung cancer), and less commonly whole-brain inaccessible loci. Because the rate of local recur-
radiation therapy. As such, treating clinicians must rence following resection of brain metastases
be creative in the postoperative setting in devising approaches 50%, SRS has been studied more
an individualized treatment plan that addresses recently as a postoperative adjuvant with positive
plans for both local and distant brain control. results. A 2014 phase II trial found high rates of
As previously discussed, WBRT was histori- local control for small (<3 cm), deep-seated meta-
cally considered standard of care for patients static lesions, with higher rates of local failure for
with a limited number of cerebral metastases large, superficial tumors [49]. The results of a
until mounting evidence from randomized con- recent phase III trial by Brown et al. reported sim-
trolled trials (RCT) confirmed limited survival ilar findings and concluded SRS should be the
benefit and demonstrated a contribution to cogni- standard of care over WBRT in this patient popu-
tive decline [25–27]. A 1998 RCT reported lation due to less frequently associated decline in
improved rates of local and distant control in cognitive function [27]. Moreover, Mahajan et al.
patients receiving resection and WBRT com- reported lower risk of recurrence with SRS to the
pared to resection alone [47]. Though WBRT resection cavity and concluded that adjuvant SRS
remains an option for patients who are not candi- is an efficacious alternative to WBRT following
dates for resection or radiosurgery, due to high resection of brain metastases (Fig. 30.4) [50].
a b
Fig. 30.4 (a) Axial T1-weighted postcontrast MRI demonstrating a large right cerebellar metastasis. (b) Post-resection
stereotactic radiosurgery plan to treat the resection cavity
Conclusion References
The treatment of patients presenting with 1. Fox BD, Cheung VJ, Patel AJ, Suki D, Rao
G. Epidemiology of metastatic brain tumors.
brain metastases has evolved over the past few Neurosurg Clin N Am. 2011;22(1):1–6, v.
years. WBRT, once a staple therapeutic tool 2. Cagney DN, Martin AM, Catalano PJ, Redig AJ,
for the treatment of multiple metastatic lesions Lin NU, Lee EQ, et al. Incidence and prognosis of
by decreasing the overall tumor burden, has patients with brain metastases at diagnosis of sys-
temic malignancy: a population-based study. Neuro
been shown in a number of RCTs to have a Oncol. 2017;19(11):1511–21.
limited impact on patient survival with 3. Patel AJ, Suki D, Hatiboglu MA, Abouassi H, Shi
increased frequency of associated decline in W, Wildrick DM, et al. Factors influencing the risk
cognitive function [27]. The advent of preop- of local recurrence after resection of a single brain
metastasis. J Neurosurg. 2010;113(2):181–9.
erative diagnostic tools has made surgery safer 4. Shaw AT, Gandhi L, Gadgeel S, Riely GJ, Cetnar J,
and aided in preoperative planning, allowing West H, et al. Alectinib in ALK-positive, crizotinib-
surgeons to anticipate functional cortical and resistant, non-small-cell lung cancer: a single-
subcortical regions in the brain. The use of group, multicentre, phase 2 trial. Lancet Oncol.
2016;17(2):234–42.
awake craniotomies in conjunction with direct 5. Lin NU, Carey LA, Liu MC, Younger J, Come
cortical stimulation is perhaps considered the SE, Ewend M, et al. Phase II trial of lapatinib
gold standard for surgical resection of metas- for brain metastases in patients with human epi-
tases, especially when they appear to be in elo- dermal growth factor receptor 2-positive breast
cancer. J Clin Oncol Off J Am Soc Clin Oncol.
quent brain regions. 2008;26(12):1993–9.
Even with combination therapy, resected brain 6. Tawbi HA, Forsyth PA, Algazi A, Hamid O, Hodi FS,
metastases show high rates of recurrence, partic- Moschos SJ, et al. Combined nivolumab and ipilim-
ularly for large tumors resected in piecemeal umab in melanoma metastatic to the brain. N Engl J
Med. 2018;379(8):722–30.
fashion. Current evidence demonstrates an en 7. Sperduto PW, Kased N, Roberge D, Xu Z, Shanley R,
bloc approach to be safe with lower rates of local Luo X, et al. Summary report on the graded prognostic
tumor recurrence and associated leptomeningeal assessment: an accurate and facile d iagnosis-specific
disease. Additionally, minimally invasive proce- tool to estimate survival for patients with brain metas-
tases. J Clin Oncol. 2012;30(4):419–25.
dures such as LITT and radiosurgery have further 8. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ,
expanded the repertoire of interventions available Maruyama Y, Kryscio RJ, et al. A randomized trial of
to neurosurgeons, particularly in the recurrent surgery in the treatment of single metastases to the
post-irradiated setting. These approaches, when brain. N Engl J Med. 1990;322(8):494–500.
9. Vecht CJ, Haaxma-Reiche H, Noordijk EM,
applicable, have allowed for both faster recovery Padberg GW, Voormolen JH, Hoekstra FH, et al.
times and the treatment of previously inaccessi- Treatment of single brain metastasis: radiotherapy
ble lesions. alone or combined with neurosurgery? Ann Neurol.
Surgical treatment of brain metastases contin- 1993;33(6):583–90.
10. Mintz AH, Kestle J, Rathbone MP, Gaspar L,

ues to evolve. While there will always be a role Hugenholtz H, Fisher B, et al. A randomized trial
for the resection of single metastases or symp- to assess the efficacy of surgery in addition to radio-
tomatic dominant metastases to improve symp- therapy in patients with a single cerebral metastasis.
toms and local control, surgery may also be Cancer. 1996;78(7):1470–6.
11. Wronski M, Arbit E, McCormick B. Surgical treat-
indicated to acquire tissue and updated molecular ment of 70 patients with brain metastases from breast
profiling in the future. Brain metastases have carcinoma. Cancer. 1997;80(9):1746–54.
been shown to be similar yet genomically distinct 12. Iwadate Y, Namba H, Yamaura A. Significance of
to their primary tumor. Targeted intervention in surgical resection for the treatment of multiple brain
metastases. Anticancer Res. 2000;20(1b):573–7.
combination with systemic therapy may become 13. Bindal RK, Sawaya R, Leavens ME, Lee JJ. Surgical
the mainstay of treatment in patients with multi- treatment of multiple brain metastases. J Neurosurg.
ple brain metastases. 1993;79(2):210–6.
14. Arbit E, Wronski M, Burt M, Galicich JH. The treat- 27. Brown PD, Ballman KV, Cerhan JH, Anderson SK,
ment of patients with recurrent brain metastases. A Carrero XW, Whitton AC, et al. Postoperative ste-
retrospective analysis of 109 patients with nonsmall reotactic radiosurgery compared with whole brain
cell lung cancer. Cancer. 1995;76(5):765–73. radiotherapy for resected metastatic brain dis-
15. Al-Zabin M, Ullrich WO, Brawanski A, Proescholdt ease (NCCTG N107C/CEC.3): a multicentre, ran-
MA. Recurrent brain metastases from lung can- domised, controlled, phase 3 trial. Lancet Oncol.
cer: the impact of reoperation. Acta Neurochir. 2017;18(8):1049–60.
2010;152(11):1887–92. 28. Rahmathulla G, Recinos PF, Kamian K, Mohammadi
16. Sarang A, Dinsmore J. Anaesthesia for awake craniot- AM, Ahluwalia MS, Barnett GH. MRI-guided laser
omy—evolution of a technique that facilitates awake interstitial thermal therapy in neuro-oncology: a
neurological testing. Br J Anaesth. 2003;90(2):161–5. review of its current clinical applications. Oncology.
17. Brastianos PK, Carter SL, Santagata S, Cahill DP, 2014;87(2):67–82.
et al. Genomic characterization of brain metastases 29. Sharma M, Balasubramanian S, Silva D, Barnett GH,
reveals branched evolution and potential therapeutic Mohammadi AM. Laser interstitial thermal therapy
targets. Cancer Discov. 2015 Nov;5(11):1164–77. in the management of brain metastasis and radiation
18. Clouston PD, DeAngelis LM, Posner JB. The spec- necrosis after radiosurgery: an overview. Expert Rev
trum of neurological disease in patients with systemic Neurother. 2016;16(2):223–32.
cancer. Ann Neurol. 1992;31(3):268–73. 30. Forster BB, MacKay AL, Whittall KP, Kiehl KA,
19. Sawaya R, Hammoud M, Schoppa D, Hess KR, Wu Smith AM, Hare RD, et al. Functional magnetic
SZ, Shi WM, et al. Neurosurgical outcomes in a mod- resonance imaging: the basics of blood-oxygen-level
ern series of 400 craniotomies for treatment of paren- dependent (BOLD) imaging. Can Assoc Radiol J.
chymal tumors. Neurosurgery. 1998;42(5):1044–55; 1998;49(5):320–9.
discussion 55–6. 31. Yamaguchi F, Takahashi H, Teramoto A. Navigation-
20. Kubicek GJ, Turtz A, Xue J, Patel A, Richards G, assisted subcortical mapping: intraoperative motor
LaCouture T, et al. Stereotactic radiosurgery for poor tract detection by bipolar needle electrode in combi-
performance status patients. Int J Radiat Oncol Biol nation with neuronavigation system. J Neurooncol.
Phys. 2016;95(3):956–9. 2009;93(1):121–5.
21. McGirt MJ, Woodworth GF, Brooke BS, Coon AL, 32. Qiu TM, Gong FY, Gong X, Wu JS, Lin CP, Biswal
Jain S, Buck D, et al. Hyperglycemia independently BB, et al. Real-time motor cortex mapping for
increases the risk of perioperative stroke, myocardial the safe resection of glioma: an intraoperative
infarction, and death after carotid endarterectomy. resting-state fMRI study. AJNR Am J Neuroradiol.
Neurosurgery. 2006;58(6):1066–73; discussion −73. 2017;38(11):2146–52.
22. McGirt MJ, Woodworth GF, Coon AL, Frazier JM, 33. Dym RJ, Burns J, Freeman K, Lipton ML. Is func-
Amundson E, Garonzik I, et al. Independent predic- tional MR imaging assessment of hemispheric
tors of morbidity after image-guided stereotactic brain language dominance as good as the Wada test?: a
biopsy: a risk assessment of 270 cases. J Neurosurg. meta-analysis. Radiology. 2011;261(2):446–55.
2005;102(5):897–901. 34. Roux FE, Boulanouar K, Lotterie JA, Mejdoubi

23. Hatiboglu MA, Wildrick DM, Sawaya R. The role of M, LeSage JP, Berry I. Language functional mag-
surgical resection in patients with brain metastases. netic resonance imaging in preoperative assessment
Ecancermedicalscience. 2013;7:308. of language areas: correlation with direct cortical
24. Muacevic A, Wowra B, Siefert A, Tonn JC, Steiger stimulation. Neurosurgery. 2003;52(6):1335–45.. dis-
HJ, Kreth FW. Microsurgery plus whole brain irra- cussion 45-7
diation versus Gamma Knife surgery alone for treat- 35. Alexander AL, Lee JE, Lazar M, Field AS. Diffusion
ment of single metastases to the brain: a randomized tensor imaging of the brain. Neurotherapeutics.
controlled multicentre phase III trial. J Neurooncol. 2007;4(3):316–29.
2008;87(3):299–307. 36. Potgieser ARE, Wagemakers M, van Hulzen ALJ,
25. Soon YY, Tham IW, Lim KH, Koh WY, Lu JJ. Sur- de Jong BM, Hoving EW, Groen RJM. The role of
gery or radiosurgery plus whole brain radio- diffusion tensor imaging in brain tumor surgery: a
therapy versus surgery or radiosurgery alone for review of the literature. Clin Neurol Neurosurg.
brain metastases. Cochrane Database Syst Rev. 2014;124:51–8.
2014(3):CD009454. 37. Wu JS, Zhou LF, Tang WJ, Mao Y, Hu J, Song YY,
26. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF, et al. Clinical evaluation and follow-up outcome of
Kornguth DG, et al. Neurocognition in patients with diffusion tensor imaging-based functional neuronavi-
brain metastases treated with radiosurgery or radio- gation: a prospective, controlled study in patients with
surgery plus whole-brain irradiation: a randomised gliomas involving pyramidal tracts. Neurosurgery.
controlled trial. Lancet Oncol. 2009;10(11):1037–44. 2007;61(5):935–48; discussion 48–9.
38. Takahashi S, Vajkoczy P, Picht T. Navigated transcra- for solid tumor metastasis to the posterior fossa. J
nial magnetic stimulation for mapping the motor cor- Neurosurg. 2008;108(2):248–57.
tex in patients with rolandic brain tumors. Neurosurg 45. Jun Hyong Ahn, Sang Hyun Lee, Sohee Kim,

Focus. 2013;34(4):E3. Jungnam Joo, Heon Yoo, Seung Hoon Lee, et al. Risk
39. Krieg SM, Picht T, Sollmann N, Bährend I, Ringel for leptomeningeal seeding after resection for brain
F, Nagarajan SS, et al. Resection of motor eloquent metastases: implication of tumor location with mode
metastases aided by preoperative nTMS-based motor of resection. J Neurosurg. 2012;116(5):984–93.
maps—comparison of two observational cohorts. 46. Yoo H, Kim YZ, Nam BH, Shin SH, Yang HS, Lee
Front Oncol. 2016;6:261. JS, et al. Reduced local recurrence of a single brain
40. Saito T, Muragaki Y, Maruyama T, Tamura M, Nitta metastasis through microscopic total resection. J
M, Okada Y. Intraoperative functional mapping and Neurosurg. 2009;110(4):730–6.
monitoring during glioma surgery. Neurol Med Chir. 47.
Patchell RA, Tibbs PA, Regine WF, et al.
2015;55(1):1–13. Postoperative radiotherapy in the treatment of single
41. Chua TH, Qi See AA, Ang BT, Kam King
metastases to the brain: a randomized trial. JAMA.
NK. Awake craniotomy for resection of brain 1998;280(17):1485–9.
metastases: a systematic review. World Neurosurg. 48. McPherson CM, Suki D, Feiz-Erfan I, Mahajan A,
2018;120:e1128. Chang E, Sawaya R, et al. Adjuvant whole-brain radi-
42. Patel TR, Knisely JP, Chiang VL. Management of ation therapy after surgical resection of single brain
brain metastases: surgery, radiation, or both? Hematol metastases. Neuro Oncol. 2010;12(7):711–9.
Oncol Clin North Am. 2012;26(4):933–47. 49. Brennan C, Yang TJ, Hilden P, Zhang Z, Chan K,
43. Patel AJ, Suki D, Hatiboglu MA, Rao VY, Fox
Yamada Y, et al. A phase 2 trial of stereotactic radio-
BD, Sawaya R. Impact of surgical methodology on surgery boost after surgical resection for brain metas-
the complication rate and functional outcome of tases. Int J Radiat Oncol Biol Phys. 2014;88(1):130–6.
patients with a single brain metastasis. J Neurosurg. 50. Mahajan A, Ahmed S, McAleer MF, Weinberg JS, Li
2015;122(5):1132–43. J, Brown P, et al. Post-operative stereotactic radiosur-
44. Suki D, Abouassi H, Patel AJ, Sawaya R, Weinberg gery versus observation for completely resected brain
JS, Groves MD. Comparative risk of leptomeningeal metastases: a single-centre, randomised, controlled,
disease after resection or stereotactic radiosurgery phase 3 trial. Lancet Oncol. 2017;18(8):1040–8.
Intraoperative Brachytherapy
for Resected Brain Metastases
31
A. Gabriella Wernicke, Sean S. Mahase,
and Theodore H. Schwartz
Introduction metastases, with breast cancers, melanoma, and

colon cancers, respectively, accounting for
Brain metastases, the most common intracranial approximately 20%, 10%, and 5% of all brain
neoplasms in adults [1], develop in approxi- metastasis primaries [2]. Epidemiologically,
mately 30% of all cancer patients and are the these primaries are also among the most common
cause of death in up to 50% of these individuals malignancies in the United States. Conversely,
[2]. They are commonly located at the gray– small-cell lung cancers, melanoma, germ cell
white matter interface where the blood vessel tumors, and choriocarcinomas demonstrate pro-
caliber decreases, and their dissemination corre- portionally high neurotropism rates.
sponds with blood flow: 80% of patients develop Symptomatic management options include
multiple intracranial metastases, with 80% occur- corticosteroids and supportive care [5].
ring in the cerebral hemispheres, 15% in the cer- Chemotherapeutic agents historically demon-
ebellum, and 5% in the brainstem [3]. Incidence strated little efficacy in treating brain metastases
rates are expected to rise with the emergence of owing to the inability to enter the central nervous
increasingly effective systemic agents that, while system. However, the utility of targeted agents
conferring improved systemic control translating and immunotherapy in the context of multidisci-
to increased survival, possess limited ability to plinary treatment strategies is currently an area of
bypass the blood–brain barrier thus making the active investigation. Commonly utilized treat-
central nervous system a sanctuary site [4]. As ment options include whole-brain radiation ther-
only 10% of patients become symptomatic from apy (WBRT), stereotactic radiosurgery (SRS),
brain metastases, incidence rates are also increas- and surgical resection.
ing with improved surveillance [2]. Primary lung WBRT was the initial standard therapy and
cancers account for over 50% of intracranial continues to play a pivotal role in treating brain
metastases, particularly in the setting of multiple
lesions, and in the presence of recurrent metasta-
A. G. Wernicke · S. S. Mahase ses or leptomeningeal disease. The first Radiation
Department of Radiation Oncology, NewYork-
Presbyterian/Weill Cornell Medicine, Therapy Oncology Group (RTOG) randomized
New York, NY, USA trials established WBRT as an effective modality
T. H. Schwartz (*) for patients with favorable performance status
Department of Neurological Surgery, NewYork- and/or well-controlled primary disease. However,
Presbyterian/Weill Cornell Medicine, these initial studies reported overall survival (OS)
New York, NY, USA rates of only a few months [6]. Patchell et al.
e-mail: schwarh@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_31
442 A. G. Wernicke et al.
improved on these outcomes in a randomized 12, 17]. Given concerns of neurocognitive

trial comparing WBRT to surgical resection fol- decline following WBRT, the paradigm is shift-
lowed by WBRT among patients with a single ing to postoperative SRS [17–19]. The relative
brain metastasis. They demonstrated surgery benefits and disadvantages of giving SRS preop-
followed by WBRT improved OS to 40 weeks, eratively is also under investigation, with a retro-
compared to 15 weeks with WBRT alone [7]. spective multi-institutional study comparing
Patchell and colleagues subsequently random- preoperative SRS to postoperative SRS showed
ized 95 patients who underwent surgical resec- no differences in OS and local recurrence at 2
tion to observation or postoperative WBRT and years [20]. The only prospective trial evaluating
reported no significant difference in OS among preoperative SRS demonstrated an 85.6% 1-year
the cohorts. However, tumor recurrence was LC rate without radionecrosis [21], and one trial
reduced from 46% in the observation group to (NCT02514915) is currently accruing.
10% in the WBRT group, as well as a reduction Another appealing option to improve postop-
in new brain metastases and death due to neuro- erative LC rates and obviate the need for adjuvant
logical causes in the WBRT group, thus estab- radiation and commute for postoperative radia-
lishing postoperative WBRT as the standard of tion treatments entails intracavitary brachyther-
care for brain metastases at that time [8]. apy. This chapter discusses the rationale,
SRS is a minimally invasive option for patients technique, outcomes, evidence, and future direc-
who are not surgical candidates; however, there tions regarding the use of intracavitary brachy-
are a dearth of appropriately powered random- therapy as an adjunct treatment to surgical
ized control trials comparing surgical resection treatment. We will discuss various types of
and SRS alone for brain metastases. RTOG 9508 brachytherapy and radioactive isotopes available
demonstrated that an SRS boost improves local for this procedure, as well as the benefits of the
control (LC) following WBRT with no difference radioisotope Cesium-131 (Cs-131), which offers
in survival [9]. Conversely, several randomized great promise as the radioisotope of choice in the
trials evaluating SRS alone versus SRS with future.
WBRT demonstrated that the latter may improve
local and distant tumor control, but OS rates
remained the same as using SRS alone [10–12]. Considerations in Selecting
Patients undergoing WBRT were also more likely an Adjuvant Radiotherapy Modality
to exhibit neurological or cognitive decline [13,
14]. These findings, alongside studies showing Changes in Resection Cavity Volume
SRS may be an appropriate option for patients
with multiple brain metastases [15], promotes its An early retrospective study of 72 patients
increasing utilization over WBRT. treated with postoperative SRS, in which the
Surgical resection remains the preferred ini- resection cavity was targeted without a margin,
tial treatment modality for patients with good demonstrated that LC was significantly higher
performance status and a limited number intra- among those with less conformal plans [22]. A
cranial lesions who require pathological confir- subsequent study targeting the resection cavity
mation, have a large (greater than 2 cm) with a 2 mm margin improved 1-year local fail-
metastasis, or who are experiencing mass effect ure rates from 16% to 3% without a significant
or neurological symptoms refractory to steroids increase in toxicity (3% with a 2 mm margin ver-
[16]. However, surgical resection without any sus 8% without a margin) [23]. The rationale to
adjuvant intracranial treatment has a 1- to 2-year incorporate an additional margin stemmed from
LC rates of 47–59%; thus adjuvant radiotherapy more conformal plans increasing the risk of mar-
is typically given in an effort to maximize LC [8, ginal miss due to difficulty contouring the post-
31 Intraoperative Brachytherapy for Resected Brain Metastases 443
operative cavity. Numerous studies have recurrence. Preoperative SRS may provide one
explored resection cavity volume dynamics with avenue to avoid some of these potential compli-
respect to time from resection on subsequent cations. However, some patients may present too
SRS planning. In 68 metastases treated with sur- urgently to allow time to safely perform MRI
gical resection and postoperative SRS, Atalar scanning, CT simulation, development of a SRS
et al. reported a median pre-resection tumor vol- plan, and radiotherapy delivery prior to surgery.
ume of 14.5 cm3, and a median resection cavity This option may also not be feasible for patient
volume of 10.1 cm3, corresponding to a 29% needing pathological confirmation or with large
reduction volume. Of note, MRI imaging showed metastases.
shrinkage in 72% of resection cavities, but also
showed increased cavity size in 26% of cases
[24]. In a study comparing MRI scans obtained Time to Radiotherapy
preoperatively, 24 hours following surgery, and 1
week prior to SRS in 43 resected brain metasta- Due to a plethora of considerations discussed
ses, 46.5% of cavities remained stable in size above, including temporal changes in resection
(defined as <2 cm3 change in size), whereas cavity dynamics and technical and practical con-
23.3% shrank by over 2 cm3, and another 30.2% siderations involved in adjuvant SRS planning,
increased in size by over 2 cm3 [25]. Thus, the there have been concerns related to the effect of
resection cavity experiences significant, unpre- adjuvant radiotherapy timing on treatment effi-
dictable changes following surgery, which may cacy. While there is a dearth of data exploring the
impact SRS planning. These changes may lead effect of time to adjuvant SRS on local recur-
to inaccuracies in planning and potentially pro- rence, Seymour and colleagues reviewed patient
mote marginal misses or excessive irradiation of demographics, clinical outcomes, and workflow
adjacent normal brain parenchyma. timing, including time from MRI and CT simula-
tion, insurance authorization, and consultation to
start of SRS for intact brain metastases. They
ractical Considerations in Adjuvant
P reported 6- and 12-month local freedom from
Treatment Planning progression rates of 95% and 75% for metastasis
with an interval of <14 days from MRI to SRS,
The treatment of brain metastases is a multidisci- compared to 56% and 34% for metastases with
plinary endeavor between neurosurgery, radia- MRI 14 days after treatment, suggesting a LC
tion oncology, and medical oncology. Each benefit in expediting treatment [28].
patient’s baseline performance status, recovery
from surgery, potential inpatient complications,
primary site and systemic disease management, Radiobiological Considerations
and social factors impact the time between resec-
tion and postoperative SRS. The interval between resection and delivery of
The SRS planning process is inherently more postoperative SRS may allow tumor repopulation
technically involved than WBRT from a radiation within the cavity. Furthermore, it can be postulated
planning perspective, which may also increase that radiotherapy is more likely to produce a sus-
time to treatment. Accordingly, several postoper- tained effect if there is a smaller residual tumor
ative SRS studies quote a median time from sur- volume to target. As SRS is typically administered
gery to SRS of 4–5 weeks [23, 26, 27]. in a single fraction, or in three fractions given over
Additionally, there may be concern that some consecutive days, it is possible that cells in more
patients may not follow-up for their adjuvant radioresistant phases of the cell cycle may be
radiotherapy, putting them at higher risk for spared. Postoperative SRS may also be delivered
to a relatively hypoxic resection cavity, impairing start systemic agents whose administration is
radiation efficacy. This phenomenon occurs sec- typically avoided concurrently with
ondary to radiation primarily acting through gen- radiotherapy.
erating oxygen-based free radicals, which in turn SRS requires a 2 mm planning target volume
induce single-strand and double-strand DNA expansion, increasing the volume of normal brain
breaks that ultimately cause cell death. Based on receiving 10–12 Gray of radiation, which has
this reasoning, referred to as the oxygen enhance- been correlated with an increased risk of radia-
ment ratio [29], preoperative SRS studies are tion necrosis [31–33]. Conversely, utilizing intra-
advocating for a 20% dose reduction from the operative brachytherapy, neurosurgeons can
standard postoperative doses used in RTOG 9005 decide upon the number of sources required to
[30]. Furthermore, postoperative radiation might adequately cover the resected volume, thus pro-
be associated with tumor cell repopulation which viding a well-defined target encompassing the
may take place during the postsurgical recovery resection cavity and areas of microscopic dis-
and treatment planning phases. ease, while enabling avoidance of deliberately
extending the treatment volume into normal brain
parenchyma. Several brachytherapy studies
Rationale for Perioperative described below report 80–95% LC rates [34–
Brachytherapy 39], which may be attributed to interstitial
brachytherapy possessing a higher conformality
Brachytherapy, entailing the implantation of a index than post-op SRS [40, 41].
radioactive source within the tumor resection Brachytherapy is an alternative salvage option
cavity at the time of surgery, has several appeal- for recurrent metastases as repeat SRS may pro-
ing advantages over WBRT and SRS with regards duce subpar outcomes in this setting [33].
to the plethora of concerns described above and Brachytherapy is also an option for larger resec-
in Tables 31.1 and 31.2. Perioperative brachy- tion cavities, given reports indicating reduced effi-
therapy offers an immediately available radio- cacy and increased risk of radionecrosis with SRS
therapy option that avoids tumor cell repopulation in this setting [8, 33, 42, 43]. Metastases 3 cm or
as radiotherapy treatment begins immediately greater treated with SRS generally require dose
intraoperatively. This treatment option does not reduction to minimize risk of radionecrosis [30],
require extensive preoperative or postoperative resulting in subpar outcomes in this cohort [44–
delays for scanning and treatment planning. 47]. Ebner et al. evaluated 343 patients with 754
Thus, the planning accuracy is not subject to total brain metastases treated with SRS, of which
changes in the resection cavity shape and size 93 had large tumors. The tumor size was 3–3.5,
that may occur in the interval between postopera- 3.5–4, and 4 cm or greater in 29%, 32%, and 39%
tive planning MRI and delivery of of these patients. The LC of large metastases was
SRS. Alternatively, patients who require urgent inferior compared to smaller tumors, with 1-year
surgical resection can still receive immediate LC rates of 68% versus 86%, respectively
adjuvant treatment, which would not be feasible (p < 0.001) [48]. The potential advantage of
with preoperative SRS. Providing adjuvant treat- brachytherapy of SRS in the postoperative man-
ment at the time of surgery removes possible agement of large metastases may be extrapolated
delays in treatment planning due to postoperative from the glioblastoma literature, in which resec-
course and recovery, concerns of patient compli- tion cavities and irradiation volumes are consider-
ance, and avoids the need to undergo the entire ably larger. In a retrospective analysis comparing
planning process and workflow entailed SRS and interstitial brachytherapy for recurrent
within postoperative SRS. This option is particu- gliomas, Shrieve and colleagues reported similar
larly appealing for patients who may have diffi- median survivals of 10.2 months and 11.5 months
culty with transportation, and who may need to among each respective cohort. However, the
Table 31.1 Neoadjuvant and adjuvant radiotherapy options for resected brain metastases
Modality Margins Advantages Disadvantages
Post-op WBRT 1.5–2.0 cm “flash” Technically easy to plan Higher risk of neurocognitive
anteriorly, superiorly, Lowers risk of distant intracranial sequela [19]
and posteriorly, with metastases
inferior field Treatment of choice for
extended to C1 or c2 leptomeningeal disease [16]
Low risk of radionecrosis
Pathology available
Post-op SRS 2 mm margin most Feasible for multiple lesions [15] May be difficult to define resection
commonly used [19] Limited neurocognitive effects [19] cavity [22–25]
Pathology available Requires involved planning, pre- and
postoperative scans and outpatient
visits [28]
Delays between surgery, scanning,
and radiation delivery [24, 26–28]
Possible hypoxic tumor
bed irradiation
Risk of radionecrosis
Not advantageous for irregularly
shaped cavities and large lesions [66]
Pre-op SRS No margin [21] Easy to define target volume Limited adjuvant treatment options if
Delivered to well-oxygenated tumor subtotal resection follows SRS
environment Requires involved planning,
Limited neurocognitive effects preoperative scans
similar to post-op SRS No pathology available at time of
radiotherapy
Temporary No margin Irradiation begins immediately from Requires second surgery for implant
brachytherapy time of placement removal [22]
Limited neurocognitive effects Patient is radioactive
Delivered to well-oxygenated tumor Dependent on technical expertise
environment
Can effectively target irregularly-
shaped cavities
Effective for larger lesions
Pathology available
Permanent 5 mm from surface Performed intraoperatively; reduces Patient is radioactive
brachytherapy [63] subsequent patient visits relative to Potential for seed migration
postop SRS Dependent on technical expertise
Irradiation begins immediately from High rates of radionecrosis with
time of placement I-125 [37]
Limited neurocognitive effects [67]
Delivered to well-oxygenated tumor
environment
Can effectively target irregularly
shaped cavities [66]
Effective for larger lesions [39, 43]
Pathology available
brachytherapy cohort had higher radiation necro- and a significantly larger average tumor volume
sis rates and 2-year reoperation rates (65% versus (29 cc versus 10.1 cc) than the SRS group, sug-
48%) than those receiving SRS. While there were gesting a preference towards brachytherapy with
no difference in age and performance status among larger cavities, as well as a question regarding
the cohorts, the brachytherapy group had a longer what the radionecrosis rates would be if SRS were
median follow-up (43 months versus 17.5 months) used in lesions of equal comparable size [49].
Table 31.2 Outcomes and radionecrosis rates by radiation modality in prospective trials
446
Distant Time between surgery

Modality Study Local control control Overall survival Radionecrosis Treatment dose and volume and radiation (days)
Post-op WBRT Brown 2017 87.1% (1 year) 89.2% 11.6 months None 30 Gy in 10 fractions Not reported
[19] (1 year) (median) 37.5 Gy in 15 fractions
Patchell 1990 80% 80% 10 months Not assessed 36 Gy in 12 fractions Within 14 days of
[7] (median) surgery
Patchell 1998 90% 86% 12 months Not assessed 50.4 Gy in 28 fractions Within 28 days of
[8] (median) surgery
Post-op SRS Brennan 2014 78% (1 year) 56% 14.7 months 7 cavities Cavity with 2 mm margin Median 31 days
[26] (1 year) (17.5%) ≤ 2.0 cm: 22 Gy
2.1–3.0 cm: 18 Gy
3.1–4.0 cm: 15 Gy
Mahajan 2017 72% (1 year) 42% 17 months None Cavity with 1 mm margin Within 30 days of
[17] (1 year) (median) ≤ 10 cc: 16 Gy surgery
10.1–15 cc: 14Gy
> 15 cc: 12 Gy
Soltys 2008 79% (1 year) 47% 15.1 months 3 patients Cavity with 2 mm margin Not reported
[22] (1 year) (median) (4.1%) Dosing per RTOG 90–05 (median
marginal dose 18 Gy)
Brown 2017 61.8% (1 year) 64.7% 12.2 months 1.1% (1 Cavity with 2 mm margin Not reported
[19] (1 year) patient) < 4.2 cc: 20 Gy
4.2–7.9 cc: 18 Gy
8.0–14.3 cc: 17 Gy
14.4–19.9 cc: 15 Gy
20.0–29.9 cc: 14 Gy
≥ 30.0 cc: 12 Gy
Pre-op SRS Asher 2014 85.6% (1 year) 67.2% 60% at 1 year None Gross tumor with no margin Median 1 day
[21] (overall) Median 14Gy to 80% isodose line
I-125 Ruge 2011 [75] 93.3% (1 year) 54.5% 14.8 months 10% (3 Prescription dose 50 Gy at surface of Intraoperative
brachytherapy (1 year) (median) patients) target volume over 42 days
Cs-131 Wernicke 2014 100% (1 year); defined 48.4% 9.9 months None Prescription dose 80Gy at 5 mm depth Intraoperative
brachytherapy [63] at recurrence within (1 year) (median) from resection cavity surface
5 mm of cavity
Wernicke 2017 100% (1 year); defined 52% 15.1 months None Prescription dose 80Gy at 5 mm depth Intraoperative
[69] at recurrence within (1 year) (median) from resection cavity surface
5 mm of cavity
RTOG Radiation Therapy Oncology Group
A. G. Wernicke et al.
Utilizing brachytherapy may be radiobio- operative SRS cohorts (61% versus 49%;
logically advantageous compared with SRS. As p = 0.137) [54]. Thus, intraoperative brachy-
radioactive implants are placed at the time of therapy may be a comparably effective but
surgery, irradiation of the resection bed begins more cost- effective, radiotherapy modality
at day 0; if radiation-induced cell death exceeds than SRS for patients requiring resection for
proliferation, potential repopulation of residual brain metastases.
tumor cells is minimized. Radiotherapy’s thera-
peutic index relies on the ability of normal cells
to repair sublethal damage more effectively I-125 Brachytherapy
than adjacent tumor cells. Brachytherapy
exposes the resection cavity to constant irradia- While several radioisotopes options exist for
tion, promoting accumulation of cellular dam- brachytherapy (Table 31.3), including palladium-
age to residual tumor cells, and increasing the 103 (17 day half-life) and gold-198 (2.7 day half-
probability of mitotic catastrophe, whereby life), historically, iodine-125 (I-125) was the
cells with an abundance of DNA damage most common radioactive source utilized in CNS
undergo apoptosis due to failure to divide. tumors and is administered using either tempo-
Continuous irradiation also enables eventual rarily placed interstitial catheters or implants, or
targeting of tumor cells undergoing redistribu- as permanent implants. I-125 possesses a half-
tion from radioresistant phases to more radio- life of 60.2 days. Temporary implants are reus-
sensitive phases (G2/M) [50]. able sources with an activity of 10–20 mCi per
As the incidence of brain metastases con- source, photon energies of 27–35 keV, and a dose
tinues to rise, an increasingly practical consid- rate of 40–60 cGy per hour [55]. Implanted I-125
eration entails the hospital costs involved in sources possess a half-value layer of 0.025 mm of
treatment (Table 31.1). Comparative analyses lead and are typically housed in 4 mm long by
suggest SRS is more cost-effective than resec- 0.8 mm diameter titanium capsules. Utilizing the
tion alone [51, 52], as well as WBRT [53]. higher-dose rate of temporary I-125 implants,
Wernicke et al. retrospectively reviewed treat- continuous irradiation can be delivered over the
ment records of 24 patients undergoing sur- next ~100 hours following resection and removed
gery and intraoperative Cs-131 brachytherapy on postoperative day 4 [56]. Temporary implant
and 25 patients undergoing surgery and post- dose distribution can be assessed both pre- and
operative SRS with the purpose of evaluating postoperatively. While temporary implants can
the cost-effectiveness of each radiotherapy be reused and enable irradiation at higher dose
modality. They reported a direct hospital cost rates, they are a less attractive option as they must
for surgery and intraoperative Cs-131 brachy- be removed, subjecting patients to a second sur-
therapy of $19,271, whereas surgery and post- gery, and the ensuing preoperative planning,
operative SRS cost $44,219. Additionally, additional operative costs, and potential postop-
there was no significant difference in 1-year erative complications. Thus, permanent, low-
survival rates among brachytherapy and post- activity implants are becoming more favored.
Table 31.3 Commonly used isotopes in brachytherapy

Average photon Exposure rate constant
Isotope Half-life energy (MeV) Half-value in lead (mm) (R-cm2/mCi-h)
Iodine-125 59.4 days [76] 0.028 [76] 0.025 [76] 1.46 [76]
Cesium-131 9.7 days [77] 0.029 [77] 0.0262 [78] 0.679 [78]
Cobalt-60 5.3 years [76] 1.17, 1.33 [76] 11.0 [76] 13.07 [76]
Iridium-192 73.8 days [76] 0.38 [76] 2.5 [76] 4.69 [76]
Palladium-103 17.0 days [76] 0.021 [76] 0.008 [76] 1.48 [76]
Permanent I-125 implants are nonreusable resections and 13 had prior WBRT) treated with
sources with an activity of 0.725 mCi per source temporary high-activity I-125 sources, Prados
and a dose rate of 11 cGy per hour [57]. The et al. reported a median survival of 20 months,
insertion of permanent radioactive sources at the with stable responses in 8 patients and radione-
time of resection exploits the precipitous but pre- crosis in 2 patients [59]. Ostertag and Kreth eval-
dictable drop in dose as a function of distance uated the efficacy of interstitial high-activity
from the source, referred to as the inverse square I-125 in 93 patients with brain metastases ≤4 cm
law. Predicting irradiation exposure of the imme- in diameter. Patients were either treated with
diately adjacent 1 cm of normal brain paren- interstitial brachytherapy to 60 Gy to the tumor
chyma can be difficult due to production of periphery plus external beam radiotherapy to
secondary photons. However, optimizing the 40 Gy, or interstitial brachytherapy alone to
1 cm distance traversed by I-125’s low-energy 60 Gy. All patients with tumor recurrence or prior
photons enables delivery of highly conformal irradiation were treated with the latter regimen.
dose distributions to maximize targeting of resid- Median survival was 17 months in the combina-
ual tumor with relative sparing of surrounding tion radiotherapy group, 15 months among those
normal brain parenchyma [58]. I-125 sources are with newly diagnosed metastases treated only
placed along the walls of the resection cavity in interstitially, and 6 months among those with
the form of either free sources or embedded in an recurrent metastases. Interstitial brachytherapy
absorbable suture and held in place with a liquid plus external beam radiotherapy did not prove to
adhesive. Unlike SRS techniques, the surgical be superior to interstitial brachytherapy alone,
staff is exposed to radiation during implantation and no patients developed symptomatic radione-
and required to don lead gloves, vests, and thy- crosis [38].
roid shields. Another potential shortcoming is
that seed positioning is confirmed 1–2 days fol-
lowing implantation, with the potential risks of ermanent Low-Activity I-125
P
inadequate dose coverage. Additionally, it is pos- Brachytherapy
sible for the seed to migrate over time or inade-
quately target the cavity if it changes size or Schulder et al. reviewed a small series of 13
shape. patients with large recurrent metastatic brain
tumors following initial WBRT who underwent
resection and permanent low-activity I-125 seed
I -125 Brachytherapy Outcomes implantation. Implant dose ranged from 43 Gy to
in Brain Metastases 132 Gy, with a mean dose of 83 Gy. The entire
cohort had a median survival of 9 months, with
High-Activity I-125 Brachytherapy durable LC achieved in 9 patients, and one case
of radionecrosis [39]. In another study of 40
Bernstein and colleagues used high-activity I-125 patients with metastases deemed too large
seeds to treat 10 patients with brain metastases (>2.5 cm in diameter) for SRS who underwent
that recurred following initial treatment with cra- resection and placement of permanent I-125
niotomy and WBRT. I-125 seeds (20–40 mCi) seeds, Huang et al. reported a median survival of
with a mean dose rate of 67.3 cGy per hour were 11.3 months (12 months among patients with
implanted with 70 Gy prescribed to the tumor. newly diagnosed metastases and 7.3 months in
Implant volumes ranged from 12.1 cc to 99.0 cc those with recurrent metastases). There were 3
(mean: 44.5 cc; median: 36.4 cc). Four patients local failures and 13 distant recurrences, with
were alive 2 years following the procedure with symptomatic necrosis developing at a median of
the caveat of potential selection bias towards 19.5 months in 9 patients (6 with pathological
favorable histology and good performance status confirmation) [37]. Petr et al. evaluated the effi-
[34]. In a study of 14 patients with recurrent met- cacy of surgical resection and permanent low-
astatic brain lesions (4 patients had prior surgical activity I-125 seeds in 72 patients with newly
diagnosed single brain metastases. At a median Thus, with proper technique, permanent I-125
of 16 months, they reported a 93% LC rate, and brachytherapy may be a favorable adjuvant treat-
a 23% distant failure rate, with four patients ment option for selected brain metastases, includ-
developing radionecrosis [60]. A retrospective ing large and recurrent lesions, albeit with high
study of two institutions’ experience treating a risk of radiation necrosis.
single metastasis with gross-total resection fol-
lowed by permanent low-activity I-125 implants
reported a 1-year LC rate of 96%, with two Limitations and Complications
patients developing symptomatic radiation with I-125 Brachytherapy
necrosis requiring intervention [36]. Bogart
et al. treated 15 patients with solitary brain Acute side effects of interstitial brachytherapy
metastases from primary non- small cell lung include seizures, infection, impaired periopera-
cancer with low dose-rate I-125, and reported a tive healing, hemorrhage, and other neurological
14 month median survival with no in-site recur- sequelae, which are more common with high-
rences and two recurrences adjacent to the origi- activity temporary implants. Radionecrosis is
nal metastases. They reported one death from a also a major concern, with reported rates as high
postoperative fungal infection, but no cases of as 29% [37]. The largest criticism of permanent
symptomatic radionecrosis [35]. I-125 brachytherapy is its relatively long half-
Raleigh et al. recently reported on the out- life, which subjects the patient to radiation for a
comes of 95 patients with 105 brain metastases prolonged period, and may potentially expose
who underwent resection followed by place- surgical staff to radiation in the case of a repeat
ment of permanent I-125 implants. I-125 sources surgery. As described above, resection cavities
were placed 6–10 mm apart and secured in place undergo significant dynamic changes following
with fibrin glue. Postoperative stereotactic com- resection, and larger resection cavities, which
puterized tomography scanning for dosimetric were often selected for brachytherapy in the
calculations was generally performed within aforementioned trials, are especially subject to
24 hours of surgery, and the prescription was postsurgical changes in volume and shape [24].
determined based on source activity, calibration Taken in context with I-125’s long half-life, these
date, and implantation date. Forty-seven percent changes may affect the dosimetry of the seeds,
of the lesions were new metastases and 53% potentially decreasing tumor dose or increasing
were recurrent lesions, of which 40% were pre- normal brain tissue exposure. As the dosimetry is
viously treated with SRS, 25% with prior dependent on seed placement and fixation to the
WBRT, and 17% with a prior resection. The cavity, brachytherapy is dependent on the techni-
median metastasis volume was 13.5 cm3 (range: cal expertise of the physician performing the pro-
0.2–76 cm3), a median of 28 sources were used, cedure. Additionally, intracranial brachytherapy
and the median source activity was 0.73 mCi. is not as common as SRS, and training in this
The median brachytherapy dose was 540 Gy at technique is highly variable among academic
3 mm, 263 Gy at 5 mm, and 135 Gy at 10 mm hospitals.
depth into brain tissue (measured outward from
the resection cavity edge), which corresponded
to median treatment volumes of 6.8 cm3, Cesium-131 (Cs-131) Brachytherapy
12.8 cm3, and 33 cm3, respectively. The median
OS was 12 months, with over 22% of patients Since obtaining FDA approval in 2003, Cs-131
surviving beyond 2 years after intervention. The has been utilized as radioactive permanent seed
authors reported a LC rate of 90%. The median implants for treatment of prostate, head and neck,
time to radionecrosis was 1 year, with 15 and lung malignancies. Cs-131 has a half-life of
reported cases, of which 11 underwent prior 9.69 days, a dose rate of 0.342 Gy per hour, and
SRS, and the remaining 4 were newly diagnosed an average energy of 30.4 KeV. Comparative
metastases [61]. studies of radioactive seeds used in prostate
brachytherapy suggested Cs-131 has preferable and fibrin glue. All patients underwent a post-
dose homogeneity, required fewer seeds to pro- implant CT scan to determine dose distribution
vide comparable prostate coverage, and enabled 1–2 days following the procedure. In this series,
superior sparing of the rectum and urethra com- the median resected tumor volume was 10.31 cc
pared to Pd-103 or I-125 [62]. (range: 1.77–87.11 cc), and a median of 12 seeds
(range: 4–35) were implanted, with median
activity per seed of 3.82 mCi (range: 3.31–
s131 Brachytherapy Outcomes
C 4.83 mCi) and total activity of 46.91 mCi (range:
in Brain Metastases 15.31–130.70 mCi). At a median follow-up of
19.3 months, the median survival was 9.9 months,
Wernicke et al. evaluated the safety, feasibility, with a 100% LC rate (defined as no recurrence
and efficacy of permanent intraoperative Cs-131 within 5-mm from the resection cavity). There
brachytherapy following resection in a prospec- was one regional recurrence (>5 mm from the
tive phase I/II study of 24 patients with newly resection cavity) and a distant control rate of
diagnosed brain metastases. Cs-131stranded 48.4%. There were no reported cases of symp-
seeds were placed with a planned dose of 80 Gy tomatic radiation necrosis [63].
to a 5 mm depth from the resection cavity sur- The high radionecrosis rates in the aforemen-
face. Each Cs-131 suture-stranded string con- tioned I-125 brachytherapy trials have been
tained 10 seeds (0.5 cm inter-seed spacing), were attributed to a combination of I-125’s long half-
cut into shorter segments as dictated by cavity life, and tumor cavity shrinkage affecting radio-
size, and permanently placed along the cavity in active seed placement, altering their dosimetry,
a tangential pattern to maintain a 7–10 mm spac- and exposing normal brain parenchyma to exces-
ing between seeds (Figs. 31.1 and 31.2). The sive radiation [24, 64, 65]. In an attempt to
seeds were subsequently secured with Surgicel improve on these shortcomings, Wernicke and
a b
Fig. 31.1 (a) Preoperative and (b) Postoperative gadolinium enhanced MRI scans show resected single brain metasta-
sis with the cavity lined with cesium-131 brachytherapy seeds
patients in either group developed radiation

necrosis [66]. Pham et al. prospectively evalu-
ated the neurocognitive impact of intraoperative
Cs-131 in the patients from the previously
described phase I/II trial by Wernicke et al. [24].
The mini-mental status examination (MSSE)
and functional assessment of cancer therapy-
brain (FACT-Br) questionnaire were performed
pretreatment, and at 2, 4, 6, and 12 months fol-
lowing treatment. The authors reported an
improvement from baseline In MMSE score at
4–12 months (30 versus 29, p = 0.017; 30 versus
29, p = 0.001, respectively), and in FACT-BR
Fig. 31.2 Intraoperative photograph shows the seeds on
a blue vicryl suture lining the resection cavity score at 4 and 6 months (162 versus 143,
p = 0.004; 164 versus 143, p = 0.005, respec-
tively). They noted several limitations in this
colleagues prospectively evaluated the efficacy analysis, including a heterogeneous patient pop-
of combining the short half-life Cs-131 with a ulation, the original study not being powered to
“seeds-on-a-string” technique with fibrin glue to compare neurocognitive outcomes, the cohort
stabilize cavity volume and minimize shrinkage. having a high MSSE (≥27) at baseline, reflecting
In this study, 30 patients with brain metastases a healthier population, and the limited follow-up
(and 6 more included from the previously time [67]. Additionally, this study does not
described phase I/II study [24]) underwent surgi- account for neurocognitive effects relating to
cal resection and intraoperative Cs-131 implan- metastatic disease control, steroid use and sys-
tation, and were compared with 30 patients who temic therapies. A summary of prospective trials
underwent postoperative SRS. Cs-131 stranded evaluating neurocognitive effects of various
seeds were implanted approximately 1 cm apart radiation modalities is provided in Table 31.4.
within the cavity with a planned dose of 80 Gy Menachem et al. evaluated radiation exposure
prescribed to a 5 mm depth from the surface of to medical personnel involved in the surgical
the resection cavity. The seeds were implanted resection and intraoperative Cs-131 implantation
akin to “barrel staves” to utilize the string’s ten- of 20 patients with brain tumors (16 with brain
sile strength to maintain the resection cavity metastases). Cs-131 stranded seeds were used
shape with the intent of preventing seed move- with a planned dose of 80 Gy to a 5 mm depth
ment for at least 1 month, at which point almost from the surface of the resection cavity. Surgeons
90% of the prescribed dose would be delivered. and radiation oncologists wore dosimetry badges
Over a median follow-up period of 110 days, on the leaded aprons and rings underneath the
resection cavities significantly shrank among leaded gloves, and measured radiation dose equiv-
those receiving Cs-131 (56.5% median volume alent at the levels of “eye” (ocular lens), “shallow”
reduction) and SRS (84.8% median volume (hands/skin), and “deep” (whole-body). The dose
reduction). However, the Cs-131 cohort demon- rate in the room was also measured following the
strated a nonsignificant amount of cavity shrink- procedure to approximate exposure to other per-
age during the first month (22% volume sonnel having patient contact who do not routinely
reduction; p = 0.06), compared with the SRS wear protective equipment. Postoperatively, the
cohort (46.7%; p = 0.42%), possibly reflecting median dose rate to surface, 35 cm and 100 cm
the effects of the brachytherapy “seeds-on-a- distances were 0.2475 mSv per hour, 0.01 mSv per
string” and fibrin glue on cavity dynamics. No hour, and 0.001 mSv per hour, respectively. At
Table 31.4 Neurocognitive effects of various radiotherapy modalities

Dose and
Modality Study fractionation Assessment Outcomes
Post-op WBRT Brown 37.5 Gy in 15 Cognitive-deterioration-free survival: Median 3 months to cognitive
2017 fractions or Drop of >1 standard deviation in decline
[19] 30 Gy in 10 HVLT-R immediate recall, verbal
fractions fluency COWAT, TMT-A, TMT-B,
HVLT-R delayed recall, HVLT-R
recognition
Post-op SRS Brown 12–20 Gy in 1 Cognitive-deterioration-free survival Median 3.7 months to
2017 fraction HVLT-R immediate recall, verbal cognitive decline
[19] fluency COWAT, TMT-A, TMT-B,
HVLT-R delayed recall, HVLT-R
recognition
Pre-op SRS No trials to date evaluating neurocognitive effects
I-125 No trials to date evaluating neurocognitive effects
brachytherapy
Cs-131 Pham 80 Gy to 5 mm MMSE and FACT-BR FACT-BR improvement at 4
brachytherapy 2016 depth and 6 months compared to
[67] pre-treatment (162 vs. 143,
P = 0.004; 164 vs. 143,
P = 0.005, respectively)
MMSE improvement at 4 and
up to 12 months compared to
pre-treatment (30 vs. 29,
P = 0.017; 30 vs. 29,
P = 0.001, respectively)
HVLT-R Hopkins Verbal Learning Test-Revised, COWAT Controlled Oral Word Association Test, TMT-A Trail Making
Test part A, TMT-B Trail Making Test part B, MMSE Mini-Mental Status Examination, FACT-BR Functional Assessment
of Cancer Therapy-Brain
30 days following implantation, the dose rates the resection cavity surface was used. The median
were 0.0298 mSv per hour, 0.0012 per hour, and OS was 7 months, with a 1 year freedom from
0.0001 mSv per hour, respectively. Based on the local progression of 83.3% and one case of
National Council on Radiation Protection guide- asymptomatic radionecrosis [69]. Wernicke et al.
lines, the authors concluded that the dose equiva- prospectively evaluated the efficacy of surgical
lent from permanent intracranial Cs-131 resection and Cs-131 implantation in 42 patients
brachytherapy maintains safe levels of exposure to with 46 metastases ≥2.0 cm in diameter (median
medical personnel and family [68]. preoperative lesion diameter was 3.0 cm). One
Cs-131 has also been evaluated in the setting year OS was 58%, with a 1 year regional failure
of recurrent metastases and in treating large from progression of 89% (80% in tumors <3.0
metastases. A retrospective review was per- cm), and a 52% distant failure from progression
formed on the outcomes of 13 patients with 15 rate. Lesion size was not significantly associated
brain metastases who underwent salvage resec- with any endpoint on multivariate analysis, and
tion and Cs-131 brachytherapy for recurrence there were no cases of radionecrosis [70].
following an initial radiotherapy course (10
patients had prior SRS to the lesion and 5 had
prior SRS and WBRT). The median resected I-125 Versus Cs-131 Brachytherapy
tumor diameter was 2.9 cm, which is signifi-
cantly larger than median cavity size in many of Cs-131 has several physical and radiobiological
the aforementioned postoperative SRS studies. A advantages over I-125 for brain brachytherapy.
prescription dose of 80 Gy at 5 mm depth from The intrinsically lower Cs-131 seed activity, jux-
taposed with lower dose prescriptions in the technical expertise of the physicians placing and
aforementioned studies, enables excellent LC securing the implants. Significant residual dis-
rates while minimizing the incidence of radiation ease is unlikely to be controlled with brachyther-
necrosis. Cs-131 has a higher dose rate than I-125 apy. It must also be noted that there is a small risk
(0.342 Gy per hour versus 0.069 Gy per hour), of seed migration following implantation [71–
translating to 90% Cs-131 dose absorption within 73]. Seed migration may occur when resections
33 days of implantation, whereas only 32% of are in close proximity to the ventricles (if opened
I-125’s dose would be delivered at this juncture. during surgery) or when seeds migrate out of the
Cs-131’s higher mean energy (29 keV) enables resection cavity into the subdural/subarachnoid
adequate dosimetry with the use of fewer seeds space [71]. If seeds are still radioactive at the
per given volume [63]. Han et al. used modelling time of migration, the treated cavity may be
methods to compare the effects of resection cav- underdosed and normal tissues in distant regions
ity changes on I-125 and Cs-131 implant dosim- will be exposed to radiation.
etry. The model was based on a single point Despite promising results in the aforemen-
source. Dose distributions were estimated via tioned studies, Cs-131 remains a less investigated
TG-43 calculations, and biological effective dose isotope than I-125, which has led to less frequent
calculations were compared for both radioactive routine utilization outside of clinical trials [74].
isotopes. They reported that resection cavities
reach their 50% reduction point at an average
3.4 months following surgery, resulting in signifi- Future Directions
cant differences between I-125 and Cs-131
dosimetry. In comparison to the cavity at time of Trials are needed to directly compare the efficacy
implantation, I-125 exhibited a 31.8% and 30.5% of I-125 to Cs-131 as well as to directly compare
increase in dose to 90% and 10% of the target the efficacy of intraoperative brachytherapy to
volume, respectively. Conversely, Cs-131 exhib- preoperative and postoperative SRS. These trials
ited a 1.44% and 0.64% increase in dose to 90% may be paradigm-changing in the setting of large
and 10% of the target volume, respectively, sug- metastases, recurrent disease or for patients
gesting changes in resection volumes affect whose individual social and medical contexts
Cs-131 dose distribution significantly less than hinder access to adjuvant radiation therapies.
that of I-125 for permanent brain implants [70].
Additionally, Cs-131 has a significantly
shorter half-life than I-125 (9.69 days versus References
59.4 days), denoting a shorter average life of the
seeds (Table. 31.3). Faster radiotherapy comple- 1. Gavrilovic IT, Posner JB. Brain metastases: epi-
tion (within 1 month of implantation) may impact demiology and pathophysiology. J Neurooncol.
a patient’s overall multidisciplinary management, 2005;75:5–14.
2. Nichols EM, Patchell RA, Regine WF, Kowk
as many patients possess extracranial metastases Y. Palliation of the brain and spinal cord metastases.
warranting systemic therapy that cannot be In: Halperin EC, Wazer DE, Perez CA, Brady LW,
safely administered simultaneously with radia- editors. Perez and Brady’s principles of radiation
tion. Furthermore, the choice of Cs-131 brachy- oncology. 6th ed. Philadelphia: Lippincott Williams
& Wilkins; 2013. p. 1765–72.
therapy seeds potentially exposes the patient and 3. Delattre JY, Krol G, Thaler HT, et al. Distribution of
those in their proximity to less radiation, as well brain metastases. Arch Neurol. 1988;45:741–4.
as hospital staff in the event of a subsequent neu- 4. Steeg PS, Camphausen KA, Smith QR. Brain metas-
rosurgical procedure than would be the case if tases as preventive and therapeutic targets. Nat Rev
Cancer. 2011;11(5):352–63.
I-125 seeds were chosen for a permanent implant 5. Mulvenna P, Nankivell M, Barton R, Faivre-Finn C,
following a metastasis resection. Wilson P, McColl E, Moore B, Brisbane I, Ardron
Both are subject to the ability to achieve gross D, Holt T, Morgan S, Lee C, Waite K, Bayman
total resection of the metastasis, as well as the N, Pugh C, Sydes B, Stephens R, Parmar MK,
Langley RE. Dexamethasone and supportive care 18. Roberge D, Parney I, Brown PD. Radiosurgery to the
with or without whole brain radiotherapy in treat- postoperative surgical cavity: who needs evidence?
ing patients with non-small cell lung cancer with Int J Radiat Oncol Biol Phys. 2012;83(2):486–93.
brain metastases unsuitable for resection or ste- 19.
Brown PD, Ballman KV, Cerhan JH, et al.
reotactic radiotherapy (QUARTZ): results from a Postoperative stereotactic radiosurgery compared
phase 3, non-inferiority, randomised trial. Lancet. with whole brain radiotherapy for resected metastatic
2016;388(10055):2004–14. brain disease (NCCTG N107C/CEC·3): a multicentre,
6. Borgelt B, Gelber R, Kramer S, et al. The palliation randomised, controlled, phase 3 trial. Lancet Oncol.
of brain metastases: final results of the first two stud- 2017;18(8):1049–60.
ies by the Radiation Therapy Oncology Group. Int J 20. Patel KR, Burri SH, Asher AL, et al. Comparing pre-
Radiat Oncol Biol Phys. 1980;6(1):1–9. operative with postoperative stereotactic radiosur-
7. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized gery for resectable brain metastases. Neurosurgery.
trial of surgery in the treatment of single metastases to 2016;79(2):279–85.
the brain. N Engl J Med. 1990;322(8):494–500. 21. Asher AL, Burri SH, Wiggins WF, Kelly RP, Boltes
8. Patchell RA, Tibbs PA, Regine WF, et al. Postoperative MO, Mehrlich M, Norton HJ, Fraser RW. A new
radiotherapy in the treatment of single metastases to treatment paradigm: neoadjuvant radiosurgery before
the brain. JAMA. 1998;280(17):1485–9. surgical resection of brain metastases with analysis of
9. Andrews DW, Scott CB, Sperduto PW, et al. Whole local tumor recurrence. Int J Radiat Oncol Biol Phys.
brain radiation therapy with or without stereotactic 2014;88(4):899–906.
radiosurgery boost for patients with one to three brain 22. Soltys SG, Adler JR, Lipani JD, et al. Stereotactic
metastases: phase III results of the RTOG 9508 ran- radiosurgery of the postoperative resection cavity
domised trial. Lancet. 2004;363(9422):1665–72. for brain metastases. Int J Radiat Oncol Biol Phys.
10. Aoyama H, Shirato H, Tago M, et al. Stereotactic 2008;70(1):187–93.
radiosurgery plus whole-brain radiation therapy vs 23. Choi CY, Chang SD, Gibbs IC, et al. Stereotactic
stereotactic radiosurgery alone for treatment of brain radiosurgery of the postoperative resection cavity
metastases. JAMA. 2006;295(21):2483–91. for brain metastases: prospective evaluation of tar-
11. Chang EL, Wefel JS, Hess KR, et al. Neurocognition get margin on tumor control. Int J Radiat Oncol Biol
in patients with brain metastases treated with radio- Phys. 2012;84(2):336–42.
surgery or radiosurgery plus whole-brain irradia- 24. Atalar B, Choi CY, Harsh GR, et al. Cavity volume
tion: a randomised controlled trial. Lancet Oncol. dynamics after resection of brain metastases and tim-
2009;10(11):1037–44. ing of postresection cavity stereotactic radiosurgery.
12. Kocher M, Soffietti R, Abacioglu U, et al. Adjuvant Neurosurgery. 2013;72(2):180–5; discussion 185.
whole-brain radiotherapy versus observation after 25. Jarvis LA, Simmons NE, Bellerive M, et al. Tumor
radiosurgery or surgical resection of one to three cere- bed dynamics after surgical resection of brain metas-
bral metastases: results of the EORTC 22952-26001 tases: implications for postoperative radiosurgery. Int
Study. J Clin Oncol. 2011;29(2):134–41. J Radiat Oncol Biol Phys. 2012;84(4):943–8.
13. Brown PD, Jaeckle K, Ballman KV, et al. Effect 26. Brennan C, Yang TJ, Hilden P, et al. A phase 2 trial of
of radiosurgery alone vs radiosurgery with whole stereotactic radiosurgery boost after surgical resection
brain radiation therapy on cognitive function in for brain metastases. Int J Radiat Oncol Biol Phys.
patients with 1 to 3 brain metastases. JAMA. 2014;88(1):130–6.
2016;316(4):401–9. 27. Prabhu R, Shu HK, Hadjipanayis C, et al. Current
14. Soffietti R, Kocher M, Abacioglu UM, et al. A
dosing paradigm for stereotactic radiosurgery alone
European Organisation for Research and Treatment after surgical resection of brain metastases needs to
of Cancer phase III trial of adjuvant whole-brain be optimized for improved local control. Int J Radiat
radiotherapy versus observation in patients with one Oncol Biol Phys. 2012;83(1):e61–6.
to three brain metastases from solid tumors after sur- 28. Seymour ZA, Fogh SE, Westcott SK, Braunstein S,
gical resection or radiosurgery: quality-of-life results. Larson DA, Barani IJ, Nakamura J, Sneed PK. Interval
J Clin Oncol. 2013;31(1):65–72. from imaging to treatment delivery in the radiation
15. Yamamoto M, Serizawa T, Shuto T. Stereotactic radiosurgery age: how long is too long? Int J Radiat Oncol
surgery for patients with multiple brain metastases Biol Phys. 2015;93(1):126–32.
(JLGK0901): a multi-institutional prospective obser- 29. Ward JF. The complexity of DNA damage: rele-

vational study. Lancet Oncol. 2014;15(4):387–95. vance to biological consequences. Int J Radiat Biol.
16. NCCN Guidelines Version 1.2018 Central Nervous 1994;66(5):427–32.
System Cancers, https://www.nccn.org/professionals/ 30. Shaw E, Scott C, Souhami L, et al. Single dose radio-
physician_gls/default.aspx. Accessed 6 May 2020. surgical treatment of recurrent previously irradiated
17. Mahajan A, Ahmed S, McAleer MF, et al. Post-
primary brain tumors and brain metastases: final
operative stereotactic radiosurgery versus observation report of RTOG protocol 90-05. Int J Radiat Oncol
for completely resected brain metastases: a single- Biol Phys. 2000;47(2):291–8.
centre, randomised, controlled, phase 3 trial. Lancet 31. Blonigen BJ, Steinmetz RD, Levin L, Lamba MA,
Oncol. 2017;18(8):1040–8. Warnick RE, Breneman JC. Irradiated volume as a
predictor of brain radionecrosis after linear accelera- for a multidisciplinary approach. Int J Radiat Oncol
tor stereotactic radiosurgery. Int J Radiat Oncol Biol Biol Phys. 2012;84:688–93.
Phys. 2010;77(4):996–1001. 45. Lee CC, Yen CP, Xu Z, et al. Large intracranial

32. Minniti G, Clarke E, Lanzetta G, et al. Stereotactic metastatic tumors treated by Gamma Knife sur-
radiosurgery for brain metastases: analysis of out- gery: outcomes and prognostic factors. J Neurosurg.
come and risk of brain radionecrosis. Radiat Oncol. 2014;120:52–9.
2011;6(1):48. 46. Baschnagel AM, Meyer KD, Chen PY, et al.

33. Sneed PK, Mendez J, Vemer-van den Hoek JG, et al. Tumor volume as a predictor of survival and local
Adverse radiation effect after stereotactic radiosur- control in patients with brain metastases treated
gery for brain metastases: incidence, time course, and with Gamma Knife surgery. J Neurosurg. 2013
risk factors. J Neurosurg. 2015;123(2):373–86. Nov;119:1139–44.
34. Bernstein M, Cabantog A, Laperriere N, Leung P, 47.
Sheehan JP, Sun MH, Kondziolka D, et al.
Thomason C. Brachytherapy for recurrent single Radiosurgery for non-small cell lung carcinoma meta-
brain metastasis. Can J Neurol Sci. 1995;22(1):13–6. static to the brain: long-term outcomes and prognos-
35. Bogart JA, Ungureanu C, Shihadeh E, Chung TC, tic factors influencing patient survival time and local
King GA, Ryu S, Kent C, Winfield JA. Resection and tumor control. J Neurosurg. 2002;97(6):1276–81.
permanent I-125 brachytherapy without whole brain 48. Ebner D, Rava P, Gorovets D, Cielo D, Hepel

irradiation for solitary brain metastasis from non-small JT. Stereotactic radiosurgery for large brain metasta-
cell lung carcinoma. J Neurooncol. 1999;44(1):53–7. ses. J Clin Neurosci. 2015;22(10):1650–4.
36. Dagnew E, Kanski J, McDermott MW, Sneed
49. Shrieve DC, Alexander E 3rd, Wen PY, et al.

PK, McPherson C, Breneman JC, Warnick Comparison of stereotactic radiosurgery and brachy-
RE. Management of newly diagnosed single brain therapy in the treatment of recurrent glioblastoma
metastasis using resection and permanent iodine-125 multiforme. Neurosurgery. 1995;36:275–84.
seeds without initial whole-brain radiotherapy: 50. Knox SJ, Sutherland W, Goris ML. Correlation of
a two institution experience. Neurosurg Focus. tumor sensitivityto low-dose-rate irradiation with
2007;22(3):E3. G2/M-phase bloc and other radiobiological param-
37. Huang K, Sneed PK, Kunwar S, Kragten A, Larson eters [Erratum appears in Radiat Res 1993; 136:439].
DA, Berger MS, Chan A, Pouliot J, McDermott Radiat Res. 1993;135:24–31.
MW. Surgical resection and permanent iodine-125 51. Rutigliano MJ, Lunsford LD, Kondziolka D, Strauss
brachytherapy for brain metastases. J Neurooncol. MJ, Khanna V, Green M. The cost effectiveness of
2009;91(1):83–93. stereotactic radiosurgery versus surgical resection
38. Ostertag CB, Kreth FW. Interstitial iodine-125 radio- in the treatment of solitary metastatic brain tumors.
surgery for cerebral metastases. Br J Neurosurg. Neurosurgery. 1995;37:445–55.
1995;9(5):593–603. 52. Mehta M, Noyes W, Craig B, Lamond J, Auchter R,
39. Schulder M, Black PM, Shrieve DC, Alexander E French M, Johnson M, Levin A, Badie B, Robbins
3rd, Loeffler JS. Permanent low-activity iodine-125 I, Kinsella T. A cost-effectiveness and cost-utility
implants for cerebral metastases. J Neurooncol. analysis of radiosurgery versus resection for single-
1997;33(3):213–21. brain metastases. Int J Radiat Oncol Biol Phys.
40. Limbrick DD Jr, Lusis EA, Chicoine MR, Rich KM, 1997;39:445–54.
Dacey RG, Dowling JL, et al. Combined surgical 53. Lee WY, Cho DY, Lee HC, Chuang HC, Chen CC,
resection and stereotactic radiosurgery for treatment Liu JL, Yang SN, Liang JA, Ho LH. Outcomes and
of cerebral metastases. Surg Neurol. 2009;71:280–8. cost-effectiveness of gamma knife radiosurgery and
41. Mathieu D, Kondziolka D, Flickinger JC, Fortin D, whole brain radiotherapy for multiple metastatic brain
Kenny B, Michaud K, et al. Tumor bed radiosurgery tumors. J Clin Neurosci. 2009;16:630–4.
after resection of cerebral metastases. Neurosurgery. 54. Wernicke AG, Yondorf MZ, Parashar B, Nori D,

2008;62:817–23. Clifford Chao KS, Boockvar JA, Pannullo S, Stieg
42. Wegner RE, Leeman JE, Kabolizadeh P, Rwigema JC, P, Schwartz TH. The cost-effectiveness of surgical
Mintz AH, Burton SA, et al. Fractionated stereotac- resection and cesium-131 intraoperative brachy-
tic radiosurgeryfor large brain metastases. Am J Clin therapy versus surgical resection and stereotactic
Oncol. 2015;38:135–9. radiosurgery in the treatment of metastatic brain

43. Wernicke AG, Hirschfeld CB, Smith AW, Taube S, tumors. J Neurooncol. 2016;127(1):145–53.
Yondorf MZ, Parashar B, Nedialkova L, Kulidzhanov 55.
Sneed PK, Lamborn KR, Larson DA, et al.
F, Trichter S, Sabbas A, Ramakrishna R, Pannullo Demonstration of brachytherapy boost dose-response
S, Schwartz TH. Clinical outcomes of large brain relationships in glioblastoma multiforme. Int J Radiat
metastases treated with neurosurgical resection and Oncol Biol Phys. 1996;35:37–44.
intraoperative Cesium-131 brachytherapy: results 56. McDermott MW, Sneed PK, Gutin PH. Interstitial
of a prospective trial. Int J Radiat Oncol Biol Phys. brachytherapy for malignant brain tumors. Semin
2017;98(5):1059–68. Surg Oncol. 1998;14(1):79–87.
44. Choi CY, Chang SD, Gibbs IC, et al. What is the opti- 57. Halligan JB, Stelzer KJ, Rostomily RC, et al.

mal treatment of large brain metastases? An argument Operation and permanent low activity 125I brachy-
therapy for recurrent high grade astrocytomas. Int J brachytherapy: a prospective trial. J Neurooncol.
Radiat Oncol Biol Phys. 1996;35:541–7. 2016;127(1):63–71.
58. Schulz RJ, Chandra P, Nath R. Determination of the 68. Yondorf M, Schwartz T, Boockvar J, Pannullo S,

exposure rate constant for 125I using a scintillation Stieg P, Sabbas A, Pavese A, Trichter S, Nedialkova L,
detector. Med Phys. 1980;7:355–61. Parashar B, Nori D, Chao KS, Wernicke A. Radiation
59. Prados M, Leibel S, Barnett CM, Gutin P. Interstitial exposure and safety precautions following 131Cs
brachytherapy for metastatic brain tumors. Cancer. brachytherapy in patients with brain tumors. Health
1989;63(4):657–60. Phys. 2017;112:403–8.
60. Petr MJ, McPherson CM, Breneman JC, Warnick
69. Wernicke AG, Smith AW, Taube S, Yondorf MZ,

RE. Management of newly diagnosed single brain Parashar B, Trichter S, Nedialkova L, Sabbas A,
metastasis with surgical resection and permanent Christos P, Ramakrishna R, Pannullo SC, Stieg PE,
I-125 seeds without upfront whole brain radiotherapy. Schwartz TH. Cesium-131 brachytherapy for recur-
J Neurooncol. 2009;92(3):393–400. rent brain metastases: durable salvage treatment for
61. Raleigh DR, Seymour ZA, Tomlin B, Theodosopoulos previously irradiated metastatic disease. J Neurosurg.
PV, Berger MS, Aghi MK, Geneser SE, Krishnamurthy 2017 Apr;126(4):1212–9.
D, Fogh SE, Sneed PK, MW MD. Resection and brain 70. Han DY, Ma L, Braunstein S, Raleigh D, Sneed

brachytherapy with permanent iodine-125 sources for PK, McDermott M. Resection cavity contraction
brain metastasis. J Neurosurg. 2017;126:1749–55. effects in the use of radioactive sources (I-125 versus
62. Yang R, Wang J, Zhang H. Dosimetric study of
Cs-131) for intra-operative brain implants. Cureus.
CS-131, I-125 and Pd-103 seeds for permanent pros- 2018;10(1):e2079.
tate brachytherapy. Cancer Biother Radiopharm. 71. Hirschfeld CB, Schwartz TH, Parashar B, Wernicke
2009;24:701–5. AG. Seed migration to the spinal canal after postre-
63. Wernicke AG, Yondorf MZ, Peng L, Trichter S,
section brachytherapy to treat a large brain metastasis.
Nedialkova L, Sabbas A, Kulidzhanov F, Parashar B, Brachytherapy. 2016;15(5):637–41.
Nori D, Clifford Chao KS, Christos P, Kovanlikaya 72. Brahimaj B, Lamba M, Breneman JC, Warnick

I, Pannullo S, Boockvar JA, Stieg PE, Schwartz RE. Iodine-125 seed migration within brain paren-
TH. Phase I/II study of resection and intraoperative chyma after brachytherapy for brain metastasis: case
cesium-131 radioisotope brachytherapy in patients report. J Neurosurg. 2016;125:1–4.
with newly diagnosed brain metastases. J Neurosurg. 73. Larson DA, Suplica JM, Chang SM, et al. Permanent
2014;121(2):338–48. iodine 125 brachytherapy in patients with progressive
64. Dale RG, Jones B, Coles IP. Effect of tumour shrinkage or recurrent glioblastoma multiforme. Neuro Oncol.
on the biological effectiveness of permanent brachy- 2004;6:119–26.
therapy implants. Br J Radiol. 1994;67(799):639–45. 74. Butler WM. Counterpoint: Cesium-131: not ready for
65. Dale RG, Jones B. Enhanced normal tissue doses prime time. Brachytherapy. 2009;8:4–6; discussion 7.
caused by tumor shrinkage during brachytherapy. Br J 75. Ruge MI, et al. Stereotactic biopsy combined with
Radiol. 1999;72(857):499–501. stereotactic (125)iodine brachytherapy for diagnosis
66. Wernicke AG, Lazow SP, Taube S, Yondorf MZ,
and treatment of locally recurrent single brain metas-
Kovanlikaya I, Nori D, Christos P, Boockvar JA, tases. J Neurooncol. 2011;105:109–18.
Pannullo S, Stieg PE, Schwartz TH. Surgical tech- 76. Khan FM, Gibbons JP. Khan’s the physics of radia-
nique and clinically relevant resection cavity dynam- tion therapy. Philadelphia: Lippincott Williams and
ics following implantation of cesium-131 (Cs-131) Wilkins; 2014. p. 310.
brachytherapy in patients with brain metastases. Oper 77. Kehwar TS. Use of Cesium-131 radioactive seeds

Neurosurg. 2016;12(1):49–60. in prostate permanent implants. J Med Phys.
67. Pham A, Yondorf MZ, Parashar B, Scheff RJ,
2009;34:191–3.
Pannullo SC, Ramakrishna R, Stieg PE, Schwartz TH, 78. Smith DS, Stabin MG. Exposure rate constants and
Wernicke AG. Neurocognitive function and quality of lead shielding values for over 1,100 radionuclides.
life in patients with newly diagnosed brain metasta- Health Phys. 2012;102:271–91.
sis after treatment with intra-operative cesium-131
Laser Interstitial Thermal Therapy
for Brain Metastases and Radiation
32
Necrosis
Jeffrey I. Traylor, Ahmed Habib, Vittorio Stumpo,

Dhiego Chaves de Almeida Bastos,
and Sujit S. Prabhu
Introduction destruction potential with minimal collateral

damage to surrounding tissue. These results pre-
History cipitated new efforts in the development of laser
ablation therapy in the subsequent decades lead-
Laser interstitial thermal therapy (LITT) is a mining to data on the effects of various light wave-
imally invasive operative technique that delivers lengths and optic fiber probe tips on surrounding
ablation under magnetic resonance imaging neural tissue [3, 4]. However, it was not until
(MRI) guidance. The principle of LITT is derived 1995 with the advent of magnetic resonance
from animal experiments in the 1960s describing (MR) thermography, that the potential applica-
the ablation of melanomas and sarcomas using a tions of LITT therapy were further examined as
neodymium laser [1]. These earliest observations real-time imaging guidance became a reality [5].
sparked clinical trials that demonstrated potential Since that time, studies have shown promise for
for the technique; however, contemporaneous LITT in the management of a wide range of sur-
limitations in laser delivery systems and techni- gical disorders.
cal difficulties in operation contributed to the In the earliest stages of LITT experimentation,
arrest of further development as a therapeutic ablation was delivered to skin surface tumors by
alternative. In 1983, almost two decades after the glass fibers, though Bown and colleagues did
original animal models, Bown et al. described the describe the potential benefit of flexible fiber
factors influencing the interaction of laser light transmission for interstitial delivery with
with living tissue based on tissue models and the Nd:YAG lasers [2]. However, a limiting element
utilities of three laser varieties: CO2, argon, and that prevented large-scale utilization and investi-
neodymium-doped yttrium aluminum garnet gation was the poor method of estimation of the
(Nd:YAG) [2]. Specifically, the authors found thermal damage zone, conventionally done by
that Nd:YAG and argon lasers had greater foci postoperative imaging, which made LITT too
risky for use in eloquent regions [6]. The modern
development of durable optic fibers for treatment
J. I. Traylor · A. Habib · D. C. de A. Bastos
S. S. Prabhu (*) delivery and stereotactic guidance has increased
Department of Neurosurgery, The University of Texas precision in the placement of the treatment probe
M.D. Anderson Cancer Center, Houston, TX, USA [5]. In addition, MR thermometry is critical to the
e-mail: sprabhu@mdanderson.edu viability of LITT as real-time monitoring of tem-
V. Stumpo perature and tissue damage allows for optimiza-
Università Cattolica del Sacro Cuore School of tion of ablation temperatures to the region of
Medicine and Surgery, Rome, Italy

https://doi.org/10.1007/978-3-030-42958-4_32
interest while minimizing collateral damage to Types of Lasers and Probes

nearby tissue [7]. The combination of better laser
delivery systems, stereotactic techniques, and Currently, two lasers are predominately used for
real-time MR thermometry have manifested a LITT: continuous wave Nd:YAG, first described
new era of clinical research. This chapter will by Bown et al., and diode lasers [2, 15]. With
describe the operative technique for LITT as well wavelengths within the infrared spectrum
as the current evidence for the management of (between 1000 and 1100 nm), Nd:YAG lasers
cerebral metastases and radiation necrosis. have the highest penetration potential and are
indicated for highly vascularized soft tissues
[16]. On the other hand, diode lasers have the
LITT Mechanism ability to deliver energy more rapidly, ablating
lesions in less time due to a higher water absorp-
The therapeutic benefits of LITT rely on the tion coefficient [15, 17]. LITT delivery relies on
components of high-intensity electromagnetic optic fibers composed of either quartz or sap-
radiation (EMR), light power density, wave- phire, with the terminal probe composed of a
length, exposure duration, and exposure method heat-resistant flexible material that does not
(surface vs. interstitial) [4]. Tissue properties, absorb light between 200 and 2000 nm. In addi-
such as water and hemoglobin content, affect tion, the recent development of fluid and gas
the absorption of laser light and contribute to cooling systems for LITT probes have decreased
the vulnerability of various lesions to LITT [8]. probe adherence to ablated tissues, improving
In addition, optical properties of various intra- reliability and control [17].
cranial structures, such as the absorption coef-
ficients, scattering coefficients, and anisotropy
factors contribute to laser penetration [9]. The Current LITT Applications
earliest experiments investigating LITT pri-
marily utilized surface exposure, whereas mod- For the management of neurosurgical disorders,
ern interstitial exposure delivers LITT directly LITT probes are often combined with stereotac-
to the center of the target lesion, minimizing tic navigation, making it suitable for the ablation
damage to the surrounding tissue. Thermal of deep-seated, otherwise inaccessible, lesions.
damage is the primary mechanism of destruc- Additionally, LITT has served as an alternative
tion resulting in enzyme induction, coagulation for the management of radioresistant tumors and
necrosis, protein denaturation, and vessel scle- ablation for epileptogenic foci in adults and chil-
rosis [10–12]. Histologically, edema, neuronal dren [18, 19]. LITT has also been used for the
swelling, and cell membrane disruption can be treatment of deep-seated tumors in particular
seen and contribute to LITT- i nduced tissue with some success [16].
necrosis [13]. Three tissue zones have been
described surrounding the LITT probe. The
first zone nearest the probe undergoes coagula- ommercially Available Delivery
C
tion necrosis, the second zone contains some Systems
tissue necrosis as well as edema, and the third
zone contains injured cells with an intact abil- Two systems for LITT delivery are currently
ity to undergo repair [13]. These zones are commercially available: the NeuroBlate System
demarcated particularly well on T1-weighted (Monteris Medical, Inc., Winnipeg, Manitoba,
magnetic resonance imaging (MRI), though Canada) and the Visualase Thermal Therapy
the probe tract is best seen on T2-weighted System (Medtronic Inc., Minneapolis, MN,
images [14]. USA) (Table 32.1). NeuroBlate uses an Nd:YAG
32 Laser Interstitial Thermal Therapy for Brain Metastases and Radiation Necrosis 459
laser delivered by optical fiber. The probe tips are coordinated yellow, blue, and white corre-
available in 3.2 mm and 2.1 mm diameters and sponding with the previously described zonal
are cooled by a CO2-gas system [19]. Monteris architecture of tissue following laser hyperther-
has developed the M-Vision software for real- mia [21]. Tissue demarcated by the white TDT
time stereotactic guidance which allows the user line represents tissue heated to 43 °C for
to define the target region, map probe trajectory, 60 minutes and has undergone coagulative
and monitor temperature changes in the ablated necrosis (Fig. 32.1a). The blue line demarcates
tissue. Within this software suite, the extent of tissue that has sustained severe damage from
ablation is represented by thermal-damage- 10 minutes at 43 °C (Fig. 32.1b). The yellow
threshold (TDT) lines based on the Arrhenius line represents transient tissue injury with
rate process model [7]. Specifically, this model 2 minutes at 43 °C while tissue beyond this
establishes a first-order relationship between margin is assumed undamaged (Fig. 32.1c).
temperature, time, and cell injury and is used to The NeuroBlate system also employs a robotic
predict thermal tissue damage [20]. Accordingly, arm and side-fire probe that enables remote
increased time or temperature will result in a changes to the directionality of the ablation tip
greater extent of tissue ablation. intraoperatively.
Within the M-Vision suite, the TDT lines The Visualase system employs a 980 nm diode
derived from the Arrhenius equation are color- laser instead of Nd:YAG for lesion ablation [22].
The probe tip is cooled by circulating sterile,
Table 32.1 Comparison between the NeuroBlate and room temperature saline in the closed system. The
Visualase systems
location of the LITT probe is superimposed upon
NeuroBlate Visualase a preoperative MRI in the Visualase software suite
Integrated platform Cart-based platform workstation allowing for real-time guidance and
DICOM image co-registration
measurement of thermographic feedback. Though
3D outline of thermal therapy 2D only
zone and critical structures this system does not utilize the TDT line system
Dedicated head fixation 3rd party fixation favored by the NeuroBlate system, it produces
Software actuated laser rotation Manual laser probe unique, color-coded images to delineate thresh-
and depth control manipulation olds of thermal damage based on the same
Choice of 2 gas-cooled probes: Liquid-cooled,
Arrhenius model [7]. An additional feature of the
directional or diffusing diffusing
Multi-slice/multi-plane thermal Single-slice/ Visualase system is an automatic “trip-switch”
monitoring single-plane that deactivates the laser if the temperature sur-
3D display of thermal dose 2D display of thermal passes a predesignated threshold at “safety points”
contours dose contours set by the user based on the preoperative MRI.
a b c
Fig. 32.1 The white thermal damage threshold (TDT) tissue ablated at 43 °C for 10 minutes, and the yellow
line (a) delineates the area of tissue ablated at 43 °C for TDT line (c) delineates the area of tissue ablated at 43 °C
60 minutes, the blue TDT line (b) delineates the area of for 2 minutes
Operative Technique acquisition gradient-echo sequence (MP-RAGE)

with the results uploaded to the supplemental
Preoperative Preparation software suite (e.g., M-Vision with NeuroBlate).
At this stage, the lesion can be defined as well as
Patients scheduled to undergo LITT must receive potential trajectories for the LITT probe.
volumetric MRI sequences for procedure plan-
ning. Functional MRI (fMRI) with diffusion ten-
sor imaging (DTI) sequences are also Operative Procedure: Pre-LITT
recommended for patients with lesions adjacent
to white matter tracts. This additional analysis, Once the patient’s head is registered within the
particularly DTI with tractography, further stereotactic navigation suite, superimposed upon
defines the region of interest and allows the sur- the preoperative MP-RAGE MRI, the surgical
geon to plan a precise trajectory avoiding elo- site can be prepped and draped in sterile fashion
quent white matter tracts. The common approach according to hospital protocol. The interface
to trajectory planning superimposes the potential platform can be aligned with the proposed probe
thermal ablation zone on the preoperative MR trajectory to ensure an unopposed entry of the
images using a planning software. Then, a trajec- probe through the frame and the head immobili-
tory is established avoiding eloquent structures to zation ring. An incision is made with the number
the region of interest, taking into account the and trajectory of probes in mind (1 cm for a sin-
directionality of the probe tip [23]. If the volume gle probe). The interface platform can then be
of the region of interest is greater than 3 cm, mounted to the skull with stereotactic guidance
more than one trajectory will have to be employed and anchored by screws. Alternatively, a small
as the diameter of thermal ablation is 1.5 cm from (5 mm) burr hole is created with a pneumatic
the probe tip, thus influencing the size of the orig- drill and the dura opened and dilated. Then, a
inal incision. Alternatively, multiple probe tips cannulated bolt is placed under image guidance
may be used to ensure adequate tumor ablation. using the VarioGuide system by Brainlab
The procedure itself can take place in an (Brainlab, Munich, Germany). Based on the
intraoperative or diagnostic MRI suite. The planned trajectory, a 4.3 mm non-skiving drill bit
NeuroBlate system, in particular, is compatible is used to make a single burr hole, through which
with several MRI system manufacturers, includ- a 4 mm skull bolt is attached to the skull
ing IMRIS (Winnipeg, Manitoba, Canada), Tesla (Fig. 32.2a). The pre-measured laser probe is
Siemens (Erlangen, Germany), and GE then passed through the bolt and anchored
Healthcare (Waukesha, WI, USA). Following (Fig. 32.2b–d). This system has significant
induction of anesthesia, patients should receive advantages to the previous AxiiS system and
10 mg of intravenous corticosteroids and be simplifies the surgical process.
positioned on a stabilization system. NeuroBlate The LITT software suite (e.g., M-Vision) can
utilizes the AtamA system for this purpose which be used to determine the distance of the deepest
employs a head immobilization ring with three margin of the lesion from the burr hole. This will
to four pins, allowing supine, prone, or lateral allow the surgeon to select the shortest probe that
positioning of the patient. After final arrange- can access the deepest margin of the lesion. The
ments are made to ensure stabilization, including probe driver commander is placed into the inter-
sufficient reduction of risk for neuropathic and face platform with the probe driver follower
vasculopathic complications, sterile fiducials are placed into the central bore of the apparatus. The
placed on the surgical site and immobilization probe can now be guided through the mini-frame
ring (AtamA for the NeuroBlate system) for ste- and burr hole following the selection of the depth
reotactic orientation. A preoperative MR image stop based on lesion margin measurements. Once
is performed with magnetization prepared rapid the laser probe is seated into the probe driver
a b
c d
Fig. 32.2 Images of LITT procedure at M.D. Anderson erative MRI scanner with the delivery probe in place (c)
Cancer Center. A cannulated bolt is placed in the patient’s and a pre-ablation T1-weighted MRI is obtained confirm-
skull (a) followed by placement of the LITT probe (b) ing correct placement of the probe (d)
through the bolt. The patient is then placed in the intraop-
another MRI of the patient is taken to confirm the During treatment, the software suite will display
correct orientation of the probe based on the coronal and sagittal plane images as well as three
planned trajectory and to guide position re- axial plane images with real-time feedback of the
adjustments if necessary. probe location. Once the probe is inserted to the
desired depth within the lesion, corresponding to
the fused MRIs, the thermography sequences can
Operative Procedure: LITT begin. Depending on the type of probe and deliv-
ery system used, the direction of laser fire may
With the probe in an acceptable location at start- require selection at this point that will best be
ing depth, the MRIs are fused together and the contained within the margins of the lesion. Eight
probe coordinates superimposed over the planned cycles, every eight seconds, of scanning for tem-
trajectory created within the software suite. perature reference points must be done prior to
laser activation followed by cooling of the probe. LITT for Brain Metastases

The operator activates a switch on the software
suite screen which arms the foot switch for laser Background
activation. Total treatment time correlates with
tumor size, number of trajectories, and type of Brain metastases occur in 10–20% of adults with
laser (e.g., diode lasers have shorter ablation underlying malignancy and are estimated to be
times) as well as tissue hydration, directionality ten times more prevalent than primary intracra-
of the probe tip, and proximity to eloquent cortex nial tumors [24]. Conventional treatment modali-
or white matter tracts [7]. ties include surgical resection, whole brain
radiotherapy (WBRT) and stereotactic radiosur-
gery (SRS), or a combination of these. Treatment
Operative Procedure: Post-LITT of choice should be individualized according to
clinical (age, Karnofsky Performance Scale
The protocol followed at M.D. Anderson Cancer [KPS] score, primary tumor control, extracranial
Center calls for a final MRI before withdrawing metastases), pathological (primary tumor histol-
the probe following ablation sequences along all ogy), and radiological aspects (number of brain
trajectories, at which point the probe driver and metastasis, functional location, deep-seated
interface platform are removed. The skull bolt is lesions, etc.) [25]. Patient preference and esti-
removed using a hex tool and the wound is irri- mated quality of life resulting from treatment in
gated followed by hemostasis, then the skin is the setting of terminal metastatic disease should
closed with a single suture and dressed. Following also be considered; the optimal therapeutic
arousal from anesthesia, a neurological exam is approach must balance risks and benefits as well
performed on the patient to determine any as patient particularities. Rapidly improving sys-
changes from the preoperative condition. On temic therapies have prolonged the survival of
postoperative day one, an MRI is recommended cancer patients subsequently increasing the inci-
to evaluate residual tumor volume and extent of dence of brain metastases, yet the poor penetra-
ablation (Fig. 32.3). For uncomplicated cases, tion of the blood–brain barrier by most of these
hospital stay is typically one day from the time of agents contributes to limited efficacy [26]. While
operation. The taper of corticosteroids can be an increasing amount of basic and clinical
based on the extent of postoperative edema at the research has made progress in delineating the
surgeon’s discretion. genetics, tumor microenvironment, mechanisms
Fig. 32.3 T1-weighted post-contrast MR images showing a metastatic lesion preceding LITT (a), immediately after
LITT (b), and at one-month follow-up (c)
of leptomeningeal spread, and effects on neuro- who received LITT [29]. In both patients, LITT
cognition, local treatment with surgical resection, was utilized for the management of recurrent
SRS, WBRT, either alone or in combination metastases and in both cases the tumor returned
remains the cornerstone of therapy for patients and required additional resection. Although these
with brain metastases. Since the introduction of results were suboptimal, the authors noted that
LITT, several case reports and case series have failure reporting for LITT is required to properly
been published, describing the efficacy of this define the utility of this procedure.
technique for the management of brain metasta- In 2016, Ali et al. reported on the first multi-
ses (Table 32.2). center study of the treatment of LITT for post-
SRS recurrent cerebral metastases in a cohort of
23 patients with 26 total lesions ranging in vol-
Current Evidence ume from 0.4 to 28.9 cm3 [30]. Disease control
was obtained in 17 cases while 9 lesions (35%)
In 2008, Carpentier and colleagues published showed disease progression after LITT. Notably,
pilot results of the first phase I study utilizing this only occurred in lesions that received <80%
MR-guided LITT for the management of patients ablation. The authors concluded that LITT can be
with cerebral metastases [22]. The patient cohort considered an effective treatment when tumor
primarily consisted of four patients with unre- ablation exceeds 80% but highlighted the impor-
sectable intracranial metastases refractory to tance of risk evaluation for complications that
multiple treatments (chemotherapy, WBRT, and may ensue following treatment of larger lesions
SRS). The authors utilized the Visualase system (defined as >20 cm3).
and reported positive results; all patients toler- In 2018, Eichberg and colleagues reported the
ated the procedure well and were discharged results of a pilot study of LITT for four patients
within 14 hours postoperatively. All lesions were with metastatic lesions in the posterior fossa [31].
observed to increase in volume at immediate Like previous studies, lesions volumes were ini-
follow-up, followed by a gradual decrease in tially increased before gradually decreasing. The
size. No lesion recurrences occurred at any point authors observed no complications and no clini-
during the 7, 15, 30, or 180-day follow-ups. The cal or radiographic evidence of tumor progres-
authors concluded LITT to be a safe, effective sion. They thus concluded LITT to be safe and
treatment for focal metastatic disease [22]. effective for cerebellar metastases. These find-
Carpentier again investigated the feasibility of ings were echoed the same year by Razavi et al.
the Visualase system in a cohort of seven in a study of eight patients who underwent LITT
patients, reporting similar results, with a median treatment, three of which had metastatic lesions
overall survival of 19.8 months [27]. in the posterior fossa [32].
Hawasli et al. provided additional evidence In the largest study on the subject to date,
for LITT in a 2013 prospective study of 17 Beechar et al. performed a volumetric analysis of
patients, 5 of which had cerebral metastases [28]. recurrent lesions managed with LITT following
The authors reported an initial increase in lesion SRS [33]. Using T1 post-contrast and T2 fluid-
size at follow-up with subsequent steady volume attenuated inversion recovery (FLAIR) MRI
decrease. The pooled analysis of LITT for pri- sequences for evaluation of edema, 50 total
mary brain tumors and metastases reduces the lesions from 36 patients were treated with LITT
reliability of this data for guiding LITT for brain with a significant overall reduction in lesion size.
metastases, specifically. However, the authors However, 37% of lesions demonstrated an upward
concluded LITT to be a viable treatment option trend overall on follow-up MRI. The authors con-
for cerebral metastases in selected patients. cluded that pre-treatment tumor volume plays a
Fabiano et al. reported different findings in a significant role in determining LITT response,
series of two patients with cerebral metastases with preferable responses in smaller lesions.
Table 32.2 Studies of LITT for brain metastases
464
Year No. of Tumor Primary Lesion diameter/volumeb

Author published patients locationa histology (cm/cm3) Outcome Complications
Carpentier et al. 2008 4 Frontal (1) Breast (5) NR Peripheral recurrence at 3 None
Temporal (2) NSCLC (1) months (3)
Parietal (2)
Occipital (1)
Carpentier et al. 2011 7 NR Breastc Range 1–3 cm Median OS: 19.8 months Probe misplacement (1)
NSLCc Mean PFS: 3.8 months Cerebellar syndrome (1)
Transient aphasia (1)
Jethwa et al. 2012 20 Parietal (2) NR Median 7.0 cm3 NR None (BM patients)
Cerebellar (1)
Frontal (1)
Hawasli et al. 2013 17 Insula (1) Colon (1) Mean 11.6 cm3 Median PFS: 5.8 months Transient aphasia (3)
Frontal (2) Melanoma (1) Median OS: 5.8 months Transient hemiparesis (3)
Parietal (1) SCLC (2) Transient hyponatremia
FP (1) Ovarian (1) (2)
DVT (1)
Fatal meningitis (1)
Ali et al. 2016 23 Frontal (10) Breast (6) Median 4.9 cm3 Recurrence (9) Transient hemiparesis (3)
Parietal (4) Lung (6) Local control (17) Hydrocephalus (1)
Occipital (2) Melanoma (5) Malignant cerebral edema
1 lt motor Colon (2) (1)
strip Ovarian (1)
Insular (1) Bladder (1)
BG (2) Esophagus (1)
Cerebellar (1) Sarcoma (1)
PO (2)
Thalamic (4)
Beechar et al. (both CRN 2017 36 NR NSLC (8) Median 5.1 cm3 ↑ Lesion size transient Motor disturbance (9)
and BM) SCLC (2) (19) Gait disturbance (8)
Breast (8) ↑ Lesion size sustained Visual disturbance (5)
Esophagus (1) (14) Sensory disturbance (2)
SCC (1) ↓ Lesion size over time Aphasia (2)
RCC (1) (31) Memory difficulty (1)
Rectal (1) Headache (1)
Sarcoma (2)
Melanoma (15)
Bladder (1)
J. I. Traylor et al.
Year No. of Tumor Primary Lesion diameter/volumeb
Author published patients locationa histology (cm/cm3) Outcome Complications
Eichberg et al. 2017 4 Cerebellar (4) Breast (3) Median 3.4 cm3 Stable (4) None
Ovarian (1)
Chaunzwa et al. (both 2018 30 Frontal (16) Lung (16) Median 7.6 cm3 Survival Intraoperative
CRN and BM) Parietal (4) Melanoma (5) 6 months (15) Hemorrhage (13%)
Occipital (5) Breast (3) 12 months (7)
Temporal (3) Colon (1) 18 months (4)
Insular (1) Gynecological 25 months (1)
BG (1) (2) 30 months (1)
RCC (1) Local control
Other (2) 92.6 months (92.9%)
Overall (83%)
Ahluwalia et al. 2018 42 Frontal (41%) Breast (10%) Mean 7.1 cm3 Survival Postop
Parietal (29) NSCSLC (50%) 12 weeks (71%) Complications (5%)
Cerebellar SCLC (5%) 26 weeks (64.5%) Intracerebral hemorrhage
(14%) Melanoma (5%)
Otherd (16%) (10%) Weakness (5%)
Otherd (25%)
Hernandez et al. 2018 45 NR NSCLC (31) Mean 3.4 cm3 Local control (83.1%) Complications (25%)
Breast (17) Recurrence (10)
Colon (2)
RCC (2)
Melanoma (3)
Testicular (2)
Cervical (1)
SCLC (1)
Razavi et al. 2018 8 Cerebellar (3) Colon (2) Median 5.4 cm3 Recurrence at 7.5 months CN 6 palsy (1)
32 Laser Interstitial Thermal Therapy for Brain Metastases and Radiation Necrosis
NSCLC (1) (1)

Stable (2)
NR not reported, MI myocardial infarction, PE pulmonary embolus, FP frontoparietal, CC corpus callosum, BG basal ganglia, PV periventricular, TP temporoparietal, FT fron-
totemporal, NSCLC non-small cell lung cancer, SCLC small cell lung cancer, RCC renal cell carcinoma, SCC squamous cell carcinoma, Pts patients, OS overall survival, PFS
progression-free survival, CN cranial nerve
a
Some patients have more than one tumor
b
Articles vary in describing lesion diameter or volume
c
Number not recorded
d
Occipital lobe, temporal lobe, thalamus, and other deep nuclei
465
Ahluwalia et al. reported on the results of the illustrate a role for this therapy in the manage-
first multicenter phase II trial of LITT for patients ment of metastases not amenable to SRS,
with radiographic progression after SRS for namely, those >3 cm in size [33].
intracranial metastases as part of the Laser We stress the need for prospective collection
Ablation After Stereotactic Radiosurgery clinical of QOL and cognition data in future studies to
trial (LAASR study, NCT01651078) [34]. Of 42 provide evidence for the role of this novel thera-
patients enrolled in the trial, 20 were confirmed peutic in allowing terminally ill patients to retain
to have a recurrence of intracranial metastases. In QOL after salvage treatment. It has been reported
addition to being well powered, this study was that when total ablation can be performed, KPS,
significant in addressing the diagnostic and man- cognitive status, and QOL can be preserved but
agement conundrum of lesion recurrence follow- additional prospective studies are needed to con-
ing SRS and the authors reported improved short firm these observations [34]. Complications asso-
term overall and progression-free survival in ciated with LITT are significantly less when
patients with radiation necrosis compared to compared to open cranial procedures and thus
cerebral metastases treated with LITT. Ultimately, acceptable in this patient population but can be
this trial provided evidence for LITT manage- associated with increased length of hospital stay.
ment with resultant stabilization of KPS, cogni-
tion, and quality-of-life (QOL) as well as a
reduction in steroid use. LITT for Radiation Necrosis
In light of the previously described diagnostic
and management conundrum associated with Background
post-SRS lesion recurrence, Hernandez et al. pro-
posed the radiographic definition of progressive Cerebral radiation necrosis (CRN) is a known
enhancing inflammatory reactions for unknown consequence of brain tumor management, affect-
lesions following SRS based on their results of a ing between 3% and 24% of patients receiving
retrospective study of 59 patients with 74 total cranial radiotherapy [14, 36]. The pathophysiol-
lesions [35]. Given the demonstrated efficacy and ogy of CRN is not fully understood, although a
safety reported on LITT for both conditions, the few theories have been reported in the literature.
authors argue that careful discrimination between One of the most accepted of these states that
these two conditions is unnecessary as good local CRN is driven by vascular endothelial damage
control was achieved for the ambiguous lesions leading to coagulation necrosis and reactive glio-
in a majority of the patients. sis in response to severe hypoxic insults by high
cumulative doses of radiation [37]. This is sup-
ported by the thickening of the endothelium and
Recommendations lymphocytic infiltration seen on histopathology
as well as the positive outcomes for CRN patients
The current body of work describing the safety associated with bevacizumab, an inhibitor of
and efficacy of LITT for cerebral metastases angiogenesis [38]. A second hypothesis suggests
which have failed radiotherapy is still in the that acute phase reactant cytokines in response to
early stages. The case series and small clinical radiation therapy may drive immune-mediated
trials have provided pilot data to evidence the damage to surrounding tissue that subsequently
utility of this therapy while noting some associ- precipitates inflammation, gliosis, and vasogenic
ated phenomena such as the initial increase in edema [39]. Though the exact molecular mecha-
lesion size before gradual volume reduction. nism is not yet fully described, researchers and
Though Beechar et al. found better LITT clinicians alike postulate that disruption of the
response in smaller metastatic tumors, the blood–brain barrier ultimately defines the patho-
results of other studies describing positive genesis [40]. Thus, a better understanding of the
results with different lesion sizes potentially molecular processes that contribute to this dis-
ease process can guide the development of more underlying malignancy and another patient
targeted therapies for treatment and prevention. required an additional craniotomy for lesion
The gold standard for diagnosis of CRN is regrowth. No complications occurred during the
biopsy, though MRI has limited diagnostic value procedure and the authors concluded that LITT is
[41]. There are often difficulties in distinguish- a feasible alternative for the treatment of lesion
ing between CRN and other pathologic pro- “regrowth” following SRS. It is important to
cesses on MRI, although some radiologic note, however, that stereotactic biopsy has an
techniques have been described [42]. CRN can intrinsic sample bias and refractory cases consid-
usually be managed conservatively with cortico- ered to be CRN may in fact correspond to tumor
steroids for associated edema followed by vari- progression within this setting.
ous experimental drugs if symptoms persist. Of In 2014, Fabiano and colleagues reported
these, bevacizumab has been reported to have on the case of a man who received SRS for a
some benefit, and anticoagulant/antiplatelet brain metastasis from lung adenocarcinoma.
medications have been shown to improve out- However, despite medical management, the
comes in some patients based on the ability to lesion continued to progress on imaging. A deci-
interfere with attributable underlying vascular sion for LITT was made based on the deep-
changes [43–48]. In addition, hyperbaric oxygen seated location of the lesion and resulted in a
has been shown to have some efficacy in the marked improvement in symptoms. Despite
management of these patients [49]. With conser- being described as CRN, no biopsy was per-
vative therapy, however, a subset of patients will formed to confirm the diagnosis; though it is
either fail to improve or experience progression plausible the lesion represented tumor recur-
of CRN, requiring a more aggressive manage- rence. Although it is unclear whether CRN was
ment strategy. Recently, case reports and patient the target of LITT in this case, the positive out-
series have illuminated a possible role for LITT come of the patient provides evidence, albeit
in cases of CRN refractory to rehabilitation and marginal, for the management of ambiguous
pharmacotherapy (Table 32.3). lesions in deep-seated loci.
Rahmathulla and colleagues were the first to The same year, Rao et al. published the results
describe LITT for the management of CRN in a of a cohort study investigating the utility of LITT
2012 case report [50]. Following SRS for man- for either tumor recurrence or CRN after SRS
agement of multiple brain metastases, a CRN [52]. In this retrospective cohort study, 16 patients
lesion was observed in the left centrum semiovale received SRS for metastatic intracranial tumors
with worsening edema refractory to high-dose with new onset of symptoms and MRI findings
glucocorticoid therapy. The authors performed consistent with either tumor recurrence or
LITT as the location of the lesion was not ame- CRN. These patients then received LITT for the
nable to resection which resulted in a successful management of these ambiguous recurrent
reduction in size at 7-week follow-up. The lesions (either tumor recurrence and/or CRN). Of
authors concluded that LITT is an option for the 15 patients with reliable follow-up, two expe-
patients with refractory CRN not amenable to rienced lesion recurrence again at 6 and 18 weeks,
surgical decompression [50]. respectively. Five patients died of extracranial
One year later, Torres et al. reported on the disease progression and one died of intracranial
results of six patients who underwent SRS for disease progression at a different locus. The
brain metastasis and were discovered to have authors concluded that LITT is a well-tolerated
lesion regrowth, later confirmed to be CRN on procedure that may be effective in treating tumor
biopsy [51]. LITT was performed to prevent fur- recurrence and/or CRN. This study provides
ther progression of neurologic symptoms and additional evidence for the utility of LITT in
edema. Four out of six patients treated with LITT managing CRN, though it again highlights the
had an improvement of neurologic symptoms. diagnostic conundrum of these lesions following
One patient died as a result of the progression of SRS.
Table 32.3 Studies of LITT for cerebral radiation necrosis

Lesion
No. of diameter/
Year patients volumeb
Author published (CRN) Tumor locationa (cm/cm3) Outcome Complications
Rahmathulla 2012 1 Motor cortex 2 cm or ↓ Lesion size and None
et al. 5.4 cm3 edema, ↓ steroid
requirement
Torres- 2013 6 Frontal (3), 0.68– ↑ Lesion size at NR
Reveron Cerebellum (2), 3.03 cm 2 weeks to 3 months,
et al. Parieto-occipital (1) then ↓ lesion size
4.5–6 months
Fabiano 2014 1 Frontal 1.8 cm ↓ Volume at 10 weeks NR
et al.
Rao et al. 2014 15 Frontal (6), 0.46– ↑ Lesion size at New-onset
Cerebellar (6), 25.45 cm3 24 hrs (12) ↓ lesion transient left-sided
Cerebellar peduncle size at 24 hrs (2), weaknessc (1)
(1), Temporal (1), lesion volume ≤ 10%
Parietal (1) pre-treatment at
16–44 weeks (7)
Smith et al. 2016 25 Frontal (11), NR Transient weakness
Cerebellum (1) (2), permanent
Temporal (5), weakness (1),
Parietal (2), steroid
Thalamus (1), complication (1)
Occipital (1), PV
(1), TP (1), FT (1),
CC (3), FP (2)
Ahulwalia 2018 19 NR 0.4– Stabilized KPS, Complete
et al. 13.2 cm3 preserved QOL hemiparesis (1),
↓ Steroid requirement headache (1),
hemineglect and
weakness (1)
Rammo et al. 2018 10 Frontal (4), 1.62 cm3 ↑ Lesion size at Intractable
Temporal (2) (mean) 1–2 weeks, ↓ lesion seizuresd (1), PE
Parietal (2), Frontal size at 6 months (1), MI (1)
thalamic (1), Transient delayed
Frontal medial (1) neurologic deficit
(3)
NR not reported, MI myocardial infarction, PE pulmonary embolus, FP frontoparietal, CC corpus callosum, PV peri-
ventricular, TP temporoparietal, FT frontotemporal
a
Some patients have more than one tumor
b
Articles vary in describing lesion diameter or volume
c
Patient has residual left-hand weakness
d
Patient had preceding seizure disorder worsened by LITT
Smith and colleagues demonstrated the out- The previously discussed phase II trial pub-
comes of LITT for biopsy-proven CRN in a cohort lished by Ahluwalia et al. in 2018 was the first
of 25 patients [53]. In this retrospective study, study of its kind and magnitude investigating
patients treated for primary and metastatic brain LITT for metastases and biopsy-proven radiation
tumors received LITT following stereotactic nee- necrosis [34]. Of 42 patients enrolled in the trial,
dle biopsy of recurrent lesions confirming CRN. No 19 had biopsy-confirmed CRN treated with
complications occurred during the procedure and LITT. In this study, the authors compared out-
overall survival and progression-free survival were comes of LITT for CRN and cerebral metastases
comparable to standard craniotomy and resection. and found longer progression-free and overall
survival rates at 12-week follow-up for patients Conclusion

with CRN, although this difference was not statis-
tically significant at 26 weeks. In this subset, LITT is a minimally invasive ablation technique
LITT stabilized the KPS score, preserved QOL which has recently seen a surge in research inves-
and cognition, and had a steroid-sparing effect. tigations and clinical applications for the treat-
The authors concluded that LITT is a low-risk ment of radiation necrosis and cerebral
procedure for patients with few alternative options metastases. The role of LITT in neurosurgical
for salvage treatment that can minimize cognitive oncology is evolving and well-powered, prospec-
decline, stabilize QOL and functional status, and tive studies are needed to fully establish its poten-
allow cessation of steroids in some cases. tial [13, 28, 55–61]. However, LITT appears to be
Rammo et al. reported on the most recent a safe modality in the management of lesion
study of LITT for CRN to date [54]. Ten patients recurrence following SRS, irrespective of the
with biopsy-proven CRN were retrospectively ultimate diagnosis.
reviewed to assess the outcome. Four patients
had neurologic deficits which resolved in three.
The authors concluded LITT to be a relatively
References
safe treatment for CRN with the added benefit of
being both diagnostic and therapeutic. Like the 1. McGuff PE, Deterling RA Jr, Gottlieb LS, Fahimi
previous study, Rammo and colleagues provide HD, Bushnell D. Surgical applications of laser. Ann
additional evidence for LITT management of Surg. 1964;160(4):765–77.
biopsy-proven CRN. 2. Bown SG. Phototherapy in tumors. World J Surg.
1983;7(6):700–9.
3. Cheng MK, McKean J, Boisvert D, Tulip J, Mielke
BW. Effects of photoradiation therapy on normal rat
Recommendations brain. Neurosurgery. 1984;15(6):804–10.
4. Elias Z, Powers SK, Atstupenas E, Brown
JT. Hyperthermia from interstitial laser irradiation
Since the original case report described by in normal rat brain. Lasers Surg Med. 1987;7(4):
Rahmathulla et al., LITT has been used as a sal- 370–5.
vage therapy for deep-seated lesions otherwise 5. De Poorter J. Noninvasive MRI thermometry with the
inaccessible by conventional resection tech- proton resonance frequency method: study of sus-
ceptibility effects. Magn Reson Med. 1995;34(3):
niques [50]. A number of small case series of 359–67.
patients with recurrent lesions after SRS without 6. Bettag M, Ulrich F, Schober R, Furst G, Langen KJ,
biopsy-proven CRN were published with good Sabel M, et al. Stereotactic laser therapy in cerebral
local control. These studies concluded that LITT gliomas. Acta Neurochir Suppl (Wien). 1991;52:
81–3.
is a safe and effective therapy for recurrence fol- 7. McNichols RJ, Gowda A, Kangasniemi M, Bankson
lowing SRS. JA, Price RE, Hazle JD. MR thermometry-based feed-
For patients with medically refractory CRN, back control of laser interstitial thermal therapy at
LITT offers a number of advantages in compari- 980 nm. Lasers Surg Med. 2004;34(1):48–55. https://
doi.org/10.1002/lsm.10243.
son with traditional resection techniques. 8. Mensel B, Weigel C, Hosten N. Laser-induced thermo-
Namely, the procedure itself is less invasive than therapy. Recent Results Cancer Res. 2006;167:69–75.
conventional craniotomy. In addition, patients 9. Yaroslavsky AN, Schulze PC, Yaroslavsky IV,
can resume their chemotherapy regimens soon Schober R, Ulrich F, Schwarzmaier HJ. Optical prop-
erties of selected native and coagulated human brain
after LITT as there is a theoretical advantage to tissues in vitro in the visible and near infrared spectral
the disruption of the blood–brain barrier by the range. Phys Med Biol. 2002;47(12):2059–73.
procedure. Although multicenter prospective 10. Stafford RJ, Fuentes D, Elliott AA, Weinberg JS,

studies are needed before detailed guidelines for Ahrar K. Laser-induced thermal therapy for tumor
ablation. Crit Rev Biomed Eng. 2010;38(1):79–100.
the management of refractory CRN are devel- 11. Sapareto SA, Dewey WC. Thermal dose determina-
oped, LITT has been shown to be an effective tion in cancer therapy. Int J Radiat Oncol Biol Phys.
treatment for these patients. 1984;10(6):787–800.
12. Nakagawa M, Matsumoto K, Higashi H, Furuta T, 24. Lin X, DeAngelis LM. Treatment of brain metasta-
Ohmoto T. Acute effects of interstitial hyperthermia ses. J Clin Oncol. 2015;33(30):3475–84. https://doi.
on normal monkey brain--magnetic resonance imag- org/10.1200/JCO.2015.60.9503.
ing appearance and effects on blood-brain barrier. 25. Soffietti R, Abacioglu U, Baumert B, Combs SE,
Neurol Med Chir (Tokyo). 1994;34(10):668–75. Kinhult S, Kros JM, et al. Diagnosis and treatment
13. Schulze PC, Vitzthum HE, Goldammer A, Schneider of brain metastases from solid tumors: guidelines
JP, Schober R. Laser-induced thermotherapy of neo- from the European Association of Neuro-Oncology
plastic lesions in the brain--underlying tissue altera- (EANO). Neuro Oncol. 2017;19(2):162–74. https://
tions, MRI-monitoring and clinical applicability. doi.org/10.1093/neuonc/now241.
Acta Neurochir. 2004;146(8):803–12. https://doi. 26. Arvold ND, Lee EQ, Mehta MP, Margolin K,
org/10.1007/s00701-004-0293-5. Alexander BM, Lin NU, et al. Updates in the
14. Rahmathulla G, Recinos PF, Kamian K, Mohammadi management of brain metastases. Neuro Oncol.
AM, Ahluwalia MS, Barnett GH. MRI-guided 2016;18(8):1043–65. https://doi.org/10.1093/neuonc/
laser interstitial thermal therapy in neuro- now127.
oncology: a review of its current clinical applica- 27. Carpentier A, McNichols RJ, Stafford RJ, Guichard
tions. Oncology. 2014;87(2):67–82. https://doi. JP, Reizine D, Delaloge S, et al. Laser thermal ther-
org/10.1159/000362817. apy: real-time MRI-guided and computer-controlled
15. Schmidt MH, Bajic DM, Reichert KW 2nd, Martin procedures for metastatic brain tumors. Lasers Surg
TS, Meyer GA, Whelan HT. Light-emitting diodes Med. 2011;43(10):943–50. https://doi.org/10.1002/
as a light source for intraoperative photodynamic lsm.21138.
therapy. Neurosurgery. 1996;38(3):552–6; discussion 28. Hawasli AH, Bagade S, Shimony JS, Miller-Thomas
6–7. M, Leuthardt EC. Magnetic resonance imaging-
16. Norred SE, Johnson JA. Magnetic resonance-guided guided focused laser interstitial thermal therapy
laser induced thermal therapy for glioblastoma multi- for intracranial lesions: single-institution series.
forme: a review. Biomed Res Int. 2014;2014:761312. Neurosurgery. 2013;73(6):1007–17. https://doi.
https://doi.org/10.1155/2014/761312. org/10.1227/NEU.0000000000000144.
17. Missios S, Bekelis K, Barnett GH. Renaissance of 29. Fabiano AJ, Qiu J. Delayed failure of laser-induced
laser interstitial thermal ablation. Neurosurg Focus. interstitial thermotherapy for postradiosurgery brain
2015;38(3):E13. https://doi.org/10.3171/2014.12. metastases. World Neurosurg. 2014;82(3–4):e559–
Focus14762. 63. https://doi.org/10.1016/j.wneu.2014.06.007.
18. Wellmer J, Kopitzki K, Voges J. Lesion focused 30. Ali MA, Carroll KT, Rennert RC, Hamelin T, Chang

stereotactic thermo-coagulation of focal cortical L, Lemkuil BP, et al. Stereotactic laser ablation as
dysplasia IIB: a new approach to epilepsy surgery? treatment for brain metastases that recur after stereo-
Seizure. 2014;23(6):475–8. https://doi.org/10.1016/j. tactic radiosurgery: a multiinstitutional experience. J
seizure.2014.01.024. Neurosurg. 2016;41(4):E11. https://doi.org/10.3171/2
19. Mohammadi AM, Schroeder JL. Laser interstitial
016.7.Focus16227.
thermal therapy in treatment of brain tumors – the 31. Eichberg DG, VanDenBerg R, Komotar RJ, Ivan
NeuroBlate System. Expert Rev Med Devices. ME. Quantitative volumetric analysis following
2014;11(2):109–19. https://doi.org/10.1586/1743444 magnetic resonance-guided laser interstitial thermal
0.2014.882225. ablation of cerebellar metastases. World Neurosurg.
20. Pearce J, Thomsen S. Rate process analysis of ther- 2018;110:e755–e65. https://doi.org/10.1016/j.
mal damage. In: Welch AJ, Van Gemert MJC, editors. wneu.2017.11.098.
Optical-thermal response of laser-irradiated tissue. 32. Borghei-Razavi H, Koech H, Sharma M, Krivosheya
Boston: Springer; 1995. p. 561–606. https://doi. D, Lee BS, Barnett GH, et al. Laser interstitial thermal
org/10.1007/978-1-4757-6092-7_17. therapy for posterior fossa lesions: an initial experi-
21. Schober R, Bettag M, Sabel M, Ulrich F, Hessel
ence. World Neurosurg. 2018;117:e146–e53. https://
S. Fine structure of zonal changes in experimen- doi.org/10.1016/j.wneu.2018.05.217.
tal Nd:YAG laser-induced interstitial hyperthermia. 33. Beechar VB, Prabhu SS, Bastos D, Weinberg JS,
Lasers Surg Med. 1993;13(2):234–41. Stafford RJ, Fuentes D, et al. Volumetric response of
22. Carpentier A, McNichols RJ, Stafford RJ, Itzcovitz progressing post-SRS lesions treated with laser inter-
J, Guichard J-P, Reizine D, et al. Real-time mag- stitial thermal therapy. J Neurooncol. 2018;137(1):57–
netic resonance-guided laser thermal therapy for 65. https://doi.org/10.1007/s11060-017-2694-3.
focal metastatic brain tumors. Oper Neurosurg. 34. Ahluwalia M, Barnett GH, Deng D, Tatter SB, Laxton
2008;63(suppl_1):ONS21–ONS9. https://doi. AW, Mohammadi AM, et al. Laser ablation after ste-
org/10.1227/01.NEU.0000311254.63848.72. reotactic radiosurgery: a multicenter p rospective study
23. Yeniaras E, Fuentes DT, Fahrenholtz SJ, Weinberg in patients with metastatic brain tumors and radiation
JS, Maier F, Hazle JD, et al. Design and initial evalu- necrosis. J Neurosurg. 2018;130:1–8. https://doi.
ation of a treatment planning software system for org/10.3171/2017.11.Jns171273.
MRI-guided laser ablation in the brain. Int J Comput 35. Hernandez RN, Carminucci A, Patel P, Hargreaves
Assist Radiol Surg. 2014;9(4):659–67. https://doi. EL, Danish SF. Magnetic resonance-guided laser-
org/10.1007/s11548-013-0948-x. induced thermal therapy for the treatment of progres-
sive enhancing inflammatory reactions following 48. Matuschek C, Bolke E, Nawatny J, Hoffmann TK,
stereotactic radiosurgery, or PEIRs, for metastatic Peiper M, Orth K, et al. Bevacizumab as a treat-
brain disease. Neurosurgery. 2018;85:84. https://doi. ment option for radiation-induced cerebral necrosis.
org/10.1093/neuros/nyy220. Strahlenther Onkol. 2011;187(2):135–9. https://doi.
36. Ruben JD, Dally M, Bailey M, Smith R, McLean CA, org/10.1007/s00066-010-2184-4.
Fedele P. Cerebral radiation necrosis: incidence, out- 49. Bui QC, Lieber M, Withers HR, Corson K, van

comes, and risk factors with emphasis on radiation Rijnsoever M, Elsaleh H. The efficacy of hyper-
parameters and chemotherapy. Int J Radiat Oncol Biol baric oxygen therapy in the treatment of radiation-
Phys. 2006;65(2):499–508. https://doi.org/10.1016/j. induced late side effects. Int J Radiat Oncol Biol
ijrobp.2005.12.002. Phys. 2004;60(3):871–8. https://doi.org/10.1016/j.
37. Panagiotakos G, Alshamy G, Chan B, Abrams R,
ijrobp.2004.04.019.
Greenberg E, Saxena A, et al. Long-term impact of 50. Rahmathulla G, Recinos PF, Valerio JE, Chao S,

radiation on the stem cell and oligodendrocyte precur- Barnett GH. Laser interstitial thermal therapy for focal
sors in the brain. PLoS One. 2007;2(7):e588. https:// cerebral radiation necrosis: a case report and literature
doi.org/10.1371/journal.pone.0000588. review. Stereotact Funct Neurosurg. 2012;90(3):192–
38. Xiang-Pan L, Yuxin C, Xiao-Fei W, Na L, Tang-
200. https://doi.org/10.1159/000338251.
Peng X, Xiao-Tao X, et al. Bevacizumab alleviates 51. Torres-Reveron J, Tomasiewicz HC, Shetty A,

radiation-
induced brain necrosis: a report of four Amankulor NM, Chiang VL. Stereotactic laser
cases. J Cancer Res Ther. 2015;11(2):485–7. https:// induced thermotherapy (LITT): a novel treatment
doi.org/10.4103/0973-1482.140782. for brain lesions regrowing after radiosurgery. J
39. Kureshi SA, Hofman FM, Schneider JH, Chin
Neurooncol. 2013;113(3):495–503. https://doi.
LS, Apuzzo ML, Hinton DR. Cytokine expres- org/10.1007/s11060-013-1142-2.
sion in radiation- induced delayed cerebral injury. 52. Rao MS, Hargreaves EL, Khan AJ, Haffty BG,

Neurosurgery. 1994;35(5):822–9; discussion 9–30. Danish SF. Magnetic resonance-guided laser abla-
40. Soussain C, Ricard D, Fike JR, Mazeron JJ, Psimaras tion improves local control for postradiosurgery
D, Delattre JY. CNS complications of radiotherapy recurrence and/or radiation necrosis. Neurosurgery.
and chemotherapy. Lancet. 2009;374(9701):1639–51. 2014;74(6):658–67;. discussion 67. https://doi.
https://doi.org/10.1016/s0140-6736(09)61299-x. org/10.1227/neu.0000000000000332.
41. Giglio P, Gilbert MR. Cerebral radiation necro-
53. Smith CJ, Myers CS, Chapple KM, Smith KA. Long-
sis. Neurologist. 2003;9(4):180–8. https://doi. term follow-up of 25 cases of biopsy-proven radiation
org/10.1097/01.nrl.0000080951.78533.c4. necrosis or post-radiation treatment effect treated with
42. Kumar AJ, Leeds NE, Fuller GN, Van Tassel P,
magnetic resonance-guided laser interstitial thermal
Maor MH, Sawaya RE, et al. Malignant gliomas: therapy. Neurosurgery. 2016;79(Suppl 1):S59–s72.
MR imaging spectrum of radiation therapy- and https://doi.org/10.1227/neu.0000000000001438.
chemotherapy- induced necrosis of the brain after 54. Rammo R, Asmaro K, Schultz L, Scarpace L,

treatment. Radiology. 2000;217(2):377–84. https:// Siddiqui S, Walbert T, et al. The safety of magnetic
doi.org/10.1148/radiology.217.2.r00nv36377. resonance imaging-guided laser interstitial thermal
43. Glantz MJ, Burger PC, Friedman AH, Radtke RA, therapy for cerebral radiation necrosis. J Neurooncol.
Massey EW, Schold SC Jr. Treatment of radiation- 2018;138(3):609–17. https://doi.org/10.1007/
induced nervous system injury with heparin and war- s11060-018-2828-2.
farin. Neurology. 1994;44(11):2020–7. 55. Kahn T, Bettag M, Ulrich F, Schwarzmaier HJ,

44. Levin VA, Bidaut L, Hou P, Kumar AJ, Wefel JS, Schober R, Fürst G, et al. MRI-guided laser-induced
Bekele BN, et al. Randomized double-blind placebo- interstitial thermotherapy of cerebral neoplasms. J
controlled trial of bevacizumab therapy for radiation Comput Assist Tomogr. 1994;18(4):519–32.
necrosis of the central nervous system. Int J Radiat 56. Leonardi MA, Lumenta CB, Gumprecht HK, von

Oncol Biol Phys. 2011;79(5):1487–95. https://doi. Einsiedel GH, Wilhelm T. Stereotactic guided laser-
org/10.1016/j.ijrobp.2009.12.061. induced interstitial thermotherapy (SLITT) in gliomas
45. Wong ET, Huberman M, Lu XQ, Mahadevan
with intraoperative morphologic monitoring in an open
A. Bevacizumab reverses cerebral radiation necro- MR-unit. Minim Invasive Neurosurg. 2001;44(1):37–
sis. J Clin Oncol. 2008;26(34):5649–50. https://doi. 42. https://doi.org/10.1055/s-2001-13581.
org/10.1200/jco.2008.19.1866. 57. Schwarzmaier HJ, Eickmeyer F, von Tempelhoff W,
46. Gonzalez J, Kumar AJ, Conrad CA, Levin VA. Effect Fiedler VU, Niehoff H, Ulrich SD, et al. MR-guided
of bevacizumab on radiation necrosis of the brain. Int laser-induced interstitial thermotherapy of recurrent
J Radiat Oncol Biol Phys. 2007;67(2):323–6. https:// glioblastoma multiforme: preliminary results in 16
doi.org/10.1016/j.ijrobp.2006.10.010. patients. Eur J Radiol. 2006;59(2):208–15. https://doi.
47. Liu AK, Macy ME, Foreman NK. Bevacizumab
org/10.1016/j.ejrad.2006.05.010.
as therapy for radiation necrosis in four children 58. Carpentier A, Chauvet D, Reina V, Beccaria K,

with pontine gliomas. Int J Radiat Oncol Biol Phys. Leclerq D, McNichols RJ, et al. MR-guided laser-
2009;75(4):1148–54. https://doi.org/10.1016/j. induced thermal therapy (LITT) for recurrent glio-
ijrobp.2008.12.032. blastomas. Lasers Surg Med. 2012;44(5):361–8.
https://doi.org/10.1002/lsm.22025.
59. Jethwa PR, Barrese JC, Gowda A, Shetty A, Danish NeuroBlate System first-in-humans Phase I clini-
SF. Magnetic resonance thermometry-guided laser- cal trial for recurrent glioblastoma: clinical article. J
induced thermal therapy for intracranial neoplasms: Neurosurg. 2013;118(6):1202–19. https://doi.org/10.
initial experience. Neurosurgery. 2012;71(1 Suppl 3171/2013.1.Jns1291.
Operative):133–44; 44–5. https://doi.org/10.1227/ 61. Thomas JG, Rao G, Kew Y, Prabhu SS. Laser inter-
NEU.0b013e31826101d4. stitial thermal therapy for newly diagnosed and recur-
60. Sloan AE, Ahluwalia MS, Valerio-Pascua J, Manjila rent glioblastoma. Neurosurg Focus. 2016;41(4):E12.
S, Torchia MG, Jones SE, et al. Results of the https://doi.org/10.3171/2016.7.Focus16234.
Preventing Cranial Wound
Complications in Cancer Patients
33
James C. Lee, Jimmy Xia, Rohan Ramakrishna,
and David M. Otterburn
Introduction Anatomy
Patients undergoing resection for intracranial neu- Anatomic Layers

rosurgical tumors pose unique issues for the
reconstructive team, and consideration needs to The scalp consists of five anatomic layers
be given for not only the initial resection but also which are often described by the mnemonic
for possible future resections. A strong under- “SCALP” (Fig. 33.1). Starting from the most
standing of the anatomy of the overlying soft tis- superficial layer to the deepest layer, the five
sue is important to prevent ischemic complications. layers consist of:
Patients undergoing re-operative cases are at
higher risk for infections, wound dehiscence, and • Skin
skin necrosis which all stem from decreased blood • Connective tissue
flow and tension in the scalp from prior scarring. • Galea aponeurotica
The combination of poor nutrition, immunosup- • Loose areolar tissue
pressive agents, anti-angiogenic agents, and radi- • Pericranium
ation all pose specific risks to the postoperative
patient which needs to be considered during oper- The skin is connected to the galea aponeuro-
ative planning. In this chapter, we will discuss tica through a network of tight connective tissue
these issues and highlight how to minimize tissue bands, allowing the galea aponeurotica and skin
ischemia with appropriate planning of incisions to move as a single unit during surgical manipu-
through assessment of scalp perfusion. We will lation. This subcutaneous connective tissue layer
also discuss the management of the patient in the also contains much of neurovascular structures,
immediate postoperative period. and dissection in this plane can result in signifi-
cant bleeding. The galea aponeurotica is a fibrous
layer continuous with the superficial musculo-
J. C. Lee · J. Xia · D. M. Otterburn (*) aponeurotic system (SMAS) of the face and is
Department of Plastic Surgery, NewYork-Presbyterian/
Weill Cornell Medicine,
controlled by the frontalis muscle anteriorly and
New York, NY, USA the occipitalis posteriorly. Under the galea
e-mail: dmo9004@med.cornell.edu aponeurotica is the loose areolar tissue, which
R. Ramakrishna allows for movement of the galea over the peri-
Department of Neurological Surgery, cranium. The loose areolar tissue is a relatively
NewYork-Presbyterian Hospital, Weill Cornell avascular plane that allows for easy dissection
Medicine, New York, NY, USA

https://doi.org/10.1007/978-3-030-42958-4_33
474 J. C. Lee et al.
Fig. 33.1 The Epicranial Aponeurosis

anatomical layers of the
scalp Connective tissue
Loose areolar connective tissue
Skin
Periosteum
during surgical exposure of the pericranium with branches. The medial branch of the supraorbital
minimal bleeding. The pericranium is the perios- nerve enters the corrugator supercilii and fronta-
teum of the skull that separates relatively easily lis muscles, while the lateral branch enters the
from the underlying bone except at the cranial galea aponeurotica. Posteriorly, the greater
sutures. The pericranium provides nutrition to the occipital nerve, originating from the C2 spinal
skull and can be elevated as a flap for coverage nerve, provides innervation from the occiput to
and lining. the vertex. The lesser occipital nerve originates
from C2 and C3 spinal nerves and innervates the
region of the scalp posterior to the ear. The auric-
Innervation ulotemporal nerve, a branch of the mandibular
division of the trigeminal nerve, innervates the
Anteriorly, the scalp is innervated by the supra- tragus, the area anterior to the ear, and the poste-
trochlear and supraorbital nerves, both of which rior portion of the temple region. In the temporal
are derived from the ophthalmic division of the scalp region, special attention should be paid to
trigeminal nerve. The supratrochlear nerve inner- the frontal branch of the facial nerve as it passes
vates the lower part of the forehead, traveling cephalad over the zygomatic arch, running just
beneath the frontalis as it ascends. The supraor- deep to the superficial temporal fascia as it inner-
bital nerve originates from the supraorbital notch vates the frontalis, orbicularis oculi, corrugator
or foramen and terminates in medial and lateral supercilii, and auriculares anterior and superior.
33 Preventing Cranial Wound Complications in Cancer Patients 475
Blood Supply with the supratrochlear and supraorbital vessels

anteriorly and the posterior auricular and occipi-
Understanding the rich blood supply of the scalp tal vessels posteriorly. The occipital artery is the
is crucial to avoiding wound complications dur- main blood supply of the posterior scalp, and it
ing cranial surgery. The five main paired arteries also arises from the external carotid artery sys-
that supply the scalp come together in a rich arte- tem. It runs deep to the neck muscles posteriorly
rial network that runs throughout the subcutane- from the external carotid artery before turning
ous connective tissue layer (Fig. 33.2). From the cephalad from the posterior scalp up to the ver-
ophthalmic branch of the internal carotid artery, tex. The posterior auricular artery, the smallest of
the supratrochlear and supraorbital arteries arise the scalp arteries, branches off of the external
from the superior orbital rim and supply much of carotid artery and provides bloody supply to the
the anterior scalp. The supratrochlear artery pro- posterior ear and mastoid region.
vides much of the blood supply to the midline The venous drainage pattern of the scalp fol-
forehead, while the supraorbital artery reaches as lows veins that run with the arterial blood supply.
far up as the vertex of the scalp. Laterally, the Additionally, emissary veins on the cranium also
superficial temporal artery arises from the exter- contribute by draining blood into the dural
nal carotid artery and runs anterior to the ear sinuses. Lymphatic vessels also run in the subcu-
before splitting into frontal and parietal branches. taneous connective tissue layer, draining into
The superficial temporal artery is typically the parotid, preauricular, postauricular, upper cervi-
largest of the scalp vessels, and it anastomoses cal, and occipital lymph node basins.
Suprathrochlear artery
Suprathrochlear nerve (VI)
Supraorbital artery
Supraorbital nerve (VI)
Zygomaticotemporal nerve (V2)

Superficial temporal artery
Auriculotemporal nerve (V3)
Lesser occipital nerve (V3)
Posterior auricular artery
Greater occipital nerve (C2, C3)
Occipital nerve (C3) Occipital artery
Fig. 33.2 The five main paired arteries of the scalp (right) and the neural innervation of the scalp (left)
Risk Factors for Complications collagen production [2]. It is also recommended

that albumin and prealbumin levels are within
Medical History normal levels prior to surgical intervention, and
supplemental nutrition should be provided as
Several patient-specific factors can put a patient needed to help patients reach their nutritional
at elevated risk for scalp complications postop- goals. Enteral nutrition is preferred over paren-
eratively. A comprehensive preoperative history teral nutrition, as enteral nutrition is more effi-
and physical is essential to note and control for cient, has fewer metabolic complications, costs
these relevant factors. Factors such as smoking less, and helps promote the growth and develop-
history and diabetes mellitus can affect the vas- ment of gastrointestinal mucosal tissue. When
cularity of the scalp and increase the risk of enteral nutrition is not possible, parenteral nutri-
wound complications. Nicotine in smoke acts as tion should be considered in malnourished
a vasoconstrictor while also increasing platelet patients. Regardless of nutritional route, all
adhesiveness [1]. Carbon monoxide reduces oxy- patients should be offered adequate nutritional
gen transport and hydrogen cyanide impairs oxi- support prior to surgical intervention in order to
dative metabolism. These substances in cigarette avoid wound complications in this high-risk
smoke can lead to tissue ischemia, thrombotic population.
microvascular occlusion, and impaired healing
[1]. Similarly, diabetes mellitus contributes to
poor wound healing in multiple ways, including Perfusion
microvascular ischemia, impaired immune func-
tion, decreased growth factor production, and Tissue perfusion is essential to wound healing,
reduced fibroblast proliferation. and any conditions that may impair adequate per-
Nutrition is also important to take into confusion of the scalp can increase the risk of cranial
sideration, especially in cancer patients who wound complications. During incision planning,
may be cachectic or have poor oral intake. previous scars should be taken into consideration,
Malnutrition has been well-documented as a risk and adequate inflow from at least one of the five
factor for poor wound healing, infectious com- major blood supplies should be preserved. The
plications, and other sources of operative mor- vascular network of the scalp runs in the subcuta-
bidity and mortality. In critically ill patients or neous connective tissue layer superficial to the
patients under stress, basal energy expenditures galea aponeurotica so great care should be taken
and caloric requirements are increased, necessi- when performing galeal scoring maneuvers.
tating a more aggressive approach to maintain- Adequate blood pressure and hemoglobin should
ing proper nutrition. Proper nutrition includes be maintained in the perioperative period to
not only adequate intake of all macronutrients ensure sufficient perfusion and oxygenation for
such as carbohydrates, fats, proteins, and fluids wound healing. Postoperative dressings should
but also a sufficient supply of micronutrients [2]. allow for expected edema and not be tight enough
These micronutrients include amino acids; vita- to restrict blood flow to the surgical site.
mins A, C, D, and E; and minerals such as zinc,
selenium, and iron. Vitamin C is well-known for
its role in collagen polymerization and cross- Tension
linking. Vitamin A is an important cofactor in the
inflammatory phase of wound healing, promot- Care should be taken to avoid tension over the
ing phagocytic activity and immunologic func- scalp closure in order to prevent wound compli-
tion. Although the mechanisms of vitamin E and cations such as dehiscence, skin necrosis, and
zinc are not as well-defined in the literature, infection. Experimental studies have also dem-
there are data that support their importance in onstrated increased incidence of hypertrophic
scarring and scar widening when wounds are ation therapy is often preferred to allow for a
subjected to excessive tension during the early period of healing prior to initiating the negative
wound healing period [3]. Techniques to avoid effects of radiation therapy on wound healing
tension during incision closure include wide [6]. Clinical studies have demonstrated lower
undermining of the scalp in the loose areolar tis- rates of wound complications when postopera-
sue layer, performing galeal scoring maneuvers, tive radiotherapy is used, and this may be an
and utilizing local flaps as needed. Scalp tissue important consideration when recurrence rates
should be closed in a layered fashion with are similar with preoperative and postoperative
sutures in the galea aponeurotica offloading radiation therapy [5, 6]. Some of these concerns
most of the tension from the cutaneous closure can be mitigated with radiosurgical techniques
[4]. Tissue expansion can sometimes be used either in the pre- or postoperative setting. Given
prior to oncologic resection if there is an antici- the high conformality and ability to limit scalp
pated deficit of scalp tissue and sufficient lead dose, the concerns related to wound healing are
time prior to surgery. minimized. In our practice, we are comfortable
with either preoperative or rapid postoperative
radiosurgery given the advantages conferred by
Radiation this radiation technique.
The negative effects of radiation on wound heal-

ing have been well-documented in the literature. Chemotherapy
The inflammatory phase of wound healing, char-
acterized by the infiltration of macrophages and Similar to radiation therapy, chemotherapy is an
neutrophils, is delayed and inhibited in irradi- important component of cancer treatment but can
ated tissues. The formation of granulation tissue negatively impact wound healing via several
is also slowed as fibroblast activity and collagen mechanisms. The effects of chemotherapeutic
formation are reduced. Lastly, epithelialization agents are linked to their ability to impair DNA
in irradiated tissue is delayed and the overall replication, interfere with metabolic processes,
healing time of wounds is prolonged. The effect and prevent cell division. While these effects dis-
of external beam radiation on the scalp is charac- proportionately impact rapidly growing tissues
terized by early skin changes followed by such as cancer cells, they can also impact immune
chronic damage long after radiation therapy has cells, epithelial tissue, neovascularization, and
been completed. Acute findings include skin ery- fibroblasts that are important in the wound heal-
thema, tenderness, warmth, epidermolysis, and ing process.
ulceration. These effects are often dose-depen- Alkylating agents such as cyclophosphamide
dent and reversible. Long-term effects of radia- at high doses have been shown to increase wound
tion can include tissue fibrosis, sebaceous gland complications by impairing neovascularization
dysfunction, loss of hair follicles, microvascular during the proliferative phase of wound healing.
compromise, skin necrosis, and secondary carci- Thiotepa and mechlorethamine have also been
nogenesis. These effects are often irreversible demonstrated to impair wound healing in animal
and result in a higher risk of delayed wound models by inhibiting fibroblast function and col-
healing, infection, hardware exposure, skin lagen production. Cisplatin has also been proven
necrosis, and flap failure [5]. in multiple animal studies to decrease wound
Preoperative radiation, when indicated, healing by impairing fibroblast proliferation,
should be performed at least 3–6 weeks prior to inhibiting neovascularization, and reducing con-
surgery in order to avoid wound complications. nective tissue proliferation [6].
This is especially important when doses larger Chemotherapeutic antibiotics such as bleomy-
than 50 Gy are administered. Postoperative radi- cin, doxorubicin, and mitomycin C have also
been found to have an impact on wound healing better define the impact of vitamin A co-
in animal models. Bleomycin inhibits the pro- administration [6].
duction of collagen by fibroblasts, thus decreas- Given the variability of chemotherapeutic
ing wound tensile strength postoperatively. agents and their effect on wound healing, it is
Doxorubicin also interferes with DNA important to keep the timing and dosing of che-
transcription and has been found to decrease
motherapy in mind when considering surgical
wound tensile strength in animal models. intervention. If possible, delaying the initiation
Mitomycin C, though most often used topically, of chemotherapy in surgical patients for
has also been demonstrated to have a negative 7–10 days may decrease the risk of wound com-
impact on wound healing in rat models [6]. plications in this population. Furthermore, it is
The use of antimetabolites such as methotrex- important to ensure patients are not neutropenic
ate and 5-fluorouracil at higher doses has also prior to surgery. Careful consideration should be
demonstrated some decreased wound tensile taken to control for other wound healing risk fac-
strength in animal models. The effects of azathio- tors before surgical intervention.
prine and 6-mercaptopurine on wound healing
are still unclear and require further study.
Similarly, plant alkaloids such as vincristine and Incision Planning
vinblastine have shown mixed results on wound
tensile strength in animal studies [6]. Careful consideration should be taken when
Anti-angiogenesis agents such as bevaci- planning cranial incisions in order to minimize
zumab provide a unique challenge to the healing the risk of postoperative wound complications.
wound. Bevacizumab is a monoclonal antibody Incisions should be selected in a fashion that
which targets VEGF, preventing neovasculariza- would allow for wide exposure of the target sur-
tion. It has been widely used in multiple cancers, gical site as well as flexibility to extend the inci-
including neurological cancers. It has a more sion for subsequent surgeries if needed. With
direct effect on the healing wound than any other cranial surgery, the incision choice should reflect
agent currently in use and has been shown to the goals of surgery and potential for future sur-
cause wound dehiscence, hematomas, and wound geries in that patient. For example, patients with
infection. As the half-life is 20 days, recommen- gliomas often recur within 2 cm of a previous
dations in the literature include waiting 6 weeks resection cavity. As such, incisions should reflect
after the last therapy prior to surgical intervention an understanding of possible future recurrence
[7]. Patients should be counseled that the rate of such that the same incision can be used or easily
wound complications following bevacizumab modified in the future without causing vascular
therapy is considerable, particularly if the wound compromise to the scalp. When cranial hardware
has been previously irradiated. is used, we try to limit the amount of hardware
Corticosteroids, while not necessarily consid- directly underneath the incision. We have found
ered a chemotherapeutic agent, are also often this technique helps avoid delayed hardware
used in cancer patients to alleviate pain and exposure, particularly in patients with atrophic
inflammation. Steroids are well-known for hav- scalp tissue or those that subsequently undergo
ing deleterious effects on wound healing, and scalp irradiation.
studies have shown increased rates of wound If pre-existing surgical scars are present, an
complications and dehiscence in patients on cor- attempt should be made to incorporate those
ticosteroid therapy in the perioperative period. scars in the new incision to avoid leaving bridges
The administration of vitamin A has been shown of devascularized scalp tissue. Regardless of inci-
to mitigate some of the negative effects on wound sion design, all remaining segments of scalp tis-
healing, although further studies are needed to sue once old scars and new incisions are taken
into account must be contiguous with at least one lateral anterior skull base and middle cranial
of the five main paired arteries (supratrochlear, fossa, although it can be modified to reach the
supraorbital, superficial temporal, posterior posterior cranium, as well. The anterior flap is
auricular, and occipital) in order to survive. All most often supplied by the superficial temporal
attempts should be made to avoid acute angles artery although the ipsilateral supraorbital or
between incisions as that often leads to devascu- supratrochlear arteries may contribute depend-
larized distal segments of the scalp. New inci- ing on the design of the incision. The posterior
sions can either be an extension of an old incision flap remains vascularized on the posterior auric-
or take off at a 90° angle from an existing scar ular and occipital arteries. The “question mark”
(Fig. 33.3). incision limits access to the contralateral hemi-
Traditional incisions such as the coronal or sphere and posterior cranium. If exposure of the
bitemporal incision allow for wide access to the contralateral anterior cranial vault is needed, a
anterior cranial vault, forehead, and facial skele- contralateral “question mark” incision can be
ton. The coronal incision can be reopened multi- made with the midline scalp preserved as a
ple times to allow for repeated exposure in the bipedicle flap. If access to the posterior cranium
case of recurrent disease or complication. In the is needed, a sagittal incision can be made per-
coronal approach, the anterior flap is vascular- pendicular to the curve of the “question mark”
ized by the supraorbital, supratrochlear, and and extended posteriorly, similar to Kempke’s
superficial temporal arteries, while the posterior “T-bar” incision.
flap is supplied by the posterior auricular and The midline posterior skull base approach
occipital arteries. If access to the posterior cra- allows access to the posterior cranium and expo-
nium is needed, the coronal incision can be sure for the classical suboccipital craniotomy. In
extended with a midline sagittal incision oriented this incision, all major scalp arteries are pre-
perpendicular to the coronal incision. served; however, blood flow across the midline is
The lateral skull base approach with the Al disrupted in the posterior scalp. The midline pos-
Kayat and Bramley modification of the preau- terior scalp incision allows for much flexibility in
ricular incision, often referred to as the “ques- extending the incision anteriorly as needed to
tion mark” incision, is often used to access the gain further exposure. This incision can also be
Question Mark
Coronal Incision Posterior Incision
Incision
Fig. 33.3 Common neurosurgical incisions (1) and example extensions to avoid wound complications (2 and 3)
converted to a “T-bar” incision if needed. Access replaced with well-vascularized adjacent tissue
to the anterior cranial vault can also be achieved or soft tissue flaps as needed [9].
through separate incisions using the traditional
approaches described above.
Delayed Flaps
Prevention of Complications When only a few weeks of lead time are avail-
able, flap delay is a technique that can be used to
Tissue Expansion maximize the survival of anticipated scalp flaps.
The delay phenomenon, also known as ischemic
In patients who are high risk for postoperative preconditioning, involves partially disrupting the
cranial wound complications, several techniques vascular supply to a flap at the anticipated inci-
can be used to maximize the chance of a success- sion sites a few days or weeks prior to transfer of
ful reconstruction. When there is an existing skin the flap. This allows for the opening of choke
deficit or resection of large portions of scalp tis- vessels in the remaining flap pedicle, propagation
sue is anticipated, preoperative tissue expansion of collateral circulation, and increased tolerance
can be utilized to increase the surface area of to ischemia that can improve the survivability of
scalp tissue available and reduce tension on the the flap after transfer. This technique is useful in
closure. Up to 50% of the scalp can be recon- patients with a history of multiple cranial opera-
structed with expanded scalp tissue, providing tions with high-risk incisions that may benefit
stable, potentially hair-bearing soft tissue cover- from concomitant scalp flap reconstruction after
age over cranial hardware or compromised bone. intracranial surgery [8].
Tissue expansion, however, requires that there is
adequate lead time prior to the procedure to place
tissue expander devices and inflate them. Other Therapies
Furthermore, tissue expansion requires an experi-
enced surgeon as it can be associated with com- Other modalities have also been described in the
plication rates as high as 25% [8]. literature to salvage compromised scalp tissue.
Hyperbaric oxygen therapy is a treatment that
utilizes 100% oxygen at pressures greater than
Laser Angiography atmospheric pressure in order to raise tissue oxy-
genation levels. Some studies have suggested that
Intraoperatively, laser angiography can be uti- elevated tissue oxygen levels may improve the
lized to assess the viability of scalp tissue and healing and oxygen-dependent antibiosis of cer-
prevent potential wound complications. Modern tain wounds such as delayed radiation injuries,
laser angiography technology uses indocyanine burns, compromised flaps, diabetic ulcers, and
green as a fluorescence agent to provide real- soft tissue infections. The efficacy of hyperbaric
time assessment of tissue perfusion. Indocyanine oxygen therapy in salvaging ischemic scalp flaps,
green binds to plasma proteins and has a half- however, is still unproven in the literature as the
life of 3–5 minutes. It is administered intrave- evidence consists of mostly case reports [10].
nously and excreted by the liver into the bile so Nitroglycerin ointment has also been described
there is no risk for nephrotoxicity. Furthermore, as a therapy to help salvage ischemic skin flaps.
indocyanine green fluorescence is viewed using Topical application of nitroglycerin has been
a laser diode array emitting light in the near- shown to increase local blood flow to the skin by
infrared wavelength so no protective eyewear is acting as a vasodilator for both arteries and veins.
needed and no harmful radiation is produced. Early studies have demonstrated benefits of 2%
Areas of scalp demonstrating poor tissue perfu- nitroglycerin ointment in the healing of anal fis-
sion on laser angiography should be excised and sures, pressure sores, and peripheral tissue
ischemia in neonates. In the plastic surgical lit- sure, and infection of the underlying contents. An
erature, nitroglycerin paste has been shown to important principle of cranial wound reconstruc-
decrease mastectomy skin flap necrosis in pro- tion is that nothing replaces scalp tissue as well
spective and randomized controlled trials with no as scalp tissue. When possible, reconstruction
increase in complication rates [11]. Although not should strive for a cosmetically appealing result
specifically studied in ischemic scalp tissue, in addition to achieving coverage by restoring
nitroglycerin ointment can be considered as a normal anatomy and paying close attention to
therapeutic option to help salvage or limit skin hair growth patterns and hairlines. Small deficits
necrosis in compromised cranial closures [12]. of scalp tissue can potentially be addressed
through undermining of adjacent scalp in the
loose areolar plane and performing galeal scoring
Management of Complications techniques. When larger skin deficits are present,
scalp rotation flaps or transposition flaps can be
Scalp Reconstruction Algorithm utilized to recruit tissue from areas of laxity
(Fig. 33.4). When needed, large rotation flaps can
When cranial wound complications do occur, be transferred to the area of concern, and skin
reconstruction is often indicated to prevent desic- grafting can be performed over the donor site to
cation of the bone, osteomyelitis, hardware expo- achieve greater coverage (Fig. 33.5). Skin graft-
Rhomboid
Rotation Flap Double Rotation Flap Flap
Fig. 33.4 Common local scalp flaps used to reconstruct cranial soft tissue defects
Orticochea Flap
Fig. 33.5 Large scalp flaps can be designed to reconstruct bigger cranial soft tissue defects
ing can also provide permanent or temporary controlled trials, but some studies have suggested
coverage over areas of the scalp with intact peri- that the risk of postoperative infection is lower
osteum. If no periosteum is present and a pericra- with autologous bone reconstructions [13].
nial flap is not available, the outer table of the Autologous bone, however, is subject to bone
cranium can be burred down to the diploic space resorption and is not immune to infection of the
in order to accept a skin graft. This serves as a devitalized bone segment. Methyl methacrylate is
viable option for immediate coverage of a scalp a popular synthetic material used in cranioplasty
defect. In the long run, hair-bearing coverage of as it is malleable, lightweight, and radiolucent. It
up to 50% of the scalp can be achieved with tis- is often in conjunction with titanium mesh to pro-
sue expansion in the subgaleal plane. Tissue vide enhanced structural support. Hydroxyapatite,
expansion requires staged operations with a form of calcium phosphorus naturally present in
lengthy interval periods and complication rates bone tissue, is also frequently used in cranio-
varying from 6% to 25%. However, oftentimes plasty. The advantage of hydroxyapatite is its
the best aesthetic results can only be achieved strong osteointegrative ability, minimal tissue
with this technique. When even larger defects are reaction, and enhanced bone healing. Its biggest
present or if local tissue quality is poor due to disadvantage, however, is its propensity to break
radiation therapy, strong consideration should be under mechanical stress. As a result, hydroxyapa-
given to free tissue transfer as a reconstructive tite is also often used in conjunction with titanium
option [8]. mesh reconstruction. More recently, poly-
etheretherketone (PEEK) has become a popular
material used in cranioplasty, especially as a com-
Cranioplasty Materials puter-designed implant requiring little to no intra-
operative molding [14].
In some circumstances, patients may require a
cranioplasty either due to decompressive craniec-
tomy, resection of cranial bone, trauma, or prior Timing of Cranioplasty
surgical complication. Cranioplasty may be per-
formed with autologous bone graft, synthetic Regardless of material used, achieving stable soft
material, or biosynthetic material. The cranio- tissue coverage over the cranioplasty implant is
plasty material of choice is somewhat controver- of utmost importance. When there is concern
sial given the paucity of quality randomized, regarding the quality of soft tissue coverage or
potential contamination of the field, it is often allows dural defects to seal and thus obliterate the
best to delay the cranioplasty procedure and CSF leak. Finally, if a CSF collection persists
allow for complete healing of the surgical site despite these maneuvers, the patient should be
before introducing devitalized or synthetic mate- evaluated for either hydrocephalus or a meningi-
rial. Various interval periods have been advocated tis (possible aseptic). It is vitally important to
for in the literature, but no definitive period has ensure that scalp collections do not cause tension
been proven to be superior. Most studies on the wound as leaks through the wound will
recommend waiting anywhere from 6 weeks to increase the rate of postoperative infections.
3 months and as long as 6 months to 1 year if
there is any evidence of infection. Ultimately,
timing of cranioplasty is an individualized physi- Conclusion
cian choice that must account for infection risk
and wound healing capability. The soft tissue concerns for patients undergoing
cranial extirpative surgeries can be complex and
layered. Appropriate preoperative planning,
Use of Drains intraoperative decision-making, and postopera-
tive care can decrease perioperative morbidity.
Little reliable evidence exists in the literature Appropriate soft tissue coverage is necessary to
regarding the use of subgaleal drains to prevent allow for appropriate timing of radiation ther-
cranial wound complications. Some retrospective apy and adjuvant chemotherapy. In our practice,
studies have suggested that the use of scalp drains neurosurgeons routinely involve plastic sur-
significantly reduces the seroma rate in patients geons preoperatively in patients with high-risk
undergoing craniofacial surgery. Other studies wounds both for incisional planning and intra-
have noted subjective improvement in facial operative closure. This team approach has been
swelling and decreased length of stay with the highly effective at preventing postoperative
use of subgaleal drains, but those findings have wound complications.
not been corroborated in the literature [15]. Most
studies on drain use have not shown a statistically
significant effect on infection rate, hematoma
formation, transfusion requirement, or other
References
postoperative complications [16]. Although some 1. Silverstein P. Smoking and wound healing. Am J
have questioned whether there is an association Med. 1992;93(1A):22S–4S.
between drain use and infection risk, there is 2. Quain AM, Kardori NM. Nutrition in wound care
minimal supporting evidence that closed-suction management: a comprehensive overview. Wounds.
2015;27(12):327–35.
drain use increases the risk of surgical site infec- 3. Burgess LP, Morin GV, Rand M, Vossoughi J,
tion [17]. As a result, we recommend the judi- Hollinger DC. Wound healing relationship of wound
cious use of closed-suction drains in cases with closing tension to scar width in rats. Arch Otolaryngol
an elevated risk of postoperative seroma and Head Neck Surg. 1990;116(7):798–802.
4. Barnes LA, Clement DM, Leavitt T, Hu MS,
timely removal of drains once no longer needed. Moore AL, Gonzalez JG, Longaker MT, Gurtner
In cases where drains are aspirating the cerebro- GC. Mechanical forces in cutaneous wound healing:
spinal fluid (CSF) because of a defect in a dural emerging therapies to minimize scar formation. Adv
repair, it is important that these drains be removed Wound Care (New Rochelle). 2018;7(2):47–56.
5. Gu Q, Wang D, Cui C, Gao Y, Xia G, Cui X. Effects of
to prevent intracranial hypotension and subse- radiation on wound healing. J Environ Pathol Toxicol
quent subdural hematomas. Should a CSF collec- Oncol. 1998;17(2):117–23.
tion develop under the scalp, these can be 6. Payne WG, Naidu DK, Wheeler CK, Barkoe D, Mentis
percutaneously tapped and often they are self- M, Salas RE, Smith DJ Jr, Robson MC. Wound heal-
ing in patients with cancer. Eplasty. 2008;8:e9.
limiting. In cases of recalcitrant CSF collections, 7. Gordon CR, Rojavin Y, Patel M, Zins JE, Grana
lumbar drains combined with percutaneous aspi- G, Kann B, Simons R, Atabek U. A review on
ration can be used for several days which often bevacizumab and surgical wound healing: an
important warning to all surgeons. Ann Plast Surg. 12. Rohrich RJ, Cherry GW, Spira M. Enhancement of
2009;62(6):707–9. skin-flap survival using nitroglycerin ointment. Plast
8. Leedy JE, Janis JE, Rohrich RJ. Reconstruction of Reconstr Surg. 1984;73(6):943–8.
acquired scalp defects: an algorithmic approach. Plast 13. Chang V, Hartzfeld P, Langlois M, Mahmood A,

Reconstr Surg. 2005;116(4):54e–72e. Seyfried D. Outcomes of cranial repair after craniec-
9. Gurtner GC, Jones GE, Neligan PC, Newman MI, tomy. J Neurosurg. 2010;112(5):1120–4.
Phillips BT, Sacks JM, Zenn MR. Intraoperative 14. Aydin S, Kucukyuruk B, Abuzayed B, Aydin S, Sanus
laser angiography using the SPY system: review GZ. Cranioplasty: review of materials and techniques.
of the literature and recommendations for use. J Neurosci Rural Pract. 2011;2(2):162–7.
Ann Surg Innov Res. 2013;7(1):1. https://doi. 15. Vasudevan K, Oh A, Tubbs RS, Garcia D, Reisner A,
org/10.1186/1750-1164-7-1. Chern JJ. Jackson-Pratt drainage in pediatric craniofa-
10.
Mesfin FB, Burton MR, Ngnitewe RA, Litt cial reconstructive surgery: is it helping or hurting? J
JS. Hyperbaric oxygen therapy of ischemic cranial Neurosurg Pediatr. 2017;20(4):341–6.
skin flap: case report and review of the literature. Case 16. Tong JW, Emelin JK, Wong R, Meltzer HS, Cohen
Reports Clin Med. 2017;6(10):250–4. SR. Subgaleal drain placement improves surgical
11. Gdalevitch P, Van Laeken N, Bahng S, Ho A, Bovill outcomes after primary cranioplasty in craniosynos-
E, Lennox P, Brasher P, Macadam S. Effects of tosis patients. J Craniofac Surg. 2015;26(6):1963–6.
nitroglycerin ointment on mastectomy flap necro- 17. Reiffel AJ, Barie PS, Spector JA. A multi-disciplinary
sis in immediate breast reconstruction: a ran- review of the potential association between closed-
domized controlled trial. Plast Reconstr Surg. suction drains and surgical site infection. Surg Infect.
2015;135(6):1530–9. 2013;14(3):244–69.
Part V
Spinal Metastases: Foundations
Introduction to Spinal Metastases
34
Ibrahim Hussain, Brenton H. Pennicooke,
and Ali A. Baaj
Introduction excellent local control rates for the majority of

patients. Nonetheless, there continues to be a
The spine is the most common skeletal site for void in our understanding of genetic tropism for
metastatic disease and up to 10% of all patients certain primary tumors to metastasize to the spine
with cancer develop spinal metastases during the and what mutations portend a more favorable
course of their disease [1–3]. Most tumors spread prognosis based on available treatment options,
to the spine via hematogenous venous circula- which is an active area of research. Currently, the
tion; however, local invasion from close proxim- focus of management involves timely diagnosis,
ity tumors is also observed. In concordance with close observation, and treatment when radio-
relative bone mass, the thoracic spine is the most graphic findings or clinical symptoms become
common site for spread, followed by the lumbo- burdensome.
sacral and cervical spine, respectively. With the
aging population and robust development of sys-
temic treatment options for various cancers, the Categorization of Spinal Metastases
prevalence of spine metastases is likely to
increase over the coming years. While the vast Spinal metastases can be categorized based on
majority are confined to the bony elements of the various parameters which have implications on
spine, those with epidural extension or intradural management. One broad category involves pri-
location often require treatment to preserve neu- mary tumor histology and more specifically if the
rologic status and quality of life. As with sys- primary pathology is of solid tumor versus hema-
temic treatment options, various advances in tologic malignancy. Solid tumors that most com-
multi-modality of treatment of these tumors have monly metastasize to the spine are similar to their
accelerated over the past 20 years and resulted in prevalence in the general population. These
include breast, lung, prostate, colorectal, and
renal cell [2, 4, 5] (Fig. 34.1). Other common pri-
mary tumor pathologies that commonly metasta-
size to the spine include thyroid, melanoma,
sarcoma, gastrointestinal, and hepatocellular.
I. Hussain (*) · B. H. Pennicooke · A. A. Baaj These vast majority of these tumors spread to the
Department of Neurological Surgery, vertebral column and through epidural extension
Weill Cornell Medical College/New York cause mass effect on underlying neural struc-
Presbyterian Hospital, New York, NY, USA
tures. A subset of these tumors is particularly
e-mail: ibh9004@nyp.org

https://doi.org/10.1007/978-3-030-42958-4_34
488 I. Hussain et al.
a b c
Fig. 34.1 Examples of spinal metastases. (a) T1-post metastasis resulting in kyphotic deformity and severe cer-
contrast MRI demonstrating an L3 colorectal metastasis vical spinal cord compression. (c) CT scan of mid-thoracic
with pathological collapse and epidural extension result- prostate metastases demonstrating the blastic nature of
ing in severe cauda equine compression. (b) T2-weighted bony involvement
MRI demonstrating a C4 burst fracture due to a breast
notorious for being highly vascularized, such as exact location of the tumor in relation to the cen-
renal cell carcinoma, thyroid carcinoma, and tral nervous system. These include epidural,
paragangliomas. Preoperative embolization has intradural extramedullary, and intradural intra-
been shown to reduce intraoperative blood loss medullary. The latter two can further be classified
during resection of these tumors [6–11]. Of the with or without the presence of leptomeningeal
hematogenous malignancies, multiple myeloma disease. Location-wise, 98% of metastatic tumors
is the most common and should not to be con- are extradural, developing within the vertebral
fused with a solitary plasmacytoma which typi- column itself, including the body, pedicles, facet
cally does not require systemic treatment [12–14]. joints, or spinous processes [17, 18]. With epi-
Patients with metastatic involvement of multiple dural extension, these tumors extend beyond the
myeloma may demonstrate extensive lytic lesions confines of the cortical wall and can exhibit com-
and compression fractures which may benefit pression of the spinal cord or cauda equina.
from vertebroplasty/kyphoplasty or percutaneous Intradural metastatic tumors are rare. These
stabilization. Lymphoma and leukemia metasta- tumors are often dural-based lesions that exert
ses similarly can involve bone and have epidural mass effect on the spinal cord. Even rarer is iso-
extension. Even with high-grade compressive lated intradural intramedullary metastases, which
lesions on imaging, surgical intervention is only are reported in less than 1% of cases. Any intra-
required for acute neurologic deterioration when dural metastases can potentially be associated
radiation therapy is not immediately available, with leptomeningeal disease, in which there is
for stabilization of unstable spines, or for diag- spread of malignant cells throughout the cerebro-
nostic purposes. This is due to the face that hema- spinal fluid that subsequently coat the brain, spi-
tologic malignancies are highly radiosensitive nal cord, and nerve roots (Fig. 34.2). Of note, a
and respond rapidly to conventional external distinct entity within leptomeningeal disease of
beam radiation [15, 16]. the spine encompasses tumors referred to as drop
Another broad category which has implica- metastases, which have spread to the spine from
tions on spinal metastatic management is the primary intracranial brain tumors rather than
34 Introduction to Spinal Metastases 489
patient as an initial screen as it will highlight

compression deformities and pathologic frac-
tures as well as gross misalignment.
Computed tomography (CT) scans provide
optimal assessment of the osseous structures of
the spine, identifying lytic and blastic metastases
and pathologic fractures. Cortical destruction can
be appreciated months earlier than can be
detected using X-rays, and epidural masses with
displacement of the underlying thecal sac can
also be identified in most cases [30]. CT imaging
is also particularly useful when evaluating
patients for cement augmentation procedures
(e.g., vertebroplasty or kyphoplasty) to determine
amenable levels as well as those undergoing sur-
gical stabilization to determine the size and
health of pedicles that may be instrumented. CT
myelogram is an alternative to magnetic reso-
nance imaging (MRI) in those patients that can-
Fig. 34.2 T1-post contrast MRI demonstrating abnormal not under MRI for various reasons including
enhancement around the spinal cord in the cervicothoracic
non-MRI compatible implants (e.g., pacemaker,
region consistent with leptomeningeal disease
spinal cord stimulator, and intrathecal pump),
those with extremely large body habitus, or those
extra-CNS primary cancers. Glioblastoma, with severe claustrophobia. A lumbar puncture is
medulloblastoma, ependymomas, primitive neu- performed in the lower lumbar spine caudal to
roectodermal tumors (PNETs), germinomas, and the conus medullaris, and the injection of a con-
choroid plexus tumors are among the primary trast agent such as iohexol is administered.
CNS tumors that can result in drop metastases Changes in the normal cylindrical shape of the
[19–28]. Any form of leptomeningeal disease thecal sac or obstruction of contrast flow suggest
carries a poor prognosis, with a median survival epidural compression, whereas dural-based
of less than 3 months [29]. hypodensities or expansion of the spinal cord
may suggest intradural and intramedullary metas-
tases, respectively.
Radiographic Evaluation MRI is the gold standard for the diagnosis of
spinal metastases. For patients with normal glo-
A number of imaging modalities can be used in merular filtration rates (GFR), gadolinium-based
the diagnostic workup for spinal metastases. intravenous administration increases the sensi-
Plain radiographs have become increasingly lim- tivity of detecting these tumors and differentiat-
ited for diagnostic and therapeutic planning pur- ing them from normal variations in bone marrow
poses but do have a role in the assessment of intensity. Typically, most tumors are T1 precon-
load-sensitive deformities in the subaxial spine. trast hypointense and enhance with gadolinium,
Computed tomography (CT) and magnetic reso- though the latter is not always the case and may
nance imaging (MRI) are more advanced high- confound assessment when tumors are isoin-
resolution studies and are standard in the initial tense to normal marrow signal. However, for
evaluation of patients with known primary can- intradural tumor and leptomeningeal disease,
cers. In fact, many spinal metastases are discov- contrast enhancement is required for optimal
ered on staging body CT scans. Plain X-rays, detection. T2 sequences are useful for determin-
however, can still be useful in the symptomatic ing the extent of epidural cord compression. Fat
suppression sequences are also highly sensitive mary tumor histology, the presence of mechani-
for diagnosing even small lesions confined to cal pain, and the burden of systemic disease in
bone, such as short tau inversion recovery (STIR) consideration of whether a patient is a surgical
sequences which demonstrate hyperintense candidate [16, 32, 33].
lesion consistent with tumors. T2 and STIR In those patients that present with pain, it is
sequences are also sensitive for bone edema, important to differentiate the exact type of pain
which can be indicative of acute pathologic from a treatment standpoint. Biologic pain is usu-
fractures. ally described as nighttime pain that improves
MRI-based grading schemes of epidural com- during the course of the day. This is usually due
pression are used frequently for guiding therapy. to the physiologic cyclical nature of endogenous
The epidural spine cord compression, developed steroid production, which is highest in the morn-
by Bilsky et al. [31], divides tumors into grades ing and steadily drops over the course of the day,
0, 1A, 1B, 1C, 2, and 3. Tumors with grade 0 are with lowest levels during nighttime. The effects
confined to bone with no epidural extension. of endogenous steroid production results in
Tumors with ESCC of 1 demonstrate varying decreased inflammation, which in turn explains
degrees of epidural extension, with thecal sac the improvement of pain in the morning versus
impingement without compression (1A), thecal nighttime for patients with spinal metastases.
sac deformation without spinal cord abutment Mechanical instability is distinct type of pain
(1B), and thecal sac deformation with spinal that specifically occurs with movement and usu-
abutment but no compression (1C). ESCC grade ally a byproduct of pathologic fracture or com-
2 tumors compress the spinal cord or cauda pression deformity caused by tumor invasion
equine nerve roots with preservation of CSF sig- [34]. Based on the location of the tumor within
nal on a representative axial cut. ESCC 3 tumors the canal, different symptoms may be described.
have cord or cauda equina compression without In the craniocervical region, rotary head motions
visible CSF flow. ESCC grade 0 and 1 can typi- may exacerbate pain. In the subaxial cervical
cally managed conservatively, whereas ESCC spine, neck or upper extremity radicular pain
grade 2 and 3 tumors may require more aggres- may result with neck flexion/extension/lateral
sive surgical consideration even in the absence of bending/rotation. In the thoracic spine, lying flat
symptoms. with radiating band of pain around the chest or
The role of more advanced imaging options torso may occur. And in the lumbosacral spine,
for spinal metastases is discussed further in axial loading with maneuvers such as going
Chap. 38. from sitting to standing, ambulating, and walk-
ing stairs may result. These instances of
mechanical pain are usually caused by destruc-
Clinical Evaluation tion of important force-sustaining bony regions
that are in close relation to exiting nerve roots
Spinal metastases can often be asymptomatic, which are compressed during various move-
and only discovered through imaging studies per- ment-related activities. The spinal instability
formed for other organs or symptoms. Of those neoplastic score (SINS) was developed to facili-
patients that are symptomatic, the most common tate diagnosis of this phenomena and is com-
manifestation is pain, but some will present with prised of six categories, five radiographic and
a focal neurologic deficit or myelopathic features one clinical [35–37]. Radiographic criteria
in the absence of pain. The NOMS framework, include tumor location within the spinal col-
discussed in more detail later in this section, is a umn, intrinsic nature of bony pathology (e.g.,
useful algorithm that takes into account the neu- lytic vs. blastic), segmental alignment, percent
rologic status of the patient and degree of thecal vertebral body collapse (> or <50%), and poste-
sac compression, the radiosensitivity of the pri- rior element involvement. The sole clinical
component is the presence of movement-related metastases has grown. As such, it is vitally impor-
pain. Cumulative scores range from 0 to 18, tant for the CNS metastases experts to individual-
with SINS 0–6 considered stable, 7–12 indeter- ize care based on patient symptoms and treatment
minate (impending instability), and 13–18 goals. Recently, radiosurgery has advanced con-
unstable. For scores of 7 or above, evaluation by siderably and is a good, noninvasive treatment
a spine surgeon is recommended. option for patients with spinal metastases that do
The distinction between biological and not require interruption of systemic therapy, even
mechanical pain has critical implications for in traditionally “radioresistant” pathologies. The
treatment. Biologic pain can usually be managed use of intraoperative radiation utilizing P32
conservatively with anti-inflammatories, ste- plaques among other radioisotopes as well as
roids, and radiation therapy. However, mechani- improvements in hypofractionated regimens has
cal pain does not respond long term to these also resulted in better local control rates with
treatments [38]. Previous studies have shown that many primary tumor types. Advances in spinal
stabilization via kyphoplasty or instrumented instrumentation have been critical as well. Screw
fusions improves pain and functional disability and rod titanium constructs, interbody devices,
faster than noninterventional therapies [39, 40]. bone substitute allograft, and percutaneous sys-
As with all patients with spinal disease, a thor- tems for stabilization when open decompression
ough neurological evaluation is required. This is not required have all changed the way surgeons
includes full sensorimotor assessment of the approach patients with spinal metastases. Cement
extremities, evaluation for long-tract signs sug- augmentation through kyphoplasty/vertebro-
gestive of myelopathy (Hoffman’s sign, Babinski plasty and the recent development of fenestrated
sign, clonus, deep tendon reflexes), propriocep- pedicle screws for cement augmented screws has
tion evaluation, and rectal tone when concern for also led to less hardware failure and the develop-
cauda equina or cord compression. Patients can ment of pseudarthrosis. Pain-related options such
often have proximal lower extremity weakness as spinal cord stimulation and opiate-based intra-
with pain that radiates from or to the hip. It is thecal pain pumps have also become excellent
vital to keep in mind that diffuse skeletal metas- palliative options for those patients who cannot
tases may be present, and that the hip joint is usu- undergo or benefit from tumor resection or
ally not included in most spinal MRI windows. stabilization.
The FABER maneuver (flexion abduction exter-
nal rotation) of the involved leg can demonstrate
acute exacerbation of hip pain which is more sug- Conclusion
gestive of acetabular pathology than from periph-
eral nerve radicular issues. AP and lateral X-rays In conclusion, spinal metastases are becoming
of the pelvis and femur should be strongly con- more common and treating surgeons must indi-
sidered in these patients. vidualize care to provide the optimal patient out-
come. Subsequent chapters will expand on the
topics highlighted in this chapter including epi-
Treatment Options demiology, bone metabolism, clinical trial
results, and decision-making algorithms for sur-
There have been enormous strides in the treat- gical intervention. Later chapters will delve into
ment algorithms for patients with spinal metasta- more details regarding treatment-specific options,
ses over the past 20 years (Fig. 34.3). As which include separation surgery, radiation ther-
traditional systemic therapies, immunotherapies apy, cement augmentation techniques, and laser
and other biologic agents have prolonged the life therapies among others as well as the manage-
expectancy of those with metastatic disease, the ment of complications that result from these
incidence of central nervous system and skeletal interventions.
a b
Fig. 34.3 Examples of treatment options for spinal instrumented stabilization. (c) Adjuvant radiation plan-
metastases. (a) Later lumbar X-ray in a patient with mul- ning in a patient following “separation surgery” with
tiple myeloma and multiple compression fractures follow- tumor decompression off of the spinal cord and taking
ing stabilization with multilevel kyphoplasties. (b) into account the use of intraoperative radiation (P32
Postoperative CT scan depicting the recent development plaque) in the contouring of target volumes
of fenestrated screws for cement augmentation during
References 15. Rades D, Fehlauer F, Schulte R, Veninga T, Stalpers LJ,

Basic H, et al. Prognostic factors for local control and
survival after radiotherapy of metastatic spinal cord
1. Kakhki VR, Anvari K, Sadeghi R, Mahmoudian AS, compression. J Clin Oncol. 2006;24(21):3388–93.
Torabian-Kakhki M. Pattern and distribution of bone 16. Barzilai O, Fisher CG, Bilsky MH. State of the art
metastases in common malignant tumors. Nucl Med treatment of spinal metastatic disease. Neurosurgery.
Rev Cent East Eur. 2013;16(2):66–9. 2018;82(6):757–69.
2. Zacharia B, Subramaniam D, Joy J. Skeletal 17. Ecker RD, Endo T, Wetjen NM, Krauss WE. Diagnosis
Metastasis-an Epidemiological Study. Indian J Surg and treatment of vertebral column metastases. Mayo
Oncol. 2018;9(1):46–51. Clin Proc. 2005;80(9):1177–86.
3. Laufer I, Sciubba DM, Madera M, Bydon A, 18. Dunning EC, Butler JS, Morris S. Complications in
Witham TJ, Gokaslan ZL, et al. Surgical manage- the management of metastatic spinal disease. World J
ment of metastatic spinal tumors. Cancer Control. Orthop. 2012;3(8):114–21.
2012;19(2):122–8. 19. Buhl R, Barth H, Hugo HH, Hutzelmann A, Mehdorn
4. Choi D, Crockard A, Bunger C, Harms J, Kawahara HM. Spinal drop metastases in recurrent glioblastoma
N, Mazel C, et al. Review of metastatic spine tumour multiforme. Acta Neurochir. 1998;140(10):1001–5.
classification and indications for surgery: the con- 20.
Carlsen JG, Tietze A, Lassen YA, Rosendal
sensus statement of the Global Spine Tumour Study F. Paraplegia due to drop metastases from ana-
Group. Eur Spine J. 2010;19(2):215–22. plastic oligodendroglioma. Br J Neurosurg.
5. Sohn S, Chung CK, Han KD. Jung JH. Kim J, et al. A 2012;26(1):94–5.
Nationwide Study of Surgery in a Newly Diagnosed 21. Chandra P, Purandare N, Shah S, Agrawal A,

Spine Metastasis Population. J Korean Neurosurg Rangarajan V. “Drop” metastases from an operated
Soc: Hyeun JH; 2018. case of intracranial anaplastic ependymoma identified
6. Robial N, Charles YP, Bogorin I, Godet J, Beaujeux on fluoro-2-deoxyglucose positron emission tomog-
R, Boujan F, et al. Is preoperative embolization a pre- raphy/computed tomography. Indian J Nucl Med.
requisite for spinal metastases surgical management? 2017;32(1):68–70.
Orthop Traumatol Surg Res. 2012;98(5):536–42. 22. Domingues RC, Taveras JM, Reimer P, Rosen

7. Smit JW, Vielvoye GJ, Goslings BM. Embolization BR. Foramen magnum choroid plexus papilloma
for vertebral metastases of follicular thyroid carci- with drop metastases to the lumbar spine. AJNR Am J
noma. J Clin Endocrinol Metab. 2000;85(3):989–94. Neuroradiol. 1991;12(3):564–5.
8. Suzuki H, Kondo T, Kuwatsuru R, Wada K, Kubota 23. Hayes LL, Jones RA, Palasis S, Aguilera D, Porter
M, Kobayashi H, et al. Decompressive surgery in DA. Drop metastases to the pediatric spine revealed
combination with preoperative transcatheter arterial with diffusion-weighted MR imaging. Pediatr Radiol.
embolization: successful improvement of ambulatory 2012;42(8):1009–13.
function in renal cell carcinoma patients with meta- 24. Pande SB, Pavithran K. Drop metastases to the spi-
static extradural spinal cord compression. Int J Urol. nal cord from infratentorial glioblastoma multi-
2011;18(10):718–22. forme in post-temozolomide era. J Cancer Res Ther.
9. Wilson MA, Cooke DL, Ghodke B, Mirza 2015;11(4):1039.
SK. Retrospective analysis of preoperative emboli- 25.
Raaijmakers C, Wilms G, Demaerel P, Baert
zation of spinal tumors. AJNR Am J Neuroradiol. AL. Pineal teratocarcinoma with drop metastases:
2010;31(4):656–60. MR features. Neuroradiology. 1992;34(3):227–9.
10. Awad AW, Almefty KK, Ducruet AF, Turner JD, 26. Solomou AG. Magnetic resonance imaging of pineal
Theodore N, McDougall CG, et al. The efficacy and tumors and drop metastases: a review approach. Rare
risks of preoperative embolization of spinal tumors. J Tumors. 2017;9(3):6715.
Neurointerv Surg. 2016;8(8):859–64. 27.
Yeamans CL, Gutierrez-Quintana R, Haley A,
11. Patsalides A, Leng LZ, Kimball D, Marcus J,
Lamm CG. Magnetic resonance imaging and clini-
Knopman J, Laufer I, et al. Preoperative catheter cal findings associated with choroid plexus spinal
spinal angiography and embolization of cervical cord “drop” metastases. J Am Anim Hosp Assoc.
spinal tumors: outcomes from a single center. Interv 2017;53(5):265–9.
Neuroradiol. 2016;22(4):457–65. 28. Yu H, Yao TL, Spooner J, Stumph JR, Hester R,
12. Lasocki A, Gaillard F, Harrison SJ. Multiple myeloma Konrad PE. Delayed occurrence of multiple spi-
of the spine. Neuroradiol J. 2017;30(3):259–68. nal drop metastases from a posterior fossa choroid
13. Latif T, Hussein MA. Advances in multiple myeloma plexus papilloma. Case report. J Neurosurg Spine.
and spine disease. Clin Lymphoma Myeloma. 2006;4(6):494–6.
2005;6(3):228–33. 29. Chamberlain MC. Leptomeningeal metastases in the
14. Kim SI, Kim YH, Ha KY, Lee JW, Lee JW. Surgical MRI era. Neurology. 2011;76(2):200; author reply −1.
roles for spinal involvement of hematological malig- 30. Shah LM, Salzman KL. Imaging of spinal metastatic
nancies. J Korean Neurosurg Soc. 2017;60(5):534–9. disease. Int J Surg Oncol. 2011;2011:769753.
31. Bilsky MH, Laufer I, Fourney DR, Groff M, Schmidt plastic score: an analysis of reliability and validity
MH, Varga PP, et al. Reliability analysis of the epi- from the spine oncology study group. J Clin Oncol.
dural spinal cord compression scale. J Neurosurg 2011;29(22):3072–7.
Spine. 2010;13(3):324–8. 37. Versteeg AL, Verlaan JJ, Sahgal A, Mendel E,

32. Laufer I, Rubin DG, Lis E, Cox BW, Stubblefield MD, Quraishi NA, Fourney DR, et al. The spinal instabil-
Yamada Y, et al. The NOMS framework: approach to ity neoplastic score: impact on oncologic decision-
the treatment of spinal metastatic tumors. Oncologist. making. Spine (Phila Pa 1976). 2016;41:S231–S237.
2013;18(6):744–51. 38. Huisman M, van der Velden JM, van Vulpen M,
33.
Moussazadeh N, Laufer I, Yamada Y, Bilsky van den Bosch MA, Chow E, Oner FC, et al.
MH. Separation surgery for spinal metastases: effect Spinal instability as defined by the spinal instabil-
of spinal radiosurgery on surgical treatment goals. ity neoplastic score is associated with radiother-
Cancer Control. 2014;21(2):168–74. apy failure in metastatic spinal disease. Spine J.
34. Fisher CG, DiPaola CP, Ryken TC, Bilsky MH,
2014;14(12):2835–40.
Shaffrey CI, Berven SH, et al. A novel classification 39. Berenson J, Pflugmacher R, Jarzem P, Zonder J,

system for spinal instability in neoplastic disease: an Schechtman K, Tillman JB, et al. Balloon kyphoplasty
evidence-based approach and expert consensus from versus non-surgical fracture management for treat-
the Spine Oncology Study Group. Spine (Phila Pa ment of painful vertebral body compression fractures
1976). 2010;35(22):E1221–9. in patients with cancer: a multicentre, randomised
35. Fisher CG, Schouten R, Versteeg AL, Boriani S, Varga controlled trial. Lancet Oncol. 2011;12(3):225–35.
PP, Rhines LD, et al. Reliability of the spinal instabil- 40. Hussain I, Barzilai O, Reiner AS, DiStefano N,

ity neoplastic score (SINS) among radiation oncolo- McLaughlin L, Ogilvie S, et al. Patient-reported out-
gists: an assessment of instability secondary to spinal comes after surgical stabilization of spinal tumors:
metastases. Radiat Oncol. 2014;9:69. symptom-based validation of the Spinal Instability
36. Fourney DR, Frangou EM, Ryken TC, Dipaola CP, Neoplastic Score (SINS) and surgery. Spine J.
Shaffrey CI, Berven SH, et al. Spinal instability neo- 2017;17:S181.
Epidemiology of Spinal Metastatic
Disease
35
John Berry-Candelario, Mark H. Bilsky, Ilya Laufer,
C. Rory Goodwin, and Ori Barzilai
Introduction tate cancer 15%, lung cancer 14%, renal cancer

12%, and multiple myeloma 6%.
With the advent of new and more selective cancer
Most common spinal metastases Incidence (%)
therapies, the intricacies of spine metastases care
Breast cancer 19
have reached the forefront of medicine. Spine
Prostate cancer 15
metastasis remains the most prevalent form of Lung cancer 14
spinal neoplasia with the likelihood of further Renal cancer 12
growth as these patients live longer with chroni- Multiple myeloma 6
cally controlled cancer. While the most common
sites of metastases remain the liver and lungs, In about 10% of patients, SM is the initial
bone is next in line. Spine metastasis (SM) repre- manifestation of the primary disease. In the
sents the largest proportion of this population United States, there are approximately 120,000
with estimates that up to 40% of all cancer cases per year with 20% or 25,000 cases pre-
patients will have SM in their lifetime and almost senting with spinal cord compression [5]. The
20% of patients diagnosed with SM develop vast majority of these lesions occur in the verte-
symptomatic spinal cord compression [1, 2]. bral body or contiguous marrow and epidural
Postmortem studies have shown that though not spaces, whereas 5% will present as an intradu-
always diagnosed, up to 90% of cancer patients ral lesion and only 1% with intramedullary
may have microscopic evidence of SM [3, 4]. metastases [6]. While all segments of the spinal
Furthermore, 70–90% of patients with breast or column are susceptible, the thoracic spine is the
prostate cancer have some form of skeletal metas- most frequent site (70%), followed by the lum-
tases. The major contributors and their incidence bosacral spine (25%), and then the cervical
to spine metastases are breast cancer 19%, pros- spine (5%). This is thought to be the function of
bone mass and blood flow [6]. Patients with spi-
nal metastasis have been reported to have a
J. Berry-Candelario · M. H. Bilsky · I. Laufer median survival of 7 months overall. Those
O. Barzilai (*) with epidural disease spread and leptomenin-
Department of Neurosurgery, Memorial Sloan
geal disease may have even worse survival
e-mail: barzilao@mskcc.org ranging from 3 to 6 months [7]. However,
recent advances in cancer treatment point to
C. R. Goodwin
Department of Neurosurgery, Duke University longer overall survival and crystalizing the
Medical Center, Durham, NC, USA need for optimal care [8].

https://doi.org/10.1007/978-3-030-42958-4_35
496 J. Berry-Candelario et al.
Over the last decade, there has been an explo- ing in bone remodeling. Some of them include
sion in the identification of tumor-specific molec- PTHRP, IL-11, IL-6, TNF-alpha, and granulocyte-
ular signatures and associated targeted therapies. macrophage colony-stimulating factor [11, 12].
Combined with multidisciplinary care and tar- This, in turn, induces the release of other growth
geted therapy, these aforementioned advance- factors that lead to a vicious cycle of bone destruc-
ments have facilitated improved survival, tion and growth of local tumor.
improved progression-free survival, and in some Morphological changes to the bone can result
cases even cure. In this chapter, we will explore in biologic pain, and often, the osteolytic process
some of the newest epidemiological data on some results in vertebral body or posterior element
of the more prevalent spinal metastatic diseases fractures, often requiring surgical stabilization.
with specific cancer markers and their correlates
to care. We will focus our attention on the epide-
miology of those cancers with the highest inci- Clinical Presentation
dence of spine metastasis.
Regardless of the type of the primary tumor, the
typical clinical presentation of spine metastasis
Pathophysiology includes either biologic bone pain (most com-
mon) or mechanical (movement-related) back
Cancer metastasizes to the bone through different pain. Mechanical back pain results from spinal
ways of propagation, yet most frequently through instability secondary to pathologic fracture and
hematogenous spread. In a review, Massagué illus- can result in radiculopathy and myelopathy due
trates with perspicacious clarity the critical steps to abnormal spinal motion. The presenting symp-
of cancer metastasis [9]. Those basic steps com- toms are dictated by the tumor or fracture loca-
mence with local invasion, extravasation, survival tion and rate of growth [13]. Classically, biologic
in circulation, intravasation, and colonization. The pain is thought to be due to an inflammatory
innate defenses against metastasis are overcome response to tumor expansion in the vertebral
via a set of general genetic harbingers for infiltra- body that is worse at night when diurnal levels of
tion. Among the gene classes involved are regu- cortisol are lowest which typically controls for
lated transcription factors TWIST1, SNAI1, and inflammation. Pathologic fractures produce acute
SNAI2 that allow for invasion. Furthermore, meta- and subacute pain secondary to bony and some-
static growth is initiated by the suppression of non- time ligamentous destruction. Furthermore, the
coding RNAs, like miR-126. Their work degree of tumor extension may produce cord or
hypothesizes further that beyond traits like cell root compression resulting in neurologic sequelae
motility and membrane degradation, tumor cells including paresthesias, dysesthesias, radiculopa-
develop an organ-specific infiltrative advantage thy, motor weakness, and/or bladder/bowel com-
that mediates adhesion and penetration to organs promise. Significant spinal cord compression
like the bone [9]. Venous spread, primarily through may lead to spinal cord edema, myelopathy, and
Batson’s plexus [6], is considered the principal ischemia/infarction [14] with resultant deficits in
process of metastases to the spinal column. This neurologic function and ambulation.
contrasts with arterial spread to other osseous sites As mentioned, the most common symptom at
such as the shoulder and pelvis (proximal) fol- initial presentation is pain. Motor dysfunction is
lowed by the elbow and knee (distal). Less fre- the next most common presentation with between
quently lesions spread by contiguity and even less 35 and 75% of all patients. Patients typically
frequently via lymphatic spread (the role of which complain of arm/leg fatigue or heaviness. This is
is not well defined) [10]. Once cancer cells have not always accompanied by definitive weakness.
invaded the bone, they produce growth factors that Sensory complaints will lag motor findings
stimulate osteoblastic or osteolytic activity result- unless there is nerve root compression. It is rare
35 Epidemiology of Spinal Metastatic Disease 497
to find bowel/bladder dysfunction in isolation P rognostication and Health-Related

based on the innervation pattern [14]. True spine- Quality of Life
related bowel/bladder incontinence will typically
be accompanied by severe back pain. Several scoring systems such as the Tokuhashi
revised score [20], the Tomita score [21], and the
Bauer modified score [22, 23] have been devel-
Diagnostic Evaluation oped to estimate expected survival in patients
with spinal metastases. In the era of modern can-
Typically, the SM patient is known to the oncol- cer care, their reliability and utility have been
ogy service prior to presentation. It is important questioned [24, 25]. New prediction models,
to recognize that cancer patients with back pain such as the Skeletal Oncology Research Group
or neurologic complaints need prompt evaluation (SORG) nomogram [26], attempt to overcome
with imaging and laboratory studies [15]. Plain the shortcomings of these models by identifying
X-rays remain useful particularly in evaluating more prognostic factors associated with out-
instability, and full-length imaging aids in the comes. Physicians should refrain from strictly
assessment of sagittal imbalance. Computed adhering to these prediction models, and patients
tomography (CT) with reconstructions provides should be considered for surgery if reasonable
useful information of the bony elements of the systemic therapy is available.
spine. For example, lytic versus sclerotic changes Treatment goals for metastatic spine disease
and vertebral body collapse are well evaluated by are palliative. In the past, outcomes of SM
CT scans [16]. Positron emission tomography patients relied on clinician-based measures, yet
(PET) combined with CT is also useful and can recently there has been an increase in utilization
help with the assessment of treatment response. of patient-reported outcome (PRO) tools as they
Finally, magnetic resonance imaging (MRI) is express a direct measure of the value of care as
the workhorse in diagnostic evaluation with its perceived by the recipient [27]. Several generic
high-resolution multi-planar imaging, clear outcome measures have been widely used for
assessment of soft tissue and osseous structures, PRO reporting in the spinal oncology popula-
and ability to precisely define the relationship tion, including EuroQol 5-D (EQ-5D), Oswestry
between the tumor and surrounding neurovascu- Disability Index (ODI), Visual Analog Scale
lar structures, bony elements, and viscera [17]. (VAS), and Short Form 36 (SF-36) [28]; how-
Every patient with spinal metastases should ever, none of these instruments focus on spine
undergo a total spine MRI as occult lesions are a cancer-specific symptoms. To address this need,
common occurrence and are often easily treated the Spine Oncology Study Group Outcome
when promptly diagnosed [18]. Current MR Questionnaire (SOSGOQ) was created and rep-
technologies, such as ultrafast data acquisition resents the only PRO instrument fully focused
and high-performance gradient systems, have on the assessment of patients with spinal tumors
made total spine examinations tolerable with [29, 30]. Consistent use of validated health-
considerably shortened examination times. related quality of life (HRQoL) tools facilitates a
Additional sequences such as diffusion-weighted common language in communication and report-
imaging (DWI) and short tau inversion recovery ing as we continue to evaluate patient outcomes
(STIR) offer information about local microstruc- in the current era of spine cancer therapy. An
tural differences and the presence of any patho- abundance of recent data demonstrate the benefit
logical alterations. Dynamic contrast-enhanced of validated PRO-based evaluations following
(DCE) MRI perfusion imaging provides func- spinal surgery for both open surgery and mini-
tional information on tumor vascularity and mally invasive surgery, as well as for patients
hemodynamics and can be used as a surrogate for with oligometastatic and widespread systemic
determining tumor progression [19]. disease [31–35].
Targeted Treatment Paradigms dermal growth factor receptor 2 (HER2)-positive

breast cancers. Tamoxifen, an ER antagonist, has
Early successes in identifying and targeting indi- led to a marked improvement in survival [42].
vidual oncogenic drivers, together with the Trastuzumab, a humanized monoclonal antibody,
increasing feasibility of sequencing tumor against HER2 has promising results [43].
genomes, have brought forth the promise of Few studies have investigated the value of
genome-driven oncology care [36]. Currently, these targeted therapies in spinal cancer. A study
advancement in the understanding of the genetic from Johns Hopkins looked at breast cancer-
basis of diseases is changing the way we diag- specific parameters for spinal metastases and
nose and treat spine cancer [37]. Genomic found estrogen receptor positivity to be associ-
sequencing drives clinical management of tumors ated with longer median survival, but metastatic
such as melanoma, sarcomas, and carcinomas of tumors in the cervical spine were associated
the lung, breast, thyroid, ovary, and colon [36]. with shorter median survival [44]. Perhaps, it is
These tools have been studied mostly in non- due to the highly morbid symptoms associated
spinal tumors, yet interest on the effect they may with spinal cord compression at this level.
have on spine cancer care is growing [38, 39]. Interestingly, the presence of visceral disease or
Traditionally, the effect of systemic therapy on >1 bony metastasis was not found to be prog-
osseous metastases has been limited. A recent nostic [45–47].
trial showed favorable response of osseous RCC
metastases treated with cabozantinib, a small
molecule tyrosine kinase inhibitor, indicating Lung Cancer
that new systemic therapy agents may offer local
tumor control for osseous metastases [40]. With One of the most profound and most common can-
the ongoing expansion of precision medicine, cers to the spine is lung cancer. With over 1.8
surgeons treating cancer patients will need to million newly diagnosed cases per year, roughly
increase their familiarity with the genomic and 70% will have locally advanced or metastatic dis-
molecular oncology landscape to make informed ease at the time of initial presentation [48–51].
decisions and maintain a leadership role in patient The largest group is non-small cell lung cancers
care. Although the concept of targeted therapy is which can be subdivided into large cell carci-
similar, the results and treatment response remain noma, adenocarcinoma, and squamous cell carci-
cancer-specific and will be discussed separately. noma. They have a median overall survival (OS)
of 8–11 months when presenting with advanced
disease burden [52, 53]. However, recent studies
Breast Cancer have reported improved survival with surgical
intervention with RT. Weiss found over half of
Breast cancer has a strong predisposition to their patients who underwent surgical resection
metastasize to bone. Despite being the second recovered at least 1 Frankel grade [54]. The big-
leading cause of cancer-related deaths, the gest advancement in treatment involves the
median survival after metastasis diagnosis is development of therapeutics based on molecular
almost 2 years. This has led providers to advocate markers. Endothelial growth factor receptor
for aggressive treatment strategies to provide pal- (EGFR) mutations play a major role in NSCLC,
liation of pain and preservation and/or improve- and certain EGFR inhibitors (erlotinib and gefi-
ment of neurological function [41]. The breadth tinib) work through the tyrosine kinase inhibitor
of molecular knowledge has transformed the dis- pathway with improved survival up to
ease course and prognosis for this patient popula- 24–36 months [55]. Their utility has been chal-
tion. Research has advanced with therapies for lenged by tumor resistance through the upregula-
estrogen receptor (ER+)-positive and human epi- tion of other tyrosine kinase receptors. A literature
review evaluating survival among patients with conventional external beam radiation or radiosur-
NSCLC metastatic to the spine found that while gery. The role of surgery for SM has decreased
the overall survival of patients with lung cancer with time but remains substantial for those with
metastases to the spine was 3.6–9 months, the progressive neurologic deficits or those with spi-
median survival of NSCLC patients with targe- nal instability.
table EGFR mutations was 18 months [56].
Another area of advancement has been in immu-
notherapy. Cytotoxic T-lymphocyte-associated Renal Cell Cancer
protein 4 inhibitors (CTLA4, e.g., ipilimumab),
anti-programmed death 1 (anti-PD1, e.g., While renal cell carcinoma (RCC) only accounts
nivolumab), and anti-programmed death ligand 1 for 2.5% of all cancers [60, 63], about 40% of
(anti-PD-L1, e.g., BMS-936559) are three path- bony RCC metastasis occur in the spine. At that
ways that have shown moderate success with point, the median survival is estimated at
marginally improving survival [55]. These 10 months. In fact, almost one-third of patients
improved survival data suggest that patients with have advanced or metastatic RCC at the time of
good performance scores may have extended sur- initial diagnosis. Renal cell is known to be radio-
vival in the era of targeted therapies. As such, resistant to conventional EBRT; however,
palliation of spinal metastases-related symptoms response rates with radiosurgery have proven
is warranted and should be addressed early. promising. The advancement in the molecular and
genomic knowledge of the disease has resulted in
the approval of several targeted therapies for the
Prostate Cancer treatment of metastatic RCC (mRCC). Some of
those agents include cytokine actors (IL-2); tyro-
Prostate cancer is the second most common cause sine kinase receptor inhibitors, like sunitinib and
of cancer-related deaths in men. Metastatic pros- axitinib; mTOR inhibitors (e.g., temsirolimus);
tate cancer most commonly affects the spine, and and VEGF inhibitors, to name a few. With only
the 1-year survival after SM diagnosis has been 10% of mRCC patients living to 5 years, the
reported between 73 and 83%, with a median OS advent of multitargeted therapies has resulted in
after diagnosis of spinal metastasis of 24 months PFS up to 27 months and OS to 40 months [63].
[57–59]. A major advancement in therapy of Given the recent introduction of these therapies, it
prostate cancer was the discovery of androgen is not yet clear if the 5-year survival rate has
receptor antagonists that have greatly improved meaningfully changed. Immunotherapeutic
patient outcomes [57, 59]. The treatment of agents such as anti-PD1, anti-PD-L1, and anti-
castration-resistant prostate cancer is proven dif- CTLA-4 antibodies have been explored in pre-
ficult; however, several targeted therapies are liminary studies with a reported 30% overall
available and include cabozantinib (an MET and response rate and 20–25% prolonged response
VEGFR2 inhibitor), cetuximab (a monoclonal rate [64]. They also identified tools that could be
antibody against EGFR), gefitinib/erlotinib particularly useful for prognostication of mRCC
(small tyrosine kinase inhibitors), and ipilim- to the spine such as the initial Fuhrman grade,
umab (anti-CTLA4) [58, 60, 61]. While studies Tokuhashi score, and Memorial Sloan Kettering
of these and other agents are ongoing, they have Cancer Center (MSKCC/Motzer) score.
shown marginal improvement in clinical studies. Traditionally, surgical excision of renal cell
Radiation therapy remains the mainstay of treat- spinal metastases was routinely performed as
ment for spinal metastases from prostate cancer. these tumors are resistant to conventional radia-
Fortunately, prostate cancer is considered relation. This is particularly challenging due to the
tively radiosensitive [62], and hence, prostate vascularity of these lesions, and preoperative
spinal metastases can be treated effectively with embolization is often utilized. Moreover, solitary
RCC spinal metastases were surgically removed 7. Loblaw D. A population-based study of malignant
“en bloc.” The integration of spinal radiosurgery spinal cord compression in Ontario. Clin Oncol (R
Coll Radiol). 2003;15(4):211–7.
has changed the management of these tumors 8. Bydon M, et al. Impact of smoking on compli-
significantly, and currently, though still debat- cation and pseudarthrosis rates after single- and
able, there is little role for wide or total excisions 2-level posterolateral fusion of the lumbar. Spine.
since high dose per fraction radiation provides 2014;39(21):1765–70.
9. Nguyen DX, Bos PD, Massague J. Metastasis: from
excellent local control rates with minimal associ- dissemination to organ-specific colonization. Nat Rev
ated morbidity [65]. Despite the improved local Cancer. 2009;9(4):274–84.
control with SBRT and advancement in targeted 10. Maccauro G, et al. Physiopathology of spine metasta-
therapies, surgery still plays a pivotal role in the sis. Int J Surg Oncol. 2011;2011:1–8.
11. Yin JJ, et al. TGF-beta signaling blockade inhib-

management of mRCC spinal metastases, partic- its PTHrP secretion by breast cancer cells and
ularly in solitary and oligometastatic disease and bone metastases development. J Clin Invest.
for those with high-grade spinal cord compres- 1999;103(2):197–206.
sion requiring separation surgery or patients with 12. Mundy GR. Metastasis to bone: causes, consequences
and therapeutic opportunities. Nat Rev Cancer.
progressive neurologic deficits. 2002;2(8):584–93.
13. Helweg-Larsen S, Sørensen PS. Symptoms and signs
in metastatic spinal cord compression: a study of pro-
Conclusion gression from first symptom until diagnosis in 153
patients. Eur J Cancer. 1994;30a(3):396–8.
14. Witham TF, et al. Surgery Insight: current manage-
This chapter provides a brief overview of the epi- ment of epidural spinal cord compression from
demiology of spinal metastasis. In the era of pro- metastatic spine disease. Nat Clin Pract Neurol.
longed cancer survivors, it is likely that the 2006;2(2):87–94.
15. Bilsky MH, Laufer I, Burch S. Shifting paradigms in
magnitude of patients with spinal metastases will the treatment of metastatic spine disease. Spine (Phila
increase. With a better understanding of the Pa 1976). 2009;34(22 Suppl):S101–7.
molecular mechanisms of metastatic cancer biol- 16. Jacobs WB, Perrin RG. Evaluation and treatment of
ogy, there is a broader understanding that the spinal metastases: an overview. Neurosurg Focus.
2001;11(6):e10.
treatment of metastatic disease requires personal- 17. Andreula C, Murrone M. Metastatic disease of the
ization and a multidisciplinary effort. While we spine. Eur Radiol. 2005;15(3):627–32.
chose to focus on advancements in the more com- 18. Laufer I, et al. The NOMS framework: approach to
mon spine metastatic cancer types, these develop- the treatment of spinal metastatic tumors. Oncologist.
2013;18(6):744–51.
ments are proving true for cancer in general. 19. Kumar KA, et al. A pilot study evaluating the use
of dynamic contrast-enhanced perfusion MRI
to predict local recurrence after radiosurgery on
spinal metastases. Technol Cancer Res Treat.
References 2017:1533034617705715.
20. Tokuhashi Y, et al. A revised scoring system for pre-
1. Delank K-S, et al. The treatment of spinal metastases. operative evaluation of metastatic spine tumor prog-
Dtsch Arztebl Int. 2011. nosis. Spine (Phila Pa 1976). 2005;30(19):2186–91.
2. Ortiz Gomez JA. The incidence of vertebral body 21. Tomita K, et al. Surgical strategy for spinal metasta-
metastases. Int Orthop. 1995;19(5):309–11 ses. Spine (Phila Pa 1976). 2001;26(3):298–306.
3. Cobb CA, Leavens ME, Eckles N. Indications for 22. Bauer HCF, Wedin R. Survival after surgery for spi-
nonoperative treatment of spinal cord compression nal and extremity metastases: prognostication in 241
due to breast cancer. J Neurosurg. 1977:653–658. patients. Acta Orthop Scand. 1995;66(2):143–6.
4. Wong DA, Fornasier VL, MacNAB I. Spinal metasta- 23. Leithner A, et al. Predictive value of seven preopera-
ses: the obvious, the occult, and the impostors. Spine tive prognostic scoring systems for spinal metastases.
(Phila Pa 1976). 1990;15(1):1–4. Eur Spine J. 2008;17(11):1488–95.
5. Ecker RD, et al. Diagnosis and treatment of ver- 24. Dardic M, et al. Evaluation of prognostic scor-

tebral column metastases. Mayo Clin Proc. ing systems for spinal metastases in 196
2005;80(9):1177–86. patients treated during 2005–2010. Eur Spine J.
6. Batson OV. The function of the vertebral veins and 2015;24(10):2133–41.
their role in the spread of metastases. Ann Surg. 25. Zoccali C, et al. The Tokuhashi score: effectiveness
1940;112(1):138–49. and pitfalls. Eur Spine J. 2016;25(3):673–8.
26. Paulino Pereira NR, et al. Development of a

41. Patchell RA, et al. Direct decompressive surgical

prognostic survival algorithm for patients with resection in the treatment of spinal cord compres-
metastatic spine disease. J Bone Joint Surg Am. sion caused by metastatic cancer: a randomised trial.
2016;98(21):1767–76. Lancet. 2005;366(9486):643–8.
27. Bilsky MH, et al. Reliability analysis of the epidural 42. Banin Hirata BK, et al. Molecular markers for breast
spinal cord compression scale. J Neurosurg Spine. cancer: prediction on tumor behavior. Dis Markers.
2010:324–328. 2014;2014:1–12.
28. DeVine J, et al. Evaluating the correlation and respon- 43. Monteiro Ide P, et al. Targeting HER family in HER2-
siveness of patient-reported pain with function and positive metastatic breast cancer: potential biomark-
quality-of-life outcomes after spine surgery. Spine ers and novel targeted therapies. Pharmacogenomics.
(Phila Pa 1976). 2011;36:S69–74. 2015;16(3):257–71.
29. Street J, et al. Introducing a new health-related qual- 44. Zadnik PL, et al. Prolonged survival follow-

ity of life outcome tool for metastatic disease of the ing aggressive treatment for metastatic breast can-
spine: content validation using the international clas- cer in the spine. Clin Exp Metastasis. 2014;31(1):
sification of functioning, disability, and health; on 47–55.
behalf of the Spine Oncology Study Group. Spine 45. Sciubba DM, et al. Positive and negative prog-

(Phila Pa 1976). 2010;35(14):1377–86. nostic variables for patients undergoing spine sur-
30. Versteeg AL, et al. Psychometric evaluation and
gery for metastatic breast disease. Eur Spine J.
adaptation of the Spine Oncology Study Group 2007;16(10):1659–67.
Outcomes Questionnaire to evaluate health-related 46. Shehadi JA, et al. Surgical treatment strategies and
quality of life in patients with spinal metastases: outcome in patients with breast cancer metastatic
validity and reliability of the SOSGOQ. Cancer. to the spine: a review of 87 patients. Eur Spine J.
2018;124(8):1828–38. 2007;16(8):1179–92.
31. Barzilai O, et al. Hybrid surgery-radiosurgery therapy 47. Walcott BP, et al. Surgical treatment and outcomes of
for metastatic epidural spinal cord compression: a metastatic breast cancer to the spine. J Clin Neurosci.
prospective evaluation using patient-reported out- 2011;18(10):1336–9.
comes. Neurooncol Pract. 2018;5(2):104–13. 48. Kalia M. Biomarkers for personalized oncology:

32. Barzilai O, et al. Minimal access surgery for spinal recent advances and future challenges. Metabolism.
metastases: prospective evaluation of a treatment 2015;64(3):S16–21.
algorithm using patient-reported outcomes. World 49. Lauro S, et al. The use of bevacizumab in non-

Neurosurg. 2018;120:e889–901. small cell lung cancer: an update. Anticancer Res.
33. Barzilai O, et al. Predictors of quality of life
2014;34(4):1537–45.
improvement after surgery for metastatic tumors 50. Raparia K, et al. Molecular profiling in non–small
of the spine: prospective cohort study. Spine J. cell lung cancer: a step toward personalized medicine.
2018;18(7):1109–15. Arch Pathol Lab Med. 2013;137(4):481–91.
34. Barzilai O, et al. Survival, local control, and health- 51. Tobin NP, et al. The importance of molecular

related quality of life in patients with oligometastatic markers for diagnosis and selection of targeted
and polymetastatic spinal tumors: a multicenter, inter- treatments in patients with cancer. J Intern Med.
national study. Cancer. 2019;125(5):770–8. 2015;278(6):545–70.
35. Fehlings MG, et al. Survival and clinical outcomes 52. Marzuka A, et al. Melanoma treatments: advances
in surgically treated patients with metastatic epi- and mechanisms. J Cell Physiol. 2015;230(11):
dural spinal cord compression: results of the pro- 2626–33.
spective multicenter AOSpine study. J Clin Oncol. 53.
Menzies AM, Long GV. Recent advances in
2016;34(3):268–76. melanoma systemic therapy. BRAF inhibitors,
36. Hyman DM, Taylor BS, Baselga J. Implementing
CTLA4 antibodies and beyond. Eur J Cancer.
genome-driven oncology. Cell. 2017;168(4):584–99. 2013;49(15):3229–41.
37. Goodwin CR, et al. Molecular markers and targeted 54. Weiss RJ, Wedin R. Surgery for skeletal metasta-

therapeutics in metastatic tumors of the spine: chang- ses in lung cancer: complications and survival in 98
ing the treatment paradigms. Spine (Phila Pa 1976). patients. Acta Orthop. 2011;82(1):96–101.
2016;41(Suppl 20):S218–23. 55. Helissey C, Champiat S, Soria J-C. Immune check-
38. Caruso JP, et al. Stereotactic radiosurgery and immu- point inhibitors in advanced nonsmall cell lung can-
notherapy for metastatic spinal melanoma. Neurosurg cer. Curr Opin Oncol. 2015;27(2):108–17.
Focus. 2015;38(3):E6. 56. Batista N, et al. Emerging and established clinical,
39. Shankar GM, et al. Effect of immunotherapy status on histopathological and molecular parametric prognos-
outcomes in patients with metastatic melanoma to the tic factors for metastatic spine disease secondary to
spine. Spine (Phila Pa 1976). 2017;42(12):E721–5. lung cancer: Helping surgeons make decisions. J Clin
40. Choueiri TK, et al. Cabozantinib versus everolimus Neurosci. 2016;34:15–22.
in advanced renal cell carcinoma (METEOR): final 57. Drzymalski DM, et al. Predictors of survival in

results from a randomised, open-label, phase 3 trial. patients with prostate cancer and spinal metastasis. J
Lancet Oncol. 2016;17(7):917–27. Neurosurg Spine. 2010:789–794.
58. Fu W, et al. Progress of molecular targeted thera- sion framework: neurologic, oncologic, mechanicals
pies for prostate cancers. Biochim Biophys Acta. stability, and systemic disease. J Clin Oncol.
2012;1825(2):140–52. 2017;35(21):2419–27.
59. Toren P, Zoubeidi A. Targeting the PI3K/Akt path- 63. Escudier B. Emerging immunotherapies for renal
way in prostate cancer: challenges and opportunities cell carcinoma. Ann Oncol. 2012;23(suppl 8):
(review). Int J Oncol. 2014;45(5):1793–801. viii35–40.
60. Combe P, et al. Trial watch: therapeutic vaccines in 64. Carlo MI, Voss MH, Motzer RJ. Checkpoint

metastatic renal cell carcinoma. Oncoimmunology. inhibitors and other novel immunotherapies for
2015;4(5):e1001236. advanced renal cell carcinoma. Nat Rev Urol. 2016;
61. Gerritsen WR. The evolving role of immunother-
13(7):420–31.
apy in prostate cancer. Ann Oncol. 2012;23(suppl 65. Yamada Y, et al. The impact of histology and deliv-
8):viii22–7. ered dose on local control of spinal metastases treated
62. Barzilai O, et al. Integrating evidence-based medi- with stereotactic radiosurgery. Neurosurg Focus.
cine for treatment of spinal metastases into a deci- 2017;42(1):E6.
Bone Metabolism in Cancer
36
Tilman D. Rachner, Lorenz C. Hofbauer,
and Andy Göbel
Introduction bone. However, under phases of increased strain

or stress, bone adapts to the changing require-
Basics of Bone Physiology ments by increasing bone formation. Vice versa,
if there is a reduction in physical strain, for exam-
In adult bone, a physiological level of bone ple, due to temporary immobilization after surgi-
remodeling is required to maintain long-term cal procedures, bone adapts by reducing its mass
structural quality and strength. The activity of and architecture [3].
bone-forming osteoblasts and bone-resorbing Locally and systemically, bone metabolism is
osteoclasts is coupled, and under healthy condi- regulated by numerous hormones, cytokines, as
tions, the level of bone formation and resorption well as physical factors [1, 2]. The most promi-
is balanced, resulting in a stable bone mass. The nent hormones, which regulate bone acquisition
coupling process is mediated by direct cell con- as well as the differentiation and lifespan of bone
tacts, ligand-receptor interactions, and a variety cells, are sex hormones. Estrogens suppress bone
of soluble factors. In addition, mechanical force resorption by inducing osteoclast apoptosis and
and microcracks are key drivers of bone remodel- inhibiting osteoclastogenesis. On the other hand,
ing where the old or damaged bone matrix is they diminish the apoptosis of osteoblasts and
resorbed and replaced by newly formed tissue support the differentiation and maturation of
[1]. The bone remodeling process is responsible osteoblast precursor cells [4]. The decline of sys-
for the renewal of 5–10% of bone of the mature temic estrogen in women after menopause is a
skeleton each year [2]. Stretching and compres- primary cause of postmenopausal osteoporosis
sion of the bone tissue during locomotion or characterized by an impaired bone mass and
microcracks are detected by osteocytes, which increased risk of developing fragility fractures
are terminally differentiated osteoblasts and ful- [5]. Testosterone can be converted to estrogen by
fill a sensing function of mechanical forces and aromatization and also mediates bone-protective
metabolic signals in the bone microenvironment. effects [2]. Further important hormones with
A balanced remodeling is a hallmark of healthy divergent effects on bone cells and remodeling
are thyroid and parathyroid hormones, growth
hormone, and corticosteroids [2, 6, 7].
T. D. Rachner (*) · L. C. Hofbauer · A. Göbel
Department of Medicine III, Division for In addition to hormones and physical factors,
Endocrinology, Diabetes and Bone Disorders, multifaceted interactions of bone and immune
University Hospital “Carl Gustav Carus”, cells have been identified. For example, inflam-
Dresden, Germany
matory cytokines like tumor necrosis factor alpha
e-mail: tilman.rachner@uniklinikum-dresden.de

https://doi.org/10.1007/978-3-030-42958-4_36
504 T. D. Rachner et al.
(TNFα) or interleukins (IL)-1 and IL-6 directly precursors. Osteoprotegerin (OPG) is a decoy
activate osteoclastogenesis and osteoclastic bone receptor for RANKL. Physiologically, the ratio
resorption [8]. These mechanisms are key con- of RANKL and OPG determines the level of
tributors to bone damage and loss in rheumatoid osteoclast activity. In malignant bone disease, a
arthritis, an inflammatory joint disease [9]. local increase in RANKL and lower OPG levels
result in an imbalanced ratio promoting osteo-
clast activity [15].
Pathophysiology of Bone Metastases Tumor cells not only stimulate osteoclasto-
genesis but also impair the differentiation and
Bone is a site of metastasis in several human activation of bone-forming osteoblasts. The mat-
malignancies including breast, lung, and pros- uration of these cells is dependent on multiple
tate cancer [10]. The metastatic process under- pathways, of which the Wnt pathway is the most
lying the occurrence of bone lesions is complex, important. Wnt ligands mediate a complex cas-
requiring a tight interaction of cancer cells and cade of signaling processes within osteoblast pre-
cells from the bone microenvironment. The precursor cells which lead to the activation of key
disposition of certain cells to metastasize to genes of osteoblastogenesis. These steps are
bone has been well recognized for over a cen- strongly suppressed by cancer-derived inhibitors
tury. Stephen Paget was first to describe the of the Wnt pathway. One of the most prominent
“seed and soil” hypothesis in 1889 [11]. This examples in this regard is Dickkopf-1, which can
hypothesis is based on the assumption that be found in osteolytic bone metastases [16].
growth factors and cytokines stored within the
bone provide a growth- promoting microenvi-
ronment for cancer cells. The vicious cycle of Prevalence of Bone Metastases
bone metastases provides a simplified explana-
tion for the process of local bone destruction In general, up to 30% of patients with breast can-
and increased tumor growth in the bone, where cer will develop metastatic disease [17]. Bone
cancer cells that have successfully migrated to metastases are a common late-stage complication
the bone secrete factors that directly and indi- of prevalent malignancies like prostate or breast
rectly promote bone resorption by increasing cancer. But also other cancer entities like renal,
osteoclast activity and inhibiting osteoblasts lung, or thyroid have a high tendency to metasta-
[12]. The increased bone resorption in turn size to bone. The incidence of bone metastases in
results in an increased release of growth factors advanced stages of the disease is highly depen-
stored in the bone matrix, which promotes the dent on the primary tumor type. The highest risk
local proliferation of the tumor cells [13]. of developing bone metastases can be found in
Bone metastases can be differentiated patients with prostate cancer with a 5-year inci-
according to their radiographic morphology into dence of 24.5%, followed by lung (12.4%), renal
sclerotic or lytic lesions. While prostate cancer- (8.4%), and breast (6%) cancers [18].
derived bone lesions are often sclerotic, breast
cancer bone metastases are typically osteolytic.
While the bone quality is inferior in both cases, General Treatment Approach
patients with predominantly osteolytic lesions to Bone Metastases
are at a higher risk of fractures [14]. In osteo-
lytic bone lesions (Fig. 36.1), the RANKL/ To achieve an optimal outcome, bone metastases
RANK/osteoprotegerin (OPG) system is a major require a multidisciplinary treatment approach
contributor to the progression of disease. consisting of surgeons, radiologists, oncologists,
Receptor activator of nuclear factor kappa-Β as well as doctors specializing in pain and nuclear
ligand (RANKL) binds to its receptor RANK medicine as well as osteology. Following the
which is expressed on osteoclasts and osteoclast diagnosis of bone metastases, an individual treat-
36 Bone Metabolism in Cancer 505
Mesenchymal
stem cell
Hematopoietic
stem cell Osteoblast
precursor
RANKL
Osteoclast
precursor RANK
OPG
CIRCULUS
VITIOSUS
BONE Tumor cells
Osteoclast Osteoblast Osteocyte
Fig. 36.1 Effects of osteolytic tumor cells on bone osteoprotegerin (OPG), which is also produced by osteo-
remodeling. The balanced remodeling of the bone is blasts. Osteocytes are terminally differentiated osteo-
ensured by the tightly controlled actions of three main blasts which are embedded in the bone matrix and serve
cell types: osteoblasts, osteoclasts, and osteocytes. as mechanosensors and support osteoclastic differentia-
Osteoblasts originate from mesenchymal stem cells and tion by the production of RANKL. In the presence of
osteoblast precursors. They are responsible for the de osteolytic tumor cells, osteoclastic bone resorption is
novo formation of bone matrix. Bone-resorbing osteo- increased by tumor cell-secreted factors that favor the
clasts derive from hematopoietic stem cells and osteo- production of RANKL, while OPG production is dimin-
clast precursors. Apart from other factors, osteoclastic ished. Growth factors and calcium stored within the bone
differentiation is mainly dependent on the receptor acti- matrix which are released by resorption in turn support
vator of nuclear factor kappa-Β ligand (RANKL) which tumor cell growth within the bone (circulus vitiosus). In
is produced by osteoblasts and binds the receptor RANK addition, osteoblastogenesis and osteoblastic functions
on osteoclast precursors and mature osteoclasts. The are inhibited by tumor cell-derived factors
action of RANKL is limited by its natural antagonist
ment strategy should be provided to each patient offered to all patients. Pharmacological treatment
in an interdisciplinary tumor conference. Aspects options for bone metastases will be extensively
to consider when proposing a treatment scheme discussed in the following paragraphs.
are the localization, extent, and operability of the
lesion. In addition, age, general health, and con-
current medical conditions need also to be con- Pharmacological Approach
sidered. Bone metastases are still generally to Treating Patients with Bone
considered palliative, but increasingly curative Metastases
approaches for single lesions in otherwise healthy
patients are considered. While the initial deci- Pharmacological approaches to treat bone metas-
sions on operative or radiotherapy procedures are tases generally consist of a potent antiresorptive
required, pharmacological treatment should be therapy. Of note, antiresorptive therapy is
recommended independent of cancer entity or makes them very stable and robust.
morphologic appearance of the bone lesion. Two Pharmacokinetic properties are largely determined
main classes of antiresorptive drugs are currently by additional side groups. Aminobisphosphonates
approved to treat bone metastases, namely, are named after a nitrogen atom included in the
bisphosphonates and denosumab (Fig. 36.2). side group and are more potent than traditional
bisphosphonates. These exert their antiresorptive
properties by inhibiting the mevalonate pathway
Bisphosphonates [19]. Several bisphosphonates are now approved
for the treatment of bone metastases and/or
Bisphosphonates are a class of antiresorptive myeloma. Of these, zoledronic acid, which is
drugs that are widely used to treat benign and also considered the most potent bisphosphonate,
malignant bone conditions. Since their initial dis- has been most extensively investigated in clini-
covery in the 1960s, several generations of cal trials [20].
bisphosphonates have evolved with increasing
bone affinity and antiresorptive potency. Breast Cancer
All bisphosphonates share a P-C-P structure Several randomized trials have compared the
(two phosphate groups with a carbon atom), which use of bisphosphonates to placebo in preventing
Mesenchymal
stem cell
Denosumab
Hematopoietic
stem cell Osteoblast
precursor
RANKL
Osteoclast
precursor RANK
OPG
Bisphosphonates
BONE Tumor cells
Osteoclast Osteoblast Osteocyte
Fig. 36.2 Denosumab and bisphosphonates as antire- reducing osteoclastogenesis and bone resorption. (Amino)
sorptive therapies in osteolytic bone metastases. bisphosphonates are inhibitors of the mevalonate pathway
Denosumab is a neutralizing monoclonal antibody against and inducers of apoptosis in bone-resorbing osteoclasts.
receptor activator of nuclear factor kappa-Β ligand In addition, they exert direct antitumor effects on osteo-
(RANKL). It reduces the binding of RANKL to RANK on lytic tumor cells
osteoclast precursors and mature osteoclasts, thereby
skeletal-related events (SREs) in patients with The adjuvant use of bisphosphonates in breast
breast cancer. A reduction or significant delay of cancer has only yielded positive results in post-
SREs has been confirmed for clodronate, pami- menopausal women or those that had menopause
dronate, ibandronate, and zoledronic acid (sum- induced by GnRH analogues like goserelin. A
marized in [21]). Meta-analyses have confirmed recent meta-analysis supported the anticancer
the benefit of bisphosphonates in reducing the effect of bisphosphonates, with a decrease in the
risk for fractures, surgery, and hypercalcemia, incidence of bone recurrence by 34% and breast
but not for spinal compression [22]. Importantly, cancer-specific mortality by 17% [29]. Based on
effects were time-dependent, and treatment had these findings, the use of an antiresorptive ther-
to be performed for at least 6 months to see pos- apy should be considered in all postmenopausal
itive effects on skeletal morbidity outcomes women with early breast cancer. Adjuvant
[22]. Fewer trials were conducted to directly bisphosphonates in women with early BC are
compare different bisphosphonates. In a com- now recommended (joint position statement of
parative trial, intravenous pamidronate (90 mg, the IOF, CABS, ECTS, IEG, ESCEO IMS, and
monthly) appeared more effective than oral clo- SIOG) [27].
dronate in controlling bone symptoms and sup-
pressing bone resorption [23]. In breast cancer Prostate
patients, zoledronic acid was shown to be supe- Trials investigating bisphosphonates in meta-
rior to pamidronate in reducing the rate of SREs static prostate cancer are scarcer than in breast
in patients who had at least one osteolytic lesion cancer. In a study of 643 men with metastatic
(48% vs. 58%) and significantly reduced the prostate cancer, zoledronic acid significantly
time to first SRE (p = 0.013) [24]. reduced the risk of at least one SRE by 11% (39%
A Cochrane analysis came to the conclusion vs. 49%, p = 0.028) while reducing the overall
that bisphosphonates reduce the risk of SREs in risk of skeletal complications by 36% after
women with breast cancer and clinically evident 24 months [30]. In another trial, the effect of a
bone metastases [25]. In addition, a more recent single infusion of ibandronate (6 mg) was com-
meta-analysis concludes that bisphosphonates pared against single-dose radiotherapy with
provided an overall survival benefit independent regard to bone pain. At 4–12 weeks, the pain
of bone metastases (HR 0.91, 95% CI 0.83–0.99; response was not statistically different, with
p = 0.04). However, subgroup analysis by meno- comparable reductions in pain score in both
pausal status showed a survival benefit from groups [31].
bisphosphonates in postmenopausal women only
(HR 0.77, 95% CI 0.66–0.90; p = 0.001) but no Myeloma
survival benefit for premenopausal women (HR As a systemic hematological disease derived
1.03, 95% CI 0.86–1.22; p = 0.78) [26]. from the bone marrow, the pathogenesis of mul-
An important aspect in breast cancer patients tiple myeloma is distinctly different from solid
with hormone receptor-positive tumors is the tumors. However, osteolytic lesions are a hall-
negative impact that adjuvant cancer therapies mark of multiple myeloma [32]. Several larger
may exert on bone health [27]. In postmeno- studies have investigated the use of different
pausal women, aromatase inhibitors are cur- bisphosphonates in patients with multiple
rently used to achieve maximal suppression of myeloma.
residual estrogen levels. While effective in Clodronate represents the only non-
reducing the risk of disease relapse, aromatase aminobisphosphonate approved for the treatment
inhibitors cause a rapid decline in bone mass and of lytic bone lesions derived from myeloma.
increase fracture risk. Several studies have inves- Clinical trials in the 1990s using clodronate were
tigated the adjuvant use of antiresorptive agents the first to confirm a reduction in SREs [33, 34].
in this setting [28]. These trials did not show a general survival
benefit for clodronate, although subset analyses To reduce the risk of these complications, regu-
revealed that among the subgroup with no s keletal lar assessment of renal function as well as ade-
fractures at presentation, survival was signifi- quate calcium supplementation is recommended.
cantly improved (59 vs. 37 months, p = 0.006) Osteonecrosis of the jaw (ONJ) is a recognized
[35]. While oral pamidronate (300 mg/daily) complication of bisphosphonate therapy. While
failed to reduce SREs [36], intravenous pamidro- the prevalence in the osteoporosis setting is
nate significantly reduced the rate of skeletal very low, about 1–5% of patients treated for
events compared to placebo (24% vs. 41%, bone metastases may develop ONJ [43]. To
p < 0.001). Survival in a subgroup of patients decrease the risk of ONJ, dental assessment
with more advanced disease was significantly prior to the initiation of therapy and good den-
longer (median survival 21 vs. 14 months, tal hygiene in addition to perioperative antibiot-
p = 0.041) [37]. Several trials have investigated ics are recommended.
zoledronic acid in multiple myeloma. Compared
to pamidronate, zoledronic acid was at least com-
parable in reducing the rate of SREs [38]. Denosumab
Compared to clodronate, zoledronic acid reduced
the risk of SREs by 26%. Importantly, zoledronic Denosumab is a monoclonal antibody against
acid reduced the rate of SREs in patients with and RANKL. Denosumab is approved for the treat-
without detectable bone lesions [39]. ment of osteoporosis (60 mg every 6 months) and
Furthermore, zoledronic acid in addition to stan- for the treatment of bone metastases secondary to
dard therapy reduced the risk of death by 16% solid cancers and myeloma (120 mg every
(p = 0.012) and prolonged median overall sur- 4 weeks). Following successful phase 2 trials,
vival by 5.5 months from 44.5 to 50.0 months denosumab was investigated in three large phase
[39]. Meta-analyses of different bisphosphonates 3 trials which led to the initial approval of deno-
showed no overall survival benefit for individual sumab for the treatment of solid tumors.
agents, but zoledronic acid was superior to pla- Denosumab was assessed in a head-to-head
cebo (and etidronate) in improving survival [40]. trial against zoledronic acid in patients with pros-
Based on their clear efficacy in reducing SREs, it tate cancer. In 1904 men with metastatic prostate
is recommended to consider bisphosphonates in cancer, denosumab was superior to zoledronic
all myeloma patients [41]. acid in delaying the time to first and subsequent
SREs. The time to first SRE was delayed by
Safety of Bisphosphonates 3.6 months (17.1 vs. 20.7 months; HR: 0.82,
Bisphosphonates are generally considered as 95%CI 0.71–0.95; p = 0.008 for superiority) in
safe and well-tolerated drugs. When discussing the denosumab arm. Denosumab significantly
adverse events, it is important to distinguish reduced the risk of first and subsequent SREs by
between events that occur in the lower doses 18% compared to zoledronic acid (p = 0.001 for
used for the treatment of osteoporosis and those superiority) [44]. Overall survival was similar in
that occur in patients treated with higher doses the two treatment arms. In a separate breast can-
for bone metastases. Following parental appli- cer trial, 2046 patients were randomized to receive
cation, some patients may experience an acute- denosumab or zoledronic acid. Denosumab was
phase reaction with flu-like symptoms. superior to zoledronic acid in delaying time to
Acute-phase reactions are described in up to first on-study SRE (HR 0.82; 95%CI 0.71–0.95;
one-third of patients and are caused by the acti- p = 0.01 superiority) and time to first and subse-
vation and proliferation of gamma-delta-T-cells quent (multiple) on-study SREs (p = 0.001).
[42]. Self-limiting symptoms generally resolve Overall survival and disease progression were
within 1–2 days and occur predominantly after similar between groups [45].
the first infusion. Bisphosphonates increase the In a third trial, denosumab was compared to
risk of renal complications and hypocalcemia. zoledronic acid in the treatment of bone metasta-
ses in patients with advanced solid cancer other denosumab in the adjuvant treatment of breast
than breast and prostate cancer and myeloma. cancer. The ABCSG-18 trial compared the effects
Denosumab was non-inferior to zoledronic acid of denosumab or placebo on fracture incidence in
in preventing or delaying first on-study SRE (HR 3425 postmenopausal breast cancer patients on
0.84; 95% CI, 0.71–0.98; p = 0.0007). No superi- aromatase inhibitor therapy. Denosumab signifi-
ority for denosumab could be shown in this trial, cantly reduced the risk of any clinical fracture
and a post hoc analysis revealed a potential nega- (HR 0.50; 95% CI 0.39–0.65; p < 0.0001).
tive effect for the myeloma subgroup [46]. Interestingly, the reduction in fracture rate was
Notably, exploratory analyses of this trial seen in both patients with and without baseline
revealed a significant survival benefit for patients T-scores below −1 as well as age above and
with lung cancer (non-small cell lung cancer and below 65 [49]. Recently, data from the D-Care
small cell lung cancer) treated with denosumab. trial (NCT01077154), which aimed to establish
Denosumab was associated with improved the ability of denosumab to prevent the occur-
median overall survival versus ZA in 811 patients rence of bone metastases in BC patients with a
with any lung cancer (8.9 vs. 7.7 months; hazard high risk of developing metastatic bone disease,
ratio [HR] 0.80) and in 702 patients with NSCLC were presented. In this trial, RANKL inhibition
(9.5 vs. 8.0 months; HR 0.78) (p = 0.01, each with denosumab failed to reduce the rate of bone
comparison). Further analysis of NSCLC by his- metastases [50].
tological type showed a median survival of
8.6 months for denosumab versus 6.4 months for
ZA in patients with squamous cell carcinoma Summary
(HR 0.68; p = 0.035). Based on these results, a
subsequent myeloma trial was initiated. In this The pathophysiology of bone metastases is
trial, 1718 patients were randomized to receive complex and treatment requires an interdisci-
subcutaneous denosumab 120 mg plus intrave- plinary approach. Antiresorptive therapy is an
nous placebo every 4 weeks or intravenous zole- essential aspect of providing optimal treatment.
dronic acid 4 mg plus subcutaneous placebo Bisphosphonates and denosumab have proven
every 4 weeks. The primary endpoint, non- efficacy in reducing the occurrence of SREs and
inferiority of denosumab to zoledronic acid for improving life quality by targeting the bone-
time to first skeletal-related event was achieved resorbing osteoclast. Future research will need
(HR 0.98, 95% CI 0.85–1.14; non-inferiority to identify how metastatic cells home to bone
p = 0·01). Notably, renal toxicity was reported in and engraft with the aim of identifying both pre-
less (10%) patients in the denosumab group com- ventative and therapeutic strategies.
pared to 17% in the zoledronic acid group [47].
ptimizing Bone Health in Patients

O References
with Cancer at High Risk 1. Sims NA, Martin TJ. Coupling the activities of bone
of Developing Bone Metastases formation and resorption: a multitude of signals
within the basic multicellular unit. Bonekey Rep.
In patients with nonmetastatic prostate cancer, 2014;3:1–10.
2. Walsh JS. Normal bone physiology, remodelling and
the adjuvant use of denosumab (120 mg monthly)
its hormonal regulation. Surgery. 2014;33:1–6.
delayed the occurrence of bone metastases by a 3. Crockett JC, Rogers MJ, Coxon FP, Hocking LJ,
median of 4.2 months (HR 0.085, 95%CI 0.71– Helfrich MH. Bone remodelling at a glance. J Cell
0.98, p = 0.032) [48]. While overall survival did Sci. 2011;124:991–8.
4. Almeida M, Laurent MR, Dubois V, Claessens F,
not differ between denosumab and placebo, the
Brien CAO, Bouillon R, et al. Estrogens and andro-
rate of ONJ was significantly increased by deno- gens in skeletal physiology and pathophysiology.
sumab (5% vs. 0%). Two trials have investigated Physiol Rev. 2017;97:135–87.
5. Rachner TD, Khosla S, Hofb LC. Osteoporosis: now 23.

Jagdev S, Purohit P, Herling C, Coleman
and the future. Lancet. 2011;377:1276–87. R. Comparison of the effects of intravenous pami-
6. Lips P, Van Schoor NM. The effect of vitamin D on dronate and oral clodronate on symptoms and bone
bone and osteoporosis. Best Pract Res Clin Endocrinol resorption in patients with metastatic bone disease.
Metab. 2011;25:585–91. Ann Oncol. 2001;12:1433–8.
7. Bassett JHD, Williams GR. Role of thyroid hormones 24. Rosen LS, Gordon DH, Dugan W, Major P, Eisenberg
in skeletal development and bone maintenance. PD, Provencher L, et al. Zoledronic acid is superior to
Endocr Rev. 2016;37:135–87. pamidronate for the treatment of bone metastases in
8. Schett G. Effects of inflammatory and anti- breast carcinoma patients with at least one osteolytic
inflammatory cytokines on the bone. Eur J Clin lesion. Cancer. 2003;100:36–43.
Investig. 2011;41:1361–6. 25. Pavlakis N, Rl S. Bisphosphonates for breast cancer
9. Smolen JS, Redlich K, Zwerina J, Aletaha D, Steiner (review). Cochrane Database Syst Rev. 2005;3:1–49.
G, Schett G. Pro-inflammatory cytokines in rheuma- 26. O’Carrigan B, Wong M, Willson M, Stockler M,

toid arthritis pathogenetic and therapeutic aspects. Pavlakis N, Goodwin A. Bisphosphonates and other
Clin Rev Allergy Immunol. 2005;28:239–48. bone agents for breast cancer (review). Cochrane
10. Mundy GR. Metastasis to bone: causes, consequences Database Syst Rev. 2017;10:CD003474.
and therapeutic opportunities. Nat Rev Cancer. 27. Hadji P, Aapro MS, Body J, Gnant M, Luisa M, Yves
2002;2:584–93. J, et al. Management of aromatase inhibitor-associated
11. Paget S. The distribution of secondary growths in can- bone loss (AIBL) in postmenopausal women with
cer of the breast. Lancet. 1889;133:571–3. hormone sensitive breast cancer: joint position state-
12. Hofbauer LC, Rachner T, Singh SK. Fatal attraction: ment of the IOF, CABS, ECTS, IEG. J Bone Oncol.
why breast cancer cells home to bone. Breast Cancer 2017;7:1–12.
Res. 2008;10:101. 28. Rachner T, Coleman R, Hadji P, Hofbauer L. Bone
13. Roodman GD. Mechanisms of bone metastasis. N health during endocrine therapy for cancer. Lancet
Engl J Med. 2004;350:1655–64. Diabetes Endocrinol. 2018;6(11):901–10.
14. Guise TA, Mohammad KS, Clines G, Stebbins EG, 29. Early Breast Cancer Trialists’ Collaborative Group
Wong DH, Higgins LS, et al. Basic mechanisms E. Adjuvant bisphosphonate treatment in early breast
responsible for osteolytic and osteoblastic bone cancer: meta-analyses of individual patient data from
metastases. Clin Cancer Res. 2006;12:6213–7. randomised trials. Lancet. 2015;386:1353–61.
15. Weilbaecher KN, Guise TA, LK MC. Cancer to bone: 30. Saad F, Gleason DM, Murray R, Venner P, Lacombe
a fatal attraction. Nat Rev Cancer. 2011;11:411–25. L, Chin JL, et al. Long-term efficacy of zoledronic
16. Rachner TD, Göbel A, Benad-Mehner P, Hofbauer acid for the prevention of skeletal complications in
LC, Rauner M. Dickkopf-1 as a mediator and novel patients with metastatic hormone- refractory prostate
target in malignant bone disease. Cancer Lett. cancer for the zoledronic acid prostate cancer study
2014;346:172–7. group. J Natl Cancer Inst. 2004;96:879–82.
17. Soni A, Ren Z, Hameed O, Chanda D, Morgan CJ, 31. Hoskin P, Sundar S, Reczko K, Forsyth S, Mithal N,
Siegal GP, et al. Breast cancer subtypes predispose Sizer B, et al. A multicenter randomized trial of iban-
the site of distant metastases. Am J Clin Pathol. dronate compared with single-dose radiotherapy for
2015;143:471–8. localized metastatic bone pain in prostate cancer. J
18. Hernandez RK, Wade SW, Reich A, Pirolli M, Liede Natl Cancer Inst. 2015;107:1–9.
A, Lyman GH. Incidence of bone metastases in 32. van Driel M, van Leeuwen JPTM. Cancer and

patients with solid tumors: analysis of oncology elec- bone: a complex complex. Arch Biochem Biophys.
tronic medical records in the United States. BMC 2014;561:159–66.
Cancer. 2018;18:1–11. 33. Lahtinen R, Laakso M, Palva I, Virkkunen P, Elomaa
19. Luckman SP, Hughes DE, Coxon FP, Graham
I. Randomised, placebo-controlled multicentre
R, Russell G, Rogers MJ. Nitrogen-containing trial of clodronate in multiple myeloma. Lancet.
bisphosphonates inhibit the mevalonate pathway 1992;340:1049–52.
and prevent post-translational prenylation of GTP- 34. McCloskey E, MacLennan I, Drayson M, Chapman
binding proteins, including Ras. J Bone Miner Res. C, Dunn J, Kanis J. A randomized trial of the effect of
1998;13(4):581–9. clodronate on skeletal morbidity in multiple myeloma.
20. Polascik T, Mouraviev V. Zoledronic acid in the man- Br J Haematol. 1998;100:317–25.
agement of metastatic bone disease. Ther Clin Risk 35. McCloskey EV, Dunn JA, Kanis JA, MacLennan ICM,
Manag. 2008;4:261–8. Drayson MT. Long-term follow-up of a prospective,
21.
Petrut B, Simmons C, Broom R, Trinkaus double-blind, placebo-controlled randomized trial
M. Pharmacotherapy of bone metastases in in of clodronate in multiple myeloma. Br J Haematol.
breast cancer patients. Expert Opin Pharmacother. 2001;113:1035–43.
2008;6566:937–45. 36. Brincker H, Westin J, Abildgaard N, Gimsing P,

22. Ross JR, Saunders Y, Edmonds PM, Patel S, Broadley Turesson I, Hedenus M, et al. Failure of oral pami-
KE, Johnston SRD. Systematic review of role of dronate to reduce skeletal morbidity in multiple
bisphosphonates on skeletal morbidity in metastatic myeloma: a double-blind placebo-controlled trial. Br
cancer. BMJ. 2003;327:1–6. J Haematol. 1998;101:280–6.
37. Berenson J, Lichtenstein A, Porter L, Dimopoulos M, treatment of bone metastases in men with castration-
Bordoni R, George S, et al. Efficacy of pamidronate resistant prostate cancer: a randomised, double-blind
in reducing skeletal events in patients with advanced study. Lancet. 2011;377:813–22.
multiple myeloma. N Engl J Med. 1996;334:488–93. 45. Stopeck AT, Lipton A, Body J-J, Steger GG, Tonkin
38. Rosen L, Gordon D, Kaminski M, Howell A, Belch K, de Boer RH, et al. Denosumab compared with
A, Mackey J, et al. Zoledronic acid versus pamidro- zoledronic acid for the treatment of bone metastases
nate in the treatment of skeletal metastases in patients in patients with advanced breast cancer: a randomized,
with breast cancer or osteolytic lesions of multiple double-blind study. J Clin Oncol. 2010;28:5132–9.
myeloma: a phase III, double-blind, comparative trial. 46. Henry DH, Costa L, Goldwasser F, Hirsh V, Hungria
Cancer J. 2001;7:377–87. V, Prausova J, et al. Randomized, double-blind study
39. Morgan GJ, Child JA, Gregory WM, Szubert AJ,
of denosumab versus zoledronic acid in the treatment
Cocks K, Bell SE, et al. Effects of zoledronic acid of bone metastases in patients with advanced cancer
versus clodronic acid on skeletal morbidity in patients (excluding breast and prostate cancer ) or multiple
with newly diagnosed multiple myeloma (MRC myeloma. J Clin Oncol. 2011;29:1125–32.
myeloma IX): secondary outcomes from a randomised 47. Raje N, Terpos E, Willenbacher W, Shimizu K, García-
controlled trial. Lancet Oncol. 2011;12:743–52. sanz R, Durie B, et al. Denosumab versus zoledronic
40. Mhaskar R, Redzepovic J, Wheatley K, Oac C,
acid in bone disease treatment of newly diagnosed
Miladinovic B, Glasmacher A, et al. Bisphosphonates multiple myeloma: an international, double- blind,
in multiple myeloma: a network meta- analysis double-dummy, randomised, controlled, phase 3
(review). Cochrane Database Syst Rev. 2012;5:1–99. study. Lancet Oncol. 2018;19:370–81.
41. Terpos E, Morgan G, Dimopoulos MA, Drake MT, 48. Smith MR, Saad F, Coleman R, Shore N, Fizazi K,
Lentzsch S, Shimizu K, et al. International myeloma Tombal B, et al. Denosumab and bone-metastasis-
working group recommendations for the treatment free survival in men with castration-resistant prostate
of multiple myeloma – related bone disease. J Clin cancer: results of a phase 3, randomised, placebo-
Oncol. 2013;31:2347–59. controlled trial. Lancet. 2012;379:39–46.
42. Kunzmann V, Bauer E, Feurle J, Weissinger F, Tony 49. Gnant M, Pfeiler G, Dubsky PC, Hubalek M, Greil
HP, Wilhelm M. Stimulation of gammadelta T cells R, Jakesz R, et al. Adjuvant denosumab in breast
by aminobisphosphonates and induction of anticancer (ABCSG-18): a multicentre, randomised,
plasma cell activity in multiple myeloma. Blood. double-blind, placebo-controlled trial. Lancet.
2000;96:384–92. 2015;6736:1–11.
43. Reyes C, Hitz M, Prieto-alhambra D, Abrahamsen 50. Coleman R, Finkelstein D, Barrios C, Martin M, Iwata
B. Risks and benefits of bisphosphonate therapies. J H, Glaspy J, et al. Adjuvant denosumab in early breast
Cell Mol Med. 2016;28:20–8. cancer: first results from the international multicenter
44. Fizazi K, Carducci M, Smith M, Damião R, Brown J, randomized phase III placebo controlled D-CARE
Karsh L, et al. Denosumab versus zoledronic acid for study. J Clin Oncol. 2018;36:(suppl; abstr 501).
Systemic Therapies for Patients
with Metastatic Spinal Disease
37
Panagiotis J. Vlachostergios and Ashish Saxena
Epidemiology formance status, the extent of metastatic disease,

and the primary tumor type [2, 3].
The spine can be considered to be composed of In general, the comprehensive assessment of
the spinal cord of the central nervous system these parameters requires the involvement of
(CNS) and the bony vertebral column. Spinal multiple different specialties, including surgery
metastases occur in 30–90% of patients with (neurosurgery, orthopedic surgery), radiation
malignant tumors of advanced stage. Of those, oncology, medical oncology, interventional radi-
20% ultimately result in symptomatic compres- ology, pain and rehabilitation medicine, and oth-
sion of the spinal cord [1, 2]. The most common ers. To ensure the optimal use of treatment in the
location of spinal metastatic disease is the tho- right patient at the right time, a framework of
racic spine (70%), followed by the lumbar (20%) decision points on four sentinel aspects of the
and cervical (10%) spine [1, 2]. disease was developed, also known as NOMS
(neurologic, oncologic, mechanical instability,
systemic disease) [1]. The role of the initial onco-
ncologic Parameters Affecting
O logic assessment within the NOMS framework is
Treatment to determine the likelihood of response to radia-
tion and systemic therapy [1]. According to the
The management of spinal metastases with sys- established algorithm, radiosensitive tumors may
temic therapy is essentially palliative, focusing be effectively addressed by external beam radia-
on symptomatic relief, preserving neurological tion therapy (EBRT) or stereotactic radiosurgery
function, and restoring or maintaining the stabil- (SRS), whereas radioresistant or previously radi-
ity of the spine [2, 3]. Beyond the acuity and ated tumors usually necessitate a surgical
severity of the presenting symptoms, several approach first, particularly in the presence of
other clinical characteristics are involved in cord compression and/or myelopathy [1, 4].
determining the most appropriate therapeutic Systemic anticancer therapies include classic
management, including the patient’s age and per- chemotherapeutic agents, targeted therapies, and
immunotherapy which promotes the activity of a
patient’s native immune system against their
P. J. Vlachostergios · A. Saxena (*) tumor (at present most commonly represented by
Department of Medicine, Division of Hematology & programmed death-1 (PD-1) or programmed
Medical Oncology, NewYork-Presbyterian Hospital/ death-ligand 1 (PD-L1) immune checkpoint
inhibitors). Despite recent advances, systemic
e-mail: ahs9018@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_37
514 P. J. Vlachostergios and A. Saxena
therapy is not yet clearly integrated into the regimens [7]. In the advanced setting, which is
NOMS framework. This apparent discrepancy most relevant for the development and progres-
may be due to multiple reasons. First, chemother- sion of bone disease, including spinal metastases,
apy is generally considered to be more effective taxane-based regimens or capecitabine are the
for visceral than for osseous disease [1]. most commonly used agents whenever visceral
Intriguingly, several chemotherapeutic drugs disease is also present and a rapid tumor regres-
may actually negatively affect bone health by act- sion is needed [8]. Other approved chemotherapy
ing through different mechanisms, including drugs include gemcitabine, vinorelbine, eribulin,
inhibition of osteoblasts (doxorubicin, cyclo- pemetrexed, and platinum agents [8]. Despite the
phosphamide), stimulation of osteoclasts (cyclo- fact that several trials have reported on the activ-
sporine), reduction of vitamin D levels (5FU, ity of these agents (single or combined), there is
leucovorin), phosphate wasting through the kid- a paucity of contemporary studies on bone-
neys (ifosfamide), hypomagnesemia (platinum specific effects of chemotherapy in metastatic
agents), and ovarian suppression-induced meno- breast cancer. Older cumulative data from 12 tri-
pause [5]. Additionally, there are only limited, als supported a tendency for soft tissue lesions to
premature data on the effect of newer systemic have a higher response rate (55–60%) compared
therapies on bone metastases. to visceral and bone metastases (31–44%) [9]. In
the era of targeted therapies, including HER-2-
directed therapies, anti-hormonal agents, and cell
I mpact of Systemic Anticancer cycle inhibitors (discussed separately in
Therapies on Spinal Metastatic “Vertebral Metastases and Targeted Therapy”
Disease section), patient selection is becoming key in
order to improve site-specific and overall antitu-
Systemic therapies have transformed the natural mor responses.
course of several types of cancers. These agents Prostate cancer (PC) is also responsive to che-
are used to treat solid tumor metastases including motherapy, particularly to taxanes. Docetaxel is
sites in the spinal cord and vertebral column. an established treatment for both castration-
However, the cord- and bone-specific effects of resistant prostate cancer (CRPC) and high-
these drugs on the outcomes of metastatic tumors volume hormone-sensitive prostate cancer
are yet to be fully determined. This is particularly (HSPC) [10, 11]. In updated results from the
true for metastases to the spinal cord, where local CHAARTED trial, HSPC patients with high-
therapies such as radiation and surgery currently volume disease (defined as presence of visceral
have a more significant role. metastases and/or ≥ four bone metastases with at
least one outside of the vertebral column and pel-
vis) had a significant survival benefit from treat-
Chemotherapy ment with docetaxel plus androgen deprivation
therapy (ADT) compared with ADT alone.
Chemotherapeutic agents are usually effective in However, those with low-volume disease did not
treating chemosensitive tumors with or without have an overall survival (OS) benefit with the
spinal cord or osseous metastases, such as small addition of docetaxel, providing evidence for the
cell lung cancer (SCLC), Ewing’s sarcoma, and activity of the drug in osseous disease [11].
neuroblastoma [3]. However, their role is limited Cabazitaxel is approved for the treatment of met-
in chemoresistant malignancies. The most com- astatic CRPC after progression on docetaxel [12].
mon primary tumors that demonstrate bone tro- Because the majority of patients studied in trials
pism are breast, prostate, and lung cancers [6]. with these drugs had osseous metastases, charac-
Breast cancer is sensitive to several chemo- terizing the effect of chemotherapy on markers of
therapeutic agents, including anthracyclines and bone metabolism became a relevant endpoint. In
taxanes, which comprise the backbone of many a prospective randomized study of docetaxel/
37 Systemic Therapies for Patients with Metastatic Spinal Disease 515
estramustine versus the bisphosphonate zole- Table 37.1 Examples of chemotherapy agents with CNS
dronic acid (discussed in “EGFR-Directed penetration
Therapies” section), no significant difference was Drug Tumor type used for
found in median change of any measured bone Temozolomide Glioma, melanoma, breast cancer,
SCLC
turnover markers in patients given zoledronic
Ifosfamide Ewing sarcoma, lymphoma
acid when compared to chemotherapy, including Pemetrexed NSCLC
interleukin-6 (IL-6), urinary deoxypyridinoline Topotecan SCLC
to serum creatinine ratio (DpD), tartrate-resistant Methotrexate ALL
acid phosphatase (TRAPC), bone-specific alka- Capecitabine Breast cancer
line phosphatase (BAP), intact osteocalcin
(OCN), osteoprotegerin (OPG), and ligand for
receptor activator of nuclear-factor κB (RANKL) In patients with cancer metastatic to the CNS,
[13]. This suggests that docetaxel may lead to a of which the spinal cord is a component, a num-
similar degree of bone turnover as zoledronic ber of chemotherapy drugs have activity by virtue
acid in CRPC bone metastases. Additionally, of their ability to penetrate into this part of the
there was a significant reduction of IL-6 levels by body (Table 37.1). Nevertheless, this does not
35% in prostate-specific antigen (PSA) respond- often translate into an increase in overall survival
ers compared to non-responders, suggesting that as chemotherapeutics usually lack sufficient con-
IL-6 could serve as a surrogate for clinical centrations for adequate duration to cause a sig-
response of bone metastases to docetaxel-based nificant antitumor effect [18]. Additionally, with
chemotherapy [13]. Within the same context, few exceptions (e.g., high-dose methotrexate for
CaBone is an ongoing single-arm phase II study prevention of acute lymphoblastic leukemia
of cabazitaxel in metastatic CRPC patients with (ALL) CNS relapse), the use of CSF drug con-
bone-only metastases assessing bone progression- centrations as a surrogate for brain and/or tumor
free survival as the primary endpoint as well as penetration is not informative due to the com-
several bone-specific secondary endpoints, plexity of the CSF barriers and drug kinetics. It
including time to skeletal-related event (SRE), will be critical for future studies to determine a
time to bone pain progression, bone pain more precise assessment of CNS extent of dis-
response, and bone turnover markers (ALP, bone ease and to optimize systemic and/or intrathecal
ALP, LDH, serum CTx, iPTH, and 1,25 dosing to improve the likelihood of prolonging
(OH)2D3) [14]. survival in this setting.
Newly diagnosed, metastatic non-small cell
lung cancers (NSCLC) with bone metastases and
no targetable driver mutations (such as in EGFR ertebral Metastases and Targeted
V
or ALK), and low (<50%) expression of PD-L1, Therapy
are commonly treated with platinum-based che-
motherapy doublets. The standard of care was to Hormonal Therapy
use these regimens alone, until the most recent
phase III trials which showed improved outcomes Aromatase Inhibitors
by combining platinum-based chemotherapy Aromatase inhibitors (AIs) play an important
with the checkpoint inhibitor pembrolizumab, role in the treatment of postmenopausal women
regardless of PD-L1 expression on tumors [15, with estrogen or progesterone receptor-positive
16]. Although data on specific effects of chemo- breast carcinoma, both in the adjuvant and meta-
therapy on bone metastases from NSCLC are static settings. However, natural bone loss is
limited or not reported, patients with bone metas- accelerated due to AI-induced estrogen deple-
tases who receive chemotherapy have a more tion. Additionally, there is an inherent difficulty
favorable prognosis (11.4 vs. 7.5 months median in evaluation of response of osseous metastases
overall survival) [17]. with current imaging modalities (bone scan, CT,
MRI) [19]. The % change in the maximum stan- PC in both castration-sensitive and resistant set-
dardized uptake value (SUVmax) on PET/CT at tings. A post hoc analysis of the landmark
8 weeks compared to baseline, either with [18F] COU-AA-301 phase III trial compared the time
fluorodeoxyglucose (FDG) or [18F]-fluoride to first occurrence of skeletal-related events, spi-
(NaF), appears to be the most promising metric nal cord compression, palliative radiation to
as a predictor of clinical progression-free sur- bone, or bone surgery in metastatic CRPC
vival (PFS) by week 24; however, validating patients who received AA plus prednisone vs.
studies are warranted [20]. Interlesional response prednisone alone [25]. AA plus prednisone
heterogeneity, flare phenomena, and non-FDG resulted in faster and more pronounced allevia-
avid osseous metastases, which may occur in up tion of skeletal pain and prolonged time to occur-
to 40% of cases, are remaining challenges that rence of first skeletal-related event compared to
need to be overcome [21, 22]. prednisone alone (25 vs. 20 months) [25]. The
PREVAIL study was another key study that
Androgen-Directed Therapies resulted in the approval of the androgen receptor
Luteinizing hormone-releasing hormone inhibitor enzalutamide for metastatic CRPC [26].
(LHRH) agonists and antagonists are an essen- When focusing on bone-specific endpoints,
tial component of ADT for advanced PC. In most enzalutamide significantly delayed the occur-
patients, ADT results in regression of metastases rence of first SRE and self-reported pain in
and in serum prostate-specific antigen (PSA) chemotherapy-naive men with metastatic CRPC
response. For skeletal metastases, the total area [26]. Overall, both abiraterone and enzalutamide
of hot spots on bone scan (<3 vs. ≥3 lumbar ver- have demonstrated the ability to delay bone pro-
tebral bodies) combined with the percentage of gression resulting in improvements in bone-
the total scan (<75% vs. ≥75% or superscan) has related endpoints in patients with metastatic
been used to calculate a score, termed the CRPC [27]. This benefit correlates at the molecu-
Soloway score. The latter was prognostic of lar level with pretreatment tumor nuclear AR
early ADT failure, defined as death from meta- overexpression (>75%) and CYP17 expression
static prostate cancer within 12 months after the (>10%) [28].
start of ADT [23]. At present, the most widely
used imaging marker of bone response is HER2-Directed Therapies
included in the Prostate Cancer Clinical Trials Bone metastases are equally common in all
Working Group 3 criteria [24]. Progression of breast cancer subtypes, including HER2-positive
osseous metastatic disease by these criteria disease [29]. There is indirect evidence to sup-
requires at least two new lesions on first post- port a positive role for anti-HER2 monoclonal
treatment scan, with at least two additional antibodies (trastuzumab) or small molecule
lesions on the next scan (2 + 2 rule). In contrast, inhibitors (lapatinib) in the outcome of these
changes in intensity of uptake alone do not con- patients. Presence of bone metastases is associ-
stitute either progression or regression [24]. ated with long-term survival in patients with
Similar to endocrine therapies for breast can- metastatic breast cancer who receive anti-HER2
cer, the benefits of ADT in bone metastatic PC treatment [30].
come at the cost of secondary osteoporosis,
which has a variable prevalence across different
studies (9–53%) but highlights the importance of EGFR-Directed Therapies
preventative and early diagnostic approaches
(calcium/vitamin D supplementation, exercise, The pattern of metastases at diagnosis is associ-
bone densitometry) as part of the standard care of ated with the tumor molecular status in certain
these patients [24]. cancers [31]. In NSCLC, more than half (54%) of
Abiraterone acetate (AA) is an androgen bio- stage IV patients with EGFR mutations (specifi-
synthesis inhibitor with strong activity against cally exon 19 deletions and exon 21 L858R

m utations) have bone metastases [30]. from NSCLC. This was evidenced by prolonga-
Additionally, the presence of mutations in onco- tion of bone-specific time-to-progression and
genes including EGFR, ALK, MET, and ROS1 in reduction in the frequency of SREs in Bev-treated
NSCLC patients with spinal metastases is associ- patients compared to non-Bev-treated patients
ated with increased overall survival [32]. [38]. In patients with bone-predominant meta-
Targeting these alterations may be beneficial for static breast cancer, the level of soluble VEGF
prevention of skeletal-related events. Indeed, in a receptor 2 (VEGFR2) was prognostic of OS in
retrospective analysis of NSCLC patients with those treated with the combination of the VEGF
bone metastases, EGFR mutation status was pre- receptor targeting agent vandetanib and endo-
dictive of treatment efficacy with an EGFR tyro- crine therapy with fulvestrant [39].
sine kinase inhibitor (TKI) [33]. The addition of An interesting observation in patients with
bisphosphonates (discussed in “Bone-targeted vertebral metastases of different primaries
Therapies” section) to EGFR TKIs can further (breast, lung, kidney) undergoing palliative
enhance their antitumor effect in patients with radiotherapy (RT) is that concomitant use of bev-
EGFR-mutated NSCLC and bone metastases, acizumab is tolerable. Thus, patients already
supported by a longer PFS compared to patients receiving bevacizumab as part of their systemic
only receiving EGFR TKIs [34]. Continuation of antitumor treatment should not be excluded from
the EGFR TKIs beyond skeletal progression of emergency RT if indicated [40].
pre-existing lesions in such patients, combined
with adequate local treatment, results in long
post-skeletal metastasis progression survival, Bone-Targeted Therapies
which is more evidence of the activity of anti-
EGFR therapy in this setting [35]. The third-generation bisphosphonate zoledronic
acid (ZA) and the monoclonal antibody against
the receptor activator of nuclear factor kappa-Β
Anti-angiogenic Therapies ligand (RANKL) are the two most widely used
systemic agents for targeting bone metastases in
The paradigm of angiogenesis-driven tumors is several different cancers. Mechanistically, they
renal cell carcinoma (RCC). Despite the fact that impair osteoclast-mediated bone resorption and
the presence of bone metastases is an adverse reduce tumor-associated osteolysis [41].
prognostic factor in metastatic RCC patients, the Numerous studies have demonstrated a statisti-
use of vascular endothelial growth factor (VEGF) cally significant reduction in the rate of patho-
pathway inhibitors (sunitinib, pazopanib, logic fractures, pain and analgesic consumption,
axitinib, cabozantinib, lenvatinib, sorafenib) in and improvement in the quality of life with these
the TKI era has significantly improved the agents compared to placebo [42].
median OS compared to historical controls or In a pivotal trial comparing ZA with the
other systemic therapies in the pre-TKI era [36]. second- generation bisphosphonate pamidronate
Notably, concurrent use of VEGF pathway- in patients with metastatic breast cancer or
targeting TKIs with bisphosphonates (discussed myeloma, fewer patients treated with the former
in “Bone-Targeted Therapies” section) in meta- required RT (19 vs. 24%) [42]. An additional
static RCC patients with bone metastases can fur- 25% reduction in the mean % of annual SREs
ther prolong median PFS and OS [37]. and skeletal morbidity was seen with ZA, com-
Less is known about the potential priming role pared to that achieved by pamidronate [43]. The
of antiangiogenic therapy in bone-related out- optimal dosing of ZA was studied in a random-
comes of patients with other tumor types. The ized, open-label clinical trial comparing adminis-
monoclonal anti-VEGF antibody bevacizumab tration every 4 weeks with every 12 weeks in
(Bev) enhances the activity of first-line platinum- patients with bone metastases due to breast can-
based chemotherapy against bone metastases cer, prostate cancer, or multiple myeloma [44].
This study suggested that a longer interval may resulted not only in delay of time to first SRE
be an acceptable treatment option since the risk but also in OS benefit compared with placebo
of skeletal events was not increased over 2 years (14 vs. 11.2 months) [50].
compared to the conventional dosing [44].
Despite ZA treatment, more than one third of
patients with bone metastases will still develop Vertebral Metastases
SREs [42]. The discovery of the osteoprotegerin and Immunotherapy
(OPG)–RANKL–RANK pathway led to the
development of denosumab, which is a recom- While the use of immune checkpoint inhibitors,
binant RANKL antagonist [45]. Several clinical particularly PD-1 and PD-L1 antagonists, has
studies examined the effects of this drug in been established in different cancer types, little is
patients with bone metastases from different known about these new agents and their efficacy
primaries and compared denosumab with in treating metastatic spine lesions. Recent stud-
ZA. Across numerous randomized comparisons ies in tumor/bone mouse models suggest an
and post hoc analyses, denosumab appears to be osteoclast-independent role of CD8+ T cells in
more effective in the prevention or delay of the negative regulation of bone metastases [51].
SREs and in pain control compared to ZA. In Thus, the immune microenvironment of tumors
metastatic breast cancer, denosumab delayed has emerged as a putative regulator of bone
the time to first on-study SRE by 18% when metastasis. These findings coupled with clinical
compared with ZA and further reduced the risk observations on favorable outcomes of patients
of subsequent SREs by 23% [46]. Denosumab with bone and/or bone marrow involvement (e.g.,
was also superior to ZA in delaying skeletal in melanoma) treated with PD-1/PD-L1 inhibi-
events in metastatic CRPC patients (20.7 vs. tors support a potential synergistic activity of
17.1 months) [47]. In another phase III trial PD-1/PD-L1 inhibitors and bisphosphonates or
comparing these two bone-targeted therapies in denosumab [52]. Another strategy that supports
patients with advanced cancer (excluding breast the activity of immunotherapy agents in the treat-
and prostate cancer) or multiple myeloma, ment of osseous metastatic disease involves com-
denosumab outperformed ZA in prevention of bination with concurrent RT. Not only is it safe
skeletal complications, including risk of radia- and tolerable, but it can also result in a decrease
tion to bone, worsening of pain, and frequency in the tumor growth rate of bone lesions [53].
of a shift from no/weak opioid analgesic use to
strong opioids [48]. In general, no significant
differences in PFS or OS were observed between NS Metastases and Targeted
C
ZA and denosumab across different primaries. Therapies/Immunotherapy
One exception was NSCLC, whereby an OS
benefit was demonstrated in the entire cohort The treatment of spinal metastatic disease with
(8.9 vs. 7.7 months) and in the subgroup of CNS involvement using systemic therapies is
patients with squamous cell carcinoma (8.6 vs. challenging. However, promising prospective
6.4 months) [49]. data on the safety and efficacy of targeted thera-
Another bone-targeted modality is pies and immunotherapy in these patients are
Radium-223 (Ra223), which exerts an antitu- beginning to emerge [54] (Table 37.2). One
mor effect via alpha particle emitting radiation. explanation for this is that newer targeted agents
Ra223 is administered intravenously, has a high have better CNS penetration. In the absence of
affinity for the bone matrix by virtue of its validated response criteria specific for CNS
chemical properties, and was studied in the metastases, it is sometimes difficult to fully
phase III ALSYMPCA trial in CRPC patients assess and compare the CNS-specific activity of
with bone metastases. Treatment with Ra223 these systemic therapies.
Table 37.2 Examples of targeted therapies and immuno- (96 mg/day) versus low-dose steroids (16 mg/
therapies with CNS penetration day) [53]. For instance, comparison of initial
Tumor type doses of 10 mg IV bolus versus 100 mg IV bolus
Drug Target used for showed no outcome differences regarding pain,
Gefitinib, EGFR NSCLC
ambulation, or bladder function [56].
erlotinib,
osimertinib
Ceritinib, ALK NSCLC
alectinib, Analgesia
brigatinib
Lapatinib, HER2 Breast cancer
trastuzumab,
Metastatic bone pain results in decreased quality
T-DM1 of life, function, and mood. Symptomatic relief
Vemurafenib, BRAF Melanoma with use of pain medications until other antican-
dabrafenib cer local or systemic interventions are introduced
Sunitinib, VEGF pathway RCC or take effect is essential. Analgesics are usually
pazopanib
Dasatinib BCR-ABL Chronic
administered in a ladder approach, starting with
myeloid non-opioid agents (e.g., nonsteroidal anti-
leukemia, ALL inflammatory drugs and paracetamol) [57]. For
Rituximab CD20 Non- mild-to-moderate breakthrough pain, opioids
Hodgkin’s
such as codeine and tramadol are recommended.
lymphoma
Ipilimumab CTLA4 Melanoma For severe breakthrough pain, opioids such as
(immunotherapy) morphine, oxycodone, hydromorphone, and
Nivolumab, PD-1 Melanoma, transdermal fentanyl should be started, slowly
pembrolizumab (immunotherapy) NSCLC titrated, and rotated to ensure adequate analgesia
while minimizing the risk for overdose [57].
Adjuvant analgesics can be added depending on
Supportive Care the type of pain, including gabapentin or pregab-
alin for neuropathic pain. Corticosteroids are also
Corticosteroids active in inflammatory pain, and bisphosphonates
can reduce bone pain [57].
Corticosteroids are the mainstay of pharmaco-
logical therapy for pain associated with vertebral
metastases and for the acute neurological dete- Conclusions
rioration that often accompanies metastatic epi-
dural spinal cord compression (MESCC). Spinal metastases are an important source of
Corticosteroids decrease tumor-associated morbidity and mortality in advanced cancer
inflammation (analgesia effect), decrease spinal patients. Improvement of existing multidisci-
cord edema (thereby improving short-term neu- plinary assessment models and algorithms is
rological function), and may be directly onco- essential to improve spine-specific and general
lytic in certain malignancies, including outcomes (PFS, OS). Systemic antitumor thera-
lymphoma, multiple myeloma, and breast cancer pies alone are generally reserved for asymptom-
[55]. Experimental animal models have con- atic or minimally symptomatic spinal metastases
firmed the clinical observations that those treated and in situations where considered therapies have
with dexamethasone improved motor function good CNS and bone penetration. Combination
faster than in untreated controls. Currently, there approaches, such as the paradigm of TKIs or
is no optimal dosing regimen for corticosteroids checkpoint inhibitors with radiation, hold prom-
used with MESCC, and no consensus data is ise for addressing patients presenting with sub-
available to recommend high-dose steroids acute symptoms and signs. Clinical and molecular
biomarkers of bone and spinal cord response to 12. de Bono JS, Oudard S, Ozguroglu M, Hansen S,

systemic therapies are urgently needed. As the Machiels JP, Kocak I, et al. Prednisone plus caba-
zitaxel or mitoxantrone for metastatic castration-
number of new systemic therapies increases with resistant prostate cancer progressing after docetaxel
better molecular characterization of tumors, treatment: a randomised open-label trial. Lancet.
assessing their specific impact on patients with 2010;376(9747):1147–54.
spinal metastases should be further elucidated 13. Ignatoski KM, Friedman J, Escara-Wilke J, Zhang X,
Daignault S, Dunn RL, et al. Change in markers of
and integrated into revised comprehensive treat- bone metabolism with chemotherapy for advanced
ment models. prostate cancer: interleukin-6 response is a poten-
tial early indicator of response to therapy. J Interf
Cytokine Res. 2009;29(2):105–12.
14. Santini D, Morelli F, Bertoldo F, Facchini G,

References Rizzi D, Gatti D, et al. Impact of cabazitaxel on
metastatic bone health in patients with castra-
1. Barzilai O, Laufer I, Yamada Y, Higginson DS, tion resistant prostate cancer previously treated
Schmitt AM, Lis E, et al. Integrating evidence-based with docetaxel: CaBone Study. J Clin Oncol.
medicine for treatment of spinal metastases into a 2018;36(6_suppl):TPS405.
decision framework: neurologic, oncologic, mechani- 15. Gandhi L, Rodríguez-Abreu D, Gadgeel S, et al.

cals stability, and systemic disease. J Clin Oncol. Pembrolizumab plus chemotherapy in meta-
2017;35(21):2419–27. static non-small-cell lung cancer. N Engl J Med.
2. Sciubba DM, Petteys RJ, Dekutoski MB, Fisher CG, 2018;378(22):2078–92.
Fehlings MG, Ondra SL, et al. Diagnosis and man- 16. Paz-Ares L, Luft A, Vicente D, Tafreshi A, Gümüş M,
agement of metastatic spine disease. A review. J Mazières J, et al. Pembrolizumab plus chemotherapy
Neurosurg Spine. 2010;13(1):94–108. for squamous non-small-cell lung cancer. N Engl J
3. Sciubba DM, Gokaslan ZL. Diagnosis and man- Med. 2018;379(21):2040–51.
agement of metastatic spine disease. Surg Oncol. 17. Kuchuk M, Addison CL, Clemons M, Kuchuk I,

2006;15(3):141–51. Wheatley-Price P. Incidence and consequences of
4. Laufer I, Rubin DG, Lis E, et al. The NOMS frame- bone metastases in lung cancer patients. J Bone
work: approach to the treatment of spinal metastatic Oncol. 2013;2(1):22–9.
tumors. Oncologist. 2013;18:744–51. 18. Jacus MO, Daryani VM, Harstead KE, Patel YT,

5. Kommalapati A, Tella SH, Esquivel MA, Correa Throm SL, Stewart CF. Pharmacokinetic proper-
R. Evaluation and management of skeletal dis- ties of anticancer agents for the treatment of central
ease in cancer care. Crit Rev Oncol Hematol. nervous system tumors: update of the literature. Clin
2017;120:217–26. Pharmacokinet. 2016;55(3):297–311.
6. Budczies J, von Winterfeld M, Klauschen F, 19. Woolf DK, Padhani AR, Makris A. Assessing

Bockmayr M, Lennerz JK, Denkert C, et al. The land- response to treatment of bone metastases from breast
scape of metastatic progression patterns across major cancer: what should be the standard of care? Ann
human cancers. Oncotarget. 2015;6(1):570–83. Oncol. 2015;26(6):1048–57.
7. Bines J, Earl H, Buzaid AC, Saad ED. Anthracyclines 20. Azad GK, Taylor BP, Green A, Sandri I, Swampillai
and taxanes in the neo/adjuvant treatment of breast A, Harries M, et al. Eur J Nucl Med Mol
cancer: does the sequence matter? Ann Oncol. Imaging. 2018;46:821. https://doi.org/10.1007/
2014;25(6):1079–85. s00259-018-4223-9.
8. Zheng R, Han S, Duan C, Chen K, You Z, Jia J, 21. Iagaru A, Mittra E, Mosci C, Dick DW, Sathekge M,
et al. Role of taxane and anthracycline combination Prakash V, et al. Combined 18F-fluoride and 18F-
regimens in the management of advanced breast can- FDG PET/CT scanning for evaluation of malignancy:
cer: a meta-analysis of randomized trials. Medicine results of an international multicenter trial. J Nucl
(Baltimore). 2015;94(17):e803. Med. 2013;54(2):176–83.
9. Kamby C, Vestlev PM, Mouridsen HT. Site-specific 22. Michaels AY, Keraliya AR, Tirumani SH, Shinagare
effect of chemotherapy in patients with breast cancer. AB, Ramaiya NH. Systemic treatment in breast
Acta Oncol. 1992;31(2):225–9. cancer: a primer for radiologists. Insights Imaging.
10. Tannock IF, de Wit R, Berry WR, Horti J, Pluzanska 2016;7(1):131–44.
A, Chi KN, et al. Docetaxel plus prednisone or mito- 23. Varenhorst E, Klaff R, Berglund A, Hedlund PO,
xantrone plus prednisone for advanced prostate can- Sandblom G. Scandinavian Prostate Cancer Group
cer. N Engl J Med. 2004;351(15):1502–12. (SPCG) Trial No. 5. Predictors of early androgen
11. Kyriakopoulos CE, Chen YH, Carducci MA, Liu
deprivation treatment failure in prostate cancer with
G, Jarrard DF, Hahn NM, et al. Chemohormonal bone metastases. Cancer Med. 2016;5(3):407–14.
therapy in metastatic hormone-sensitive prostate can- 24. Lassemillante AC, Doi SA, Hooper JD, Prins JB,

cer: long-term survival analysis of the randomized Wright OR. Prevalence of osteoporosis in pros-
phase III E3805 CHAARTED trial. J Clin Oncol. tate cancer survivors: a meta-analysis. Endocrine.
2018;36(11):1080–7. 2014;45(3):370–81.
25. Logothetis CJ, Basch E, Molina A, Fizazi K, North administration in EGFR-mutated non-small cell lung
SA, Chi KN, et al. Effect of abiraterone acetate cancer with skeletal metastasis. Cancer Res Treat.
and prednisone compared with placebo and predni- 2016;48(3):1110–9.
sone on pain control and skeletal-related events in 36. Kalra S, Verma J, Atkinson BJ, Matin SF, Wood CG,
patients with metastatic castration-resistant pros- Karam JA, et al. Outcomes of patients with meta-
tate cancer: exploratory analysis of data from the static renal cell carcinoma and bone metastases in
COU-AA-301 randomised trial. Lancet Oncol. the targeted therapy era. Clin Genitourin Cancer.
2012;13(12):1210–7. 2017;15(3):363–70.
26. Loriot Y, Miller K, Sternberg CN, Fizazi K, De Bono 37. Beuselinck B, Wolter P, Karadimou A, Elaidi R,

JS, Chowdhury S, et al. Effect of enzalutamide on Dumez H, Rogiers A, et al. Concomitant oral tyrosine
health-related quality of life, pain, and skeletal- kinase inhibitors and bisphosphonates in advanced
related events in asymptomatic and minimally symp- renal cell carcinoma with bone metastases. Br J
tomatic, chemotherapy-naive patients with metastatic Cancer. 2012;107(10):1665–71.
castration-resistant prostate cancer (PREVAIL): 38. Tokito T, Shukuya T, Akamatsu H, Taira T, Ono

results from a randomised, phase 3 trial. Lancet A, Kenmotsu H, et al. Efficacy of bevacizumab-
Oncol. 2015;16(5):509–21. containing chemotherapy for non-squamous non-
27. Rizzo S, Galvano A, Pantano F, Iuliani M, Vincenzi small cell lung cancer with bone metastases. Cancer
B, Passiglia F, et al. The effects of enzalutamide Chemother Pharmacol. 2013;71(6):1493–8.
and abiraterone on skeletal related events and bone 39. Addison CL, Pond GR, Cochrane B, Zhao H, Chia
radiological progression free survival in castration SK, Levine MN, et al. Correlation of baseline bio-
resistant prostate cancer patients: an indirect compari- markers with clinical outcomes and response to
son of randomized controlled trials. Crit Rev Oncol fulvestrant with vandetanib or placebo in patients
Hematol. 2017;120:227–33. with bone predominant metastatic breast cancer:
28. Efstathiou E, Titus M, Wen S, Hoang A, Karlou
an OCOG ZAMBONEY sub-study. J Bone Oncol.
M, Ashe R, et al. Molecular characterization of 2015;4(2):47–53.
enzalutamide- treated bone metastatic castration- 40. Mbagui R, Langrand-Escure J, Annede P, Mery B,
resistant prostate cancer. Eur Urol. 2015;67(1):53–60. Ceccaldi B, Guy JB, et al. Safety of spinal radio-
29. Wu Q, Li J, Zhu S, Wu J, Chen C, Liu Q, et al. Breast therapy in metastatic cancer patients receiving
cancer subtypes predict the preferential site of dis- bevacizumab therapy: a bi-institutional case series.
tant metastases: a SEER based study. Oncotarget. Anti-Cancer Drugs. 2015;26(4):443–7.
2017;8(17):27990–6. 41. Coleman RE. Bone cancer in 2011: prevention and
30. Harano K, Lei X, Gonzalez-Angulo AM, Murthy RK, treatment of bone metastases. Nat Rev Clin Oncol.
Valero V, Mittendorf EA, et al. Clinicopathological 2011;9(2):76–8.
and surgical factors associated with long-term 42. Gralow JR, Biermann JS, Farooki A, Fornier MN,
survival in patients with HER2-positive meta- Gagel RF, Kumar R, et al. NCCN task force report:
static breast cancer. Breast Cancer Res Treat. bone health in cancer care. J Natl Compr Canc Netw.
2016;159(2):367–74. 2013;11(Suppl 3):S1–50; quiz S51.
31. Kuijpers CCHJ, Hendriks LEL, Derks JL, Dingemans 43. Rosen LS, Gordon D, Kaminski M, Howell A, Belch
AC, van Lindert ASR, van den Heuvel MM, et al. A, Mackey J, et al. Long-term efficacy and safety of
Association of molecular status and metastatic zoledronic acid compared with pamidronate disodium
organs at diagnosis in patients with stage IV nonin the treatment of skeletal complications in patients
squamous non-small cell lung cancer. Lung Cancer. with advanced multiple myeloma or breast carcinoma:
2018;121:76–81. a randomized, double-blind, multicenter, comparative
32. Choi BD, Shankar GM, Sivaganesan A, Van Beaver trial. Cancer. 2003;98(8):1735–44.
LA, Oh K, Shin JH. Implication of biomarker muta- 44. Himelstein AL, Foster JC, Khatcheressian JL,

tions for predicting survival in patients with meta- Roberts JD, Seisler DK, Novotny PJ, et al. Effect
static lung cancer to the spine. Spine (Phila Pa 1976). of longer- interval vs standard dosing of zole-
2018;43(21):E1274–80. dronic acid on skeletal events in patients with bone
33. Bittner N, Balikó Z, Sárosi V, László T, Tóth E, Kásler metastases: a randomized clinical trial. JAMA.
M, et al. Bone metastases and the EGFR and KRAS 2017;317(1):48–58.
mutation status in lung adenocarcinoma--the results 45. Lacey DL, Boyle WJ, Simonet WS, Kostenuik PJ,
of three year retrospective analysis. Pathol Oncol Res. Dougall WC, Sullivan JK, et al. Bench to bedside:
2015;21(4):1217–21. elucidation of the OPG-RANK-RANKL pathway
34. Zhang G, Cheng R, Zhang Z, Jiang T, Ren S, Ma and the development of denosumab. Nat Rev Drug
Z, et al. Bisphosphonates enhance antitumor effect Discov. 2012;11(5):401–19.
of EGFR-TKIs in patients with advanced EGFR 46. Stopeck AT, Lipton A, Body JJ, Steger GG, Tonkin
mutant NSCLC and bone metastases. Sci Rep. K, de Boer RH, et al. Denosumab compared with
2017;7:42979. zoledronic acid for the treatment of bone metas-
35. Hong SH, Kim YS, Lee JE, Kim IH, Kim SJ, Han D, tases in patients with advanced breast cancer: a
et al. Clinical characteristics and continued epider- randomized, double-blind study. J Clin Oncol.
mal growth factor receptor tyrosine kinase inhibitor 2010;28(35):5132–9.
47. Fizazi K, Carducci M, Smith M, Damião R, Brown J, 52. Rosner S, Sen F, Postow M. Response after treatment
Karsh L, et al. Denosumab versus zoledronic acid for with pembrolizumab in a patient with myelophthisis
treatment of bone metastases in men with castration- due to melanoma: the role of checkpoint inhibition in
resistant prostate cancer: a randomised, double-blind the bone. J Immunother Cancer. 2017;5:34.
study. Lancet. 2011;377(9768):813–22. 53. Levy A, Massard C, Soria JC, Deutsch E. Concurrent
48. Vadhan-Raj S, von Moos R, Fallowfield LJ, Patrick irradiation with the anti-programmed cell death
DL, Goldwasser F, Cleeland CS, et al. Clinical benefit ligand-1 immune checkpoint blocker durvalumab:
in patients with metastatic bone disease: results of a single centre subset analysis from a phase 1/2 trial.
phase 3 study of denosumab versus zoledronic acid. Eur J Cancer. 2016;68:156–62.
Ann Oncol. 2012;23(12):3045–51. 54. Di Lorenzo R, Ahluwalia MS. Targeted therapy of
49. Scagliotti GV, Hirsh V, Siena S, Henry DH, Woll PJ, brain metastases: latest evidence and clinical implica-
Manegold C, et al. Overall survival improvement in tions. Ther Adv Med Oncol. 2017;9(12):781–96.
patients with lung cancer and bone metastases treated 55. Skeoch GD, Tobin MK, Khan S, Linninger AA,

with denosumab versus zoledronic acid: subgroup Mehta AI. Corticosteroid treatment for metastatic
analysis from a randomized phase 3 study. J Thorac spinal cord compression: a review. Global Spine J.
Oncol. 2012;7(12):1823–9. 2017;7(3):272–9.
50. Sartor O, Coleman R, Nilsson S, Heinrich D, Helle 56. Vecht CJ, Haaxma-Reiche H, van Putten WL, de Visser
SI, O’Sullivan JM, et al. Effect of radium-223 dichlo- M, Vries EP, Twijnstra A. Initial bolus of conventional
ride on symptomatic skeletal events in patients with versus high-dose dexamethasone in metastatic spinal
castration-resistant prostate cancer and bone metasta- cord compression. Neurology. 1989;39(9):1255–7.
ses: results from a phase 3, double-blind, randomised 57. Spratt DE, Beeler WH, de Moraes FY, Rhines

trial. Lancet Oncol. 2014;15(7):738–46. LD, Gemmete JJ, Chaudhary N, et al. An inte-
51. Zhang K, Kim S, Cremasco V, Hirbe AC, Collins L, grated multidisciplinary algorithm for the man-
Piwnica-Worms D, et al. CD8+ T cells regulate bone agement of spinal metastases: an International
tumor burden independent of osteoclast resorption. Spine Oncology Consortium report. Lancet Oncol.
Cancer Res. 2011;71(14):4799–808. 2017;18(12):e720–30.
The Role of Advanced Imaging
in Spinal Metastases
38
Sasan Karimi, Nicholas S. Cho, Kyung K. Peck,
and Andrei I. Holodny
eneral Overview of Conventional

G onventional Magnetic Resonance
C
Methods Imaging
The spine is the most common site for skeletal MRI has been commonly utilized for spinal can-
metastases. With spinal metastases being present cer because it is the only modality that can
in up to 70% of cancer patients at autopsy [1], the directly image the bone marrow with high spatial
need for detecting spinal metastases and assess- resolution. In particular, T1-weighted, T2-
ing their treatment response remains a priority weighted, and short tau inversion recovery
when treating patients with cancer. The evalua- (STIR) MRI sequences have been commonly
tion for spinal metastases is largely performed employed for detecting spinal metastases [2].
using conventional imaging methods including T1-weighted spin-echo sequences are successful
bone scan, computed tomography (CT), and in detecting spinal metastases as hypointense
magnetic resonance imaging (MRI). lesions compared to healthy bone marrow and
Positron emission tomography (PET) has vertebral discs [3]. T1-weighted post-contrast
been combined with CT technology and recently sequences can further improve tumor detection
with MRI to optimize localization, diagnostic because spinal metastases often show increased
accuracy, treatment planning, and follow-up by enhancement after contrast injection [4]. T2-
combining the imaging modalities. weighted spin-echo sequences can also detect
spinal metastases because of their high water
content, and they often present with a ring of
S. Karimi (*) · A. I. Holodny bright T2 enhancement, often referred to as a
Department of Radiology, Memorial Sloan Kettering
halo sign [5]. Finally, STIR MRI is able to sup-
e-mail: karimis@mskcc.org press the signal from fat through a 180° inversion
pulse, and it sums the contrast effects of T1 and
N. S. Cho
Department of Radiology, Memorial Sloan Kettering T2 to improve tumor detection [6].
Cancer Center, New York, NY, USA However, these traditional MRI methods also
Medical Scientist Training Program, David Geffen have their limitations in the management of spi-
School of Medicine at UCLA, nal metastases. For example, there is a risk of
Los Angeles, CA, USA false-negative results if a lesion is too small or
K. K. Peck early in its progression to cause a significant
Department of Medical Physics and Radiology, alteration in local cell composition, which may
Memorial Sloan Kettering Cancer Center,
yield no observable difference in MR signal
New York, NY, USA

https://doi.org/10.1007/978-3-030-42958-4_38
524 S. Karimi et al.
intensities on the resulting image [2, 7]. The sur- permeability constant (Ktrans), which is a measure
rounding location of the lesion itself may also of vasculature leakiness [12] in addition to semi-
pose challenges in its diagnosis. Some metastatic quantitative parameters including area under the
tumors exhibit similar T1 and T2 signal intensi- curve of contrast uptake. DCE-MRI is already
ties as healthy hematopoietic bone marrow, established as an imaging method for brain
which is found abundantly in the axial skeleton tumors [14, 15] and has been recently used for
among young patients. These methods also have diagnostic imaging and treatment monitoring in
a risk of false-positive results in diagnostically patients with spinal metastasis.
challenging spinal lesions. Lesions involving
infarction, edema, fibrosis, infection, or compres-
sion fractures as well as vertebral hemangiomas, Imaging Protocol
the most common type of benign spinal tumors,
have been known to resemble malignancies on d-DTPA Contrast Agent
G
conventional imaging [7, 8]. Another drawback The DCE-MRI perfusion measurements are
of these methods is their difficulty in assessing obtained using the injected contrast agent
the treatment response of spinal metastases. Gd-DTPA (gadolinium diethylenetriaminepenta-
Tumor progression is defined as an increase in acetic acid), which provides higher physiological
the size of the lesion, but with traditional MRI, and tissue contrast compared to endogenous con-
evaluation of positive treatment response is lim- trast techniques such as arterial spin labeling
ited to simply observing stability of the lesion (ASL). Gd- DTPA has already been utilized in
size after treatment [9–11]. other imaging modalities for assessing tumors,
for increased blood vessel permeability in tumors
can be detected by 2D static imaging of contrast
Dynamic Contrast-Enhanced agent accumulation in a time window following
Magnetic Resonance Imaging administration. For example, T1-weighted post-
contrast sequences have shown increased
Background enhancement of lesions after the injection of
Gd-DTPA compared to images taken prior to
Dynamic contrast-enhanced magnetic resonance injection [4]. Another technique is dynamic sus-
imaging (DCE-MRI) is an emerging imaging ceptibility contrast perfusion MRI (DSC MRI),
method for spinal tumors [12]. DCE-MRI can which has been used in the brain to assess tumor
provide a direct, quantitative measurement of the diagnosis and progression by measuring the rela-
tumor microvasculature, making this technique tive cerebral blood volume (rCBV) between
very valuable for assessing spinal metastases. tumors and healthy tissue [16, 17]. However,
Despite their varying histology, spinal metas- limitations of this technique include that the per-
tases are well-known to secrete proangiogenic fusion measurements are context-variant, the
growth factors upon their localization in the spine images have poor spatial resolution, and the anal-
[13]. The resulting aberrant neoangiogenesis ysis is relative and user-dependent.
leads to the development of highly fragile and
permeable blood vessels in the tumor microenvi- Pharmacokinetic Two-Compartment
ronment, which can be characterized by DCE- Model
MRI. This T1-weighted perfusion MRI technique During DCE-MRI, the patient is injected with
noninvasively assesses the vascular microenvi- Gd-DTPA and scanned periodically for several
ronment and the hemodynamic information of minutes before, during, and after the contrast
the tumor through quantitative parameters such agent accumulates in the microenvironment of
as plasma volume (Vp), which is related to the the region of interest (ROI). Analysis of
number of blood vessels in the tumor, and the dynamic data can be used to study tissue
38 The Role of Advanced Imaging in Spinal Metastases 525
perfusion and vascular permeability. Voxel- (3) peak enhancement signal percentage change
wise tracer kinetic analysis is accomplished by ([signalmax – signalbase]/signalbase × 100%) [19,
applying a pharmacokinetic two-compartment 20]. Chen et al. have also established a qualitative
model by Tofts et al. where the two compart- classification of five general types of TIC mor-
ments are (1) the intravascular space (blood phologies: (1) Type A (nearly flat), (2) Type B
plasma) and (2) the extracellular extravascular (slowly rising contrast enhancement), (3) Type C
space (EES) [18]. (rapid wash-in followed by a plateau), (4) Type D
Pharmacokinetic modeling of contrast agent (rapid wash-in followed by a wash-out), and (5)
uptake is applied to the measured signal inten- Type E (rapid wash-in followed by a second
sity changes (ΔSI) over time, allowing for the slowly rising contrast enhancement) [20].
quantitative estimation of vascular characteris-
tics (Fig. 38.1): Vp estimates tumor vascularity MR Acquisition
through the blood vessel compartment, and Ktrans The 3D T1-weighted spoiled gradient recalled
estimates vessel permeability through the vol- echo (SPGR) sequence is the most widely used
ume transfer constant per minute from the blood method for DCE-MRI data acquisition. 3D acqui-
vessel to the EES. sition can be used to improve the image resolu-
Quantitative analysis includes detection of the tion and coverage. Post-contrast images are
arterial input function (AIF) from the aorta within acquired every 4–5 s to provide sufficient data to
the imaging volume. Appropriate shape of the model the contrast concentration-time-intensity
AIF curve is usually confirmed based on pixels curve. Low flip angles between 15 and 25° can be
with a large change in signal intensity, with a used to improve the measurement of signal
rapid change immediately after bolus injection, change due to contrast injection. Short repetition
and with an early peak in intensity. Further semi- time and short echo time should also be used to
quantitative analysis can be conducted by placing improve the scan time and to remove the T2∗
ROIs and analyzing the averaged time-intensity effect of contrast, respectively. Recently, a new
curves (TIC) of the changing MR signal intensity 3D volume acquisition, called Differential Sub-
during contrast accumulation. sampling with Cartesian Ordering (DISCO), has
Three measurements from these TICs include been demonstrated with an effective temporal
(1) area under curve (AUC), (2) wash-in enhance- resolution of 3–4 s while preserving spatial reso-
ment slope ([signalmax – signalbase]/timerise), and lution [21].
Fig. 38.1 A schematic Blood Plasma

illustration showing the Volume Vp
pharmacokinetic
two-compartment
model. The tissue is
presented as two Ktrans
compartments: the
vascular plasma space
Ve, Ce
and the extracellular Intra-cellular
Extra-
extravascular space. Cp Cp Space
cellular
concentration of contrast Kep Space
agent in plasma space, Ve
extracellular volume, Vp
plasma volume, Ktrans
and Kep volume transfer
constants between Vp
and EES
Diagnostic Imaging Using DCE-MRI with the expected higher degree of neoangiogen-
esis in hypervascular metastases. Vp was also
etermining Healthy Bone Marrow
D considered the best discriminator between the
and Tumor Vascularity two groups, with hypervascular lesions having
Several DCE-MRI studies have shown highly values 1.8× higher than hypovascular lesions. Vp
promising results for the diagnosis of spinal was followed by peak enhancement signal per-
metastases with this method. Khadem et al. retro- centage change as the second-best discriminator,
spectively analyzed 26 patients with spinal metas- which was 1.64× higher in hypervascular lesions
tases using DCE-MRI and conventional MRI, and than in hypovascular lesions [22]. Finally, DCE-
DCE-MRI was able to differentiate spinal metas- MRI has been deemed as an effective, noninva-
tases from normal bone marrow through general sive surrogate to catheter spinal angiography, the
TIC morphologies alone [19]. While healthy concurrent “gold standard” for assessing tumor vas-
trols exhibited little to no contrast enhancement cularity despite its invasiveness and high cost
(Type A TIC [20]), spinal metastases exhibited [23]. This has implications for surgery because
contrast enhancement above the baseline [19]. DCE-MRI can noninvasively determine hyper-
DCE-MRI can also differentiate new spinal vascularity so that catheter spinal angiography
metastases from previously treated metastases need only be employed for preoperative tumor
despite their similar appearance on conventional embolization to reduce intraoperative blood loss
MRI (Fig. 38.2a). New metastases are easily [22]. DCE-MRI can assess biomarkers for anti-
visualized on DCE-MRI through enhancement angiogenetic treatment.
on phase-derived and Vp heat maps (Fig. 38.2b,
c), and previously treated metastases exhibit flat Malignant and Benign Vertebral
TIC morphologies like normal bone marrow Compression Fractures
(Type A TIC [20], Fig. 38.2d). Vertebral compression fractures are a common and
Moreover, DCE-MRI can differentiate growing concern in our aging patient population. In
between hypervascular and hypovascular metas- elderly patients, compression fractures are usually
tases, while conventional MRI cannot. Khadem benign as a result of osteoporosis [24]. However,
et al. found that conventional MR signal inten- elderly cancer patients are prone to developing
sity percentage changes between pre- and post- malignant compression fractures as a result of
Gd- DTPA injection for T1-weighted images osteolytic spinal metastases that can decrease bone
were not significantly different between hyper- density and structural integrity. Common treatment
vascular and hypovascular spinal metastases. regimens such as chemotherapy, radiation therapy,
Nevertheless, DCE-MRI was able to semiquanti- hormone therapy, and steroids can also affect bone
tatively distinguish these two groups, for hyper- density and lead to compression fractures [25].
vascular metastases were found to have higher Diagnosis is further complicated by malignant and
average wash-in enhancement slope (p < 0.01) benign fractures having similar appearances in con-
and higher average peak enhancement signal ventional MRI [26].
percentage change (p < 0.01) compared to hypo- DCE-MRI can differentiate between these two
vascular metastases. types of fractures through multiple perfusion
A follow-up study by Saha et al. analyzed 20 metrics. Arevalo-Perez et al. found that malig-
patients with hypervascular renal spinal metasta- nant fractures had higher Vp, Ktrans, wash-in slope,
ses and 20 patients with hypovascular prostate peak enhancement, and AUC compared to benign
spinal metastases, and DCE-MRI was able to fractures (p < 0.01) [27]. DCE-MRI also had sen-
also quantitatively distinguish these two groups. sitivity within benign fractures to distinguish
Hypervascular metastases had higher Vp between acute (edema present) and chronic frac-
(p < 0.001) and Ktrans values (p < 0.01) compared tures (no edema present), with acute fractures
to hypovascular metastases, which was in line having higher values of the aforementioned.
a b c
Time-Intensity Curves for Treated (Tx) and Untreated

d Metastases
10
8
MR Signal (A.U.)
0
0 10 20 30 40 50
-2 Phase
Lesion 1 Lesion 2 Lesion 3 Lesion 4 (Tx)
Fig. 38.2 Evaluating untreated (lesions 1–3) and treated on the phase-derived (b) and Vp (c) heat maps and exhibit
(lesion 4) metastases using DCE-MRI. Conventional MRI a rapid wash-in followed by a wash-out Type D TIC (d).
shows little difference between the two lesion types (a). (Tx = treated; A.U. = arbitrary unit)
On DCE-MRI, untreated metastases appear hyperintense
Atypical Hemangiomas [8]. DCE-MRI can quantitatively differentiate

Atypical hemangiomas are common benign spinal metastases from atypical vertebral heman-
tumors that often resemble spinal metastases and giomas. Morales et al. found that spinal
other malignant lesions in conventional MRI due metastases had higher Vp values (p < 0.01) and
to their high vascularity and low-fat composition higher Ktrans values (p < 0.01) than atypical
vertebral hemangiomas. From a qualitative anal- Application in Chordomas

ysis of the TICs, spinal metastases also had Chordoma is a rare type of spinal tumor, account-
higher signal intensities and a curve morphology ing for only 1–4% of bone cancers. However,
of rapid wash-in followed by a wash-out (Type D they are known for being very aggressive tumors
TIC [20]), while atypical vertebral hemangiomas that are also prone to recurrence. Moreover, they
generally had a curve morphology of slow wash- are diagnostically challenging using conventional
in followed by a plateau (closest to Type C TIC imaging and predominately clinically silent
[20]). Interestingly, four cases of atypical heman- before exhibiting rapid progression and becom-
giomas presented Type D TIC morphologies sim- ing highly resistant to treatment [33]. Assessing
ilar to spinal metastases, underscoring their treatment efficacy remains difficult as well
challenging diagnosis [28]. because chordomas generally do not change in
appearance in conventional imaging during posi-
tive response to therapy.
Monitoring Radiation Therapy DCE-MRI has demonstrated potential clinical
Treatment Response Using DCE-MRI utility in the treatment of chordomas. This
method has been successful in differentiating
Radiation Therapy for Spinal chordoma from giant cell tumors [34], and their
Metastases TICs exhibit a distinct rapid wash-in followed by
DCE-MRI also has the valuable capability to a second slowly rising contrast enhancement
quantitatively monitor radiation therapy treat- morphology [35] (Type E TIC [20]). Also,
ment response in spinal metastases (Fig. 38.3). decreased Vp, Ktrans, and MR signal intensities
As seen earlier in this chapter, spinal metastases have been found to be more sensitive than con-
are known for their high perfusion metrics, par- ventional imaging in determining positive treat-
ticularly in Vp. However, radiation therapy ment response after radiation therapy [35].
induces changes in the tumor microvasculature
that reduce blood flow through fibrosis, thrombo-
sis, and apoptosis [29]. DCE-MRI can detect Limitations of DCE-MRI
these changes through decreased Vp values in less
than 1 hour after high-dose image-guided radia- DCE-MRI also has several limitations.
tion therapy, and Lis et al. reported an average Application of the pharmacokinetic two-
drop of 65.2% (p < 0.05) compared to pretreat- compartment model requires accurate measure-
ment with no significant change in this value to ment of the AIF. However, the low temporal
the first clinical follow-up [30]. Decreased Vp resolution of DCE- MRI because of its high
after radiation therapy has also been found to be spatial resolution, field-of-view, and signal-to-
the best predictor of positive treatment response noise ratio may be insufficient in reliably cap-
within only 6 months of treatment, which is about turing the AIF. As a result, insufficient sampling
half the time needed to establish stable tumor size of the AIF can affect its time course and cause
in conventional MRI [31]. Moreover, qualitative saturation effects during initial wash-in of the
TIC analysis of successfully treated tumors has contrast agent [30].
shown a rapid wash-in followed by a second Quantitative analysis using the pharmacoki-
slowly rising contrast enhancement (Type E TIC netic model also relies on a set of model assump-
[20]), while unsuccessfully treated tumors tions, which may not uphold for all tumor or
retained their rapid wash-in followed by wash- tissue types [22]. Also, the physiological basis of
out (Type D TIC [20]) morphology [31]. DCE- semiquantitative parameters such as peak
MRI can also indicate local recurrence of spinal enhancement and AUC as well as of the
metastases through an increase in Vp and Ktrans mechanisms that lead to perfusion differences
values about 6 months earlier than conventional before and after radiation therapy remains
MRI [32]. unknown [22, 30].
T1-w MRI Ktrans Vp Peak Enhance
Fig. 38.3 Monitoring treatment response at baseline (a), each of the MRI scans. The lesion appears stable in con-
3 weeks (b), 13 weeks (c), 15 weeks (d), and 16 weeks (e) ventional MRI, but there are dramatic changes seen on the
using conventional MRI (first column to the left) and perfusion maps, particularly in the decreased plasma vol-
DCE-MRI (second–fourth columns to the left) of success- ume (Vp) indicating positive treatment response.
fully treated metastasis. The arrows indicate the lesion on (T1-w = T1-weighted; Vp = plasma volume)
Unlike T2∗-based DSC perfusion which is protocol with an additional application of

acquired within a minute range, DCE-MRI diffusion-sensitive gradients at multiple
requires a certain amount of scanning time to “b-values”. The b-value determines the strength
allow sampling the wash-out phase which is used of the diffusion-weighting, with higher b-values
to estimate the Ve parameter. Despite its limita- having greater sensitivity to diffusion properties
tions, DCE-MRI has already demonstrated great but lower b-values having higher SNR [38]. Scan
potential in the management of spinal metastases times are relatively quick in DWI, and this
given its recent advent as described above. Future method does not utilize an intravenous contrast
studies involving larger sample populations and agent as in perfusion MRI, making this technique
more different types of pathologies can provide safe for pregnant patients and those with allergies
further insights into the clinical applications of or poor renal function.
this emerging and promising method.
iagnostic Imaging Using DWI
D
Quantitative assessment of DWI using ADC is a
Diffusion-Weighted Imaging reliable method for diagnosing spinal lesions. A
meta-analysis conducted by Suh et al. has found
Background that ADC can differentiate between benign and
malignant vertebral bone marrow lesions with
Diffusion-weighted imaging (DWI) has com- 89% sensitivity and 87% specificity as well as
monly been used for imaging stroke patients, but differentiate between benign and malignant
it is now finding a new role in assessing tumors as compression fractures with 92% sensitivity and
well. DWI is an advanced method of MRI that is 91% specificity [39]. Multiple studies have
sensitive to free water diffusion, or the random shown that spinal metastases have lower ADC
Brownian motion of water molecules. Within tis- values than healthy bone marrow and that malig-
sue microstructure, diffusion can be impacted by nant neoplastic compression fractures have
hindered diffusion and restricted diffusion. In lower ADC values than benign osteoporotic
hindered diffusion, water molecules are impeded compression fractures [40–42]. Similarly, spinal
extracellularly by cells and other obstacles in the metastases were also found to have lower ADC
extracellular matrix. In restricted diffusion, water values than both typical and atypical vertebral
molecules are impeded intracellularly by cellular hemangiomas [43]. However, Pozzi et al.
compartments such as the cell membrane [36]. showed that DWI is unsuccessful in differentiat-
Tumors are known for their high cellularity, ing between spinal metastases and malignant
which results in increased restricted diffusion. primary spinal tumors [41].
This behavior can be detected in DWI from the
original DWI image, but it is more commonly onitoring Treatment Response Using
M
assessed by converting the image into its apparent DWI
diffusion coefficient (ADC) map, which is a quan- DWI has been successful in determining positive
titative measure related to the amount of diffusion treatment response for spinal metastases.
in a voxel. ADC is negatively correlated with cel- Radiation therapy causes necrosis of tumor cells,
lularity [37], so tumors exhibit decreased ADC which decreases tumor cellularity and increases
values and appear hypointense on ADC maps. extracellular volume fractions that can lead to
more free water diffusion [44]. This has been
used to explain how ADC values increased in
Imaging Protocol successfully treated spinal metastases and contin-
ued to decrease in unsuccessfully treated cases as
MR Acquisition early as 1 month after radiation therapy [44, 45],
DWI scans are typically acquired using a despite no significant changes in signal intensi-
T2-weighted single-shot echo-planar imaging ties being observed in conventional MRI [45].
DWI has also shown promise in assessing cord surgery [54]. DTI can also assess the integ-
positive treatment response for androgen with- rity of the spinal cord after radiation therapy,
drawal therapy, where Resichauer et al. found such as evaluating for radiation-induced myelop-
significantly higher ADC values in successfully athy [55]. This has led to DTI being employed to
treated metastases in the pelvis at 1, 2, and monitor the spinal cord during radiation therapy
3 months posttreatment [46]. for spinal metastases.
However, its use for spinal cord imaging is
still challenging due to its low spatial resolution
Limitations of DWI and the small size of the spinal cord, which
results in low SNR. Moreover, this technique is
Despite its clinical utility, DWI has several limi- sensitive to susceptibility and flow artifacts in
tations. For example, determining a strict cutoff the spine that can lead to distortion [54].
ADC value for different diagnoses is not practi- Recently, a new DTI method using a restricted
cal because ADC values are dependent on the small field-of-view (FOV) is recognized as a
field strength of the MRI scanner and the b-value promising way to acquire images of regions’ thin
of the diffusion-sensitive gradients [47]. structures like the spinal cord [56]. It consists of
DWI is not suitable for assessing sclerotic reducing the FOV in the phase- or frequency-
lesions because of their low water content, which encoding direction to shorten the echo-planar
can lead to false-negative results [48, 49]. It has readout train and to attenuate susceptibility- and
also been suggested that infections, blood prod- motion-related artifacts.
ucts, and abscess formations can lead to false-
positive results due to their decreased ADC
values [50]. Castillo et al. also found that DWI Dual-Energy Computed
did not offer any advantage over conventional Tomography
non-contrast T1-weighted MRI in detecting spi-
nal metastases, likely due to the T2 shine-through Continual improvement in computed tomogra-
effect in DWI [51]. Treatment can lead to heter- phy (CT) technology has allowed for high-quality
ogenous ADC increases and decreases in tumors imaging of the spine. CT offers exquisite detail of
that can significantly impact mean ADC analysis the bony cortex and can provide answers to most
for assessing treatment response as well [52]. clinical questions, especially for evaluating frac-
tures. It is also ideal for evaluating the integrity of
spinal hardware. The use of intrathecal contrast
Diffusion Tensor Imaging as in CT myelography can further increase its
utility as it improves visualization of the spinal
Diffusion tensor imaging (DTI) is another diffu- cord and the subarachnoid space. It is also rou-
sion MRI technique that can evaluate the integ- tinely used at many tertiary institutions as the
rity of white matter tracts and provide more modality of choice for radiation treatment plan-
sensitive information about spinal cord altera- ning and simulation as it provides very high spa-
tions, such as those originating from inflamma- tial resolution [57].
tion, trauma, neurodegenerative diseases, and One of the more recent innovations in CT
intramedullary tumors [53]. Diffusion tensor technology is dual-energy CT (DECT). DECT is
imaging utilizes a single-shot echo-planar imag- essentially simultaneous imaging at two different
ing sequence, and it has been widely used to energies. DECT takes advantage of that fact that
investigate the brain. substances exhibit varying imaging characteris-
There has been a recent implementation of tics at different X-ray energies. That information
DTI for the management of spinal tumors. For can then be used to extract various information
example, DTI can be used for pre-surgical plan- about tissue composition such as differentiating
ning to delineate tumor boundaries for spinal between soft tissue and vertebra, cystic lesions,
and crystals. It can also be used to provide virtual malignant bone marrow involvement with dynamic
unenhanced and perfusion images. With regard to contrast-enhanced magnetic resonance imaging. Ann
Oncol. 2003;14(1):152–8.
spinal imaging, DECT can be used to decrease 8. Gaudino S, Martucci M, Colantonio R, Lozupone
artifact in patients with metallic hardware which E, Visconti E, Leone A, et al. A systematic
in turn allows for better visualization of struc- approach to vertebral hemangioma. Skelet Radiol.
tures near the hardware like grafts and more 2015;44(1):25–36.
9. O’Sullivan GJ, Carty FL, Cronin CG. Imaging
importantly the spinal canal [58]. of bone metastasis: an update. World J Radiol.
2015;7(8):202–11.
10. Otake S, Mayr NA, Ueda T, Magnotta VA, Yuh

WTC. Radiation-induced changes in MR signal
Conclusion intensity and contrast enhancement of lumbosacral
vertebrae: do changes occur only inside the radiation
Conventional imaging techniques have been therapy field? Radiology. 2002;222(1):179–83.
11. Yankelevitz DF, Henschke CI, Knapp PH, Nisce L, Yi
valuable for the management of spinal metasta- Y, Cahill P. Effect of radiation therapy on thoracic and
ses, but their imaging limitations have posed lumbar bone marrow: evaluation with MR imaging.
difficulties in diagnosis and treatment monitor- AJR Am J Roentgenol. 1991;157(1):87–92.
ing. The development of advanced imaging 12. Montazel JL, Divine M, Lepage E, Kobeiter H, Breil
S, Rahmouni A. Normal spinal bone marrow in
techniques, particularly in dynamic perfusion adults: dynamic gadolinium-enhanced MR imaging.
MRI, diffusion MRI, and dual-energy CT, offers Radiology. 2003;229(3):703–9.
promising solutions for imaging spinal metasta- 13. Maccauro G, Spinelli MS, Mauro S, Perisano C, Graci
ses. Increased development and clinical imple- C, Rosa MA. Physiopathology of spine metastasis. Int
J Surg Oncol. 2011;2011:107969.
mentation of these techniques will be important 14. Arevalo-Perez J, Peck K, Young R, Holodny A,

to further improve clinical care for spinal metas- Karimi S, Lyo J. Dynamic contrast- enhanced perfu-
tases patients. sion MRI and diffusion-weighted imaging in grading
of gliomas. J Neuroimaging. 2015;25(5):792–8.
15. Thomas AA, Arevalo-Perez J, Kaley T, Lyo J,

Peck KK, Shi W, et al. Dynamic contrast enhanced
References T1 MRI perfusion differentiates pseudoprogres-
sion from recurrent glioblastoma. J Neuro-Oncol.
1. Perrin RG. Metastatic tumors of the axial spine. Curr 2015;125(1):183–90.
Opin Oncol. 1992;4(3):525–32. 16. Barajas RF, Chang JS, Sneed PK, Segal MR,

2. Solomou E, Kazantzi A, Romanos O, Kardamakis McDermott MW, Cha S. Distinguishing recurrent
D. Magnetic resonance imaging of metastatic bone intra-axial metastatic tumor from radiation necrosis
disease. In: Kardamakis D, Vassiliou V, Chow E, following gamma knife radiosurgery using dynamic
editors. Bone metastases: a translational and clinical susceptibility-weighted contrast-enhanced perfusion
approach. Dordrecht: Springer; 2009. p. 163–81. MR imaging. American Journal of Neuroradiology.
3. Carroll KW, Feller JF, Tirman PF. Useful internal stan- 2009;30(2):367.
dards for distinguishing infiltrative marrow pathology 17. Hatzoglou V, Ulaner GA, Zhang Z, Beal K, Holodny
from hematopoietic marrow at MRI. J Magn Reson AI, Young RJ. Comparison of the effectiveness of
Imaging: JMRI. 1997;7(2):394–8. MRI perfusion and Fluorine-18 FDG PET-CT for dif-
4. Breger RK, Williams AL, Daniels DL, Czervionke ferentiating radiation injury from viable brain tumor.
LF, Mark LP, Haughton VM, et al. Contrast enhance- Clin Imaging. 2013;37(3):451–7.
ment in spinal MR imaging. AJR Am J Roentgenol. 18. Tofts PS, Brix G, Buckley DL, Evelhoch JL,

1989;153(2):387–91. Henderson E, Knopp MV, et al. Estimating kinetic
5. Schweitzer ME, Levine C, Mitchell DG, Gannon parameters from dynamic contrast-enhanced T(1)-
FH, Gomella LG. Bull’s-eyes and halos: useful MR weighted MRI of a diffusable tracer: standardized
discriminators of osseous metastases. Radiology. quantities and symbols. J Magn Reson Imaging:
1993;188(1):249–52. JMRI. 1999;10(3):223–32.
6. Mirowitz SA, Apicella P, Reinus WR, Hammerman 19. Khadem NR, Karimi S, Peck KK, Yamada Y, Lis E,
AM. MR imaging of bone marrow lesions: relative Lyo J, et al. Characterizing hypervascular and hypo-
conspicuousness on T1-weighted, fat- suppressed vascular metastases and normal bone marrow of the
T2-weighted, and STIR images. AJR Am J spine using dynamic contrast-enhanced MR imaging.
Roentgenol. 1994;162(1):215–21. AJNR Am J Neuroradiol. 2012;33(11):2178–85.
7. Moulopoulos LA, Maris TG, Papanikolaou N, 20. Chen WT, Shih TT, Chen RC, Lo HY, Chou CT,
Panagi G, Vlahos L, Dimopoulos MA. Detection of Lee JM, et al. Blood perfusion of vertebral lesions
evaluated with gadolinium-enhanced dynamic MRI to predict local recurrence after radiosurgery
MRI: in comparison with compression fracture on spinal metastases. Technol Cancer Res Treat.
and metastasis. J Magn Reson Imaging: JMRI. 2017;1533034617705715
2002;15(3):308–14. 33. Walcott BP, Nahed BV, Mohyeldin A, Coumans JV,
21. Saranathan M, Rettmann DW, Hargreaves BA, Clarke Kahle KT, Ferreira MJ. Chordoma: current concepts,
SE, Vasanawala SS. Differential subsampling with management, and future directions. Lancet Oncol.
Cartesian ordering (DISCO): a high spatio-temporal 2012;13(2):e69–76.
resolution Dixon imaging sequence for multiphasic 34. Lang N, Su MY, Xing X, Yu HJ, Yuan H. Morphological
contrast enhanced abdominal imaging. J Magn Reson and dynamic contrast enhanced MR imaging features
Imaging. 2012;35(6):1484–92. for the differentiation of chordoma and giant cell
22. Saha A, Peck KK, Lis E, Holodny AI, Yamada Y, tumors in the axial skeleton. J Magn Reson Imaging:
Karimi S. Magnetic resonance perfusion characteris- JMRI. 2017;45(4):1068–75.
tics of hypervascular renal and hypovascular prostate 35. Santos P, Peck KK. T1-weighted dynamic contrast-
spinal metastases: clinical utilities and implications. enhanced MR perfusion imaging characterizes tumor
Spine. 2014;39(24):E1433–40. response to radiation therapy in chordoma. AJNR Am
23. Mazura JC, Karimi S, Pauliah M, Banihashemi MA, J Neuroradiol. 2017;38(11):2210–6.
Gobin YP, Bilsky MH, et al. Dynamic contrast- 36. White NS, McDonald C, Farid N, Kuperman J, Karow
enhanced magnetic resonance perfusion compared D, Schenker-Ahmed NM, et al. Diffusion-weighted
with digital subtraction angiography for the evalu- imaging in cancer: physical foundations and appli-
ation of extradural spinal metastases: a pilot study. cations of restriction spectrum imaging. Cancer Res.
Spine. 2014;39(16):E950–4. 2014;74(17):4638–52.
24. Jung HS, Jee WH, McCauley TR, Ha KY, Choi
37. Surov A, Meyer HJ, Wienke A. Correlation between
KH. Discrimination of metastatic from acute osteo- apparent diffusion coefficient (ADC) and cellular-
porotic compression spinal fractures with MR ity is different in several tumors: a meta-analysis.
imaging. Radiographics: a review publication of Oncotarget. 2017;8(35):59492–9.
the Radiological Society of North America, Inc. 38.
Khoo MM, Tyler PA, Saifuddin A, Padhani
2003;23(1):179–87. AR. Diffusion-weighted imaging (DWI) in mus-
25. Croarkin E. Osteopenia in the patient with cancer. culoskeletal MRI: a critical review. Skelet Radiol.
Phys Ther. 1999;79(2):196–201. 2011;40(6):665–81.
26. Verstraete KL, Van der Woude HJ, Hogendoorn
39. Suh CH, Yun SJ, Jin W, Lee SH, Park SY, Ryu

PC, De-Deene Y, Kunnen M, Bloem JL. Dynamic CW. ADC as a useful diagnostic tool for differ-
contrast-enhanced MR imaging of musculoskeletal entiating benign and malignant vertebral bone
tumors: basic principles and clinical applications. J marrow lesions and compression fractures: a sys-
Magn Reson Imaging: JMRI. 1996;6(2):311–21. tematic review and meta-analysis. Eur Radiol.
27. Arevalo-Perez J, Peck KK, Lyo JK, Holodny AI,
2018;28(7):2890–902.
Lis E, Karimi S. Differentiating benign from malig- 40. Herneth AM, Philipp MO, Naude J, Funovics M,

nant vertebral fractures using T1 -weighted dynamic Beichel RR, Bammer R, et al. Vertebral metasta-
contrast- enhanced MRI. J Magn Reson Imaging: ses: assessment with apparent diffusion coefficient.
JMRI. 2015;42(4):1039–47. Radiology. 2002;225(3):889–94.
28. Morales KA, Arevalo-Perez J. Differentiating atypi- 41. Pozzi G, Albano D. Solid bone tumors of the spine:
cal hemangiomas and metastatic vertebral lesions: diagnostic performance of apparent diffusion coeffi-
the role of T1-weighted dynamic contrast-enhanced cient measured using diffusion-weighted MRI using
MRI. AJNR Am J Neuroradiol. 2018;39(5):968–73. histology as a reference standard. J Magn Reson
29. Barker HE, Paget JTE, Khan AA, Harrington KJ. The Imaging: JMRI. 2018;47(4):1034–42.
tumour microenvironment after radiotherapy: mecha- 42. Pozzi G, Garcia Parra C, Stradiotti P, Tien TV,

nisms of resistance and recurrence. Nat Rev Cancer. Luzzati A, Zerbi A. Diffusion-weighted MR imaging
2015;15(7):409–25. in differentiation between osteoporotic and neoplas-
30. Lis E, Saha A, Peck KK, Zatcky J, Zelefsky MJ, tic vertebral fractures. Eur Spine J. 2012;21(Suppl
Yamada Y, et al. Dynamic contrast- enhanced mag- 1):S123–7.
netic resonance imaging of osseous spine metastasis 43. Shi YJ, Li XT, Zhang XY, Liu YL, Tang L, Sun

before and 1 hour after high- dose image-guided radi- YS. Differential diagnosis of hemangiomas from spi-
ation therapy. Neurosurg Focus. 2017;42(1):E9. nal osteolytic metastases using 3.0 T MRI: compari-
31. Chu S, Karimi S, Peck KK, Yamada Y, Lis E, Lyo son of T1-weighted imaging, chemical-shift imaging,
J, et al. Measurement of blood perfusion in spinal diffusion-weighted and contrast-enhanced imaging.
metastases with dynamic contrast-enhanced magnetic Oncotarget. 2017;8(41):71095–104.
resonance imaging: evaluation of tumor response to 44. Byun WM, Shin SO, Chang Y, Lee SJ, Finsterbusch
radiation therapy. Spine. 2013;38(22):E1418–24. J, Frahm J. Diffusion-weighted MR imaging of
32. Kumar KA, Peck KK, Karimi S, Lis E, Holodny
metastatic disease of the spine: assessment of
AI, Bilsky MH, et al. A pilot study evaluating response to therapy. AJNR Am J Neuroradiol.
the use of dynamic contrast-enhanced perfusion 2002;23(6):906–12.
45. Cappabianca S, Capasso R, Urraro F, Izzo A,

51.
Castillo M, Arbelaez A, Smith JK, Fisher
Raucci A, Di Franco R, et al. Assessing response LL. Diffusion-weighted MR imaging offers no
to radiation therapy treatment of bone metastases: advantage over routine noncontrast MR imaging in
short-term followup of radiation therapy treatment the detection of vertebral metastases. AJNR Am J
of bone metastases with diffusion-weighted mag- Neuroradiol. 2000;21(5):948–53.
netic resonance imaging. Journal of Radiotherapy. 52. Messiou C, Collins DJ, Giles S, de Bono JS, Bianchini
2014;2014:8. D, de Souza NM. Assessing response in bone metas-
46. Reischauer C, Froehlich JM, Koh DM, Graf N,
tases in prostate cancer with diffusion weighted
Padevit C, John H, et al. Bone metastases from MRI. Eur Radiol. 2011;21(10):2169–77.
prostate cancer: assessing treatment response by 53. Egger K, Hohenhaus M, Van Velthoven V, Heil S,
using diffusion-weighted imaging and functional Urbach H. Spinal diffusion tensor tractography for
diffusion maps--initial observations. Radiology. differentiation of intramedullary tumor-suspected
2010;257(2):523–31. lesions. Eur J Radiol. 2016;85(12):2275–80.
47. Dale BM, Braithwaite AC, Boll DT, Merkle EM. Field 54.
Choudhri AF, Whitehead MT, Klimo P Jr,
strength and diffusion encoding technique affect Montgomery BK, Boop FA. Diffusion tensor imag-
the apparent diffusion coefficient measurements in ing to guide surgical planning in intramedullary
diffusion-weighted imaging of the abdomen. Investig spinal cord tumors in children. Neuroradiology.
Radiol. 2010;45(2):104–8. 2014;56(2):169–74.
48. Hackländer T, Scharwächter C, Golz R, Mertens
55. Keřkovský M, Zitterbartová J, Pour L, Šprláková-
H. Value of diffusion-weighted imaging for diag- Puková A, Mechl M. Diffusion tensor imaging in
nosing vertebral metastases due to prostate can- radiation-induced myelopathy. J Neuroimaging.
cer in comparison to other primary tumors. Rofo. 2014;25(5):836–40.
2006;178(4):416–24. 56. Crombe A, Alberti N, Hiba B, Uettwiller M, Dousset
49. Oztekin O, Ozan E, Hilal Adibelli Z, Unal G,
V, Tourdias T. Cervical spinal cord DTI is improved
Abali Y. SSH-EPI diffusion-weighted MR imag- by reduced FOV with specific balance between the
ing of the spine with low b values: is it useful in number of diffusion gradient directions and averages.
differentiating malignant metastatic tumor infiltra- Am J Neuroradiol. 2016;37(11):2163.
tion from benign fracture edema? Skelet Radiol. 57. Sudha SP, Gopalakrishnan MS, Saravanan K. The role
2009;38(7):651–8. of CT myelography in sparing the spinal cord during
50. Subhawong TK, Jacobs MA, Fayad LM. Diffusion- definitive radiotherapy in vertebral hemangioma. J
weighted MR imaging for characterizing musculo- Appl Clin Med Phys. 2017;18(5):174–7.
skeletal lesions. Radiographics: a review publication 58. Wolman DN, Patel BP, Wintermark M, Heit JJ. Dual-
of the Radiological Society of North America, Inc. energy computed tomography applications in neuro-
2014;34(5):1163–77. intervention. J Comput Assist Tomogr. 2018;42:831.
Decision-Making Algorithms
for Surgical Treatment of Spine
39
Metastatic Disease
Brenton H. Pennicooke, Ibrahim Hussain,
and Ali A. Baaj
istory of Treating Spinal Epidural

H walk longer (median 122 days vs. 13 days,
Disease p = 0.003), regained the ability to walk if they
were unable to do so at the time of randomiza-
In the 1980s, epidural spinal cord compression tion (62% vs. 19%, p = 0.01), and had longer
(ESCC) treatment consisted of a combination of survival times (median 126 days vs. 100 days,
corticosteroids and radiotherapy [1] after a p = 0.033), when compared to the radiation
study showed that performing a laminectomy alone group.
alone or in combination with radiation resulted
in no benefit and possible harm when compared
to radiation alone. With the advancement of sur- Framework for Surgical
gical instrumentation techniques, a wide cir- Decision-Making
cumferential decompression and posterior
instrumentation with or without an anterior col- OMS (Neurologic, Oncologic,
N
umn reconstruction became the optimal treat- Mechanical, Systemic) Criteria
ment for patients with ESCC. This modality of
treatment was validated by Patchell et al. in Assessing the patient’s symptoms, tumor burden,
1990 who performed a randomized, multi-insti- life expectancy, the degree of neural element
tutional, non-blinded trial, in which patients compression, and comorbidities is essential for
with spinal cord compression caused by meta- offering surgery to patients with epidural spinal
static cancer were randomly assigned to either cord or cauda equina compression. The goal of
wide surgical decompression and appropriate any surgical intervention for patients with spine
reconstruction followed by radiotherapy metastases is palliative with the intent to preserve
(n = 50) or radiotherapy alone (n = 51) [2]. This or restore neurologic function, improve pain,
study showed that patients in the surgical group treat mechanical instability, improve quality of
were able to walk after treatment (84% vs. 57%, life, and provide durable tumor control. The
odds ratio 6.2, p = 0.001), retained the ability to NOMS criteria, as seen in Table 39.1, is a deci-
sion framework that selects those patients most
likely to benefit from surgical decompression and
instrumentation [3, 4].
B. H. Pennicooke · I. Hussain · A. A. Baaj (*)
Weill Cornell Medical Center, New York, NY, USA
e-mail: ALB9140@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_39
536 B. H. Pennicooke et al.
Table 39.1 NOMS decision framework

Neurologic Oncologic Mechanical Systemic Decision
Low-grade epidural Radiosensitive Stable Conventional external beam
spinal cord compression Radiosensitive Unstable radiation (cERT)
with no myelopathy Radioresistant Stable Stabilization followed by cERT
Radioresistant Unstable Stereotactic radiosurgery (SRS)
Stabilization followed by SRS
High-grade epidural Radiosensitive Stable Able to tolerate surgery
spinal cord compression Radiosensitive Unstable Unable to tolerate surgery
with or without Radioresistant Stable Able to tolerate surgery
myelopathy Radioresistant Stable Unable to tolerate surgery
Radioresistant Unstable
Radioresistant Unstable
Data from Laufer et al. [4]
Neurologic Status However, once these patients move into a partic-

The neurologic assessment of a patient with spi- ular position, they have an instant exacerbation of
nal metastases with or without CEC or ESCC their pain. Mechanical pain is typically a sharp,
involves the assessment of myelopathy, mechani- shooting pain that follows the dermatomal distri-
cal pain, weakness, and radiculopathy on a neu- bution of the nerve root that is being compressed
rologic exam or patient history. Additionally, the in the neuroforamina with movement.
neurologic assessment involves the evaluation of The second classification of pain is referred
the radiographic degree of ESCC or CEC. pain, which occurs at a location that is distant
from the metastasis and midline point tenderness
on palpation. Both referred pain and point tender-
Pain ness on palpation do not radiate and are indica-
tions of subclinical spinal instability. Referred
Pain is typically the presenting symptom in virtu- pain to the sacroiliac joints and iliac crest can
ally all patients (90–95%) with metastasis to their occur with upper lumbar metastatic lesions.
spine [5, 6]. Given the prevalence of pain as the Referred pain can also occur in the interscapular
presenting symptom, it is important to differenti- region from a cervical metastatic lesion.
ate the quality, form, and distribution of the pain The third classification of pain is biologic
the patient is experiencing. Though many patients pain, which is caused by invasion and irritation of
complain of local back pain, particular types of the periosteum and its innervation by the meta-
back pain are indicative of compression of the static disease. Biologic pain is often worsened
neural elements. while lying supine, possibly due to increased
Mechanical pain is one classification of pain, flow through vertebral veins, and during the early
which is characterized by pain that is exacerbated morning, due to cortisol levels being at their low-
by axial loading such as standing from a seated or est. However, this pain typically improves over
laying down position or lightly pressing on the the course of the day as cortisol levels and its cor-
top of the patients head during a physical exami- responding anti-inflammatory effect rise.
nation. This mechanical pain is typically due to
the destruction of the vertebral body or the poste-
rior bony elements, which results in compression Motor Weakness and Myelopathy
of exiting nerve roots in their neuroforamina with
normal physiologic movement. Patients with Motor weakness and/or myelopathy are clinical
mechanical pain are usually able to find particu- manifestations of neural element compression,
lar positions that alleviate their pain, such as which may require surgical decompression for
lying down if the spinal metastasis is in the lum- treatment. Aside from highly radiosensitive
bar spine or sleeping in an upright position if the tumors that have a robust response to radiation
spinal metastasis is in the thoracic spine. therapy, patients with motor weakness and/or
39 Decision-Making Algorithms for Surgical Treatment of Spine Metastatic Disease 537
myelopathy require an expedited surgical gold standard for detecting and evaluating meta-
decompression to ensure the preservation of static epidural compression, with a sensitivity of
function [2]. 98.5% and a specificity of 98.9% in detecting
Epidural compression in the lumbar spine or lesions. The amount of compression may also
the cauda equina results in lower motor neuron influence if surgical intervention is required. The
symptoms. If the compression is severe, the Bilsky epidural disease grading system is a grad-
patient will have a cauda equina syndrome, which ing scale that is helpful in delineating high-grade
is characterized by low back pain with radiation ESCC, which may require surgical decompres-
into the perineum and legs, saddle anesthesia, sion versus low-grade ESCC, which may require
hyporeflexia, legs weakness, and bladder/bowel simply radiotherapy for disease control [7]. The
dysfunction, which is typically a late finding grading system assigns a grade 0 for no epidural
unless the lesion is at the conus medullaris. disease present; grade 1a for epidural disease
Epidural compression in cervical or thoracic impinging on the thecal sac but no deformation;
spine results in upper motor neuron symptoms grade 1b for epidural disease deforming the the-
and myelopathy. Upper motor neuron symptoms cal sac but not contacting the spinal cord; grade
include neck pain with radiation into the arms 1c for epidural disease deforming the thecal sac
and hands, loss of hand dexterity characterized and contacting the spinal cord; grade 2 for epi-
by an inability to button shirts or write, difficulty dural disease with spinal cord compression with
walking/wide-based gait, hyperreflexia, and pro- CSF visible; and grade 3 for epidural disease
gressive arm/hand weakness. Additionally, if with spinal cord compression with no visible
there is focal compression of exiting nerve roots CSF. High-grade ESCC is used to describe com-
in the cervical or lumbar spine, these patients can pression that is a grade 2 or 3, as this amount of
have specific weakness of the muscle group cord compression typically requires surgical
innervated by that nerve root. intervention for adequate and expeditious decom-
pression of the spinal cord.
In addition to an MRI, a computed tomo-
Numbness/Tingling graphic (CT) scan is helpful to obtain for assess-
ing bony destruction in cases where instability is
Numbness/tingling as a manifestation of neurop- suspected. A CT scan can also guide the extent of
athy and nerve root irritation along a dermatomal instrumentation needed to stabilize a patient, as
distribution are typically well tolerated by the some tumors are lytic while others are sclerotic,
patient. However, the presence of this clinical though in both cases bone will be suboptimal.
finding is an indicator that particular nerve roots Lastly, in cases for which an MRI is contraindi-
are being compressed or irritated by the spinal cated or previous hardware will complicate
metastasis. Therefore, though numbness/tingling imaging assessment, a CT myelography can be
in isolation is not an absolute indication for spi- used to image any spinal cord compression.
nal decompression for the preservation of func-
tion, it may indicate impending spinal instability
or worsening function should the offending Oncologic Assessment
metastasis not be treated promptly.
This subcategory of the NOMS criteria assesses
the likelihood of local tumor control with radia-
Imaging tion and chemotherapy alone versus surgical
decompression followed by radiation and che-
In patients who have new-onset or progressive motherapy. Tumors such as lymphoma, multiple
back or neck pain with a cancer history, a myeloma, and plasmacytoma are highly radio-
contrast-enhanced MRI scan must be obtained sensitive, and bulky metastasis is effectively
for a full clinical assessment and to guide further controlled or eliminated by chemotherapy with-
management. The contrast-enhanced MRI is the out the need for surgical decompression.
However, tumors that are classically radioresis- well as in bone quality. A tool used to classify
tant, such as renal cell carcinoma, non-small cell spinal instability specifically in the oncology
lung carcinoma (NSCLC), thyroid, melanoma, population is the Spine Instability Neoplastic
and hepatocellular carcinoma, require higher Score (SINS) created by the Spine Oncology
radiation doses to treat these tumors effectively Study Group (SOSG). SOSG, an international
[8, 9]. A common radiation dose used for con- group of 30 spinal oncologists, defines spine
ventional external beam radiation is 30 Gy in 10 instability as “a loss of spinal integrity as a
fractions [8–10]. With conventional radiation, result of a neoplastic process that is associated
the spinal cord is within the radiation field; thus, with movement-related pain, symptomatic or
the dose of radiation is limited due to potential progressive deformity, and/or neural compro-
toxicity and radiation injury [11]. Recently, ste- mise under physiological loads” [21, 22]. The
reotactic radiosurgery, which is defined as SINS score, as seen in Table 39.2, is composed
>10 Gy per fraction in typically <5 fractions of six subcategories, including mechanical pain,
[12–20], can be used to effectively treat radiore- spinal alignment, spine location, bone lesion
sistant tumors after surgery to circumferentially quality, spinal alignment, and posterolateral
decompress the thecal sac and allow for contour- involvement of spinal elements. After each sub-
ing around the spinal cord. category is scored, the total summation is used
For patients who do not have an established to guide surgeons to instrument patients with
cancer diagnosis and do not have an acute neuro- higher scores, who likely have spinal instability.
logic compromise, a percutaneous biopsy may The SINS ranges are as follows: total scores of
be warranted to assess if the offending tumor is
radiosensitive. Nevertheless, patients with radio-
sensitive tumors may require decompression and Table 39.2 Spinal Instability Neoplastic Score (SINS)
stabilization, if they demonstrate mechanical Score
stability (to be discussed below). Of note, Location Junctional 3
Mobile Spine 2
although the Patchell study was overwhelmingly Semirigid 1
positive in favor of the surgical group, it excluded Rigid 0
patients with radiosensitive tumors (e.g., lym- Pain Yes 3
phoma and multiple myeloma), multiple (non- Occasional pain but 1
not mechanical 0
contiguous) areas of spinal cord compression, or Pain-free lesion
total paraplegia for longer than 48 hours [2]. Bone lesion Lytic 2
Additionally, it excluded patients with <3 months Mixed (lytic/blastic) 1
of expected survival time due to tumor burden. Blastic 0
Thus, a comprehensive framework for assessing Radiographic spinal Subluxation/ 4
alignment translation present 2
patients with these attributes is needed to decide De novo deformity 0
which patients benefit from surgical decompres- (kyphosis/scoliosis)
sion and instrumentation. Normal alignment
Vertebral body collapse > 50% collapse 3
< 50% collapse 2
No collapse with 1
Mechanical >50% body involved 0
None of the above
The mechanical assessment of spinal instability Posterolateral Bilateral 3
due to tumor invasion and resulting pathologic involvement of spinal Unilateral 1
elements None of the above 0
fractures are the strongest indicator for surgical Total score Stable 0–6
intervention. Spinal instability due to a neoplas- Indeterminate 7–12
tic process differs from traumatic injuries in the Unstable 13–18
pattern of bony and soft tissue involvement, as Data from Fisher et al. [22]
39 Decision-Making Algorithms for Surgical Treatment of Spine Metastatic Disease 539
0–6 indicate spinal “stability,” scores of 7–12 Conclusion

indicate “indeterminate (possibly impending)
instability,” and scores of 13–18 indicate spinal There has been a significant advancement in radi-
“instability.” Thus, the authors recommend ation therapy and chemotherapy in the treatment
instrumentation of patients who have a SINS of metastatic disease; however, surgery has
>13 but allow the treating surgeon discretion on evolved as a major treatment option for patients
instrumenting patients with SINS between 7 and with spinal metastases. Surgical decompression
12 [21, 22]. is particularly useful in treating patients with
Though SINS is a relatively granular tool for high-grade ESCC and CEC, and instrumentation
grading spinal instability and the need for surgi- is useful for treating spinal instability. The use of
cal stabilization, typically the strongest motiva- NOMS framework provides a treatment algo-
tor to decompress and instrument a patient is a rithm that facilitates coordination between a mul-
drastic increase in their pain and decrease in tidisciplinary team to offer safe, reliable, and
their functional status. The primary role for sur- reproducible treatment for patients with meta-
gery in patients with metastatic disease is pallia- static spinal disease.
tive with a goal to prevent the patient from
becoming increasingly immobile and unable to
complete activities of daily living if their pain or References
neurologic compromise is not adequately treated.
Immobility results in increased risk of deep 1. Gilbert RW, Kim J, Posner JB. Epidural spinal cord
venous thrombosis/pulmonary embolisms, atel- compression from metastatic tumor: diagnosis and
treatment. Ann Neurol Off J Am Neurol Assoc Child
ectasis resulting in pneumonia, muscle atrophy, Neurol Soc. 1978;3(1):40–51.
and emotional distress. All of these factors ulti- 2. Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ,
mately reduce the patient’s likelihood of sur- Maruyama Y, Kryscio RJ, et al. A randomized trial of
vival. Thus, our group weighs the patients’ surgery in the treatment of single metastases to the
brain. N Engl J Med. 1990;322(8):494–500.
change in functional status as the strongest rela- 3. Barzilai O, Fisher CG, Bilsky MH. State of the art
tive indicator for surgical decompression and treatment of spinal metastatic disease. Neurosurgery.
instrumentation. 2018;82(6):757–69.
4. Laufer I, Rubin DG, Lis E, Cox BW, Stubblefield MD,
Yamada Y, et al. The NOMS framework: approach to
the treatment of spinal metastatic tumors. Oncologist.
Systemic 2013;18(6):744–51.
5. Helweg-Larsen S, Sørensen PS. Symptoms and signs
The systemic assessment involves the evaluation in metastatic spinal cord compression: a study of pro-
gression from first symptom until diagnosis in 153
of the patient’s burden of disease and the overall patients. Eur J Cancer. 1994;30(3):396–8.
patient’s survival. It allows the surgeon and 6. Bach F, Larsen BH, Rohde K, Børgesen SE, Gjerris F,
oncology team to assess the patient as a whole, Bøge-Rasmussen T, et al. Metastatic spinal cord com-
considering the patient’s medical comorbidities, pression. Acta Neurochir. 1990;107(1–2):37–43.
7. Bilsky MH, Laufer I, Fourney DR, Groff M, Schmidt
the number of systemic tumor metastases, and MH, Varga PP, et al. Reliability analysis of the epi-
their baseline functional status [12, 23, 24]. These dural spinal cord compression scale. J Neurosurg
factors indicate the ability of a patient to tolerate Spine. 2010;13(3):324–8.
a proposed procedure, inform the risk-benefit 8. Mizumoto M, Harada H, Asakura H, Hashimoto T,
Furutani K, Hashii H, et al. Radiotherapy for patients
ratio of treatment, and characterize the likelihood with metastases to the spinal column: a review of 603
of survival outside of the perioperative period. A patients at Shizuoka Cancer Center Hospital. Int J
high likelihood of death within the perioperative Radiat Oncol Biol Phys. 2011;79(1):208–13.
period and high systemic metastatic disease bur- 9. Maranzano E, Latini P. Effectiveness of radiation
therapy without surgery in metastatic spinal cord
den are relative contraindications for surgical compression: final results from a prospective trial. Int
intervention. J Radiat Oncol Biol Phys. 1995;32(4):959–67.
10. Gerszten PC, Mendel E, Yamada Y. Radiotherapy

metastatic sarcomas involving the spine. J Neuro-
and radiosurgery for metastatic spine disease: what Oncol. 2012;107(3):551–7.
are the options, indications, and outcomes? Spine. 18. Song CW, Kim M-S, Cho LC, Dusenbery K, Sperduto
2009;34(22S):S78–92. PW. Radiobiological basis of SBRT and SRS. Int J
11. Lovelock DM, Zhang Z, Jackson A, Keam J,
Clin Oncol. 2014;19(4):570–8.
Bekelman J, Bilsky M, et al. Correlation of local fail- 19. Garcia-Barros M, Paris F, Cordon-Cardo C, Lyden D,
ure with measures of dose insufficiency in the high- Rafii S, Haimovitz-Friedman A, et al. Tumor response
dose single-fraction treatment of bony metastases. Int to radiotherapy regulated by endothelial cell apopto-
J Radiat Oncol Biol Phys. 2010;77(4):1282–7. sis. Science. 2003;300(5622):1155–9.
12. Leithner A, Radl R, Gruber G, Hochegger M, Leithner 20. Yamada Y, Katsoulakis E, Laufer I, Lovelock M,

K, Welkerling H, et al. Predictive value of seven Barzilai O, McLaughlin LA, et al. The impact of his-
preoperative prognostic scoring systems for spinal tology and delivered dose on local control of spinal
metastases. Eur Spine J. 2008;17(11):1488–95. metastases treated with stereotactic radiosurgery.
13. Gerszten PC, Burton SA, Quinn AE, Agarwala
Neurosurg Focus. 2017;42(1):E6.
SS, Kirkwood JM. Radiosurgery for the treatment 21. Fourney DR, Frangou EM, Ryken TC, DiPaola CP,
of spinal melanoma metastases. Stereotact Funct Shaffrey CI, Berven SH, et al. Spinal instability neo-
Neurosurg. 2005;83(5–6):213–21. plastic score: an analysis of reliability and validity
14. Ghia AJ, Chang EL, Bishop AJ, Pan HY, Boehling from the spine oncology study group. J Clin Oncol.
NS, Amini B, et al. Single-fraction versus multi- 2011;29(22):3072–7.
fraction spinal stereotactic radiosurgery for spinal 22. Fisher CG, DiPaola CP, Ryken TC, Bilsky MH,

metastases from renal cell carcinoma: secondary Shaffrey CI, Berven SH, et al. A novel classification
analysis of phase I/II trials. J Neurosurg Spine. system for spinal instability in neoplastic disease:
2016;24(5):829–36. an evidence-based approach and expert consen-
15. Gerszten PC, Burton SA, Ozhasoglu C, Vogel WJ, sus from the Spine Oncology Study Group. Spine.
Welch WC, Baar J, et al. Stereotactic radiosurgery 2010;35(22):E1221–9.
for spinal metastases from renal cell carcinoma. J 23. Dardic M, Wibmer C, Berghold A, Stadlmueller

Neurosurg Spine. 2005;3(4):288–95. L, Froehlich EV, Leithner A. Evaluation of prog-
16. Zelefsky MJ, Greco C, Motzer R, Magsanoc JM, Pei nostic scoring systems for spinal metastases in 196
X, Lovelock M, et al. Tumor control outcomes after patients treated during 2005–2010. Eur Spine J.
hypofractionated and single-dose stereotactic image- 2015;24(10):2133–41.
guided intensity-modulated radiotherapy for extracra- 24. Pereira NRP, Janssen SJ, van Dijk E, Harris MB,
nial metastases from renal cell carcinoma. Int J Radiat Hornicek FJ, Ferrone ML, et al. Development of a
Oncol Biol Phys. 2012;82(5):1744–8. prognostic survival algorithm for patients with meta-
17. Chang UK, Cho WI, Lee DH, Kim MS, Cho CK, Lee static spine disease. JBJS. 2016;98(21):1767–76.
SY, et al. Stereotactic radiosurgery for primary and
Part VI
Surgical Treatment of Spinal Metastases
Biomechanics of Spinal Fixation
in Metastatic Disease
40
Allen L. Ho and Atman M. Desai
Introduction tion is necessary in both of these scenarios to treat

or prevent the symptoms of spinal instability.
Spinal bony metastases are the most common Complications following surgical resection
neoplasms encountered in the spine. They have a and stabilization of spinal metastasis are sig-
slight male predominance, peak between the ages nificant, ranging from 10% to 52% [1–5].
40 and 60, and occur most frequently in the tho- However, a randomized controlled trial pub-
racolumbar spine. Typical clinical presentation lished by Patchell et al. in 2005 demonstrated
includes a combination of symptoms arising the superiority of decompression with instru-
from mechanical pain, radiculopathy, instability, mented stabilization where patients had evi-
and neurologic deficits. Treatment of spinal dence of spinal instability, followed by external
metastases is typically multimodal with options beam radiation therapy over external beam
including surgery, radiation therapy, and sys- radiation therapy alone for symptomatic
temic chemo- or immune-therapy. patients with spinal metastasis causing spinal
When surgically treating spinal metastases, cord compression. Patients undergoing surgery
surgical goals typically include tissue diagnosis had significant improvements in ambulation,
and reduction of oncological burden, decompres- bowel and bladder function, and overall sur-
sion of neural elements, and restoration of spinal vival [6]. This has led to the wide adoption of
stability and spinal alignment. Mechanical insta- surgical decompression and stabilization for
bility of the spine secondary to metastatic disease symptomatic spinal metastasis. In parallel, over
can cause significant pain, neurological deficit, the past two decades, the development of ste-
spinal deformity, and disability. Such instability reotactic radiosurgery (SRS) has led to the
can arise from the osteolytic nature of the meta- newer paradigm of combined judicious use of
static lesions themselves – pathological frac- surgical and radiosurgical approaches. This
tures – or may be iatrogenic in origin as a result of strategy, commonly referred to as “separation
tumor resection. Spinal fixation with instrumenta- surgery,” typically utilizes surgery for rapid
treatment of epidural cord compression and spi-
nal instability followed by SRS to contain the
bulk of nonepidural metastatic disease. Similar
A. L. Ho · A. M. Desai (*)
outcomes of symptom improvement and tumor
Department of Neurosurgery, Stanford University
School of Medicine, Stanford, CA, USA control can be achieved while minimizing sur-
e-mail: Atman@stanford.edu gical morbidity with this approach [7].

https://doi.org/10.1007/978-3-030-42958-4_40
544 A. L. Ho and A. M. Desai
Diagnosis and Decision-Making such as the cervical spine receives more points

than those in stable segments such as the thoracic
Modern decision-making in the treatment of spi- spine and sacrum. Mechanical pain receives more
nal metastasis needs to incorporate all possible points than local biologic pain from periosteal
treatment modalities and approaches. The most stretching. Any subluxation, translation, or verte-
comprehensive framework for decision-making bral body collapse greater than 50% can lead to
is the NOMS decision framework that consists of instability. Finally, evaluation of posterior element
four fundamental components [8]: involvement, including bilateral pedicle, facet, or
costovertebral joints, will receive more points. The
• N – Neurologic: includes presence of myelop- score ranges from 0 to 18 points, with 0–6 classi-
athy, functional radiculopathy, and the degree fied as stable, 7–12 as potentially unstable, and
of epidural cord compression appreciated on 13–18 as unstable [11, 12] (Table 40.1).
imaging
Table 40.1 Spinal instability neoplastic score (SINS)
• O – Oncologic: includes ability to achieve scoring
local, dural control and, thus, reflects the
Element of SINS Score
radio/chemosensitivity of tumor type
Location
• M – Mechanical: assesses the degree of insta- Junctional (occiput– 3
bility or the spine’s ability to withstand physi- C2, C7–T2, T11–L1,
ologic loads without pain, deformity, or L5–S1)
neurologic deficit Mobile spine (C3–C6, 2
L2–L4)
• S – Systemic” assesses both the degree of sys-
Semirigid (T3–T10) 1
temic tumor burden and systemic medical dis- Rigid (S2–S5) 0
ease and comorbidities Pain relief with recumbency and/or pain with
movement/loading of the spine
The vertebral body is the most common loca- Yes 3
tion for seeding of spinal metastasis. As the pri- No (occasional pain 1
but not mechanical)
mary axial-load bearing structure in the spine,
Pain-free lesion 0
bony replacement and destruction of the vertebral Bone lesions
body by metastatic lesions lead to instability and Lytic 2
compression fracture. The most important predic- Mixed (lytic/blastic) 1
tors of instability from vertebral body metastasis Blastic 0
include the cross-sectional area of remaining Radiographic spinal alignment
tumor-free vertebral body, tumor size, and bone Subluxation/ 4
translation present
mineral density [9, 10]. Mechanical back pain is
De novo deformity 2
the most common presenting symptom for spinal (kyphosis/scoliosis)
instability related to metastatic disease. While bio- Normal alignment 0
logic back pain related to periosteal stretching Vertebral body collapse
from the tumor itself is mainly nocturnal or early >50% collapse 3
morning pain that improves throughout the day, <50% collapse 2
No collapse with 1
patients with mechanical back pain have pain that >50% of body
worsens with movement and is localized to the involved
level of involvement. The Spinal Instability None of the above 0
Neoplastic Score (SINS) was devised to help sur- Posterolateral involvement of the spinal elements
geons as well as radiation and medical oncologists (facet, pedicle, or CV joint fracture or replacement with
tumor)
predict the degree of spinal instability based on six
Bilateral 3
components: location, pain, bone lesion, radio- Unilateral 1
graphic spinal alignment, vertebral body collapse, None of the above 0
and posterolateral element involvement. Tumor From Fisher et al. [91]. Reprinted with permission from
location in more mobile segments of the spine Wolters Kluwer Health, Inc.
40 Biomechanics of Spinal Fixation in Metastatic Disease 545
Biomechanics of Surgical a posterior-only or combined 360-degree

Stabilization approach with a thoracotomy or retroperitoneal
approach. Lumbar levels are access via retroperi-
In the past, decompressive laminectomy was the toneal, transabdominal, or posterior transpedicu-
standard procedure to address neurologic com- lar approaches. Finally, sacral lesions may require
pression from symptomatic spinal metastasis. most complex posterior approaches or even ante-
However, this approach fails to address patho- rior approaches through the pelvis.
logic fractures and may paradoxically increase Given the degree of vertebral body resection
instability by removing posterior elements of the required for complete or meaningful resection of
spinal column. Therefore, aggressive surgical symptomatic and/or compressive spinal metasta-
decompression with spinal stabilization has sis, many cervical and generally all thoracolum-
become the main surgical approach [13, 14] and bar resections of spinal metastases will be
remains the most effective technique for reducing inherently destabilizing and require careful con-
pain and improving quality of life (QOL) in sideration of reconstruction strategies. The large
patients with symptomatic spinal metastasis [15]. gap created by spondylectomies creates a chal-
These lesions most commonly arise anterior to lenge of arthrodesis, and a robust construct
the spinal cord in the vertebral body and progress design is essential to withstand the mechanical
posteriorly to cause compression of the spinal stresses on the spinal column until arthrodesis
cord and/or exiting nerve roots. These lesions are can be achieved. Recreation and augmented sup-
easily accessed via an anterior approach in the port of all three columns of the spine is impera-
cervical spine. Anterior decompression followed tive [21]. Anterior column reconstruction is most
by anterior fixation and supplemented with pos- commonly performed with utilization of carbon
terior instrumentation (with additional decom- fiber or titanium expandable cages or polymeth-
pression if necessary) is generally the most ylmethacrylate (PMMA) [22–24]. Posterior spi-
common strategy for subaxial cervical lesions. nal stabilization can help ensure posterolateral
Anterior approaches in the thoracic spine are arthrodesis and is performed with standard screw
more challenging, especially in the upper tho- and rod fixation [25, 26]. Most of these patients
racic spine (T1–T4) due to the anatomic relation- will need additional postoperative treatment with
ship to the sternum that may necessitate a adjuvant radiation therapy and imaging surveil-
sternotomy or thoracotomy [16]. Similarly, lance for recurrence. Thus, careful implant selec-
because of the anatomic location of the aortic tion to minimize interference with therapy and
arch and great vessels in relation to thoracic lev- imaging artifact is also important.
els T5 through T10, a right-sided thoracotomy is
recommended for anterolateral approaches to the
spine unless the target lesion is exclusively on the Interbody Grafts
left side of the spine [17, 18]. Because of the dif-
ficulty of pure anterior approaches, development Although autograft bone remains the gold stan-
of posterior-only approaches to the ventral tho- dard construct material for arthrodesis, most ven-
racic spine (transpedicular, costotransversec- tral constructs for reconstruction of vertebral
tomy, and lateral extracavitary) is increasingly body removed from metastatic lesion resection
utilized [19]. However, these posterior approaches are now a combination of synthetic and allograft
often involve division of nerve roots and segmen- material [27]. Long-term solid bony fusion may
tal vessels that can increase risk of ischemic not be necessary if life expectancy is less than
injury to the spinal cord. Animal studies have 18 months, and autogenous donor sites may also
shown that interruption of bilateral segmental have tumor involvement. Moreover, in patients
arteries of four or more consecutive level risked with metastatic disease, bone autograft carries
ischemic spinal cord damage [20]. The remaining significant short-term donor-site morbidity [28].
thoracolumbar junction is typically accessed via Postoperative local irradiation and chemotherapy
can also interfere with bone remodeling and ity to bone, which decreases its risk of subsidence
fusion [29, 30]. Local bone autograft can be plen- in comparison to titanium. PEEK is also radiolu-
tiful, especially in posterior and posterolateral cent and also has magnetic resonance imaging
exposures of the spine that include significant (MRI) compatibility without artifact. Both PEEK
bony decompression. Without incurring the mor- and carbon fiber improve visualization on post-
bidity of iliac crest harvest, local bone can be just operative imaging without artifact, and while
as effective in achieving short-segment lumbar PEEK has a more favorable modulus of elasticity
fusion [31, 32]. Cadaveric fibular allograft is pre- compared to carbon fiber, carbon fiber is more
ferred to autologous iliac crest since it avoids osteoinductive and may allow for an increased
complications associated with iliac crest harvest. degree of cellular integration that can help sup-
Additionally, it can be tailored to any length and port fusion [38–40]. Finally, osteoconductive
provides a central packing channel for local bone substitutes can be either mixed with a local-
autologous bone graft or other cancellous bone ized autograft bone or crushed cancellous
substitute to enhance fusion. However, it carries a allograft bone to form a packing material for
higher modulus of elasticity which confers a sig- interbody implants and as a substrate to facilitate
nificant risk for “pistoning” or subsidence. dorsal bony fusion. Demineralized bone matrix
Though allograft fibula is slower to incorporate (DBM) is the most common of these substitutes
than autologous iliac crest [33], there has been no that supplies potent bone morphogenetic proteins
significant difference in pseudoarthrosis rates (BMPs) to the fusion bed on a collagenous con-
between the two identified [34, 35]. ductive substrate. High-dose recombinant human
Interbody cages provide anterior column sta- BMPs, primarily rhBMP-2, can be quite effective
bilization with synthetic materials such as stain- in promoting bony fusion but are contraindicated
less steel or titanium. Titanium alloys provide a in tumor surgery given the oncogenic properties
high tensile strength while retaining a reasonable of these agents. The most common hardware
degree of malleability and biocompatibility. They complication encountered with graft placement is
also have less imaging artifact compared to stain- a result of dislodging at either the proximal or
less steel. Titanium mesh cages allow for easy distal ends where the graft sits against adjacent
selection for cage length and can be filled with spinal segments. Placing the graft under com-
autologous graft material. Although the modulus pression serves the dual purpose of stabilizing the
of elasticity is more rigid than vertebral bone graft in the interbody space and promoting fusion
with some risk of subsidence, clinically signifi- by ensuring a stable contact surface for bony
cant subsidence causing deformity or pseudoar- ingrowth to occur [41].
throsis is rare [36]. Expandable titanium cages
allow for deployment of full cage lengths after
placement in the vertebral body cavity. This Pedicle Screws
allows for ease of placement with a smaller pro-
file implant that is especially advantages in poste- Pedicle screws apply force to the spine by fixed
rior and posterolateral approaches to the ventral moment arm cantilever beam fixation. The ped-
spinal column where the working corridor is icle screw represents the fixation point that sup-
smaller than the resection cavity. Expansion of ports the cantilever, either a rod or plate that
the ventral graft in situ allows for additional sim- rigidly buttresses the spine, thereby resisting
ple distraction fixation to resist axial loads and axial loads. With fixed pedicle screw and rod/
well as correct any cervical deformity be restor- plate constructs, there is no load sharing with
ing natural lordosis [37]. the anterior column, and the bulk of the stress is
Other synthetic materials, such as poly- borne by the screw/rod or screw/plate junction
etheretherketone (PEEK) and carbon fiber cages, which can lead to failure [42]. Dynamic, or non-
are also increasing utilized. PEEK is a semicrys- fixed, pedicle screw fixation systems allow for
talline polyaromatic polymer with similar elastic- some toggling of the screws and constitute a
nonfixed moment arm cantilever beam fixation. normal, instrumentation failure with pedicle
There is some axial-load transfer onto the ante- screws in patients with osteoporosis remains
rior column, which decreases stress at the screw relatively common. A longer construct can
rod/plate junction. However, the toggling of the supply more points of fixation and distribute
screw leads to increased failure via screw pull- the load over more segments in osteoporotic
out. Thus, fixed moment arm systems are more patients, decreasing the pullout load at any
effective at resisting sagittal translation and specific screw site. Consideration may be
especially useful for deformity correction. given to the judicious use of cement augmen-
Different screw modifications have been devised tation of pedicle screws. An intact cortical sur-
to combat the problem of screw pullout. Screw face must exist prior to injection of PMMA
strength is directly proportional to the cube of into a pedicle screw track to prevent leakage.
the core or minor screw diameter. Screw pullout Pedicle screws must then be rapidly inserted
strength is directly proportional to the volume after injection of PMMA. Alternatively, newer
of bone between screw threads. This is deter- fenestrated screws allow for direct injection of
mined by the screw thread depth (outside diam- PMMA down the hollow bore of these screws
eter) and pitch (distance from one thread to to extrude out of the shaft at fenestration
another). Increasing the pitch and thread depth points. PMMA may increase the pullout
will increase the pullout strength by increasing strength of pedicle screws up to threefold in
the bone volume between screw threads. The osteoporotic vertebrae [46, 47].
angulation or shape of the thread affects the
bone volume and pullout strength. Varying the
core (minor) diameter to increase screw depth Biomechanics of Cervical Fixation
along the distal end but increasing diameter to
increase screw strength where they are most like Occipitocervical Fixation
to fail (conical shaped screws) near the tulip
head can be another strategy for increasing pull- Occipitocervical fixation is typically required
out strength. Undertapping of pedicle trajecto- when there is destruction of one or both occipital
ries or utilization of self-tapping screws can also condyles. This may be from osteolytic tumor or a
increase the pullout resistance of a screw. Axial far lateral approach to the foramen magnum. This
loading can lead to screw failure due to the par- approach is generally taken to improve visualiza-
allelogram-like translational motion created. tion to the ventral or ventrolateral craniocervical
Directing screw trajectories medially (“toeing junction [48]. Generally, up to 70% resection of
in”) and utilization of transverse connectors can one occipital condyle (resection up to the hypo-
help limit this motion and prevent screw pullout glossal canal) is well-tolerated without the need
with axial loading [43]. Trajectories aligned for fusion. Any further resection increases the
along cortical bone surfaces have also shown to likelihood of occipitocervical instability. Fixation
have superior pullout strength, given implanta- is achieved most commonly with occipital plat-
tion of screw threads within more rigid cortical ing. Several indications have been identified for
bone versus cancellous bone within the verte- fusion in these cases: identification of a painful
bral body [44, 45]. head tilt, instability on flexion–extension radio-
Bone quality plays an important role in ped- graphs, or complete resection of the occipital
icle screw pullout. Patients with spinal metasta- condyle [49].
ses are particularly vulnerable to poor bone
quality as they may have tumor at multiple lev-
els affecting bone quality or simply have con- Cervical Spine Fixation
current osteopenia or osteoporosis. Though
pedicle screws are superior to hooks and sub- More complete resections of metastatic lesions
laminar wires when bone mineral density is of the cervical spine generally require either
partial or complete corpectomy, causing ventral lesional or iatrogenic disruption of posterior

instability. Cervical constructs need to be elements [27]. It can also be helpful in prevent-
designed to correct this instability while providing postlaminectomy kyphosis [52]. Especially
ing a necessary substrate for osseous fusion. The with multilevel corpectomy, there is evidence
main modes of application for this purpose to suggest increased fusion rates and less
include simple distraction and cantilever beam kyphosis with the addition of posterior instru-
fixation. Interbody grafts create simple distrac- mentation [53, 54]. Generally, tumors that
tion fixation by applying a distraction force ven- invade both ventral and dorsal neural elements
trally to resist axial loads and kyphosis. In this and cause kyphotic deformity are good candi-
way, these grafts reconstitute the ventral load- dates for a combined approach [55]. Lateral
bearing column of the cervical spine. There is mass screw and rod constructs generally behave
also some stability imparted in flexion, exten- as nonfixed moment arm cantilever beam fix-
sion, axial rotation, and lateral bending move- ators, while also providing some dorsal tension-
ments [50]. However, some sort of fixation, band fixation. These devices restore stiffness to
either via ventral plating or posterior instrumen- the cervical spine in flexion, extension, and tor-
tation, is necessary to provide full stability, given sion [56]. Lateral mass screws provide superior
the degree of vertebral body removal necessary flexion/extension stability and torsion resis-
to resect most symptomatic metastatic lesions. tance compared to wiring constructs [57–59].
The principle of cantilever beam fixation is Lateral mass fixation can be achieved from C1
achieved with ventral cervical plating systems to T1 (Fig. 40.1), but modern systems allow for
with locking screws as well as rigid posterior lat- integration with various fixation techniques
eral mass/rod instrumentation. The fixed move- across the occipitocervical and cervicothoracic
ment arm of these cantilever beam devices junctions. Similar to ventral plating, posterior
allows for axial-load sharing across the con- screw instrumentation should be placed only in
struct. Ventral plate and screw constructs provide normal stable bone free of metastatic disease.
immediate internal stability, recreate the ventral Adequate spinal alignment must be achieved
tension band, and provide stabilization and resis- before instrumentation because lateral mass
tance to abnormal motion, especially flexion and screws are not optimized for adjustments in spi-
extension [51]. They should be applied to the nal alignment. The normal 3.5-mm diameter
completely intact vertebral bodies above and lateral mass screws utilized are not large enough
below the level(s) of interest, spanning the to correct kyphotic deformity or reduce signifi-
entirety of the interbody graft. Plating systems cant translation or subluxation. Cancellous
that utilize nonlocking, variable angle screws are screws provide better purchase than those with
more dynamic implants that allow graft exposure cortical threads. Though not mandatory, safe
to more continuous axial loading that may facili- bicortical fixation can be achieved with 14- to
tate bone fusion. Generally, though bicortical 16-mm screws. Rescue screw placement with
screw placement achieves greater pullout slightly larger diameter screws can improve
strength, the added complexity and risk of plac- bony purchase in patients with osteoporotic
ing these screws in the cervical spine have made bone. However, larger screws run the risk of
unicortical screw placement the standard for cer- fracturing the lateral mass [60]. Alternatively, a
vical plate fixation. small amount of polymethyl methacrylate
Dorsal cervical stabilization following cer- (PMMA) may also be infused into the hole
vical metastasis resection is most commonly prior to screw placement. Finally, cervical
achieved with lateral mass-based instrumenta- transfacet screws can also serve as salvage fixa-
tion. Dorsal decompression of the spinal cord tion [61]. In addition to traditional dorsal peri-
can be advantageous in cases of circumferential osteal decortication and placement of bone
epidural spinal cord compression, and posterior graft to encourage arthrodesis, placement of
stabilization is necessary if there is excessive interfacet grafts can help encourage fusion
Fig. 40.1 Fixation following resection of C2 metastasis. ittal image [top right]). Patient underwent multilevel cer-
MRI imaging of a patient with destructive colon cancer vical decompression, reduction of C1/2 fracture, and
metastasis with a C2 metastasis causing a pathologic frac- C1–C6 posterior cervical fusion followed by Cyberknife
ture at C2 with atlantoaxial instability and epidural spinal radiosurgery to both lesions (postoperative lateral XR
cord compression, and a separate lesion causing cervical [bottom right]). Cervical fixation extended to cover all
stenosis at C5/6 (T1-weighted contrast-enhanced sagittal levels of involvement given extent of disease and need for
[top left] and axial [bottom left] images, T2-weighted sag- potentially destabilizing adjuvant radiosurgery
across the facets. These spacers can also help ally range from 3 to 3.5 mm and are easily con-
increase foraminal area and provide additional toured to fit. Excessive force in rod persuasion
stiffness to cervical constructs to enhance or in set-screw final tightening with antitorque
fusion [62, 63]. Titanium rod diameters gener- should be avoided since the lateral mass is more
fragile than the pedicle and more prone to frac- side-to-side domino connector extending
ture or screw pullout. Screwheads are generally between two separate rods of the same diameters
polyaxial to allow for maximal degrees of free- (3.5 and 5.5 mm). Utilization of a side-to-side
dom for accommodation of rod fixation. connector was found to have similar stiffness but
Anatomically, the C7 and T1 lateral masses are lower ultimate and yield force [71]. Finally, the
smaller than other cervical levels and, gener- length of the construct must be chosen carefully
ally, pedicle screws are preferred at these lev- by tailoring to the curves surrounding the junc-
els. Cervical pedicle screw placement at levels tion. Focal and gradual curves, as well as apical
C6 and above carries a higher risk of vertebral and neutral vertebrae in both sagittal and coronal
artery injury and is, thus, utilized only in exten- planes, must be identified. Apical vertebrae are
uating circumstances [64]. those located at the apex of curvatures in the
spine, while neutral vertebrae are the least angu-
lated and typically located between the curves
Cervical–Thoracic Junction [72]. Constructs should not end at or near apical
vertebrae because angulation of the endogenous
Metastasis located at the cervicothoracic junction curve at that level will increase the loads experi-
provides a unique challenge due to the critical enced in relation to the construct and accelerate
surrounding anatomic structures and the unique adjacent level breakdown. Similarly, constructs
biomechanical considerations of transitioning should not end at the level of the junction since it
from a mobile and lordotic cervical spine to a is prone to angular and translational deformation
rigid and kyphotic thoracic spine. Generally, between the relatively flexible cervical spine and
decompression without fusion at the cervicotho- the rigid thoracic spine stabilized by a ventral rib
racic junction predisposes toward postlaminec- cage. Other considerations include avoiding end-
tomy kyphotic deformity [65]. Laminectomy ing at levels with advanced degenerative disease
disrupts the posterior tension band and shifts the and/or spinal stenosis, as well as extending con-
weight-bearing axis ventrally, putting the dorsal structs in patients with poor bone quality and
muscle groups at a significant mechanical disad- density for extra points of fixation [73].
vantage [66]. Thus, posterior instrumentation and
fixation across the junction should be strongly
considered. Several studies comparing the meth- Biomechanics of Thoracolumbar
ods of fixation at the cervicothoracic junction Fixation
find that lateral mass screws, pedicle screws, and
ventral interbody grafting with or without plating Anterior Fixation
can all provide adequate stabilization across the
junction [67–71]. The degree of three-column Approaches to metastasis in the thoracolumbar
involvement is crucial to selecting the correct sta- spine encompass anterior, anterolateral, lateral,
bilization strategy. In biomechanical studies, dor- posterolateral, and posterior techniques.
sal fixation alone is sufficient to stabilize a dorsal Approximately two-thirds of spinal metastasis
two-column injury but not with involvement of are found in the vertebral body and pedicles
the anterior column [69]. Thus, ventral and dorsal necessitating access to the ventral spinal column.
instrumentation, mostly common in the form of a Anterior, anterolateral, or lateral approaches
ventral interbody graft plus posterior screw and involve a thoracotomy or retroperitoneal expo-
rod fixation, should be utilized for three-column sure to complete a corpectomy. If the posterior
injuries [70]. In terms of rod choice, there were elements are intact, then posterior instrumenta-
no significant differences in flexion bending and tion following these approaches may not be nec-
axial rotation between a transitional dual- essary from T1 to T9 levels as they are buttressed
diameter rod (3.5 and 5.5 mm) versus a solid by the rib cage. Supplemental instrumentation
may help provide additional support in cases umn via transpedicular, costotransversectomy,
where the tumor or tumor resection has disrupted or lateral extracavitary exposures, allowing sur-
the anterior and middle columns [74, 75]. These geons to avoid the morbidity associated with a
include single-rod [76], double-rod [77], or ante- thoracotomy or retroperitoneal anterior expo-
rior plate and screw constructs. An anterior dual- sures (Fig. 40.2).
rod construct allows for rigid stabilization against These approaches generally necessitate pedi-
axial compression, flexion, extension, and rota- cle screw fixation of at least two levels above and
tion. Single-rod construct provides a lesser below the involved segments to provide for mul-
degree of stabilization against flexion, exten- tiple points of fixation above and below what is
sion, and rotational forces [78]. Anterior plate essentially a three-column injury created by the
and screw fixation are similar to dual-rod con- lesion and resultant surgical exposure. Though
structs in that it provides added resistance short-segment (one level above and below)
against flexion, extension, and rotation by rec- fusions have shown efficacy in thoracolumbar
reating a portion of the ventral tension band and trauma [82, 83], loss of anterior column integrity
employing fixed moment arm cantilever beam with metastatic lesions leads to higher rates of
fixation. However, they lack the ability to dis- short-segment fixation failure when not supple-
tract or compress the vertebral bodies in order mented with anterior column reconstruction or
to accommodate graft reconstruction or com- extension of posterior fusion constructs [84, 85].
press a graft. Single-rod constructs may be nec- Longer constructs also distribute the load over
essary when the vertebral bodies are small or more segments, which is particularly helpful in
partially destroyed by tumor [79]. Above T10, patients with osteoporotic bone. Constructs
vertebral body size makes anterolateral screw should be designed to avoid ending at intermedi-
placement more difficult, though in select cases ate junctions (cervicothoracic and thoracolumbar
screws may be placed as high as T6. Below L4, junctions) given the transitional anatomy present
the iliac veins and origin of the Inferior vena at these points. Ending long constructs at these
cava (IVC) also impede safe placement of junctions leads to higher implant loads, higher
anterolateral screws. failure rates, and a greater likelihood of adjacent
segment disease. More rigid fixation is required
at the terminal ends of the spine, and a higher
Posterior Fixation flexion–extension bending moment exists at the
lumbosacral junction. Thus, multiple points of
Posterior fixation via pedicle screw and rod fixation are required for long constructs ending at
instrumentation has become the standard for tho- L5–S1 and addition of iliac screws, or S2-alar-
racolumbar tumor resection. If there is posterior iliac screws can decrease the strain on S1 screws
element or dorsal/circumferential epidural and improve fusion rates [86, 87] (Fig. 40.3).
involvement, then posterior instrumentation will Longer rod constructs also lead to greater tor-
be necessary since the posterior column will be sional forces on rods that can lead to rod fracture
disrupted to achieve tumor resection and com- and pseudoarthrosis. Cross-fixation between
plete neural element decompression. For lesions long-segment rods can improve the torsional sta-
involving levels T10 and below, there is little or bility and lateral bending stiffness of a construct
no additional support from the rib cage and a [88]. This is more crucial when using hook
greater degree of extension with spinal motion. anchors to improve hook stability and not as
Supplemental posterior instrumentation is gen- advantageous with distal pedicle screw fixation
erally required to prevent excessive motion than [89]. A box construct of two cross-links is the
can lead to graft extrusion [80, 81]. Finally, optimal configuration, ideally with links placed
advances in posterolateral approaches can also at the junction of the middle and terminal thirds
afford increasing access to the ventral spinal col- of the construct [90].
Fig. 40.2 Fixation following resection of thoracic metas- was taken to resect the lesion and complete a corpectomy
tasis. Total invasion of T3 vertebral body with metastasis at this level with removal of disc above and below. An
leading to complete collapse and vertebra plana causing a expandable titanium cage was utilized to correct the
severe kyphotic deformity and ventral cord impingement kyphotic deformity and deployed and expanded gradually
(T1-weighted contrast-enhanced sagittal [top left] and endplate to endplate. Thoracic pedicle screws were placed
axial [bottom left] images, T2-weighted sagittal image two levels above and below for stabilization and fusion
[top right]). A right-sided costotransversectomy approach (postoperative lateral XR [bottom right])
Fig. 40.3 Fixation following resection of lumbar metasta- and support prior to complete destabilization of the spine via
sis. Patient with low back pain and bilateral lower extremity decompression of the L5 level. A left-sided transpedicular
radiculopathy was found to have a liposarcoma at the L5 decompression with complete removal of the facet joint
level with invasion of the epidural space causing central and allowed complete access to the vertebral body and ventral to
foraminal stenosis that was unresponsive to chemotherapy the thecal sac. Total resection of the epidural lesion was
(T1-weighted contrast-enhanced sagittal [top left] and axial achieved and partial corpectomy of the L5 level revealed
[bottom left] images, T2-weighted sagittal image [top only partial tumor invasion and normal bone margins anteri-
right]). Patient was taken to the operating room for potential orly. The decision was made to then leave the majority of the
full L5 corpectomy and multilevel fusion with pelvic fixa- vertebral body in place for adjuvant radiosurgery, and the
tion. Pedicle screws were placed at L2, L3, L4, S1, and the lumbar spine was stabilized with screw and rod instrumenta-
pelvis prior to decompression of the level for stabilization tion (postoperative lateral XR [bottom right])
Conclusion patients with spinal metastases: a systematic review.

Int J Oncol. 2011;38:5–12.
11. Fourney DR, Frangou EM, Ryken TC, et al. Spinal
An understanding of biomechanical principles is instability neoplastic score: an analysis of reliability
essential toward individualizing fixation strate- and validity from the spine oncology study group. J
gies in patients with metastatic spine lesions. Clin Oncol. 2011;29:3072–7.
12. Fisher CG, Schouten R, Versteeg AL, et al. Reliability
Patients with spinal metastases have unique man- of the Spinal Instability Neoplastic Score (SINS)
agement issues related to osteoporosis and per- among radiation oncologists: an assessment of insta-
formance status that necessitate considered bility secondary to spinal metastases. Radiat Oncol.
planning of stabilization strategies. Finally, given 2014;9:69.
13. Witham TF, Khavkin YA, Gallia GL, et al. Surgery
the long-term survival of many patients with spi- insight: current management of epidural spinal cord
nal metastases, fixation strategies that are biome- compression from metastatic spine disease. Nat Clin
chanically sound will mitigate against future Pract Neurol. 2006;2:87–94; quiz 116
hardware failure or adjacent segment disease. 14. Quraishi NA, Gokaslan ZL, Boriani S. The sur-

gical management of metastatic epidural com-
pression of the spinal cord. J Bone Joint Surg Br.
2010;92–B:1054–60.
15. Schwab JH, Gasbarrini A, Cappuccio M, et al.

References Minimally invasive posterior stabilization improved
ambulation and pain scores in patients with plasma-
1. Harrington KD. Anterior cord decompression and spi- cytomas and/or metastases of the spine. Int J Surg
nal stabilization for patients with metastatic lesions of Oncol. 2011;2011:1–5.
the spine. J Neurosurg. 1984;61:107–17. 16. Cohen ZR, Fourney DR, Gokaslan ZL, et al. Anterior
2. Cooper PR, Errico TJ, Martin R, et al. A system- stabilization of the upper thoracic spine via an “inter-
atic approach to spinal reconstruction after anterior aortocaval subinnominate window”: case report and
decompression for neoplastic disease of the thoracic description of operative technique. J Spinal Disord
and lumbar spine. Neurosurgery. 1993;32:1–8. Tech. 2004;17:543–8.
3. Bauer HC, Wedin R. Survival after surgery for spi- 17. Fourney DR, Gokaslan ZL. Anterior approaches for
nal and extremity metastases. Prognostication in 241 thoracolumbar metastatic spine tumors. Neurosurg
patients. Acta Orthop Scand. 1995;66:143–6. Clin N Am. 2004;15:443–51.
4. Walsh GL, Gokaslan ZL, McCutcheon IE, et al. 18. Stulík J, Vyskocil T, Bodlák P, et al. Injury to major
Anterior approaches to the thoracic spine in patients blood vessels in anterior thoracic and lumbar spinal
with cancer: indications and results. Ann Thorac Surg. surgery. Acta Chir Orthop Traumatol Cechoslov.
1997;64:1611–8. 2006;73:92–8.
5. Mazel C, Balabaud L, Bennis S, et al. Cervical and 19. Lubelski D, Abdullah KG, Mroz TE, et al. Lateral
thoracic spine tumor management: surgical indica- extracavitary vs. costotransversectomy approaches
tions, techniques, and outcomes. Orthop Clin North to the thoracic spine: reflections on lessons learned.
Am. 2009;40:75–92, vi–vii Neurosurgery. 2012;71:1096–102.
6. Patchell RA, Tibbs PA, Regine WF, et al. Direct 20. Kato S, Kawahara N, Tomita K, et al. Effects on spinal
decompressive surgical resection in the treatment of cord blood flow and neurologic function secondary
spinal cord compression caused by metastatic can- to interruption of bilateral segmental arteries which
cer: a randomised trial. Lancet (London, England). supply the artery of Adamkiewicz: an experimen-
2005;366:643–8. tal study using a dog model. Spine (Phila Pa 1976).
7. Laufer I, Iorgulescu JB, Chapman T, et al. Local dis- 2008;33:1533–41.
ease control for spinal metastases following “sepa- 21. Denis F. The three column spine and its significance
ration surgery” and adjuvant hypofractionated or in the classification of acute thoracolumbar spinal
high-dose single-fraction stereotactic radiosurgery: injuries. Spine (Phila Pa 1976). 1983;8:817–31.
outcome analysis in 186 patients. J Neurosurg Spine. 22. Snell BE, Nasr FF, Wolfla CE. Single-stage thora-
2013;18:207–14. columbar vertebrectomy with circumferential recon-
8. Bilsky M, Smith M. Surgical approach to epidural spi- struction and arthrodesis: surgical technique and
nal cord compression. Hematol Oncol Clin North Am. results in 15 patients. Neurosurgery. 2006;58:ONS-
2006;20:1307–17. 263–8; discussion ONS-269.
9. Taneichi H, Kaneda K, Takeda N, et al. Risk factors 23. Sciubba DM, Gallia GL, McGirt MJ, et al. Thoracic
and probability of vertebral body collapse in metasta- kyphotic deformity reduction with a distractible
ses of the thoracic and lumbar spine. Spine (Phila Pa titanium cage via an entirely posterior approach.
1976). 1997;22:239–45. Neurosurgery. 2007;60:223–30.. discussion 230–1
10. Weber MH, Burch S, Buckley J, et al. Instability and 24. Hofstetter CP, Chou D, Newman CB, et al. Posterior
impending instability of the thoracolumbar spine in approach for thoracolumbar corpectomies with
expandable cage placement and circumferential 38. Brantigan JW, Steffee AD, Lewis ML, et al. Lumbar
arthrodesis: a multicenter case series of 67 patients. J interbody fusion using the Brantigan I/F cage for
Neurosurg Spine. 2011;14:388–97. posterior lumbar interbody fusion and the variable
25. Fourney DR, Abi-Said D, Lang FF, et al. Use of ped- pedicle screw placement system: two-year results
icle screw fixation in the management of malignant from a Food and Drug Administration investigational
spinal disease: experience in 100 consecutive proce- device exemption clinical trial. Spine (Phila Pa 1976).
dures. J Neurosurg. 2001;94:25–37. 2000;25:1437–46.
26. Wang JC, Boland P, Mitra N, et al. Single-stage
39. Christensen FB, Hansen ES, Eiskjaer SP, et al.

posterolateral transpedicular approach for resection Circumferential lumbar spinal fusion with Brantigan
of epidural metastatic spine tumors involving the cage versus posterolateral fusion with titanium
vertebral body with circumferential reconstruction: Cotrel-Dubousset instrumentation: a prospective, ran-
results in 140 patients. Invited submission from the domized clinical study of 146 patients. Spine (Phila
Joint Section Meeting on Disorders of the Spine and Pa 1976). 2002;27:2674–83.
Peripheral Nerves, March 2004. J Neurosurg Spine. 40. Chou Y-C, Chen D-C, Hsieh WA, et al. Efficacy
2004;1:287–98. of anterior cervical fusion: comparison of tita-
27. Altaf F, Weber M, Dea N, et al. Evidence-based
nium cages, polyetheretherketone (PEEK) cages
review and survey of expert opinion of reconstruction and autogenous bone grafts. J Clin Neurosci.
of metastatic spine tumors. Spine (Phila Pa 1976). 2008;15:1240–5.
2016;41:S254–61. 41. Wolfe SA, Kawamoto HK. Taking the iliac-bone

28. Ryken TC, Heary RF, Matz PG, et al. Techniques graft. J Bone Joint Surg Am. 1978;60:411.
for cervical interbody grafting. J Neurosurg Spine. 42.
Yoganandan N, Larson SJ, Pintar F, et al.
2009;11:203–20. Biomechanics of lumbar pedicle screw/plate fixation
29. Gerster JC, Bossy R, Dudler J. Bone non-union after in trauma. Neurosurgery. 1990;27:873–80.. discus-
osteotomy in patients treated with methotrexate. J sion 880–1
Rheumatol. 1999;26:2695–7. 43.
Krag MH, Weaver DL, Beynnon BD, et al.
30. Gal TJ, Munoz-Antonia T, Muro-Cacho CA, et al. Morphometry of the thoracic and lumbar spine related
Radiation effects on osteoblasts in vitro: a potential to transpedicular screw placement for surgical spinal
role in osteoradionecrosis. Arch Otolaryngol Head fixation. Spine (Phila Pa 1976). 1988;13:27–32.
Neck Surg. 2000;126:1124–8. 44. Santoni BG, Hynes RA, McGilvray KC, et al. Cortical
31. Inage K, Ohtori S, Koshi T, et al. One, two-, and bone trajectory for lumbar pedicle screws. Spine J.
three-level instrumented posterolateral fusion of the 2009;9:366–73.
lumbar spine with a local bone graft: a prospective 45. Calvert GC, Lawrence BD, Abtahi AM, et al. Cortical
study with a 2-year follow-up. Spine (Phila Pa 1976). screws used to rescue failed lumbar pedicle screw
2011;36:1392–6. construct: a biomechanical analysis. J Neurosurg
32. Ito Z, Imagama S, Kanemura T, et al. Bone union rate Spine. 2015;22:166–72.
with autologous iliac bone versus local bone graft in 46. Burval DJ, McLain RF, Milks R, et al. Primary pedi-
posterior lumbar interbody fusion (PLIF): a multi- cle screw augmentation in osteoporotic lumbar verte-
center study. Eur Spine J. 2013;22:1158–63. brae. Spine (Phila Pa 1976). 2007;32:1077–83.
33. Eleraky MA, Llanos C, Sonntag VK. Cervical corpec- 47. Elder BD, Lo S-FL, Holmes C, et al. The biomechan-
tomy: report of 185 cases and review of the literature. ics of pedicle screw augmentation with cement. Spine
J Neurosurg. 1999;90:35–41. J. 2015;15:1432–45.
34. Nirala AP, Husain M, Vatsal DK. A retrospective
48. Bassiouni H, Ntoukas V, Asgari S, et al. Foramen
study of multiple interbody grafting and long segment magnum meningiomas: clinical outcome after
strut grafting following multilevel anterior cervical microsurgical resection via a posterolateral sub-
decompression. Br J Neurosurg. 2004;18:227–32. occipital retrocondylar approach. Neurosurgery.
35. Ikenaga M, Shikata J, Tanaka C. Anterior corpec- 2006;59:1177–85.. discussion 1185–7
tomy and fusion with fibular strut grafts for mul- 49. Bejjani GK, Sekhar LN, Riedel CJ. Occipitocervical
tilevel cervical myelopathy. J Neurosurg Spine. fusion following the extreme lateral transcondylar
2005;3:79–85. approach. Surg Neurol. 2000;54:109–15; discussion
36. Jang J-W, Lee J-K, Lee J-H, et al. Effect of posterior 115–6
subsidence on cervical alignment after anterior cer- 50. Schulte K, Clark CR, Goel VK. Kinematics of the
vical corpectomy and reconstruction using titanium cervical spine following discectomy and stabilization.
mesh cages in degenerative cervical disease. J Clin Spine (Phila Pa 1976). 1989;14:1116–21.
Neurosci. 2014;21:1779–85. 51. Traynelis VC, Donaher PA, Roach RM, et al.

37. Waschke A, Kaczor S, Walter J, et al. Expandable Biomechanical comparison of anterior Caspar plate
titanium cages for anterior column cervical recon- and three-level posterior fixation techniques in a human
struction and their effect on sagittal profile: a review cadaveric model. J Neurosurg. 1993;79:96–103.
of 48 cases. Acta Neurochir. 2013;155:801–7.. dis- 52. Albert TJ, Vacarro A. Postlaminectomy kyphosis.
cussion 807 Spine (Phila Pa 1976). 1998;23:2738–45.
53. Fraser JF, Härtl R. Anterior approaches to fusion of 69. Kreshak JL, Kim DH, Lindsey DP, et al. Posterior
the cervical spine: a metaanalysis of fusion rates. J stabilization at the cervicothoracic junction: a
Neurosurg Spine. 2007;6:298–303. biomechanical study. Spine (Phila Pa 1976).
54. Andaluz N, Zuccarello M, Kuntz C. Long-term fol- 2002;27:2763–70.
low-up of cervical radiographic sagittal spinal align- 70. Prybis BG, Tortolani PJ, Hu N, et al. A compara-
ment after 1- and 2-level cervical corpectomy for the tive biomechanical analysis of spinal instability and
treatment of spondylosis of the subaxial cervical spine instrumentation of the cervicothoracic junction: an
causing radiculomyelopathy or myelopathy: a retro- in vitro human cadaveric model. J Spinal Disord Tech.
spective study. J Neurosurg Spine. 2012;16:2–7. 2007;20:233–8.
55. McAfee PC, Bohlman HH, Ducker TB, et al. One- 71. Tatsumi RL, Yoo JU, Liu Q, et al. Mechanical com-
stage anterior cervical decompression and posterior parison of posterior instrumentation constructs for
stabilization. A study of one hundred patients with spinal fixation across the cervicothoracic junction.
a minimum of two years of follow-up. J Bone Joint Spine (Phila Pa 1976). 2007;32:1072–6.
Surg Am. 1995;77:1791–800. 72. Lapsiwala S, Benzel E. Surgical management of cer-
56. Anderson PA, Henley MB, Grady MS, et al. Posterior vical myelopathy dealing with the cervical-thoracic
cervical arthrodesis with AO reconstruction plates and junction. Spine J. 2006;6:268S–73S.
bone graft. Spine (Phila Pa 1976). 1991;16:S72–9. 73.
Yamagata M, Kitahara H, Minami S, et al.
57. Mihara H, Cheng BC, David SM, et al. Biomechanical Mechanical stability of the pedicle screw fixation
comparison of posterior cervical fixation. Spine (Phila systems for the lumbar spine. Spine (Phila Pa 1976).
Pa 1976). 2001;26:1662–7. 1992;17:S51–4.
58. Omeis I, DeMattia JA, Hillard VH, et al. History of 74. Shono Y, Kaneda K, Yamamoto I. A biomechanical
instrumentation for stabilization of the subaxial cervi- analysis of Zielke, Kaneda, and Cotrel-Dubousset
cal spine. Neurosurg Focus. 2004;16:E10. instrumentations in thoracolumbar scoliosis. A calf
59. Murakami H, Jarrett C, Rhee JM, et al. Spinous process spine model. Spine (Phila Pa 1976). 1991;16:1305–11.
wiring versus lateral mass fixation for the treatment 75. An HS, Lim TH, You JW, et al. Biomechanical evalua-
of anterior cervical pseudarthrosis: a biomechanical tion of anterior thoracolumbar spinal instrumentation.
comparison. J Surg Orthop Adv. 2011;20:220–4. Spine (Phila Pa 1976). 1995;20:1979–83.
60. Lovick DS, Ryken TC, Traynelis VC, et al.
76. Turi M, Johnston CE, Richards BS. Anterior correc-
Assessment of primary and salvage lateral mass screw tion of idiopathic scoliosis using TSRH instrumenta-
insertion torque in a cadaveric model. J Spinal Disord. tion. Spine (Phila Pa 1976). 1993;18:417–22.
1997;10:431–5. 77. Saraph VJ, Krismer M, Wimmer C. Operative treat-
61. Klekamp JW, Ugbo JL, Heller JG, et al. Cervical ment of scoliosis with the Kaneda anterior spine sys-
transfacet versus lateral mass screws: a biomechani- tem. Spine (Phila Pa 1976). 2005;30:1616–20.
cal comparison. J Spinal Disord. 2000;13:515–8. 78. Shimamoto N, Kotani Y, Shono Y, et al. Static and
62. Goel A, Shah A. Facetal distraction as treatment for dynamic analysis of five anterior instrumentation
single- and multilevel cervical spondylotic radicu- systems for thoracolumbar scoliosis. Spine (Phila Pa
lopathy and myelopathy: a preliminary report. J 1976). 2003;28:1678–85.
Neurosurg Spine. 2011;14:689–96. 79. Reddy CG, Magnetta M, Dahdaleh NS, et al. An

63. Tan LA, Gerard CS, Anderson PA, et al. Effect
in vitro biomechanical comparison of single-rod, dual-
of machined interfacet allograft spacers on cervi- rod, and dual-rod with transverse connector in ante-
cal foraminal height and area. J Neurosurg Spine. rior thoracolumbar instrumentation. Neurosurgery.
2014;20:178–82. 2012;70:1017–23.. discussion 1023
64. Yoshihara H, Passias PG, Errico TJ. Screw-related 80. Kostuik JP, Errico TJ, Gleason TF, et al. Spinal stabi-
complications in the subaxial cervical spine with lization of vertebral column tumors. Spine (Phila Pa
the use of lateral mass versus cervical pedicle 1976). 1988;13:250–6.
screws: a systematic review. J Neurosurg Spine. 81. Manabe S, Tateishi A, Abe M, et al. Surgical treat-
2013;19:614–23. ment of metastatic tumors of the spine. Spine (Phila
65. Steinmetz MP, Miller J, Warbel A, et al. Regional Pa 1976). 1989;14:41–7.
instability following cervicothoracic junction surgery. 82. Mahar A, Kim C, Wedemeyer M, et al. Short-segment
J Neurosurg Spine. 2006;4:278–84. fixation of lumbar burst fractures using pedicle fixa-
66. Pal GP, Sherk HH. The vertical stability of the cervi- tion at the level of the fracture. Spine (Phila Pa 1976).
cal spine. Spine (Phila Pa 1976). 1988;13:447–9. 2007;32:1503–7.
67. Chapman JR, Anderson PA, Pepin C, et al. Posterior 83. Guven O, Kocaoglu B, Bezer M, et al. The use of
instrumentation of the unstable cervicothoracic spine. screw at the fracture level in the treatment of tho-
J Neurosurg. 1996;84:552–8. racolumbar burst fractures. J Spinal Disord Tech.
68. Albert TJ, Klein GR, Joffe D, et al. Use of cervico- 2009;22:417–21.
thoracic junction pedicle screws for reconstruction 84. McLain RF. The biomechanics of long versus short
of complex cervical spine pathology. Spine (Phila Pa fixation for thoracolumbar spine fractures. Spine
1976). 1998;23:1596–9. (Phila Pa 1976). 2006;31:S70–9.
85. Viljoen SV, DeVries Watson NA, Grosland NM, et al. lumbar constructs. A biomechanical analysis. Spine J.
Biomechanical analysis of anterior versus posterior 2001;1:373–9.
instrumentation following a thoracolumbar corpec- 89. Wood KB, Wentorf FA, Ogilvie JW, et al. Torsional
tomy. J Neurosurg Spine. 2014;21:577–81. rigidity of scoliosis constructs. Spine (Phila Pa 1976).
86. Alegre GM, Gupta MC, Bay BK, et al. S1 screw bend- 2000;25:1893–8.
ing moment with posterior spinal instrumentation 90. Ruland CM, McAfee PC, Warden KE, et al.

across the lumbosacral junction after unilateral iliac Triangulation of pedicular instrumentation. A
crest harvest. Spine (Phila Pa 1976). 2001;26:1950–5. biomechanical analysis. Spine (Phila Pa 1976).
87. Kuklo TR, Bridwell KH, Lewis SJ, et al. Minimum 1991;16:S270–6.
2-year analysis of sacropelvic fixation and L5-S1 91. Fisher CG, et al. A novel classification system for
fusion using S1 and iliac screws. Spine (Phila Pa spinal instability in neoplastic disease: an evidence-
1976). 2001;26:1976–83. based approach and expert consensus from the Spine
88. Brodke DS, Bachus KN, Mohr RA, et al. Segmental Oncology Study Group. Spine (Phila Pa 1976).
pedicle screw fixation or cross-links in multilevel 2010;35(22):E1221–9.
Separation Surgery for Spinal
Metastases
41
Robert J. Rothrock, Ori Barzilai, Ilya Laufer,
and Mark H. Bilsky
Introduction postoperative SBRT in order to minimize treat-

ment morbidity and to maximize local control
Advances in systemic treatment for metastatic and safety [4]. Hybrid therapy describes the com-
cancer have led to longer patient survival, para- bination of separation surgery promptly followed
doxically increasing lifetime risk for develop- by SBRT to treat remaining noncompressive
ment of distant metastases [1]. As the incidence osseous and paraspinal disease. The surgical
of spinal disease increases, so does the impor- strategy of separation surgery provides circum-
tance of effective management of spinal metasta- ferential decompression of the spinal cord and
ses and surgical strategies that minimize the need stabilization of the spinal column, without the
for interruption of systemic therapy [2]. Surgical goal of gross total resection of osseous and para-
care for spinal metastases has evolved over time spinal tumor. Local tumor control is dependent
in concert with other oncologic improvements. on the response to SBRT rather than cytoreduc-
Historically, in the setting of lacking alternative tion. The surgery, in turn, provides optimal con-
strategies for local control, surgery for metastatic ditions for SBRT, allowing the safe delivery of
epidural spinal cord compression (MESCC) cen- tumoricidal radiation dosing. Although the surgi-
tered on gross total tumor resection, coupled with cal management of spinal metastatic disease is
mechanical reconstruction. This often involved palliative by nature, separation surgery allows for
extended circumferential approaches with multi- effective treatment of spinal metastatic tumor
level vertebrectomy, prolonged surgical duration, with demonstrated improvement in local recur-
risk of significant perioperative complications, rence, increased patient-reported quality-of-life
and, importantly, prolonged recovery times for indicators, and preservation or restoration of
patients [3]. patient mobility [5–8].
Stereotactic body radiotherapy (SBRT) pro-
vides reliable and durable local tumor control,
which has significantly changed the paradigm for Indications for Surgery
management of metastatic spinal disease. Hybrid
therapy combines and optimizes surgical and The NOMS decision framework provides a com-
prehensive assessment of four sentinel decision
R. J. Rothrock · O. Barzilai · I. Laufer points: neurologic, oncologic, mechanical stabil-
M. H. Bilsky (*) ity, and systemic disease [9]. This framework
Department of Neurosurgery, Memorial Sloan standardizes the assessment of patients with
e-mail: bilskym@MSKCC.ORG
metastatic spine tumors and allows for the

https://doi.org/10.1007/978-3-030-42958-4_41
560 R. J. Rothrock et al.
incorporation of evidence-based medicine, a proposed intervention, taking into account the

which promotes the rational use of new radia- overall expected survival and the ability of a
tion, surgical, interventional radiology, and sys- patient to tolerate spine-specific treatment.
temic therapies. The NOMS decision framework allows for
The neurologic assessment evaluates both clini- flexible, multifactorial decision-making to help
cal and radiologic parameters, including the pres- define the appropriate balanced treatment plan
ence of myelopathy, functional radiculopathy, and for a given metastatic cancer patient. Using this
the radiographic degree of epidural tumor exten- framework, indications for separation surgery
sion, and spinal cord compression. A validated include patients with radioresistant tumor histol-
magnetic resonance–based epdirual spinal cord ogy with high-grade ESCC, who can tolerate sur-
compression (ESCC) scoring system, known as the gery from a medical and systemic perspective
Bilsky score, is used to define the extent of epidural [12]. There is also a role for stabilization surgery
spinal cord compression [10]. Patients with Bilsky in patients with mechanical instability without
grades 0 and 1 have tumors that either are confined overt spinal cord compression, which is dis-
to bone or exhibit epidural extension without cussed elsewhere in this textbook [13].
displacement or compression of the spinal cord.
Patients with Bilsky grades 2 and 3 have tumors
that displace, deform, or frankly compress the Rationale for Approach
spinal cord. Patients with Bilsky grades 0 and 1
have low-grade MESCC, and patients with Bilsky The rationale for hybrid therapy is based on
grades 2 and 3 have high-grade MESCC. numerous studies demonstrating both a benefit to
The oncologic consideration is based on the surgical decompression of high-grade ESCC and
expected local tumoral response, principally to durable local tumor control with prompt postop-
conventional external beam radiation therapy erative SBRT. Patchell et al.’s randomized con-
(cEBRT) and systemic therapy. cEBRT provides trolled trial comparing patients receiving
local control for radiosensitive tumors such as fractionated conventional external beam radia-
lymphoma, prostate, and breast adenocarcinoma. tion therapy (cEBRT) alone versus surgical
Remaining solid tumor metastases generally decompression followed by cEBRT for spinal
exhibit radioresistance when treated with cord compression due to metastatic cancer was
cEBRT. Thus, tumors exhibit a range of radiore- stopped early when significantly more patients
sistance with primary tumor histology as the from the surgical arm of the study were able to
most commonly used predictor of sensitivity. walk after treatment (84% vs 57%, odds ratio 6.2
However, SBRT, through the delivery of high- [95% CI 2.0–19.8] p = 0.001) [14]. The pre-
dose conformal radiotherapy, can overcome scribed dose of radiation was standardized in the
radioresistance, providing durable local control trial to 30 Gy in 10 fractions in each arm.
regardless of tumor histology and volume. The Although it was not the primary outcome, there
neurologic and oncologic assessments are com- was an overall survival advantage in the surgical
bined to determine the optimal radiation strategy arm of the study. Patchell’s trial is credited with
to achieve tumor control and/or the need for a establishing surgery as the standard of care for
surgical decompression of the spinal cord. single-level MESCC in a symptomatic patient
Mechanical instability is a separate consider- with solid tumor malignancy who has an accept-
ation and is generally defined according to the able life expectancy based on extent of disease,
Spinal Instability Neoplastic Score (SINS) crite- systemic treatment options, and medical comor-
ria [11]. Patients with mechanical instability typi- bidities [15].
cally require stabilization with bone cement or With SBRT providing effective local control
spinal instrumentation. The fourth consideration regardless of tumor volume, gross total tumor
is the extent of systemic disease and medical excision is no longer necessary to optimize tumor
comorbidities that affect the risk–benefit ratio of control. In the absence of spinal cord c ompression,
41 Separation Surgery for Spinal Metastases 561
SBRT can be used as definitive therapy to deliver undergoing aggressive resection followed by
an ablative dose to the entire tumor volume. conventional external beam radiation with 1-year
SBRT doses commonly used are 18–24 Gy in local failures up to 70% [19].
single fraction or 24–30 Gy in three fractions. In Separation surgery requires spinal instrumen-
a recent series reporting single fraction outcomes, tation to treat existing spinal instability and pre-
a median dose of 22.4 Gy resulted in 98% 4-year vent iatrogenic instability. Patients undergoing
local control rates even for radioresistant tumors separation surgery require spinal instrumentation
such as renal cell carcinoma and sarcoma metas- and fixation, since anterior and middle column
tases [15, 16]. integrity is usually compromised by tumoral
The ability to deliver an ablative radiation invasion, and decompression requires removal of
dose to the entire tumor volume, particularly at the lamina and pedicle/joint complex (posterior
the epidural margin, is limited in the setting of column) [11]. In addition, the need for multiple
high-grade spinal cord compression. The spinal levels of decompression and adjacent level
cord is the most critical organ at risk (OAR), involvement are not uncommon, requiring larger
which limits radiation dose to the dural margin constructs. Complicating matters further, poten-
without the risk of iatrogenic spinal cord injury. tial bony fusion, is severely compromised in
However, when the tumor is separated from the oncologic patients due to poor bone quality, radi-
spinal cord by 2–3 mm, the entire tumor volume ation and chemotherapy effects, and overall
can be treated with an effective SBRT dose with- expected survival [20]. Based on these features
out exceeding the accepted spinal cord con- and the risk of tumor extension to adjacent levels,
straints. Due to the improved local tumor control posterior spinal instrumentation has usually
observed with SBRT, the oncologic goals of extended at least two levels above and below the
achieving local tumor control have transitioned surgical index level(s), and sometimes greater if
from gross total excision to simple separation crossing a junctional area or in the setting of
surgery. The goal of separation surgery is circum- markedly poor bone quality [11].
ferential excision of epidural tumor to reconsti-
tute the thecal sac creating a 2-mm margin for the
safe delivery of an ablative radiation dose. Surgical Considerations
Most centers use a 1.5- to 2-mm margin to the
thecal sac as a planning OAR volume. The cumu- There are many involved preoperative consider-
lative acceptable point exposure dose to the spi- ations for patients undergoing separation sur-
nal cord is considered 10 Gy to 10% of the gery. By definition, most cancer patients have an
epidural volume or a cord Dmax of 14 Gy [17]. American Society of Anesthesiologists (ASA)
Al-Omair et al. demonstrated that the degree of Score of IV or greater, placing them at higher
resection of epidural disease (surgical downgrad- perioperative risk for mortality as validated by
ing of Bilsky grade for MESCC) has a significant numerous clinical studies [21–23]. Given this
impact on long-term local tumor control in the increased risk, optimization for surgery demands
context of hybrid therapy [18]. Thus, thorough an interdisciplinary discussion, generally
separation surgery not only directly addresses involving the patient’s oncologist in order to
spinal cord compression but also allows a safe determine the availability of further systemic
corridor to effectively treat remaining osseous therapy, provide perioperative risk stratification,
and paraspinal tumor. In our previous analysis of and confer with the anesthesia team. Patients
186 patients undergoing hybrid therapy, the with poor pulmonary function and significant
cumulative incidence of local failure was 16.4% liver tumor burden generally represent the
for 1 year after SBRT [5]. In patients receiving a highest-risk patient populations. Furthermore,
24-Gy single fraction or 24–30 Gy in three frac- extensive tumor infiltration of the bone mar-
tions, the 1-year local failure rate was less than row or the effects of chemotherapy may lead
10%. This is far superior to historical controls to chronic thrombocytopenia. Finally, cancer
predisposes patients to the d evelopment of deep it is crucial to avoid transmitting pressure to the
venous thrombosis, with 9.5% of patients under- spinal cord during decompression. Our practice
going spinal surgery in this setting having pre- is to drill the posterior elements using a high-
operative DVT and 24% of nonambulatory speed 3-mm matchstick burr. The laminae are
patients having a DVT [24]. egg-shelled, and the remaining bone and liga-
Several metastatic tumor types, such as renal mentum flavum are resected away from the spi-
cell carcinoma and solitary fibrous tumor, have a nal cord. Depending on tumoral location, a
robust vascular supply, which may lead to signifi- surgical corridor to the ventral epidural space is
cant intraoperative blood loss. Therefore, preop- created via unilateral or bilateral removal of the
erative embolization is used to minimize the risk facet joint(s) and pedicle(s) using the drill.
of severe blood loss in the setting of patients with Normal anatomical planes above and below the
vascular tumors. Furthermore, patients with prior tumor level are defined prior to tumor excision in
radiation to the surgical field or on active cyto- order to facilitate safe separation of the tumor
toxic chemotherapy at the time of surgery gener- from the dura. We use a combination of tenotomy
ally benefit from involvement of plastic surgeons scissors, Penfield dissectors, forceps, and pitu-
in the surgical closure or reconstruction. itary rongeurs to resect the epidural tumor and to
maintain a safe epidural plane.
To ensure circumferential decompression is
Surgical Approach achieved, the ventral epidural component of the
tumor must be visualized and dissected away
Patients are sedated under general anesthesia, from the dura. Delineation of the posterior longi-
and an arterial line and Foley catheter are placed. tudinal ligament (PLL) is crucial in adequately
Intraoperative neurophysiological monitoring visualizing the ventral epidural tumor that is gen-
(IONM) is routinely used, including EMGs, erally deep to the PLL. The PLL is sectioned
SSEPs, and MEPs. Following prone positioning using tenotomy scissors, providing exposure of
on a four-post radiolucent table, fluoroscopic the ventral epidural tumor and the vertebral body,
localization is used to plan a midline linear skin and a Woodson dissector is used to clear the ven-
incision. Midline subperiosteal exposure of the tral dural margin and to decompress the spinal
posterior spinal elements is performed using cord. A partial vertebrectomy is performed to
monopolar cautery and Cobb periosteal elevators maintain a safe corridor, and usually approxi-
[4]. Our practice is to place instrumentation prior mately 20% removal of the involved vertebral
to beginning decompression and resection of the body is sufficient. Once a ventral cavity has been
epidural tumor. Pedicle or lateral mass screws are created, a Woodson dissector can be used to fur-
placed via anatomical freehand technique or by ther separate the tumor from the dura and to
various navigational guidance systems [25]. As ensure adequate ventral epidural decompression.
described above, due to both the need for bilat- If a large portion of the vertebral body is
eral facetectomies at the index level and the removed or compromised, anterior column sup-
inherently compromised bone quality in these port can be achieved by inserting poly-methyl-
patients, it is our practice to incorporate at least methacrylate (PMMA) into the anterior vertebral
two levels above and below the tumoral level into cavity as previously described [13]. In cases of
the final surgical construct when open surgical extended vertebral body removal, an expandable
approaches are used. Rods are contoured to or stackable cage may be used for anterior col-
approximate the anatomical kyphosis or lordosis umn reconstruction. If a cage is used, then either
depending on the spinal segment, and screw caps a polyetheretherketone (PEEK) or a Harms tita-
are tightened to lock the construct. nium mesh cage is preferred to minimize radio-
Next, attention is turned to posterolateral graphic magnetic resonance imaging (MRI)
decompression of the spinal canal. In the setting artifact. Importantly, aggressive or gross-total
of high-grade epidural spinal cord compression, resection of the vertebral body or paraspinal
a b d
Fig. 41.1 (a, b) Seventy-four-year-old woman who pre- T1 to T7. (c) Postoperative CT myelography was obtained
sented with high-grade malignant epidural spinal cord on postoperative day 2 with demonstration of circumfer-
compression (Bilsky grade 3) at T3–T4 from non-small- ential decompression. (d) Postoperative X-ray demon-
cell lung adenocarcinoma with associated back pain and strates the surgical construct. She was treated with 27 Gy
ataxic gait. She underwent separation surgery with in three-fraction SBRT beginning approximately 2 weeks
decompression from T3 to T4 and instrumentation from after separation surgery
tumor is not required since postoperative SBRT ligament (PLL) via the epidural ligaments of
will effectively treat these tumor components. Hoffman, usually requiring resection of the PLL
Meticulous hemostasis is achieved, and the to ensure complete decompression [26]. Because
wound is irrigated copiously with antibiotic irri- ventral decompression can be difficult to directly
gation. The facet joints and transverse processes visualize, intraoperative ultrasound can be a use-
are decorticated, and autologous bone graft is ful confirmatory adjunct, allowing visualization
used to augment bony fusion. Vancomycin pow- of ventral cerebrospinal fluid (CSF) pulsatility
der is left in the operative bed for prophylaxis. At and dural planes [27].
least one subfascial drain is left in place to full
suction, and the incision is closed in multiple
anatomical layers (Figs. 41.1 and 41.2). Vertebroplasty
In cases where there is a related pathologic frac-

Intraoperative Adjuvants ture or compression deformity, intraoperative
vertebroplasty can be a useful adjunct in the
Ultrasound Guidance treatment of mechanical pain [28]. Although vio-
lation of the posterior wall of the vertebral body
The primary goal of separation surgery is obtain- by tumor has been cited as a relative contraindi-
ing adequate ventral decompression. The anterior cation to vertebroplasty, there is evidence that it
dura is connected to the posterior longitudinal can still be safely performed in this setting [29].
a b d e
Fig. 41.2 (a, b) Fifty-year-old man who presented with on postoperative day 2 with demonstration of circumfer-
severe upper back pain from a pathologic fracture and ential decompression. (d) Postoperative X-ray demon-
high-grade malignant epidural spinal cord compression strates the surgical construct. He was treated with 27 Gy
(Bilsky grade 3) at T3–T5 from non-small-cell lung ade- in three-fraction SBRT beginning approximately 2 weeks
nocarcinoma. He underwent separation surgery with after separation surgery. (e) Three-month follow-up tho-
decompression from T3 to T5 and instrumentation from racic MRI demonstrates durable local tumor control
T2 to T6. (c) Postoperative CT myelography was obtained
Fenestrated Screws/Cement tial compression of the thecal sac. Often in the

Augmentation case of previous radiation, the spinal cord has
already been exposed to substantial radiation
In patients with widespread bony metastases, dose, and further exposure might place it at risk
bone quality and screw purchase can be severely of toxicity [17]. The predominant pattern of dis-
affected. In these cases, cement augmentation of ease recurrence after postoperative SBRT is
the screws or the anterior column can aid in bony within the epidural space. In their series,
purchase and decrease the risk of hardware fail- Al-Omair et al. demonstrated that when treat-
ure [13, 30]. Cement injection through fenes- ment failure occurred, it was exclusively in the
trated screws provides a facile way to epidural space in two-thirds of patients [18].
cement-augment the osseous screw purchase in In the setting of circumferential tumor infil-
cancer patients [31]. Intraoperative kyphoplasty tration and previously irradiated targets, one
can augment the anterior column structure even solution is to deliver a therapeutic radiation
in patients with compromised posterior vertebral dose to the dural margin with single-dose intra-
cortex. operative brachytherapy [32]. The P32 brachy-
therapy plaque delivers a high radiation dose
(median 27 Gy to 1 mm) with a steep dose-fall
P32 Brachytherapy off (3 mm) making it an ideal dural radiation
plaque. The plaque is brought into the operat-
One of the greatest challenges for hybrid therapy ing room, laid directly into the targeted epi-
is in the case of recurrent tumor that was previ- dural space, and removed after therapeutic
ously irradiated and presents with circumferen- dose has been delivered. In several series, P32
brachytherapy has been shown to be a useful radiosurgery and subsequent postoperative

adjunct to surgical intervention following epi- guidelines provide the basis for treatment plan-
dural decompression [33]. ning [15, 35, 36]. Target volumes are defined
according to the definitions set by the International
Commission on Radiation Units and
Postoperative Management Measurements Report 50 [35, 37]. Gross tumor
volume (GTV) describes observed disease at sur-
The second phase of hybrid therapy is the postop- gery or gross tumor seen on imaging. Clinical
erative SBRT delivery. SBRT is defined as “the target volume (CTV) is the region of potential
precise delivery of highly conformal and microscopic spread of tumor cells that includes
image-
guided hypo-fractionated external beam the GTV and represents the total desired treat-
radiotherapy, delivered in a single or few ment volume. The planning target volume (PTV)
fraction(s), to an extra-cranial body target with is a geometric construct that encompasses the
doses at least biologically equivalent to a radical CTV and adds an additional margin of tissue to
course when given over a conventionally fraction- ensure that the CTV receives the intended dose.
ated (1.8–3.0 Gy/ fraction) schedule” [16]. This margin takes into account factors that are
Circumferential decompression achieved via sepa- difficult to control, like patient positioning,
ration surgery allows for tumoricidal SBRT doses motion during treatment, physical errors of the
to be delivered to the entire tumor volume within treatment machinery, and other random errors
the constraints of spinal cord tolerance. Planning that can occur. In modern stereotactic spine
for SBRT typically begins while the patient is still radiosurgery, a typical PTV expansion on the
in the hospital and recovering from surgery. CTV is 2 mm.
Treatment planning is ultimately a compro-
mise between the prescribed dose and the allow-
Simulation able dose to surrounding normal structures
(OARs). In general for spine radiosurgery, dose
In our practice, simulation is performed on post- uniformity within the target volume is sacrificed
operative days 2 and 3, with the goal of SBRT for steep dose gradients immediately outside the
treatment approximately 2 weeks following sepa- target volume to allow maximal sparing of OARs
ration surgery. Because MRI-related artifact from such as the spinal cord or esophagus. Radiation
hardware can limit radiosurgical treatment plan- “hot spots” over 130% of the prescribed dose are
ning, we utilize CT myelography to better visual- allowed. An ideal treatment plan would be able to
ize the neural elements, surgical construct, and cover at least 90% of the PTV with the prescribed
organs at risk (OARs) [34, 35]. Patients are dose, but better than 80% coverage of the PTV
immobilized during simulation in a reproducible with the prescribed dose would be still consid-
manner using a patient-specific positioning ered acceptable. Due to the complexity of
frame. Preoperative images (usually MRI with decision-making in the setting of postoperative
and without contrast) are used to delineate to the SBRT, we utilize a multidisciplinary conference
preoperative tumor volume, and this volume is between radiation oncology and neurosurgery for
outlined on the postoperative simulation imaging treatment planning.
(CT myelogram) for accurate delineation of
tumor target, OARs, and treatment planning.
Complications
SBRT Immediate
The Spine Radiosurgery Consensus Consortium Immediate complications following separation

contouring guidelines for spinal stereotactic surgery are similar to those for all instrumented
spinal surgery. In patients with highly vascular Conclusion

metastatic disease, such as renal cell carcinoma or
hepatocellular carcinoma, there can be high rates Hybrid therapy for spinal metastatic disease –
of intraoperative blood loss with resultant postop- concomitant separation surgery and SBRT – is an
erative anemia [38]. Intraoperative durotomy may effective, tolerable, and reproducible treatment.
be repaired using muscle patching and fibrin glue. Separation surgery provides rapid decompres-
In cases of postoperative cerebrospinal fluid (CSF) sion, stabilization, and continuation of treatment,
leaks through the incision and pseudomeningocele generally without a prolonged recovery period.
formation, placement of a lumbar drain usually Assuring adequate circumferential epidural
results in resolution of the leak. In cases of persis- decompression is crucial and allows for optimal
tent leaks, plastic surgeons can help to provide SBRT dosing and durable local tumor control.
extended muscle coverage of the dural defect and
soft-tissue reconstruction. Cancer patients are at
increased risk of poor wound healing due to poor References
nutritional status, utilization of systemic therapy
that impairs wound healing, and extensive use of 1. Mokdad AH, Dwyer-Lindgren L, Fitzmaurice C,
radiation [39]. As discussed above, this risk can be Stubbs RW, Bertozzi-Villa A, Morozoff C, et al.
Trends and patterns of disparities in cancer mor-
mitigated by preoperative identification of patients tality among US counties, 1980–2014. JAMA.
who might benefit from plastic surgery-assisted 2017;317(4):388–406.
closure. Lau et al. examined 106 adult patients 2. Klimo P Jr, Thompson CJ, Kestle JRW, Schmidt
undergoing surgery for spinal metastatic disease MH. A meta-analysis of surgery versus conventional
radiotherapy for the treatment of metastatic spinal
and found that age greater than 65 years and the epidural disease. Neuro-Oncology. 2005;7(1):64–76.
presence of contiguous disease in three or more 3. Tomita K, Kawahara N, Murakami H, Demura
spinal levels were independent predictors of com- S. Total en bloc spondylectomy for spinal tumors:
plications from surgery [40]. improvement of the technique and its associated basic
background. J Orthop Sci. 2006;11(1):3–12.
4. Barzilai O, Laufer I, Robin A, Xu R, Yamada Y,
Bilsky MH. Hybrid therapy for metastatic epidural
Delayed spinal cord compression: technique for separation
surgery and spine radiosurgery. Oper Neurosurg
(Hagerstown) [Internet]. 2018 Jun 8. Available from:
As patient survival improves with advances in sys- https://doi.org/10.1093/ons/opy137.
temic cancer therapies, a new set of delayed post- 5. Laufer I, Iorgulescu JB, Chapman T, Lis E, Shi W,
operative complications have emerged in the Zhang Z, et al. Local disease control for spinal metas-
setting of separation surgery. The highly effective tases following “separation surgery” and adjuvant
hypofractionated or high-dose single-fraction stereo-
tumoricidal doses of SBRT can result in profound tactic radiosurgery: outcome analysis in 186 patients.
osteonecrosis of the vertebral body, resulting in J Neurosurg Spine. 2013;18(3):207–14.
delayed fracture progression and hardware failure 6. Bate BG, Khan NR, Kimball BY, Gabrick K, Weaver
[41]. Delayed vertebral body fracture can be J. Stereotactic radiosurgery for spinal metastases with
or without separation surgery. J Neurosurg Spine.
treated with salvage vertebroplasty/kyphoplasty to 2015;22(4):409–15.
avoid larger revision surgery in this high surgical 7. Barzilai O, Amato M-K, McLaughlin L, Reiner AS,
risk population [42]. In addition, de novo disease Ogilvie SQ, Lis E, et al. Hybrid surgery-radiosurgery
in adjacent segments can be challenging to treat, therapy for metastatic epidural spinal cord compres-
sion: a prospective evaluation using patient-reported
especially if there is overlap with previous radia- outcomes. Neurooncol Pract. 2018;5(2):104–13.
tion treatment (i.e., dose constraints). Given the 8. Barzilai O, McLaughlin L, Amato M-K, Reiner AS,
low rates of solid arthrodesis, pseudoarthrosis and Ogilvie SQ, Lis E, et al. Predictors of quality of life
delayed instability can result in debilitating pain improvement after surgery for metastatic tumors
of the spine: prospective cohort study. Spine J.
[43]. Esophageal perforation is a rare but known 2018;18(7):1109–15.
complication of SBRT to the cervical and upper- 9. Laufer I, Rubin DG, Lis E, Cox BW, Stubblefield MD,
thoracic spine that can be severely morbid [44]. Yamada Y, et al. The NOMS framework: approach to
the treatment of spinal metastatic tumors. Oncologist. 22. Hopkins TJ, Raghunathan K, Barbeito A, Cooter M,
2013;18(6):744–51. Stafford-Smith M, Schroeder R, et al. Associations
10. Bilsky MH, Laufer I, Fourney DR, Groff M, Schmidt between ASA Physical Status and postoperative mor-
MH, Varga PP, et al. Reliability analysis of the epi- tality at 48 h: a contemporary dataset analysis com-
dural spinal cord compression scale. J Neurosurg pared to a historical cohort. Perioperative Medicine
Spine. 2010;13(3):324–8. [Internet]. 2016;5(1). Available from: https://doi.
11. Fourney DR, Frangou EM, Ryken TC, Dipaola CP, org/10.1186/s13741-016-0054-z.
Shaffrey CI, Berven SH, et al. Spinal instability neo- 23. Hackett NJ, De Oliveira GS, Jain UK, Kim JYS. ASA
plastic score: an analysis of reliability and validity class is a reliable independent predictor of medical
from the spine oncology study group. J Clin Oncol. complications and mortality following surgery. Int J
2011;29(22):3072–7. Surg. 2015;18:184–90.
12. Barzilai O, Laufer I, Yamada Y, Higginson DS,
24. Zacharia BE, Kahn S, Bander ED, Cederquist GY,
Schmitt AM, Lis E, et al. Integrating evidence-based Cope WP, McLaughlin L, et al. Incidence and risk fac-
medicine for treatment of spinal metastases into a tors for preoperative deep venous thrombosis in 314
decision framework: neurologic, oncologic, mechani- consecutive patients undergoing surgery for spinal
cals stability, and systemic disease. J Clin Oncol. metastasis. J Neurosurg Spine. 2017;27(2):189–97.
2017;35(21):2419–27. 25. Costa F, Dorelli G, Ortolina A, Cardia A, Attuati

13. Moussazadeh N, Rubin DG, McLaughlin L, Lis
L, Tomei M, et al. Computed tomography-based
E, Bilsky MH, Laufer I. Short-segment percutane- image-guided system in spinal surgery: state of the
ous pedicle screw fixation with cement augmenta- art through 10 years of experience. Neurosurgery.
tion for tumor-induced spinal instability. Spine J. 2015;11(Suppl 2):59–67; discussion 67–8
2015;15(7):1609–17. 26. Tardieu GG, Fisahn C, Loukas M, Moisi M, Chapman
14. Patchell RA, Tibbs PA, Regine WF, Payne R, Saris J, Oskouian RJ, et al. The epidural ligaments (of
S, Kryscio RJ, et al. Direct decompressive surgical Hofmann): a comprehensive review of the literature.
resection in the treatment of spinal cord compres- Cureus. 2016;8(9):e779.
sion caused by metastatic cancer: a randomised trial. 27.
Vasudeva VS, Abd-El-Barr M, Pompeu YA,
Lancet. 2005;366(9486):643–8. Karhade A, Groff MW, Lu Y. Use of intraoperative
15.
Redmond KJ, Lo SS, Fisher C, Sahgal ultrasound during spinal surgery. Global Spine J.
A. Postoperative Stereotactic Body Radiation 2017;7(7):648–56.
Therapy (SBRT) for spine metastases: a critical 28. Fourney DR, Schomer DF, Nader R, Chlan-Fourney
review to guide practice. Int J Radiat Oncol Biol J, Suki D, Ahrar K, et al. Percutaneous vertebroplasty
Phys. 2016;95(5):1414–28. and kyphoplasty for painful vertebral body fractures
16. Sahgal A, Roberge D, Schellenberg D, Purdie TG, in cancer patients. J Neurosurg Spine. 2003;98:21–30.
Swaminath A, Pantarotto J, et al. The Canadian 29. Alvarez L, Pérez-Higueras A, Quiñones D, Calvo E,
Association of Radiation Oncology scope of practice Rossi RE. Vertebroplasty in the treatment of vertebral
guidelines for lung, liver and spine stereotactic body tumors: postprocedural outcome and quality of life.
radiotherapy. Clin Oncol. 2012;24(9):629–39. Eur Spine J. 2003;12(4):356–60.
17. Yamada Y, Bilsky MH, Lovelock DM, Venkatraman 30.
Frankel BM, Jones T, Wang C. Segmental
ES, Toner S, Johnson J, et al. High-dose, single- polymethylmethacrylate- augmented pedicle screw
fraction image-guided intensity-modulated radiother- fixation in patients with bone softening caused by
apy for metastatic spinal lesions. Int J Radiat Oncol osteoporosis and metastatic tumor involvement: a
Biol Phys. 2008;71(2):484–90. clinical evaluation. Neurosurgery. 2007;61(3):531–7;
18. Al-Omair A, Masucci L, Masson-Cote L, Campbell discussion 537–8
M, Atenafu EG, Parent A, et al. Surgical resection 31. Barzilai O, McLaughlin L, Lis E, Reiner AS, Bilsky
of epidural disease improves local control following MH, Laufer I. Utility of cement augmentation via per-
postoperative spine stereotactic body radiotherapy. cutaneous fenestrated pedicle screws for stabilization
Neuro-Oncology. 2013;15(10):1413–9. of cancer related spinal instability. Oper Neurosurg
19. Klekamp J, Samii H. Surgical results for spinal metas- (Hagerstown) [Internet]. 2018 Dec 3. Available from:
tases. Acta Neurochir. 1998;140(9):957–67. https://doi.org/10.1093/ons/opy186.
20. Rades D, Fehlauer F, Schulte R, Veninga T, Stalpers 32. Folkert MR, Bilsky MH, Cohen GN, Zaider M,

LJA, Basic H, et al. Prognostic factors for local Dauer LT, Cox BW, et al. Intraoperative 32P high-
control and survival after radiotherapy of meta- dose rate brachytherapy of the dura for recurrent pri-
static spinal cord compression. J Clin Oncol. mary and metastatic intracranial and spinal tumors.
2006;24(21):3388–93. Neurosurgery. 2012;71(5):1003–10; discussion
21. Lakomkin N, Zuckerman SL, Stannard B, Montejo 1010–1
J, Sussman ES, Virojanapa J, et al. Preoperative risk 33. Folkert MR, Bilsky MH, Cohen GN, Voros L, Oh JH,
stratification in spine tumor surgery – a comparison of Zaider M, et al. Local recurrence outcomes using the
the modified Charlson Index, Frailty Index, and ASA 32P intraoperative brachytherapy plaque in the man-
score. Spine [Internet]. 2018 Dec 19. Available from: agement of malignant lesions of the spine involving
https://doi.org/10.1097/BRS.0000000000002970. the dura. Brachytherapy. 2015;14(2):202–8.
34. Sahgal A, Bilsky M, Chang EL, Ma L, Yamada Y, 39. Payne WG, Naidu DK, Wheeler CK, Barkoe D,

Rhines LD, et al. Stereotactic body radiotherapy for Mentis M, Salas RE, et al. Wound healing in patients
spinal metastases: current status, with a focus on its with cancer. Eplasty. 2008;8:e9.
application in the postoperative patient. J Neurosurg 40. Lau D, Leach MR, Than KD, Ziewacz J, La Marca
Spine. 2011;14(2):151–66. F, Park P. Independent predictors of complication
35. Redmond KJ, Robertson S, Lo SS, Soltys SG,
following surgery for spinal metastasis. Eur Spine J.
Ryu S, McNutt T, et al. Consensus contouring 2013;22(6):1402–7.
guidelines for postoperative stereotactic body 41. Mantel F, Flentje M, Guckenberger M. Stereotactic
radiation therapy for metastatic solid tumor malig- body radiation therapy in the re-irradiation situation –
nancies to the spine. Int J Radiat Oncol Biol Phys. a review. Radiat Oncol. 2013;8(1):7.
2017;97(1):64–74. 42. Xu R, O’Connor K, Krol G, Yamada Y, Bilsky M,
36. Cox BW, Spratt DE, Lovelock M, Bilsky MH, Lis E, Laufer I, et al. Cement salvage of instrumentation-
Ryu S, et al. International spine radiosurgery consor- associated vertebral fractures. AJNR Am J
tium consensus guidelines for target volume defini- Neuroradiol. 2014;35(11):2197–201.
tion in spinal stereotactic radiosurgery. Int J Radiat 43. Zhang M, Appelboom G, Ratliff JK, Soltys SG, Adler
Oncol Biol Phys. 2012;83(5):e597–605. JR, Park J, et al. Radiographic rate and clinical impact
37. Website [Internet]. [cited 2019 Jan 8]. Available from: of pseudarthrosis in spine radiosurgery for metastatic
https://icru.org/home/reports/prescribing-recording- spinal disease. Cureus [Internet]. 2018. Available
and-reporting-photon-beam-therapy-report-50. from: https://doi.org/10.7759/cureus.3631.
38. Quraishi NA, Purushothamdas S, Manoharan SR,
44. Yoshimura S, Mori K, Kawasaki K, Tanabe A, Aikou
Arealis G, Lenthall R, Grevitt MP. Outcome of S, Yagi K, et al. A surgical case of radiotherapy
embolised vascular metastatic renal cell tumours induced esophageal perforation accompanying pyo-
causing spinal cord compression. Eur Spine J. genic spondylodiscitis: a case report. Surg Case Rep.
2013;22(Suppl 1):S27–32. 2017;3(1):98.
Vertebrectomy for Spinal
Metastases
42
Samuel Kalb and Juan S. Uribe
Introduction of cases. More than 50% of patients with spinal

metastasis have multiple levels involved. Lung
Approximately two-thirds of patients with malig- and breast cancers metastasize preferably into
nant tumors will develop bone metastasis. the thoracic spine since the venous drainage of
Primary tumors that most often lead to bone the breast through the azygos vein communi-
metastasis in the order of highest incidence are as cates with the plexus of Batson in the thoracic
follows: prostate, breast, kidney, lung, and thy- region [3].
roid cancer. Up to 70% of patients with breast or Spinal metastases are classified according to
prostate cancers and 15–30% of patients with their anatomical location. Almost 95% of spinal
lung, colon, bladder, or kidney cancers develop metastases are extradural lesions. The remaining
bone metastasis [1]. lesions are either intradural extramedullary or
The spine is the most common site of bone intramedullary metastases. Lesions located
metastasis with an estimated incidence of over purely within the epidural compartment and
10%. The most common initial anatomic location without bone involvement account for only a
of metastases within vertebrae is the posterior small fraction of extradural metastases [4].
portion of the body [2]. CT scans of affected indi- The management of patients with symptom-
viduals usually show the body of the vertebra atic metastatic spinal lesions is carried out to
being involved before the pedicles, although relieve pain and to preserve or restore neurologi-
destruction of the pedicles is the most common cal function. Life expectancy is often fairly short,
finding on plain X-ray films. Destruction of the with median survival ranging from 4 to
pedicles occurs only in combination with the 15 months. Since a cure is not a realistic expecta-
involvement of the vertebral body. tion, palliation is usually the aim of therapy.
Symptomatic lesions occur more frequently Treatment options vary based on the presenta-
in the thoracic region (70%), while the cervical tion of complete or incomplete neurologic defi-
spine is the least involved with only about 10% cits. Traditionally, decompressive laminectomy
was performed in an attempt to alleviate cord
S. Kalb compression. However, simple laminectomy pro-
Department of Neurological Surgery, Barrow vides insufficient decompression in cases where
Neurological Institute, Phoenix, AZ, USA tumors are either anteriorly or laterally located.
J. S. Uribe (*) Laminectomy alone is also likely to aggravate
Division of Spinal Disorders, Department of mechanical spinal instability, especially in the
Neurological Surgery, Barrow Neurological Institute,
case of vertebral collapse.
Phoenix, AZ, USA

https://doi.org/10.1007/978-3-030-42958-4_42
570 S. Kalb and J. S. Uribe
Improvement of minimally invasive surgery the intercostal nerve or intraforaminal radicu-

(MIS) techniques, such as the anterior or lateral lomedullary artery. In addition, the extrapleu-
exposures, along with the ongoing development ral nature of the approach decreases the risk of
and evolution of spinal stabilization instrumenta- injury to the aorta, vena cava, and sympathetic
tion, has greatly improved the efficacy and mor- plexus, as well as reducing the risk of develop-
bidity/mortality of surgical intervention. ing a pleural CSF fistula. The MIS version of
Numerous studies have validated the efficacy of the lateral approach allows for a significantly
MIS approaches to alleviate pain and improve smaller incision with smaller amount of rib
functional outcomes in the setting of metastatic retraction, ultimately resulting in decreased
vertebral lesions [5]. blood loss, postoperative pain, time to mobili-
Vertebrectomy constitutes the foundation for zation, and reduction in the length of hospital
restoration of the ventral spinal column during stay [6].
surgery for metastatic tumors. This operative This chapter focuses on the lateral minimal
strategy facilitates correction of deformity and invasive approach to access the spine when verte-
immediate stabilization. With the advances in brectomy for tumor is warranted.
chemotherapy and radiation therapy, the indica-
tions for vertebrectomy in the setting of meta-
static spine tumor have decreased over time. Preoperative Planning
Nonetheless, current indications to perform a
vertebrectomy as part of the surgical treatment The preoperative planning starts with a com-
include the following: plete history and physical examination. Most
patients will complain of back pain, which is
1. Oligo metastatic disease (one lesion and no either localized or in a radicular distribution.
systemic disease), therefore, taking out the Neurological deficits, which include motor
single and only tumor can eliminate tumor weakness and/or sensory derangement as well
burden. as bowel/bladder insufficiency or retention,
2. Expected long survival and risk of potential may or may not be present depending on the
hardware failure. Anterior column reconstruc- patient pathology. When present, one should
tion offers better long-term support. suspect compression of the spinal cord or nerve
3. Significant kyphotic deformity associated to roots, especially for intradural tumors. In addi-
the lesion (pathological fracture leading to tion, the presence of spinal deformity should
deformity). be taken into account.
4. Highly vascular lesion, thus, the whole tumor Radiographic evaluation should always
needs to be resected in order to control bleed- accompany any patient with suspected verte-
ing as with renal cell or melanoma. bral or spinal cord involvement in the setting of
metastatic disease. Magnetic resonance imag-
Vertebrectomy can be achieved through dif- ing (MRI) with and without contrast is the
ferent surgical approaches including anterior, ideal imaging technique. If MRI is not possi-
posterior, lateral, or a combination of each. In ble, a computer tomography (CT) myelogram
an effort to reduce morbidity related to a thora- is recommended. In both cases, imaging is nec-
cotomy or the extensive tissue damage from a essary to delineate the extent of the lesion,
posterior approach, lateral-based approaches determine the anatomical involvement, and
have gained popularity in recent years. This evaluate the degree of neural compression. In
technique allows direct view of the neural ele- addition, a plane CT is recommended to deter-
ments without the need to dissect or sacrifice mine the extent of vertebral involvement, and
42 Vertebrectomy for Spinal Metastases 571
standing scoliosis films to evaluate for any small incision, little blood loss, and short con-
form of deformity are essential for surgical valescence. However, this is technically a
planning [6]. demanding methodology. An understanding of
In the setting of known metastasis, radioiso- regional neurovascular and visceral anatomy
tope bone scans can be used to detect small bone is vital, and experience with small working
lesions, as it is sensitive in detecting osteolytic or corridors, tubular retractors, and minimally
osteoblastic activity. Angiography is beneficial invasive instrumentation are required.
when a hypervascular lesion is suspected. It is However, the overall outcomes have been
both a diagnostic tool to determine the blood remarkably well with overall complications of
supply as well as a therapeutic option in order to 12.5% [8].
initiate preoperative embolization with the aim of The surgical techniques for the MIS lateral
reducing intraoperative blood loss. Examples of approach for access to the thoracic spine begins
metastatic lesion that benefit from angiography with the patient positioned under fluoroscopic
include renal cell carcinoma, melanoma, and guidance in a true and direct lateral decubitus
chordoma. position on a flexible radiolucent surgical
table. For procedures involving only thoracic
levels, the patient is positioned with the table
Surgical Techniques break under the mid-surgical level. The side of
the approach is chosen depending on the loca-
inimal Invasive Lateral
M tion of the tumor, surrounding viscera, and the
Retropleural/Transthoracic Approach vertebral level. Under fluoroscopic guidance,
the index vertebral body level and tumor are
The minimal invasive retropleural approach to located and marked on the skin. A 3- to 6-cm
the thoracolumbar junction is considered a vari- oblique incision is marked parallel to the rib
ant of the lateral retropleural thoracotomy. In its traversing the pathologic vertebral body at the
essence, it combines many of the features of midaxillary line.
both the anterolateral transthoracic and the lat- The incision is made obliquely over the rib
eral extracavitary approaches. It grants the sur- across the region delineated by the skin mark-
geon the ability to remain outside the pleura ings. Dissection is carried down through the sub-
while achieving a ventral decompression of the cutaneous tissue to the ribs or intercostal space.
dural sac. Five to seven centimeters of the immediately
The advantage of having lateral exposure of underlying rib, directly over the lesion, are dis-
the thoracolumbar spine is that it allows the sur- sected in a subperiosteal fashion. Using a rib dis-
geon to visualize the thecal sac during the sector or Cobb elevator, the rib is removed from
approach to the tumor. The surgeon will then the underlying pleura and neurovascular bundle,
have control of both the thecal sac and pathology, removed, and saved for autograft at the end of the
as compared to more ventral approaches in which case. The intercostal muscles and parietal pleura
the thecal sac is not visualized until the disease are incised to enter the thoracic cavity for a trans-
process is resected [7]. thoracic approach, while the parietal pleura is
More recently, a mini-open anterolateral swept anteriorly with blunt finger dissection for a
approach to the thoracolumbar has been retropleural approach. Further rib resection may
described as a method to access the spine. The be required if a larger exposure is needed. The rib
potential advantages of this MIS approach resected for access to the thoracolumbar junction
include independence from an access surgeon, usually corresponds to 2 levels above the desired
vertebral level (i.e., 10th rib for access to T12, longitudinal ligament are preserved to protect
11th rib for L1, and 12th rib for L2). mediastinal and thoracic structures.
Once the rib is removed, an index finger is Once the corpectomy is done and decompres-
used to enter the pleural space (for a transpleural sion of the thecal sac when necessary is com-
approach) or the plane between the endothoracic pleted, ventral reconstruction is performed using
fascia and pleura (for a retropleural approach). expandable titanium cages, biological allograft,
The appropriate plane is developed, and dia- and the rib autograft harvested during the
phragm and/or lung are mobilized anteriorly approach. Spinal instrumentation is completed
using a finger and/or sponge stick until the lateral using ventrolateral plate/screw fixation through
face of the vertebral body, pedicle, and adjacent the expandable retractor and/or percutaneous
intervertebral discs are exposed. For access to the posterior pedicle screw/rod fixation. Dural repair,
thoracolumbar junction, it should be noted that when necessary after resection of intradural
removal of the diaphragmatic-costal attachment tumor or iatrogenic CSF leak, is performed with
may be required. Because of the lateral (costal) a running 5-0 suture. The dural repair is rein-
diaphragmatic insertion, and for access to L1, the forced with fibrin glue, and CSF is drained
lumbar or posterior attachments of the diaphragm through a lumbar catheter.
must be sharply transected off the transverse pro- Following a transthoracic approach or in the
cess of L1. The intervening attachment between event of a pleural violation air must be removed
the medial and lateral arcuate ligaments must from the pleural cavity, which is traditionally
also be cut to fully expose the lateral vertebral accomplished by placement of a chest tube.
body. If more anterior exposure of the vertebral Alternatively, a red rubber catheter can be situ-
body is needed, the ipsilateral crus, which extend ated in the pleural space through the wound, and
along the anterolateral spine to L2 on the left and placed under a water trap (i.e., with the distal end
L3 on the right, may also be transected. submerged under water). The surgical wound is
For a left-sided approach, the aorta and hemia- closed in standard fashion, including the muscu-
zygos vein are also retracted anteriorly. Segmental lar and fascial layers. The red rubber catheter is
vessels are ligated as proximally as possible. secured with a purse-string stitch, and a Valsalva
Sequential tubular dilators are then inserted, and maneuver with end-inspiratory hold is performed
an expandable retractor system is inserted over until no more air bubbles are observed to ema-
the largest dilator and secured with a flexible nate from the submerged distal end of the cathe-
table-mounted arm assembly. ter, representing evacuation of all air from the
With the retractor placed and adequate expo- thoracic cavity. The red rubber catheter is
sure obtained, the next step before proceeding removed as the purse string is tied. This tech-
with the corpectomy is to expose the dura by nique obviates the use of a chest tube.
removing the pedicle with rongeurs and a high- A chest radiograph is obtained immediately
speed drill. The intervertebral discs above and after surgery and on the morning of postoperative
below the vertebral body of interest are then day 1, to verify the absence of pneumothorax if
removed, and osteotomes are used to delineate the aforementioned red rubber technique was
the area of the corpectomy. At this point, bony used, or to verify placement and position of a
removal can be achieved using a combination of chest tube if one was placed intraoperatively. In
rongeurs, curettes, high-speed drills, and osteo- this case, it is initially placed on suction and
tomes. A thin layer of bone on the ventral and weaned to water seal. Serial chest radiographs
contralateral sides of the body and the anterior are obtained to confirm re-expansion of the lung
42 Vertebrectomy for Spinal Metastases 573
before removal of the chest tube. Declining oxy- Case Example

gen saturation or recurrence of a pneumothorax
warrants further evaluation and, if necessary, sur- Preoperative and postoperative images of a
gical re-exploration. The patients are encouraged young (mid-30s) female patient with known
to ambulate postoperatively with thoracolumbo- metastatic breast cancer who underwent mini-
sacral orthoses. Obtaining upright radiographs mally invasive lateral retropleural T10 corpec-
are recommended to verify hardware placement tomy with cage fixation and fusion (Figs. 42.1,
and stability. 42.2, and 42.3).
Fig. 42.2 Postoperative T2 sagittal MRI images show

adequate decompression of the spinal cord
Fig. 42.1 Preoperative T2 sagittal MRI images reveal
tumor involving T10 vertebra in all three columns. Spinal
and paraspinal enhancement represent recurrent tumor
and postsurgical effects
a b
Fig. 42.3 Sagittal (a) and (b) Coronal CT scan show T10 corpectomy with cage and plate instrumentation
5. Molina C, Gokaslan Z, Sciubba D. A systematic

References review of the current role of minimally invasive spine
surgery in the management of metastatic spine dis-
1. Roodman GD. Mechanisms of bone metastasis. N ease. Int J Surg Oncol. 2011;2011:598148.
Engl J Med. 2004;350(16):1655–64. 6. Park MS, Deukmedjian AR, Uribe JS. Minimally
2. Algra PR, Heimans JJ, Valk J, Nauta JJ, Lachniet M, invasive anterolateral corpectomy for spinal tumors.
Van Kooten B. Do metastases in vertebrae begin in the Neurosurg Clin N Am. 2014;25(2):317–25.
body or the pedicles? Imaging study in 45 patients. 7. Uribe JS, Dakwar E, Cardona RF, Vale FL. Minimally
AJR Am J Roentgenol. 1992;158(6):1275–9. invasive lateral retropleural thoracolumbar
3. Togawa D, Lewandrowski KU. The pathophysiology approach: cadaveric feasibility study and report of
of spinal metastases. In: RF ML, Lewandrowski KU, 4 clinical cases. Neurosurgery. 2011;68(1 Suppl
Markman M, Bukowski RM, Macklis R, Benzel EC, Operative):32–9.
editors. Cancer in the spine. Current clinical oncol- 8. Baaj AA, Dakwar E, Le TV, Smith DA, Ramos E,
ogy. Totowa: Humana Press; 2006. Smith WD, Uribe JS. Complications of the mini-open
4. Jacobs WB, Perrin R. Evaluation and treatment of spi- anterolateral approach to the thoracolumbar spine. J
nal metastases. Neurosurg Focus. 2001;11(6):e10. Clin Neurosci. 2012;19(9):1265–7.
Minimally Invasive Surgery
for Spinal Metastases
43
Robert J. Rothrock, Ori Barzilai, Mark H. Bilsky,
and Ilya Laufer
Introduction While decompression surgery remains a stan-

dard approach for the treatment of symptomatic,
Surgical care for spinal metastases has evolved spinal cord compression from epidural metastatic
over time in concert with other cancer-treatment disease, other surgical indications have emerged
improvements. Historically, surgery for spinal for treatment of spinal metastatic disease.
metastases centered on gross total or en-bloc Pathologic fractures from tumor erosion and
resection, coupled with mechanical reconstruc- infiltration are a debilitating source of pain in
tion. This often involved combined approaches patients with metastatic spinal disease [6].
(i.e., front-back) with multilevel vertebrectomy, Although they do not necessarily have associated
prolonged operative times, relatively high periop- neurologic deficits, especially in the absence of
erative complication rates, and, importantly, pro- compressive epidural disease, the pain associated
longed recovery times for patients with systemic with spinal metastatic disease can decrease
cancer and thus increased peri-operative morbid- patient mobility, with deleterious effect on sur-
ity [1–4]. With the advent of stereotactic body vival and negative effects on quality of life [7, 8].
radiation therapy (SBRT) and improved tumor For patients without high-grade spinal cord com-
control, through better radiation and systemic pression or in whom epidural disease can be
therapies, the paradigm for management of meta- effectively treated with SBRT, but still have
static epidural spinal disease has shifted to hybrid tumor-related instability, minimally invasive sur-
therapy [5]. Hybrid therapy describes the combi- gical stabilization can be a useful and efficacious
nation of separation surgery promptly followed treatment [9, 10].
by stereotactic body radiation therapy (SBRT) to
treat remaining noncompressive osseous and
paraspinal disease. Separation surgery describes inimally Invasive Surgery
M
circumferential decompression of the spinal neu- for Spinal Metastases
ral elements and stabilization of the spinal col-
umn, without the goal of gross total resection of Minimally invasive surgical (MIS) approaches
osseous and paraspinal disease. have gained popularity in treating spine trauma,
deformity, and degenerative disease. In cancer
R. J. Rothrock · O. Barzilai · M. H. Bilsky patients, spinal MIS techniques might offer some
I. Laufer (*) advantages over open techniques [11]. Smaller
Department of Neurosurgery, Memorial Sloan incisions help to minimize risk of intraoperative
Kettering Cancer Center, New York, NY, USA and postoperative blood loss, and there is some
e-mail: lauferi@mskcc.org

https://doi.org/10.1007/978-3-030-42958-4_43
evidence that they have less associated postoper- metastatic cancer patient. It utilizes four sentinel
ative pain [10]. Importantly within the context of decision points to assess disease: neurologic,
cancer care, MIS techniques facilitate the return oncologic, mechanical stability, and systemic dis-
to early systemic and radiation therapy with ease [17, 18]. This framework standardizes assess-
smaller incisions and less healing time [12]. MIS ment and allows for the incorporation of
approaches may offer benefit to patients with evidence-based medicine and the rational use of
advanced systemic disease and higher periopera- new radiation, surgical, interventional radiology,
tive risk who might not tolerate more extensive and systemic therapies. The neurologic assess-
intervention. ment evaluates both clinical and radiologic param-
Although the goal with instrumented spine sur- eters, including the presence of myelopathy,
gery is usually solid arthrodesis, the combination functional radiculopathy, and the degree of epi-
of poor bone quality, radiation, and chemotherapy dural spinal cord compression (ESCC). A vali-
severely undermines the potential for osseous dated magnetic resonance–based scoring system
healing in cancer patients [13]. Given that instru- (known as the Bilsky Grade) is used to define the
mentation for cancer-associated spinal instability extent of epidural spinal cord compression, and
does not necessarily have the goal of eventual patients are dichotomized into high-grade and
solid bony fusion, it is well suited to minimally low-grade ESCC groups [19]. The oncologic con-
invasive techniques [10]. As in all relevant areas sideration is based on the expected tumoral
of spine surgery, however, MIS techniques must response, principally to radiation, but also to sys-
be implemented with a clear understanding of the temic therapy. Tumor histology is categorized
surgical goals and without compromising the abil- according to the response to conventional external
ity to safely accomplish them. beam radiation therapy (cEBRT) as radiosensitive
or radioresistant. The neurologic and oncologic
assessments are combined to determine the opti-
Contraindications to MIS Approach mal radiation strategy to achieve tumor control
and/or the need for a surgical intervention.
Multi-level tumors and high-grade spinal cord Mechanical instability is a separate consider-
compression present significant challenges in the ation, and is generally defined according to the
application of MIS techniques [14]. Sometimes, Spinal Instability Neoplastic Score (SINS) criteria
however, a minimal access type approach can be [20]. In this classification system, tumor-related
employed to achieve circumferential separation instability is assessed by adding together six indi-
coupled with percutaneous instrumentation. vidual component scores: spine location, pain,
Highly vascularized tumors such as renal cell lesion bone quality, radiographic alignment, verte-
carcinoma and solitary fibrous tumors also favor bral body collapse, and posterolateral involvement
open surgical approaches to allow open access of the spinal elements [21]. The minimum score is
for hemostasis and rapid tumor removal [14]. 0 and the maximum is 18. A score of 0–6 denotes
MIS approaches for metastatic spinal disease stability, 7–12 denotes indeterminate (possibly
have been largely limited to the thoracic and lum- impending) instability, and 13–18 denotes instabil-
bar spine, and have not been widely utilized in ity. Patients with mechanical instability typically
the cervical spine [15]. There are reported cases, require stabilization with spinal instrumentation or
however, of percutaneous, navigated instrumen- cement. Spinal instability serves as a separate sur-
tation in the cervical spine [16]. gical indication as there is a role for stabilization
surgery in patients with mechanical instability but
without overt spinal cord compression [12]. We
NOMS Framework and SINS have previously published our algorithm for mini-
mally invasive treatments for pathologic vertebral
The NOMS decision framework allows for flexi- compression fractures (Fig. 43.1) [9].
ble, multifactorial decision-making to help define The fourth consideration is the extent of sys-
the appropriate balanced treatment plan for a given temic disease and medical comorbidities that
43 Minimally Invasive Surgery for Spinal Metastases 577
Fig. 43.1 Memorial
Sloan Kettering Cancer Vertebral compression
Center (MSKCC) fracture
treatment algorithm for
metastatic
thoracolumbar
compression fracture Percutaneous instrumented
Yes
Decompression required stabilization +
“mini-open” decompression
No
Yes Percutaneous instrumented

Mechanical radiculopathy stabilization +
unilateral facetectomy
No
Cord compression Yes Percutaneous instrumented

amenable to RT stabilization
No
Yes Percutaneous instrumented

Posterior elements
stabilization +
involved
kyphoplasty
No
Kyphoplasty
affect the risk–benefit ratio of a proposed inter- tomy and transpedicular ventral epidural decom-
vention, taking into account the overall expected pression can be performed via a mini-open
survival and the ability of a patient to tolerate midline incision or via a tubular retraction sys-
spine-specific treatment. tem. This can then be coupled with transfascial or
percutaneous instrumentation. This mini-open
approach can allow for circumferential decom-
inimal Access Surgery (MAS)
M pression, and spare the more extensive muscle
for Decompression dissection required for open pedicle screw place-
and Stabilization ment traditionally performed with long segment
fixation. For patients with compromised bone
Using the NOMS framework, indications for sep- quality, fenestrated screws with cement can be
aration surgery include patients with radioresis- utilized to augment fixation in both minimally
tant tumor histology with high-grade ESCC with invasive and open surgical approaches [12].
or without mechanical instability, who can toler-
ate surgery from a medical and systemic perspec-
tive [22]. For patients who require separation AS Facetectomy for Mechanical
M
surgery for preserved ambulation, minimal access Radiculopathy
techniques coupled with percutaneous instru-
mentation offers an efficacious alternative with Patients with lumbar burst fractures with exten-
demonstrated improvement in patient reported sion into the pedicle and facet frequently experi-
outcomes [9, 10]. In this case, bilateral laminec- ence mechanical radiculopathy [23]. Mechanical
radiculopathy is a clinical syndrome in which a significant improvement in Eastern Cooperative

radicular pain results from lumbar instability Oncology Group (ECOG) status at 3-month fol-
associated with metastatic disease. In this syn- low up [23].
drome, radicular pain occurs with axial loading Our paradigm for minimally invasive treat-
and/or ambulation, but is absent with recumbence ment of mechanical radiculopathy includes per-
or rest. The presence of mechanical radiculopa- cutaneous short-segment instrumented
thy will also correspond with a high intermediate stabilization with fenestrated screw cement aug-
or unstable SINS [10]. In our experience, effec- mentation, kyphoplasty at the level of the burst
tive treatment of the radiculopathic component fracture, and minimal access surgery (MAS) fac-
requires instrumented stabilization and thorough etectomy with pediculectomy or decompression
decompression of the thecal sac and exiting nerve of the affected nerve root. The MAS facetectomy
root, usually requiring ipsilateral facetectomy is performed using an expandable retractor
and pediculectomy (Fig. 43.2). In a series of 55 inserted on the side ipsilateral to the radiculopa-
patients undergoing surgical decompression and thy, usually through one of the incisions created
stabilization for mechanical radiculopathy, 98% for screw placement, and the facet and pedicle
of patients had a significant improvement in are resected until full decompression of the exit-
visual analog scale (VAS) scores and 41.5% had ing and traversing nerve roots is achieved [9].
a b d e
Fig. 43.2 (a) A 71-year-old female with metastatic lung percutaneous bilateral instrumented stabilization from L1
adenocarcinoma who presented with severe low back pain to L3. Axial CT demonstrates the amount of right L2 ped-
and right lower extremity radiculopathy from a metastatic icle removal necessary to thoroughly decompress the dor-
lesion to L2. Both her back pain and leg pain completely sal root ganglion. (d) Intraoperative kyphoplasty was
resolved with recumbency, and her clinical exam and syn- performed at L2 to augment the lytic pathologic fracture.
drome were consistent with mechanical radiculopathy. Her (e) Standing lateral X-ray demonstrates the final surgical
SINS was 8 and there was Bilsky Grade 1c epidural dis- construct. The patient underwent CT myelography for
ease at L2. (b) Axial CT demonstrates the lytic lesion at simulation on postoperative day 2 and was treated with
L2. (c) Patient underwent right L2-L3 MAS facetectomy 27 Gy in three fractions SBRT to L2 beginning approxi-
with excision of the L2 pedicle and cement-augmented mately 2 weeks after surgery
ercutaneous Pedicle Screw

P index treatment level) can be combined with
Fixation for Tumor-Induced Spinal screw cement augmentation by using fenestrated
Instability screws [12]. In addition, in patients with mechan-
ically unstable fractures without significant epi-
In patients with mechanical instability without dural extension, balloon kyphoplasty can be a
high-grade ESCC or with highly radiosensitive useful adjuvant to the percutaneous instrumented
tumor with ESCC, percutaneous pedicle fixation constructs, providing additional anterior column
can offer durable stability without open surgical support [9, 25]. Intraoperative imaging- either a
arthrodesis [12]. In the case of highly radiosensi- navigation system or standard fluoroscopy- is
tive epidural tumor such as lymphoma and mul- used for localization and incision planning, can-
tiple myeloma, conventional radiation therapy nulation of pedicles, and screw placement.
offers effective tumoricidal control without the Fluoroscopy is required for monitoring the
need for decompressive surgery [24]. In the cement injection (Fig. 43.3).
absence of high-grade ESCC, patients with
mechanical instability due to extensive cortical
destruction or a fracture extending into the poste- Vertebroplasty/Kyphoplasty
rior elements can be treated with percutaneous
instrumented stabilization, since kyphoplasty Vertebroplasty and Kyphoplasty are both effec-
would not provide adequate stabilization. Short tive procedures utilized in the treatment of pain
constructs (one level above, one level below the from pathologic vertebral body fractures.
a b c d
Fig. 43.3 (a) 35-year-old BRCA1-positive female with with intraoperative kyphoplasty at L1 to augment the lytic
metastatic breast adenocarcinoma who presented with pathologic fracture. (d) Standing lateral X-ray demon-
severe axial low back pain, prompting MRI and revealing strates the final surgical construct. The patient underwent
a pathologic fracture at L1. (b) CT L-spine demonstrates CT simulation on postoperative day 2 and was treated
a lytic lesion at L1 with SINS 10 and Bilsky grade 1a epi- with 27 Gy in three fractions SBRT to L1 beginning
dural disease. (c) Patient underwent cement augmented approximately 2 weeks after surgery
percutaneous bilateral instrumentation from T12 to L2
In kyphoplasty, a balloon is inflated in the verte- cord and nerve roots during treatment of extra-
bral body to create a cavity into which bone vertebral lesions [31, 32].
cement can be injected, while in vertebroplasty, MRI-guided laser interstitial thermal therapy
bone cement is injected into the vertebral body (LITT) has the advantage of real-time thermal
without the balloon [26]. In a randomized, con- monitoring to help prevent direct injury to adja-
trolled trial of balloon kyphoplasty versus non- cent tissues during treatment. MRI thermography
surgical treatment of symptomatic pathologic enables noninvasive, real-time monitoring of the
fractures in patients with one to three lesions, ablation zone [33]. Using this technology, the
patients undergoing intervention had a significant surgeon can monitor heat intensity and spread in
improvement in Roland-Morris disability ques- real time to customize treatment. In select cases,
tionnaire (RDQ) score at 1 month compared to spinal LITT has been used as a minimally inva-
controls [27]. Vertebral bone-cement augmenta- sive approach to treat asymptomatic high-grade
tion can be used in conjunction with SBRT and compressive epidural tumor without interruption
can be performed without interruption of chemo- of systemic therapy [33].
therapy or other systemic therapy [28]. We utilize
isolated vertebroplasty/ kyphoplasty in the set-
ting of painful compression fractures without Conclusion
involvement of the posterior elements (Fig. 43.1).
Surgical intervention in the metastatic cancer
population is palliative, and thus these patients
Ablation should be considered for less-invasive procedures
that limit the interruption of systemic therapy and
Image-guided ablation therapies for spinal metas- allow for the delivery of early adjuvant radiation.
tases have been introduced as a minimally inva- As in all relevant areas of spine surgery, however,
sive alternative to conventional surgical MIS techniques must be implemented with a clear
interventions for patients who are not good surgi- understanding of the surgical goals and without
cal candidates [29]. CT/fluoroscopic-guided tech- compromising the ability to safely accomplish
niques include radiofrequency ablation (RFA), them. MIS approaches may offer benefits over
cryoablation (or cryotherapy), and microwave open approaches to patients with advanced sys-
ablation [29]. MRI-guided techniques include temic disease and higher perioperative risk who
laser interstitial thermal therapy (LITT) and might not tolerate more extensive intervention.
focused ultrasound [29]. In these procedures, a
probe is directly inserted into the targeted tissue
and activated to directly injure tumor. References
CT/fluoroscopic-guided thermal ablation can
1. Mokdad AH, Dwyer-Lindgren L, Fitzmaurice C,
be performed in conjunction with vertebroplasty/
Stubbs RW, Bertozzi-Villa A, Morozoff C, et al.
kyphoplasty to treat pathologic fractures with Trends and patterns of disparities in cancer mor-
the advantage of combining both treatments into tality among US counties, 1980–2014. JAMA.
one outpatient procedure. Combination RFA and 2017;317(4):388–406.
2. Railton C. Perioperative use of cardiac medications in the
vertebral augmentation has been shown to be a
high-risk patient. In: Anesthesia for the high risk patient.
safe and effective therapy for palliation of pain- Cambridge: Cambridge University Press; p. 58–67.
ful spinal metastases in carefully selected https://doi.org/10.1017/CBO9780511576652.006.
patients [30]. RFA and cryoablation (X-ray 3. Audisio RA, Ramesh H, Longo WE, Zbar AP, Pope
D. Preoperative assessment of surgical risk in oncoge-
guided techniques) are usually reserved for
riatric patients. Oncologist. 2005;10(4):262–8.
lesions within the vertebral body as there have 4. Tomita K, Kawahara N, Murakami H, Demura
been reports of thermal injuries to the spinal S. Total en bloc spondylectomy for spinal tumors:
improvement of the technique and its associated basic 17. Patchell RA, Tibbs PA, Regine WF, Payne R, Saris
background. J Orthop Sci. 2006;11(1):3–12. S, Kryscio RJ, et al. Direct decompressive surgical
5. Barzilai O, Laufer I, Robin A, Xu R, Yamada Y, resection in the treatment of spinal cord compres-
Bilsky MH. Hybrid therapy for metastatic epidural sion caused by metastatic cancer: a randomised trial.
spinal cord compression: technique for separation Lancet. 2005;366(9486):643–8.
surgery and spine radiosurgery. Oper Neurosurg 18. Laufer I, Rubin DG, Lis E, Cox BW, Stubblefield MD,
(Hagerstown) [Internet]. 2018 Jun 8. Available from: Yamada Y, et al. The NOMS framework: approach to
https://doi.org/10.1093/ons/opy137 the treatment of spinal metastatic tumors. Oncologist.
6. Vassiliou V, Chow E, Kardamakis D, Lauzon 2013;18(6):744–51.
N. Natural history, prognosis, clinical features 19. Bilsky MH, Laufer I, Fourney DR, Groff M, Schmidt
and complications of metastatic bone disease. MH, Varga PP, et al. Reliability analysis of the epi-
In: Cancer metastasis – biology and treatment. dural spinal cord compression scale. J Neurosurg
Dordrecht: Springer; 2013. p. 19–36. https://doi. Spine. 2010;13(3):324–8.
org/10.1007/978-1-4020-9819-2_4. 20. Fourney D, DiPaola C, Fisher C. P153. A novel clas-
7. Reyes-Gibby CC, Anderson KO, Merriman KW, Todd sification system for spinal instability in neoplastic
KH, Shete SS, Hanna EY. Survival patterns in squa- disease: an evidence based approach and expert con-
mous cell carcinoma of the head and neck: pain as sensus from the Spine Oncology Study Group. Spine
an independent prognostic factor for survival. J Pain. J. 2009;9(10):193S.
2014;15(10):1015–22. 21. Fourney DR, Frangou EM, Ryken TC, Dipaola CP,
8. Campbell G, Hagan T, Gilbertson-White S, Houze Shaffrey CI, Berven SH, et al. Spinal instability neo-
M, Donovan H. Cancer and treatment-related symp- plastic score: an analysis of reliability and validity
toms are associated with mobility disability in women from the spine oncology study group. J Clin Oncol.
with ovarian cancer: a cross-sectional study. Gynecol 2011;29(22):3072–7.
Oncol. 2016;143(3):578–83. 22. Barzilai O, Laufer I, Yamada Y, Higginson DS,

9. Barzilai O, McLaughlin L, Amato M-K, Reiner AS, Schmitt AM, Lis E, et al. Integrating evidence-based
Ogilvie SQ, Lis E, et al. Minimal access surgery for medicine for treatment of spinal metastases into a
spinal metastases: prospective evaluation of a treat- decision framework: neurologic, oncologic, mechani-
ment algorithm using patient-reported outcomes. cals stability, and systemic disease. J Clin Oncol.
World Neurosurg. 2018;120:e889–901. 2017;35(21):2419–27.
10. Molina CA, Gokaslan ZL, Sciubba DM. A system- 23. Moliterno J, Veselis CA, Hershey MA, Lis E, Laufer
atic review of the current role of minimally invasive I, Bilsky MH. Improvement in pain after lumbar sur-
spine surgery in the management of metastatic spine gery in cancer patients with mechanical radiculopa-
disease. Int J Surg Oncol. 2011;2011:1–9. thy. Spine J. 2014;14(10):2434–9.
11. Zuckerman SL, Laufer I, Sahgal A, Yamada YJ,
24. Osborn VW, Lee A, Yamada Y. Stereotactic body radi-
Schmidt MH, Chou D, et al. When less is more. Spine. ation therapy for spinal malignancies. Technol Cancer
2016;41:S246–53. Res Treat. 2018;17:1533033818802304.
12. Moussazadeh N, Rubin DG, McLaughlin L, Lis
25. Sun G, Li L, Jin P, Liu X-W, Li M. Percutaneous ver-
E, Bilsky MH, Laufer I. Short-segment percutane- tebroplasty for painful spinal metastasis with epidural
ous pedicle screw fixation with cement augmenta- encroachment. J Surg Oncol. 2014;110(2):123–8.
tion for tumor-induced spinal instability. Spine J. 26.
Mathis JM, Orlando Ortiz A, Zoarski
2015;15(7):1609–17. GH. Vertebroplasty versus kyphoplasty: a compari-
13. Rades D, Fehlauer F, Schulte R, Veninga T, Stalpers son and contrast. In: Percutaneous vertebroplasty and
LJA, Basic H, et al. Prognostic factors for local kyphoplasty. Dordrecht: Springer; 2004. p. 145–56.
control and survival after radiotherapy of meta- 27. Berenson J, Pflugmacher R, Jarzem P, Zonder

static spinal cord compression. J Clin Oncol. J, Schechtman K, Tillman JB, et al. Balloon
2006;24(21):3388–93. kyphoplasty versus non-surgical fracture man-
14. Rose PS, Clarke MJ, Dekutoski MB. Minimally inva- agement for treatment of painful vertebral body
sive treatment of spinal metastases: techniques. Int J compression fractures in patients with cancer: a mul-
Surg Oncol. 2011;2011:494381. ticentre, randomised controlled trial. Lancet Oncol.
15. Wong AP, Lall RR, Dahdaleh NS, Lawton CD, Smith 2011;12(3):225–35.
ZA, Wong RH, et al. Comparison of open and mini- 28. Barzilai O, DiStefano N, Lis E, Yamada Y, Lovelock
mally invasive surgery for intradural-extramedullary DM, Fontanella AN, et al. Safety and utility of kypho-
spine tumors. Neurosurg Focus. 2015;39(2):E11. plasty prior to spine stereotactic radiosurgery for
16. Schaefer C, Begemann P, Fuhrhop I, Schroeder M, metastatic tumors: a clinical and dosimetric analysis.
Viezens L, Wiesner L, et al. Percutaneous instrumen- J Neurosurg Spine. 2018;28(1):72–8.
tation of the cervical and cervico-thoracic spine using 29. Kurup A, Callstrom M. Image-guided percutane-

pedicle screws: preliminary clinical results and analy- ous ablation of bone and soft tissue tumors. Semin
sis of accuracy. Eur Spine J. 2011;20(6):977–85. Intervent Radiol. 2010;27(03):276–84.
30. Wallace AN, Greenwood TJ, Jennings JW. Radio- 32. Nakatsuka A, Yamakado K, Maeda M, Yasuda M,
frequency ablation and vertebral augmentation for Akeboshi M, Takaki H, et al. Radiofrequency abla-
palliation of painful spinal metastases. J Neuro- tion combined with bone cement injection for the
Oncol. 2015;124(1):111–8. treatment of bone malignancies. J Vasc Interv Radiol.
31. Nakatsuka A, Yamakado K, Takaki H, Uraki J, Makita 2004;15(7):707–12.
M, Oshima F, et al. Percutaneous radiofrequency 33. Tatsui CE, Lee S-H, Amini B, Rao G, Suki D, Oro M,
ablation of painful spinal tumors adjacent to the spinal et al. Spinal laser interstitial thermal therapy: a novel
cord with real-time monitoring of spinal canal tem- alternative to surgery for metastatic epidural spinal
perature: a prospective study. Cardiovasc Intervent cord compression. Neurosurgery. 2016;79(Suppl
Radiol. 2008;32(1):70–5. 1):S73–82.
Deformity Secondary to Vertebral
Body Metastases
44
Zach Pennington, A. Karim Ahmed,
and Daniel M. Sciubba
Introduction broken into neurologic and structural indica-

tions. The former topic is covered extensively
Neoplastic disease is the second most common elsewhere; in this chapter, we will focus on
cause of death in the United States, claiming structural indications, namely, the instability
nearly 600,000 lives in 2017 [1]. Yet advances in and deformity that may occur secondary to bony
radiation and chemotherapeutic regimens con- destruction. We begin with a brief overview of
tinue to drastically increase life expectancies the biomechanics of the mobile spine, followed
across all stages of malignancy [2]. by a description of the etiology of bony destruc-
Consequently, more and more patients are living tion, and finishing with diagnosis and treatment
with the long-term stigmata of cancer, namely, of mechanical instability and deformity second-
metastatic disease, which is now reported in up ary to metastatic disease.
to 64% of patients at the time of diagnosis [3–
6]. Within bony metastases, the most common
site for metastatic disease is the mobile spine [7, Biomechanical Model of the Spine
8] with some evidence suggesting that up to
70% of patients have spinal involvement at the verview and the Basis of Focal
O
time of death [9], most commonly in the tho- Kyphosis
racic (70%) and lumbar regions (20%) [10–18].
Though these metastases are often asymptom- Analysis of the biomechanics of the bony spine
atic or so clinically indolent as to be masked by requires considering both the bony and non-bony
a patient’s other symptoms, more than 1 in 10 elements, including the ligamentous complex and
patients will have associated symptoms severe paraspinal musculature. Additionally, though
enough to require surgery [12–15, 19–28]. uninvolved in the support of the spine itself, the
Operative indications for these lesions can be truncal soft tissues – namely, the thoracic contents
and abdominal viscera – must be considered for
their ability to alter the forces applied to each bony
Z. Pennington · A. K. Ahmed
Department of Neurosurgery, Johns Hopkins level. Much of the data describing the contribu-
Hospital, Baltimore, MD, USA tions of each of these elements to the structural
D. M. Sciubba (*) integrity of the spine has derived from the trauma
Spine Tumor and Spine Deformity Surgery, literature, and so our knowledge of how each con-
Department of Neurosurgery, Johns Hopkins tributes to the spine integrity results from examin-
Hospital, Baltimore, MD, USA ing the instability created by its disruption.
e-mail: dsciubb1@jhmi.edu

https://doi.org/10.1007/978-3-030-42958-4_44
584 Z. Pennington et al.
In all models, biomechanics of the spine can them experience smaller decreases in joint
be best described by treating each vertebra as an thickness though, producing an asymmetrical
individual segment subject to a finite set of settling of the vertebrae and induction of a small
forces and torques. The vertebra is broken into kyphosis at each level. This results in two
the large vertebral body, which supports most of effects. First, the realized axis of rotation of the
the loading from the superior and inferior verte- vertebra shifts backward toward the facet joint,
bral levels under physiological conditions, and and second, a greater proportion of the static
the posterolateral elements comprised of the forces that each vertebra applies to the other is
pedicles, facets, laminae, and spinous process. applied through the facet joints (Fig. 44.1).
Compared to the forces supported by the verte- When said changes occur at multiple levels,
bral body, those supported by the facet joints are there is a global kyphosing of the spine, giving
trivial in young patients. As patients age though, rise to the so-called Dowager’s hump. If this
the intervertebral disks that allow force transfer process occurs in only one segment, as might
between the vertebral bodies degenerate, com- occur with a neoplastic or fragility-related com-
pressing the vertebral bodies together. The pos- pression fracture, then a focal kyphosis will
terior elements and the facet joints that divide develop, producing a de novo deformity.
a Normal aging
Anterior
Anterior
θ intervertebral
intervertebral
spacing
spacing
Age-related
Poterior degeneration of
intervertebral intervertebral
spacing discs
b Compression fracture (metastatic disease)
Anterior Anterior
θ intervertebral
intervertebral
spacing spacing
Poterior Compression
intervertebral fracture 2° to
spacing osteolytic lesion
Fig. 44.1 Diagram illustrating age-related kyphosing of axis of rotation within each vertebra. This kyphosing (rep-
the thoracic spine. As the intervertebral discs degrade (a) resented by θ, right diagrams) is exacerbated by collapse
there is a reduction of the anterior intervertebral spacing of the vertebral body (b) such as occurs in osteoporosis or
without concomitant decreases in posterior intervertebral metastatic disease of the spine, resulting in de novo
spacing. This leads to an increased shift of force onto the deformity
facet joints and posterior movement of the instantaneous
44 Deformity Secondary to Vertebral Body Metastases 585
enis Three-Column Model and Its

D The result of this kyphosis may take the
Application to Metastatically form of one of two clinical manifestations. The
Involved Vertebrae most obvious is that there is a shortening of the
spinal column, which patients may report as
The most well-known model of spine biomechan- decreased height, a steady downward deflec-
ics is that proposed by Francis Denis who used a tion of gaze (“looking at my feet”), and/or the
series of several hundred thoracolumbar injuries formation of a “hump” as the anterior and mid-
to develop what is now known as the three-col- dle columns shorten relative to the posterior
umn model (Fig. 44.2a) [29]. Denis’ model built column. The second consequence of progres-
on the older model of Holdsworth [30, 31] and sive kyphosing of the vertebral bodies is a sig-
evaluated spinal biomechanics by considering nificant increase in axial back pain. Milder
each vertebra as a separate segment in a larger forms of anterior wedging, as are common
construct. Using this system, each vertebra is among the aging population [32], are often
divided into anterior, middle, and posterior col- unassociated with pain. But more severe com-
umns, respectively, comprised of the ALL and pression fractures are generally symptomatic
anterior half of the vertebral body, the posterior [33] and may require surgical intervention.
half of the vertebral body and PLL, and the pos- Furthermore, in the context of metastatic dis-
terolateral elements (pedicle, laminae, spinous ease, extensive involvement of the vertebrae
process, and associated soft tissues of the poste- may also create a baseline oncological pain
rior tension band). Using this model, compression that is then punctuated by a new mechanical
fractures – those most commonly seen in the con- component.
text of metastasis-related deformity – were
defined as failures of the anterior column. Burst
fractures – the other common fracture type in the Biomechanical Changes
metastatic spine – were classified as failures of the to the Spine in the Context
anterior and middle column, suggesting that the of Osteolysis
two fracture types may simply represent different
levels of severity on the same spectrum. As will be discussed later in this chapter, the
Upon examination of the force diagrams in bone of metastatically involved vertebrae can be
Fig. 44.2b, it is apparent that weakening or desta- divided into osteoblastic and osteolytic lesions.
bilization of the anterior column allows the mid- Osteoblastic lesions, also referred to radiographi-
dle column to act as a hinge. Torques applied by cally as sclerotic lesions, are characterized by
the prevertebral soft tissue mass and loading of stimulation of the local osteoblasts and deposi-
the superior endplate of the anterior column tion of new bone [34]. By contrast, osteolytic or
enable this rotation and collapse the structurally radiolucent lesions, are characterized by progres-
compromised anterior column, producing ante- sive destruction of local bone and stimulation of
rior wedging, eventually followed by complete osteoclasts. Osteolytic lesions are generally more
collapse of the vertebral body (vertebrae plana). relevant to the discussion of metastasis-effected
In heavily involved vertebrae, the middle column deformity as they cause greater destabilization of
is also compromised, preventing it from acting as the vertebrae. In both cases though, metastasis
a hinge. Here, the instantaneous axis of rotation – formation relies upon interplay between osteo-
normally found within the middle column – is blasts (the bone-forming cells) and osteoclasts
forced dorsally to the zygapophyseal joints. (the bone-resorbing cells) which is by and large
These joints then serve as the fulcrum about controlled by the receptor activator of NFκB
which the axial plane of the vertebra rotates, cre- ligand (RANKL) and its noncompetitive antago-
ating focal kyphosis. nist osteoprotegerin.
Column
a Posterior Middle Anterior
Paraspinals and
Static torques
posterior tension
PLL band
Lordosing
b torques
FN
anterior
column –
cephalad VB
Local osseoligamentous Anterior
anatomy (simplified) Superior soft tissues Inferior facets –
facets –
cephalad
cephalad
Kyphosing vertebra
vertebra
FN Anterior ALL torques
column –
cephalad VB
Static Forces
FN
FN cephalad cephalad
Instantaneous vertebral facets
axis of body
rotation FT cephalad
segment
posterior
tension
band
FN caudad FN
vertebral caudad
body facets
FT caudad
segment posterior
tension band
Fig. 44.2 (a) Three-column model in thoracic vertebrae cant kyphosing torques being applied by the anterior soft
as described by Denis. (b) Simplified model of locore- tissue mass. Compromise of the structural rigidity of the
gional osseoligamentous anatomy (left) with static force anterior vertebral body without diminution of this torque
(top right) and torque diagrams (bottom right). The instan- increases the risk of anterior vertebral body wedging with
taneous axis of rotation of the vertebrae runs through the subsequent kyphosis and deformity
middle column of the vertebral body with the most signifi-
asic Science of Osteoblastic Lesion

B signaling in osteocytes [44], with decreased
Formation PTHrP-1 signaling leading to decreases in
osteoclastic activity and subsequent sclerosis of
Like all bone metastases, osteoblastic lesions local bone. In addition to PTHrP-1 signaling
form when circulating tumor cells invade the downregulation, osteoblastic lesions may
bone marrow sinusoids and extravasate into the secrete bone morphogenic proteins 4, 6, and 7,
marrow [35]. Once in the bone marrow, the cells insulin-like growth factors 1 and 2, endothe-
release VEGF, which (1) stimulates adhesion to lin-1, and platelet-derived growth factor [38,
the surrounding matrix through upregulation of 45]. These secreted factors stimulate osteo-
adhesion molecules and (2) promotes angiogen- blasts and thereby promote bone deposition.
esis, generating a vascular supply for the cells of The osteoblasts in turn release VEGF, mono-
the nascent metastasis [36]. As lesions destined cyte chemotactic protein-1 (MCP-1), IL-6, and
to be osteoblastic progress – such as those seen in macrophage inflammatory protein-2 (MIP-2),
prostate cancer – tumor cells begin to upregulate factors that promote further metastatic cell
dickkopf-1 (DKK-1) [37], an inhibitor of the invasion and tumor growth [38].
Wnt-Frizzled pathway, which promotes bone
turnover. Then, as the lesions mature, DKK-1
expression decreases, disinhibiting the Wnt path- asic Science of Osteolytic Lesion
B
way and leading to increases in osteoblastic Formation
activity [35, 36]. This differential upregulation in
osteoblastic activity produces a sclerotic or As alluded to in the introduction to this section,
osteoblastic lesion that has higher density than the formation of osteolytic lesions is more com-
the surrounding bone. Despite this net increase in mon than the formation of osteoblastic lesions,
bone density, osteoblastic lesions have abnormal However, like osteoblastic lesions, osteolytic
bony architecture due to gross increases in both lesions rely upon the activity of both osteo-
osteoblastic and osteoclastic activity relative to blasts and osteoclasts. Stimulation of osteo-
normal bone [38]. This is akin to what is seen in blasts by cancer cells leads to the release of
Paget’s disease of bone, resulting in dense but multiple inflammatory cytokines, including
brittle bone that may be mechanically unstable MCP-1, IL-6, IL-8, MIP-2, and VEGF, all of
relative to normal bone [39] and may have which are osteoclastogenic. Activated osteo-
decreased pullout strength. Additionally, in blasts, such as those stimulated by cancer cell-
patients with both osteoblastic and osteolytic derived PTHrP, also release RANKL, which
lesions, the difference in bulk moduli between activates receptors on osteoclasts, promoting
adjacent osteoblastic and osteolytic lesions may bony resorption.
increase the risk of compression fracture within These secreted factors, among others, mediate
the osteolytic vertebra. bony destruction through both RANKL-
Previous evidence has suggested that circu- dependent and RANKL-independent mecha-
lating parathyroid hormone (PTH) levels may nisms [39]. TGF-β notably upregulates the
also influence the development of osteoblastic production of PTHrP by tumor cells [46]. PTHrP
metastases by altering the osteoblast-to-osteo- in turn activates PTH receptors on osteoblasts,
clast activity ratio [38]. Evidence for this is still stimulating RANKL release and osteoclast acti-
inconclusive though, as some groups [40] have vation [38, 46, 47]. The osteoclasts fuse into mul-
presented evidence supporting an anti-meta- tinucleated cells and form a ruffled border that
static role for PTH, whereas others have sug- releases H+ and cathepsin K to degrade the sur-
gested it to promote the formation of rounding bone [47, 48]. TGF-β also stimulates
osteoblastic [41] and/or osteolytic lesions [42, metastatic cells to release IL-8 and IL-11, which
43]. The reason for this heterogeneity of results increases osteoclast formation in a RANKL-
may be tied to differences in PTHrP-1 receptor independent fashion [39, 47, 49]. The aforemen-
tioned mechanisms beget a vicious cycle, as predisposes to compression fractures, and in

increased bone resorption stimulates osteoblast many cases, one or more segments may have
differentiation, in turn promoting further RANKL undergone such trauma by the time the patient
release and osteoclast activation. This leads to comes to clinical attention. Underlying the pro-
formation of overt lesions and also promotes pensity to suffer compression fractures is the
release of TGF-β and other growth factors as preference of metastatic cells to lodge in the vas-
bone matrix is progressively resorbed [47]. The cular trabecular bone. Trabecular bone micro-
elevated local TGF-β inhibits osteoblast differen- structure is essential for bone to resist repetitive
tiation and promotes progression of the osteolytic axial compression loads [58]. Consequently, its
metastasis [38]. destruction – seen in osteoporosis as gross bony
Tumor cells may also directly release RANKL loss and metastatic disease as tumor-initiated
through suppression of the Wnt/β-catenin signal- resorption of local bone – diverts forces into the
ing pathway, as seen in models of multiple cortical shell (now responsible for up to 97% of
myeloma that overexpress DKK-1 and scleros- normal load forces) [59] with a disproportionate
tin – two negative regulators of this pathway [50]. decrease in the ability of the vertebral body to
As with osteoblast-derived RANKL, myeloma- resist axial loading [58]. For this reason, com-
derived RANKL increases bone resorption and pression fractures are also known as trabecular
turnover, giving rise to an osteolytic lesion [36, fractures [58].
51, 52]. The key role of RANKL in progression Studies of human cadaveric vertebrae have
of these and other osteolytic lesions has been also demonstrated that uniform bone loss is
confirmed clinically through the administration associated with significantly higher levels of
of denosumab, an anti-RANKL antibody [53, weakening in compression [58]. Vertebrae of
54]. Because it addresses a key step in the major equivalent density but with different levels of
pathway of osteolysis, denosumab use may even intervertebral heterogeneity will display distinct
be superior to bisphosphonates [55] – the current yield strengths, with higher heterogeneity being
standard of care – in preventing vertebral com- associated with higher yield strength [58]. As
pression fractures in involved segments [56]. applied to metastatic vertebrae, this suggests that
mixed osteoblastic/osteolytic lesions will be dis-
proportionately stronger than purely lytic lesions
Lessons from the Osteoporotic of equivalent bone mineral density.
Spine Given that all subunits of the bone then deter-
mine fracture risk in aggregate, it is also germane
The majority of systems for classifying spinal to consider the changes that have occurred in the
instability to date have been established to healthy bone of patients with metastatic disease,
describe traumatic injuries [57]. These injuries as this provides the bulk of mechanical strength
involve a combination of damage to the bony and in the affected vertebra. To be succinct, this
soft tissue elements of the spine. By contrast, healthy bone is often compromised, as most
metastatic disease rarely affects the soft tissues; patients with metastatic spinal disease are in the
the ligaments, muscles, and cartilaginous ele- sixth decade or beyond [60], and so this healthy
ments of the spine are seldom involved. Because bone has already begun to undergo changes char-
of this, the biomechanics of the metastatic spine acteristic of the aging process. This includes
are highly similar to those of the osteoporotic depletion of trabecular bone and cortical thinning
spine, which is also characterized by almost [61], which decrease fracture toughness and
exclusively bony degradation. In both osteolytic compressive strength by 10% and 2% per decade,
lesions and osteoporotic vertebrae, there is a respectively [58, 62]. The increased fracture risk
gross decrease in spinal bone mineral density is also attributable to denaturation of bone colla-
with significant involvement of the vertebral gen [63], increases in cortical bone porosity [64,
body. This decrease in vertebral body integrity 65], and micro-damage to the bone [66].
Degradation of collagen fibrils and decreased mise – in cases of posterior wall blowout [71].
heterogeneity of collagen fibril orientation may The propensity for blowout lesions is directly
also impair the ability of the collagen network to correlated to increasing cellular content within
disperse energy applied to the vertebral body the lesion, as this is negatively correlated with
[67]. This decreases the elastic modulus of the tumor bulk modulus [71]. Weakening of the cor-
healthy bone component, thereby increasing the tical bone also increases the chances of a blowout
risk of bony fracture in response to non-axial injury. Such structural changes may already be
forces, such as may be experienced during ambu- occurring in this population secondary to nor-
lation [68–70]. mal aging [64, 65] and are liable to be further
Also gleaned from the osteoporosis literature compromised by tumoral involvement of cortical
is the fact that loads are distributed unevenly over bone, though this is uncommon in most non-lung
the vertebral body itself. With normal aging, a primary pathologies [72].
greater proportion of axial compressive loads are
shifted to the posterior half of the vertebral body
with greater dependence upon cortical bone and Determination of Mechanical
the paracortical trabecular bone [58]. This sug- Instability
gests that metastatic segments with solely ante-
rior column involvement have greater intrinsic Animal and Cadaveric Work
stability than do equivalently sized lesions
involving the middle column, regardless of poste- Several biomechanical studies have been per-
rior element involvement. formed looking at structural instability in
cadaver and animal models. One of the first
series was described by Silva et al., who used
nique Features of the Metastatic
U cadaveric thoracic vertebrae to test axial-flexion
Spine loading in simulated transcortical defects [73].
They found bicortical involvement significantly
In early stages of metastatic disease, the vertebral decreased failure loads, but unlike later series
lesions remain small and biomechanically they and the contemporary series of McGowan et al.
may be considered reasonably akin to osteopo- [74], Silva and colleagues failed to document
rotic bone due to their preferential destruction of an influence of tumor size on failure strength.
trabecular bone. However, as the lesion evolves, Dimar et al. used an essentially identical model
it progressively destroys trabeculae in the sur- to demonstrate that compressive strength was
rounding bone, carving out a cavity completely determined by the interaction of bone min-
devoid of normal bony architecture. These osteo- eral density and the proportion of the vertebral
lytic lesions act very much like an incompress- cross-section affected by the bony destruction,
ible semisolid [71], and so vertebrae involved by suggesting that patient age and vertebral lesion
these lesions may be thought of as a soft-boiled size are the best determinants of instability [75,
egg. The structural integrity of these lesions 76]. Whyne et al. also used cadaveric vertebrae
is solely dependent upon unaffected cortical with simulated osteolytic defects to test a com-
bone, and axial pressures applied to the endplate puterized model of mechanical stability of lum-
are diverted to the lateral cortical walls by the bar vertebrae secondary to tumoral involvement
tumoral mass. Axial loading pressurizes the ver- [77]. Across all variables considered, they found
tebral body contents, causing the incompressible that tumor size was the most important predictor
medullary soft tissue mass to deform, redirecting of instability, though overall bone density, and
force into the surrounding vertebral cortex. These the magnitude of axial loading were also sig-
laterally displaced forces can blowout the verte- nificant predictors of instability. This decrease
bral sidewalls and produce wedging – in cases of in axial loading strength had been previously
anterior wall blowout – or spinal canal compro- demonstrated by Windhagen et al. [78] to pre-
dict mechanical stability in involved vertebral CT can be used to evaluate both size and loca-
segments. tion of the tumor within the vertebra. Tumor size
Ebihara presented the first animal model of on CT has been shown to be the most important
simulated osteolytic metastatic spine involve- predictor of metastatic spine instability [71].
ment by generating trabecular and/or cortical Additionally, finite element analysis has demon-
defects in fresh ovine thoracic vertebrae using a strated that posterior displacement of the tumor
high-speed burr [79]. They found that lesion size within the vertebral body increases the risk of
had a significant negative correlation with failure burst fracture with subsequent canal compromise
load upon axial compression. Additionally, in [71]. By contrast, displacement into the anterior
lesions involving greater than 40% of the verte- column increases the risk for compression frac-
bral body, concomitant involvement of the costo- ture and subsequent wedging with de novo
vertebral joint was independently associated with kyphosis [71].
a decrease in failure strength, demonstrating the It has been suggested that MR has higher sen-
rib cage to significantly contribute to stabiliza- sitivity and diagnostic accuracy than multidetec-
tion of the metastatic spine. The same year, Hong tor CT for the identification of osseous metastases
et al. used simulated lytic lesions in whale verte- [84]. Use of this modality relies upon unenhanced
brae to demonstrate that the strength of the patho- T1-weighted and STIR sequences. Bony metas-
logic vertebrae is set by the weakest cross-section tases are generally T1-hypointense and demon-
through the vertebrae [80]. strate increased STIR signal due to low fat
As computing power has progressed, comput- content relative to surrounding marrow. Lesions
erized modeling software has been used to per- also frequently enhance on gadolinium-enhanced
form finite element analysis of mechanical T1-weighted lesions due to high vascularity [84].
instability in simulated vertebrae. Tschirhart Yet MR does not provide evaluation of the qual-
et al. used this model to demonstrate that lesion ity of osseous invasion by the tumor, that is,
location and tumoral morphology best predict whether the tumor results in osteoblast or
failure method, with upper thoracic location and osteoclast-dominated changes. Consequently,
bicortical involvement decreasing the risk of MR may be useful for initial identification of
burst fracture [81]. osseous lesions, but CT provides an overall better
assessment of potential instability.
Osteoblastic Versus Osteolytic

Lesions and CT Imaging SINS Framework
Currently, computed tomography (CT) imaging The focus of this chapter is correction of defor-
is considered the gold standard for noninvasive mity; however, the goal of care in the patient with
assessment of spinal instability [71]. It classifies metastatic disease of the spine is to identify at-
lesions as osteolytic or osteoblastic depending risk vertebrae prior to the onset of deformity.
upon whether they are characterized by Vertebrae at risk for deformation are termed
increased or decreased radiolucency, respec- mechanically unstable and are far more common
tively. CT imaging demonstrating radiolucent or than vertebrae which have undergone pathologic
osteolytic lesions has been correlated with sig- collapse – the inciting event for metastasis-
nificant decreases in bone density [82]. related spinal deformity. Diagnosis of mechani-
Additionally, mathematical modeling using CT cal instability relies on a combination of
images from healthy patients and those with radiographic findings and clinical presentation;
vertebral metastases has demonstrated that patients typically present with complaints of
thresholds for osteoblastic and osteolytic lesions axial (with or without radicular) pain that is
can be generated allowing them to be classified worsened with activity and loading of the spine.
quantitatively [83]. Mechanically unstable segments also typically
demonstrate extensive osteolysis of the vertebral ing management plan formulation for potentially
body, occasionally with involvement of the pedi- unstable lesions without associated neural ele-
cles and rarely the posterior tension band. ment compression. Consequently, the SINS
Because of the relative commonness of mechani- score, while valuable in presenting a standard-
cal instability in the metastatic spine – pathologic ized means of assessing mechanical stability,
fractures occur in 10–30% of all cancer patients cannot be employed as a definitive decision-
[34] – extensive work has been put into develop- making tool; ultimately, the decision of when and
ing methods for identifying and classifying the how to intervene must be made based upon the
mechanical stability of involved vertebral seg- experience and clinical acumen of the treating
ments. The most widely used system is the Spinal physician.
Instability Neoplastic Score created by the Spinal
Oncology Study Group [85]. The system scores
lesions based upon the location of the metastasis, Interventions for Mechanical
quality and presence of pain associated with the Instability
lesion, the quality of the bone in the affected ver-
tebra, the gross alignment at that segment, the Stabilization
degree of vertebral body involvement, and the
degree of posterolateral element involvement Instrumented fusion is one of the oldest interven-
(Table 44.1). Based upon these factors, lesions tions for spinal metastases and is considered the
are identified as stable (0–6), unstable (13–18), gold standard for treatment of these pathologies
or potentially unstable (7–12) and recommended owing to class I evidence demonstrating improved
for conservative management (stable lesions) or survival in patients receiving this intervention
stabilization procedures – cement or pedicle [25]. It is also the only intervention capable of
screw augmentation (unstable lesions). Several correcting significant deformity secondary to
studies have been performed demonstrating osteolytic disease. Current constructs typically
interobserver reliability [86–89]; however, there employ pedicle screw instrumentation with
still remains uncertainty among providers regard- anchors placed two levels above and below the
Table 44.1 The Spinal Instability Neoplastic Score [87]

Metric Score Metric Score
Location Radiographic spinal alignment
Junctional (O-C2, C7-T2, T11-L1, L5-S1) 3 Subluxation/translation present 4
Mobile (C3–6, L2–4) 2 De novo deformity (kyphosis/scoliosis) 2
Semirigid (T3–10) 1 Normal alignment 0
Rigid (S2-S5) 0
Pain quality Vertebral body collapse
Mechanical 3 >50% collapse 3
Oncologic/nonmechanical 1 <50% collapse 2
Pain-free lesion 0 No collapse with >50% body involved 1
None of the above 0
Bone lesion quality Posterolateral involvement (pedicles, laminae,
costovertebral joints)
Lytic 2 Bilateral 3
Mixed (lytic/blastic) 1 Unilateral 1
Blastic 0 None of the above 0
Total
Stable 0–6
Potentially unstable 7–12
Unstable 13–18
lesion, or in the case of concomitant corpectomy, tancy greater than 3 months [16, 21, 25, 26,
three levels above and below the lesion. As part 102–109] and have intractable pain [107, 110,
of preoperative evaluation, computed tomogra- 111], spinal instability [25, 110, 112], or meta-
phy scans should be acquired of the locoregional static epidural spinal cord compression
spine as multi-ostotic disease is common. In (MESCC) that is causing progressive neurologi-
cases with adjacent segment involvement, screw cal dysfunction [12, 14, 15, 17, 21, 103, 107,
purchase becomes questionable as the pullout 108, 110, 111, 113–115]. Patients with expected
strength of tumor is substantially less than that of survival less than 3 months without acute-onset
normal, healthy bone. In many cases, the bone metastasis-related neurological dysfunction
integrity may be so compromised as to preclude should be considered nonsurgical and recom-
screw placement at this level. If the purchase is mended for cement augmentation (in the case of
only questionable in the posterior elements, ante- mechanical instability) with or without concom-
rior constructs may be employed, especially if itant focused radiotherapy. Patients with acute-
corpectomy and decompression are indicated. onset neurological symptoms with limited
Said constructs usually bracket the involved survival or poor performance status should be
level, only extending one segment above and recommended for less invasive surgical tech-
below the metastatic vertebra. However, anterior niques. For example, these patients may need
approaches cannot be reasonably adopted in only a minor decompression with percutaneous
patients with intraperitoneal involvement. pedicle screw fixation to address associated
Given the similarities between osteolytic and instability. Rapid procedures like this separation
osteoporotic bone, techniques for improving surgery minimize recovery times and allow
screw pullout strength in patients operated for resumption of other adjuvant therapies [116].
metastatic disease can be borrowed from the But in cases of gross malalignment where
osteoporosis literature. These include proper tri- patients have good prognoses, limited deformity
angulation of pedicle screws [90–92], larger correction may be indicated. No guidelines or
diameter screws [93], bicortical purchase [94], series exist to describe the optimal alignment
cement augmentation [95, 96], use of an expand- for these patients, yet given the greater frailty of
able screw design [96], and decreasing pilot hole these patients and generally poorer- quality
diameter [97]. Placing screws without tapping bone, we recommend less aggressive correc-
may also increase pullout strength [98, 99] tions (i.e., SVA >5 cm). Persistent, mild
though it may also decrease screw placement malalignment is likely to make only minor con-
accuracy, especially in the thoracic spine [100]. tributions to the patient’s quality of life relative
In the past decade, cannulated screws have to their systemic disease as compared to instru-
become available that allow for cement fixation mentation failure that could occur secondary to
after screw placement, offering the opportunity overly aggressive deformity correction. The lat-
to revise screw placement intraoperatively prior ter may significantly impair these patients and is
to reinforcement. However, there is some evi- realistically an unacceptable risk given the pos-
dence that the pullout strength of these implants sibility that patients may be too moribund at the
is reduced relative to solid screws placed in time of failure to undergo a revision procedure.
tapped screw tracts pre-filled with cement [95].
In cases where solid screws are placed, cement
volumes of 1 mL in thoracic spine and 3 mL in Case Example
lumbar spine have been shown to be safe [101].
Because of the difficulties associated with A 73-year-old woman with history of multiple
surgery on the metastatic spine, along with the myeloma presented to the clinic of the senior
overall high morbidity of this patient popula- author with severe mechanical back pain of
tion, surgery is reserved for only select patients. greater than 6 months duration localizing the
In general, surgical candidates have a life expec- apex of his thoracic spine. The patient had
a b d e
Fig. 44.3 A 73-year-old woman presented with multi- PI = 78.32°, PI – LL = 6.90°) though a high degree of pel-
ostotic multiple myeloma and mechanical pain localizing vic tilt (30.56°) and large thoracic kyphosis (T1–
to the mid-thoracic spine. Preoperative imaging demon- 12 = 55.91°) were noted secondary to a focal kyphosis at
strated extensive destruction of the T7 vertebra on CT the fractured level (T6–8 = 41.44°). Postoperative films
with 80% vertebral height loss (a), bipedicular involve- (e) demonstrated similar overall coronal (CVA = 0 cm)
ment (b), and epidural disease without abutment or com- and sagittal alignment (SVA = −0.54 cm) with improve-
pression of the spinal cord at that level (c). Standing films ment of the pelvic tilt (26.80°) and thoracic kyphosis
(d) demonstrated grossly normal coronal (CVA = 0 cm) (T1–12 = 49.25°), following correction of the focal
and sagittal alignment (SVA = 0.33 cm, LL = 71.42°, kyphosis (T6–8 = 24.27°)
received both chemotherapy and radiation for his screw pullout strength as she had diffuse osteope-
myeloma. Radiographs demonstrated a notable nia throughout the thoracic spine. The patient had
kyphosis at T7 (41.44°) secondary to metastatic an uneventful inpatient stay and stayed in an
involvement (SINS score = 14; unstable) inpatient rehabilitation unit for 4 days before
(Fig. 44.3a–d) along with extensive destruction being discharged to a subacute rehabilitation
of the C3 vertebral, though the pain associated facility. At 3 months postoperative, the patient
with this lesion was nonmechanical (SINS score had significant improvement in her thoracic spine
11; potentially unstable). He was neurologically alignment (Fig. 44.3e) and complete relief of her
intact and scheduled for surgical intervention. mid-thoracic spine pain.
The patient underwent a staged surgery with
independent C2-T2 and T5–9 instrumented
fusions. Facet-based osteotomies (Schwab 1) Cement Augmentation:
were performed at C7/T1 and T1/2 to maintain Vertebroplasty and Kyphoplasty
cervical lordosis, as well as at T5/6, T6/7, T7/8,
and T8/9 to reduce the thoracic kyphosis. The For patients too ill for surgical intervention or
collapsed T7 vertebral body was corpectomized unwilling to accept the morbidity of surgical
via a bilateral posterolateral approach, and a tita- intervention, vertebral body augmentation – ver-
nium cage was placed in the vertebrectomy site tebroplasty and kyphoplasty – may be an option
both to provide anterior and middle column sup- to reinforce unstable segments. It is performed
port and to reduce the focal thoracic kyphosis. percutaneously and can be executed on an outpa-
Cannulated screws were employed at T5, T6, T8, tient basis, meaning that patients need not stop
and T9, and roughly 1 mL of polymethylmethac- their adjuvant chemotherapy regimens to undergo
rylate cement was placed bilaterally to increase these procedures. Ideal candidates are patients
with predominately mechanical axial pain tion with kyphoplasty is 3–8° per level though the
relieved in recumbency, who are neurologically durability of this correction is unclear.
intact and have no evidence of (1) compromise of
the posterior vertebral body cortex or (2) epidural
cord compression [117, 118]. In these patients, rophylaxis Against Mechanical
P
pain relief has been reported in ≥80% Instability
[118–134].
Models of cementoplasty for vertebral Radiation
metastases have demonstrated its ability to sta-
bilize the metastatic spine. Recent finite ele- Radiation is a commonly employed interven-
ment analysis by Berton et al. demonstrated tion for the treatment of symptomatic spinal
that prophylactic vertebroplasty was able to metastases. It can lead to good local control
completely prevent vertebral height collapse and relief of pain in the majority of patients
and circumferential bulging that occur second- [145–147], and for those with epidural com-
ary to axial compression of a vertebral body pression without acute neurological findings or
[135]. The augmented vertebrae were also mechanical instability, it is the treatment of
found to increase forces exerted on the endplate choice. However, radiation cannot be used to
of adjacent vertebrae in the osteoporotic model remedy mechanical instability, limiting its role
though, suggesting that vertebroplasty may to prophylaxis.
increase the risk of adjacent segment break- Though radiation provides good local tumor
down in those with osteoporotic spines. Studies control, it is not without costs though. In addi-
in cadaveric spine have echoed these results tion to damage to non-oncologic tissues
[136, 137], and other computerized models involved in the treatment field, radiation
have suggested that posterior placement of the destroys collagen fibers within the irradiated
cement within the vertebral body decreases the bone [71]. These fibers help impart much of the
risk of burst fracture [138]. tensile strength to bone that is responsible for
Contraindications to cementoplasty include a its durability. Their destruction then increases
history of coagulation disorders, significant neu- the fracture susceptibility of the irradiated
ral element compression, and complete or near- bone. Previous animal work finds this effect to
complete vertebral body collapse [139], though be greatest for hypofractionated regimens,
recent evidence suggests that it may be possible which decrease maximum axial loading in
to safely treat patients with either vertebral body rodent long bones without changing the bone
collapse [140] or posterior cortex compromise mineral content, consistent with collagenous
[141]. Vertebroplasty is also of limited utility in damage [148]. Radiation may also catalyze the
patients with focal kyphosis. In these patients, formation of pathological cross-links between
balloon kyphoplasty may be able to provide some adjacent collagen fibrils that prevent remodel-
correction or at least stabilization of the defor- ing and increase bone brittleness, predisposing
mity. Kyphoplasty balloons increase vertebral bone to fracture [149]. The most recent
body height [135] and in doing so create a cavity research though has suggested that non-
in the vertebral body capable of receiving the collagenous injury may also characterize irra-
injected cement. The vertebral height correction diated bone, with decreases in both trabecular
addresses the deformity, and the formation of a bone density and cortical bone thickness being
receiving cavity may decrease the likelihood of noted in the limbs of irradiated rodents [150–
cement leakage from the vertebral body [142, 152]. Many of these changes may be reversible
143], which is the precipitating event of the most with ambulation and other weight-bearing
significant complications of vertebral augmenta- exercises, though evidence is currently limited
tion [133], namely, pulmonary embolism and [150]. Lastly, there is some suggestion that
nerve compression [144]. The reported correc- even focal irradiation may produce systemic
decreases in bone mineral density [152], sup- Bisphosphonates, including zoledronic acid/
porting the notion that irradiation of one spine zoledronate, ibandronate, risedronate, and alen-
metastasis may destabilize other, remote bony dronate, function to inhibit metastasis in a com-
areas. pletely different fashion. After absorption into the
systemic circulation, these pyrophosphate ana-
logs [155] enter bone matrix and bind to hydroxy-
Anti-osteolysis Drugs: apatite within the matrix [47]. The bisphosphonate
The Bisphosphonates is then phagocytosed by osteoclasts along with
and Denosumab the bony matrix. Once inside osteoclasts, they
bind to farnesyl diphosphate synthase, a key regu-
As stated numerous time in this chapter, one of lator of cholesterol synthesis and protein prenyl-
the goals with metastatic disease of the spine is ation, resulting in improper intracellular protein
to stop or delay mechanical destabilization of the localization and osteoclast apoptosis [155]. They
vertebral column prior to the onset of de novo may also promote accumulation of the ATP ana-
deformity. Radiation, while relatively focal in logue triphosphoric acid 1-adenosin-5-yl ester
the context of SBRT, fails to address the underly- 3-(3-methylbut-3-enyl ester), which inhibits the
ing metabolic changes mediating osteolysis in mitochondrial ADP/ATP translocase, impairing
the affected segments. For this reason, it fails to cellular metabolism and inducing apoptosis.
treat any lesions not within the irradiated field, Combined, these proapoptotic effects have been
and in many cases, widespread irradiation is an demonstrated to reduce progressive osteolysis and
unrealistic option due to the side effects caused formation of bony metastases [155]. Additionally,
by irradiation of healthy tissues. A commonly recent systematic reviews demonstrated that they
used alternative is anti-osteolytic class of medi- may improve survival [156, 157] and lower
cations, namely, denosumab (an anti-RANKL skeletal-
related events (e.g., compression frac-
ligand monoclonal antibody) and the bisphos- ture) [157] in select populations. Most of the stud-
phonates (a group of small phosphorous-based ies considered focused on zoledronic acid
salts that encourage osteoclast death). The two administration, which is the most potent of the
medications work through complementary path- nitrogen-containing bisphosphonates and is cur-
ways. Denosumab (trade names Xgeva® and rently standard of care for skeletal-related event
Prolia®) increases the osteoprotegerin- to- prophylaxis in patients with metastatic disease
RANKL ratio, thus favoring osteoblastic over (4 mg q3–4wk) [155].
osteoclastic activity and stalling progressive
osteolysis. Research in women with aging-
related osteoporosis has demonstrated that it Conclusions
decreases the risk of vertebral compression frac-
ture by over threefold [153]. Though the response Metastatic involvement of the vertebral column –
in patients with osteolytic lesions is likely to be most notably osteolytic lesions – is characterized
less owing to accessory RANKL- independent by progressive destabilization that can result in
mechanisms of bone resorption, it has also been debilitating de novo spinal deformity. As with
demonstrated to reduce the rate of skeletal events neoplastic disease itself, the best intervention is
in this population [53, 56] presumably through prophylaxis against the formation of osteolytic
an attenuation of bone resorption. Class I evi- lesions using adequate systemic therapy with
dence evaluating its effect on patient survival concomitant bisphosphonates or denosumab
have recently been published (NCT01077154), administration. Once formed, lesions lead to
demonstrating no influence on disease recur- decreases in axial loading and sheer strength of
rence or overall survival in patients already being involved vertebrae secondary to destruction of
treated with standard of care locoregional and both trabecular and cortical bone. When stresses
systemic therapies [154]. exceed the strength of these segments, compres-
sion and/or burst fractures develop that can gener- age, cancer stage, and grade, the United States,
ate de novo deformity. The most effective means 2001–2007. Prostate Cancer. 2012;2012:1–8.
7. Kakhki VRD, Anvari K, Sadeghi R, Mahmoudian A,
of correcting this deformity at present is surgical Torabian-Kakhki M. Pattern and distribution of bone
intervention; though there is some suggestion that metastases in common malignant tumors. Nucl Med
kyphoplasty may adequately address minor defor- Rev Cent East Eur. 2013;16(2):66–9.
mity, especially in patients too moribund to 8. Nakamoto Y, Osman M, Wahl RL. Prevalence and
patterns of bone metastases detected with positron
undergo surgical management. Current literature emission tomography using F-18 FDG. Clin Nucl
evaluating standard alignment parameters in Med. 2003;28(4):302–7.
patients with metastatic spine disease is sparse 9. Fornasier VL, Horne JG. Metastases to the vertebral
(e.g., sagittal vertical axis and lumbar lordosis- column. Cancer. 1975;36(2):590–4.
10. Togawa D, Lewandrowski K. The pathophysiology
pelvic incidence mismatch), given the historical of spinal metastases. In: RF ML, Lew RK, Markman
poor survival of patients with metastatic disease M, Bukowski RM, Macklis R, et al., editors. Cancer
and subsequent reluctance to perform significant in the spine: comprehensive care. Totowa, NJ:
deformity correction in these patients. However, Humana Press; 2006. p. 17–23.
11. Tokuhashi Y, Matsuzaki H, Toriyama S, Kawano H,
with increasing numbers of long-term survivors, Ohsaka S. Scoring system for the preoperative eval-
this viewpoint may need to be reexamined in an uation of metastatic spine tumor prognosis. Spine
effort to reduce mechnical complications in (Phila Pa 1976). 1990;15(11):1110–3.
patients with spinal metastases who undergo sur- 12. Klimo P Jr, Thompson CJ, Kestle JRW, Schmidt
MH. A meta-analysis of surgery versus conventional
gical instrumentation. Accordingly, future direc- radiotherapy for the treatment of metastatic spinal
tions should focus on establishing interventions to epidural disease. Neuro-Oncology. 2005;7(1):64–76.
prevent spinal destabilization and deformity, sur- 13. Sioutos PJ, Arbit E, Meshulam CF, Galicich
gical strategies to improve construct stability that JH. Spinal metastases from solid tumors.
Analysis of factors affecting survival. Cancer.
can overcome compromised bone, and guidelines 1995;76(8):1453–9.
describing the impact of alignment on quality of 14. Cole JS, Patchell RA. Metastatic epidural spinal cord
life in those with deformity-inducing fractures. compression. Lancet Neurol. 2008;7(5):459–66.
15. Miscusi M, Polli FM, Forcato S, Ricciardi L, Frati
A, Cimatti M, et al. Comparison of minimally inva-
sive surgery with standard open surgery for verte-
References bral thoracic metastases causing acute myelopathy
in patients with short- or mid-term life expectancy:
1. Forman-Hoffman VL, Ault KL, Anderson WL, surgical technique and early clinical results. J
Weiner JM, Stevens A, Campbell VA, et al. Disability Neurosurg Spine. 2015;22(5):518–25.
status, mortality, and leading causes of death in the 16. Smith ZA, Yang I, Gorgulho A, Raphael D, De
United States community population. Med Care. Salles, Antonio AF, Khoo LT. Emerging techniques
2015;53(4):346–54. in the minimally invasive treatment and manage-
2. Wright JD, Chen L, Tergas AI, Patankar S, Burke ment of thoracic spine tumors. J Neuro-Oncol.
WM, Hou JY, et al. Trends in relative survival for 2012;107(3):443–55.
ovarian cancer from 1975 to 2011. Obstet Gynecol. 17. Eleraky M, Papanastassiou I, Vrionis
2015;125(6):1345–52. FD. Management of metastatic spine disease. Curr
3. Alberts SR, Cervantes A, van de Velde CJH. Gastric Opin Support Palliat Care. 2010;4(3):182–8.
cancer: epidemiology, pathology and treatment. Ann 18. Zaikova O, Giercksky K, Fosså SD, Kvaløy S,
Oncol. 2003;14 Suppl 2:36. Johannesen TB, Skjeldel S. A population-based study
4. Cheng TD, Cramb SM, Baade PD, Youlden DR, of spinal metastatic disease in South-East Norway.
Nwogu C, Reid ME. The international epide- Clin Oncol (R Coll Oncol). 2009;21(10):753–9.
miology of lung cancer: latest trends, dispari- 19. Molina CA, Gokaslan ZL, Sciubba DM. A system-
ties, and tumor characteristics. J Thorac Oncol. atic review of the current role of minimally invasive
2016;11(10):1653–71. spine surgery in the management of metastatic spine
5. Verdial FC, Etzioni R, Duggan C, Anderson disease. Int J Surg Oncol. 2011;2011:598148.
BO. Demographic changes in breast cancer inci- 20. Fürstenberg CH, Wiedenhöfer B, Gerner HJ, Putz
dence, stage at diagnosis and age associated with C. The effect of early surgical treatment on recovery
population-based mammographic screening. J Surg in patients with metastatic compression of the spinal
Oncol. 2017;115(5):517–22. cord. J Bone Joint Surg Br. 2009;91B(2):240–4.
6. Li J, Djenaba JA, Soman A, Rim SH, Master 21. Kaloostian PE, Yurter A, Zadnik PL, Sciubba DM,
VA. Recent trends in prostate cancer incidence by Gokaslan ZL. Current paradigms for metastatic spi-
nal disease: an evidence-based review. Ann Surg prostate cancer to the skeleton. Prostate Cancer.
Oncol. 2014;21(1):248–62. 2013;2013:418340.
22. Loblaw DA, Laperriere NJ, Mackillop WJ. A 37. Hall CL, Daignault SD, Shah RB, Pienta KJ, Keller
population- based study of malignant spinal cord ET. Dickkopf-1 expression increases early in pros-
compression in Ontario. Clin Oncol (R Coll Oncol). tate cancer development and decreases during pro-
2003;15(4):211–7. gression from primary tumor to metastasis. Prostate.
23. Loblaw DA, Perry J, Chambers A, Laperriere 2008;68(13):1396–404.
NJ. Systematic review of the diagnosis and manage- 38. Ottewell PD. The role of osteoblasts in bone metas-
ment of malignant extradural spinal cord compres- tasis. J Bone Oncol. 2016;5(3):124–7.
sion: the cancer care Ontario practice guidelines 39. Coelho RM, Lemos JM, Alho I, Valério D,
initiative‘s neuro-oncology disease site group. J Clin Ferreira AR, Costa L, et al. Dynamic modeling of
Oncol. 2005;23(9):2028–37. bone metastasis, microenvironment and therapy:
24. Mak KS, Lee LK, Mak RH, Wang S, Pile-Spellman Integrating parathyroid hormone (PTH) effect, anti-
J, Abrahm JL, et al. Incidence and treatment pat- resorptive and anti-cancer therapy. J Theor Biol.
terns in hospitalizations for malignant spinal cord 2016;391:1–12.
compression in the United States, 1998–2006. Int J 40. Swami S, Johnson J, Bettinson LA, Kimura T, Zhu
Radiat Oncol Biol Phys. 2011;80(3):824–31. H, Albertelli MA, et al. Prevention of breast cancer
25. Patchell RA, Tibbs PA, Regine WF, Payne R, Saris skeletal metastases with parathyroid hormone. JCI
S, Kryscio RJ, et al. Direct decompressive surgical Insight. 2017;2(17).
resection in the treatment of spinal cord compres- 41. Iddon J, Bundred NJ, Hoyland J, Downey SE, Baird
sion caused by metastatic cancer: a randomised trial. P, Salter D, et al. Expression of parathyroid hormone-
Lancet. 2005;366(9486):643–8. related protein and its receptor in bone metastases
26. Quraishi NA, Gokaslan ZL, Boriani S. The sur- from prostate cancer. J Pathol. 2000;191(2):170–4.
gical management of metastatic epidural com- 42. Ritchie CK, Thomas KG, Andrews LR, Tindall
pression of the spinal cord. J Bone Joint Surg Br. DJ, Fitzpatrick LA. Effects of the calciotrophic
2010;92(8):1054–60. peptides calcitonin and parathyroid hormone on
27. Tomycz N, Gerszten P. Minimally invasive treat- prostate cancer growth and chemotaxis. Prostate.
ments for metastatic spine tumors: vertebroplasty, 1997;30(3):183–7.
kyphoplasty, and radiosurgery. Neurosurg Q. 43. Schwartz GG. Prostate cancer, serum para-
2008;18(2):104–8. thyroid hormone, and the progression of skel-
28. Witham TF, Khavkin YA, Gallia GL, Wolinsky J, etal metastases. Cancer Epidemiol Biomark Prev.
Gokaslan ZL. Surgery insight: current manage- 2008;17(3):478–83.
ment of epidural spinal cord compression from 44. Saini V, Marengi DA, Barry KJ, Fulzele KS,
metastatic spine disease. Nat Clin Pract Neurol. Heiden E, Liu X, et al. Parathyroid hormone
2006;2(2):87–94. (PTH)/PTH-related peptide type 1 receptor (PPR)
29. Denis F. Spinal instability as defined by the three- signaling in osteocytes regulates anabolic and
column spine concept in acute spinal trauma. Clin catabolic skeletal responses to PTH. J Biol Chem.
Orthop Relat Res. 1984;189:65–76. 2013;288(28):20122–34.
30. Holdsworth F. Fractures, dislocations, and fracture- 45. Ibrahim T, Flamini E, Mercatali L, Sacanna E,
dislocations of the spine. J Bone Joint Surg Am. Serra P, Amadori D. Pathogenesis of osteoblas-
1970;52(8):1534–51. tic bone metastases from prostate cancer. Cancer.
31. Azam MQ, Sadat-Ali M. The concept of evolution of 2010;116(6):1406–18.
thoracolumbar fracture classifications helps in surgi- 46. Guise TA. Molecular mechanisms of osteolytic bone
cal decisions. Asian Spine J. 2015;9(6):984–94. metastases. Cancer. 2000;88(12 Suppl):2892–8.
32. Alexandru D, So W. Evaluation and manage- 47. Chen Y, Sosnoski DM, Mastro AM. Breast cancer
ment of vertebral compression fractures. Perm J. metastasis to the bone: mechanisms of bone loss.
2012;16(4):46–51. Breast Cancer Res. 2010;12(6):215.
33. Fink HA, Milavetz DL, Palermo L, Nevitt MC, 48. Naylor K, Eastell R. Bone turnover markers: use in
Cauley JA, Genant HK, et al. What proportion of osteoporosis. Nat Rev Rheumatol. 2012;8(7):379–89.
incident radiographic vertebral deformities is clini- 49. Bendre MS, Montague DC, Peery T, Akel NS,
cally diagnosed and vice versa? J Bone Miner Res. Gaddy D, Suva LJ. Interleukin-8 stimulation of
2005;20(7):1216–22. osteoclastogenesis and bone resorption is a mecha-
34. Macedo F, Ladeira K, Pinho F, Saraiva N, Bonito N, nism for the increased osteolysis of metastatic bone
Pinto L, et al. Bone metastases: an overview. Oncol disease. Bone. 2003;33(1):28–37.
Rev. 2017;11(1):321. 50. Mariz K, Ingolf J, Daniel H, Teresa NJ, Erich-Franz
35. Suva LJ, Washam C, Nicholas RW, Griffin RJ. Bone S. The Wnt inhibitor dickkopf-1: a link between
metastasis: mechanisms and therapeutic opportuni- breast cancer and bone metastases. Clin Exp
ties. Nat Rev Endocrinol. 2011;7(4):208–18. Metastasis. 2015;32(8):857–66.
36. Roberts E, Cossigny DAF, Quan GMY. The role 51. Yavropoulou MP, van Lierop AH, Hamdy NAT,
of vascular endothelial growth factor in metastatic Rizzoli R, Papapoulos SE. Serum sclerostin levels in
Paget’s disease and prostate cancer with bone metas- 68. Martin RB, Ishida J. The relative effects of collagen
tases with a wide range of bone turnover. Bone. fiber orientation, porosity, density, and mineralization
2012;51(1):153–7. on bone strength. J Biomech. 1989;22(5):419–26.
52. Tian E, Zhan F, Walker R, Rasmussen E, Ma Y, 69. Martin RB, Boardman DL. The effects of collagen
Barlogie B, et al. The role of the Wnt-signaling fiber orientation, porosity, density, and mineraliza-
antagonist DKK1 in the development of osteo- tion on bovine cortical bone bending properties. J
lytic lesions in multiple myeloma. N Engl J Med. Biomech. 1993;26(9):1047–54.
2003;349(26):2483–94. 70. Riggs CM, Vaughan LC, Evans GP, Lanyon LE,
53. Steger GG, Bartsch R. Denosumab for the treatment Boyde A. Mechanical implications of collagen fibre
of bone metastases in breast cancer: evidence and orientation in cortical bone of the equine radius.
opinion. Ther Adv Med Oncol. 2011;3(5):233–43. Anat Embryol. 1993;187(3):239–48.
54. Yuasa T, Yamamoto S, Urakami S, Fukui I, Yonese 71. Whyne CM. Biomechanics of metastatic dis-
J. Denosumab: a new option in the treatment of bone ease in the vertebral column. Neurol Res.
metastases from urological cancers. Onco Targets 2014;36(6):493–501.
Ther. 2012;5:221–9. 72. Greenspan A, Norman A. Osteolytic cortical
55. McClung MR, Lewiecki EM, Cohen SB, Bolognese destruction: an unusual pattern of skeletal metasta-
MA, Woodson GC, Moffett AH, et al. Denosumab ses. Skelet Radiol. 1988;17(6):402–6.
in postmenopausal women with low bone mineral 73. Silva MJ, Hipp JA, McGowan DP, Takeuchi T, Hayes
density. N Engl J Med. 2006;354(8):821–31. WC. Strength reductions of thoracic vertebrae in the
56. Gül G, Sendur MAN, Aksoy S, Sever AR, Altundag presence of transcortical osseous defects: effects of
K. A comprehensive review of denosumab for bone defect location, pedicle disruption, and defect size.
metastasis in patients with solid tumors. Curr Med Eur Spine J. 1993;2(3):118–25.
Res Opin. 2016;32(1):133–45. 74. McGowan DP, Hipp JA, Takeuchi T, White AA,
57. Filis AK, Aghayev KV, Doulgeris JJ, Gonzalez- Hayes WC. Strength reductions from trabecu-
Blohm SA, Vrionis FD. Spinal neoplastic instabil- lar destruction within thoracic vertebrae. J Spinal
ity: biomechanics and current management options. Disord. 1993;6(2):130–6.
Cancer Control. 2014;21(2):144–50. 75. Dimar JR, Voor MJ, Zhang YM, Glassman SD. A
58. Osterhoff G, Morgan EF, Shefelbine SJ, Karim human cadaver model for determination of patho-
L, McNamara LM, Augat P. Bone mechanical logic fracture threshold resulting from tumor-
properties and changes with osteoporosis. Injury. ous destruction of the vertebral body. Spine.
2016;47(Suppl 2):11. 1998;23(11):1209–14.
59. Yang KH, King AI. Mechanism of facet load trans- 76. Mizrahi J, Silva MJ, Hayes WC. Finite element stress
mission as a hypothesis for low-back pain. Spine. analysis of simulated metastatic lesions in the lumbar
1984;9(6):557–65. vertebral body. J Biomed Eng. 1992;14(6):467–75.
60. Aebi M. Spinal metastasis in the elderly. Eur Spine J. 77. Whyne CM, Hu SS, Lotz JC. Burst fracture in
2003;12 Suppl 2:202. the metastatically involved spine: development,
61. Tong X, Burton IS, Isaksson H, Jurvelin JS, Kröger validation, and parametric analysis of a three-
H. Cortical bone histomorphometry in male femo- dimensional poroelastic finite-element model. Spine.
ral neck: the investigation of age- association 2003;28(7):652–60.
and regional differences. Calcif Tissue Int. 78. Windhagen H, Hipp JA, Hayes WC. Postfracture
2015;96(4):295–306. instability of vertebrae with simulated defects can be
62. Burstein AH, Reilly DT, Martens M. Aging of bone predicted from computed tomography data. Spine.
tissue: mechanical properties. J Bone Joint Surg Am. 2000;25(14):1775–81.
1976;58(1):82–6. 79. Ebihara H, Ito M, Abumi K, Taneichi H, Kotani
63. Wang X, Bank RA, TeKoppele JM, Agrawal CM. The Y, Minami A, et al. A biomechanical analysis of
role of collagen in determining bone mechanical metastatic vertebral collapse of the thoracic spine.
properties. J Orthop Res. 2001;19(6):1021–6. Spine (Phila Pa 1976). 2004;29(9):994–9.
64. Keaveny TM, Hayes WC. A 20-year perspective 80. Hong J, Cabe GD, Tedrow JR, Hipp JA, Snyder
on the mechanical properties of trabecular bone. J BD. Failure of trabecular bone with simulated lytic
Biomech Eng. 1993;115(4B):534–42. defects can be predicted non-invasively by structural
65. Zebaze RMD, Ghasem-Zadeh A, Bohte A, Iuliano- analysis. J Orthop Res. 2004;22(3):479–86.
Burns S, Mirams M, Price RI, et al. Intracortical 81. Tschirhart CE, Finkelstein JA, Whyne
remodelling and porosity in the distal radius and CM. Biomechanics of vertebral level, geometry,
post-mortem femurs of women: a cross-sectional and transcortical tumors in the metastatic spine. J
study. Lancet. 2010;375(9727):1729–36. Biomech. 2007;40(1):46–54.
66. Schaffler MB, Choi K, Milgrom C. Aging and 82. Vassiliou V, Kalogeropoulou C, Petsas T, Leotsinidis
matrix microdamage accumulation in human com- M, Kardamakis D. Clinical and radiological evalua-
pact bone. Bone. 1995;17(6):521–5. tion of patients with lytic, mixed and sclerotic bone
67. Nair AK, Gautieri A, Chang S, Buehler metastases from solid tumors: is there a correlation
MJ. Molecular mechanics of mineralized collagen between clinical status of patients and type of bone
fibrils in bone. Nat Commun. 2013;4:1724. metastases? Clin Exp Metastasis. 2007;24(1):49–56.
83. Whyne C, Hardisty M, Wu F, Skrinskas T, Clemons ent pedicle screw designs and augmentation tech-
M, Gordon L, et al. Quantitative characterization of niques in an osteoporotic bone model. Asian Spine
metastatic disease in the spine. Part II. Histogram- J. 2018;12(1):3–11.
based analyses. Med Phys. 2007;34(8):3279–85. 97. Chatzistergos PE, Sapkas G, Kourkoulis SK. The
84. Shah LM, Salzman KL. Imaging of spinal metastatic influence of the insertion technique on the pullout
disease. Int J Surg Oncol. 2011;2011:769753. force of pedicle screws: an experimental study.
85. Fisher CG, DiPaola CP, Ryken TC, Bilsky MH, Spine. 2010;35(9):332.
Shaffrey CI, Berven SH, et al. A novel classification 98. Chen L, Tai C, Lai P, Lee D, Tsai T, Fu T, et al.
system for spinal instability in neoplastic disease: Pullout strength for cannulated pedicle screws with
an evidence-based approach and expert consensus bone cement augmentation in severely osteoporotic
from the spine oncology study group. Spine (Phila bone: influences of radial hole and pilot hole tapping.
Pa 1976). 2010;35(22):E1229. Clin Biomech (Bristol, Avon). 2009;24(8):613–8.
86. Fourney DR, Frangou EM, Ryken TC, DiPaola CP, 99. Pfeiffer FM, Abernathie DL, Smith DE. A compari-
Shaffrey CI, Berven SH, et al. Spinal Instability son of pullout strength for pedicle screws of differ-
Neoplastic Score: an analysis of reliability and ent designs: a study using tapped and untapped pilot
validity from the spine oncology study group. J Clin holes. Spine. 2006;31(23):867.
Oncol. 2011;29(22):3072–7. 100. Erkan S, Hsu B, Wu C, Mehbod AA, Perl J,
87. Campos M, Urrutia J, Zamora T, Román J, Canessa Transfeldt EE. Alignment of pedicle screws with
V, Borghero Y, et al. The spine instability neoplastic pilot holes: can tapping improve screw trajectory in
score: an independent reliability and reproducibility thoracic spines? Eur Spine J. 2010;19(1):71–7.
analysis. Spine J. 2014;14(8):1466–9. 101. Leichtle CI, Lorenz A, Rothstock S, Happel J, Walter
88. Fisher CG, Schouten R, Versteeg AL, Boriani S, F, Shiozawa T, et al. Pull-out strength of cemented
Varga PP, Rhines LD, et al. Reliability of the Spinal solid versus fenestrated pedicle screws in osteopo-
Instability Neoplastic Score (SINS) among radiation rotic vertebrae. Bone Joint Res. 2016;5(9):419–26.
oncologists: an assessment of instability second- 102. Tokuhashi Y, Ajiro Y, Umezawa N. Outcome of

ary to spinal metastases. Radiat Oncol (London, treatment for spinal metastases using scoring sys-
England). 2014;9(1):69. tem for preoperative evaluation of prognosis. Spine
89. Fox S, Spiess M, Hnenny L, Fourney DR. Spinal (Phila Pa 1976). 2009;34(1):69–73.
Instability Neoplastic Score (SINS): reliability 103. Yang SB, Cho W, Chang U. Analysis of prog-
among spine fellows and resident physicians in nostic factors relating to postoperative survival
orthopedic surgery and neurosurgery. Global Spine in spinal metastases. J Korean Neurosurg Soc.
J. 2017;7(8):744–8. 2012;51(3):127–34.
90. Ruland CM, McAfee PC, Warden KE, Cunningham 104. Finkelstein JA, Zaveri G, Wai E, Vidmar M, Kreder
BW. Triangulation of pedicular instrumenta- HJ, Chow E. A population-based study of surgery
tion. A biomechanical analysis. Spine. 1991;16(6 for spinal metastases. Survival rates and complica-
Suppl):270. tions. J Bone Joint Surg Br. 2003;85(7):1045–50.
91. Hadjipavlou AG, Nicodemus CL, al-Hamdan FA, 105. Hosono N, Ueda T, Tamura D, Aoki Y, Yoshihawa
Simmons JW, Pope MH. Correlation of bone equiva- H. Prognostic relevance of clinical symptoms in
lent mineral density to pull-out resistance of trian- patients with spinal metastases. Clin Orthop Relat
gulated pedicle screw construct. J Spinal Disord. Res. 2005;436:196–201.
1997;10(1):12–9. 106. Laufer I, Sciubba DM, Madera M, Bydon A, Witham
92. Barber JW, Boden SD, Ganey T, Hutton TJ, Gokaslan ZL, et al. Surgical management of meta-
WC. Biomechanical study of lumbar pedicle screws: static spinal tumors. Cancer Control. 2012;19(2):122–8.
does convergence affect axial pullout strength? J 107. Sciubba D, Gokaslan Z, Suk I, Suki D, Maldaun
Spinal Disord. 1998;11(3):215–20. M, McCutcheon I, et al. Positive and negative
93. Lai D, Shih Y, Chen Y, Chien A, Wang J. Effect of prognostic variables for patients undergoing spine
pedicle screw diameter on screw fixation efficacy in surgery for metastatic breast disease. Eur Spine J.
human osteoporotic thoracic vertebrae. J Biomech. 2007;16(10):1659–67.
2018;70:196–203. 108. Sciubba D, Yurter A, Ju D, Gokaslan Z, Fisher C,
94. Ponnusamy KE, Iyer S, Gupta G, Khanna Rhines L, et al. A systematic review of clinical out-
AJ. Instrumentation of the osteoporotic spine: comes and prognostic factors for patients undergo-
biomechanical and clinical considerations. Spine ing surgery for spinal metastases secondary to breast
J. 2011;11(1):54–63. cancer. Global Spine J. 2015;5(1_suppl):1554402.
95. Chen L, Tai C, Lee D, Lai P, Lee Y, Niu C, et al. 109. Zadnik P, Hwang L, Ju D, Groves M, Sui J, Yurter A,
Pullout strength of pedicle screws with cement aug- et al. Prolonged survival following aggressive treat-
mentation in severe osteoporosis: a comparative ment for metastatic breast cancer in the spine. Clin
study between cannulated screws with cement injec- Exp Metastasis. 2014;31(1):47–55.
tion and solid screws with cement pre-filling. BMC 110. Pointillart V, Vital J, Salmi R, Diallo A, Quan
Musculoskelet Disord. 2011;12:33. GMY. Survival prognostic factors and clinical out-
96. Kiyak G, Balikci T, Heydar AM, Bezer comes in patients with spinal metastases. J Cancer
M. Comparison of the pullout strength of differ- Res Clin Oncol. 2011;137(5):849–56.
111. Tomita K, Kawahara N, Kobayashi T, Yoshida A, in cancer patients. J Neurosurg. 2003;98(Spine

Murakami H, Akamaru T. Surgical strategy for spi- 1):21–30.
nal metastases. Spine. 2001;26(3):298–306. 124. Alvarez L, Pérez-Higueras A, Quiñones D, Calvo E,
112. North RB, LaRocca VR, Schwartz J, North CA, Rossi R. Vertebroplasty in the treatment of vertebral
Zahurak M, Davis RF, et al. Surgical management tumors: postprocedural outcome and quality of life.
of spinal metastases: analysis of prognostic factors Eur Spine J. 2003;12(4):356–60.
during a 10-year experience. J Neurosurg Spine. 125. Ambrosanio G, Lavanga A, Vassallo P, Izzo

2005;2(5):564–73. R, Diano AA, Muto M. Vertebroplasty in the
113. Ju D, Zadnik P, Groves M, Hwang L, Kaloostian P, treatment of spine disease. Interv Neuroradiol.
Wolinksy J, et al. Factors associated with improved 2005;11(4):309–23.
outcomes following decompressive surgery for pros- 126. Barr JD, Barr MS, Lemley TJ, McCann
tate cancer metastatic to the spine. Neurosurgery. RM. Percutaneous vertebroplasty for pain relief and
2013;73(4):657–66. spinal stabilization. Spine. 2000;25(8):923–8.
114. Katagiri H, Okada R, Takagi T, Takahashi M, Murata 127. Caudana R, Renzi Brivio L, Ventura L, Aitini E,
H, Harada H, et al. New prognostic factors and scor- Rozzanigo U, Barai G. CT-guided percutaneous ver-
ing system for patients with skeletal metastasis. tebroplasty: personal experience in the treatment of
Cancer Med. 2014;3(5):1359–67. osteoporotic fractures and dorsolumbar metastases.
115. Quraishi NA, Rajagopal TS, Manoharan SR, Elsayed Radiol Med. 2008;113(1):114–33.
S, Edwards KL, Boszczyk BM. Effect of timing of 128. Chen L, Ni R, Liu S, Liu Y, Jin Y, Zhu X, et al.
surgery on neurological outcome and survival in Percutaneous vertebroplasty as a treatment for pain-
metastatic spinal cord compression. Eur Spine J. ful osteoblastic metastatic spinal lesions. J Vasc
2013;22(6):1383–8. Interv Radiol. 2011;22(4):525–8.
116. Laufer I, Iorgulescu JB, Chapman T, Lis E, Shi 129. Kobayashi T, Arai Y, Takeuchi Y, Nakajima Y,
W, Zhang Z, et al. Local disease control for spinal Shioyama Y, Sone M, et al. Phase I/II clinical
metastases following “separation surgery” and adju- study of percutaneous vertebroplasty (PVP) as pal-
vant hypofractionated or high-dose single-fraction liation for painful malignant vertebral compression
stereotactic radiosurgery: outcome analysis in 186 fractures (PMVCF): JIVROSG-0202. Ann Oncol.
patients. J Neurosurg Spine. 2013;18(3):207. 2009;20(12):1943–7.
117. Burton AW, Rhines LD, Mendel E. Vertebroplasty 130. Lee B, Franklin I, Lewis JS, Coombes RC,
and kyphoplasty: a comprehensive review. Leonard R, Gishen P, et al. The efficacy of percu-
Neurosurg Focus. 2005;18(3):e1. taneous vertebroplasty for vertebral metastases
118. Pizzoli AL, Brivio LR, Caudana R, Vittorini associated with solid malignancies. Eur J Cancer.
E. Percutaneous CT-guided vertebroplasty in the 2009;45(9):1597–602.
management of osteoporotic fractures and dor- 131. Nirala AP, Vatsal DK, Husain M, Gupta C, Chawla
solumbar metastases. Orthop Clin North Am. J, Kumar V, et al. Percutaneous vertebroplasty:
2009;40(4):458, vii. an experience of 31 procedures. Neurol India.
119. Berenson J, Pflugmacher R, Jarzem P, Zonder 2003;51(4):490–2.
J, Schechtmna K, Tillman JB, et al. Balloon 132. Sun G, Cong Y, Xie Z, Jin P, Li F, Yi Y, et al.
kyphoplasty versus non-surgical fracture man- Percutaneous vertebroplasty using instruments and
agement for treatment of painful vertebral body drugs made in China for vertebral metastases. Chin
compression fractures in patients with cancer: a mul- Med J. 2003;116(8):1207–12.
ticentre, randomised controlled trial. Lancet Oncol. 133. Sun G, Li L, Jin P, Liu X, Li M. Percutaneous verte-
2011;12(3):225–35. broplasty for painful spinal metastasis with epidural
120. Dalbayrak S, Onen MR, Yilmaz M, Naderi S. Clinical encroachment. J Surg Oncol. 2014;110(2):123–8.
and radiographic results of balloon kyphoplasty for 134. Xie P, Zhao Y, Li G. Efficacy of percutaneous ver-
treatment of vertebral body metastases and multiple tebroplasty in patients with painful vertebral metas-
myelomas. J Clin Neurosci. 2010;17(2):219–24. tases: a retrospective study in 47 cases. Clin Neurol
121. Pflugmacher R, Beth P, Schroeder R, Schaser K, Neurosurg. 2015;138:157–61.
Melcher I. Balloon kyphoplasty for the treatment 135. Berton A, Salvatore G, Giambini H, Ciuffreda M,
of pathological fractures in the thoracic and lum- Longo UG, Denaro V, et al. A 3D finite element
bar spine caused by metastasis: one-year follow-up. model of prophylactic vertebroplasty in the meta-
Acta Radiol. 2007;48(1):89–95. static spine: vertebral stability and stress distribu-
122. Qian Z, Sun Z, Yang H, Gu Y, Chen K, Wu tion on adjacent vertebrae. J Spinal Cord Med.
G. Kyphoplasty for the treatment of malignant ver- 2020;43(1):39–45.
tebral compression fractures caused by metastases. J 136. Oakland RJ, Furtado NR, Timothy J, Hall RM. The
Clin Neurosci. 2011;18(6):763–7. biomechanics of vertebroplasty in multiple
123. Fourney DR, Schomer DF, Nader R, Chlan-Fourney myeloma and metastatic bladder cancer: a prelimi-
J, Suki D, Ahrar K, et al. Percutaneous vertebroplasty nary cadaveric investigation. J Neurosurg Spine.
and kyphoplasty for painful vertebral body fractures 2008;9(5):493–501.
137. Oakland RJ, Furtado NR, Timothy J, Hall RM. A are the options, indications, and outcomes? Spine.
preliminary cadaveric study investigating the bio- 2009;34(22 Suppl):78.
mechanical effectiveness of vertebroplasty in treat- 148. Nyaruba MM, Yamamoto I, Kimura H, Morita
ing spinal metastases and multiple myeloma. Orthop R. Bone fragility induced by X-ray irradiation in
Proc. 2009;91-B(SUPP_III):497. relation to cortical bone-mineral content. Acta
138. Tschirhart CE, Roth SE, Whyne CM. Biomechanical Radiol. 1998;39:43–6.
assessment of stability in the metastatic spine fol- 149. Gong B, Oest ME, Mann KA, Damron TA, Morris
lowing percutaneous vertebroplasty: effects of MD. Raman spectroscopy demonstrates pro-
cement distribution patterns and volume. J Biomech. longed alteration of bone chemical composition
2005;38(8):1582–90. following extremity localized irradiation. Bone.
139. Hide IG, Gangi A. Percutaneous vertebroplasty: 2013;57(1):252–8.
history, technique and current perspectives. Clin 150. Govey PM, Zhang Y, Donahue HJ. Mechanical
Radiol. 2004;59:461–7. loading attenuates radiation-induced bone loss
140. Hentschel SJ, Rhines LD, Shah HN, Burton AW, in bone marrow transplanted mice. PLoS One.
Mendel E. Percutaneous vertebroplasty in vertebra 2016;11(12):e0167673.
plana secondary to metastasis. J Spinal Disord Tech. 151. Oest ME, Policastro CG, Mann KA, Zimmerman
2004;17(6):554–7. ND, Damron TA. Longitudinal effects of single
141. Amoretti N, Diego P, Amélie P, Andreani O, Foti P, hindlimb radiation therapy on bone strength and
Schmid-Antomarchi H, et al. Percutaneous vertebro- morphology at local and contralateral sites. J Bone
plasty in tumoral spinal fractures with posterior ver- Miner Res. 2018;33(1):99–112.
tebral wall involvement: feasibility and safety. Eur J 152. Wright LE, Buijs JT, Kim H, Coats LE, Scheidler
Radiol. 2018;104:38–42. AM, John SK, et al. Single-limb irradiation induces
142. Eck JC, Nachtigall D, Humphreys SC, Hodges local and systemic bone loss in a murine model. J
SD. Comparison of vertebroplasty and balloon Bone Miner Res. 2015;30(7):1268–79.
kyphoplasty for treatment of vertebral compression 153. Cummings SR, San Martin J, McClung MR, Siris
fractures: a meta-analysis of the literature. Spine J. ES, Eastell R, Reid IR, et al. Denosumab for pre-
2008;8(3):488–97. vention of fractures in postmenopausal women with
143. Hulme PA, Krebs J, Ferguson SJ, Berlemann osteoporosis. N Engl J Med. 2009;361(8):756–65.
U. Vertebroplasty and kyphoplasty: a sys- 154. Coleman RE, Finkelstein D, Barrios CH, Martin
tematic review of 69 clinical studies. Spine. M, Iwata H, Glaspy JA, et al. Adjuvant denosumab
2006;31(17):1983–2001. in early breast cancer: first results from the inter-
144. Siemionow K, Lieberman I. Vertebral augmenta- national multicenter randomized phase III placebo
tion in osteoporosis and bone metastasis. Curr Opin controlled D-CARE study. 2018 ASCO Annual
Support Palliat Care. 2007;1(4):323–7. Meeting 2018 June 4.
145. Gerszten PC, Ozhasoglu C, Burton SA, Vogel WJ, 155. Holen I, Coleman RE. Bisphosphonates as treat-
Atkins BA, Kalnicki S, et al. CyberKnife frame- ment of bone metastases. Curr Pharm Des.
less stereotactic radiosurgery for spinal lesions: 2010;16(11):1262–71.
clinical experience in 125 cases. Neurosurgery. 156. O’Carrigan B, Wong MH, Willson ML, Stockler
2004;55(1):99. MR, Pavlakis N, Goodwin A. Bisphosphonates
146. Gerszten PC, Burton SA, Ozhasoglu C, Welch and other bone agents for breast cancer. Cochrane
WC. Radiosurgery for spinal metastases: clinical Database Syst Rev. 2017;10:CD003474.
experience in 500 cases from a single institution. 157. Mhaskar R, Kumar A, Miladinovic B, Djulbegovic
Spine. 2007;32(2):193–9. B. Bisphosphonates in multiple myeloma: an
147. Gerszten PC, Mendel E, Yamada Y. Radiotherapy updated network meta-analysis. Cochrane Database
and radiosurgery for metastatic spine disease: what Syst Rev. 2017;12:CD003188.
Postoperative Complications
and Spinal Metastases
45
Bushra Yasin and Michael S. Virk
Background treatment team. Operative indications for meta-

static spine tumors are directed at stabilization of
The vertebral column is the leading site of skele- the vertebral column, decompression of the neu-
tal metastasis making spinal metastases the most ral elements and optimizing targets for radiation
common spinal malignancy. As many as 30% of therapy. Because these patients have metastatic
patients with solid organ malignancies will disease, such procedures are palliative with the
develop spinal metastatic disease [1], and this ultimate goal of treating pain, optimizing func-
rate continues to increase as multimodality ther- tion and ambulation, improving quality of life
apy is extending life expectancy of cancer and preserving continence of bowel and bladder.
patients. Patients frequently present with neck or Such procedures can range from vertebro- or
back pain. Pain can result from inflammatory kyphoplasty to percutaneous instrumentation for
mediators associated with tumor growth that is stabilization of the posterior tension band to open
often responsive to steroids or due to spinal procedures for separation surgery or en-bloc
mechanical instability that is provoked by axial resection with spinal reconstruction.
loading and/or movement. Pathologic fractures in Postoperative complications after spinal sur-
the vertebral bodies or posterior elements can gery have been reported to fluctuate widely in the
further contribute to pain. Tumor extension into literature, from 10% to 52% [2, 3]. Complications
the epidural compartment may cause compres- may be medical or surgical and occur intraopera-
sion of neural elements resulting in radicular tively or postoperatively. Common medical com-
pain, numbness, weakness and, in the most severe plications include deep vein thrombosis and
settings, myelopathy or cauda equina syndrome. pulmonary embolism, pneumonia, stroke, myo-
Management of spinal metastasis requires an cardial infarction, pressure ulcers, urinary tract
increasingly multidisciplinary approach. infection, sepsis or septicemia and persistent
Enhanced diagnostic imaging modalities, CT- pain. Surgical-related complications include
guided biopsy, chemotherapeutics, radiation postoperative hematoma, surgical site infection,
therapy, and spine surgery may all play a role wound dehiscence, dural tear with persistent CSF
thus incorporating five specialty services into the leak, hardware failure, and neurological injury
(Table 45.1). Risk factors for postoperative com-
B. Yasin (*) · M. S. Virk
plications are associated with advanced age, pre-
Department of Neurological Surgery, operative radiotherapy, multilevel spinal
Weill Cornell Medicine, New York Presbyterian metastasis, and the burden of comorbidities.
Hospital, New York, NY, USA Proper risk stratification involves evaluation of
e-mail: buy2002@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_45
604 B. Yasin and M. S. Virk
Table 45.1 Common postoperative complications and management

Adverse event Risk factors Management strategy
Surgical:
Hardware failure Preoperative radiation, Identify high risk patients, use of fenestrated screws,
osteopenia/osteoporosis, rib PMMA vertebral augmentation, increase number of
resection, >6 instrumented instrumented levels
segments
Durotomy and Previous surgery or irradiation, Repair primarily with autologous graft, dural substitute,
persistent CSF leak size and location of durotomy fibrin glue; place lumbar drain, subfascial epidural drain
management, positioning; local vascularized,
myocutaneous flap closures
Wound Previous surgery or irradiation, Modify preoperative risk factors, Infection prevention:
complications instrumentation, DM, smoking, Preoperative antibiotics, application of intrawound
chronic steroids, low serum vancomycin powder, treatment: Systemic antibiotics,
albumin, obesity, age, packing or wound VAC, reoperation with local
neurological disability vascularized, myocutaneous flap closures
Neurologic deficit Coagulopathy, multilevel spinal Identify etiology, imaging, critical care setting, epidural
surgery hematoma evacuation, blood pressure augmentation,
possible steroids
Intra−/postoperative Thombocytopenia, Preop risk assessment: Thrombocytopenia/coagulopathy,
hemorrhage coagulopathy, bone marrow use of anti-thrombotic, preop angiography
suppression, highly vascularized +/− embolization
tumor
Medical:
Thromboembolism: Immobility, vasculopathy, Prevention: Patient assessment for coagulation test and
PE coagulopathy reviewing medical history, placement of IVC filter,
DVT pneumatic intermittent compression and compression
stockings, postoperative pharmacological prophylaxis.
Treatment: Observation and thrombolytic therapy
LMWH. Thrombectomy in massive PE
CSF cerebrovascular fluid, PMMA polymethylmethacrylate, DM diabetes mellitus, VAC vacuum-assisted closure, PE
pulmonary embolism, DVT deep venous thrombosis, IVC inferior vena cava, LMWH low molecular weight heparin
each individual patient’s medical and surgical liver. Thrombocytopenia due to blood count
history, co morbidities, and physical exam prior nadirs following treatment with certain chemo-
to offering surgical intervention. therapeutics or bone marrow suppression second-
ary to wide-field radiation or significant
metastatic disease should be identified and rem-
Intraoperative and Postoperative edied prior to surgery. While medications that
Hemorrhage cause thrombocytopenia can often be held until
counts return, marrow suppression may ulti-
Multiple factors affecting intraoperative tumor mately be a contraindication to surgery. Bone
bleeding must be carefully considered prior to marrow biopsy may be a helpful diagnostic
proceeding with surgical management of meta- adjunct to determine the etiology, severity, and
static spine tumors. Among these are thrombocy- implications of failed synthesis.
topenia, coagulopathy, bone marrow suppression, Certain metastatic tumors pose increased risk
and factor deficiencies. These same risk factors of intraoperative hemorrhage that may prove
contribute to the likelihood of postoperative challenging to control or result in high blood loss
hematomas. Effectively addressing these con- during surgery. Tumor histologies incorporating
cerns requires a multidisciplinary approach with “angio” or originating from vascular organs such
hematology and oncology among others. as the thyroid, liver, or kidney are often highly
Coagulopathies can result from factor deficien- vascular. Performing preoperative digital
cies, clotting disorders, hepatocellular carci- subtraction angiography assists with determining
noma, or large metastatic tumor burden in the the extent of tumor vascularization, identifying
45 Postoperative Complications and Spinal Metastases 605
vascular anatomy including significant feeding d iabetes, smoking, systemic therapy (e.g., ste-
vessels, and potentially embolizing the tumor to roids, immunosuppressive adjuvant therapy),
decrease intraoperative hemorrhage [4, 5]. neutropenia, low serum albumin level, higher
Tumors with deep contrast blushes are more vas- number of fused vertebrae, intraoperative
cular and should be considered for embolization bleeding in excess of 2000 ml, obesity, age,
with particles (polyvinyl alcohol), liquid embol- neurological disability and ASA >3 [11–14].
ics (NBCA) and/or platinum coils. Vascular anat- Staphylococcus aureus, including methicillin-
omy should be considered carefully during the resistant organisms (MRSA), is the leading iso-
diagnostic phase in order to avoid the artery of lated pathogen causing infection in these patients
Adamkiewicz or radiculomedullary feeders prior and is estimated to account for 50% of cases
to making the decision to embolize. Reductions [10]. Additional common pathogens include
in intraoperative blood loss of 50% have been Streptococci, Enterococcus faecalis,
demonstrated following effective embolization Psuedomonas aeruginosa, Escherichia coli, and
[6, 7]. Patients should undergo surgery between Klebsiella pneumoniae [10]. Propionibacterium
24 and 72 hours following embolization or there acnes is a low virulence, anaerobic bacteria,
is risk of tumor revascularization [8]. By identi- comprising skin flora, however, is an underesti-
fying risk factors for hemorrhage and performing mated cause of SSI. It may be the most common
tumor embolization, surgeons may reduce blood cause of late postoperative infections in implant-
loss, decrease surgical time, prevent transfusions, able devices and has been reportedly involved in
and avoid hypotensive episodes. up to 45% of spinal implant infections [15]. The
Postoperative hemorrhage is an additional management of severe SSI caused by P. acnes
worry in patients with metastatic spine disease. generally involves antimicrobial treatment
In patients who undergo large decompressions including long-term suppression. The treatment
and reconstructions, there is significant potential of spinal implant infection is variable. While
space. As such, hematoma development can debridement, washout, and local flap closure
cause new neurologic injury given the now- with well-vascularized muscle are generally the
decompressed thecal sac. This complication can treatment of choice, there may be cases where
be minimized by preoperative and intraoperative implant removal is necessary [16, 17]. Preventing
techniques. Preoperatively, the patient should be SSIs during the index procedure is the preferred
optimized hematologically and poor surgical strategy. Prophylactic antibiotic dosing within
candidates should not be offered surgery. 1 hour of incision, in addition to applying vanco-
Intraoperatively, meticulous hemostasis, drain mycin powder directly into the wound prior to
use, and closure of dead space (see later chapters closure have proven to be effective in SSI preven-
in this book related to complex wound closure) tion [18, 19]. Regional application of Vancomycin
represent important surgical techniques. powder before wound closure has been shown to
decrease the rate of SSI [18, 20]. Several cohort
studies have demonstrated the effect of vancomy-
Wound Infection and Dehiscence cin in SSI reduction in thoracolumbar fusion sur-
gery [20, 21], posterior fusion after trauma [18],
Reported complication rates for patients under- and posterior surgical decompression and fusion
going surgery for spinal metastases are higher surgeries [22]. Godil et al. [23] found a signifi-
than for equivalent surgeries for nontumor indi- cant improvement in the rates of SSI by using
cations [9, 10]. Surgical site infections (SSI) are intrawound vancomycin (13% vs. 0%). In addi-
the most common complication after instru- tion, this study also showed that the cost of treat-
mented spinal metastasis surgery and are asso- ing postoperative infections was significantly
ciated with prolonged hospital stay as well as reduced by the use of vancomycin powder.
increased morbidity and mortality [9–11]. Risk One approach to address wound complica-
factors contributing to SSI include spinal tions prophylactically is the use of soft tissue
instrumentation, previous radiation, reoperation, reconstruction techniques during the index
surgery. Patients with spinal instrumentation delayed fashion [24, 27]. For complex failures
undergoing revision surgery, those undergoing where local tissue is of questionable viability,
surgery in a previously irradiated field, smok- specialized closure techniques including local
ers and those with diabetes may be appropriate muscle advancement, rotational or transposi-
candidates. The strategy, often coordinated tional tissue flaps may be necessary in order to
with plastic surgery, is to mobilize well-vascu- increase the vascularity and tissue coverage over
larized local muscle flaps to close potential the defect. In addition to reducing dead space and
dead space and provide vascularized coverage preventing seroma cavities, procedures that relo-
to spinal instrumentation. Chang et al. demon- cate vascularized tissue can provide necessary
strated a decreased incidence of wound com- hardware coverage, facilitate wound healing, and
plications from 45% to 20% with such an accelerate bacterial clearance [29]. Flap closure
approach [24]. A group from MD Anderson has been shown to decrease both the number of
compared their rate of major wound complica- debridements as well as to decrease the need to
tions when using this prophylactic strategy and remove hardware [30].
found a decrease from 38% to 12% [25].
For patients returning with suspected SSI, it is
important to determine whether it is located in a CSF Leak
superficial or deep (to the paraspinal muscle fas-
cia) compartment. While superficial infections Surgical approaches to spinal metastases fre-
may be effectively managed with a course of quently involve resecting tumor in a circumferen-
antibiotics, wound packing or wound vacuums, tial fashion from the epidural space with the
deep infections can require more aggressive potential need to sacrifice a nerve root. These
treatment. Moreover, deep infections can cause maneuvers can result in unintended durotomy
osteomyelitis, discitis, epidural abscesses with with subsequent cerebrospinal fluid (CSF) egress.
compression of the neural elements and coloniza- CSF leaks occur in spine surgery at a reported
tion of the hardware. Patients presenting with incidence of 0.3–35% [31–33]. Complications
high suspicion for infection based on pain, ery- resulting from CSF leak include positional head-
thema, tenderness to palpation, fluid collection, ache, pseudomeningoceles, meningitis, arach-
or drainage from wound should have a vitals and noiditis, CSF fistula through the dermis,
labs including complete blood count, ESR, CRP, neurological symptoms resulting from nerve root
pro-calcitonin, and cultures sent. High suspicion or spinal cord compression, failed wound heal-
should prompt an MRI with gadolinium contrast. ing, and surgical site infection [34, 35].
Contrast enhancing collections can be aspirated Successful management of durotomies is
under image guidance in order to obtain gram related to both the size and early detection of the
stain and culture. Based on these results, surgical defect. Small durotomies detected intraopera-
intervention including drainage, wound debride- tively should be closed primarily and may be
ment, placement of drains and plastics-assisted augmented with fibrin sealants, fat, muscle or
closure should be considered (Fig. 45.1), [26– fascial grafts or gelatin sponges. When sacrific-
28]. The selection of antibiotic agents and dura- ing a nerve root, it should be performed proximal
tion of treatment should be determined by an to the dorsal root ganglion and ligated with silk
infectious disease specialist. Patients are gener- suture or vessel clips where the root diverges
ally treated with intravenous broad-spectrum from the common dural sac. Closures can be
antibiotics until a pathogen is identified from cul- challenged intraoperatively by requesting a
ture and then treatment is narrowed. Valsava maneuver from the anesthesiologist to
Patients that have had prior radiation, particu- determine whether they are water tight and
larly conventional, and/or treatment with certain durable.
chemotherapeutic agents, such as bevacizumab, Larger durotomies may require dural grafts to
may develop wound complications in a more be sutured to native dural margins to form a
a b
c d
e f
Fig. 45.1 Wound reconstruction with local muscle flap. toward the midline and imbricated to cover the overlying
(a) Wound infection at the cervicothoracic junction with dead space between vertebrae and instrumentation. (e)
dehiscence and exposed hardware. (b) Incision is Subcutaneous drains are placed between the paraspinous
reopened prior to debridement of necrotic tissue. (c) muscle layers and in the epidural subfacial space. The tra-
Paraspinous musculature and trapezius are dissected and pezius muscles are approximated. (f) Skin is closed and
elevated bilaterally. (d) Paraspinous muscles are advanced drains are secured
patch. Dural substitutes are made of a variety of approaches, other onlay strategies may be
materials including bovine pericardium, porcine employed. These consist of dural slings made of
intestinal mucosa or processed collagen matrices. dural substitutes, fascia lata, muscle, gelatin
The suture line can then be covered with fibrin sponges, fibrin sealants with or without buttress-
glue. In difficult-to-access regions, such as ante- ing by hardware, interbody grafts, or other
rior defects encountered during posterior implants. Location of drains and whether or not
they are placed to suction is a matter of debate. In leaks are kept upright [31]. In complex cases
a series of 25 patients undergoing intentional refractory to conservative treatment, revision sur-
durotomy and placement of subfascial epidural gery may be the necessary management strategy
drain, no patient developed postoperative CSF in order to prevent postentially severe delayed
cutaneous leak, symptomatic pseudomeningo- sequelae: the formation of chronic fistula, pseu-
cele, or complication associated with closed suc- domeningocele, and delayed wound healing [45].
tion drains [36]. Alternatively, epidural drains
can be placed to passive or gravity drainage if
there is concern for the integrity of the dural Hardware Failure
repair. Tenuous dural closures may benefit from
placement of a lumbar drain intraoperatively with Hardware failure is the second most common
postoperative drainage and positional restric- adverse event that necessitates reoperation [46].
tions. This is discussed in greater detail below. This complication is characterized by broken
Postoperative detection of CSF leaks is of rods, cage migration, loosened or pulled out
equal importance. Patients presenting with pos- screws, and displacement of implanted hardware
tural symptoms of headache, dizziness, nausea requiring revision surgery. Previous radiation
and vomiting that resolve when recumbent should therapy is the most significant risk factor associ-
be approached with suspicion. Further indicators ated with hardware failure [47]. Other risk factors
include persistent high output clear drainage from include extensive tumor involvement of the pedi-
wound drains, clear fluid from the wound or blot- cle and vertebral body, prior or concomitant rib
table collection. More severe intrathecal hypoten- resection leading to chest wall instability, instru-
sion may result in hygromas, subdural hematomas, mentation construct involving more than six ver-
and cerebellar tonsillar descent to the foramen tebral segments [48], poor bone quality associated
magnum. Postoperative neurological symptoms with metastatic involvement, and postmeno-
resulting from compression secondary to pseudo- pausal or androgen-blockade-induced osteoporo-
meningocele and meningitis should also be ruled sis [49]. Metastatic tumor histology also
out. MRI may be useful to detect CSF collections, contributes to the risk of hardware failure. Among
determine if they are exerting mass effect, and symptomatic hardware failure patients, breast,
whether they are communicating with the intra- and prostate cancer represent the most common
thecal space. MRI studies of the brain will dem- source of primary tumors, while lung cancer was
onstrate pachymeningeal enhancement in the the most common in the group of patients who
setting of intracranial hypotension [37]. did not suffer from hardware failure [47]. Of
With symptomatic CSF leaks, placement of note, the same study revealed that survival time
lumbar drains and positional restrictions with bed for patients without hardware failure is twice as
rest should be considered. If CSF is draining long as for patients with hardware failure.
through the wound, the wound can be oversewn Hardware failure requires revision surgery if
[38]. The decision to put a patient on bed rest the patient becomes symptomatic. Strategies to
generally remains surgeon preference. Indeed, prevent screw loosening (Fig. 45.2a, b) or pullout
Gautschi et al. demonstrated that in a mixed pop- include cement augmentation of pedicle screws
ulation 175 spine surgeons, 14.9% do not use bed with polymethyl methacrylate (PMMA) in the
rest, 35% endorse 24-hr bed rest, 28% endorse thoracic and lumbar spine (Fig. 45.2b, c) [50–52].
48-hr bed rest and 6.3% use 72-hr bed rest [39, Recently, fenestrated screws through which
40]. Drainage parameters range from 5 to 15 cc/ cement can be injected into the vertebral body
hr for 4–5 days. This was reported to be effective have been approved by the US Food and Drug
in 83–100% of cases [41–44]. While patients Administration (FDA) [53]. This technique can
with leaks occurring in the distal thoracic or lum- serve as an anchor to increase screw pull out
bar spine are generally placed flat while on bed strength and may also ward off vertebral body
rest, patients with proximal thoracic or cervical compression fractures. Additionally, extending
Fig. 45.2 Hardware
failure and cement a b
augmentation via
fenestrated pedicle
screws. (a, b) T5 – T11
posterior spinal fusion
with T9 transpedicular
decompression for
metastatic renal cell
carcinoma. (a) Lateral
radiograph obtained on
routine follow up
demonstrates regions of
lucency around right
T10 and T11 pedicle
screws (arrowheads)
prompting follow up CT.
(b) Sagittal CT confirms
osteolysis around right
T10 and T11 pedicle
screws. (c, d) T8 – T12
posterior spinal fusion
with T10 transpedicular
decompression for
metastatic lung
adenocarcinoma in
osteoporotic patient with
metastases at adjacent
segments. (c) Lateral
radiograph showing c d
multi-level cement
augmentation through
fenestrated pedicle
screws. (d) Sagittal CT
demonstrates extent of
PMMA cement around
the tip of pedicle screw
in attempt to prevent
screw loosening as well
as vertebral body
compression fracture
(arrow)
construct length to two or more levels below and neurological examination and can be confirmed
above vertebrae infiltrated with tumor for with imaging. This is an operative emergency
posterior-only approaches may mitigate the risk of and the patient should be taken for evacuation
future hardware failure. Where possible, inclusion and decompression immediately to optimize
of anterior column support in the reconstruction neurologic recovery [60]. In patients who
process may be useful as 360-degree stabilization emerge from anesthesia with radicular pain or
distributes axial loading forces and decreases the neurologic deficit, imaging should be acquired
likelihood of future hardware failure. immediately to rule out misplaced hardware,
compressive lesion, overcorrection, or
malalignment. Spinal cord infarcts are rare in
Venous Thromboembolism (VTE) this patient population but if suspected, MRI
with perfusion and diffusion imaging can be
Venous thromboembolism (VTE), including deep used for diagnosis with subsequent blood pres-
vein thrombosis (DVT) and pulmonary embolism sure management in the ICU setting. Finally, in
(PE), represents one of the most significant causes patients who sustained a change in intraopera-
of morbidity and mortality in cancer patients with tive neurophysiological monitoring and wake
a prevalence ranging from 0.3% to 15.5% [54]. with a neurological deficit that has no other eti-
The preoperative prevention and early detection ology confirmed with imaging, intensive care
of thrombosis including, screening and assess- should be pursued. Mean arterial pressure goals
ment of coagulation cascade is necessary for early greater than 85 mm Hg and possible steroid
intervention and risk stratification. Placement of administration can be considered for 5–7 days
inferior vena cava (IVC) filter in spinal metastasis [61–63].
patients with positive ultrasonographic screening
for DVT has significantly reduced the incidence
of postoperative PE associated with DVT [55]. In Conclusion
addition to IVC filter, placement of mechanical
devices such as pneumatic intermittent compres- The high incidence of spinal metastatic lesions
sion boots and compression stockings has also calls for a thorough understanding of surgical
reduced the rate of postoperative DVT [55, 56]. treatment strategies as well as complication rec-
Postoperative prophylactic subcutaneous unfrac- ognition and management in order to optimize
tionated heparin is critical in this population and patient outcomes. Surgical treatment is employed
can be used safely without significantly increased as a palliative measure to alleviate pain, improve
risk of postoperative hemorrhage [57]. In patients function, and optimize quality of life in symp-
diagnosed with acute postoperative PE who are tomatic patients. Minimizing morbidity associ-
hemodynamically stable, clinical observation ated with complications is critical in this patient
may be sufficient. However, should patient population. Complications can be categorized
become hemodynamically unstable, thrombolytic into medical such as DVT, pneumonia, and
therapy or mechanical pulmonary thrombectomy wound infection. Surgical complications include
should be considered to prevent complications excess intraoperative bleeding, postoperative
such as cardiopulmonary arrest [58]. hematoma, wound dehiscence, CSF leak, and
hardware failure. Individual patient risk profiles
should be constructed based on comorbidities
Neurological Deterioration and prognosis should be carefully considered
prior to offering surgery. Based on risk assess-
Neurologic deterioration after spinal surgery is ment, some patients may benefit from open
a potential complication and has an incidence surgery while others may be more appropriate
of 2–4% [59]. Epidural hematoma should be for minimally invasive or percutaneous
suspected in patients with a rapidly declining approaches. Still others may not be appropriate
surgical candidates and most appropriately 14. Sebaaly A, et al. Surgical site infection in spi-

treated with chemotherapy and/or radiation nal metastasis: incidence and risk factors. Spine J.
2018;18(8):1382–7.
alone. Careful preoperative optimization with 15. Sampedro MF, et al. A biofilm approach to detect
attention to risk mitigation facilitates desirable bacteria on removed spinal implants. Spine (Phila Pa
postoperative outcomes. Finally, early recogni- 1976). 2010;35(12):1218–24.
tion of postoperative complications is the first 16. Levi AD, Dickman CA, Sonntag VK. Management of
postoperative infections after spinal instrumentation.
step in effective management and will serve to J Neurosurg. 1997;86(6):975–80.
minimize long-term morbidity in this vulnerable 17.
Weinstein MA, McCabe JP, Cammisa FP Jr.
patient population. Postoperative spinal wound infection: a review of
2,391 consecutive index procedures. J Spinal Disord.
2000;13(5):422–6.
18. O’Neill KR, et al. Reduced surgical site infections in
References patients undergoing posterior spinal stabilization of
traumatic injuries using vancomycin powder. Spine J.
1. Kakhki VR, et al. Pattern and distribution of bone 2011;11(7):641–6.
metastases in common malignant tumors. Nucl Med 19. Okafor R, et al. Intrawound vancomycin pow-

Rev Cent East Eur. 2013;16(2):66–9. der for spine tumor surgery. Global Spine J.
2. Campbell PG, et al. Patient comorbidity score pre- 2016;6(3):207–11.
dicting the incidence of perioperative complica- 20. Hey HW, et al. Is intraoperative local vancomycin
tions: assessing the impact of comorbidities on powder the answer to surgical site infections in spine
complications in spine surgery. J Neurosurg Spine. surgery? Spine (Phila Pa 1976). 2017;42(4):267–74.
2012;16(1):37–43. 21. Sweet FA, Roh M, Sliva C. Intrawound application of
3. Reis RC, et al. Risk of complications in spine surgery: vancomycin for prophylaxis in instrumented thoraco-
a prospective study. Open Orthop J. 2015;9:20–5. lumbar fusions: efficacy, drug levels, and patient out-
4. Nair S, et al. Preoperative embolization of hyper- comes. Spine (Phila Pa 1976). 2011;36(24):2084–8.
vascular thoracic, lumbar, and sacral spinal column 22. Pahys JM, et al. Methods to decrease postoperative
tumors: technique and outcomes from a single center. infections following posterior cervical spine surgery.
Interv Neuroradiol. 2013;19(3):377–85. J Bone Joint Surg Am. 2013;95(6):549–54.
5. Robial N, et al. Is preoperative embolization a pre- 23. Godil SS, et al. Comparative effectiveness and cost-
requisite for spinal metastases surgical management? benefit analysis of local application of vancomycin
Orthop Traumatol Surg Res. 2012;98(5):536–42. powder in posterior spinal fusion for spine trauma:
6. Prince EA, Ahn SH. Interventional management of clinical article. J Neurosurg Spine. 2013;19(3):331–5.
vertebral body metastases. Semin Intervent Radiol. 24. Chang DW, Friel MT, Youssef AA. Reconstructive
2013;30(3):278–81. strategies in soft tissue reconstruction after resec-
7. Wilson MA, et al. Retrospective analysis of preop- tion of spinal neoplasms. Spine (Phila Pa 1976).
erative embolization of spinal tumors. AJNR Am J 2007;32(10):1101–6.
Neuroradiol. 2010;31(4):656–60. 25. Garvey PB, et al. Immediate soft-tissue reconstruc-
8. Hong CG, et al. Preoperative embolization in tion for complex defects of the spine following
patients with metastatic spinal cord compres- surgery for spinal neoplasms. Plast Reconstr Surg.
sion: mandatory or optional? World J Surg Oncol. 2010;125(5):1460–6.
2017;15(1):45. 26. Chahoud J, Kanafani Z, Kanj SS. Surgical site infec-
9. Wise JJ, et al. Complication, survival rates, and risk tions following spine surgery: eliminating the con-
factors of surgery for metastatic disease of the spine. troversies in the diagnosis. Front Med (Lausanne).
Spine (Phila Pa 1976). 1999;24(18):1943–51. 2014;1:7.
10. Omeis IA, et al. Postoperative surgical site infec-
27. Mesfin A, et al. Changing the adverse event profile in
tions in patients undergoing spinal tumor surgery: metastatic spine surgery: an evidence-based approach
incidence and risk factors. Spine (Phila Pa 1976). to target wound complications and instrumenta-
2011;36(17):1410–9. tion failure. Spine (Phila Pa 1976). 2016;41 Suppl
11. Demura S, et al. Surgical site infection in spinal
20:S262–s270.
metastasis: risk factors and countermeasures. Spine 28. Janssen DMC, et al. A retrospective analysis of deep
(Phila Pa 1976). 2009;34(6):635–9. surgical site infection treatment after instrumented
12. Kumar N, et al. Blood loss and transfusion require- spinal fusion with the use of supplementary local anti-
ments in metastatic spinal tumor surgery: evalu- biotic carriers. J Bone Joint Infect. 2018;3(2):94–103.
ation of influencing factors. Ann Surg Oncol. 29. Vitaz TW, et al. Rotational and transpositional flaps
2016;23(6):2079–86. for the treatment of spinal wound dehiscence and
13. Kumar S, et al. Risk factors for wound infec-
infections in patient populations with degenerative
tion in surgery for spinal metastasis. Eur Spine J. and oncological disease. J Neurosurg. 2004;100(1
2015;24(3):528–32. Suppl Spine):46–51.
30. Chieng LO, et al. Reconstruction of open wounds as 48. Amankulor NM, et al. The incidence and pat-

a complication of spinal surgery with flaps: a system- terns of hardware failure after separation surgery
atic review. Neurosurg Focus. 2015;39(4):E17. in patients with spinal metastatic tumors. Spine J.
31. Menon SK, Onyia CU. A short review on a complica- 2014;14(9):1850–9.
tion of lumbar spine surgery: CSF leak. Clin Neurol 49. Moon BJ, et al. Polymethylmethacrylate-augmented
Neurosurg. 2015;139:248–51. screw fixation for stabilization of the osteoporotic
32. Ghobrial GM, et al. Iatrogenic neurologic deficit
spine : a three-year follow-up of 37 patients. J Korean
after lumbar spine surgery: a review. Clin Neurol Neurosurg Soc. 2009;46(4):305–11.
Neurosurg. 2015;139:76–80. 50.
Frankel BM, Jones T, Wang C. Segmental
33. Weber C, Piek J, Gunawan D. Health care costs of polymethylmethacrylate- augmented pedicle screw
incidental durotomies and postoperative cerebrospi- fixation in patients with bone softening caused by
nal fluid leaks after elective spinal surgery. Eur Spine osteoporosis and metastatic tumor involvement: a
J. 2015;24(9):2065–8. clinical evaluation. Neurosurgery. 2007;61(3):531–7;
34. Guerin P, et al. Incidental durotomy during spine sur- discussion 537–8.
gery: incidence, management and complications. A 51. Jang JS, et al. Polymethylmethacrylate-augmented

retrospective review. Injury. 2012;43(4):397–401. screw fixation for stabilization in metastatic spi-
35. Tafazal SI, Sell PJ. Incidental durotomy in lumbar nal tumors. Technical note. J Neurosurg. 2002;96(1
spine surgery: incidence and management. Eur Spine Suppl):131–4.
J. 2005;14(3):287–90. 52. Amendola L, et al. Fenestrated pedicle screws for
36. Niu T, et al. Postoperative cerebrospinal fluid leak cement-augmented purchase in patients with bone
rates with subfascial epidural drain placement after softening: a review of 21 cases. J Orthop Traumatol.
intentional durotomy in spine surgery. Global Spine 2011;12(4):193–9.
J. 2016;6(8):780–5. 53. Fransen P. Increasing pedicle screw anchoring in the
37. Pannullo SC, et al. MRI changes in intracranial hypo- osteoporotic spine by cement injection through the
tension. Neurology. 1993;43(5):919–26. implant. Technical note and report of three cases. J
38. Tosun B, et al. Management of persistent cerebrospi- Neurosurg Spine. 2007;7(3):366–9.
nal fluid leakage following thoraco-lumbar surgery. 54. Yoshioka K, et al. Prevalence and risk factors for
Asian Spine J. 2012;6(3):157–62. development of venous thromboembolism after
39. Gautschi OP, et al. Incidental durotomy in lumbar degenerative spinal surgery. Spine (Phila Pa 1976).
spine surgery – is there still a role for flat bed rest? 2015;40(5):E301–6.
Spine J. 2014;14(10):2522–3. 55. Zacharia BE, et al. Incidence and risk factors for pre-
40. Gautschi OP, et al. Incidental durotomy in lumbar operative deep venous thrombosis in 314 consecutive
spine surgery–a three-nation survey to evaluate its patients undergoing surgery for spinal metastasis. J
management. Acta Neurochir. 2014;156(9):1813–20. Neurosurg Spine. 2017;27(2):189–97.
41. Hu P, et al. A circumferential decompression-
56. Ferree BA, Wright AM. Deep venous thrombosis fol-
based surgical strategy for multilevel ossification lowing posterior lumbar spinal surgery. Spine (Phila
of thoracic posterior longitudinal ligament. Spine J. Pa 1976). 1993;18(8):1079–82.
2015;15(12):2484–92. 57. Gerlach R, et al. Postoperative nadroparin admin-
42. Hu PP, Liu XG, Yu M. Cerebrospinal fluid leak-
istration for prophylaxis of thromboembolic
age after thoracic decompression. Chin Med J. events is not associated with an increased risk
2016;129(16):1994–2000. of hemorrhage after spinal surgery. Eur Spine J.
43. Mazur M, et al. Management of cerebrospinal fluid 2004;13(1):9–13.
leaks after anterior decompression for ossification of 58. Fukuda W, et al. Management of pulmonary throm-
the posterior longitudinal ligament: a review of the boembolism based on severity and vulnerability to
literature. Neurosurg Focus. 2011;30(3):E13. thrombolysis. Ann Vasc Dis. 2017;10(4):371–7.
44. Cho JY, et al. Management of cerebrospinal fluid leak- 59. Luksanapruksa P, et al. Perioperative complica-

age after anterior decompression for ossification of tions of spinal metastases surgery. Clin Spine Surg.
posterior longitudinal ligament in the thoracic spine: 2017;30(1):4–13.
the utilization of a volume-controlled pseudomenin- 60. Scavarda D, et al. [Postoperative spinal extra-

gocele. J Spinal Disord Tech. 2012;25(4):E93–102. dural hematomas. 14 cases]. Neurochirurgie.
45. Fang Z, et al. Subfascial drainage for management of 1997;43(4):220–7.
cerebrospinal fluid leakage after posterior spine sur- 61. Ziewacz JE, et al. The design, development, and imple-
gery–a prospective study based on Poiseuille’s law. mentation of a checklist for intraoperative neuromoni-
Chin J Traumatol. 2016;19(1):35–8. toring changes. Neurosurg Focus. 2012;33(5):E11.
46. Quraishi NA, et al. Reoperation rates in the surgical 62. Ryken TC, et al. The acute cardiopulmonary manage-
treatment of spinal metastases. Spine J. 2015;15(3 ment of patients with cervical spinal cord injuries.
Suppl):S37–43. Neurosurgery. 2013;72(Suppl 2):84–92.
47. Pedreira R, et al. Hardware failure in patients with 63. Yue JK, et al. Update on critical care for acute spinal
metastatic cancer to the spine. J Clin Neurosci. cord injury in the setting of polytrauma. Neurosurg
2017;45:166–71. Focus. 2017;43(5):E19.
Vertebral Augmentation
Procedures for Treatment
46
of Pathologic Vertebral Body
Fractures
Justin Schwarz, Alejandro Santillan,
Adham Mushtak, and Athos Patsalides
Introduction or compression of neural elements that can cause

neurologic deficits. Vertebral augmentation pro-
The spine is a common site for metastasis in can- cedures (VAP) are minimally invasive treatment
cer patients, and spinal metastases are observed options which have been shown to reduce pain,
in approximately 60–70% of patients with sys- improve mobility, and stabilize vertebral bodies
temic cancer [1]. In the United States alone, more in patients with refractory back pain from patho-
than 350,000 cases of bony spinal metastasis are logic compression fractures [4].
reported each year due to prostate, breast, kidney,
lung, and thyroid cancers [2]. Symptomatic
pathologic vertebral compression fractures Procedural Details
(VCF) can be debilitating for patients, causing
pain that significantly affects quality of life. As Vertebroplasty and kyphoplasty are effective per-
with osteoporotic compression fractures, patho- cutaneous minimally invasive techniques for
logic fractures historically have been treated con- treating pathologic VCF and ameliorating symp-
servatively with pain medication and bracing [3]. toms [4]. In both procedures, polymethylmethac-
While conservative treatment is adequate for rylate (PMMA) is injected into the fractured bone
some patients, others continue to suffer from where it hardens and congeals the fracture frag-
debilitating pain that affects their ability to care ments, providing immediate stability and pain
for themselves, live independently, and signifi- relief [5–7]. Kyphoplasty includes an additional
cantly decreases their quality of life. Depending step prior to PMMA injection where a balloon is
on the degree of tumor invasion of the vertebral gently inflated within the fractured vertebral
body, these malignant fractures can significantly body to create a cavity for the PMMA and restore
worsen with conservative measures alone, lead- vertebral body height [8].
ing to fracture progression, exacerbation of pain, VAP is most efficiently performed using
biplane fluoroscopy but can also be performed
using single-plane fluoroscopy in either an inter-
J. Schwarz (*) · A. Santillan · A. Patsalides
Department of Neurological Surgery, NewYork- ventional procedure room or a traditional operat-
Presbyterian/Weill Cornell Medicine, ing room. Patients are placed in the prone position
New York, NY, USA on the procedure table prior to the administration
e-mail: jts2004@nyp.org of moderate sedation or following endotracheal
A. Mushtak intubation and administration of general anesthe-
Weill Cornell Medical College in Qatar, sia. Pressure points including the forearms,
Ar-Rayyan, Qatar

https://doi.org/10.1007/978-3-030-42958-4_46
614 J. Schwarz et al.
elbows, knees, and abdomen are appropriately (10-gauge) from a posterior approach through the
secured and supported (Fig. 46.1a). Patients are skin and soft tissues and into the fractured verte-
sterilely prepped in typical fashion, and a single bral body. The cannula is directed through or just
dose of perioperative intravenous antibiotics is lateral to the corresponding vertebral body’s ped-
administered just prior to the procedure icle so that the center of fractured vertebral body
(Fig. 46.1b). is ultimately reached (Figs. 46.2d, e and 46.3e).
VAP are minimally invasive percutaneous The trajectory of the cannula is carefully chosen
procedures that only require one or two small to ensure that the neural elements, neuroforamen,
skin incisions per treated vertebral body level, and spinal canal are not encountered. This mini-
depending upon if a unilateral or bilateral mizes the risk of direct neural element injury or
approach is utilized (Fig. 46.1c). Intermittent cerebrospinal fluid leak, both of which are
fluoroscopy is used to direct a narrow cannula exceedingly rare but potential complications.
Once the cannulas are appropriately positioned,
core biopsies of the fractured vertebral body are
a collected so that tissue can be sent for pathologic
and histologic analysis, especially for patients
with questionable cancer recurrence or a new
cancer diagnosis (Fig. 46.2e). Then, for patients
undergoing kyphoplasty, balloons are placed into
the vertebral body and carefully inflated under
fluoroscopy (Fig. 46.2f). After balloon inflation
for kyphoplasty or just after cannula placement
for vertebroplasty, PMMA is injected into the
fractured vertebral body (Figs. 46.2g and 46.3f).
PMMA injection proceeds under fluoroscopy to
b prevent an inadvertent leakage of cement out of
the vertebral body, either into venous structures
or into the spinal canal. If this situation is encoun-
tered, cement injection is stopped immediately.
Following the successful injection of PMMA, the
cannulas are removed and sterile dressings are
applied.
Post-Procedural Care
c
The postoperative care for VAP is limited, and an
emphasis is placed on early mobilization. Patients
are encouraged to ambulate as much as reason-
ably possible in order to prevent complications
from inactivity, such as pneumonia and deep
venous thromboses. However, patients are coun-
seled to avoid strenuous activity, lifting objects
over 5–10 pounds, and excessive bending or
twisting until they are seen at their follow-up
appointment 2 weeks following VAP. Patients
Fig. 46.1 (a) Prone positioning for VAP with pressure
points adequately padded. (b) Sterilely prepped and
with lower thoracic and lumbar VCF are encour-
draped patient. (c) VAPs require small incisions aged to use a thoracic lumbar sacral orthosis
46 Vertebral Augmentation Procedures for Treatment of Pathologic Vertebral Body Fractures 615
a b c
d e
f g
Fig. 46.2 Pathologic T8 VCF from lung adenocarci- transpedicular VAP approach to T8 VCF with 10-gauge
noma. (a) Non-contrast CT sagittal pre-VCF. (b) Non- cannulas. (e) Left unilateral core biopsy needle in place in
contrast CT sagittal post-VCF. (c) MRI sagittal T1 without T8 vertebral body. (f) Bilateral balloon inflation for T8
contrast demonstrates T8 tumor infiltration. (d) Bilateral kyphoplasty. (g) T8 post-kyphoplasty
(TLSO) for 2 weeks following vertebroplasty or care required. The procedural sterile dressings
kyphoplasty when ambulating, but not while sit- are typically removed 24–48 hours after the VAP,
ting or lying down. These procedures do not and the small incisions can remain uncovered
require an extensive hospital stay and are usually after that time. Patients are instructed to avoid
performed as an outpatient, with patients leaving any submerging of the incisions, but they are
1–2 hours after the completion of the procedure. encouraged to shower normally after the dress-
There is also minimal postoperative incisional ings are removed. VAP also minimizes the time
a b c d
e f
Fig. 46.3 Pathologic L3 VCF from breast adenocarci- enhancement. (d) Frontal and lateral X-rays are inade-
noma. (a) MRI sagittal T2 STIR hyperintensity within L3 quate to demonstrate VCF or tumor infiltration. (e) Frontal
suggests an acute to subacute fracture. (b) MRI sagittal T1 and lateral fluoroscopy demonstrates bilateral transpedic-
without contrast demonstrates L3 tumor infiltration. (c) ular approach with bilateral radiofrequency ablation
MRI sagittal T1 with contrast demonstrates typical tumor probes. (f) Post-vertebroplasty of L3 pathologic VCF
that patients need to be off of their antiplatelet or being treated, the patient is unable to cooperate,
anticoagulation medication. These procedures or systemic medical issues necessitate endotra-
can be performed while patients are on single cheal intubation with general anesthesia.
antiplatelet therapy, such as aspirin. Dual anti-
platelet therapy and anticoagulation are typically
held prior to the procedure but can be resumed on Patient Selection
postoperative day 1.
As with any other invasive treatment, patient
selection is important. VCF with symptoms that
Anesthesia Care are not adequately controlled on oral pain medica-
tions are considered for kyphoplasty or vertebro-
VAP is performed with the assistance of an anes- plasty. The typical presentation of a symptomatic
thesiologist and is typically performed under pathologic compression fracture is an acute onset
monitored anesthesia care (MAC) with generous of mechanical back pain that roughly correlates
use of local anesthetic to minimize the amount of with the vertebral level of the VCF.
sedative medications required. Endotracheal
intubation with general anesthesia is used in
select patients where MAC is not appropriate. Pain Quality and Characteristics
MAC is used for relatively healthy and coopera-
tive patients undergoing a one- or two-level Pathologic compression fractures cause signifi-
kyphoplasty or vertebroplasty. General anesthe- cant pain due to the relative instability of the frac-
sia is used if three or more vertebral levels are tured vertebral body. This pain is often intensified
with any axial loading of the compression frac- past radiologic studies, including previous X-rays,
ture and brought on by movement. VCF pain is CT, or MRI (Fig. 46.2a–c). If no comparison
typically experienced in the region of the frac- imaging is available, an MRI is obtained to deter-
ture. For instance, a lower thoracic pathologic mine acuity of the fracture. Short tau inversion
compression fracture will cause pain in the lower recovery (STIR) hyperintensity within the verte-
thoracic region corresponding to the fractured bral body of interest suggests a relatively recent
level. Sometimes this pain can be reproduced on fracture and identifies a vertebral body that is
physical examination by palpation of the midline amenable to intervention (Fig. 46.3a). Subacute
at the level of suspected fracture. It is common and acute compression fractures respond well to
for patient with pathologic VCF to also have sig- VAP, whereas chronic fractures are unlikely to
nificant pain from spasm of the paraspinal mus- have a favorable response. Pathologic fractures or
cles. This pain is often described as sharp and at risk vertebral levels can be identified by regions
episodic and travels rostrally and caudally just of T1 hypointensity, which is typically exception-
lateral to midline. It does not directly improve ally sensitive at identifying tumor infiltration
following VAP because it is muscular in etiology. (Figs. 46.2c and 46.3b).
Muscle spasm pain is typically improved by mus-
cle relaxant medications and by increasing physi-
cal activity. Patients with symptomatic Contraindications
compression fractures may also experience radic-
ular pain at the associated level, especially for Vertebroplasty and kyphoplasty are typically
thoracic compression fractures. The loss of verte- reserved for patients with pathologic VCF that
bral body height can cause irritation of the asso- are mechanically stable and are not causing
ciated nerve roots leading to pain that can radiate symptomatic spinal cord or nerve root compres-
in a radicular distribution to the anterior chest sion. Patients with unstable fractures do not have
wall. Less frequently, this can occur in the lum- significant pain relief following VAP and usually
bar spine, leading to radicular pain in the corre- require surgical stabilization. A qualified practi-
sponding nerve root distribution into the legs. tioner, such as a neurosurgeon or an orthopedic
spine surgeon, should be consulted if there is any
uncertainty about the mechanical stability of a
Diagnosis VCF. Retropulsion with symptomatic cord, conus
medullaris, cauda equina, or nerve root compres-
Many patients who suffer from pathologic VCF sion is an absolute contraindication to VAP. These
have multiple medical comorbidities, such as patients require open surgical intervention for
arthritis, spinal stenosis, or other bony or visceral decompression of the neural elements and possi-
metastases that may make the assessment of pain ble stabilization. VAP in these circumstances
difficult. In such cases, it may be difficult to deter- may worsen the compression of neuronal struc-
mine if a newly diagnosed compression fracture is tures in these patients and worsen their neuro-
truly symptomatic. A thorough clinical history is logic status. Asymptomatic patients with
necessary to determine if a VCF is the etiology of retropulsion are still candidates for vertebro-
a patient’s pain. Determining the chronicity of plasty. In this patient population, PMMA injec-
symptoms and correlating the clinical history tion is performed cautiously to avoid any
with physical exam and radiographic findings are worsening of retropulsion into the spinal canal
important. A dedicated spinal CT or MRI is pre- and prevent any neurologic deterioration. Other
ferred for diagnosis (Figs. 46.2a–c and 46.3a–c). absolute contraindications to VAP include active
Plain X-rays are inadequate for proper diagnosis osteomyelitis at the fracture area or an allergy to
but have limited utility as an initial screening tool polymethylmethacrylate. Patients being consid-
(Fig. 46.3d). The chronicity of injury can be ered for VAP should not be thrombocytopenic,
determined by comparing current imaging with leukopenic, or coagulopathic at the time of the
procedure because these clinical scenarios fluoroscopy for these levels. In certain situations,
increase the risk of post-procedural hematoma or T3 and T4 can be visualized well enough to
infection. Typically, these situations can be attempt VAP, but this is dependent upon the
addressed by waiting for the patient’s leukopenia patient’s body habitus and anatomy. In these situ-
or thrombocytopenia to resolve following che- ations, the operator will often not know if the T3
motherapy administration. If this is not possible, or T4 level can be successfully visualized to per-
transfusion or administration of bone marrow form VAP until the patient is positioned.
stimulant medications can be considered. These Pathologic cervical fractures are usually not
situations require direct communication between treated with VAP, but these procedures can be uti-
the interventionist, primary care team, and oncol- lized in certain situations, especially for patho-
ogist to determine the best course of action. logic fractures of C2 [17–19]. Symptomatic
sacral metastases can also be treated with VAP,
but it is much less common than treatment for
Potential Complications lumbar and thoracic spine metastases [20].
The risk of VAP is exceptionally low, and the

potential benefit of these procedures is significant Kyphoplasty and Vertebroplasty
[4]. There is a risk of infection in VAP, as with any Results
surgical procedure, but the risk of infection in
VAP has consistently been reported to be less than Multiple studies have demonstrated that verte-
1% [9]. While the rate of infection is low, an broplasty and kyphoplasty are well tolerated,
infection involving PMMA can result in substan- provide pain relief, and improve functional out-
tial morbidity and necessitate extensive surgical comes in patients with painful neoplastic spinal
procedures, including laminectomies, corpecto- fractures. A single randomized study of 134
mies, and spinal fusions [10]. The majority of patients with bone metastases resulting from
patients with post-VAP infections have a recent solid tumors and multiple myeloma demonstrated
preoperative history of infection, including osteo- that treatment of VCF with kyphoplasty was
myelitis, discitis, or urinary tract infections [11]. associated with durable and clinically meaning-
Therefore, patients must be carefully evaluated ful improvements in physical functioning, back
preoperatively to rule out any active infective pro- pain, and quality of life when compared to non-
cesses prior to consideration of VAP. The leakage surgical management [21]. A meta-analysis of
of PMMA outside of the vertebral body is not seven nonrandomized studies of patients with
uncommon, but it is usually not of clinical signifi- multiple myeloma or osteolytic metastasis
cance [12]. Leakage of PMMA into the surround- revealed that kyphoplasty was associated with
ing soft tissues or the intervertebral disc space is reduced pain and improved functional outcomes,
typically asymptomatic [13]. PMMA can travel which were maintained up to 2 years post-
into venous structures and lead to pulmonary procedure. Kyphoplasty also improved early ver-
embolism, but the risk of a symptomatic PMMA tebral height loss, but these effects were not long
pulmonary embolism is exceedingly low [14]. term [22]. Similarly, a retrospective review of 67
Post-procedural hematoma or PMMA leakage patients with multiple myeloma-related VCF
into the neuroforamina or spinal canal can also demonstrated that vertebroplasty provided clini-
occur, but these are rarely symptomatic [15, 16]. cally meaningful improvements in physical func-
tioning, pain, and mobility throughout a year of
follow-up [23]. Several small nonrandomized
Treatable Vertebral Levels studies of VAP including kyphoplasty and verte-
broplasty have generated comparable results
Fractures involving T5 through the lumbar spine [24–26]. The role of vertebroplasty for patients
are amenable to VAP because the vertebral anat- with myeloma, however, remains debatable in the
omy is easily visualized using frontal and lateral absence of prospective data because two random-
ized trials failed to show any benefit with verte- pathologic VCFs with minimal risk.
broplasty in patients with osteoporotic fractures Radiofrequency ablation is a promising adjuvant
versus conservative therapy [27–29]. or stand-alone therapy for pathologic VCFs for
Furthermore, a meta-analysis of 59 studies, pain control and local tumor control, but this
including 56 case series, showed that kypho- treatment modality needs additional prospective
plasty seemed to be more effective than vertebro- studies to confirm its efficacy.
plasty in relieving pain secondary to cancer-related
VCF [30]. These results taken in aggregate sug-
gest that kyphoplasty should be performed for References
symptomatic pathologic VCF when possible.
1. Shah LM, Salzman KL. Imaging of spinal metastatic
Vertebroplasty is potentially useful in situations
disease. Int J Surg Oncol. 2011;2011:769753. https://
where kyphoplasty is contraindicated, such as in doi.org/10.1155/2011/769753. Epub 2011 Nov 3.
patients with VCFs associated with significant 2. Mundy GR. Metastasis to bone: causes, conse-
retropulsion or neural element compression. quences, and therapeutic opportunities. Nat Rev
Cancer. 2002;2:584–93.
3. Audat ZA, Hajyousef MH, Fawareh MD, Alawneh
KM, Odat MA, Barbarawi MM, Alomari AA, Jahmani
Radiofrequency Ablation RA, Khatatbeh MA, Assmairan MA. Comparison
if the addition of multilevel vertebral augmenta-
tion to conventional therapy will improve the out-
Radiofrequency ablation has been proposed as a
come of patients with multiple myeloma. Scoliosis
stand-alone and adjuvant therapy to vertebro- Spinal Disord. 2016;11:47. https://doi.org/10.1186/
plasty and kyphoplasty. RFA utilizes a high- s13013-016-0107-6.
frequency alternating current that is passed from 4. Kasperk C, Haas A, Hillengass J, et al. Kyphoplasty
in patients with multiple myeloma a retrospec-
a needle electrode into the surrounding tissue,
tive comparative pilot study. J Surg Oncol.
resulting in heating and eventual coagulative tis- 2012;105(7):679–86.
sue necrosis [31]. Some reports have suggested 5. Voormolen MHJ, Mali WPTM, Lohle PNM, et al.
that combined RFA and vertebroplasty is a safe Percutaneous vertebroplasty compared with opti-
mal pain medication treatment: short-term clini-
and efficacious procedure for not only pain man-
cal outcome of patients with subacute or chronic
agement but also local tumor control in spinal painful osteoporotic vertebral compression frac-
metastasis [32]. While RFA and vertebroplasty tures. The VERTOS Study. Am J Neuroradoiol.
are independently effective in pain palliation in 2007;28(3):555–60.
spinal metastasis, some studies suggest that the 6. Wardlaw D, Cummings SR, Van Meirhaeghe J, et al.
Efficacy and safety of balloon kyphoplasty compared
combination of RFA and vertebroplasty may with non-surgical care for vertebral compression frac-
have a synergistic effect on pain management tures (FREE): a randomized controlled trial. Lancet
[33–41]. The majority of these studies are single- (Lond Engl). 2009;373(9668):1016–24. https://doi.
org/10.1016/S0140-6736(09)60010-6.
arm observational studies, and there is a need for
7. Klazen CAH, Lohle PNM, de Vries J, et al.
additional studies evaluating combined RFA and Vertebroplasty versus conservative treatment in
VAPs for efficacy in regard to pain relief and acute osteoporotic vertebral compression fractures
local tumor control. (Vertos II): an open-label randomized trial. Lancet
(Lond Engl). 2010;376(9746):1085–92. https://doi.
org/10.1016/S0140-6736(10)60954-3.
8. Van Meirhaeghe J, Bastian L, Boonen S, et al. A
Conclusion randomized trial of balloon kyphoplasty and non-
surgical management for treating acute vertebral
compression fractures: vertebral body kyphosis
Pathologic VCFs are relatively common in can- correction and surgical parameters. Spine.
cer patients and are often painful and potentially 2013;38(12):971–83. https://doi.org/10.1097/
debilitating for patients. Their prompt diagnosis BRS.0b013e31828e8e22.
and treatment is essential. Vertebroplasty and 9. Saracen A, Kotwica Z. Complications of percutane-
ous vertebroplasty: an analysis of 1100 procedures
kyphoplasty are minimally invasive procedures performed in 616 patients. Medicine (Baltimore).
that can significantly improve the quality of life 2016;95(24):e3850. https://doi.org/10.1097/
and functional status of patients suffering from MD.0000000000003850.
10. Abdelrahman H, Siam AE, Shawky A, Ezzati A,

23. McDonald RJ, Trout AT, Gray LA, et al. Vertebroplasty
Boehm H. Infection after vertebroplasty or kypho- in multiple myeloma: outcomes in a large patient
plasty. A series of nine cases and review of lit- series. AJNR Am J Neuroradiol. 2008;29:642–8.
erature. Spine J. 2013;13(12):1809–17. https://doi. 24. Huber F, McArthur N, Tanner M, et al. Kyphoplasty
org/10.1016/j.spinee.2013.05.053. for patients with multiple myeloma is a safe surgical
11. Walker DH, Mummaneni P, Rodts GE. Infected verte- procedure: results from a large patient cohort. Clin
broplasty: report of two cases and review of the litera- Lymphoma Myeloma. 2009;9:375–80.
ture. Neurosurg Focus. 2004;17(6):E6. 25. Zou J, Mei X, Gan M, et al. Kyphoplasty for spi-
12. Saracen A, Kotwica Z. Treatment of multiple
nal fractures from multiple myeloma. J Surg Oncol.
osteoporotic vertebral compression fractures by 2010;102:43–7.
percutaneous cement augmentation. Int Orthop. 26. Dalbayrak S, Onen M, Yilmaz M, et al. Clinical

2014;38(11):2309–12. and radiographic results of balloon kyphoplasty for
13. Kotwica Z, Saracen A. Early and long-term outcomes treatment of vertebral body metastases and multiple
of vertebroplasty for single osteoporotic fractures. myelomas. J Clin Neurosci. 2010;17:219–24.
Neurol Neurochir Pol. 2011;45(5):431–5. 27. Chew C, Craig L, Edwards R, et al. Safety and effi-
14. Luetmer MT, Bartholmai BJ, Rad AE, Kallmes
cacy of percutaneous vertebroplasty in malignancy: a
DF. Asymptomatic and unrecognized cement pulmo- systematic review. Clin Radiol. 2011;66:63–72.
nary embolism commonly occurs with vertebroplasty. 28. Buchbinder R, Osborne RH, Ebeling PR, et al.

AJNR Am J Neuroradiol. 2011;32(4):654–7. https:// A randomized trial of vertebroplasty for pain-
doi.org/10.3174/ajnr.A2368. ful osteoporotic vertebral fractures. N Engl J Med.
15. Taylor RS, Fritzell P, Taylor RJ. Balloon kyphoplasty 2009;361:557–68.
in the management of vertebral compression fractures: 29. Kallmes DF, Comstock BA, Heagerty PJ, et al. A ran-
an updated systematic review and meta-analysis. Eur domized trial of vertebroplasty for osteoporotic spinal
Spine J. 2007;16(8):1085–100. fractures. N Engl J Med. 2009;361:569–79.
16. Robinson Y, Tschöke SK, Stahel PF, Kayser R,
30. Bhargava A, Trivedi D, Kalva L, et al. Management
Heyde CE. Complications and safety aspects of of cancer-related vertebral compression fracture:
kyphoplasty for osteoporotic vertebral fractures: comparison of treatment options—A literature meta-
a prospective follow-up study in 102 consecutive analysis. J Clin Oncol. 2009;27:15S, e20529.
patients. Patient Saf Surg. 2008;2:2. https://doi. 31. Halpin RJ, Bendok BR, Liu JC. Minimally invasive
org/10.1186/1754-9493-2-2. treatments for spinal metastases: vertebroplasty,
17. De la Garza-Ramos R, Benvenutti-Regato M, Caro- kyphoplasty, and radiofrequency ablation. J Support
Osorio E. Vertebroplasty and kyphoplasty for cer- Oncol. 2004;2(4):339–51.
vical spine metastases: a systematic review and 32. Wallace AN, Robinson CG, Meyer J, Tran ND,

meta-analysis. Int J Spine Surg. 2016;10(7) https:// Gangi A, Callstrom MR, et al. The metastatic spine
doi.org/10.14444/3007. disease multidisciplinary working group algorithms.
18. Blondel B, Adetchessi T, Demakakos J, Pech-Gourg Oncologist. 2015;20(10):1205–15.
G, Dufour H, Fuentes S. Anterolateral kyphoplasty in 33. Madaelil TP, Wallace AN, Jennings
the management of cervical spinal metastasis. Orthop JW. Radiofrequency ablation alone or in combination
Traumatol Surg Res. 2012;98(3):341–5. with cementoplasty for local control and pain pal-
19. Sun G, Jin P, Li M, et al. Percutaneous vertebro- liation of sacral metastases: preliminary results in 11
plasty for treatment of osteolytic metastases of patients. Skelet Radiol. 2016;45(9):1213–9.
the C2 vertebral body using anterolateral and pos- 34. Munk PL, Rashid F, Heran MK, Papirny M, Liu

terolateral approach. Technol Cancer Res Treat. DM, Malfair D, et al. Combined cementoplasty and
2010;9(4):417–22. radiofrequency ablation in the treatment of pain-
20. Shah RV. Sacral kyphoplasty for the treatment of
ful neoplastic lesions of bone. J Vasc Interv Radiol.
painful sacral insufficiency fractures and metastases. 2009;20(7):903–11.
Spine J. 2012;12(2):113–20. https://doi.org/10.1016/j. 35. Lane MD, Le HB, Lee S, Young C, Heran MK,

spinee.2012.01.019. Badii M, et al. Combination radiofrequency abla-
21. Berenson J, Pflugmacher R, Jarzem P, et al. Balloon tion and cementoplasty for palliative treatment of
kyphoplasty versus non-surgical fracture man- painful neoplastic bone metastasis: experience with
agement for treatment of painful vertebral body 53 treated lesions in 36 patients. Skelet Radiol.
compression fractures in patients with cancer: a mul- 2011;40(1):25–32.
ticentre, randomised controlled trial. Lancet Oncol. 36. Reyes M, Georgy M, Brook L, Ortiz O, Brook
2011;12:225–3596. A, Agarwal V, et al. Multicenter clinical and
22. Bouza C, López-Cuadrado T, Cediel P, Saz-Parkinson imaging evaluation of targeted radiofrequency
Z, Amate JM. Balloon kyphoplasty in malignant ablation (t-RFA) and cement augmentation of
spinal fractures: a systematic review and meta- neoplastic vertebral lesions. J Neuro Interv
analysis. BMC Palliat Care. 2009;8:12. https://doi. Surg. 2018;10:176–82. https://doi.org/10.1136/
org/10.1186/1472-684X-8-12. neurintsurg-2016-012908.
37. Halpin RJ, Bendok BR, Sato KT, Liu JC, Patel JD, for pain relief and functional recovery in painful bone
Rosen ST. Combination treatment of vertebral metas- metastases. Skelet Radiol. 2013;42(1):25–36.
tases using image-guided percutaneous radiofre- 40. Toyota N, Naito A, Kakizawa H, Hieda M, Hirai N,
quency ablation and vertebroplasty: a case report. Tachikake T, et al. Radiofrequency ablation therapy
Surg Neurol. 2005;63(5):469–74. combined with cementoplasty for painful bone metas-
38. Schaefer O, Lohrmann C, Markmiller M, Uhrmeister tases: initial experience. Cardiovasc Intervent Radiol.
P, Langer M. Combined treatment of a spinal metas- 2005;28(5):578–83.
tasis with radiofrequency heat ablation and vertebro- 41. Hoffmann RT, Jakobs TF, Trumm C, Weber C,

plasty. Am J Roentgenol. 2003;180(4):1075–7. Helmberger TK, Reiser MF. Radiofrequency ablation
39. Clarençon F, Jean B, Pham H-P, Cormier E, Bensimon in combination with osteoplasty in the treatment of
G, Rose M, et al. Value of percutaneous radiofrequency painful metastatic bone disease. J Vasc Interv Radiol.
ablation with or without percutaneous vertebroplasty 2008;19(3):419–25.
Spinal Laser Interstitial Thermal
Therapy for Metastatic Tumors
47
Linton T. Evans, Rafael A. Vega,
and Claudio E. Tatsui
Introduction spinal metastases continues to grow as advance-

ments in radiation and systemic therapy have
Metastatic epidural spinal cord compression prolonged survival in individuals with metastatic
(MESCC) is a significant source of morbidity cancer. Treatment is fundamentally palliative,
and impairment in quality of life in individuals focused on neurologic preservation, restora-
with cancer [1]. Approximately 40% of patients tion of spinal stability, pain relief, and durable
with a systemic malignancy will develop spi- local tumor control [3]. Due to the palliative
nal osseous metastases, and up to 10% present intent of therapy, however, any intervention must
with symptomatic spinal cord compression [2]. minimize treatment- related morbidity or com-
Not all tumors exhibit the same predilection or plications, leading to a relatively narrow thera-
tropism for bone; frequent offenders include peutic window. The practice of spinal oncology
prostate, lung, and breast carcinoma followed by is becoming increasingly complex as innovations
lymphoma, renal cell carcinoma, and multiple in surgical technology, immunotherapy, targeted
myeloma. The distribution of metastases along chemotherapies, and radiation therapy change
the spinal axis reflects the relative bone mass of the therapeutic landscape. Furthermore, patients
each segment and the regional blood flow. Most with metastatic disease frequently have multiple
spine metastases are found within the thoracic medical comorbidities in the face of progres-
spine (60%) followed by the lumbosacral (25%) sive systemic disease. The clinical management
and cervical spine (15%). Multiple synchro- of these patients is multidisciplinary at its core,
nous sites of disease in the spine are common, requiring discussions between surgeons, medi-
an important fact to consider during evaluation cal oncologists, and radiation oncologists. The
and treatment of these patients. The burden of demands on the surgeon are to provide effective
surgical intervention associated with minimal
L. T. Evans
morbidity, shortest hospitalization, and least dis-
Department of Neurosurgery, Dartmouth-Hitchcock ruption to systemic therapy.
Medical Center, Lebanon, NH, USA Historically patients with MESCC were
R. A. Vega treated with high-dose glucocorticoids and frac-
Department of Neurosurgery, Beth Israel Deaconess tionated radiation therapy [4]. Initial efforts at
Medical Center, Boston, MA, USA surgical intervention were focused on posterior-
C. E. Tatsui (*) only decompression of the spinal canal and were
Department of Neurosurgery, MD Anderson commonly associated with worsened neurologic
e-mail: cetatsui@mdanderson.org
and functional outcomes compared with radiation

https://doi.org/10.1007/978-3-030-42958-4_47
624 L. T. Evans et al.
alone. In retrospect this surgical strategy led to radiation therapy (SBRT), delivers radiation to
further destabilization of an already compromised a contoured volume with a steep dose gradient
spine by removal of the intact posterior elements. that spares surrounding tissues such as the spi-
Furthermore, the site of epidural compression nal cord, nerves, or esophagus. The biologically
and spinal metastases is typically anterior or ven- effective dose of radiation delivered with SSRS is
tral to the spinal canal. Later developments in estimated to be approximately three times greater
spine stabilization and instrumentation, as well than with cEBRT, leading to more extensive
as methods of circumferential decompression of DNA damage, irrecoverable endothelial damage,
the spinal canal, revitalized the role of surgery in and potentially enhanced immune environment
the management of spine metastases. In a pivotal with T-cell activation and pro-inflammatory cyto-
study by Patchell et al. [5], individuals with soli- kines [7]. Radiosurgery effectively overcomes
tary and symptomatic MESCC were randomized the previously held histology-specific radioresis-
to circumferential decompression/stabilization tance, with 12-month local control rates of 85%
followed by conventional external beam radiation in even notoriously difficult tumor types such as
therapy (cEBRT) or cEBRT alone. Patients in the RCC [8]. Furthermore, due to the conformality of
surgical cohort experienced significant improve- SSRS and relative sparing of surrounding tissues,
ment in rates of ambulation, functional ability, it is possible to use as a salvage therapy in the
pain control, urinary continence, and survival. setting of prior radiation failures for local recur-
This study established that appropriately selected rence [9, 10].
surgery offers a meaningful improvement in qual- While SSRS is an effective and reliable treat-
ity of life with acceptable morbidity when added ment for spine metastases, radiation-induced spi-
to radiation therapy. The aim of surgery is to pro- nal cord injury remains a concern [11]. A widely
vide surgical stabilization and decompression of accepted dose maximum to the spinal cord is
the neural elements. Ultimately radiotherapy is 14 Gy. Using this parameter, a large multicenter
the source of local tumor control. study following over 1000 individuals treated
Tumor histology has an impact on the effi- with SSRS found only 6 patients that developed
cacy of radiation therapy, measured as the rate radiation-induced myelopathy. In the setting of
of local control. Traditionally, tumors were clas- high-grade epidural compression, the toxicity-
sified as either radiation sensitive or resistant limiting dose of the spinal cord or cauda equina
based on their response to conventional frac- requires adjustment to the prescribed treatment
tionated radiation therapy [6]. Radiosensitive dose, potentially under treating the tumor margin
histologies include lymphoma, plasmacytoma, and compromising local tumor control. Lovelock
multiple myeloma, small cell lung carcinoma, et al. [12] found that local treatment failure was
germ cell tumors, breast carcinoma, and prostate associated with tumors that received less than
carcinoma. In response to cEBRT, these tumors 15 Gy to any point in the treatment planning
have a reported 2-year local control rate of up volume. A surgical strategy designed to create
to 80–90%. In contrast, radioresistant malig- separation between the tumor and spinal cord
nancies such as lung, thyroid, hepatocellular, has emerged to facilitate the use of radiosurgery
colorectal, and renal cell carcinoma, melanoma, in the setting of epidural compression [13, 14].
and sarcomas exhibit much poorer 2-year local Surgery involves resection of epidural tumor with
control – as low as 30% following radiation ther- reconstitution of the thecal sac, followed by spi-
apy. Furthermore, symptomatic and neurologic nal stabilization as indicated. The aim of surgery
improvement is often limited to several months applied in conjunction to SSRS is (i) decompres-
in these patients. Developments in image-guided sion of the spinal cord in cases of compressive
stereotaxy and radiation therapy have enabled myelopathy; (ii) to create separation between
the delivery of highly conformal and tumoricidal tumor and the spinal cord; (iii) and to provide
doses of radiation as either a single treatment or spinal stabilization. The extent of tumor resection
hypofractionated (2–5) regimen. Spinal stereo- is not crucial to local control as long as there is
tactic radiosurgery (SSRS), or stereotactic body an adequate distance between the tumor margin
47 Spinal Laser Interstitial Thermal Therapy for Metastatic Tumors 625
and spinal cord to deliver tumoricidal doses of or radiofrequency ablation of vertebral tumors
SSRS. Separation surgery followed by SSRS rep- [15–17]. Injury to the spinal cord or nerve roots
resents a paradigm shift in spinal oncology and has been documented with radiofrequency abla-
has dramatically improved treatment of oligo- tion, and in animal studies, placement of the elec-
metastatic disease. trode immediately adjacent to the posterior cortex
of the vertebral body or pedicle leads to neural
injury [18, 19]. Concern for neurologic injury and
ationale for Laser Interstitial
R the inability to monitor tissue injury in real time
Thermal Therapy (LITT) has limited the adoption of these techniques for
the ablation of epidural tumor in close proximity
Individuals with metastatic cancer are frequently to the neural elements. Laser interstitial thermal
deconditioned and harbor a number of medi- therapy is an alternative method of percutaneous
cal comorbidities. Malnutrition, chronic anemia, ablation that has seen widespread adoption in the
chronic steroid use, systemic thromboses (DVT treatment of intracranial tumors and other pathol-
or PE), and/or prior radiation complicate surgical ogy [20, 21]. Using this technique, a small laser
intervention. Furthermore, these patients com- probe is inserted into the lesion using stereotactic
monly have rapidly progressive disease at other guidance. Energy is transferred from the laser into
sites in addition to their spine requiring concurrent the surrounding tissue producing a thermal injury
and systemic therapy with cytotoxic or targeted sufficient to lead to tumor cell death and coagu-
agents. For these individuals separation surgery lative necrosis. The amount of tissue damage is
may lead to significant morbidity and delays based on a thermal response model in which there
systemic therapy until the patient has adequately is a correlation between temperature, duration of
recovered. Percutaneous techniques have been exposure, and the ensuing damage. An advantage
developed as an alternative to open surgical pro- of this technology over others is that an intraopera-
cedures in certain scenarios to decrease morbidity, tive MRI is used to monitor in real time the heat
limit disruption of systemic therapy or antico- generation within a particular region. Using spinal
agulation, shorten hospital admissions, decrease LITT (sLITT), epidural tumor in close proximity
pain, and minimize blood loss or transfusions. to the thecal sac and spinal cord can be ablated
Currently used methods include CT-guided cryo- while ensuring that there is no thermal injury to
a 7 A 6 b
Soft Vertebral
B
tissue 8 5 body
C
9 4
Pedicle
Transverse E
process
D
10 3
11
2
12
Superior articular 1 Spinous
facet process
Right Left
Fig. 47.1 Diagram demonstrating the typical approaches cord (a). The ideal distance between the fiber and dura is
(i.e., oblique transpedicular, yellow arrow) based on the 5–7 mm, while each fiber covers a 10–12 mm radius (b)
location of the metastatic lesion in relation to the spinal
the spinal cord (Fig. 47.1) [22–24]. As an addi- and destruction of tumor within the foramina
tional source of protection, the CSF surrounding and associated nerve typically provides complete
the spinal cord and epidural venous plexus serve as resolution of the pain. For the same reason, we
a heat-sink limiting the generation of heat in close restrict the use of LITT to the thoracic spinal
proximity to the spinal cord. Regions of high- segments to avoid unintentional injury to func-
grade epidural compression can safely be ablated tional nerve roots of the cervical and lumbosacral
using sLITT. This treatment paradigm, similar to plexus. For lesions of the cervical and lumbar
separation surgery, requires adjuvant SSRS for spine, surgical decompression with visualization
effective tumor control. Similar to circumferen- and complete decompression of the functional
tial decompression, the region of necrotic tissue roots is preferred. As previously discussed, prior
following thermal ablation creates a separation conventional radiation therapy and spinal insta-
between viable tumor and the spinal cord facilitat- bility are not contraindications to sLITT. In the
ing effective doses of SSRS. For individuals that case of prior radiation, a percutaneous technique
also have spinal instability, a percutaneous stabi- such as LITT is desirable to avoid wound com-
lization can be performed following LITT in the plications. If there is spinal instability, a percu-
same setting [25]. taneous stabilization is frequently performed
following the laser ablation [25]. This can be
done during the same anesthetic or as a staged
Patient Selection procedure.
A number of metastatic tumors are notori-
Spinal laser interstitial thermal therapy is an ously vascular. These include renal cell carci-
effective and safe procedure in properly selected noma, hepatocellular carcinoma, and thyroid
patients. sLITT is a minimally invasive alterna- carcinoma. Prior to a circumferential decompres-
tive to open circumferential decompression for sion, these tumors are typically embolized pre-
patients with epidural compression that are can- operatively in an effort to decrease the amount of
didates for radiosurgery [22, 23]. High-grade blood loss. Percutaneous laser interstitial thermal
epidural compression is typically defined using ablation is associated with only minimal blood
the Bilsky scale [13] and classified as grade 1c or loss. Furthermore, an endovascular embolization
higher. In these individuals the degree of epidural is unnecessary and avoids an additional proce-
compression would limit treatment with an effec- dure in this patient cohort.
tive radiosurgery dose. Additional considerations
for patient selection include (i) medical comor-
bidities; (ii) need to continue or rapidly resume Technical Description
systemic therapy; (iii) normal neurologic exam;
(iv) thoracic spine; and (v) no contraindications At our institution sLITT is performed within an
to MRI (e.g., pacemaker or neurostimulator). For operating room suite equipped with an intraop-
patients in which MRI is contraindicated, sLITT erative MRI (iMRI) (BrainLab Inc., Feldkirchen,
cannot be performed without MRI thermography. Germany). Following induction of general anes-
Similarly, existing instrumentation at the level of thesia, the patient is placed in the prone position
ablation typically creates metallic artifact that with the upper extremities parallel to the body in
impairs the accuracy of MRI thermography and a manner that is ergonomic to the surgeon and
precludes its use. Individuals presenting with a does not interfere with the use of the C-arm flu-
neurologic deficit require surgical decompres- oroscope or iMRI [26]. Initially, we used a CT
sion and are not candidates for a percutaneous scan of the spine and C-arm for localization and
procedure such as LITT or radiosurgery alone. stereotactic placement of the laser fibers [22, 23].
Individuals with debilitating thoracic radiculopa- Currently, we are using MRI for coregistration
thy due to foraminal tumor involvement are ideal and spinal navigation and have found that this
candidates for laser ablation [24]. The ablation can be accomplished with submillimeter accu-
racy. Additionally, the MRI provides better spatial Spine navigation allows for meticulous trajec-
resolution of the tumor and its relation to the neu- tory and entry point planning. In our experience
ral elements for trajectory planning and insertion we have relied on the Weinstein-Boriani-Biagini
of the fibers. After final positioning, but prior to tumor classification to select the optimal probe
the iMRI, skin fiducials (Izi Medical Products, trajectory [27]. Typically one of three trajec-
Owing Mills, MD, USA) are placed on the region tories is used based on the location of the epi-
of interest in a unique pattern that distinguishes dural disease that is being treated. The most
right-left and rostral-caudal (Fig. 47.2a). The sur- common trajectory is an oblique transpedicular
gical site is prepped and draped, and a small inci- or transforaminal trajectory. This is well suited
sion is made with dissection proceeding to the to treat disease that is ventral to the spinal cord
level of the spinous process. Using subperiosteal or canal (zones 4–6 or 7–9). Orthogonal trans-
dissection the soft tissues are reflected away from pedicular or translaminar trajectories can also be
the spinous process, and a MRI-compatible clamp used to access different sites of disease intended
and reference array (Medtronic, Minneapolis, for treatment. In general the selected trajectory
MN) are secured to the bone (Fig. 47.2b, c). places the laser fiber approximately 6 mm from
Without disrupting or displacing the reference the dura or thecal sac, and it is assumed that
array and fiducials, a Siemens body matrix coil is each fiber can achieve a 10 mm diameter of ther-
placed over the region of interest and the patient mal injury. Depending on the extent of disease
is positioned within the MRI (Fig. 47.2d). A in the rostral-caudal plane, multiple trajectories
high-resolution T2WI is used for coregistration may be required to achieve an adequate ablation
and navigation. Following image acquisition, the (Fig. 47.4). We have used up to nine trajectories
series is transferred to a Stealth S7 workstation in a single patient. When planning multiple tra-
(Medtronic, Minneapolis, MN), and coregistra- jectories, they are placed within 10 mm of one
tion is performed using a point matching regis- another to ensure that there are no untreated seg-
tration with the fiducial markers (Fig. 47.3). The ments between successive ablations. Similarly,
accuracy is confirmed prior to proceeding with bilateral trajectories may be needed to completely
insertion of the epidural cannulas and laser fibers. treat ventral or lateral epidural disease.
a b d
Fig. 47.2 Patient in prone position on the iMRI transfer ered with a sterile plastic bag (c). MRI coil placed over the
table, with fiducial markers applied in the dorsal region plastic fiducial held by a plastic cradle to avoid fiducial
overlying the tumor (a). The skin is prepped and the spi- displacement (d)
nous process clamp is secured (b). Spinal clamp is cov-
a c
b d
Fig. 47.3 (a) Sterile reference array is attached to clamp tested inside the fiducials, midline and easily palpable spi-
under sterile conditions. A non sterile probe is used to per- nous processes. (c) Axial (d) sagital navigated inline
form surface matching of fiducials. (b). MRI 1mm acial images are used for planing trajectories for placement of
cuts are obtained and transferred to standard navigation laser catheter and pedicle screws.
system for registration, accuracy of the image guidance is
Following selection of the appropriate The laser fiber consists of a 980-nm diode
trajectory(s) and entry point(s), a navigated encased in a catheter that is connected to a
Jamshidi needle is introduced and the navigation 15-W power source. A single fiber is introduced
accuracy is confirmed. Small incisions are made to the cannula and advanced to depth. MR ther-
at the entry sites, and a Jamshidi needle (DePuy mography is based on gradient-echo acquisition
Synthes, Raynham, MA, USA) is advanced until and used throughout the ablation to monitor the
it contacts the lamina or other bone surface. heat generated within the tissue. Proton reso-
The C-arm is then used to confirm the location nance within the tissue is sensitive to temper-
of the Jamshidi needle and verify that the fluo- ature, and the difference in phases allows for
roscopy and spine navigation are commensurate modeling of the temperature within the exposed
with one another. Next, the Jamshidi is advanced tissue. 3-mm slices are acquired every 5–6 sec-
to target depth using navigation (Fig. 47.4). A onds while the laser is activated. The laser
K-wire is introduced through the Jamshidi needle is deactivated when one of two temperature
and exchanged with a 1.65 mm-diameter plastic thresholds are reached. The boundary between
catheter and stylet (Fig. 47.5). This is repeated dura and tumor is identified and set to an upper
in succession for each trajectory. Once all of the temperature limit of 48–50° (Fig. 47.7). A sec-
cannulas have been inserted, the surgical field is ond threshold is set to 90° in the tissue adjacent
covered, and another MRI is obtained to confirm to the laser fiber to prevent excessive heating
the locations (Fig. 47.6). of the tumor and tissue carbonization. The
Fig. 47.4 Fiducials are

removed; the rest of the a
skin is prepped and
draped in the usual
sterile fashion (a).
Navigated Jamshidi
needle is inserted using
image guidance, where
the diameter of the
needle (yellow) is
increased to position the
needle (blue) 5–7 mm
lateral to the dura (b).
This is repeated to with
multiple trajectories, as
needed, to achieve an
adequate ablation (c)
b c
thermal maps are sensitive to and degraded by T1WI is acquired, again with breath holding. The
motion. The spine is vulnerable to respiropha- region of coagulative necrosis will lack contrast
sic motion and demands that a breath hold be enhancement, and it appears as a hypointense or
completed during the ablation. Thus, the abla- dark area post-contrast sequence (Fig. 47.8). In
tion is performed in cycles in which the laser our experience this has been an accurate estima-
is active for up to 120 seconds during a breath tion of the ablated volume. For individuals with
hold, interrupted by periods of ventilation to concomitant spinal instability, a stabilization pro-
allow for adequate oxygenation and recovery cedure can be performed under the same anes-
from hypercapnia. Typically, the ablation time thetic or as a separate staged surgery (Fig. 47.9).
in total is up to 4 minutes at a single site. The Typically, a percutaneous instrumentation with
laser fiber is manually advanced or withdrawn cement augmentation is performed using spi-
as needed to ensure that there is ablation of the nal navigation and the reference array from the
entire intended epidural tumor. sLITT, or standard fluoroscopic techniques.
After the ablation is complete, the laser fiber Generally, our practice is to repeat a MRI of the
and cannulas are removed, and the incisions are spine in 6–12 weeks. If instrumentation is used,
closed with an absorbable suture. To visualize a CT myelogram is obtained postoperatively for
the extent of ablation, a pre- and post-contrast radiosurgery planning.
a Clinical Outcomes and Results
In conjunction with radiosurgery, spinal laser

interstitial thermal therapy provides effective and
durable local tumor control with minimal mor-
bidity. From our initial experience, we reported
outcomes of sLITT and SSRS in 19 individuals
presenting with radioresistant tumors, the major-
ity of which had progressed despite systemic
therapy [22]. Within this cohort seven patients
had Bilsky 1c epidural compression, eight had
b grade 2 compression, and four exhibited grade 3
compression. SSRS was indicated in all subjects
for oncologic control, but the degree of epidural
compression would have restricted effective dose
to the planned treatment volume. sLITT provided
a percutaneous alternative to open surgery with
the benefit of an abbreviated hospital admission
(median of 2 days) and durable tumor control.
Progression was documented in only two patients
at 16 and 33 weeks and was ultimately retreated
Fig. 47.5 K-wires are inserted through the Jamshidi, which
with a subsequent sLITT. Furthermore, there was
is exchanged to a plastic access cannula (a). A modified plas-
tic introducer is inserted into the plastic cannula to maintain a statistically significant reduction (22%) in the
the trajectories, and additional needles are inserted in tandem dimensions of epidural tumor seen at 2 months
to cover the craniocaudal extension of the epidural mass (b) and improvement in the degree of epidural com-
a b
Fig. 47.6 The access cannulas are covered with sterile magnet for fiber localization (a). Sterile towels are placed
technique and the MRI coil placed over the region of over the MRI coils, and the laser catheter is inserted into
interest, followed by transferring the patient to the MRI the access cannulas (b)
45.4
a b c
42.9 41.1
43.9
High Safety Mathematical model of

Limit = 90 C Limit = 48 C thermal damage
Fig. 47.7 MRI T2 sequence is utilized to localize the model of thermal damage monitored in real time, attained
exact axial plane for the fiber, and a high limit is placed with our imaging software (b). A monitored ventilator
lateral to the fiber and set to 90 °C (red cross), and a lower pause is performed by the anesthesiologist during the
limit is placed in the interface between the tumor and dura acquisition of thermal images, where a total of 2 minutes
mater and set to 50 °C (purple cross) (a). Mathematical is allowed for each ablation cycle (c)
Fig. 47.8 MR images

are obtained and a b
demonstrate the
immediate thermal
damage. Preoperative T1
with contrast (a),
postoperative T1 without
contrast (b)
pression. Pain scores (VAS) were also signifi- lished analysis). Median follow-up was 35 weeks
cantly improved following sLITT. Complications for the entire cohort, with time to recurrence
in this series included a transient monoparesis in measuring a mean of 26 weeks. Approximately
one patient, a wound dehiscence requiring reop- one-third of patients also underwent a subsequent
eration, and a delayed compression fracture. To stabilization procedure.
date, we have performed more than 100 proce- From this larger experience, several lessons
dures to treat a variety of tumor histologies. Local have emerged. In our current practice, we limit
tumor progression has been documented at a total treatment to lesions within the thoracic spine
of 17 treated sites – 15 were in-field recurrences, located between T2 and T12 to avoid injury to
while 2 were at the treatment margins (unpub- the cervical or lumbosacral plexus. Based on the
a b c
Fig. 47.9 Cases associated with spinal instability are scanner and standard percutaneous pedicle screws with
treated in the same day directly after the ablation is com- cement augmentation can be placed using either fluoros-
pleted (a). The patient is positioned away from the MRI copy or image guidance; lateral (b) and AP views (c)
percutaneous nature of the procedure, traversing on immunotherapy undergoing LITT for cranial
nerve roots cannot be identified and protected. tumors that subsequently develop severe edema
Initial efforts to treat lesions in the upper lum- and inflammation.
bar spine were complicated by injury to roots Although the zone of thermal injury typically
at the corresponding level. In addition to spinal measures up to 10 mm in diameter, the ablation
level, the presence of a neurologic deficit prior is not universally homogenous or predictable.
to surgery, even if subtle, is an absolute contra- Regions of tumor that are adjacent to spinal fluid,
indication. Individuals with preexisting deficits large vessels, or cystic areas are more difficult to
have increased potential for neurologic worsen- treat due to the ability for these structures to dis-
ing post-ablation. Our series includes a patient sipate heat and function as a heat sink. Similarly,
treated with mild motor weakness preoperatively vascular tumors such as renal cell carcinoma may
and renal cell carcinoma. The procedure itself require longer treatment times and multiple tra-
was uncomplicated and initially well tolerated, jectories to adequately treat an area. The area of
but unfortunately, the patient had a delayed ablation is often less homogenous compared to
neurologic decline requiring surgical decom- other tumor types such as chordoma or lung car-
pression. Interestingly, review of the pathology cinoma. Osteoblastic tumors present additional
obtained from the ablated level at the time of challenges when using sLITT, as highly calcified
reoperation consisted of necrotic tissue with no tissue presents a low MRI signal interfering or
viable tumor. A second subject included in this decreasing the quality of temperature monitoring
series required an urgent decompression in the by MRI thermography.
setting of a delayed neurologic deficit. In this
case the patient was neurologically intact prior
to laser ablation but subsequently declined. The Conclusion
patient had received concurrent immunother-
apy for RCC, and it was hypothesized that the Spine laser interstitial therapy is an emerging and
combination of LITT and immunotherapy led minimally invasive method to treat spine metas-
to a significant immune reaction and edema. tases. It provides effective and durable local con-
Individuals treated with sLITT and immunomod- trol with minimal morbidity. Compared to other
ulatory agents may require special consideration. percutaneous techniques, sLITT is unique in
Similar observations have been made in patients offering real-time monitoring of thermal injury.
Additional benefits over conventional separa- tasis and its pattern of failure. J Neurosurg Spine.
tion surgery include limited hospital admissions, 2007;7(2):151–60.
12. Lovelock DM, Zhang Z, Jackson A, Keam J,

improved pain control, and minimal blood loss. Bekelman J, Bilsky M, et al. Correlation of local
Furthermore, vascular tumors do not require pre- failure with measures of dose insufficiency in the
operative embolization, and patients with signifi- high-dose single-fraction treatment of bony metas-
cant medical comorbidities or need for continued tases. Int J Radiat Oncol Biol Phys. 2010;77(4):
1282–7.
systemic therapy can safely be treated. The tech- 13. Bilsky M, Smith M. Surgical approach to epidural spi-
nology is still early in its development and not nal cord compression. Hematol Oncol Clin North Am.
stand-alone therapy. Rather it is best used in con- 2006;20(6):1307–17.
junction with SSRS to provide symptom pallia- 14. Laufer I, Iorgulescu JB, Chapman T, et al. Local dis-
ease control for spinal metastases following “sepa-
tion and local oncologic control. ration surgery” and adjuvant hypofractionated or
high-dose single-fraction stereotactic radiosurgery:
outcome analysis in 186 patients. J Neurosurg Spine.
References 2013;18(3):207–14.
15. Nakatsuka A, Yamakado K, Takaki H, et al.

Percutaneous radiofrequency ablation of painful spi-
1. Cole JS, Patchell RA. Metastatic epidural spinal cord nal tumors adjacent to the spinal cord with real-time
compression. Lancet Neurol. 2008;7(5):459–66. monitoring of spinal canal temperature: a prospec-
2. Bach F, Larsen BH, Rohde K, et al. Metastatic spi- tive study. Cardiovasc Intervent Radiol. 2009;32(1):
nal cord compression. Occurrence, symptoms, 70–5.
clinical presentations and prognosis in 398 patients 16. Masala S, Chiocchi M, Taglieri A, et al. Combined
with spinal cord compression. Acta Neurochir. use of percutaneous cryoablation and vertebroplasty
1990;107(1–2):37–43. with 3D rotational angiograph in treatment of single
3. Laufer I, Rubin DG, Lis E, et al. The NOMS frame- vertebral metastasis: comparison with vertebroplasty.
work: approach to the treatment of spinal metastatic Neuroradiology. 2013;55(2):193–200.
tumors. Oncologist. 2013;18(6):744–51. 17. Masala S, Roselli M, Manenti G, et al. Percutaneous
4. Gilbert RW, Kim JH, Posner JB. Epidural spinal cord cryoablation and vertebroplasty: a case report.
compression from metastatic tumor: diagnosis and Cardiovasc Intervent Radiol. 2008;31(3):669–72.
treatment. Ann Neurol. 1978;3(1):40–51. 18. Goetz MP, Callstrom MR, Charboneau JW, et al.

5. Patchell RA, Tibbs PA, Regine WF, et al. Direct Percutaneous image-guided radiofrequency ablation
decompressive surgical resection in the treatment of of painful metastases involving bone: a multicenter
spinal cord compression caused by metastatic cancer: study. J Clin Oncol. 2004;22(2):300–6.
a randomised trial. Lancet. 2005;366(9486):643–8. 19. Nakatsuka A, Yamakado K, Maeda M, et al.

6. Maranzano E, Latini P. Effectiveness of radiation Radiofrequency ablation combined with bone cement
therapy without surgery in metastatic spinal cord injection for the treatment of bone malignancies. J
compression: final results from a prospective trial. Int Vasc Interv Radiol. 2004;15(7):707–12.
J Radiat Oncol Biol Phys. 1995;32(4):959–67. 20. Sharma M, Balasubramanian S, Silva D, et al.

7. Greco C, Pares O, Pimentel N, et al. Spinal metas- Laser interstitial thermal therapy in the manage-
tases: from conventional fractionated radiotherapy ment of brain metastasis and radiation necrosis after
to single-dose SBRT. Rep Pract Oncol Radiother. radiosurgery: an overview. Expert Rev Neurother.
2015;20(6):454–63. 2016;16(2):223–32.
8. Gerszten PC, Burton SA, Ozhasoglu C, et al. 2 1. Thomas JG, Rao G, Kew Y, et al. Laser interstitial
Stereotactic radiosurgery for spinal metastases from thermal therapy for newly diagnosed and recurrent
renal cell carcinoma. J Neurosurg Spine. 2005;3(4): glioblastoma. Neurosurg Focus. 2016;41(4):E12.
288–95. 22. Tatsui CE, Stafford RJ, Li J, et al. Utilization of laser
9. Gerszten PC, Burton SA, Ozhasoglu C, et al. interstitial thermotherapy guided by real-time thermal
Radiosurgery for spinal metastases: clinical experi- MRI as an alternative to separation surgery in the
ence in 500 cases from a single institution. Spine management of spinal metastasis. J Neurosurg Spine.
(Phila Pa 1976). 2007;32(2):193–9. 2015;23(4):400–11.
10. Sahgal A, Larson DA, Chang EL. Stereotactic body 23. Tatsui CE, Lee SH, Amini B, et al. Spinal laser inter-
radiosurgery for spinal metastases: a critical review. stitial thermal therapy: a novel alternative to surgery
Int J Radiat Oncol Biol Phys. 2008;71(3):652–65. for metastatic epidural spinal cord compression.
11. Chang EL, Shiu AS, Mendel E, et al. Phase I/II study Neurosurgery. 2016;79(Suppl 1):S73–82.
of stereotactic body radiotherapy for spinal metas-
24. Thomas JG, Al-Holou WN, de Almeida Bastos DC, 26. Jimenez-Ruiz F, Arnold B, Tatsui CE, et al.

Ghia A, Li J, Bishop AJ, et al. A novel use of the Perioperative and anesthetic considerations for neu-
intraoperative MRI for metastatic spine tumors: laser rosurgical laser interstitial thermal therapy ablations.
interstitial thermal therapy for percutaneous treatment J Neurosurg Anesthesiol. 2018;30(1):10–7.
of epidural metastatic spine disease. Neurosurg Clin 27.
Boriani S, Weinstein JN, Biagini R. Primary
N Am. 2017;28:513–24. bone tumors of the spine. Terminology and surgi-
25. Tatsui CE, Belsuzarri TA, Oro M, et al. Percutaneous cal staging. Spine (Phila Pa 1976). 1997;22(9):
surgery for treatment of epidural spinal cord compres- 1036–44.
sion and spinal instability: technical note. Neurosurg
Focus. 2016;41(4):E2.
Optimizing Wound Healing
in Metastatic Spine Surgery
48
Jaime L. Bernstein, Matthew A. Wright,
and Jason A. Spector
Introduction may involve decompression, debulking of the

tumor, and stabilization of the spine [4].
Neoplasms involving the spine present a chal- Despite the treatment benefits that surgical
lenging and increasingly common problem for interventions offer, wound-healing complica-
the spine surgeon. As treatment options for onco- tions have traditionally been high in patients
logic diagnoses have improved patient survival with spinal metastasis due to multiple factors
from primary malignancy, the incidence of spinal including poor nutritional status, pre- and/or
metastasis has inevitably increased. In fact, the postoperative chemotherapy or radiation, and a
axial skeleton is the third most common location history of previous operations/instrumentation.
of metastasis, and it is the most frequent site of This leaves the surgeon to close within a subop-
bony metastasis [1]. The incidence of spinal timal, atrophic wound bed following tumor
metastasis has been predicted to range from 30% extirpation and placement of instrumentation
to 90% at the time of death depending on the pri- and/or avascular graft material. Moreover, this
mary cancer diagnosis (most commonly from the wound bed is then often subjected to adjuvant
breast, prostate, and lung) [2]. Although treat- chemotherapy and/or radiation [5]. Not surpris-
ment recommendations for metastasis of the ingly, the historical rate of postoperative compli-
spine have evolved over the last decade with a cations in these complex cases is reported in the
concomitant improvement in life expectancy, sur- literature to be as high as 30–40%, with surgical-
gery remains an important component of the site infection being the most common complica-
treatment algorithm as it reduces tumor burden/ tion [5, 6].
recurrence and stabilizes the spine which helps When wound-healing issues arise in this
patients maintain an increased quality of life [3]. patient population, the consequences can be pro-
Current indications for surgical intervention in found and include reoperation, hardware removal,
cases of metastatic disease of the spine include prolonged hospital stay, critical delay in the
progressive neurologic deficit, intractable pain, delivery of adjuvant therapy, decreased patient
spinal instability, and metastasis resistant to radi- quality of life, and increased healthcare costs [4,
ation therapy. To address these issues, surgery 7–10]. The poor postoperative outcomes reported
in the literature, combined with the limited lifes-
pan remaining for many patients with metastatic
J. L. Bernstein · M. A. Wright · J. A. Spector (*) disease of the spine, have traditionally tempered
Division of Plastic Surgery, Department of Surgery, the desire for surgical treatment.
NewYork-Presbyterian/Weill Cornell Medicine, More recently, however, the literature has
New York, NY, USA
shifted toward supporting surgical intervention
e-mail: jas2037@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_48
636 J. L. Bernstein et al.
for a select cohort of patients with metastatic oncerns in the Spinal Metastasis

C
disease to the spine as data show declining Patient
complication rates and excellent functional
outcomes after surgery. However, when com- Wound healing is a complex and multifactorial
pared with nononcologic spinal surgeries, the process, and there are numerous steps where it
complication rate remains significantly higher may become impaired in the patient with spinal
[3]. Ultimately, the risk of complications must metastases either as a direct consequence of the
be weighed against the benefits of the surgery. cancer or indirectly via the treatment modalities
If surgery is selected as a treatment modality, it employed against the tumor with a resultant
is important that the surgical team focuses on increased risk for the development of wound
measures to mitigate wound-healing complica- complications [4, 5, 16, 17]. Cancer is almost
tions beginning in the preoperative period. always associated with some degree of immune
dysregulation regardless of the patient’s primary
tumor type. On a systemic level, tumor cells
Wound Healing produce cytokines which lead to a decreased
ability of the immune system to respond to
General Wound Healing either the tumor or other potential pathogens
[18, 19]. Furthermore, high levels of pro-inflam-
Healing of spinal surgical wounds follows the matory cytokines such as TNF-α, IL-1, and IL-6
traditional model beginning with hemostasis via are frequently associated with adverse clinical
vasoconstriction and formation of a platelet plug. consequences, including muscle wasting,
To briefly summarize, platelet cytokine release in cachexia, and malnutrition. Patients may also
addition to activation of the coagulation and com- become hypoalbuminemic (likely due to TNF-α
plement cascades leads to chemotaxis of inflam- inhibition of albumin production from the liver),
matory cells. Neutrophils predominate within which is associated with postoperative mortality
the first 48 hours, though macrophages become [20, 21].
the dominant cell type driving the inflammatory Another factor to consider in the patient with
response at 48–96 hours. Macrophages initially spinal metastasis is whether he or she has a his-
exhibit an “M1” phenotype, releasing cytokines tory of irradiation either for treatment of the pri-
such as tumor necrosis factor (TNF)-α, interleu- mary tumor or for the metastasis. Irradiation
kin (IL)1, and IL6, resulting in a clearing of debris results in scarring and tissue fibrosis, which may
and pathogens from the wound. Macrophages lead to fusion of tissue planes and loss of elastic-
then switch to an “M2” phenotype, commencing ity. Furthermore, patients who receive irradiation
the fibroproliferative phase. During this phase, within the weeks or months following their spinal
growth factors including transforming growth procedure face additional wound-healing chal-
factor beta (TGF-β), epidermal growth factor lenges, given that irradiation inhibits the neoan-
(EGF), and vascular endothelial growth factor giogenesis and fibroblast proliferation necessary
(VEGF) promote matrix formation, epitheliza- for proper wound healing.
tion, and neoangiogenesis. Collagen production Finally, many patients with spinal metastasis
peaks during this time period, but net collagen will receive steroids to combat cord compres-
deposition equilibrates by around week 3 due to sion. Steroids dysregulate wound healing by
concurrent resorption via collagenase. Over the multiple mechanisms, including decreased col-
following months, collagen maturation occurs lagen deposition by fibroblasts, reduced inflam-
and is characterized by collagen cross-linking mation, and decreased re-epithelization. Other
and turnover of type 3 collagen to type 1 col- comorbidities, which can inhibit proper wound
lagen. In a healthy patient, the wound achieves healing that are not necessarily specific to this
approximately 80% the prewound tensile strength patient population but which must be assessed,
by about 2 months and reaches its final plateau include advanced age, very high or low BMI,
strength at around 6 months [11–15]. tobacco use, diabetes mellitus, renal disease,
48 Optimizing Wound Healing in Metastatic Spine Surgery 637
cardiac disease, peripheral vascular disease, vas- approach to the spine in order to avoid compro-
culitis, and coagulation disorders. mising the blood supply to either the skin or
underlying muscle. Such communication helps
to ensure that the muscle flaps remain a viable
Preoperative Optimization reconstructive option.
Whenever appropriate, the patient should always

be medically and nutritionally optimized prior to Reconstructive Options
surgery. In addition to routine preoperative medi-
cal and cardiac clearance, nutritional optimiza- Although a wide array of reconstructive options
tion should be achieved, with a targeted albumin of the spine exists, the most common approach
of >4 g/dL and prealbumin >16 mg/dL, with the utilizes local muscle flaps given the efficiency of
addition of protein supplementation as necessary. harvest and minimal additional morbidity.
Patients should also be counseled on smoking Coverage with flaps supports wound healing in a
cessation, and nicotine products should be dis- multifaceted manner. First, flaps provide
continued at least 6 weeks prior to surgery. increased vascularity and perfusion to the surgi-
Patients with a history of diabetes mellitus should cal site, which facilitates wound healing, bone
have their blood glucose levels controlled as graft revascularization, and antibiotic delivery.
monitored by their hemoglobin A1c. Muscle flaps also create a soft-tissue barrier
Coagulopathies must be reversed, and any vaso- between the skin and the spine, providing con-
pressors must be discontinued as they decrease tinual protection of the spine and any hardware
perfusion to the wound bed. Finally, steroids that has been placed. In the event of a superficial
should be discontinued as rapidly as possible or dehiscence or even full-thickness skin loss,
supplemented with Vitamin A to counter steroid- wounds have the potential to heal with nonopera-
induced negative wound-healing effects. tive interventions, and salvage of the hardware is
possible given the extra protection afforded by
the layer(s) of muscle flaps from outside expo-
Multidisciplinary Approach sure and contamination (Fig. 48.1). In addition,
local muscle flaps have the added benefit of oblit-
Traditionally, the plastic surgeon becomes erating dead space, which reduces the potential
involved with the complex spinal wound after space for fluid accumulation. Finally, placing
complications have developed, with a resultant muscle tissue adjacent to the exposed spinal cord
reconstructive need that has been associated may also improve dural healing and decrease the
with a high rate of spinal wound morbidity. potential for cerebrospinal fluid (CSF) leak.
Recent literature has demonstrated lower com- Anatomically, there are superficial and deep
plication rates in patients who receive prophy- layers of muscles near the spine available for
lactic coverage of the spine with muscle flaps at soft-tissue coverage (Fig. 48.2). Closure of the
the time of the primary operation when com- wound with muscle flaps almost always includes
pared against delayed reconstruction. Patients the paraspinous muscles given their deep and
with decreased wound-healing capacity should longitudinal orientation to the spine. Depending
be specifically identified so that wound-healing on the wound and spinal levels involved, recon-
complications may be mitigated by providing struction may also call for a second layer of cov-
well-vascularized soft-tissue coverage and a ten- erage from the superficial extrinsic muscles of
sion-free closure at the time of the index opera- the back, including the trapezius, latissimus
tion [4–6, 22]. Preoperatively, through this dorsi, thoracolumbar fascia, and/or gluteus maxi-
collaborative effort, the spine and reconstructive mus [4, 22] (Fig. 48.3). All efforts should be
surgeon can discuss the location of planned sur- made to preserve the perforators supplying the
gical incisions and the anticipated soft-tissue skin overlying these muscle flaps.
In order to optimize wound healing, even in

the setting of muscle flaps, it is important to
make sure all layers of closure are tension free
with sufficient mobility to midline. Tension
reduction can be achieved through advancement
of the deep muscle layers of the spine, which
allows for greater mobility of the more superfi-
cial layers. Relaxing incisions through the lat-
eral muscle fascia can also be considered if
greater tension reduction is still needed at the
time of closure. Tension reduction is crucial as
increased wound tension leads to excess force
on sutures, causing ischemia of the wound edge,
tissue necrosis, and potentially failure of the
suture itself, thereby resulting in wound
dehiscence.
It is important to note that preservation of the
fascia surrounding the muscle flap increases the
strength of the closure, as fascia is more able to
bear tension from the sutures. Muscle alone
often does not provide the necessary strength,
Fig. 48.1 Superficial dehiscence in the early postoperative
and suturing muscle without its investing fascia
period. Because muscle flaps deep to the superficial fascia may result in “cheese-wiring” of the sutures
remain intact, the instrumentation and graft are safely through the muscle, thereby resulting in flap
sequestered even from full-thickness skin breakdown dehiscence.
Trapezius
a Posterior view b Cervical level muscle
Trapezuis
muscle
Paraspinous
muscle
c Thoracic level
Latissimus dorsi
muscle
Paraspinous
muscle
Paraspinous
muscle
Latissimus dorsi
muscle
d Lumbar level Thoracolumbar
fascia
Thoracolumbar
fascia
Paraspinous
muscle
Fig. 48.2 (a) Posterior view of superficial and deep mus- (b) cervical level, (c) thoracic level, and (d) lumbar level.
cles of the back commonly used for flap reconstruction of (With permission from Franck et al. [22], Elsevier ©
the spine. Axial cuts of two-layer flap closures through the 2018)
a b
Fig. 48.3 (a) Open cervical wound with hardware and underlying dead space. (c) Next, the superficial trapezius
bone graft in place. After the deep and superficial muscle muscle flaps are approximated to provide a secondary
planes are dissected, (b) the deep paraspinous muscle layer of coverage over the previously exposed hardware
flaps are approximated and imbricated to obliterate the
evision Spine Surgery

R ing methods of smoking cessation preoperatively,
and Associated Challenges it is important that they are aware the postopera-
tive period is also a critical time where smoking
Not surprisingly, patients who have undergone is detrimental to success of the surgery.
previous spine surgery are at higher risk for Even with preventive measures to mitigate
wound-healing complications. Reconstructive hypoxia to the wound bend, the surgical site
surgeons must consider the possibility that previ- should be closely monitored for signs of hypoxia,
ous incisions may have interrupted blood flow to especially during the early postoperative period.
the soft tissue overlying the spine. Existing mid- Skin color, temperature, turgor, and capillary
line incisions do not usually pose an issue. In refill should be regularly assessed, and any abnor-
contrast, when paramedian incisions or scars mality should be investigated as soon as possible.
deviating from midline are present, it is crucial to Beyond tissue hypoxia, there are numerous other
preoperatively plan the location of the new inci- postoperative parameters that should be moni-
sion so as to maximally preserve blood supply. tored, which can have a direct effect on wound
Previous spinal surgery is not a contraindication healing. Nutritional status should be monitored
for closure with the previously discussed spinal using a blood nutrition panel which tests param-
muscle flaps, but it does mandate close preopera- eters such as prealbumin, and deficiencies should
tive communication between the spine and recon- be promptly corrected/optimized. Protein supple-
structive surgeon. ments should be also considered. And, similar to
the preoperative period, immunosuppressive
medications should be avoided if possible. If ste-
Postoperative Optimization roids must be given, then vitamin A should be
given concurrently to counteract the detrimental
As with any postoperative period, close and regu- effect of steroids on healing.
lar monitoring of both the surgical site and the Pressure should be offloaded from the spine
patient’s medical condition (i.e., vital signs, pain, as much as possible. Using a pillow as a wedge
and mental status) should be performed. In addi- to keep the patient propped to one side or the
tion, prevention or elimination of any potential other and rotating every few hours is an effec-
barriers to basic wound-healing physiology tive way to ensure that there is no unnecessary
should be addressed. pressure being placed on a fresh incision. Both
The primary enemy of any healing wound is the nurse and the patient should be educated
tissue hypoxia. Multiple sources may contribute about pressure offloading during the postopera-
to a hypoxic wound bed, including vasoconstric- tive period. The wound should be kept clean,
tion (secondary to cold, pain, hypovolemia, ciga- and any soiled dressings should be changed
rette smoking, vasopressors), atherosclerosis or immediately.
microvascular disease, anemia, and decreased More recently, many surgeons have opted for
cardiac output (among other potential causes). an incisional subatmospheric dressing, which not
Care must be given to maximizing oxygen deliv- only ensures that the incision remains sterile but
ery to the wound – the wound should be kept also encourages wound healing. Negative pressure
warm, pain should be well managed, and the wound therapy has been shown in the literature to
blood count should be monitored for any drop in facilitate wound healing as it relieves tension from
hematocrit and acted upon as necessary. Fluid the wound edges, decreases fluid accumulation,
status should be optimized to ensure adequate removes infectious material, and promotes the
perfusion of the wound bed and muscle flaps to accumulation of granulation tissue. A newer use of
ensure sufficient oxygen delivery. Although subatmospheric pressure dressings has been over
smoking cessation is ideal in every patient, smok- closed incisions. Known as an incisional subatmo-
ers should especially refrain from doing so dur- spheric dressing, these devices have been shown to
ing the immediate postoperative period. Although decrease tension on the line of closure, lessen
patients can be counseled ahead of time regard- fluid accumulation/seroma formation, and provide
a sterile, airtight environment, all of which con- mas in addition to those causing neurologic defi-
tribute to a decreased rate of dehiscence and infec- cits should be considered for surgical or
tion [23–25]. interventional radiologic drainage.
Handling Wound Complications Cerebrospinal Fluid Leak
Although spinal closures are subject to the poten- Potential signs and symptoms of a CSF leak
tial complications of a standard surgical wound, include headache, photophobia, nausea/vomit-
including seroma, infection, and dehiscence, ing, and wound swelling. If concern for a CSF
these patients must also be monitored for addi- leak develops, all deep drains should be taken
tional complications such as CSF leak and hard- off suction and drained, instead, via gravity into
ware exposure. a bile bag which will allow for excess fluid
evacuation without exerting a suctioning force
within the wound bed. If the drainage remains
Seroma excessive and continues to result in symptoms,
the drain should then be clamped. Other stan-
Muscle flap closure reduces dead space and dard treatments such as horizontal positioning
significantly decreases the risk of seroma for- and decompressive drainage should be per-
mation, but the complication can still occur. formed as necessary.
Placement of appropriately sized drains (e.g.,
#15 Blake) beneath both the muscle and fascio-
cutaneous layers with application of suction Dehiscence/Infection
will allow fluid to drain during the early post-
operative period. A common goal prior to drain As previously mentioned, in the event of a super-
removal is for the 24-hour drain output to ficial dehiscence or even full-thickness skin loss,
remain below 20–30 mL for two consecutive wounds have the potential to heal with nonopera-
days. Our published experience demonstrates tive interventions if muscle flaps have been uti-
that drains may be left safely in place for lized. What would have likely been a deeper
2–4 weeks as necessary and do not increase the space infection that may have involved the hard-
risk of infection [4, 22]. Despite these inter- ware with a simple spinal closure remains a
ventions, seroma can still occur while the drain superficial one that may be treated only with
remains in place or after drain removal. dressing changes or subatmospheric dressings.
Seromas are most frequently present as a pain- Prevention of return to the operating room and
less fluctuant subcutaneous mass, but other salvage of the hardware is possible given the
less common signs and symptoms may include extra layers of protection from outside exposure
new or persistent pain, incisional drainage, or and contamination, which also provide vascular-
even new-onset neurologic deficits secondary ization to the deeper spaces and act as a conduit
to pressure [26]. to ensure adequate antibiotic delivery in the event
Small and clinically stable seromas can be of an infection.
drained at bedside or in the clinic. If the seroma In the event of a suspected or known infection,
persists after multiple drainages, however, surgi- antimicrobial therapy should be initiated after
cal intervention may be necessary as a mature sending cultures. The choice of antimicrobial
cystic lining may prevent fluid resorption and should be guided by cultures and sensitivity
healing of the opposing tissue interfaces. results. Infectious disease physicians should be
Drainage should be inspected closely, and any- consulted as necessary. Operative intervention
thing other than a serosanguinous output (i.e., should be undertaken if there is clear evidence of
frank blood, pus, and CSF) should prompt further involvement of the hardware or if local wound
investigation. Large or rapidly progressing sero- care is inadequate.
References 13. Janis JE, Kwon RK, Lalonde DH. A practi-

cal guide to wound healing. Plast Reconstr Surg.
2010;125(6):230e–44e.
1. Sebaaly A, Shedid D, Boubez G, Zairi F, Kanhonou
14. Broughton G 2nd, Janis JE, Attinger CE. Wound heal-
M, Yuh SJ, et al. Surgical site infection in spi-
ing: an overview. Plast Reconstr Surg. 2006;117(7
nal metastasis: incidence and risk factors. Spine J.
Suppl):1e-S–32e-S.
2018;18(8):1382–7.
15. Glat P, Longaker M. Wound healing. In: Grabb

2. Wong DA, Fornasier VL, MacNab I. Spinal metasta-
and Smith’s plastic surgery. 5th ed. Philadelphia:
ses: the obvious, the occult, and the impostors. Spine
Lippincott-Raven; 1997.
(Phila Pa 1976). 1990;15(1):1–4.
16.
Garvey PB, Rhines LD, Dong W, Chang
3. Demura S, Kawahara N, Murakami H, Nambu K,
DW. Immediate soft-tissue reconstruction for com-
Kato S, Yoshioka K, et al. Surgical site infection in
plex defects of the spine following surgery for spinal
spinal metastasis: risk factors and countermeasures.
neoplasms. Plast Reconstr Surg. 2010;125(5):1460–6.
Spine (Phila Pa 1976). 2009;34(6):635–9.
17. Nasser R, Yadla S, Maltenfort MG, Harrop JS,

4. Cohen LE, Fullerton N, Mundy LR, Weinstein
Anderson DG, Vaccaro AR, et al. Complications in
AL, Fu KM, Ketner JJ, et al. Optimizing suc-
spine surgery. J Neurosurg Spine. 2010;13(2):144–57.
cessful outcomes in complex spine reconstruc-
18. Porporato PE. Understanding cachexia as a cancer
tion using local muscle flaps. Plast Reconstr Surg.
metabolism syndrome. Oncogene. 2016;5:e200.
2016;137(1):295–301.
19. Ohm JE, Carbone DP. VEGF as a mediator of

5. Dolan RT, Butler JS, Wilson-MacDonald J, Reynolds
tumor-associated immunodeficiency. Immunol Res.
J, Cogswell L, Critchley P, et al. Quality of life and
2001;23(2):263–72.
surgical outcomes after soft-tissue reconstruction of
20. Gibbs J, Cull W, Henderson W, Daley J, Hur K, Khuri
complex oncologic defects of the spine and sacrum. J
SF. Preoperative serum albumin level as a predictor
Bone Joint Surg Am. 2016;98(2):117–26.
of operative mortality and morbidity: results from
6. Chang DW, Friel MT, Youssef AA. Reconstructive
the National VA Surgical Risk Study. Arch Surg.
strategies in soft tissue reconstruction after resec-
1999;134(1):36–42.
tion of spinal neoplasms. Spine (Phila Pa 1976).
21. Matthys P, Billiau A. Cytokines and cachexia.

2007;32(10):1101–6.
Nutrition (Burbank, Los Angeles County, Calif).
7. Savage JW, Anderson PA. An update on modifiable
1997;13(9):763–70.
factors to reduce the risk of surgical site infections.
22. Franck P, Bernstein JL, Cohen LE, Hartl R, Baaj
Spine J. 2013;13(9):1017–29.
AA, Spector JA. Local muscle flaps minimize post-
8. McCarthy IM, Hostin RA, Ames CP, Kim HJ, Smith
operative wound morbidity in patients with neo-
JS, Boachie-Adjei O, et al. Total hospital costs of
plastic disease of the spine. Clin Neurol Neurosurg.
surgical treatment for adult spinal deformity: an
2018;171:100–5.
extended follow-up study. Spine J. 2014;14(10):
23. Scalise A, Calamita R, Tartaglione C, Pierangeli M,
2326–33.
Bolletta E, Gioacchini M, et al. Improving wound
9. Pull ter Gunne AF, Cohen DB. Incidence, prevalence,
healing and preventing surgical site complications of
and analysis of risk factors for surgical site infection
closed surgical incisions: a possible role of incisional
following adult spinal surgery. Spine (Phila Pa 1976).
negative pressure wound therapy. A systematic review
2009;34(13):1422–8.
of the literature. Int Wound J. 2016;13(6):1260–81.
10. Whitmore RG, Stephen J, Stein SC, Campbell

24. Cahill C, Fowler A, Williams LJ. The application
PG, Yadla S, Harrop JS, et al. Patient comor-
of incisional negative pressure wound therapy for
bidities and complications after spinal surgery: a
perineal wounds: a systematic review. Int Wound J.
societal-based cost analysis. Spine (Phila Pa 1976).
2018;15(5):740–8.
2012;37(12):1065–71.
25. Ingargiola MJ, Daniali LN, Lee ES. Does the applica-
11.
Janis JE, Harrison B. Wound healing: part
tion of incisional negative pressure therapy to high-
I. Basic Science. Plast Reconstr Surg. 2016;138(3
risk wounds prevent surgical site complications? A
Suppl):9S–17S.
systematic review. Eplasty. 2013;13:e49.
12. Janis JE. Essentials of plastic surgery. 2nd ed. St.
26. Tan LA, Kasliwal MK, Traynelis VC. Surgical

Louis/Boca Raton: Quality Medical Publishing/CRC
seroma. J Neurosurg Spine. 2013;19(6):793–4.
Press/Taylor & Francis Group; 2014.
Contemporary Radiation for Spinal
Metastasis and Spinal Cord
49
Compression
John Roberson, Bernard Newman,
and Samuel Ryu
Introduction definitive therapy resulting in durable pallia-

tion of symptoms and effective local control has
Spine metastases are a common complication arisen. Indeed, recent reports have demonstrated
of many cancers. It is estimated that more than that patients with limited spine metastases treated
40% of all cancer patients develop spine metas- with single-fraction radiosurgery to the involved
tases during the course of their disease [1–3]. spine have 49% 1-year and 35% 3-year survival
Spine metastases, like any other bone metastasis, rates, with varying median overall survivals based
commonly present with pain. However, as they on the primary cancer site, from 1.8 months for
progress, they can cause additional issues such lung primaries to 16 months for breast prima-
as structural problems, due to vertebral com- ries [4]. An additional multi-institutional analy-
pression fractures or bony retropulsion into the sis demonstrated a median overall survival of
spinal canal or neural foramina with associated 19.5 months in a similar group of patients [5].
neurologic deficits. Traditionally, spine metas- Spine metastases commonly occur within the
tases have been considered a major sign of ter- oligometastatic state, where SBRT has shown
minal stage disease, and, therefore, the goal of both improved local control and prolonged sur-
treatment has been palliative, relieving pain, and vival. Therefore, aggressive local treatment, with
attempting to improve neurologic function and the goal of providing durable local control, may
performance status. be strongly warranted in these patients. This
With recent improvements in systemic ther- chapter will focus on providing a review of the
apy and the identification of an oligometastatic management of spinal metastases from the per-
disease state, coupled with advancements in radi- spective of a contemporary radiotherapeutic
ation delivery in the form of stereotactic radiosur- approach.
gery (SRS) or stereotactic body radiation therapy
(SBRT), the treatment paradigm has been gradu-
ally changing toward more aggressive treatment Presentation and Diagnosis
to achieve local tumor control. Furthermore, as
these advances have contributed to improve over- Spine metastases frequently present with severe
all survival for select patients, the need for more back pain that can limit a patient’s ability to
function, resulting in worsened performance
J. Roberson · B. Newman · S. Ryu (*) status. This pain is highly complex as it can be
Department of Radiation Oncology, Stony Brook due to direct tumor involvement of the verte-
bra, mechanical instability, or nerve impinge-
e-mail: Samuel.ryu@stonybrookmedicine.edu

https://doi.org/10.1007/978-3-030-42958-4_49
644 J. Roberson et al.
ment including spinal cord compression, often aforementioned neurologic signs or symptoms of
making it difficult to accurately characterize spinal cord compression should undergo emer-
and consistently assess. Pain from bone involve- gent evaluation of the spine with MRI to evaluate
ment is typically described as non-positional, for the presence of spinal cord compression or
constant, improved with movement, and steroid- canal compromise.
responsive. Mechanical pain (reflecting spinal
instability) is typically more positional and wors-
ened with increased axial load or by bending or pinal Canal Compromise Without
S
standing, though patients with kyphotic defor- Neurological Deficits
mity may instead notice increased pain when
lying flat. Radicular pain consists of a sharp, Epidural tumor causing canal compromise with-
shooting pain in a dermatomal distribution, while out neurologic abnormality can be incidentally
central neurological pain may instead be constant found during staging workup or follow-up evalu-
and escalating with associated neurological dys- ation. If left untreated, this will almost inevitably
function from epidural compression [6]. progress to symptomatic spinal cord compres-
Neurologic compromise can also be the pre- sion. In these cases, shared management deci-
senting symptom as approximately 10% of sions should be made on a relatively urgent basis
patients may ultimately develop spinal cord com- through discussion between spine surgeons,
pression and/or cauda equine syndrome. These radiation oncologists, medical oncologists, and
may manifest with weakness or paresthesia/ patients. Epidural tumor control is imperative to
anesthesia distal to the level of compression and/ prevent the development of neurologic dysfunc-
or inability to control bladder and bowel function and progression to spinal cord compression
tion, resulting in either incontinence or reten- and can be achieved by surgical resection and/
tion. To make an early diagnosis and provide or radiation. This represents an excellent clini-
treatment, patients with a known history of can- cal scenario in which contemporary spine SRS/
cer who develop new-onset spine pain warrant SBRT can play a major role. To evaluate this, a
prompt diagnostic workups with imaging studies reliable grading system (as described below) for
including CT, MRI, and PET. For patients who canal compromise can be utilized.
are found to have a lesion without an oncologic
diagnosis, biopsy or resection of the spine lesion
should also be considered. Spinal Cord Compression
For patients with overt symptomatic spinal cord

Vertebral Metastases compression, glucocorticoids should be admin-
istered immediately. This management step
Spine metastases have long been regarded as an has been shown to improve ambulation in such
oncologic emergency often treated with at least patients [7]. Trials have established the most
one session of fractionated radiation. For patients commonly used regimen consisting of dexameth-
without epidural disease, management can be asone 10 mg IV bolus followed by 16 mg per
made on a non-emergent basis with the urgency day (4 mg every 6 hours), tapered over 2 weeks
of management determined, at least in part, on [8]. MRI should be obtained promptly to visual-
the ability to control pain pharmacologically. ize the extent of spinal cord compression and to
Regardless, appropriate medical pain manage- allow for treatment planning. For patients found
ment is always required. With the use of aggres- to have spinal cord compression on a localized
sive local treatment to the involved spine, the first MRI, imaging of the complete spine to evalu-
decision that must be made when assessing spine ate for multilevel cord compression is highly
metastases is whether or not emergent treatment advised to allow for a more comprehensive treat-
is warranted. Any patient who presents with the ment decision. Multimodality spine tumor board
49 Contemporary Radiation for Spinal Metastasis and Spinal Cord Compression 645
discussion is highly recommended to coordinate along with their performance status, general
available treatment options and formulate an condition, and comorbidities, is required to
optimal, individualized management strategy. prognosticate patients’ overall oncologic status.
Radiosensitivity of the primary tumor is also
an important factor in decision-making. Many
I ncidentally Found Asymptomatic patients who present with metastatic spinal
Lesions cord compression will have known pathology,
allowing for rapid estimation of radiosensitiv-
Recently, MRIs have become more widely used ity. When there is no previous diagnosis of can-
for evaluation of spinal metastases, and due to cer, initial surgical management may be both
their increased ability to detect lesions, they have therapeutic and diagnostic. Radioresponsive or
resulted in finding incidental occult or small vol- chemoresponsive tumors include lymphomas,
ume vertebral lesions without associated symp- seminomas, small cell lung cancers, and mul-
toms. In some instances, these lesions represent tiple myelomas and can be treated with either
oligometastatic disease where SRS/SBRT can radiation or chemotherapy alone [10]. All other
actually lead to prolonged survival, thereby open- tumors will require radiation either postopera-
ing a new realm for SBRT treatment of spine tively or as the sole treatment.
lesions [9]. Though decisions regarding treatment Conventional fractionation external beam radi-
options in this particular clinical scenario are not ation therapy (cEBRT) is delivered with the most
well established, SRS/SBRT to the involved spine common regimen of 30 Gy in 10 fractions. An
can be the treatment of choice. Even so, a mul- early prospective trial of cEBRT alone for spinal
tidisciplinary approach is necessary to facilitate cord compression conducted in the 1980s–1990s
treatment decision-making and coordinate this found that patients with myeloma or lymphoma
portion of the patient’s total oncologic care. remained ambulatory in 100% of cases (10/10)
in which they were ambulatory prior to treatment
and regained ambulation in 64% of cases (7/11)
Considerations of Management in which they were non-ambulatory prior to treat-
ment [11]. As will be seen below, these numbers
Definitive treatment of spine metastases and/or are comparative to more radioresistant tumors
spinal cord compression after initial imaging and treated by either SRS/SBRT alone or surgery
initiation of glucocorticoids typically consists of followed by cEBRT. For this reason, patients
either surgery with adjuvant radiation or radia- with radioresponsive tumors have been excluded
tion alone. Unlike other bone metastases, spine from subsequent trials evaluating management
treatment evaluation should take into consider- options.
ation the presence of spinal instability, degree
of spinal cord compression, radiosensitivity of
the tumor, and duration and rapidity of symptom Prior Treatment
development. Other factors, including control of
systemic disease, number of levels involved by Accurate information regarding any previous
disease, and whether the patient has received spine treatment is also required when decid-
prior local treatment to the area, also play a role ing management options, particularly whether
in determining the management. patients have received any prior local therapy
with surgery or radiation. cEBRT treatments
typically include 30 Gy in 10 fractions delivered
Oncologic Assessment to the 1–2 spine segments above and below the
involved segments. Any previous treatments have
Assessment of the entire tumor burden includ- also delivered similar radiation doses to adjacent
ing the primary site and all metastatic sites, normal structures, including the spinal cord, lung,
esophagus, bowel, etc. Retreatment of the spine signaling poor overall surgical outcome may also
often is challenging in the setting of previous full- impact the advisability and necessity for spinal
dose spine radiation. In many instances, these stabilization.
patients may be retreated with nearly full-dose
radiation assuming that enough time has passed
since the initial treatment. Yet, in some instances, uration and Severity of Neurologic
D
patients will develop local progression or recur- Symptoms
rence after only a relatively short period after the
initial radiation treatment. Consideration of radia- The duration and severity of neurologic symp-
tion dose, time interval, and target volume are toms also play a significant role in determining
important factors when considering retreatment, the proper management of spinal metastases.
as well as patient’s symptoms, oncologic status, As noted above, whether patients with spinal
general condition, and availability of other treat- cord compression are ambulatory is an impor-
ments. As will be discussed below, current evi- tant hallmark of whether they will be ambula-
dence suggests a similar response to SRS/SBRT tory after treatment. This is seen in the literature
following initial cEBRT compared with de novo regardless of histology or treatment method with
SRS/SBRT. cEBRT alone, SRS/SBRT alone, or surgery fol-
lowed by cEBRT. Since these neurologic symp-
toms develop as a result of extrinsic compression
Spinal Stability on the spinal cord, it is believed that the longer
a patient experiences such symptoms, the more
An essential component in the assessment of likely they are to develop irreversible neurologic
spinal cord compression and vertebral column damage and the less likely they are to return to
metastases is the structural stability of the spinal their previous level of functioning. Indeed, an
column. No amount of radiation or chemother- analysis regarding the timing of surgery follow-
apy – no matter the histology – will stabilize an ing onset of neurologic symptoms due to meta-
unstable spine. The Spinal Instability Neoplastic static spinal cord compression demonstrated
Score (SINS) has been developed to provide improved outcomes if patients underwent sur-
guidance in determining which patients have the gery within 48 hours of symptom onset, and it
greatest instability and uses the following cri- also established a negative correlation between
teria: vertebral body location, presence of pain, delaying surgery and neurologic improvement
type of bone lesion, spinal alignment on imaging, [13]. Thus, both the severity and duration of
presence of vertebral body collapse and extent symptoms and the rapidity of their development
of vertebral body involvement, and involvement play a significant role in determining further
of the posterior elements, with total scores rang- management.
ing from 0 to 18. Lesions are then classified as
“stable” (0–6), “potentially unstable” (7–12),
and “unstable” (13–18) with recommendation Extent of Spinal Involvement
for surgical consultation for all lesions that are
unstable or potentially unstable (i.e., score > 6) Both surgical intervention and radiation ther-
[12]. Though the SINS may provide guidance apy are directed to the involved spinal segment,
in determining the necessity for stabilization of meaning that an important aspect in determin-
the spine, note that it is not a perfect classifica- ing treatment options is the extent of spinal
tion system and lesions classified as “stable” involvement. Furthermore, since the goal of
may actually behave as if they were “unstable.” spine SRS/SBRT is to maximize local tumor
Finally, even if a patient is determined to have control and preserve neurologic function while
an “unstable lesion,” factors like extensive sys- providing durable palliation, it is important to
temic disease or severe medical comorbidities properly identify patients who may most ben-
efit from this treatment approach (as opposed to decision-making, subdividing grade 1 into three
cEBRT). Much like in the setting of stereotactic separate grades, whereas Ryu et al. used grades
radiosurgery for brain metastases, there is no 0–V based on the extent of the epidural lesion.
clear-cut answer as to the number of spine lev- The two available grading systems are essen-
els that can be treated with spine radiosurgery. tially identical radiographically (Fig. 49.1). In
RTOG (NRG) 0631 includes different clinical both systems, involvement of the vertebral bone
scenarios: (1) solitary spine metastasis with or only is grade 0, with the other grades represent-
without epidural or paraspinal soft tissue exten- ing epidural impingement with no thecal sac
sion; (2) two contiguous spine levels; (3) non- deformation (Bilsky grade 1a = Ryu grade I),
adjacent spine metastases (generally up to three thecal sac deformation without spinal cord abut-
sites); (4) diffuse metastases along the spinal ment (Bilsky grade 1b = Ryu grade II), spinal
column; and (5) multiple spinal level involve- cord abutment (Bilsky grade 1c = Ryu grade
ment with very small “occult” lesions visible III), partial spinal cord compression with visible
only on MRI scan within the vertebral bod- CSF (Bilsky grade 2 = Ryu grade IV), and com-
ies (the size of these small lesions is defined plete block with no visible CSF (Bilsky grade
as being less than 20% of the vertebral body). 3 = Ryu grade V) [14, 15]. In order to provide
One future area of development includes defin- further assistance with making clinically ori-
ing how to best handle these occult metastases. ented treatment decisions, Ryu et al. developed
Of note, caution must also be taken with larger a dual grading system which takes into account
targets (such as scenario 3 above) as the spinal both radiographic spinal cord compression and
cord dose tends to increase more with individ- the neurological deficits [15]. In this system,
ual targets larger than 6 cm [4]. A paravertebral patients receive a radiographic grade from 0
mass extending from an involved spine can also to V, which scales the degree of canal compro-
be treated with spine radiosurgery, though there mise and cord compression, and a neurologic
is again no guideline as to the size of paraver- grade from a to e, routinely used in clinic, cor-
tebral involvement. In RTOG 0631, 5 cm was responding to the Tomita functional motor grade
used as the size cut off for paraspinal masses, [16]. Using this system, for example, grade IIIc
but it must be decided whether these paraspi- indicates the spinal cord is impinged by tumor
nal lesions can be safely included in the target and the patient has mild muscle weakness with
volumes. Such factors also affect the decision- 4/5 power. This grading system is thus useful
making to undergo surgery. for treatment decision-making and monitoring
epidural lesions and neurological status after
treatment.
pinal Cord Compression Grading
S
Systems
Treatment Options
With the routine use of MRI to evaluate epidural
extension and spinal cord compromise and the The treatment paradigm for spine metastases
more widespread availability of SRS/SBRT and and/or spinal cord compression with surgery and
sophisticated surgical methods to treat spine radiation has been changing over the last decade,
metastases, attempts have been made to develop with contemporary options including cEBRT,
grading of spinal cord compression. There are multiple fractionation schedules of radiosurgery,
two main systems developed for evaluation and various surgical or other interventional pro-
of malignant spinal cord compression, one by cedures, as well as combinations of these treat-
Bilsky et al. based on MR images and another ments in the form of multimodality therapy. It
by Ryu et al. based on both MRI configuration is thus important to understand how the various
and neurological status. Bilsky et al. developed a treatment modalities may be employed in the
four-point grading system to assist with surgical management of spine metastases.
648
Dual grading system (Ryu’s) Surgical grading system

a (Bilsky’s) b
Radiographic grade Neurological grade
a No abnormality A At spinal cord level
b Minor symptoms (eg, pain,

radiculopathy, sensory
change) 1c
1b
0 1a Grade 0: No epidural Grade I: Minimal Grade II: Thecal
c Functional paresis tumor, spine bone mets epidural lesion, thecal sac impinged, but
Muscle power ≥ 4/5. can be present sac not abutted spinal cord intact
-nerve root sign or spinal

B
cord sign present
-functional in the upper
0 Spine bone metastasis extremity muscles
No epidural lesion -ambulatory in the lower Grade Ill: Spinal cord Grade IV: Partial block Grade V: Complete
extremity muscles 2 impinged of CSF, Compression CSF block, Frank
I Minimal canal of spinal cord cord compression
compromise, Thecal can
be impinged d Non-Functional paresis At cauda level
Muscle power ≤ 3/5.
II Thecal sac compressed, C
but spinal cord is intact -non-functional in the upper
extremity Grade II: ≤50% Grade IV: >50%
III Spinal cord impinged
-non-ambulatory in the 3 canal compromise canal compromise
IV Spinal cord lower extremity
displaced/compressed,
CSF visible, Partial block
e Paralysis, Incontinence
V CSF not visible,
Complete block
Fig. 49.1 Comparison of the proposed grading systems of spinal cord compression: Ryu’s dual radiographic and neurological system (a and b with schematic drawing of
radiographic grade), and Bilski’s grading system
J. Roberson et al.
Surgical Management treatment. As in all radiation treatments, daily

or less frequent image guidance is used, though
Surgical treatment options include laminectomy, in rare circumstances patient alignment can be
separation surgery, and direct decompressive sur- done clinically. Since cEBRT uses greater mar-
gery with corpectomy. Stabilization of the spine, gins around the involved level, it may be used to
often as a required component of surgery, may treat patients in extreme uncontrolled pain who
also be accomplished through different methods are unable to comfortably lie still. Over the past
of instrumentation and reconstruction. Significant two decades, the practice of radiation oncology
bony retropulsion due to a compression fracture has changed considerably with the use of radio-
should be managed surgically rather than with surgery for spine metastases and spinal cord com-
radiation. Likewise, structural spine instability pression showing promising results, leading to
should also be evaluated and managed surgically. ongoing large randomized trials.
Separation surgeries are intended to create a gap
between the epidural tumor and the spinal cord,
not necessarily removing the entire tumor, in Surgery Followed by Radiation
order to allow for the full radiosurgery dose to be
delivered to the gross epidural tumor. Regardless Following a prospective randomized trial carried
of the type of surgery employed, postoperative out in the 1990s, surgery with direct decompres-
radiation is essential to prevent local recurrence sive corpectomy followed by cEBRT to 30 Gy in
and tumor progression. Further discussion of 10 fractions became the standard of care for med-
the surgical management of spine metastases is ically operable patients with at least 3 months
reserved for another chapter. to live and non-radiosensitive tumors [19]. The
endpoint of this trial was the ability to ambu-
late, defined as four steps with use of an assistive
onventional External Beam
C device. The trial showed significant improve-
Radiation Therapy ment in overall ambulatory rate with surgery plus
cEBRT (84%) compared to cEBRT alone (57%)
Conventional external beam radiation therapy and an improved duration of ambulation of
(cEBRT) has been used for palliation of pain and 123 days for surgery plus cEBRT versus 13 days
for spinal cord compression for decades. Studies by cEBRT alone. However, in the subset analy-
have shown that multiple fractionation regimens sis, only 62% of non-ambulatory patients became
are acceptable and provide equivalent response. ambulatory after surgery. This suggests that
The most common regimen has been 30 Gy in patients’ initial ambulatory status is an important
10 fractions using simple radiation field arrange- prognostic factor of functional outcome.
ments. It appears that patients treated with a sin- With the development of SRS/SBRT in the
gle fraction (e.g., 8 Gy in 1 fraction) as opposed 2000s, the role of postoperative SBRT was also
to multifraction regimens (e.g., 3 Gy × 10 frac- evaluated. The initial clinical experience was
tions) are more likely to require retreatment (21.5 reported in 2006 and included 18 patients treated
vs. 7.4%) and develop pathologic fractures [17]. with SRS to 14–16 Gy in a single fraction deliv-
However, a more recent randomized trial compar- ered 1–2 weeks after open surgery. This demon-
ing 8 Gy × 1 with 4 Gy × 5 showed no difference strated that postoperative SRS is well tolerated
in outcomes for patients with spinal cord compres- and associated with minimal morbidity [20].
sion [18]. In general, the target volume includes However, one of the main difficulties with post-
the involved spine segments as well as 1–2 ver- operative SRS involves delineation of the tumor
tebral bodies above and below the intended target and spinal cord in the setting of recently placed
spine levels. The procedure for cEBRT includes instrumentation hardware, resulting in poor
assessing targeting accuracy, delineating tumor CT and MR image quality. Because of this, CT
and normal tissues, and planning and delivering myelograms have also been used to provide better
delineation of the spinal cord. In an effort to help 13.3 months [28]. Others have also demonstrated
define the postoperative target volume for SRS/ similar results regarding pain control in patients
SBRT, consensus guidelines were also developed with spine metastases [27, 29–31]. Quality of
by 10 experts from high-volume institutions [21]. life also was improved following pain control
The general consensus is to cover preoperative [27]. Local tumor control at the treated spine was
sites of osseous and epidural disease irrespective achieved in 95% of patients with recurrence at
of the extent of surgical resection, including gross the immediately adjacent vertebra in less than 5%
residual disease seen on postoperative imaging [26]. Patients with oligometastatic disease had
and the adjacent tissue at risk for microscopic longer survival with more effective local treat-
disease extension, with PTV expansions vary- ment of the spine metastasis [4]. These clinical
ing from 0 to 2.5 mm. The spinal cord avoidance results support the idea that more intensive treat-
structure should also be subtracted from the final ment may be appropriate for patients with local-
PTV for treatment planning. Surgical instrumen- ized spine metastases in order to improve their
tation and the incision do not need to be included clinical outcome and overall quality of life.
unless believed to be specifically at risk for tumor The overall procedure for spine SRS/SBRT
involvement. includes patient positioning and immobiliza-
There is also concern raised about potential tion with imaging to provide targeting accuracy.
“underdosing” of portions of epidural tumor Ensuring proper immobilization is the first step
immediately adjacent to the spinal cord due to toward the clinical application of spine SRS/
the inherent radiosurgical dose gradient in the SBRT. Initial techniques included invasive pro-
spinal cord. To ensure that the full radiosurgical cedures that anchored hardware to the cervical
dose covers the epidural tumor while maintaining spine and skull or required a stereotactic frame
a safe spinal cord dose, separation surgery has to be attached under general anesthesia to the
been developed in which the strategy is both to lumbar spinous processes [32]. Another early
decompress the spinal cord to allow full radiosur- technique developed used a body frame with a
gical dosing and also to stabilize the spine. This contour mold fixation [33]. More recently, a
strategy has been recommended for patients with frameless and noninvasive positioning method,
high-grade epidural spinal cord compression or used by most institutions, has been developed
previously irradiated tumors [22]. An analysis of [25, 34]. There is no perfect method for immo-
patients treated in this manner at Memorial Sloan bilization. It is important to provide stability and
Kettering demonstrated local tumor progression support to patients in a treatment position that
of 4–9% at 1 year following treatment with either is comfortable. While breathing-related organ
24 Gy in a single fraction or 24–30 Gy in three motion exists, it does not appear to affect treat-
fractions [23, 24]. ment outcomes. Other voluntary and involuntary
movements, such as swallowing, coughing, and
pulsation, also occur, some of which may be con-
Stereotactic Radiosurgery trolled with premedication. Another important
clinical scenario involves spine pain which may
Stereotactic radiosurgery (SRS or SBRT) was ini- lead to random and unexpected patient move-
tially developed as a primary treatment modality ment. It is therefore important to properly man-
for spine metastases in the late 1990s. Multiple age spine pain with short-term pain medication
single institution experiences have reported prior to initiating the procedure.
durable and rapid pain control in the range of The radiation doses used for spine SRS/
90%. Median time to pain relief has been found SBRT also vary among investigators. An ini-
to be just 2 weeks after treatment with select tial analysis of patients at Henry Ford Hospital
patients experiencing pain relief within 24 hours treated between May 2001 and May 2003 found
[25–27]. Median duration of pain control in the a strong trend toward improved pain control with
treated region of the spine has been shown to be higher doses, particularly of 14 Gy or more [28].
Based on this experience, RTOG 0631 adopted The most critical normal structure for spine
16–18 Gy in one fraction as the standard dose to SBRT is the spinal cord. In order to delineate the
be evaluated [35]. Another analysis of patients spinal cord, T1-weighted contrast-enhanced and
treated with single-fraction SRS at Memorial T2-weighted MR images are fused to the CT simu-
Sloan Kettering suggested improved local control lation images with 1–2 mm slice thickness. Due to
for patients treated with 23–24 Gy compared to the nature of radiosurgery with rapid dose fall-off,
lower doses [36]. It is important to note that the there is a degree of radiation dose gradient within
dose prescription methods are different between the diameter of the spinal cord. The accumulated
institutions, prescribing to either the tumor mar- dose volume analysis of the spinal cord in 230
gin or the isocenter. There are other fractionated procedures by Ryu et al. demonstrated a partial-
regimens for spine SRS/SBRT, such as 24–30 Gy volume tolerance of the spinal cord of 10 Gy to the
in three fractions or 30–40 Gy in five fractions, 10% cross-sectional area of the spinal cord (cor-
while Canadian investigators are also testing responding to 0.35 cc of the spinal cord volume),
24 Gy in two fractions in a national clinical trial. provided that the spinal cord is defined as 6 mm
In terms of target delineation for SRS/SBRT, it above and below the radiosurgery target volume.
is important to recognize that each vertebra con- The reason for using this partial volume includes
sists of compact bone and marrow within a tra- the rapid dose fall-off with SBRT from the 90% to
beculated network that extends from the vertebral 50% isodose lines of 5 mm when coplanar SBRT
body into the pedicles. Thus, while imaging may beams are used [4]. When non-coplanar beams are
demonstrate gross disease within the vertebral used, the absolute volume criteria are also used.
body (i.e., GTV), this may not represent the full Other investigators developed slightly different
extent of tumor involvement. We, therefore, recom- criteria to define the spinal cord dose, including
mend using a clinical target volume (CTV) which a maximum dose of 12–14 Gy at the surface of
consists of the involved elements of each vertebral an MRI-defined spinal cord or a maximum dose
body as the target. Examples of target volumes are of 10 Gy to a myelogram- defined spinal cord
illustrated in Fig. 49.2, adapted from Ryu et al. [4], [36, 38]. Taken together, these dose constraints
showing the most common cases of vertebral body appear to be in a similar range. It is also impor-
involvement. Guidelines for the delineation of such tant to delineate other surrounding normal tissues,
targets were adopted in the RTOG 0631 trial and including the laryngopharynx, trachea, esophagus,
have been further defined by consensus among bowel, and kidneys, with recommended dose con-
experts at high-volume institutions [37]. straints having also been published [39, 40].
a b c
Fig. 49.2 Target delineation models for SRS/SBRT to more extensive lesions can be treated either with generous
spine metastases. (a) The most common form of spine margins (dotted line) or by including both anterior and
metastasis involving the vertebral body. (b) Involvement posterior elements (solid line). (c) Involvement of the pos-
of the vertebral body with extension into the pedicles; terior elements only [4]
Decision-Making Algorithm tionation [41]. Meanwhile, initial results with

radiosurgery suggest that pain control is seen in
Historically, there was not sufficient need to around 85–90%. A retrospective review of 500
develop radiation treatment algorithms since lesions treated with spine SRS at the University
spine metastases were treated palliatively with of Pittsburgh to a median of 20 Gy in 1 fraction
cEBRT with or without surgery. However, with found long-term pain control of 86% [42]. A
recent advances in systemic therapy, targeted phase II dose escalation trial carried out at Henry
therapy, immunotherapy, radiosurgery, and open Ford Hospital found increased pain control with
surgery, treatment algorithms that aid decision- doses above 14 Gy in 1 fraction with 1-year actu-
making will be essential, particularly in patients arial pain control of 84%. Duration of pain con-
with oligometastatic disease. An interdisciplin- trol after SBRT appears to be around 13 months
ary decision-making framework was developed [28]. It appears that higher doses seen with radio-
which used components of neurologic, oncologic, surgery provide superior and more durable pain
mechanical instability, and systemic disease control than conventional palliative regimens.
(NOMS) [22]. An update from the International These observations served as the basis for the
Spine Oncology Consortium provided greater phase II/III RTOG 0631 trial, comparing 8 Gy in
guidance based on the status of the systemic dis- one-fraction cEBRT with 16 Gy or 18 Gy in one-
ease, including overall performance status, tumor fraction SRS. The result of this large randomized
systemic burden, and further treatment options study showed equal rate of pain control between
[10]. When the results of ongoing large clinical the two arms due to the unexpected low rate of
trials assessing SRS/SBRT and the treatment of pain control in the radiosurgery arm (presented at
oligometastases become available, more detailed ASTRO 2019). Patients should be evaluated peri-
and practical guidelines should be developed. odically to evaluate pain control after treatment.
Providers must be diligent in distinguishing
radiation site-specific pain from other pain com-
reatment Outcomes
T plaints [6]. Patients with recalcitrant pain should
be referred to a pain management specialist.
Evaluation of treatment outcomes must take into
consideration the patient’s presenting symptoms
as well as pretreatment imaging findings. For Epidural Tumor Control
patients who present with back pain (a very com- and Neurologic Compromise
mon presenting symptom), the goal of treatment
is primarily to relieve the back pain. For patients The goals of treatment for spinal cord compres-
who present with neurologic symptoms and/ sion (or spinal canal compromise) are control of
or evidence of epidural tumor extension, treat- the epidural disease, decompression of the spinal
ment goals include decompression of the epi- cord, and preservation or improvement of neu-
dural tumor and thecal sac expansion as well as rological function. A prospective randomized
improving or maintaining neurologic function. trial carried out by Patchell et al. in the 1990s
for patients with a single site of metastatic spi-
nal cord compression with expected survival of
Control of Metastatic Spine Pain at least 3 months and who were not paraplegic
for more than 48 hours found that direct decom-
Pain response after cEBRT, regardless of the pressive surgery followed by cEBRT(30 Gy in 10
radiation dose used, appears to be approxi- fractions), compared to cEBRT alone, resulted
mately 50–60% including 30–35% with a com- in a better ambulatory rate for all patients (84%
plete response [17]. Trials also indicate that the vs. 57%), for patients who were ambulatory prior
median duration of pain control after cEBRT is to treatment (94 vs. 74%), and for patients who
around 3–4 months regardless of treatment frac- were not ambulatory prior to treatment (62 vs.
19%) [19]. While this trial defined ambulation Table 49.1 Comparison of clinical trial results of sur-
as the ability to take four steps with the use of gery versus radiosurgery
an assistive device, it is uncertain whether this Ryu’s
relates practically to actual independent walking, phase II
Patchell’s phase trial
though it is certainly important for activities such III trial [19] [15]
as transferring from bed to chair or commode and Surgery
back. This trial established surgical decompres- + cEBRT SRS
sion followed by cEBRT as the standard of care cEBRT alone alone
for this group of patients. Overall 84% 57% Overall 81%
ambulatory (42/50) (29/51) intact (50/62)
With advancements in radiation treatment, a rate rate
phase II trial was carried out by Ryu et al. in the Ambulatory 94% 74% Intact pts 94%
2000s assessing SRS to a dose of 14–20 Gy in pts remain (32/34) (26/35) remain (33/35)
a single fraction in the management of epidural ambulatory intact
spinal cord compression. The results demon- Non- 62% 19% Deficit 63%
ambulatory (10/16) (3/16) pts (17/27)
strated that the mean reduction in epidural tumor pts improve to improve
volume was 65% at 2 months with an overall ambulatory to intact
epidural tumor response rate of 80%, including cEBRT conventional external beam radiation therapy,
27% complete response, 30% partial response (> SRS stereotactic radiosurgery
50% tumor volume reduction), and 23% mini-
mal response (25–50% reduction). This resulted Table 49.1. The rates of ambulatory or neuro-
in decreased epidural tumor area at the level of logically intact patients remaining ambulatory or
most severe decompression and improved the- intact were 94% after either surgery or SRS. The
cal sac patency [15]. Importantly, the study also rates of non-ambulatory or neurological deficit
demonstrated that 94% (33 of 35) of patients patients improving to ambulatory or intact were
who were intact before radiosurgery remained 63% after either treatment. That said, treatment
intact, and 63% (17 of 27) of patients who had decisions must be individualized for each patient
neurologic deficits prior to radiosurgery demon- as discussed in this chapter and elsewhere in this
strated improvement. Excellent results for pain book. The NOMS and SINS frameworks are use-
relief and improvement in neurologic deficits ful, but providers must also integrate the patient’s
were obtained for multiple myeloma causing the- performance status, prognosis, and available sys-
cal sac compression, which commonly occurs temic treatment options in the decision regarding
in this disease [43]. An additional retrospective optimal spinal metastasis management.
analysis with high-grade spinal cord compression
(radiographic grades IV–V) treated with SRS to
18 Gy in one fraction was performed from the Re-irradiation Outcomes
Henry Ford Hospital Database. Only 18% (6 of
33) deteriorated within 2 months of treatment Approximately 50% of patients treated with
and 67% retained their ambulatory status [44]. A cEBRT require re-irradiation following the ini-
more recent effort to relax the spinal cord dose tial treatment [46]. Re-irradiation with cEBRT
in patients with spinal cord compression showed is often done at lower doses than initial treat-
encouraging results for SRS with a spinal cord ments and tends to be less effective, with overall
maximum dose up to 16 Gy in a phase I clinical response rates of 45–51% and complete response
trial [45]. rates of only 11–14% [47]. By contrast, retreat-
Although it is not possible to compare two ment with SRS/SBRT provides local control rates
separate trials, the results of surgery versus radio- >75% with a modest toxicity profile comparable
surgery are comparable in preserving or improv- to de novo SBRT [48]. For example, in an analy-
ing the neurological outcome. The subset results sis of patients treated with salvage SBRT after
of Patchell’s and Ryu’s trials are summarized in in-field failure of initial SBRT (24 of whom had
already received prior cEBRT as well), the study steroids prophylactically [52], we do not do so.
demonstrated a 1-year local control rate of 81% In fact, we taper steroids immediately after the
with no radiation-induced vertebral compression procedure for those who were already receiving
fractures or myelopathy observed [49]. SBRT steroids. The cause of the pain flare is unknown,
therefore is a good option for recurrent tumors. but it is advised to limit the radiation dose to the
spinal nerve root under 14 Gy.
Other side effects from radiation are related
Treatment Complications to incidental treatment of neighboring normal tis-
sues, including the larynx, pharynx, esophagus,
Treatment Failure bowel, lung, kidneys, etc. and should thus be
delineated. Symptoms may manifest as dyspha-
Treatment failures after spine SBRT can be gia, odynophagia, nausea, and bloating, depend-
divided into three different categories: in-field ing in large part on what part of the spine is treated.
failures (regrowth within the target volume), Of note, we limit the esophagus to 10–12 Gy in
marginal failures (progression within the region a single dose [39]. These side effects are gener-
of rapid dose fall-off surrounding the target vol- ally self-limiting and resolve within weeks after
ume), and distant failures. Each of these failures treatment, but long-term development of a tra-
may have distinctive causes: for example, in-field cheoesophageal fistula has been reported [53].
failures due to the inherent radioresistance of To avoid acute complications, efforts should be
tumors, marginal failures due to errors in patient taken to minimize radiation doses to mucosal
setup or underestimation of the target volume, structures.
and distant failures due to continued progression
of metastatic disease.
Previous studies have demonstrated low rates Long-Term Neurologic Complications
of in-field and marginal failures following SBRT
of 5–6% [26, 38]. This low incidence of in-field Radiation-induced damage to the spinal cord can
and marginal failures helps justify the use of SRS/ severely adversely affect patients’ quality of life.
SBRT. It should also be noted that persistent or It is therefore imperative to make every effort to
progressive pain may not be a good indicator for avoid unnecessary and/or excessive radiation to
tumor progression, as there may be other causes the spinal cord. As noted above, the spinal cord
including spinal instability and degenerative dis- partial-volume tolerance dose has been defined
orders that contribute to this pain. as 10 Gy to the 10% partial volume of the spinal
cord defined as including 6 mm above and below
the target volume (calculated to be equivalent to
Acute Complications 0.35 cc) [4]. The spinal cord tolerance dose can
vary depending on the fractionation scheme.
Acute exacerbation of pain occurs usually Other factors include host factors, comorbidi-
within 1–5 days after treatment, known as a pain ties, oncologic status, and previous treatments.
flare. The incidence of pain flares after SBRT is Regardless, it is always advised to minimize the
reported to be 20–60% and can occur anywhere radiation dose to any of the normal tissues.
from the day after treatment until 20 days later
[50]. Some have also found an increase in the
incidence of pain flares after single-fraction SRS Long-Term Non-neurologic
compared to multiple fractions [51]. Fortunately, Complications
pain flares are typically transient and very
responsive to a short course of low-dose gluco- Vertebral compression fracture has been reported
corticoids (e.g., dexamethasone 4 mg daily for to occur in about 10–15% of patients [54–56]. A
up to 5 days). Although some advocate using similar rate of vertebral compression fractures
After Kyphoplasty
Fig. 49.3 An example of worsening of epidural compression after kyphoplasty
is also seen for SBRT delivered after cEBRT it more difficult to treat with SRS. An example of
[48]. Analyses of predictive factors have found this case is shown in Fig. 49.3. We recommend
that patients with prior kyphotic/scoliotic defor- initial SRS/SBRT for those who have any onco-
mity, lytic lesions, and receiving higher doses (≥ logical issues unless management of the com-
20 Gy) are at a higher risk of developing verte- pression fracture or instability is more critical
bral compression fractures. In these instances, for total care of the patient after discussion at a
consideration must also be given to the extent of multidisciplinary spine tumor board.
involvement of the vertebra, the presence of other
degenerative changes, and whether the patient is
symptomatic from the compression deformity. References
Some patients are candidates for kyphoplasty
prior to SBRT when there is concern of devel- 1. Hernandez RK, Adhia A, Wade SW, et al. Prevalence
of bone metastases and bone-targeting agent use
oping a compression fracture particularly if they among solid tumor patients in the United States. Clin
have pain and no evidence of epidural extension Epidemiol. 2015;7:335–45. https://doi.org/10.2147/
[57]. We have observed worsening of spinal cord CLEP.S85496.
compression by upfront kyphoplasty in patients 2. Coleman RE. Clinical features of metastatic bone dis-
ease and risk of skeletal morbidity. Clin Cancer Res.
with an existing spinal cord compression, making
2006;12(20):6243s–9s. https://doi.org/10.1158/1078- J. 2013;22(6):1383–8. https://doi.org/10.1007/

0432.CCR-06-0931. s00586-012-2635-y.
3. Dunne EM, Fraser IM, Liu M. Stereotactic body 14. Bilsky MH, Laufer I, Fourney DR, et al. Reliability
radiation therapy for lung, spine and oligometastatic analysis of the epidural spinal cord compression
disease: current evidence and future directions. Ann scale. J Neurosurg Spine. 2010;13(3):324–8. https://
Transl Med. 2018;6(14):283. https://doi.org/10.21037/ doi.org/10.3171/2010.3.SPINE09459.
atm.2018.06.40. 15. Ryu S, Rock J, Jain R, et al. Radiosurgical decom-
4. Ryu S, Jin J-Y, Jin R, et al. Partial volume tolerance pression of metastatic epidural compression.
of the spinal cord and complications of single-dose Cancer. 2010;116:2250–7. https://doi.org/10.1002/
radiosurgery. Cancer. 2007;109(3):628–36. https:// cncr.24993.
doi.org/10.1002/cncr.22442. 16. Tomita T, Galicich JH, Sundaresan N. Radiation

5. Guckenberger M, Mantel F, Gerszten PC, et al. Safety therapy for spinal epidural metastases with complete
and efficacy of stereotactic body radiotherapy as block. Acta Radiol Oncol. 1983;22(2):135–43. http://
primary treatment for vertebral metastases: a multi- www.ncbi.nlm.nih.gov/pubmed/6310968. Accessed
institutional analysis. Radiat Oncol. 2014;9(1):226. 27 Sept 2018.
https://doi.org/10.1186/s13014-014-0226-2. 17. Sze WM, Shelley M, Held I, Mason M. Palliation
6. Thibault I, Chang EL, Sheehan J, et al. Response of metastatic bone pain: single fraction versus
assessment after stereotactic body radiotherapy for multifraction radiotherapy. Cochrane Database
spinal metastasis: a report from the SPIne response Syst Rev. 2002;2:CD004721. https://doi.
assessment in Neuro-Oncology (SPINO) group. org/10.1002/14651858.CD004721.
Lancet Oncol. 2015;16(16):e595–603. https://doi. 18. Hoskin P, Misra V, Hopkins K, et al. SCORAD III:
org/10.1016/S1470-2045(15)00166-7. Randomized noninferiority phase III trial of single-
7. Sørensen S, Helweg-Larsen S, Mouridsen H, Hansen dose radiotherapy (RT) compared to multifraction
HH. Effect of high-dose dexamethasone in carcino- RT in patients (pts) with metastatic spinal canal
matous metastatic spinal cord compression treated compression (SCC). J Clin Oncol. 2017;35(18_
with radiotherapy: a randomised trial. Eur J Cancer. suppl):LBA10004. https://doi.org/10.1200/
1994;30A(1):22–7. http://www.ncbi.nlm.nih.gov/ JCO.2017.35.18_suppl.LBA10004.
pubmed/8142159. Accessed 27 Sept 2018. 19. Patchell RA, Tibbs PA, Regine WF, et al. Direct

8. Vecht CJ, Haaxma-Reiche H, van Putten WL, de decompressive surgical resection in the treatment of
Visser M, Vries EP, Twijnstra A. Initial bolus of spinal cord compression caused by metastatic cancer:
conventional versus high-dose dexamethasone in a randomised trial. Lancet. 2005;366(9486):643–8.
metastatic spinal cord compression. Neurology. https://doi.org/10.1016/S0140-6736(05)66954-1.
1989;39(9):1255–7. http://www.ncbi.nlm.nih.gov/ 20. Rock JP, Ryu S, Shukairy MS, et al. Postoperative
pubmed/2771077. Accessed 27 Sept 2018. radiosurgery for malignant spinal tumors.
9. Palma DA, Olson RA, Harrow S, et al. Stereotactic Neurosurgery. 2006;58(5):891–8. https://doi.org/10.
ablative radiation therapy for the comprehen- 1227/01.NEU.0000209913.72761.4F.
sive treatment of oligometastatic tumors (SABR- 21. Redmond KJ, Robertson S, Lo SS, et al. Consensus
COMET): results of a randomized trial. Int J Radiat contouring guidelines for postoperative stereo-
Oncol. 2018;102(3):S3–4. https://doi.org/10.1016/j. tactic body radiation therapy for metastatic solid
ijrobp.2018.06.105. tumor malignancies to the spine. Int J Radiat
10. Spratt DE, Beeler WH, de Moraes FY, et al. An
Oncol. 2017;97(1):64–74. https://doi.org/10.1016/j.
integrated multidisciplinary algorithm for the man- ijrobp.2016.09.014.
agement of spinal metastases: an International 22. Laufer I, Rubin DG, Lis E, et al. The NOMS frame-
Spine Oncology Consortium report. Lancet Oncol. work: approach to the treatment of spinal metastatic
2017;18(12):e720–30. https://doi.org/10.1016/S1470- tumors. Oncologist. 2013;18(6):744–51. https://doi.
2045(17)30612-5. org/10.1634/theoncologist.2012-0293.
11. Maranzano E, Latini P. Effectiveness of radiation
23. Laufer I, Iorgulescu JB, Chapman T, et al. Local dis-
therapy without surgery in metastatic spinal cord ease control for spinal metastases following “sepa-
compression: final results from a prospective trial. Int ration surgery” and adjuvant hypofractionated or
J Radiat Oncol Biol Phys. 1995;32(4):959–67. http:// high-dose single-fraction stereotactic radiosurgery:
www.ncbi.nlm.nih.gov/pubmed/7607970. outcome analysis in 186 patients. J Neurosurg Spine.
12. Fisher CG, DiPaola CP, Ryken TC, et al. A novel clas- 2013;18(3):207–14. https://doi.org/10.3171/2012.11.
sification system for spinal instability in neoplastic SPINE12111.
disease. Spine (Phila Pa 1976). 2010;35(22):E1221– 24.
Moussazadeh N, Laufer I, Yamada Y, Bilsky
9. https://doi.org/10.1097/BRS.0b013e3181e16ae2. MH. Separation surgery for spinal metastases: effect
13. Quraishi NA, Rajagopal TS, Manoharan SR, Elsayed of spinal radiosurgery on surgical treatment goals.
S, Edwards KL, Boszczyk BM. Effect of timing Cancer Control. 2014;21:168–74. http://journals.sage-
of surgery on neurological outcome and survival pub.com/doi/pdf/10.1177/107327481402100210.
in metastatic spinal cord compression. Eur Spine Accessed 2 Oct 2018.
25. Ryu S, Fang Yin F, Rock J, et al. Image-guided and radiosurgery. Int J Radiat Oncol. 2012;83(5):e597–
intensity-modulated radiosurgery for patients with 605. https://doi.org/10.1016/j.ijrobp.2012.03.009.
spinal metastasis. Cancer. 2003;97(8):2013–8. https:// 38. Chang EL, Shiu AS, Mendel E, et al. Phase I/II
doi.org/10.1002/cncr.11296. study of stereotactic body radiotherapy for spinal
26. Ryu S, Rock J, Rosenblum M, Kim JH. Patterns of metastasis and its pattern of failure. J Neurosurg
failure after single-dose radiosurgery for spinal metas- Spine. 2007;7(2):151–60. https://doi.org/10.3171/
tasis. J Neurosurg. 2004;101(Suppl 3):402–5. https:// SPI-07/08/151.
doi.org/10.3171/jns.2004.101.supplement3.0402. 39. Schipani S, Wen W, Jin J-Y, Kim JK, Ryu S. Spine
27. Degen JW, Gagnon GJ, Voyadzis J-M, et al.
radiosurgery: a dosimetric analysis in 124 patients
CyberKnife stereotactic radiosurgical treatment of who received 18 Gy. Int J Radiat Oncol Biol Phys.
spinal tumors for pain control and quality of life. 2012;84(5):e571–6. https://doi.org/10.1016/j.ijrobp.
J Neurosurg Spine. 2005;2(5):540–9. https://doi. 2012.06.049.
org/10.3171/spi.2005.2.5.0540. 40. Benedict SH, Yenice KM, Followill D, et al.

28. Ryu S, Jin R, Jin J-Y, et al. Pain control by image- Stereotactic body radiation therapy: the report of
guided radiosurgery for solitary spinal metastasis. J AAPM Task Group 101. Med Phys. 2010;37(8):4078–
Pain Symptom Manag. 2008;35(3):292–8. https://doi. 101. https://doi.org/10.1118/1.3438081.
org/10.1016/j.jpainsymman.2007.04.020. 41. Gaze MN, Kelly CG, Kerr GR, et al. Pain relief and
29. Gerszten PC, Burton SA, Ozhasoglu C, et al.
quality of life following radiotherapy for bone metas-
Stereotactic radiosurgery for spinal metastases tases: a randomised trial of two fractionation sched-
from renal cell carcinoma. J Neurosurg Spine. ules. Radiother Oncol. 1997;45(2):109–16. http://
2005;3(4):288–95. https://doi.org/10.3171/ www.ncbi.nlm.nih.gov/pubmed/9423999. Accessed
spi.2005.3.4.0288. 27 Sept 2018.
30. Gerszten PC, Burton SA, Welch WC, et al. Single- 42. Gerszten PC, Burton SA, Ozhasoglu C, Welch
fraction radiosurgery for the treatment of spinal breast WC. Radiosurgery for spinal metastases. Spine
metastases. Cancer. 2005;104(10):2244–54. https:// (Phila Pa 1976). 2007;32(2):193–9. https://doi.
doi.org/10.1002/cncr.21467. org/10.1097/01.brs.0000251863.76595.a2.
31. Gerszten PC, Burton SA, Quinn AE, Agarwala 43. Jin R, Rock J, Jin J-Y, et al. Single fraction spine

SS, Kirkwood JM. Radiosurgery for the treat- radiosurgery for myeloma epidural spinal cord com-
ment of spinal melanoma metastases. Stereotact pression. J Exp Ther Oncol. 2009;8(1):35–41. http://
Funct Neurosurg. 2005;83(5–6):213–21. https://doi. www.ncbi.nlm.nih.gov/pubmed/19827269. Accessed
org/10.1159/000091952. 5 Dec 2018.
32. Hamilton AJ, Lulu BA, Fosmire H, Stea B, Cassady 44. Lee I, Omodon M, Rock J, Shultz L, Ryu S. Stereotactic
JR. Preliminary clinical experience with linear radiosurgery for high-grade metastatic epidural cord
accelerator-based spinal stereotactic radiosurgery. compression. J Radiosurg SBRT. 2014;3(1):51–8.
Neurosurgery. 1995;36(2):311–9. http://www.ncbi. http://www.ncbi.nlm.nih.gov/pubmed/29296385.
nlm.nih.gov/pubmed/7731511. Accessed 27 Sept Accessed 1 Oct 2018.
2018. 45. Ghia AJ, Guha-Thakurta N, Hess K, et al. Phase 1
33.
Lax I, Blomgren H, Näslund I, Svanström study of spinal cord constraint relaxation with single
R. Stereotactic radiotherapy of malignancies in the session spine stereotactic radiosurgery in the pri-
abdomen. Methodological aspects. Acta Oncol. mary management of patients with inoperable, pre-
1994;33(6):677–83. http://www.ncbi.nlm.nih.gov/ viously unirradiated metastatic epidural spinal cord
pubmed/7946448. Accessed 27 Sept 2018. compression. Int J Radiat Oncol Biol Phys. 2018;
34. Yin F-F, Ryu S, Ajlouni M, et al. A technique of 102(5):1481–8. https://doi.org/10.1016/j.ijrobp.2018.
intensity-modulated radiosurgery (IMRS) for spinal 07.2023.
tumors. Med Phys. 2002;29(12):2815–22. https://doi. 46. Hartsell WF, Scott CB, Bruner DW, et al. Randomized
org/10.1118/1.1521722. trial of short- versus long-course radiotherapy for pal-
35. Ryu S, Pugh SL, Gerszten PC, et al. RTOG 0631 liation of painful bone metastases. J Natl Cancer Inst.
phase 2/3 study of image guided stereotactic radio- 2005;97(11):798–804. https://doi.org/10.1093/jnci/
surgery for localized (1-3) spine metastases: phase 2 dji139.
results. Pract Radiat Oncol. 2014;4(2):76–81. https:// 47. Chow E, van der Linden YM, Roos D, et al. Single
doi.org/10.1016/j.prro.2013.05.001. versus multiple fractions of repeat radiation for pain-
36. Yamada Y, Bilsky MH, Lovelock DM, et al. High- ful bone metastases: a randomised, controlled, non-
dose, single-fraction image-guided intensity- inferiority trial. Lancet Oncol. 2014;15(2):164–71.
modulated radiotherapy for metastatic spinal lesions. https://doi.org/10.1016/S1470-2045(13)70556-4.
Int J Radiat Oncol. 2008;71(2):484–90. https://doi. 48. Myrehaug S, Sahgal A, Hayashi M, et al. Reirradiation
org/10.1016/j.ijrobp.2007.11.046. spine stereotactic body radiation therapy for spinal
37. Cox BW, Spratt DE, Lovelock M, et al. International metastases: systematic review. J Neurosurg Spine.
spine radiosurgery consortium consensus guidelines 2017;27(4):428–35. https://doi.org/10.3171/2017.2.S
for target volume definition in spinal stereotactic PINE16976.
49. Thibault I, Campbell M, Tseng C-L, et al. Salvage Oncol Biol Phys. 2012;83(5):e661–7. https://doi.
stereotactic body radiotherapy (SBRT) following org/10.1016/j.ijrobp.2012.01.080.
in-field failure of initial SBRT for spinal metastases. 54. Boyce-Fappiano D, Elibe E, Schultz L, et al. Analysis
Int J Radiat Oncol. 2015;93(2):353–60. https://doi. of the factors contributing to vertebral compres-
org/10.1016/j.ijrobp.2015.03.029. sion fractures after spine stereotactic radiosurgery.
50. Chiang A, Zeng L, Zhang L, et al. Pain flare is a Int J Radiat Oncol. 2017;97(2):236–45. https://doi.
common adverse event in steroid-naïve patients org/10.1016/j.ijrobp.2016.09.007.
after spine stereotactic body radiation therapy: a 55. Cunha MVR, Al-Omair A, Atenafu EG, et al.

prospective clinical trial. Int J Radiat Oncol Biol Vertebral compression fracture (VCF) after spine
Phys. 2013;86(4):638–42. https://doi.org/10.1016/j. stereotactic body radiation therapy (SBRT): analy-
ijrobp.2013.03.022. sis of predictive factors. Int J Radiat Oncol Biol
51. Pan HY, Allen PK, Wang XS, et al. Incidence and Phys. 2012;84(3):e343–9. https://doi.org/10.1016/j.
predictive factors of pain flare after spine stereotactic ijrobp.2012.04.034.
body radiation therapy: secondary analysis of phase 56. Sahgal A, Atenafu EG, Chao S, et al. Vertebral com-
1/2 trials. Int J Radiat Oncol. 2014;90(4):870–6. pression fracture after spine stereotactic body radio-
https://doi.org/10.1016/J.IJROBP.2014.07.037. therapy: a multi-institutional analysis with a focus
52. Chow E, Meyer RM, Ding K, Nabid A, Chabot P, on radiation dose and the spinal instability neoplastic
Wong P, et al. Dexamethasone in the prophylaxis of score. J Clin Oncol. 2013;31(27):3426–31. https://
radiation-induced pain flare after palliative radiother- doi.org/10.1200/JCO.2013.50.1411.
apy for bone metastases: a double-blind, randomised 57. Gerszten PC, Germanwala A, Burton SA, Welch

placebo-controlled, phase 3 trial. Lancet Oncol. WC, Ozhasoglu C, Vogel WJ. Combination kypho-
2015;16(15):1463–72. https://doi.org/10.1016/ plasty and spinal radiosurgery: a new treatment
S1470-2045(15)00199-0. paradigm for pathological fractures. Neurosurg
53. Cox BW, Jackson A, Hunt M, Bilsky M, Yamada Y. Focus. 2005;18(3):e8. http://www.ncbi.nlm.nih.gov/
Esophageal toxicity from high-dose, single-fraction pubmed/15771398. Accessed 2 Oct 2018.
paraspinal stereotactic radiosurgery. Int J Radiat
Intraoperative Radiation for Spinal
Metastatic Disease
50
Brandon S. Imber, Michael R. Folkert,
and Yoshiya Yamada
Introduction and Rationale and this dose typically dictates the maximum

for Intraoperative Radiotherapy possible dose of external beam radiotherapy
(EBRT), which can be offered. However, pro-
Intraoperative radiotherapy (IORT) is a general spective series with image-guided intensity
term that describes the delivery of therapeutic modulated radiation therapy (IMRT) and proton
doses of radiation during surgical resection of therapy have explored raising this constraint to
a tumor to the sites of greatest likelihood of 54 Gy at the center of the cord and up to 63 Gy
recurrence (i.e., suspicion of microscopic posi- to the surface of the cord over a length of up to
tive margins). This can involve utilization of a 5 cm using conventional fractionation [2]. The
small linear accelerator inside a shielded oper- advantage of IORT compared to EBRT is that a
ating or treatment room, which delivers rela- highly ablative dose of radiation can be deliv-
tively superficial radiation dose to a surgical ered to the area at risk with relative sparing of
cavity, generally with electrons or low-energy nearby radiosensitive structures (e.g., spinal
photons. Alternatively, IORT may also be deliv- cord and cauda equina). This can be achieved
ered using brachytherapy techniques, which are by using forms of radiation with high dose rate
defined by the placement of radioactive sources but low penetrative ability (e.g., electron beams
directly into, or in close proximity to, a tumor or very low energy photons). For brachyther-
or tumor bed. apy, nearby tissue sparing is possible due to the
The risk of radiation myelitis is one of the fact that dose rate falls off rapidly in accordance
most feared complications of spinal radio- with the inverse square law, which states that
therapy. The classic spinal cord tolerance is the intensity of light is inversely proportional to
typically felt to be around 45–50 Gy in conven- the square of the distance. This results in rapid
tional or standard fractions (e.g., 1.8–2 Gy per reductions in delivered dose within very short
fraction) or 13–15 Gy in a single fraction [1], distances from the surface of the brachytherapy
source.
IORT can also be used to supplement dose
B. S. Imber · Y. Yamada (*)
Department of Radiation Oncology, Memorial Sloan delivered with EBRT given pre- or postopera-
Kettering Cancer Center, New York, NY, USA tively, where it can be an effective strategy for
e-mail: yamadaj@mskcc.org dose escalation since it enables greater focal
M. R. Folkert dose delivery to tumors within the vertebral
Department of Radiation Oncology, University of body or epidural space [3, 4]. Brachytherapy
Texas Southwestern Medical Center,
Dallas, TX, USA

https://doi.org/10.1007/978-3-030-42958-4_50
660 B. S. Imber et al.
is particularly versatile for the following com- therapy techniques [8]. Today, numerous IORT
plex clinical situations: strategies exist to address complex intracranial
and spinal pathologies.
(a) When there is a relatively radioresistant
tumor histology (e.g., chordoma, renal cell
carcinoma or sarcoma) where effective Physics of Accelerator-Based
tumoricidal doses exceed the cord tolerance and Brachytherapy IORT
(b) When circumferential disease involvement is
present around the spinal dura Since numerous IORT approaches have been
(c) In retreatment or salvage settings when the employed for spinal disease, it is first important
spinal cord has already received high radia- to understand potential treatment considerations.
tion doses and additional external beam radi- Accelerator-based approaches require a linear
ation could result in exceeding the tolerance accelerator to generate either therapeutic X-rays
of the cord or electrons [9]. A treatment cone or applicator
is usually placed directly in a surgical cavity (at
risk for microscopic residual disease after gross
Historical Context total resection) or potentially against areas of
residual gross disease. The energy and type of
Brachytherapy is one of the oldest forms of ther- radiation generated dictates how deeply the dose
apeutic radiation and was first utilized for cen- penetrates beyond the tumor bed.
tral nervous system tumors in 1912 by Hirsch Treatment of a spinal tumor using brachyther-
with catheter injection of radium into the sella apy requires appropriate selection of a source,
turcica [5]. The neurosurgeon Harvey Cushing which is the radioactive material that delivers
implanted a “radium bomb” into an intracra- radiation dose. Table 50.1 highlights some of
nial surgical cavity [6], but his disappointment the common source isotopes utilized for spinal
with brachytherapy’s modest results led him to brachytherapy. There are several parameters that
quickly abandon the approach. Interest in central influence source selection:
nervous system brachytherapy was strengthened
by the development of stereotactic guidance in • Dose rate. Spinal IORT can utilize either high
the 1950s, which enabled precise implantation of dose rate (HDR) or low dose rate (LDR)
radioactive sources into inoperable brain tumors radioisotopes. As the name suggests, LDR
[7]. Over the ensuing decades, further advance- sources decay and deliver dose more slowly,
ment in stereotactic approaches and better image typically between 0.4 and 2 Gy per hour,
guidance enabled further refinement of brachy- whereas HDR brachytherapy by definition
Table 50.1 Common spinal brachytherapy sources and their physical properties
Radioactive Decay Mean energy of decay Dose rate HVL Half-life
isotope emission (MeV) (Gy/hour) (in water) (days)
Iridium-192 Gamma 0.380 65 mm 73.83
Iodine-125 X-ray 0.028 Permanent: 0.07 17 mm 59.4
Temporary:
0.5–0.6
Cesium-131 X-ray 0.030 Permanent: 0.34 18 mm 9.7
Phosphorus-32 Beta 0.695 40–802 Range in 14.28
(Max = 1.7) water = 7 mm
Yttrium-90 Beta 0.934 40–80 Range in 2.67
(Max = 2.27) water = 12 mm
Samarium-53 Beta 0.225 Range in bone = 1.93
1 mm
50 Intraoperative Radiation for Spinal Metastatic Disease 661
delivers dose at 12 Gy per hour or more [10]. • Source physical properties. In general,
Selection of HDR versus LDR is often patient sources release radioactivity through decay
specific and is influenced by several factors and release of photons (X-rays), neutrons,
including the size and topography of expected gamma rays, or charged particles such as
post-resection tumor, IORT dosing require- alpha or beta particles (helium nuclei or
ments, radiobiological assumptions of the electrons). Source encapsulation will greatly
tumor and adjacent normal tissues, prior radi- reduce any component of dose from alpha
ation exposure, institutional and physician or beta particles or neutrons for most clini-
expertise, workflow requirements, and radia- cally used sources—the greatest component
tion safety concerns. of dose arises from gamma rays. The physi-
• Source format. Sources can be implanted per- cal properties of the characteristic emissions
manently into a tumor or tumor cavity to of isotopes such as mean energy of the emis-
slowly release radioactivity or can be placed sion, half value-layer (HVL), and half-life are
temporarily to deliver a pre-specified amount well characterized and determine the amount
of radiation. In general, permanently surgi- of time required for treatment and anatomic
cally implanted sources are usually LDR. For penetrance of the radioactivity. For example,
spinal IORT, sources take several forms such a source with a very rapid dose fall off would
as small three-dimensional seeds or two- likely be selected for a patient whose disease
dimensional foil plaques that can be directly closely abuts the thecal sac. Figure 50.1 shows
positioned on the dural surface. the dose profiles of commonly employed iso-
100%
P-32
Y-90
90%
Pd-103
80% I-125
50 kVp (Axxent)
70% 50 kVp (Intrabeam/kypholORT)
Percent depth dose (%)
Ir-192
60%
6 MeV electrons
50%
40%
30%
20%
10%
0%
0 2 4 6 8 10 12 14 16 18 20
Depth [mm]
Fig. 50.1 Depth dose curves for various spinal IORT Axxent – surface assumed at 1.5 cm from center of source;
radiation sources are shown. Depth for all sources is spec- 6 MeV electron – depth in tissue with 1 cm bolus applied;
ified in tissue with references as follows: P-32 – surface of and I-125/Pd-103/Ir-192 – interstitial implant assumed
the foil; Y-90 surface of the plastic applicator; 50 kVp with depth specified from the plane of the sources
Intrabeam – surface of 2 cm planar applicator; 50 kVp
topes. The dose profiles of high-energy X-ray Summary of the Literature

emitting sources (e.g., Iridium-192) closely
approximate the predicted inverse square fall In the first series of 37 patients, of which 59%
off with small attenuation and scatter differ- also received neoadjuvant or adjuvant EBRT,
ences. Conversely, low-energy X-ray emit- they reported that all patients had some degree of
ting sources (e.g., Iodine-125) penetrate less improvement in pain, neurologic function status,
deeply due to photoelectric attenuation in tis- or both. No local failures were noted, and about
sue. Sources releasing charged particles (e.g., one third of patients developed distant spinal met-
Phosphorus-32) have even steeper dose pro- astatic disease. One case of radiation myelopathy
files and generally deliver the majority of their was noted in a patient where the spinal cord was
dose within millimeters of the source surface. not shielded with lead. In an updated series of
• Dose specification. For any IORT procedure 79 lesions treated with posterior decompressive
(including accelerator-based techniques), surgery followed by 20 Gy IORT, 86% clinical
the total dose of radiation to be delivered, as improvement in one domain of the Frankel clas-
well as the reference point for the delivery of sification scale with 2.5% local recurrence was
that dose, needs to be pre-specified. For reported [12]. Again, one case of myelopathy was
example, a total dose of 8 Gy might be deliv- noted in a heavily pretreated patient. In a retro-
ered at the surface of a plaque. The degree to spective review of 96 patients who were nonam-
which the surrounding tissue receives dose bulatory due to severe cord compression and who
is dictated by the aforementioned physical underwent posterior decompression with IORT,
properties. nearly 90% of patients had some degree of neuro-
logic improvement and 80% regained ambulatory
status after surgery [13] (Table 50.2).
I ntraoperative Electron Beam
Treatment
LDR Brachytherapy
Technical Considerations for the Management of Spinal
Lesions
The Tokyo Metropolitan Komagome Hospital
has reported one of the largest experiences using Technical Considerations
electron beam IORT [11–13]. Decompressive
surgery with or without preoperative emboliza- Iodine-125 (I-125) has been the most widely reported
tion was performed, though the authors specified source used for spinal LDR brachytherapy. Loose
that total resection of the tumor was often impos- seeds can be implanted using a Mick applicator
sible. After hemostasis is achieved, the patients (Mick Radio-Nuclear Instruments, Mount Vernon,
were transferred to the radiotherapy department NY) or embedded on a suture and affixed using bio-
where an electron cone (applicator) was placed logic adhesive [14, 15]. Stereotactic localization can
directly in the surgical field. A custom 3–5 mm be used to assemble the seeds into circular or linear
lead block was placed over the spinal cord to arrays. Numerous methods of securing the placed
protect it from electron dose (Fig. 50.2). Each seeds have been reported, including methyl meth-
patient received 20 Gy in a single fraction. The acrylate, staples, sutures, Gelfoam, or direct fixa-
electron energy was determined by measuring tion to surgical hardware [16]. Correct placement of
the anteroposterior thickness of the tumor by the seeds is critical, as the dose given by an I-125
magnetic resonance imaging (MRI) so that the seed near the spinal cord is very high, risking local
80% isodose line falls at least 1–2 cm below the myelitis. Reports from two groups describe wrap-
deepest aspect of the tumor. After treatment, the ping the thecal sac with layers of either platinum
patient was returned to the operating room for [14] or gold [17] metal foil to shield the spinal cord
wound closure. from radiation from the implanted seeds.
a b
d
c
Fig. 50.2 Photographic and schematic representations (c) View of the cone before attachment to the gantry with
of intraoperative radiotherapy. (a) View of the irradiation a midline lead shield for the spinal cord. (d) Schematic
room. A: anesthesia machine; B: gantry, C: cone; D: cartoon of intraoperative radiotherapy. (From Kondo
operating table and patient; E: image from the camera et al. [13]. Reprinted with permission from Wolters
located in the cone. (b) Cone set up in the surgical field. Kluwer Health, Inc.)
Hamilton and colleagues delivered 120 Gy Summary of the Literature

to an in-field thoracic spinal recurrence of chon-
drosarcoma in a 28-year-old patient who had Gutin et al. published one of the earliest reports,
previously received 45 Gy [17]. The tumor was which analyzed 14 heavily pretreated patients
resected, and the thecal sac was wrapped in two who underwent re-resection and brachytherapy
0.025 mm thick layers of gold foil, after which for recurrent paraspinal or skull base tumors [14].
point the I-125 seeds were sutured into the tumor Prescribed doses ranged from 70 to 150 Gy, and
bed. The prescription dose was 120 Gy at 5 mm common histologies included chordoma (36%)
depth, and the gold foil was predicted to reduce and meningioma (21%). Outcomes were modest;
cord dose to less than 5% over the life span of the two-thirds of evaluable patients at 6 months fol-
implant. Although the use of a foil is an excellent lowing IORT had local progression.
protection for the spinal cord, it has the disad- Kumar et al. have reported the use of I-125 in
vantage of potentially shielding the dural surface, the management of previously irradiated clival
which in many tumors with epidural extension is and sacral chordomas [18]. The sacral lesion
at risk for tumor contamination. received 160 Gy and had excellent control until
Table 50.2 Summary of the literature using IORT for malignant spinal diseases
First author Dose
(year) rate Isotope/source N Dose, Gy Notes
Gutin (1987) LDR I-125 13 70–150 Platinum foil cord shield
[14] (permanent) 14% long-term remission
Kumar (1988) LDR I-125 2 160, 400 1 patient with 3-year OS; 1 patient NED
[18] (permanent) 19 months
Armstrong LDR I-125 14 125 50% LC overall, 47% OS at 1 year, 12% at
(1991) [15] (permanent) 2 years
LDR Ir-192 21 30 NSCLC and dural involvement negatively
(temporary) prognostic for LC
Hamilton LDR I-125 1 120 Gold foil spinal cord shielding
(1995) [17] (permanent)
Rogers LDR I-125 30 50–160 2- and 3-year LC were 87% and 73%, but
(2002) [16] (permanent) most patients did not receive post-
brachytherapy imaging
Majority received adjuvant EBRT and no
myelopathy
Yao LDR I-125 24 Median D90 of 99 Salvage reirradiation treatment using
(2016) [21] (permanent) (range 90–176) percutaneous seed implantation
6- and 12-month LC rates were 52% and
40%
Delaney HDR Ir-192 3 10 1/3 NED
(2003) [25] YT-90 5 10 1/5 NED
Folkert HDR Ir-192 5 14 (12–18) 100% LC at 9 months
(2013) [30] 80% palliation reported 1–4 weeks
post-procedure
Folkert HDR P-32 68 10 LR 18.5% with plaque vs 34% without
(2015) [25] (p = 0.04)
No IORT-associated myelopathy
Cardoso HDR Sm-153 19 3 mCi of Sm-153 Kyphoplasty performed with radionuclide
(2009) [32] mixed with bone impregnated bone cement
cement No reported complications or hematologic
toxicity
100% pain reduction, no discussion of LC
Saito IORT 11–20 MeV 74 20 Posterior epidural decompression followed
(2006) [12] Electron beam by single-fraction electron beam therapy
97.5% LC
86% improvements in pain, neurologic
function, or both
Kondo IORT Electron beam 96 20–30 All patients initially nonambulatory, and
(2008) [13] 89% regained neurologic status and 80%
became ambulatory after treatment
Bludau IORT 50 kV X-rays Phase 8 Gy at 8–13 mm Phase I/II dose escalation study of
(2018) [39] I: 9 depth kypho-IORT
Phase No dose-limiting toxicities
II: 52 Significant improvements in pain on VAS
3- and 12- month LC of 98% and 94%
Abbreviations: OS overall survival, NED no evidence of disease, LC local control, EBRT external beam radiotherapy,
VAS visual analog scale
the patient ultimately succumbed to meningeal Larger LDR experiences were published by
chondromatosis at 3 years. A transnasal approach the Memorial Sloan Kettering Cancer Center
was used to deliver 400 Gy to a small clival recur- (MSKCC) [15] and the Barrow Neurologic Institute
rence using two I-125 seeds. This patient was (BNI) [16]. MSKCC reported the treatment of 35
reported to be well 19 months after the procedure. patients who underwent brachytherapy following
incomplete resection of a paraspinal lesion. IORT the use of this technique for primary paraspinal
utilized permanent I-125 seed placement (40%), or lesions [19] or as salvage therapy for reirradia-
temporary single-plane implants using Iridum-192 tion of spinal metastases following prior EBRT
(Ir-192) delivered 3–6 days after tumor resection [20, 21]. In the reirradiation setting, 26 lesions
via afterloading catheters (60%). Numerous meta- were contoured and pre-planned using simulation
static and primary histologies were treated, includ- computed tomography (CT) in the prone posi-
ing non-small-cell lung cancer (51%) and sarcoma tion 3–5 days prior to brachytherapy [21]. Dose
(26%), and 60% of patients had received previ- was prescribed as D90 (dose delivered to 90%
ous EBRT. Median doses were 30 Gy for patients of the clinical target volume), and seeds were
delivered with Ir-192 and 125 Gy for patients deliv- implanted percutaneously, in a linear arrange-
ered with I-125. Median estimated cord dose for ment, under local anesthesia into the paraspinal
Ir-192 treatments was 20 Gy. Local control (LC) lesions using a Mick applicator. Post-implant
was achieved for 51% of patients with median dosimetry revealed median actual D90 of 99 Gy
time to local failure of 1.3 years. However, over- (range 90–176) with median maximal dose to the
all survival (OS) of the cohort was poor; only two spinal cord of 39 Gy (range 6–111). At a median
patients were alive with local control at the IORT follow-up of 9.5 months, they reported actuarial
site 3 years following the procedure. Surgeries LC rates at 6 and 12 months of 52% and 40%,
requiring exposure of the dura and NSCLC histol- respectively. Nearly all patients reported some
ogy were negatively prognostic for local control. degree of pain relief following brachytherapy
The authors did not report any cases of radiation after 1–3 weeks, and the overall rate of neurologic
myelitis but acknowledge the poor OS. functional recovery or retention using American
Rogers and colleagues summarized the treat- Spine Injury Association (ASIA) grading was
ment of 30 patients at BNI who underwent para- reported as nearly 80%. In general, the brachy-
spinal surgery for metastatic cord compression therapy was well tolerated, with no myelopathy
followed by IORT with permanent I-125 seeds in reported. Three patients (13%) suffered vertebral
absorbable sutures [16]. The majority of the eval- compression fractures 3–6 months after brachy-
uable patients (56%) had received prior EBRT, therapy without concomitant tumor progression.
and most (88%) of them also underwent adju-
vant EBRT following IORT. They report 2-year
and 3-year local control rates of 87% and 73%, HDR Brachytherapy
respectively. Of note, most of the patients were for the Management of Spinal
surveilled clinically after IORT. Only 40% of Lesions
evaluable patients had posttreatment imaging, but
they report four radiographic local failures (16%) Background and Technical
at a mean time of 20 months after IORT. Three Considerations
of these failures occurred in patients who under-
went IORT as salvage therapy after previous Over the past decade, several changes have
EBRT. OS was again poor, with 2-year OS rate impacted the delivery of paraspinal radiation. First,
of 24%. They report good functional improve- improvements in image-guided radiotherapy have
ment with 84% of patients having either normal fostered the growth of u ltrahypofractionated EBRT
or improved ambulation following surgery. No techniques. These strategies, also known as stereo-
myelopathies or radiculopathies were noted. tactic body radiotherapy (SBRT) or stereotactic
A group from Peking University Third ablative radiotherapy (SABR), have enabled the
Hospital has described usage of CT-guided inter- safe delivery of very high doses of radiation in very
stitial brachytherapy using percutaneously placed close proximity to the spinal cord [22]. Second, the
I-125 seeds for a variety of paraspinal lesions. growth of conformal approaches such as proton
Their retrospective experience has described therapy enables dose escalation to resistant spinal
histologies with relative sparing of the spinal cord relative biologic effectiveness (RBE) to the
[23, 24]. Despite these advancements, the treatment cord surface and center, respectively. Of the
of contaminated dura or epidural surface remains a eight treated patients, 75% had local control at
significant clinical challenge. Specifically, the risk median follow-up of 2 years post-procedure.
of spinal cord myelitis sets an upper limit for the The group updated their experience in a more
acceptable dose at the dural edge, which is often recent abstract [26], summarizing the treatment
below the perceived tumoricidal threshold. To experience of 51 patients with primary (51%),
address this challenge, several groups have devel- recurrent (24%), or metastatic (12%) lesions.
oped short-range HDR plaques that are directly This experience included the use of multiple
affixed on the dural margin at risk. plaque-based brachytherapy sources, including
Published reports of plaques have utilized Y-90, P-32, and Ir-192. With a median follow-
beta-particle-emitting isotopes, including up of 18 months post-brachytherapy, they report
Yttrium-90 (Y-90) and Phosphorous-32 (P-32). good local control rates across indications. No
The benefit of beta emitters is rapid dose fall off acute or late myelopathy could be attributed to
a short distance away from the source. Therefore, dural plaque brachytherapy, and the authors con-
these sources enable high dose to be delivered clude this is a safe and effective means of dose
directly to the dural surface; the nearby spinal escalation for tumors with dural involvement.
cord typically receives a small fraction of the Due to the low-energy beta particles emitted by
total dose. Short-range brachytherapy sources Y-90 and P-32, the source has an added advan-
such as P-32 and I-125 can be used in any operat- tage of limited radiation exposure to operating
ing room with minimal risk to operative staff, room staff; Ir-192 emits more penetrating radia-
while more penetrating radiation such as elec- tion, requiring additional shielding or distance
trons or Ir-192 requires specially shielded operat- from the patient for staff safety.
ing rooms because of the radiation exposure. The MSKCC group utilizes a P-32 plaque
Furthermore, operative staff may need to leave (previously RIC-100, R.I. Consultants, Hudson,
the room when radiation is being delivered. NH, USA; now NucMedCor, San Francisco,
CA, USA) where the isotope is bound chemi-
cally to a flexible and transparent polymer layer
Review of the Literature and coated with silicone; the overall thickness is
approximately 0.5 mm [4, 27]. The thin plaque is
Delaney and colleagues at Massachusetts then wrapped in iodinated surgical film (Ioban,
General Hospital designed a Y-90 foil in a 3M, St. Paul, MN, USA) to reduce the chances
semicylindrical polycarbonate plaque [25] of microscopic isotope contamination of the
(Fig. 50.3). The plaque was placed directly on surgical bed. This approach has several advan-
the dural margin following gross total surgi- tages. P-32 has a similar steep dose fall off as
cal resection of primary and metastatic para- Y-90 but has a longer half-life, enabling longer
spinal masses and removed after delivery of shelf life. Dosimetric analysis suggests that the
7.5–15 Gy. Surface doses were reported as percent depth dose declines to 1% at 4 mm from
29% and 9% at 2 and 4 mm from the surface of the prescription depth [27]. The plaques can be
the foil, respectively. All patients had received cut to the appropriate shapes intraoperatively and
pre- and/or postoperative EBRT using photon do not require preoperative fabrication neces-
and/or proton beams. With the use of intensity sary for Y-90 products. The flat plaques are often
modulated radiotherapy, a dose constraint on easier to affix to a surgical contour compared
the dural surface can be applied at the time of to a curved semicylindrical construction. They
postoperative radiation treatment planning to also do not require special intraoperative shield-
account for the radiation given intraoperatively ing. P-32 plaques can also be used in conjunc-
with the plaque. The group reported a total tion with neoadjuvant or adjuvant EBRT, and the
spinal cord dose constraint to 63 and 54 Gy dose delivered to the dural surface enables better
a c e
90
Y Foil
5.0 mm
0.5 mm Window
f
b
d
Fig. 50.3 (a) Diagram of the Y-90 foil-based semicylindri- edge by the neurosurgeon and radiation oncologist as part of
cal polycarbonate plaque used for HDR brachytherapy. separation surgery procedure with hardware stabilization.
(From DeLaney et al. [25]. Reprinted with permission from (From Folkert et al. [27]. Reprinted with permission from
Elsevier.) (b) In situ positioning of the Y-90 plaque against Oxford University Press.) (e) Radiographic confirmation of
the dural surface after posterior decompressive surgery. bipedicular placement of catheters as part of Kypho-IORT
(From Folkert [42]. Reprinted with permission from Oxford system. (From Wenz et al. [33]. Open Access, Creative
University Press.) (c) Assembly of the P-32 plaque under Commons Attribution License.) (f) Treatment position of
sterile conditions in the operating room. (From Folkert et al. the Zeiss INTRABEAM system used to deliver X-ray-based
[27]. Reprinted with permission from Oxford University IORT prior to kyphoplasty. (From Wenz et al. [33]. Open
Press.) (d) Positioning of the P-32 plaque against the dural Access, Creative Commons Attribution License.)
homogeneity of the CTV coverage while satisfy- failure was significantly lower at 19% compared
ing cord constraints. with 34% for those who did not (p = 0.04). No
At MSKCC, patients often undergo separation acute or long-term complications were specifi-
surgery with a goal of partial resection to decom- cally attributed to the IORT.
press the thecal sac to enable adjuvant SBRT The approach was also utilized for a pediat-
[28]. For patients with extensive dural involve- ric patient with multiple recurrent thoracic spi-
ment, P-32 is placed intraoperatively on the sur- nal neuroblastoma, who had previously received
gical margin to deliver a dose of 10 Gy at a depth 25 Gy in five fractions [29]. The patient was noted
of 1 mm from the plaque edge. to have no evidence of local failure at 10 months
Folkert et al. reported P-32 plaque outcomes post-P-32 but unfortunately had suffered out of
for 68 patients with 69 treated lesions [25]. Most field progression.
patients (86%) had previously received at least A catheter-directed interstitial HDR approach
one prior course of EBRT and just over half had has also been described for patients with multiple
adjuvant EBRT following P-32 using single frac- relapsed spinal metastases [30]. Five patients who
tion (13%), high-dose hypofractionated (34%) or were felt to be ineligible for further EBRT due
low-dose hypofractionated (53%) image-guided to prior cord exposure underwent intraoperative
RT. At median follow-up of 10 months, local or percutaneous placement of vertebral catheters
relapse of 26% was noted at 12 months. In the into gross disease. HDR was performed using
subgroup of patients who underwent adjuvant Ir-192 to deliver 12–18 Gy in a single fraction.
EBRT after surgery and P-32 IORT, rate of local At a median follow-up of 9 months, 100% local
control was observed. The approach was effec- of a miniature X-ray generator (INTRABEAM,
tive for palliation as most patients (80%) had Carl Zeiss Surgical, Oberkochen, Germany),
complete (n = 2) or partial (n = 2) pain reduction with a maximum energy of 50 keV. At this point,
1–4 weeks post-procedure. No brachytherapy- IORT was performed to deliver 8 Gy at 5 mm dis-
related complications were observed, even in tance from the source, which was completed in
patients with surgical hardware. 90 seconds. The INTRABEAM device was then
removed and kyphoplasty was performed per the
usual approach.
I ORT for Painful or Unstable Spinal Since then, several reports showed the feasibil-
Metastases ity, safety, and efficacy of the approach [35–38].
For example, Reis et al. reported short-term out-
The management of patients with malignant comes after treatment of 18 lesions, noting radio-
mechanical spinal instability typically requires graphically stable disease in 93% of patients with
multimodality treatment [31]. Typically, this significant improvements in pain and no severe
requires a combination of a procedure to address complications [35]. A Phase I/II dose escalation
the instability (e.g., kyphoplasty, vertebroplasty, trial studied three IORT dose levels and found no
invasive surgical stabilization, or less-invasive dose-limiting toxicities [39]. Fifty- two patients
percutaneous screw placement) together with were subsequently enrolled in a Phase II portion,
radiation for further palliation and local tumor and median pain score on the visual analog scale
control. Several strategies have been reported to (VAS) significantly dropped from 5 preoperatively
combine the two interventions such that tumor to 2 at the first postoperative day (p < 0.001). Of
control and skeletal stabilization are performed 43 patients who reported a pre-interventional pain
concurrently. level of 3 or more, 30 (70%) reported a reduc-
One approach has been direct injection of tion of ≥3 points on the first postoperative day,
radionuclides into the bone at the time of kypho- and most had persistent pain reduction. The 3-,
plasty. Cardoso and colleagues have described 6-, and 12-month LC rates were excellent at 98%,
kyphoplasty using Samarium 153 (Sm-153) 94%, and 94%, respectively. The 6- and 12-month
mixed with polymethyl methacrylate (PMMA) OS were 64% and 48%, respectively. Given this
bone cement [32]. Sm-153 was selected as it promising early data, the Universitätsmedizin
has bone-seeking properties and releases beta Mannheim is currently conducting a Phase III trial,
particles to irradiate adjacent tumors. The group randomizing kypho-IORT with a single fraction of
reported no procedural complications or hemato- 8 Gy versus conventional palliative EBRT to 30 Gy
logic toxicities, and all patients had at least par- in ten fractions [40].
tial improvement in pain. Local control was not
explicitly quantified; however, Sm-153 beta has
a very short decay range and thus can only effec- Summary
tively treat a limited distance from the source (i.e.,
the bone cement), which will limit the amount of As systemic therapies continue to advance and
tumor that this approach can effectively treat to OS improves, the prevalence of spinal metastatic
within a few millimeters of the PMMA. disease will rise [41]. The management of pri-
Wenz and colleagues have described a hybrid mary spinal tumors also remains a significant
brachytherapy and kyphoplasty approach, which therapeutic dilemma. While EBRT remains a
they term kypho-IORT [33, 34]. A pilot case of mainstay of treatment, addressing recurrent dis-
a 60-year-old patient with breast cancer meta- ease remains a clinical challenge. IORT is a ver-
static to the T12 vertebra was presented. A per- satile strategy for focal treatment and retreatment
cutaneous, bipedicular approach into the vertebra of malignant spinal lesions especially when dural
was chosen with insertion of specially designed involvement is suspected. This approach is par-
metallic sleeves to guide the electron drift tube ticularly attractive for patients with prior EBRT
exposure where significant additional dose risks domas, chondrosarcomas, and other sarcomas. J Surg
Oncol. 2014;110:115–22.
spinal cord myelopathy. Spinal plaques allow for 4. Folkert MR, Bilsky MH, Cohen GN, Voros L, Oh
the delivery of focused treatment of the surface JH, Zaider M, Laufer I, Yamada Y. Local recurrence
of the thecal sac and can augment conformal outcomes using the 32P intraoperative brachyther-
SBRT for the combined delivery of high-dose apy plaque in the management of malignant lesions
of the spine involving the dura. Brachytherapy.
radiation to the full extent of paraspinal and epi- 2015;14:202–8.
dural diseases. 5. Hirsch O. Die operative Behandlung von
In terms of outcomes, IORT has proven effec- Hypophysentumoren: Nach endonasalen Methoden.
tive for palliation and can be combined with Arch Laryngol Rhinol. 1912;26:529–686.
6. Schulder M, Loeffler JS, Howes AE, Alexander E,
bone-stabilizing procedures such as kyphoplasty Black PM. Historical vignette: the radium bomb:
for patients with mechanical instability. Most Harvey Cushing and the interstitial irradiation of glio-
series do describe high rates of at least partial mas. J Neurosurg. 1996;84:530–2.
neurologic improvement that can be durable. 7. Hamel W, Köppen JA, Hariz M, Krack P, Moll CKE. The
pioneering and unknown stereotactic approach of
While the data exploring spinal IORT remains Roeder and Orthner from Göttingen. Part I Surgical tech-
limited to a small number of institutions with nique for tailoring individualized stereotactic lesions.
particular expertise in the management of com- Stereotact. Funct. Neurosurg. 2016;94:240–53.
plex spinal disease, the approach seems to be safe 8. Schwarz SB, Thon N, Nikolajek K, Niyazi M, Tonn
J-C, Belka C, Kreth F-W. Iodine-125 brachytherapy
and transferable, with relatively few examples of for brain tumours – a review. Radiat Oncol. 2012;7:30.
IORT-related myelopathy in the published lit- 9. Willett CG, Czito BG, Tyler DS. Intraoperative radia-
erature. However, interpretation of these series tion therapy. J Clin Oncol. 2007;25:971–7.
does require caution since many of the treated 10. Chassagne D, Dutreix A, Almond P, Burgers JMV, Busch
M, Joslin CA. Report 38. J ICRU os20:NP-NP. 1985.
patients suffered from metastatic cancer with a 11. Seichi A, Kondoh T, Hozumi T, Karasawa
low-baseline anticipated survival; therefore, they K. Intraoperative radiation therapy for metastatic spinal
may not live long enough to develop long-term tumors. Spine. 1999;24:470–3; discussion 474-475
sequelae of their treatment. 12. Saito T, Kondo T, Hozumi T, Karasawa K, Seichi
A, Nakamura K. Results of posterior surgery with
intraoperative radiotherapy for spinal metastases. Eur
Acknowledgments The authors wish to thank medical Spine J. 2006;15:216–22.
physicist Gil’ad Cohen for assistance with figure 13. Kondo T, Hozumi T, Goto T, Seichi A, Nakamura
preparation. K. Intraoperative radiotherapy combined with poste-
rior decompression and stabilization for n on-ambulant
Disclosures BSI – none, MRF – Varian Medical paralytic patients due to spinal metastasis. Spine.
2008;33:1898–904.
Systems (travel expenses), Augmenix, Inc. 14. Gutin PH, Leibel SA, Hosobuchi Y, Crumley
(research materials/grant), YY – Varian Medical RL, Edwards MS, Wilson CB, Lamb S, Weaver
systems, BrainLab, Vision RT Institute for Medical KA. Brachytherapy of recurrent tumors of the skull
Education (speaker); Chordoma foundation (med- base and spine with iodine-125 sources. Neurosurgery.
1987;20 VN-re:938–45.
ical advisory board). 15. Armstrong JG, Fass DE, Bains M, Mychalczak B,
Nori D, Arbit E, Martini N, Harrison LB. Paraspinal
tumors: techniques and results of brachytherapy. Int J
Radiat Oncol Biol Phys. 1991;20:787–90.
16. Rogers CL, Theodore N, Dickman CA, Sonntag

References VKH, Thomas T, Lam S, Speiser BL. Surgery and
permanent 125I seed paraspinal brachytherapy for
1. Kirkpatrick JP, van der Kogel AJ, Schultheiss malignant tumors with spinal cord compression. Int J
TE. Radiation dose–volume effects in the spinal cord. Radiat Oncol Biol Phys. 2002;54:505–13.
Int J Radiat Oncol Biol Phys. 2010;76:S42–9. 17. Hamilton AJ, Lulu B, Stea B, Cheng CW, Cassady
2. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Phase II JR. The use of gold foil wrapping for radiation pro-
study of high-dose photon/proton radiotherapy in the tection of the spinal cord for recurrent tumor therapy.
management of spine sarcomas. Int J Radiat Oncol TL – 32. Int J Radiat Oncol Biol Phys. 1995;32:507–
Biol Phys. 2009;74:732–9. 11. VN-readcube.com
3. DeLaney TF, Liebsch NJ, Pedlow FX, et al. Long- 18. Kumar PP, Good RR, Skultety FM, Leibrock
term results of phase II study of high dose photon/ LG. Local control of recurrent clival and sacral chor-
proton radiotherapy in the management of spine chordoma after interstitial irradiation with iodine-125:
new techniques for treatment of recurrent or unresect- 31. Barzilai O, Laufer I, Yamada Y, Higginson DS,

able chordomas. Neurosurgery. 1988;22:479–83. Schmitt AM, Lis E, Bilsky MH. Integrating evidence-
19. Wang J, Yuan H, Ma Q, Liu X, Wang H, Jiang Y, Tian based medicine for treatment of spinal metastases
S, Yang R. Interstitial 125I seeds implantation to treat into a decision framework: neurologic, oncologic,
spinal metastatic and primary paraspinal malignan- mechanicals stability, and systemic disease. J Clin
cies. Med Oncol. 2010;27:319–26. Oncol. 2017;35:2419–27.
20. Cao Q, Wang H, Meng N, et al. CT-guidance inter- 32. Cardoso ER, Ashamalla H, Weng L, Mokhtar B, Ali S,
stitial 125Iodine seed brachytherapy as a salvage Macedon M, Guirguis A. Percutaneous tumor curettage
therapy for recurrent spinal primary tumors. Radiat and interstitial delivery of samarium-153 coupled with
Oncol. 2014;9:301. kyphoplasty for treatment of vertebral metastases: tech-
21. Yao L, Cao Q, Wang J, Yang J, Meng N, Guo F, nical note. J Neurosurg Spine. 2009;10:336–42.
Jiang Y, Tian S, Sun H. CT-guided 125I seed inter- 33. Wenz F, Schneider F, Neumaier C, Kraus-Tiefenbacher
stitial brachytherapy as a salvage treatment for U, Reis T, Schmidt R, Obertacke U. Kypho-IORT – a
recurrent spinal metastases after external beam radio- novel approach of intraoperative radiotherapy during
therapy. Biomed Res Int. 2016;2016:1. https://doi. kyphoplasty for vertebral metastases. Radiat Oncol.
org/10.1155/2016/8265907. 2010;5:11.
22.
Katsoulakis E, Kumar K, Laufer I, Yamada 34. Schneider F, Greineck F, Clausen S, Mai S, Obertacke
Y. Stereotactic body radiotherapy in the treat- U, Reis T, Wenz F. Development of a novel method
ment of spinal metastases. Semin Radiat Oncol. for intraoperative radiotherapy during kyphoplasty for
2017;27:209–17. spinal metastases (Kypho-IORT). Int J Radiat Oncol
23. Rotondo RL, Folkert W, Liebsch NJ, et al. High- Biol Phys. 2011;81:1114–9.
dose proton-based radiation therapy in the 35. Reis T, Schneider F, Welzel G, Schmidt R, Bludau
management of spine chordomas: outcomes and clini- F, Obertacke U, Wenz F. Intraoperative radiotherapy
copathological prognostic factors. J Neurosurg Spine. during kyphoplasty for vertebral metastases (Kypho-
2015;23:788–97. IORT): first clinical results. Tumori. 2012;98:
24.
Indelicato DJ, Rotondo RL, Begosh-Mayne 434–40.
D, Scarborough MT, Gibbs CP, Morris CG, 36. Miglierini P, Dam-Hieu P, Key S, Quillevere S, Lucia
Mendenhall WM. A prospective outcomes study A-S, Pradier O. Kypho-IORT: the first French treat-
of proton therapy for chordomas and chondrosar- ment. Transl Cancer Res. 2014;3:88–93.
comas of the spine. Int J Radiat Oncol Biol Phys. 37. Gandhi S, Latefi A, Molina FD, Chen Y, Ghaly

2016;95:297–303. M. SURG-28. KYPHO-IORT: a new treatment
25. DeLaney TF, Chen GT, Mauceri TC, Munro JJ,
paradigm for pathological fractures. Neuro Oncol.
Hornicek FJ, Pedlow FX, Suit HD. Intraoperative 2017;19:vi241.
dural irradiation by customized 192iridium and 38. Pinar Sedeño B, Rodríguez Ibarria N, Mhaidli Hamdani
90yttrium brachytherapy plaques. Int J Radiat Oncol H, Fernández Varela T, San Miguel Arregui I, Macías
Biol Phys. 2003;57:239–45. Verde D, Lara Jiménez PC. First reported treatment of
26. Yip DD, DeLaney TF, Jacobson A, Hornicek
aggressive hemangioma with intraoperative radiation
FJ, Schwab JH, Mauceri TC, Chen Y. Review therapy and kyphoplasty (Kypho-IORT). Clin Transl
of Experience and Outcome of Dural Plaque Radiat Oncol. 2017;2:19–22.
Brachytherapy:2000–2013. Int J Radiat Oncol Biol 39. Bludau F, Welzel G, Reis T, et al. Phase I/II trial of
Phys. 2014;90:S758–9. combined kyphoplasty and intraoperative radio-
27. Folkert MR, Bilsky MH, Cohen GN, Zaider M,
therapy in spinal metastases. Spine J. 2018;18:
Dauer LT, Cox BW, Boland PJ, Laufer I, Yamada 776–81.
Y. Intraoperative 32P high-dose rate brachytherapy 40. Bludau F, Welzel G, Reis T, Abo-Madyan Y, Sperk
of the dura for recurrent primary and metastatic intra- E, Schneider F, et al. Combined kyphoplasty and
cranial and spinal tumors. Neurosurgery. 2012;71: intraoperative radiotherapy (Kypho-IORT) versus
1003–10. discussion 1010-1011 external beam radiotherapy (EBRT) for painful ver-
28. Barzilai O, Fisher CG, Bilsky MH. State of the art tebral metastases - a randomized phase III study.
treatment of spinal metastatic disease. Neurosurgery. BMC Cancer. 2019 May 9;19(1):430. https://doi.
2018;82:757–69. org/10.1186/s12885-019-5666-5.
29. Tong WY, Folkert MR, Greenfield JP, Yamada Y,
41. Spratt DE, Beeler WH, de Moraes FY, et al. An

Wolden SL. Intraoperative phosphorus-32 brachy- integrated multidisciplinary algorithm for the man-
therapy plaque for multiply recurrent high-risk agement of spinal metastases: an international
epidural neuroblastoma. J Neurosurg Pediatr. spine oncology consortium report. Lancet Oncol.
2014;13:388–92. 2017;18:e720–30.
30. Folkert MR, Bilsky MH, Cohen GN, Zaider M, Lis 42. Folkert, MR. Harvard-MIT Health Sciences and

E, Krol G, Laufer I, Yamada Y. Intraoperative and Technology MD thesis. Design, dosimetry, and imple-
percutaneous iridium-192 high-dose-rate brachy- mentation of customized 90-Yttrium plaque applica-
therapy for previously irradiated lesions of the spine. tors for intraoperative dural brachytherapy of spinal
Brachytherapy. 2013;12:449–56. tumors. MIT; 2009.
Part VII
Pain
Approach to Pain in Patients
with Central Nervous System
51
Metastases
Thomas Chai, Jennifer Erian, Mihir Joshi,

Larry C. Driver, and Dhanalakshmi Koyyalagunta
Introduction An Overview of Pain Processing
Pain is defined by the International Association Within the body’s somatic tissue (muscle, bone,
for the Study of Pain (IASP) as “an unpleasant joint, tendon, skin, organs, etc.), specific nerve
sensory and emotional experience associated fiber types of sensory neurons, known as A-∆
with actual or potential tissue damage…” and, (or A-delta) and C fibers, have in their periph-
therefore, involves more than the mere detection eral terminals specialized receptors that respond
of a (potentially) harmful stimulus by the body to nociceptive stimuli. These specialized recep-
(which describes nociception). Pain, rather, is a tors, called nociceptors, may be activated by
subjective, complex condition affected or chemical, thermal, and/or mechanical stimuli
modulated by many physiological and that reach the nociceptor’s high threshold for
psychological factors. In this chapter, we describe response. These specific “pain” fibers, with
briefly first the concept of “pain processing,” their cell bodies located in dorsal root ganglia
followed by an overview of the various types of (or respective cranial nerve ganglia), travel in
the pain, classified by tissue type. The remainder peripheral nerves (or cranial nerves V, VII, IX,
of this chapter describes select pharmacologic and X) to synapse with second-order neurons
agents used to manage specifically the located in the central nervous system (either dor-
neuropathic component of pain in central nervous sal horn neurons of the spinal cord or neurons
system metastases. Evidence to support each within brainstem nuclei). Release of excitatory
specific agent’s use in this particular condition neurotransmitters, such as glutamate and aspar-
will be provided, where available. tate, occurs at these nerve synapses, resulting in
travel (and modulation) of the nociceptive signals
to higher CNS centers, via ascending projections
in various tracts (the spinothalamic tract being
an important example). An important supra-
spinal structure in this ascending system is the
T. Chai (*) · J. Erian · L. C. Driver · D. Koyyalagunta thalamus, which receives the nociceptive input
Department of Pain Medicine, UT MD Anderson and sends projections further to other structures
Cancer Center, Houston, TX, USA in the brain that influence both the discriminative
e-mail: tchai@mdanderson.org and affective components of pain.
M. Joshi This entire “nociceptive system” may be
Department of Physical Medicine & Rehabilitation, modulated at multiple points along the pathway.
Baylor College of Medicine, Houston, TX, USA

https://doi.org/10.1007/978-3-030-42958-4_51
674 T. Chai et al.
For example, chronic nociceptive input (with Table 51.1 Classification of pain type by tissue
resultant release of inflammatory mediators) Nociceptive Examples
may sensitize peripheral nociceptors, leading to pain
Somatic Skin, bone, joints, connective
a lowered threshold for response or an increased
Visceral tissue, muscle
responsiveness to normal suprathreshold input Lung, liver, esophagus, pancreas,
(a condition known as peripheral sensitization). intestines, colon, bladder.
Repetitive stimulation can also result in lowered Neuropathic Examples
thresholds for response or increased suprathresh- pain
Central pain Brain, spinal cord
old response of the second-order, dorsal horn Peripheral Cranial nerves, spinal nerves and
neurons (central sensitization) or an increased pain their branches, ganglia
output: input ratio (referred to as the wind-up
phenomenon) of these neurons.
In contrast to pain facilitation as described Nociceptive Pain
above, modulation of nociceptive signals by cer-
tain descending supraspinal systems results in There are two main types of nociceptive pain –
inhibitory modulation of pain. Some of the struc- somatic and visceral. Somatic nociceptive pain is
tures associated with this descending inhibitory associated with injury to somatic, nonneural tis-
system include the periaqueductal gray, the sero- sues. Somatic nociceptors innervate somatic
toninergic raphe nucleus, and the noradrenergic structures such as, but not limited to, the skin,
locus ceruleus. These systems influence the subcutaneous tissue, joint capsules, muscles, lig-
dorsal horn neurons of the spinal cord via aments, tendons, fascia, periosteum and endos-
projections within the dorsolateral funiculus. The teum of bone, parietal pleura, and parietal
endogenous opioid system (endorphins, enkeph- peritoneum. Somatic nociceptive pain is usually
alins, and dynorphins) also exerts its pain inhibi- localizable by the patient.
tory effects at both the peripheral and central Visceral nociceptors innervate thoracic,
nervous system levels. abdominal, and pelvic viscera, and its surround-
The affective component of pain may signifi- ing connective tissue/capsule, usually not the
cantly influence the patient’s perception of the organ parenchyma proper. Visceral nociceptors
pain experience. Spinal pathways leading to both are activated by organ distention, inflammation,
limbic structures and medial thalamic nuclei pro- and ischemia, rather than stimuli such as cutting,
vide input to areas of the brain related to affect/ stabbing, or burning. Visceral pain is usually
emotion. For instance, the anterior cingulate cor- described as poorly localized and may be accom-
tex of the brain, and its association with limbic panied by autonomic symptoms. Pain from vis-
structures, appears to be intimately involved in ceral structures may refer to, and be perceived in,
conferring the emotional aspect to pain, having a a different area of the body – this is due to the
role in the sensorimotor, cognitive processing, vis- convergence of visceral afferent nociceptive
ceromotor, endocrine outflow, skeletomotor out- fibers with somatic afferent nociceptive fibers
flow, and other responses to nociceptive stimuli. onto the same dorsal horn neurons within a simi-
lar segment of the gray matter of the spinal cord.
Types of Pain
Neuropathic Pain
There are various ways by which to classify pain,
based on factors such as time (acute, chronic), Neuropathic pain is “pain caused by a lesion or
mechanism (trauma, surgical, etc.), or by tissue disease of the somatosensory nervous system,” as
type (Table 51.1), among other classification defined by the IASP. There are two subtypes of
schemes. In this chapter, we describe pain by tis- neuropathic pain – (1) central and (2) peripheral
sue type, using IASP terminology, as follows: neuropathic pain. Central neuropathic pain (or
51 Approach to Pain in Patients with Central Nervous System Metastases 675
simply “central pain”) is a type of neuropathic are described in more detail in subsequent sec-
pain “caused by a lesion or disease of the central tions of this chapter). For instance, central effects
somatosensory nervous system,” whereas periph- from opioids can produce euphoria, dysphoria,
eral pain involves the peripheral somatosensory sedation, nausea (through direct effects on the
nervous system. brainstem chemoreceptor trigger zone), cough
The quality of neuropathic pain is described as suppression, and probably the most feared com-
a burning, throbbing, electrical-shocking, or plication – respiratory depression (through direct
“pins and needles.” Neuropathic pain can be effects on the brainstem respiratory centers).
associated with abnormal sensations, spontane- Peripheral effects of opioids can result in consti-
ous or evoked, known as paresthesias, or with pation (from slowing of gastrointestinal motil-
both unpleasant and abnormal sensations, called ity), biliary smooth muscle constriction, urinary
dysesthesias. Allodynia is a condition whereby retention, and pruritis, among many other effects.
pain is experienced from a normally innocuous Below we discuss select opioid analgesics most
stimulus, for instance, light touch. commonly prescribed for cancer pain by the Pain
Service at the authors’ institutions. Evidence of
efficacy specifically on neuropathic-type cancer
elect Pharmacologic Agents
S pain in human subjects is provided in this section.
for Neuropathic Cancer Pain Table 51.2 is a sample opioid equianalgesic dos-
ing reference from the authors’ institution.
In this section, we describe the major classes of
analgesics used for neuropathic cancer pain, Morphine Sulfate
including those caused by CNS metastatic dis- Morphine is known as the prototypic opioid. It
ease. A survey of select agents from each class is is a full agonist at the mu-opioid receptor, which
described below. is the predominant analgesic receptor within
the nervous system. Morphine is absorbed well
orally, but undergoes extensive hepatic first-
Opioid Analgesics pass metabolism, and therefore, oral dosages
must be increased compared to parenteral doses.
Opioid analgesics (henceforth referred to simply Morphine undergoes glucuronidation by the
as opioids) are drugs that bind to and assert ago- liver, with the resulting major metabolite known
nist effects on the opioid receptors of the nervous as morphine-3-glucuronide (M3G). To a much
system. Opioids are considered the gold standard lesser extent, morphine-6-glucuronide (M6G) is
in the management of cancer pain, of all types – produced, this metabolite being more potent than
neuropathic and nociceptive. Opioids produce the parent compound. Excretion of morphine and
analgesic effects but may also result in other its byproducts is through the renal route. There
potentially unwanted side effects (some of which is concern, therefore, for using morphine in the
Table 51.2 Equianalgesic dosing table

Oral dose Parenteral Conversion factor for changing Conversion factor for changing oral
Opioid (PO) dose (IV) parenteral opioid to oral opioid opioid to oral morphine
Morphine 15 mg 6 mg 2.5 1
Oxycodone 10 mg N/A N/A 1.5
Hydrocodone 15 mg N/A N/A 1
Oxymorphone 5 mg 0.5 mg 10 3
Hydromorphone 3 mg 1.5 mg 2 5
Fentanyl N/A 60 mcg N/A Should be managed by clinicians
experienced in pain management
Note: Methadone should be initiated and managed by clinicians experienced in pain management
Source: UT MD Anderson Cancer Pain – Adult Practice Algorithm
676 T. Chai et al.
renal patient population, as active metabolite Fentanyl

accumulation could lead to neurotoxic and other Fentanyl is a synthetic, highly lipophilic opioid,
significant adverse effects. Morphine is often with a potency of roughly 100× that of morphine.
combined with other agents and adjuvants [1–3] Fentanyl has properties of rapid onset and short
for neuropathic cancer pain, and it is one of the duration of action and is used commonly in peri-
few drugs approved by the United States Food operative and intensive care settings. There are
and Drug Administration (FDA) for use in intra- various preparations of fentanyl for different
thecal drug delivery systems. routes of administration, including parenteral,
transmucosal, transdermal, and spinal. Fentanyl’s
Tramadol major metabolite is norfentanyl, which is inactive
Tramadol is a synthetic opioid with dual proper- and thus considered less risky to use in the renally
ties – agonist effects on the mu-opioid receptor impaired patient population.
and norepinephrine/serotonin reuptake inhibi-
tion. Tramadol undergoes hepatic metabolism, N-Methyl-D-Aspartate (NDMA)
with one of the active metabolites, desmetra- Antagonists: Methadone and Ketamine
madol, being notable for its much higher affin-
ity for the mu-opioid receptor compared to its Methadone
parent compound. Tramadol and its by-products Methadone, a synthetic opioid, is an agonist at the
are excreted renally and also must be used care- mu-receptor, but also an antagonist of the
fully in renally impaired patients. Tramadol was N-methyl-D-aspartate (NMDA) receptor, which
assessed [4] for efficacy, safety, and quality-of- is implicated in central sensitization/hyperalgesia.
life impact for patients with neuropathic pain in Methadone has highly variable pharmacokinetic
cancer. In this double-blind, placebo-controlled properties and a long half-life. In addition, metha-
study, patients were randomized to receive either done is biotransformed hepatically and may be
tramadol or placebo. Thirty-six patients were affected by other drugs that inhibit its metabo-
enrolled and equally divided into each study lism; therefore, expert prescribing and monitoring
group. Tramadol was given in the treatment arm of methadone is necessary to minimize risks of
at 1 mg/kg every 6 hours and increased to 1.5 mg/ respiratory depression. Methadone is regularly
kg every 6 hours if necessary. The group receiv- prescribed at the authors’ institution for cancer-
ing tramadol showed major improvement in pain related neuropathic pain, as there is both anec-
intensity, Karnofsky scores, sleep quality, and dotal and scientific evidence [5–7] supporting its
activities of daily living, compared to the placebo use in this condition, particularly when the neuro-
group. In this study, tramadol was concluded to pathic pain is refractory even to high-dose opi-
be a therapeutic option to control neuropathic oids. For example, Sugiyama et al. [8] performed
cancer pain and improve quality of life in the a retrospective study on the effectiveness of
cancer patient. changing patients’ opioid regimens to methadone
for cancer-related neuropathic pain. The Faces
Hydromorphone Pain Scale (FPS) was used to measure pain inten-
Hydromorphone, like morphine, undergoes sity and pain relief. Twenty-eight patients on other
metabolism by conjugation to form metabolites potent opioids were changed to methadone, and
hydromorphone-3-glucuronide (H3G), predomi- 78.6% of those patients, within 2 weeks, had a
nantly, and 6-glucuronide, which are excreted in significant reduction in their mean FPS score, and
the urine. Similarly to the morphine metabolites, 12 out of 17 patients either reduced or discontin-
these byproducts may also contribute to neuro- ued entirely adjuvant analgesics.
toxic side effects, requiring caution when pre-
scribing to the renal population. Hydromorphone Ketamine
is considered, mg to mg, about five times more Ketamine is an anesthetic that has analgesic and
potent than morphine. dissociative properties. Its analgesic property is
thought to be related to its antagonism of the wide range of available agents and routes, avail-
NMDA receptor. Although randomized clinical ability of immediate- and extended-release for-
trials show little efficacy for ketamine in manag- mulations, and efficacy in many types of pain.
ing cancer pain, there are a number of case series These benefits must be carefully considered
and open-label studies that show benefit [9]. For against the side effect profile common to most
instance, Mercadante et al. [10] published a case opioid agents, as well as the significant risk of
report on administration of ketamine as a subcu- disorders related to opioid use. Awareness of the
taneous infusion in a patient who experienced side effects and safety considerations involved in
opioid-resistant neuropathic cancer pain, with opioid therapy, as well as a proactive approach
dramatic reduction in opioid requirement and to addressing them, are imperative in effec-
continued relief after 13 months with treatment, tive risk management for patients using opioid
despite progression of disease. medications.
Ketamine is utilized in the authors’ pain clinic
practice as an intravenous infusion at 0.5 mg/kg,
over a one-hour duration; however, there is no con- Cognitive Impairment
sensus as to the optimal protocol, and, consequently,
there exist many parenteral ketamine protocols for Concurrent use of opioids with sedating agents
treating unremitting cancer pain [11–13]. may increase the risk of cognitive impairment.
Because cognitive impairment can present either
in opioid overdose or in the course of regular
Opioid Safety Considerations opioid use, it is important to readily identify
whether a patient may indeed be in overdose – a
In this section, we describe some of the most potentially fatal situation. For example, rapidly
pressing or concerning side effect and safety declining cognitive status after opioid adminis-
issues associated with opioid prescribing. As the tration is more concerning for overdose and war-
use of opioid medications for pain management rants prompt evaluation. In cases where cognitive
has increased steadily over the past decades, the impairment seems to be linked to regular opioid
incidence of opioid-related deaths has tracked use and has a more gradual onset, there are a few
closely with this trend, as reported by the Center strategies available for the prescribing provider.
for Disease Control and Prevention (CDC). Initial The first is to consider dose reduction if analge-
public acceptance of opioid medications as gen- sia is sufficient at the current dose; this is done
erally safe agents has given way to an increased with the knowledge that pain may worsen. If
awareness of the risks associated with their use. dose reduction is not a viable option, consider
Additionally, increased prescription of opioid opioid rotation or dose reduction alongside the
medications has increased the incidence of diver- addition of an adjuvant analgesic (next section of
sion and misuse. Indeed, the Center for Medicare this chapter).
and Medicaid Services (CMS) has declared the
opioid misuse epidemic a public health emer-
gency, and many policies are in place to address Opioid Overdose
this opioid crisis here in the United States. For
instance, the CDC has published guidelines on The 2017 data from the National Institutes of
opioid therapy for chronic pain. (Notably, the Health demonstrate a continued trend of increas-
CDC states that their chronic opioid prescribing ing opioid overdose deaths, with opioid pain
guidelines are not applicable for patients on relievers accounting for approximately 40% of
active cancer treatment, palliative care, or end-of- total opioid overdose deaths. Though the great
life care.) majority of these events involve diversion, co-
Opioid medications remain a major compo- ingestion, or misuse, prescribers should be aware
nent of the treatment of cancer pain, due to the of the risk and available treatment.
678 T. Chai et al.
Co-ingestion with sedating agents, including ues opioid therapy, he/she may experience a
but not limited to benzodiazepines and alcohol, withdrawal syndrome, resulting in an “autonomic
dramatically increases the risk of respiratory arousal” described as a limited period of irritabil-
depression. Additionally, any condition (pulmo- ity, agitation, lacrimation, yawning, abdominal
nary disease/compromise, sleep apnea, stroke cramping, and loose stools, among other unpleas-
history, brain injury) or prescription medication ant sensations.
that increases the patient’s risk of respiratory Despite preventative measures, opioid over-
depression must be weighed when initiating or doses continue to occur at increasing rates,
escalating opioid therapy. For example, the FDA year after year, in the United States. As part of
has in place a box warning as of 2016 regarding a broader harm-reduction initiative, the FDA
the combined use of opioids and benzodiaze- approved the opioid antagonist naloxone (trade
pines, due to evidence of the combined increased name Narcan) in 1971 for treatment of opioid
risk of respiratory depression and death when overdose. Initially available only as intravenous
these agents are used in conjunction. Prescribers or intramuscular injections, naloxone is now
should counsel patients on this risk when initiat- available as a subcutaneous injectable, intra-
ing opioid therapy for a patient already on benzo- muscular auto-injector, and intranasal spray. The
diazepines or those with comorbid conditions. A latter is seeing increased use as an effective res-
low starting dose and slow drug titration can help cue medication deployed by first responders and
minimize the risk of overdose and respiratory community bystanders to reverse opioid over-
depression. dose, and its prescribing is encouraged under the
Opioid misuse can stem from the intentional “Surgeon General’s Advisory on Naloxone and
therapeutic use of the opioid, but in an inappro- Opioid Overdose,” by the current US Surgeon
priate way. Abuse occurs when patients use opi- General, Dr. Jerome Adams, for patients who
oids for intentional nontherapeutic use to achieve are at higher risk for opioid-use disorders.
a desirable effect. Patients on daily opioid medi- Increasingly, physicians are co-prescribing nal-
cation must be counseled therefore to take their oxone with opioids for patients on nominally
medication strictly as prescribed. Daily opioid high doses, patients with preexisting risk factors
use can lead to the development of physiologic for respiratory depression, or patients where the
tolerance, a condition of diminishing analgesic risk of opioid overdose is felt to be significant
effect over time. Rapid development of tolerance [14, 15]. This measure was added to the CDC’s
is a phenomenon known as tachyphylaxis. 2016 prescribing guidelines for opioid therapy
Another concept, called the opioid-tolerant state, as a harm-reduction strategy worthy of con-
is defined as the state whereby a patient is taking sideration when initiating or escalating opioid
at least 60 mg daily of oral morphine or its equiv- therapy. Naloxone, available in easily adminis-
alent, for at least 1 week. This state is in contrast tered intranasal or intramuscular forms without
to the opioid-naive state, where the patient has no a prescription in 48 states, acts within minutes to
regular exposure to opioids, and to the opioid displace opioid agents from central mu-receptors.
non-tolerant state, where the patient is using opi- Patients who are at higher risk for an overdose
oids regularly, but not to the amount sufficient to event should be educated on the use of naloxone,
meet the criteria for the opioid-tolerant state. A and more importantly so should any individual
period of abstinence can lead to the loss of the who will be with the patient on a regular basis.
opioid-tolerant state, which can result in uninten- Like intramuscular epinephrine auto-injectors for
tional overdose when the patient attempts to patients with anaphylaxis, naloxone will often be
resume their opioid therapy. Therefore, it is administered to the patient by someone who is
advisable for physicians to check with their with them around the time of overdose.
patients at every appointment to ensure they Naloxone has proven to be extremely effica-
understand the importance of taking their medi- cious as a rescue agent, with a 2014 meta-analysis
cation as directed. If a patient abruptly discontin- [16] demonstrating an Odds Ratio (OR) 8.58 of
increased recovery from opioid overdose when that follow. The fourth, and perhaps most impor-
naloxone is administered. Its pharmacokinetic tant, is the lack of consensus on the actual
profile allows for rapid decoupling of opioid agents mechanics of opioid diversion. SAMHSA data,
from the mu-receptor, but it also dissociates itself which rely on self-reporting, show that 75% of
from the mu-receptor within minutes. Depending opioid abusers obtained medications from a fam-
on the location and response time of emergency ily member or friend. Increased activity at all lev-
services, it may be necessary to administer multiple els of law enforcement to counter street and
successive doses of naloxone to maintain internet sales of prescription pain medication has
respiratory function until first responders arrive. not addressed, therefore, what may be the most
common route of opioid diversion. While opioid
medications continue to maintain a high street
Diversion price, making them a lucrative option for patients
in financial strain, the data suggest most diver-
Diversion, either intentional or unintentional, sion is not transactional. Diversion from friends
is a major concern for physicians, patients, the and family, whether solicited or unsolicited,
healthcare system, and law enforcement agen- seemingly constitutes the major access route for
cies. A landmark 5-year national study of individuals seeking unprescribed opioids. That
diversion revealed over 64,000 reported cases said, hard data on diverting mechanisms are
[17]. Due to acknowledged study shortcom- scarce due to a variety of social and political fac-
ings, and Substance Abuse and Mental Health tors, as well as limits in effective data collection.
Administration (SAMHSA) survey data showing Regardless of routes to diversion, the fact
abuse rates of hydrocodone and oxycodone mea- agreed upon most commonly is that the major
suring 17.7 million and 13.6 million individuals, source for diverted opioids is patients who receive
respectively [18, 19], there is good reason to sus- prescriptions for opioids. The prescribing physi-
pect the actual rate of diversion is far higher. cian, then, plays a role in reducing diversion.
Several trends have emerged in diversion and This fact is reflected in increased scrutiny by fed-
prescription opioid abuse. The first is that, over- eral agencies of physicians’ prescribing prac-
all, immediate-release (IR) formulations are tices, as well as pharmacies that dispense opioids.
diverted and abused at higher rates than extended- Here is a selection of some tools physicians can
release (ER) formulations. The second is that an utilize to reduce the risk of involvement in
initial preponderance of prescription opioid diversion:
abuse in rural communities, thought to be sec-
ondary to higher availability of street drugs in • Pain Contract: In its most basic form, a pain
urban communities, has begun to level off. contract will bind the patient to three rules.
Prescription drug abuse is seen now at high levels First, that their pain physician will be his/her
in urban, suburban, and rural settings across all only source of opioid prescriptions. Second,
socioeconomic strata. The third is the importance that he/she will only use one pharmacy to fill
of cultural and employment differences between his/her prescriptions. Third, that he/she will be
rural and non-rural settings; in communities the only ones to use his/her prescribed opioid
where the majority of employed adults perform medications. Additional language may include
manual labor (e.g., coal mining, farming, log- a promise not to miss appointments or use
ging, fishing), the incidence of occupation-related other sedatives, consent to random drug
pain is higher. Thus, the prevalence of pain and screens at office visits, or restrictions on refills
the prevalence of pain medication prescribing are in the event of lost or stolen medication. This
higher on a per capita basis. The widespread document, signed by the patient and counter-
nature of prescription opioid utilization in these signed by the prescribing physician, acts as a
communities is thus more commonly accepted as code of conduct for both parties and defines
a part of life, as are the dependence and abuse the terms under which the prescribing physi-
680 T. Chai et al.
cian will continue to prescribe opioids to the Adjuvant Analgesics

patient. The contract is enforceable to the
extent that the physician is willing to stop see- In this section, we describe some of the most
ing a patient who violates its terms. commonly prescribed adjuvants for neuropathic
• Drug Screen: Used in conjunction with a con- pain. Adjuvant analgesics are drugs that with pri-
tract, random drug screens, most commonly mary indications not related to pain but are found
using hair, urine, or saliva, are a way to ensure to be useful for their pain-relieving effects. The
a patient is taking prescribed medications and specific adjuvants detailed here are ones with his-
no other agents of concern [20, 21]. Older torical benefit for a variety of neuropathic pain
drug tests could only detect opioids generally, conditions, and many belong to the class of medi-
while newer tests can detect active drug and cations used to treat seizures and depression. In
metabolites for a variety of commercially fact, anticonvulsants and antidepressants are con-
available and illicit agents. If a patient is sidered first-line agents for neuropathic pain in
diverting their prescribed medication, or if cancer, often used in combination with opioids.
they are using any prescribed agents in con- Use of these adjuvants can reduce the patient
junction, a drug screen will be able to reveal need for opioids, an effect called opioid-sparing.
this.
• Prescription Drug Monitoring Program
(PDMP): PDMP systems, which have been Anticonvulsants
developed in North America, Australia, and
some European countries, have allowed an Gabapentin and Pregabalin
increased degree of prescription monitoring. The anticonvulsant drugs most commonly
Patients are entered into a database by phar- employed for neuropathic cancer pain are gaba-
macies, listing their prescribed controlled pentin and pregabalin. These two drugs have
agents, dosing, prescriber information, and similar pharmacodynamic properties, in that they
filling pharmacy. These programs were both inhibit voltage-gated calcium channels,
started in an effort to reduce “doctor shop- through blockade of the α−2/Δ−1 subunit of these
ping,” whereby patients would go to multiple channels, which are upregulated in pain states.
physicians to get opioid prescriptions, filling Both gabapentin and pregabalin are structurally
them at multiple pharmacies to avoid raising similar to gamma-amino-butyric acid (GABA);
suspicion. Where available, PDMP data however, they are not ligands for the GABA
should be reviewed at every patient visit to receptor. These drugs are not metabolized, and
ensure fidelity with single-prescriber and drug clearance is through the renal route (urine);
single-pharmacy rules. If any discrepancies thus, dose adjustment is necessary in those with
are revealed, they should be discussed with renal insufficiency. The most common side
the patient. effects reported for these “gabapentinoids”
include dizziness, drowsiness, weight change
Opioid medications, owing to their effective- (gain), and edema of the hands and feet.
ness against multiple pain mechanisms, are Several studies support the effectiveness of
widely used in the treatment of cancer pain. gabapentinoids for neuropathic cancer-related
Effective pain management, in turn, improves pain. For example, in a prospective, open-label
quality of life for patients with cancer and also study, Ross et al. [22] studied gabapentin effec-
increases their ability to continue treatment. The tiveness in two parallel groups – 25 patients in
safety considerations involved in opioid use are the first group had cancer-treatment-related neu-
significant, and merit constant surveillance by ropathic pain, while 37 patients, assigned to the
prescribing physicians to ensure patients are other group, had tumor-related neuropathic pain.
using their medications appropriately with Gabapentin dosage was titrated to 1800 mg/day
minimal adverse effects. for patients in both groups. Pain scores per the
modified Brief Pain Inventory (BPI) were neuropathic cancer pain. Fifty-two patients were
assessed as the primary outcome measure, and assigned into one of four groups. Those in group
the results of the study showed a significant 1 were administered both gabapentin 200 mg and
reduction in “worst,” “average,” and “current” imipramine 10 mg every 12 hours; group 2, gaba-
BPI pain scores, but not the “least” score. Of the pentin 200 mg every 12 hours; group 3, gabapen-
total patients, 45.2% achieved a minimum of tin 400 mg every 12 hours; and group 4,
one-third reduction in the pain score. The authors imipramine 10 mg every 12 hours. Results
of this study concluded that gabapentin was showed that the low-dose gabapentin–imipra-
indeed effective in the treatment of cancer-related mine combination significantly reduced total
neuropathic pain. pain score, as well as daily paroxysmal pain
Caraceni et al. [23] performed a multicenter, episodes.
randomized, double-blind, placebo-controlled, Pregabalin was compared to opioids for both
parallel-design trial to determine the analgesic safety and efficacy in treating neuropathic cancer
effect of adding gabapentin to opioid therapy for pain in a prospective, head-to-head, randomized,
managing neuropathic cancer pain. A total of 121 open-label study [25]. A total of 120 patients
patients were enrolled in the study. Gabapentin were randomized into two groups, receiving
was titrated to 1800 mg/day while patients increasing doses of either oral pregabalin or
remained on stable opioid therapy. Average daily transdermal fentanyl. The main outcome measure
pain was measured by Numerical Rating Scale was pain score by VAS. A significantly higher
(NRS) score, and the whole follow-up average proportion of patients had at least 30% reduction
pain score was used as the primary outcome mea- in pain score, compared to the fentanyl group,
sure. A total of 79 patients received gabapentin and the percentage mean change (decrease) from
and 58 completed the study; 41 patients received pain baseline was significantly different for pre-
placebo, of which 31 completed the study. gabalin versus fentanyl. Secondary measures of
Analysis showed a significant difference of aver- patient-reported satisfaction were also more fre-
age pain intensity between the gabapentin group quent in the pregabalin-treated group, and
and placebo group, supporting the effectiveness adverse events and treatment discontinuation
of gabapentin in improving analgesia in neuro- were higher in the fentanyl group. This study
pathic pain cancer patients using opioids. concluded that the use of adjuvants, like pregaba-
In a similar study [3], the efficacy and safety lin, could lead to better neuropathic pain control
of pregabalin were evaluated in neuropathic can- and to opioid sparing effects.
cer pain patients who were using morphine. Forty A post hoc analysis [26] of pregabalin versus
patients were randomized into two groups: the non-pregabalin-treated patients with neuropathic
first group received pregabalin plus oral mor- cancer pain in a 2-month multicenter, prospective,
phine in Phase I and then placebo plus oral mor- epidemiologic study showed a higher satisfaction
phine in Phase II, while the latter group received rate, decreased benzodiazepine use, and decreased
the opposite in each phase. There was a 1-week total pain intensity and interference in the Brief
washout period between phases. The primary Pain Inventory for those patients treated with pre-
outcome measure was reduction in oral morphine gabalin polytherapy, compared to the non-pregab-
dose. Results showed that there was a significant alin treatment group. The study authors concluded
reduction in the mean minimal effective dose of that the addition of more specific drugs that target
morphine during treatment with pregabalin. The neuropathic pain in affected patients provides
authors concluded that pregabalin enhanced the more treatment satisfaction and better pain- and
efficacy of oral morphine, while also reducing pain interference-related outcomes.
opioid dose-related side effects, in cancer patients
with neuropathic pain. Carbamazepine and Oxcarbazepine
In another study [24], low-dose gabapentin Carbamazepine and its structural derivative,
was studied in combination with imipramine for oxcarbazepine, are sodium channel blockers that
682 T. Chai et al.
appear to selectively inhibit active A-∆ and C There are few, small studies supporting its use for
nociceptive fibers, blocking both peripheral and neuropathic cancer pain. For instance, a study by
central pathways for pain. Although the literature Banaerjee et al. [36] compared the efficacy and
is sparse in describing their effects on cancer safety of amitriptyline versus gabapentin as a co-
pain, these drugs are well established in manag- analgesic for patients receiving opioids to man-
ing other chronic pain conditions with a neuro- age cancer-related neuropathic pain. Eighty-eight
pathic component, such as trigeminal neuralgia patients with neuropathic pain in malignancy
[27] and various forms of peripheral neuropathy were randomly assigned to two groups. The first
[28, 29]. Oxcarbazepine is considered to have a group received gabapentin and tramadol, while
more favorable safety profile, with less risk for the second group received amitriptyline and tra-
hepatic or hematologic adverse reactions, com- madol. At 6 months, there was a decline in Visual
pared to carbamazepine. Analogue Scale (VAS) scores from baseline in
both treatment groups, without any statistically
significant difference between groups. The
Antidepressants authors of the study concluded that amitriptyline
could be an appropriate alternative to gabapentin
Duloxetine for managing neuropathic pain from cancer.
Duloxetine is a serotonin- and norepinephrine- In a prospective randomized study, Mishra
reuptake inhibitor (SNRI) antidepressant, et al. [37] compared the efficacy of amitriptyline,
approved by the US FDA to treat depression, gabapentin, and pregabalin for neuropathic can-
generalized anxiety disorder, and pain associated cer pain. A total of 120 patients with neuropathic
with various conditions, such as painful diabetic cancer pain were enrolled and divided into four
peripheral neuropathy, fibromyalgia, and chronic, different groups: amitriptyline group, gabapentin
multisite musculoskeletal pains. In the cancer group, pregabalin group, and placebo group. A
patient population, duloxetine has been used to significant reduction in VAS scores were seen in
manage chemotherapy-induced peripheral neu- all groups, with the authors concluding that all of
ropathy pain [30, 31] and joint pains from aroma- the anti-neuropathic drugs studied demonstrated
tase inhibitor therapy [32, 33]. Although less well effect in relieving cancer-related neuropathic
supported, duloxetine has been routinely used pain.
also to manage cancer pain with a neuropathic
component. In a small retrospective pilot study,
Matsuoka et al. [34] assessed the effectiveness of Topical Agents
duloxetine in patients with cancer-related neuro-
pathic pain refractory to opioids and gabapenti- Lidocaine
noids, finding it to be effective in reducing pain Lidocaine is a local anesthetic of the amide type.
scores in 7 of 15 patients. The same authors have Lidocaine inhibits voltage-gated sodium chan-
underway a prospective, randomized phase III nels within nerve cell membranes, preventing
study [35] to further establish evidence to support depolarization and, therefore, action potential
duloxetine use in this setting. generation. Lidocaine is available in topical form,
and it can be helpful in relieving malignant neu-
Amitriptyline ropathic pain. Lopez Ramirez [38] conducted a
Amitriptyline is a tricyclic antidepressant, with study aimed to evaluate the efficacy of lidocaine
evidence supporting its efficacy as an adjuvant 5% patch for focal neuropathic pain in patients
for neuropathic pain in conditions such as central with or without cancer. Fifteen patients were
pain related to stroke and spinal cord injuries, as recruited. Six of the fifteen patients had cancer-
well as peripheral neuropathic pain related to dia- related neuropathic pain. Eight out of the 15
betes, chemotherapy, and postherpetic neuralgia, patients treated reported a potent analgesic effect,
among many other neuropathic pain conditions. and four patients reported partial analgesia.
Fleming and O’Connor [39] retrospectively were asked which treatment arm was most bene-
audited the use of lidocaine patch 5% in a com- ficial – 60% chose the capsaicin arm, 18% chose
prehensive cancer center. Among the 97 patients the placebo arm, and 22% chose neither. The
prescribed the patch, 26 were for persistent post- authors of the study concluded that topical capsa-
surgical neuropathic pain, 24 were for posther- icin cream significantly decreased postsurgical
petic neuralgia, and 18 were for cancer-related neuropathic pain in cancer patients and was pre-
neuropathic pain. Allodynia was a feature in 60% ferred by patients over placebo by a 3:1 margin.
of these patients, and analgesic efficacy in those
with allodynia was “potent” in 35%, 38%, 39%,
respectively. Conclusion
Kern et al. [40] performed a retrospective
analysis of 68 case reports regarding 5% lido- Cancer-related neuropathic pain, such as from
caine medicated plaster for cancer pain with a CNS metastases, can be a challenging condition
neuropathic component or for trigeminal neuro- to manage. A multidisciplinary strategy, includ-
pathic pain. The plaster was found most helpful ing potential interventional pain management
for surgical- or chemotherapy-related neuro- strategies discussed elsewhere in this book, is
pathic pain, with at least 50% of those using the essential to optimize patient outcomes. Providers
plaster able to dose-reduce systemic analgesics. should consider not only opioid drugs but also
In trigeminal neuralgia, potential predictors of other adjuvants with analgesic properties such as
response to lidocaine plaster were found to be antidepressants, anticonvulsants, and local anes-
hyperalgesia, allodynia, continuous pain, among thetic classes, among others.
others.
Capsaicin
Capsaicin is the substance that gives chili pep-
References
pers the characteristic burning sensation with tis- 1. Kannan TR, Saxena A, Bhatnagar S, Barry A. Oral
sue contact. Capsaicin, along with heat, acid, and ketamine as an adjuvant to oral morphine for neuro-
other ligands, binds to transient receptor potential pathic pain in cancer patients. J Pain Symptom Manag.
vanilloid subtype 1 (TrpV1), a cation receptor 2002;23(1):60–5.
2. Mercadante S, Arcuri E, Tirelli W, Villari P,
expressed on the peripheral and central terminals Casuccio A. Amitriptyline in neuropathic cancer
of nociceptive neurons. Prolonged capsaicin pain in patients on morphine therapy: a randomized
exposure is thought to result in a paradoxical placebo-
controlled, double-blind crossover study.
desensitization of TrpV1, with subsequent anal- Tumori. 2002;88(3):239–42.
3. Dou Z, Jiang Z, Zhong J. Efficacy and safety of pre-
gesic effect. Although well studied for nonmalig- gabalin in patients with neuropathic cancer pain
nant neuropathic pain conditions [41, 42], undergoing morphine therapy. Asia Pac J Clin Oncol.
capsaicin has limited evidence in the cancer neu- 2017;13(2):e57–64.
ropathic pain patient. One study, however, of 4. Arbaiza D, Vidal O. Tramadol in the treatment of neuro-
pathic cancer pain: a double-blind, placebo-controlled
chronic postsurgical neuropathic pain in 99 can- study. Clin Drug Investig. 2007;27(1):75–83.
cer survivors [43] involved an 8-week application 5. Mannino R, Coyne P, Swainey C, Hansen LA,
of 0.075% capsaicin cream four times daily to the Lyckholm L. Methadone for cancer-related neuro-
affected painful area, followed by 8 weeks of pla- pathic pain: a review of the literature. J Opioid Manag.
2006;2(5):269–76.
cebo cream application, or vice versa. The capsa- 6. Leppert W, Kowalski G. Methadone as an additional
icin cream arm of treatment had a significant opioid for a cancer patient with severe neuropathic and
reduction in pain compared to placebo. The cap- bone pain not responsive to other opioids and adjuvant
saicin treatment arm was associated, however, analgesics. J Palliat Care. 2013;29(2):119–21.
7. Makin MK, Ellershaw JE. Substitution of another
with significantly more skin burning and redness, opioid for morphine. Methadone can be used to
but treatment arm discontinuation was similar in manage neuropathic pain related to cancer. BMJ.
both groups. At the end of the study, participants 1998;317(7150):81.
684 T. Chai et al.
8. Sugiyama Y, Sakamoto N, Ohsawa M, Onizuka M, ment of cancer-related neuropathic pain: a prospective,

Ishida K, Murata Y, et al. A retrospective study on open-label study. J Palliat Med. 2005;8(6):1118–26.
the effectiveness of switching to oral methadone for 23. Caraceni A, Zecca E, Bonezzi C, Arcuri E, Yaya

relieving severe cancer-related neuropathic pain and Tur R, Maltoni M, et al. Gabapentin for neuropathic
limiting adjuvant analgesic use in Japan. J Palliat cancer pain: a randomized controlled trial from the
Med. 2016;19(10):1051–9. Gabapentin Cancer Pain Study Group. J Clin Oncol.
9. Jonkman K, van de Donk T, Dahan A. Ketamine for 2004;22(14):2909–17.
cancer pain: what is the evidence? Curr Opin Support 24. Arai YC, Matsubara T, Shimo K, Suetomi K,

Palliat Care. 2017;11(2):88–92. Nishihara M, Ushida T, et al. Low-dose gabapentin as
10. Mercadante S, Lodi F, Sapio M, Calligara M, Serretta useful adjuvant to opioids for neuropathic cancer pain
R. Long-term ketamine subcutaneous continuous when combined with low-dose imipramine. J Anesth.
infusion in neuropathic cancer pain. J Pain Symptom 2010;24(3):407–10.
Manag. 1995;10(7):564–8. 25. Raptis E, Vadalouca A, Stavropoulou E, Argyra E,
11. Waldfogel JM, Nesbit S, Cohen SP, Dy SM. Successful Melemeni A, Siafaka I. Pregabalin vs. opioids for the
treatment of opioid-refractory cancer pain with treatment of neuropathic cancer pain: a prospective,
short-course, low-dose ketamine. J Pain Palliat Care head-to-head, randomized, open-label study. Pain
Pharmacother. 2016;30(4):294–7. Pract. 2014;14(1):32–42.
12. Loveday BA, Sindt J. Ketamine protocol for pallia- 26. Manas A, Ciria JP, Fernandez MC, Gonzalvez ML,
tive care in cancer patients with refractory pain. J Adv Morillo V, Perez M, et al. Post hoc analysis of pre-
Pract Oncol. 2015;6(6):555–61. gabalin vs. non-pregabalin treatment in patients
13. Okamoto Y, Tsuneto S, Tanimukai H, Matsuda Y,
with cancer-related neuropathic pain: better pain
Ohno Y, Tsugane M, et al. Can gradual dose titration of relief, sleep and physical health. Clin Transl Oncol.
ketamine for management of neuropathic pain prevent 2011;13(9):656–63.
psychotomimetic effects in patients with advanced 27. Di Stefano G, La Cesa S, Truini A, Cruccu G. Natural
cancer? Am J Hosp Palliat Care. 2013;30(5):450–4. history and outcome of 200 outpatients with classi-
14. Xu J, Davis CS, Cruz M, Lurie P. State naloxone cal trigeminal neuralgia treated with carbamazepine
access laws are associated with an increase in the or oxcarbazepine in a tertiary centre for neuropathic
number of naloxone prescriptions dispensed in retail pain. J Headache Pain. 2014;15:34.
pharmacies. Drug Alcohol Depend. 2018;189:37–41. 28. Razazian N, Baziyar M, Moradian N, Afshari D,

15. Lambdin BH, Davis CS, Wheeler E, Tueller S, Kral Bostani A, Mahmoodi M. Evaluation of the efficacy
AH. Naloxone laws facilitate the establishment of and safety of pregabalin, venlafaxine, and carbam-
overdose education and naloxone distribution pro- azepine in patients with painful diabetic peripheral
grams in the United States. Drug Alcohol Depend. neuropathy. A randomized, double-blind trial.
2018;188:370–6. Neurosciences (Riyadh). 2014;19(3):192–8.
16. Davis CS, Southwell JK, Niehaus VR, Walley AY, 29. Demant DT, Lund K, Vollert J, Maier C, Segerdahl
Dailey MW. Emergency medical services nalox- M, Finnerup NB, et al. The effect of oxcarbaze-
one access: a national systematic legal review. Acad pine in peripheral neuropathic pain depends on
Emerg Med. 2014;21(10):1173–7. pain phenotype: a randomised, double-blind,
17.
Davis CS, Burris S, Beletsky L, Binswanger placebo- controlled phenotype-stratified study. Pain.
IMMM. Co-prescribing naloxone does not increase 2014;155(11):2263–73.
liability risk. Subst Abus. 2016;37(4):498–500. 30. Smith EM, Pang H, Cirrincione C, Fleishman S,

18. McCabe SE, Cranford JA, West BT. Trends in pre- Paskett ED, Ahles T, et al. Effect of duloxetine on
scription drug abuse and dependence, co-occurrence pain, function, and quality of life among patients
with other substance use disorders, and treatment with chemotherapy-induced painful peripheral
utilization: results from two national surveys. Addict neuropathy: a randomized clinical trial. JAMA.
Behav. 2008;33(10):1297–305. 2013;309(13):1359–67.
19. Meyer R, Patel AM, Rattana SK, Quock TP, Mody 31. Hirayama Y, Ishitani K, Sato Y, Iyama S, Takada K,
SH. Prescription opioid abuse: a literature review of Murase K, et al. Effect of duloxetine in Japanese
the clinical and economic burden in the United States. patients with chemotherapy-induced peripheral neu-
Popul Health Manag. 2014;17(6):372–87. ropathy: a pilot randomized trial. Int J Clin Oncol.
20. Moeller KE, Kissack JC, Atayee RS, Lee KC. Clinical 2015;20(5):866–71.
interpretation of urine drug tests: what clinicians need 32. Henry NL, Banerjee M, Wicha M, Van Poznak

to know about urine drug screens. Mayo Clin Proc. C, Smerage JB, Schott AF, et al. Pilot study of
2017;92(5):774–96. duloxetine for treatment of aromatase inhibitor-
21. Moeller KE, Lee KC, Kissack JC. Urine drug screen- associated musculoskeletal symptoms. Cancer.
ing: practical guide for clinicians. Mayo Clin Proc. 2011;117(24):5469–75.
2008;83(1):66–76. 33. Henry NL, Unger JM, Schott AF, Fehrenbacher L,
22. Ross JR, Goller K, Hardy J, Riley J, Broadley K, Flynn PJ, Prow DM, et al. Randomized, multicenter,
A'Hern R, et al. Gabapentin is effective in the treat- placebo-controlled clinical trial of duloxetine versus
placebo for aromatase inhibitor-associated arthralgias blind placebo-controlled study. Am J Hosp Palliat
in early-stage breast cancer: SWOG S1202. J Clin Care. 2012;29(3):177–82.
Oncol. 2018;36(4):326–32. 38. Lopez RE. Treatment of acute and chronic focal neu-
34. Matsuoka H, Makimura C, Koyama A, Otsuka M, ropathic pain in cancer patients with lidocaine 5%
Okamoto W, Fujisaka Y, et al. Pilot study of dulox- patches. A radiation and oncology department experi-
etine for cancer patients with neuropathic pain ence. Support Care Cancer. 2013;21(5):1329–34.
non-responsive to pregabalin. Anticancer Res. 39. Fleming JA, O'Connor BD. Use of lidocaine patches
2012;32(5):1805–9. for neuropathic pain in a comprehensive cancer
35. Matsuoka H, Ishiki H, Iwase S, Koyama A,
Centre. Pain Res Manag. 2009;14(5):381–8.
Kawaguchi T, Kizawa Y, et al. Study protocol for 40. Kern KU, Nalamachu S, Brasseur L, Zakrzewska

a multi- institutional, randomised, double-blinded, JM. Can treatment success with 5% lidocaine medi-
placebo-controlled phase III trial investigating addi- cated plaster be predicted in cancer pain with neuro-
tive efficacy of duloxetine for neuropathic cancer pathic components or trigeminal neuropathic pain? J
pain refractory to opioids and gabapentinoids: the Pain Res. 2013;6:261–80.
DIRECT study. BMJ Open. 2017;7(8):e017280. 41. Crawford P, Xu Y. Topical capsaicin for treatment
36. Banerjee M, Pal S, Bhattacharya B, Ghosh B, Mondal of chronic neuropathic pain in adults. Am Fam
S, Basu J. A comparative study of efficacy and Physician. 2017;96(11):Online.
safety of gabapentin versus amitriptyline as coan- 42. Schumacher M, Pasvankas G. Topical capsaicin for-
algesics in patients receiving opioid analgesics for mulations in the management of neuropathic pain.
neuropathic pain in malignancy. Indian J Pharmacol. Prog Drug Res. 2014;68:105–28.
2013;45(4):334–8. 43. Ellison N, Loprinzi CL, Kugler J, Hatfield AK, Miser
37. Mishra S, Bhatnagar S, Goyal GN, Rana SP,
A, Sloan JA, et al. Phase III placebo-controlled trial
Upadhya SP. A comparative efficacy of amitrip- of capsaicin cream in the management of surgical
tyline, gabapentin, and pregabalin in neuropathic neuropathic pain in cancer patients. J Clin Oncol.
cancer pain: a prospective randomized double- 1997;15(8):2974–80.
Interventions for Refractory Pain
in Cancer Patients
52
Michael G. Kaplitt
Introduction echanisms of Pain in Cancer

M
Patients
Pain is a serious complication of cancer that can
severely limit quality of life as well as reduce lon- Pain is a common problem in patients with can-
gevity due to inability to perform activities essen- cer, particularly metastatic cancer. As with any
tial to maintaining health, such as physical activity pain, it can sometimes reflect a structural prob-
and proper nutrition. Medical therapy is generally lem that needs to be addressed, such as spinal
the first treatment for patients with severe cancer instability from a destructive lesion leading to
pain, often consisting of opiates. When this is mechanical back pain. This should be addressed
inadequate, however, alternatives for pain control if possible with resection and stabilization of the
are necessary. There are also increasing societal spine. However, for many cancer patients, chronic
pressures regarding responsible administration pain is a disease that needs to be addressed and
and consumption of opiates for chronic diseases. does not necessarily reflect a more proximate
In the cancer population, however, sensitivities problem that can be discretely fixed. The major
around opiate abuse must be considered in the cause of pain is generally activation of nocicep-
context of often recalcitrant pain and limited lifes- tors by cancer cells or factors released by tumors,
pan. Radiation and/or chemotherapy, to control leading to typical aching, difficult to localize
the lesion(s) responsible for the pain as a pallia- nociceptive pain [1]. This type of pain is gener-
tive measure, are also often successful at reducing ally treated with opiates, but resistance to opiates
pain adequately to relieve distress and improve can develop, and in patients with longer life
quality of life. When these treatments are unsuc- expectancies, concerns about opiate dependence
cessful or not feasible for a given situation, there and abuse are increasingly common even in the
are a variety of more interventional procedures cancer pain population. There is also a paradoxi-
which can be quite effective for patients with can- cal opiate-induced hyperalgesia that can occur
cer-related pain. Here, we will review the causes from chronic opiate use [2]. In addition to releas-
of pain in cancer and current surgical procedures ing factors which activate nociceptors, tumor
available for treating these complex patients. cells can also promote or induce inflammation,
both locally within the cancer microenvironment
M. G. Kaplitt (*) and systemically, and this can cause or worsen
Department of Neurological Surgery, NewYork- pain as well. Cancerous lesions can also irritate
Presbyterian/Weill Cornell Medicine, local neurons which induce pain. Surgical thera-
New York, NY, USA pies for nociceptive cancer pain described below
e-mail: mik2002@med.cornell.edu

https://doi.org/10.1007/978-3-030-42958-4_52
688 M. G. Kaplitt
are often tailored to patients based upon the type methacrylate) [4]. This stabilizes the local bone
and location of the malignancy, the type of pain while also potentially limiting effects of the
that they experience, and the nature of their tumor on local nerve endings. Vertebroplasty
responses and/or adverse effects to drug therapy involves simply injecting cement into the affected
for pain. vertebral body, while kyphoplasty uses balloon
Cancer therapies can cause pain through dif- inflation to restore the lost height of a fractured
ferent mechanisms than the cancers themselves vertebral body prior to injecting cement. While
[3]. These include neuropathies from radiation or these are widely used for treatment of osteopo-
chemotherapy, due to various known and rotic compression fractures, they were originally
unknown mechanisms that can permanently alter developed to treat hemangiomas and primary
the function of sensory and pain neurons, leading bone tumors and have been studied extensively in
to a more neuropathic type of pain, although treatment of pain from vertebral metastases. Both
inflammation from cancer therapy can also cause procedures have been shown to substantially
pain. Neuropathic pain is generally treated with reduce pain from metastatic vertebral disease
antiepileptic medication or antidepressants, as (60–70% or more) with substantial improve-
with non-cancer neuropathic pain; however, it ments in quality of life [5]. An ongoing random-
may be more difficult to obtain satisfactory ized phase 3 study is exploring a combination of
responses to neuropathic pain in cancer patients kyphoplasty and radiation to determine if there is
as compared to non-oncologic neuropathic pain. superior short- and long-term pain control com-
Therefore, surgical therapies for patients with pared with patients receiving radiation alone,
sufficiently severe and intractable treatment- based upon phase 1/2 data showing substantial
related pain should focus upon those therapies, promise for this approach [6].
which are most appropriate for the mechanism Ablation of specific tracts within the spinal
causing the pain, which is generally neuropathic cord has a long history in treating cancer pain
in nature. patients [7]. Cordotomy has been the most widely
studied, yet the availability of the technique has
become increasingly limited due to lack of ade-
esective and Ablative Spinal
R quately trained practitioners, leading to an unde-
Procedures for Cancer Pain rutilization of this procedure [8]. Nonetheless,
the vast majority of reports indicate that this is a
Spinal metastases are common in patients with very effective and safe procedure in appropriate
cancer, occurring in up to 10% of patients with patients with cancer pain treated by experienced
malignancies. Gross instability from destructive practitioners [9]. The goal is to lesion the lateral
lesions can lead to severe mechanical back pain, spinothalamic tract, usually between the C1 and
which is characterized by pain with motion. This C2 spinal levels, in order to interrupt nociceptive
usually requires tumor resection and surgical fibers emanating from the contralateral body
fusion with hardware in order to promote healing below the level of the lesion. This tract also sub-
and prevent spinal cord injury while also resolv- serves light touch and temperature, and so, these
ing the pain. For patients with vertebral body dis- functions may be disturbed as a result of a suc-
ease and pathological fractures, however, there cessful procedure. Given this anatomy and physi-
are less invasive options. These patients usually ology, the optimal candidate for the procedure is
do not have gross spinal instability but have pain a cancer pain with nociceptive pain, usually vis-
either from biological factors released locally ceral pain, below the level of the lesion and pref-
from the tumor or from microscopic instability. erably in one hemibody in order to avoid the need
Vertebroplasty and kyphoplasty are percutaneous for bilateral cordotomies which can be less effec-
procedures that use fluoroscopic guidance to tive and more morbid [10]. Initially, the proce-
insert a needle into the affected vertebral body in dure was performed with open surgery, but for at
order to inject bone cement (such as methyl least the past 20 years, it has been largely a per-
52 Interventions for Refractory Pain in Cancer Patients 689
cutaneous procedure performed with CT guid- chial plexus avulsion, which can lead to distorted
ance [11, 12]. The patient undergoes a CT anatomy at the DREZ region due to degenerative
myelogram to identify the space for entry above changes following the plexus injury. There have
C2. A needle enters the skin roughly just below been many isolated reports and small series
the mastoid, then penetrates the CSF, and enters where DREZ has been explored in cancer pain
the spinal cord in an anterolateral location. The [16]. Most of these studies have unsurprisingly
lesioning radiofrequency probe is then passed explored DREZ lesions for either Pancoast
through the needle and into the spinal cord. Test tumors of the upper lung, which can impinge
stimulation at high frequency confirms the pres- upon or infiltrate the brachial plexus, or for bra-
ence of paresthesias and/or temperature changes chial plexitis and other neuropathic pain syn-
in the contralateral body, while low-frequency dromes following radiation-induced injury. There
stimulation is performed to activate neurons in have been some very promising outcomes in
order to confirm that the nearby corticospinal these reports, but to date, no definitive large or
tract is not being activated at a threshold that is randomized study has been performed in cancer
too low. If this does happen and motor contrac- patients to clarify the best candidates for this
tions occur at a low-voltage threshold, this sug- treatment.
gests that the probe is too close to the corticospinal
tract, risking hemiplegia if lesioning continues,
so the needle and probe must be repositioned. A I ntrathecal Delivery of Medication
radiofrequency lesion is then performed at for Cancer Pain
roughly 70–80 °C for 60 seconds, similar to other
RF lesions such as those used to treat trigeminal Perhaps the most common procedure currently
neuralgia. Midline myelotomy is another spinal for treating intractable cancer pain is placement
cord ablation technique, which is less technically of an intrathecal pump. This is designed to deliver
challenging as this enters directly in the center of narcotics directly into the CSF, thereby limiting
the dorsal spinal cord to create a punctate lesion dose to the brain and essentially eliminating sys-
that interrupts midline posterior column fibers as temic toxicities from oral opiates [17]. For cancer
well as crossing fibers. This has been shown to be patients with a very short life expectancy of only
effective in a small series of patients with visceral a few weeks, the treatment goals can frequently
pelvic and abdominal pain, but this has not been be achieved with placement of an externalized
nearly as widely studied nor as clearly effective epidural catheter with constant epidural infusion.
as cordotomy [13]. These are not generally effective for long-term
Lesioning of the dorsal root entry zone treatment of months to years, however, due to the
(DREZ) is another ablative procedure that has likelihood of catheter obstruction when not in a
been used in neurosurgical treatment of pain for fluid compartment as well as the risk of infection
decades. This targets the neurons of the dorsal from a long-term externalized device. While
horn, as well as the lateral portion of the dorsal there is certainly a risk of both infection and cath-
root fibers and a portion of the local projections eter obstruction or malfunction with permanent
between levels known as Lissauer’s tract. A small intrathecal systems, these are very low risk even
hemilaminotomy is performed at the appropriate in medically complex late-stage cancer patients.
spinal level, followed by a durotomy to expose Such permanent systems should be considered
the spinal cord and existing dorsal roots. The dor- for those patients with evidence of response to
sal rootlets are then elevated to expose the lateral systemic narcotics who either cannot obtain
entry zone. A lesion can then be created either by adequate pain relief or have unacceptable adverse
bipolar cautery or by insertion of a probe fol- effects to these medications.
lowed by radiofrequency lesioning; laser ablation Prior to surgery, patients often undergo a trial
has also been reported [14, 15]. It has most com- of epidural or intrathecal medication to deter-
monly been used for neuropathic pain from bra- mine the likelihood of response to a permanent
690 M. G. Kaplitt
implant. While this is common for degenerative from the catheter, it will dilute along a gradient of
spine patients, we have published an algorithm CSF. As such, if the catheter is not close to the
for evaluation of cancer patients who may be target area, then it is difficult with these agents to
optimal for this treatment without the need for an achieve adequate effectiveness at the desired spi-
invasive trial [18]. Cancer patients can have nal cord target. Combining opiates with agents
blunted immune systems, coagulopathies, or such as bupivacaine or clonidine can be particu-
other problems which make any intervention larly effective in patients with a mixed picture of
somewhat risky, and therefore, the ability to iden- nociceptive and neuropathic pain.
tify candidates for an intrathecal pump without One long-term concern that is mostly relevant
the need for an invasive trial can be very helpful for patients with longer life expectancies is the
in this population. If a trial is done, we prefer to development of inflammatory masses at the tip
have externalized catheters removed the day of the catheter [19]. These usually occur after
before permanent implant to reduce the risk of many months and often years following the onset
infection. A longer period following removal of treatment. They can lead not only to obstruc-
would be desirable to further reduce any risk, but tion of the catheter tip but also to tethering of
in our experience of implanting these devices for the spinal cord at the site and eventually to frank
nearly 20 years at a major international cancer spinal cord compression with associated symp-
center, the risks of infection with this approach toms. If symptoms are mild or if the patient is
have been minimal, while the need for immediate asymptomatic, then reducing or eliminating the
implant in this particular population is usually drug that is causing the problem can lead to reso-
very high. lution. However, if the mass is large and causing
The surgical procedure for implanting a sys- spinal cord compression, then open resection as
tem is fairly straightforward. Patients are placed if the mass were a tumor can be required to pre-
in the lateral position, and fluoroscopy is used to vent permanent spinal cord injury [20]. Given the
identify the insertion point and to follow the cath- relatively shorter life expectancy of metastatic
eter during implantation. The thecal sac is entered cancer patients compared with the larger popula-
below the conus and usually at a less mobile level tion of degenerative spine patients with pumps,
such as L3/4 to minimize risk of catheter migra- this is usually thought to be less of a concern.
tion. The needle tip should be in the middle of the However, while any drug can lead to this prob-
spinal canal by fluoroscopy since CSF flow from lem, it is more common with agents that are off-
the needle can still occur if the level is only par- label or made by compounding pharmacies [19].
tially in the subarachnoid space either posteriorly Regardless of drug, though, any patient with
or anteriorly. The catheter should go in several a pump and new neurologic spinal symptoms
levels to reduce the risk of migration and extru- should prompt consideration of an inflammatory
sion, and once in a good location, the anchor catheter mass in the differential.
should be buried in the fascia, and the neck of the
anchor should be sutured to reduce the risk of
toggling that could also promote catheter extru- Spinal Stimulation for Cancer Pain
sion. The level of the catheter tip is less important
when pure opiate such as morphine are used Neuromodulation devices such as spinal cord
since they will diffuse into the CSF. However, stimulation have become very popular for treat-
when mixtures of other agents are used, particu- ment of neuropathic pain, particularly pain in the
larly when they include local anesthetics such as extremities. The most common application of the
bupivacaine, then the tip should be placed at the technology is in patients with either degenerative
level of the spinal cord with dermatomes in the spinal column disease who have failed to respond
most painful body area to be addressed. This is to either complex spine surgery or conservative
because the effects of such agents are mostly management or in patients with injuries such as
local on the spinal cord, and as the drug emanates orthopedic trauma that have led to long-term neu-
ropathic pain from nerve injury. Cancer pain is has largely been supplanted by spinal stimulation
rarely neuropathic in nature, and therefore, this since the results are largely similar without the
has not been commonly used in such patients. perceived risk of a brain implant. A second more
However, as indicated earlier, treatment for can- medial thalamic target, including the periaque-
cer, such as chemotherapy or radiation, can result ductal and periventricular gray areas, is more rel-
in neuropathies and severe neuropathic pain that evant to cancer pain, since stimulation of this area
is better treated with antiepileptics or antidepres- leads to natural opiate release, resulting in a feel-
sants than with opiates [3]. When this is inad- ing of warmth that often leads to pain relief [25].
equate, spinal stimulation can be very effective As with the more lateral target, this has also been
[21]. Spinal stimulation is usually trialed with largely replaced by intrathecal pumps which can
externalized leads placed percutaneously into increase CSF opiates without the need for brain
the epidural space, unless there are structural penetration. A third target that is still considered
problems such as scar or hardware that prevents in some centers for refractory patients is the cin-
passage of leads to the correct level. In those gulate cortex. This is a critical center for process-
cases, a surgical paddle lead can be placed at the ing affective components of pain such as distress.
target spinal level through a small laminotomy. Lesioning this area (cingulotomy) can lead to pain
Generally, pain relief of greater than 50% during relief, although it does not block either the periph-
the 5–7-day trial period is considered necessary eral pain signals or central perception of pain but
to justify proceeding with a permanent implant. rather reduces the consequences of pain [26].
This increases the likelihood of success follow- Patients often report that they still feel pain, but it
ing a permanent implant, since those with a more no longer causes them anxiety or distress, and
modest response are unlikely to do better with a overall, their quality of life is generally improved
longer period of stimulation. despite ongoing perception of pain. Therefore,
Traditional spinal stimulation devices for neu- this is generally reserved for patients with few
ropathic pain used relatively low-frequency stim- alternatives to address either the underlying cause
ulation (10–40 Hz) to induce paresthesias in the of the pain or initial processing of pain signals
area of pain, based upon the Melzak and Wall with more common procedures. Although neuro-
gate theory [22]. However, many devices now stimulation can be performed in this region,
offer higher-frequency stimulation (from lesioning of the cingulate is preferred for cancer
1000 Hz to greater than 50,000 Hz) which are patients as this does not require any device
paresthesia-free and work by theoretically differ- implant, and usually, this is used in patients who
ent mechanisms based upon the frequency range are in later stages of disease. Cingulotomy is usu-
[23]. These have only been available commer- ally performed with a minimally invasive burr
cially for a limited number of years, and so, it is hole followed by stereotactic insertion of a radio-
unclear whether these may have a greater poten- frequency probe into the anterior cingulate; how-
tial in cancer patients to treat nociceptive pain ever, radiosurgery has also been used for a less
from cancer in addition to treatment-induced invasive approach [27]. Although radiosurgery is
neuropathies. attractive, the efficacy of radiosurgery for func-
tional goals usually is not evident for 2–3 months
after treatment, since the outcome depends upon
Brain Procedures for Cancer Pain the response of target neurons and supportive
cells to radiation and the resulting radiation-
Targeting brain regions which process pain has induced cell death, which is generally not imme-
long been of interest for patients with refractory diate. This has to be considered when deciding
pain syndromes, yet they are still rarely used in upon a method for performing a cingulotomy in a
most centers [9]. One brain region used for neuro- cancer pain patient, since the life expectancy of
pathic pain is the lateral thalamus, including the the patient may influence whether a more imme-
major sensory thalamic nucleus VPl [24], but this diate response is necessary.
692 M. G. Kaplitt
A new method for immediate lesioning deep 1 year. These patients were mostly those with
brain targets without invasive surgery and with- nonmalignant causes for their pain. However, the
out radiation may ultimately hold promise for ability to provide a noninvasive option for lesion-
treating cancer pain patients who are refractory ing various brain targets associated with pain,
to or not candidates for extracranial therapies. without the need for radiation (which might also
MR-guided focused ultrasound (MRgFUS) was be a concern in patients with prior radiation to the
recently approved in many countries, including brain) and with immediate responses, may be of
the United States and Europe, to perform nonin- great utility to cancer pain patients in the future.
vasive thalamotomies for essential tremor.
Ultrasound can traverse the skull and pass
through the brain with relative safety, but it is Conclusions
generally low energy. With MRgFUS, a helmet
with an array of 1000 ultrasound transducers is Cancer patients have unique pain needs that are
placed over the head of a patient after fixation in sometimes not accommodated by traditional oral
an external frame to prevent movement of the opiate medications. In patients with short life
head during treatment [28]. The transducers are expectancy, externalized intrathecal catheters are
all focused upon a single point in the center of the a quick, effective way of alleviating pain. In
imaginary sphere of the helmet, so that the ultra- patients with longer life expectancies, implanted
sound beams from each transducer converge at intrathecal pumps are generally the mainstay of
the same target point. As a result, a large amount interventional pain techniques. Techniques like
of energy can be delivered to a deep brain target cordotomy, DREZ lesioning, spinal stimulation,
when it is matched to the focal point of the array, and others have their role, and their consideration
and the energy is sufficient to raise the tempera- should be individualized to each patient. Newer
ture of the tissue to a lesional level of technologies like MR-guided focused ultrasound
55–60 °C. Using MR thermometry, a heat map hold promise for noninvasive treatment of cancer-
can be generated showing the volume of tissue related pain.
that was raised to a particular temperature. When
the appropriate temperature is achieved, the tis-
sue is ablated, and when the lesion is therapeuti- References
cally effective, the benefits are generally observed
immediately. For tremor patients, the target is the 1. Chwistek M. Recent advances in understanding and
cerebellar receiving area of the thalamus (so- managing cancer pain. F1000Res. 2017;6:945.
called Vim nucleus), and when the target is 2. Roeckel LA, Le Coz GM, Gaveriaux-Ruff C, Simonin
F. Opioid-induced hyperalgesia: cellular and molecu-
ablated with focused ultrasound, the tremor usu- lar mechanisms. Neuroscience. 2016;338:160–82.
ally improves immediately on the table [29]. The 3. Smith EM, Bridges CM, Kanzawa G, Knoerl R,
procedure is performed entirely within the MRI Kelly JP, Berezovsky A, Woo C. Cancer treatment-
machine, and patients can usually be sent home related neuropathic pain syndromes–epidemiology
and treatment: an update. Curr Pain Headache Rep.
the same day. 2014;18:459.
While this technique has been most widely 4. Aparisi F. Vertebroplasty and kyphoplasty in vertebral
used for tremor, there is great interest in applica- osteoporotic fractures. Semin Musculoskelet Radiol.
tions for pain. One of the first clinical reports of 2016;20:382–91.
5. Sorensen ST, Kirkegaard AO, Carreon L, Rousing
MRgFUS was in fact an application of thalamot- R, Andersen MO. Vertebroplasty or kyphoplasty
omy for pain, targeting the lateral thalamic region as palliative treatment for cancer-related vertebral
described above [30, 31]. This is adjacent to the compression fractures: a systematic review. Spine J.
area of the thalamus targeted for essential tremor 2019;19:1067–75.
6. Bludau F, Welzel G, Reis T, Abo-Madyan Y, Sperk
so the procedure is technically very similar. E, Schneider F, Clausen S, Ruder AM, Obertacke U,
Lesion efficiency was good, and patients experi- Ghaly MM, et al. Combined kyphoplasty and intra-
enced roughly 40–60% improvement of pain at operative radiotherapy (Kypho-IORT) versus exter-
nal beam radiotherapy (EBRT) for painful vertebral 20. Tomycz ND, Ortiz V, McFadden KA, Urgo L,

metastases – a randomized phase III study. BMC Moossy JJ. Management of symptomatic intrathecal
Cancer. 2019;19:430. catheter-associated inflammatory masses. Clin Neurol
7. Harsh V, Viswanathan A. Surgical/radiological inter- Neurosurg. 2012;114:190–5.
ventions for cancer pain. Curr Pain Headache Rep. 21. Peng L, Min S, Zejun Z, Wei K, Bennett MI. Spinal
2013;17:331. cord stimulation for cancer-related pain in adults.
8. Raslan AM, Cetas JS, McCartney S, Burchiel Cochrane Database Syst Rev. (2015):Cd009389.
KJ. Destructive procedures for control of can- 22. Melzack R, Wall PD. Pain mechanisms: a new theory.
cer pain: the case for cordotomy. J Neurosurg. Science. 1965;150:971–9.
2011;114:155–70. 23. Ahmed S, Yearwood T, De Ridder D, Vanneste

9. Burchiel KJ, Raslan AM. Contemporary concepts of S. Burst and high frequency stimulation: underly-
pain surgery. J Neurosurg. 2019;130:1039–49. ing mechanism of action. Expert Rev Med Devices.
10. Raslan AM, Burchiel KJ. Neurosurgical advances in 2018;15:61–70.
cancer pain management. Curr Pain Headache Rep. 24. Kovanlikaya I, Heier L, Kaplitt M. Treatment of

2010;14:477–82. chronic pain: diffusion tensor imaging identification
11. Kanpolat Y, Ugur HC, Ayten M, Elhan AH. Computed of the ventroposterolateral nucleus confirmed with
tomography-guided percutaneous cordotomy for successful deep brain stimulation. Stereotact Funct
intractable pain in malignancy. Neurosurgery. Neurosurg. 2014;92:365–71.
2009;64:187–93; discussion 193–184 25. Boccard SG, Pereira EA, Aziz TZ. Deep brain

12. Raslan AM. Percutaneous computed tomography-
stimulation for chronic pain. J Clin Neurosci.
guided radiofrequency ablation of upper spinal cord 2015;22:1537–43.
pain pathways for cancer-related pain. Neurosurgery. 26.
Viswanathan A, Harsh V, Pereira EA, Aziz
2008;62:226–33; discussion 233–224 TZ. Cingulotomy for medically refractory cancer
13. Nauta HJ, Soukup VM, Fabian RH, Lin JT, Grady JJ, pain. Neurosurg Focus. 2013;35:E1.
Williams CG, Campbell GA, Westlund KN, Willis 27. Martuza RL, Chiocca EA, Jenike MA, Giriunas IE,
WD Jr. Punctate midline myelotomy for the relief of Ballantine HT. Stereotactic radiofrequency thermal
visceral cancer pain. J Neurosurg. 2000;92:125–30. cingulotomy for obsessive compulsive disorder. J
14. Nashold BS Jr, Bullitt E. Dorsal root entry zone
Neuropsychiatry Clin Neurosci. 1990;2:331–6.
lesions to control central pain in paraplegics. J 28. Levi Chazen J, Stradford T, Kaplitt MG. Cranial

Neurosurg. 1981;55:414–9. MR-guided focused ultrasound for essential tremor:
15. Sindou M, Fischer G, Goutelle A, Mansuy L. Selective technical considerations and image guidance. Clin
surgery of posterior nerve roots. First results of sur- Neuroradiol. 2019;29:351–7.
gery for pain. Neurochirurgie. 1974;20:391–408. 29. Chazen JL, Sarva H, Stieg PE, Min RJ, Ballon DJ,
16. Gadgil N, Viswanathan A. DREZotomy in the treat- Pryor KO, Riegelhaupt PM, Kaplitt MG. Clinical
ment of cancer pain: a review. Stereotact Funct improvement associated with targeted interruption
Neurosurg. 2012;90:356–60. of the cerebellothalamic tract following MR-guided
17. Ver Donck A, Vranken JH, Puylaert M, Hayek S, focused ultrasound for essential tremor. J Neurosurg.
Mekhail N, Van Zundert J. Intrathecal drug admin- 2018;129:315–23.
istration in chronic pain syndromes. Pain Pract. 30.
Moser D, Zadicario E, Schiff G, Jeanmonod
2014;14:461–76. D. MR-guided focused ultrasound technique in
18. Malhotra VT, Root J, Kesselbrenner J, Njoku I,
functional neurosurgery: targeting accuracy. J Ther
Cubert K, Gulati A, Puttanniah V, Bilsky M, Kaplitt Ultrasound. 2013;1:3.
M. Intrathecal pain pump infusions for intractable 31. Jeanmonod D, Werner B, Morel A, Michels L,

cancer pain: an algorithm for dosing without a neur- Zadicario E, Schiff G, Martin E. Transcranial mag-
axial trial. Anesth Analg. 2013;116:1364–70. netic resonance imaging-guided focused ultrasound:
19. Coffey RJ, Burchiel K. Inflammatory mass lesions noninvasive central lateral thalamotomy for chronic
associated with intrathecal drug infusion catheters: neuropathic pain. Neurosurg Focus. 2012;32:E1.
report and observations on 41 patients. Neurosurgery.
2002;50:78–86; discussion 86–77
Complementary and Integrative
Therapies (CIM) in Patients
53
with CNS Metastasis
Santhosshi Narayanan, Wenli Liu,
and Gabriel Lopez
Introduction [6]. Here we will discuss some of the integrative

treatments that can help symptoms of patients
Cancer patients with metastasis to the brain expe- with metastatic cancer to the brain.
rience a spectrum of symptoms such as headache,
nausea, vomiting, seizures, fatigue, cognitive def-
icits, drowsiness, etc. In general, the use of com- Definitions
plementary health approaches is highest among
individuals with cancer [1–3]. Patients with Complementary and alternative medicine (CAM) is
central nervous system metastasis may resort to defined by the National Center for Complementary
complementary and integrative therapies (CIM) and Alternative Medicine (NCCAM) as a group
in addition to conventional therapies for symptom of diverse medical and health systems, practices,
management or for hope of cure. In an attempt and products that are not normally considered
to meet the patient’s needs and appropriate the to be part of conventional medicine. It is classi-
use of CIM, integrative oncology programs have fied into four broad categories: natural products
developed or are under development in several (e.g., vitamins, minerals, dietary supplements,
cancer centers [4, 5]. herbs), mind and body medicine (e.g., medita-
According to a published expert consensus, tion, yoga, acupuncture), manipulative and other
integrative oncology is defined as a “patient cen- body-based practices (e.g., massage, chiropractic),
tered, evidence-informed field of cancer care that and other CAM practices (e.g., Ayurveda, tradi-
utilizes mind and body practices, natural products tional Chinese medicine, energy therapies). CAM
and/or lifestyle modifications from different tra- includes certain modalities that may or may not
ditions alongside conventional cancer treatments. have high-quality evidence. Alternative medicine
Integrative oncology aims to optimize health, is when a patient makes use of a CAM modality
quality of life, and clinical outcomes across the for which there is no evidence of efficacy in place
cancer care continuum and to empower people of conventional medicine. Complementary medi-
to prevent cancer and become active participants cine is when a patient makes use of CAM modal-
before, during and beyond cancer treatment” ity for which there may or may not be evidence
for its efficacy in combination with conventional
S. Narayanan (*) · W. Liu · G. Lopez medicine.
Palliative, Rehabilitation & Integrative Medicine, The Integrative medicine or complementary and
University of Texas MD Anderson Cancer Center, integrative medicine (CIM) uses evidence-based
Houston, TX, USA approach to merge conventional and nonconven-
e-mail: snarayanan2@mdanderson.org

https://doi.org/10.1007/978-3-030-42958-4_53
696 S. Narayanan et al.
tional therapies. The consortium of Academic alized integrative care plan which may involve a
Health Centers for integrative medicine defines combination of physical, mind-body, and social
integrative medicine as “the practice of medi- aspects of the patient’s health as illustrated in
cine that reaffirms the importance of relation- Fig. 53.1. There is a growing evidence support-
ship between practitioner and patient, focuses ing the use of CIM therapies such as acupunc-
on the whole person, is informed by evidence, ture, massage, and mind-body practices as a part
and makes use of all appropriate therapeutic of the standard of care. Some of the integrative
approaches, health professionals and disci- approaches used in management of symptoms in
plines to achieve optimal health and healing.” patients with central nervous system metastases
Integrative oncology is the application of integra- are listed below [7]. Other areas such as healing
tive medicine to the care of patients with cancer touch, homeopathy, energy therapies, and spe-
and their caregivers [6]. cial diets have insufficient evidence to support
their use as part of the standard of care. Here, we
discuss some of the commonly used integrative
Clinical Consultation approaches in caring for patients with metastatic
disease to the central nervous system.
The goal of the physician consultation is to pro-
vide patients with an integrative care plan tai-
lored to the individual and his/her unique disease Integrative Therapies for Symptom
trajectory [5]. Initial consultation involves a thor- Management in Patients with Brain
ough evaluation of the patient, which includes Metastasis
detailed history of their cancer, current treatment,
medical conditions, presenting symptoms affect- Nausea
ing physical health and emotional health, and
review of the laboratory tests and/or imaging. • Acupuncture
After a comprehensive assessment, the integra- • Mind-body (guided imagery, hypnosis, music
tive oncology physician is able to create a person- therapy, meditation)
Fig. 53.1 Integrative
medicine center model Physical
• Exercise
• Nutrition
• Acupuncture
• Oncology Massage
- Primary Oncology team
- Physical Medicine & Rehabilitation
- Supportive Care
- Cancer Pain
Oncology Con
ative su
lt
Mind-Body gr Physical
Social
at
e
ion
Int
• Health Psychology • Group Programs

• Meditation
• Music Therapy - Social Work
- Support Groups
• Tai Chi/Qi Gong
d-Body
• Yoga
Social
- Psychiatry
Min
- Spiritual Care
Op n
g
ti m a li
a l H e alt h & H e
53 Complementary and Integrative Therapies (CIM) in Patients with CNS Metastasis 697
Fatigue Dry Mouth
• Discuss energy conservation/exercise • Acupuncture

counseling.
• Consider:
–– Physical health assessment/exercise coun- Alternative Therapies
seling by physical therapy
–– Strategies to improve sleep if there is sleep • Education – discuss current evidence/risks
disturbance versus benefits.
–– Yoga and oncology massage (Category 1, • Review motivations for use, and explore other
NCCN guidelines) opportunities to help meet patient objectives
–– Acupuncture using evidence-informed approaches.
• Encourage/support ongoing communication
with conventional oncology care team.
Stress/Anxiety
• Expressive supportive counseling Acupuncture

• Consider:
–– Psychology Acupuncture, a therapy that has been used
–– Psychiatry for more than 2500 years as part of traditional
–– Meditation or other mind-body practices Chinese medicine (TCM), is now used world-
such as yoga wide. The practice of acupuncture involves diag-
–– Oncology massage nostic assessment of a patient’s symptoms based
–– Social work and support groups on TCM principles, selection of acupoints, and
insertion of fine needles into the selected acu-
points. In modern practice of acupuncture, elec-
Insomnia tric stimulation is often applied to the needles
in addition to the traditional manual stimula-
• Sleep hygiene counseling tion. Acupuncture has a well-established safety
• Consider: profile, with minor side effects of local pain
–– Psychology and cognitive behavioral therapy (3.3%), bruising (3.2%), minor bleeding (1.4%),
–– Exercise/physical therapy assessment and orthostatic problems (0.5%) [8]. Its role in
–– Meditation or other mind-body practices managing cancer- and treatment-related symp-
(tai chi, qigong, yoga) or music toms, such as pain, hot flashes, nausea/vomiting,
–– Pulmonary/sleep evaluation fatigue, and xerostomia, is well-recognized [9].
–– Medication Many comprehensive cancer centers incorpo-
rate acupuncture for cancer symptom manage-
ment [4]. Our own published experience in an
Headaches and Neck Pain outpatient cancer care setting has demonstrated
statistically and clinically significant effects of
• Yoga/meditation acupuncture on self-reported symptoms [10].
• Acupuncture Headache, nausea/vomiting, fatigue, pain, and
• Oncology massage focal deficits are common symptoms of patients
who have CNS metastasis. Acupuncture has
shown promising treatment efficacy for manag-
Neuropathy ing headaches, nausea, and tumor-related pain
and a low incidence of adverse effects [11–15].
• Acupuncture Side effects, such as somnolence, change in
• Massage mental status, and constipation, are commonly
seen with use of narcotics and antiemetic drugs. Yoga, tai chi, and qigong are movement-based
Acupuncture treatment, without the common side mind and body practices which combine physi-
effects and adverse drug interactions of pharma- cal postures or movements, breathing techniques,
cological agents, may be particularly a suitable and meditation with the goal to enhance health
adjunct modality for symptom management in and well-being. Yoga has been shown to facili-
this population. tate relief for a multitude of symptoms in cancer,
improving quality of life, sleep, and fatigue [26–
31]. Meditation, meditative movements such as
Massage yoga and qigong, and mindfulness-based stress
reduction have been shown to improve cognitive
Massage is shown to benefit symptoms such as functions in cancer patients and survivors [32–
anxiety and fatigue and leads to improved qual- 34]. Individuals affected by cancer may consider
ity of life in cancer patients [16, 17]. Oncology regular practice of a mind and body approach
massage involves modification of massage tech- in support of overall health goals during cancer
niques in cancer patients. In patients with CNS care, including cognitive benefits [35].
metastasis, precautions need to be undertaken in
the setting of recent surgery and/or history of sei-
zures. Before the massage, the therapists review Physical Well-Being
blood counts and other areas of metastasis and
modify the massage techniques by avoiding cer- Nutrition
tain sites, changing pressure, etc. If patients are
neutropenic, massage is not recommended. Patients with CNS metastasis undergo radiation,
Patients with CNS metastasis may be treated surgery, or chemotherapy or a combination of
with medications such as opioids for pain control these. During these treatments, nutritional pro-
and ondansetron for nausea which can contribute tein requirements may increase, with goals set
to constipation. Massage was shown to relieve by expert consultation with a registered dietician
constipation in several studies [18, 19]. There is (e.g., 1–1.2 g/kg body weight per day). Ketogenic
anecdotal evidence suggesting massage can help diet (KD) is a high-fat, adequate-protein, low-
provide relief for chemotherapy-induced periph- carbohydrate diet. Energy-restricted ketogenic
eral neuropathy [20]. Massage may also be inte- diet has been proposed as metabolic treatment in
grated into chemo-infusion suites to help with primary brain cancer patients, and patients often
anxiety, nausea, and pain [21]. start ketogenic diet on their own without any
supervision. It is based on the theory that tumor
cells depend on glucose for energy metabolism
Mind-Body Practices whereas normal cells in the brain can use the
ketones as a source of energy [36, 37]. However,
Mind-body practices are techniques that could there are no large trials which have shown the
help decrease distress and balance sympathetic benefit of ketogenic diet in patients with CNS
and parasympathetic nervous system [22]. These metastasis. Though anecdotal evidence suggests
include meditation, relaxation, tai chi, qigong, that side effects are minimal and keto diet is tol-
and yoga. The expressive arts such as music erated well in patients with primary brain can-
therapy, art therapy, dance therapy, and journal- cers, we do not have information on the level of
ing are also considered mind and body practices. blood glucose or ketones and amount of calorie
In addition to decreasing distress, mind-body consumption per day that are associated with
practices have additional benefits on neurotrans- antitumor effect [38–40]. Per American Institute
mitters (GABA, glutamate), balancing HPA axis, for Cancer Research (AICR) recommendations
improving immune function, and other physi- for cancer prevention, we advise patients to eat
ological benefits [23–25]. a variety of vegetables, fruits, whole grains, and
legumes such as beans, avoid sugary drinks, limit [47]. Stress leads to persistent increase in sympa-
consumption of energy-dense foods, and limit thetic nervous system activity and hypothalamic-
alcoholic drinks [41]. pituitary axis which in turn causes changes
such as increased blood pressure, heart rate,
etc. Chronic psychological stress also impairs
Exercise memory directly or through mediators of stress
as shown in a study of caregivers of patients with
Fatigue can limit patients from exercising. Cancer dementia [48]. Patients and their spouses are
itself or treatments such as radiation and che- vulnerable to experiencing distress as a result of
motherapy can contribute fatigue. Encouraging the diagnosis and treatment of CNS metastases.
patient participation in a program of regular, Distress may exacerbate memory issues in these
safe exercise, with supervision as appropriate, patients/caregivers and may also contribute to the
may be of benefit for supporting overall health development of headaches. Expressive support-
during and after treatment. Aerobic exercise has ive counseling is recommended in these patients.
neuroprotective benefits as it has been shown We recommend assessing patients for positive
that 1 year of aerobic exercise increased hippo- coping strategies such as hobbies and listening to
campal volume; this translates to higher BDNF music and negative coping strategies such as alco-
which is a mediator of memory formation and hol. Expressive supportive counseling may help
therefore may lead to improved memory func- in addition to referral to psychology or psychia-
tion [42]. Exercise helps to restore muscle mass try based on their symptoms. In addition, mind-
and strength and also helps balance and mobil- body practices may modulate pain/headache by
ity in addition to improving sleep quality [43]. other neural and cognitive mechanisms or may
Current American College of Sports Medicine indirectly influence pain by lowering stress and
recommendations include 150 minutes of aerobic anxiety [49]. It is important to note that caregiv-
exercise per week and 20 minutes of resistance ers may also be afflicted by significant stress and
exercises twice a week. However, we recom- its associated maladies; providers should assess
mend individualizing exercise regimens. Referral for caregiver stress and counsel appropriately.
to physical therapy for exercise counseling and Meditative movement such as yoga can also help
review of energy conservation techniques in the relieve distress experienced by caregivers [50].
setting of fatigue may help in the development of Another commonly reported symptom in can-
an individualized program of activity [44]. cer patients with CNS tumors includes sleep dis-
For patients who are sedentary or decondi- turbance. The root cause is often multifactorial
tioned, tai chi, Qi gong or yoga which are forms and can be related to depression, stress, anxiety,
of meditative movements can be offered at a poor exercise routine, treatment side effects, etc.
lower intensity. Tai chi or yoga may also enhance Sleep impairment can cause worsening of mem-
cognitive function [45, 46]. ory issues [51] and can also contribute to fatigue
and daytime drowsiness. Cognitive behavioral
therapy is the gold standard for management
Psychosocial Well-Being of insomnia. Medications also have a role and
should be prescribed as appropriate .Yoga or tai
Stress and anxiety are commonly reported symp- chi can be used as adjunct modality [52].
toms in patients with cancer. Stress-induced
physiological changes in the body can adversely
affect the patients in many ways. Studies in Herbs and Supplements
breast cancer patients show that patients who
receive comprehensive education for stress man- Patients often use herbs and supplements as part
agement, maintain a healthy diet, and engage in of their cancer care, typically when their cancer
regular physical activity had a survival advantage progresses despite conventional therapy. These
supplements are also used to help decrease side tic acidosis, and coma have been reported [60].
effects from conventional therapy or to augment Increased bleeding risk is associated with some
the anticancer effect of their prescribed therapies. supplements such as ginkgo biloba, fish oil, and
Some patients decline conventional therapies and garlic, and patients should be educated regarding
instead look for alternative treatment options. this risk, with supplements discontinued before
Patients often have a list of natural products that surgical procedure [61].
they currently take or are interested in taking. There are also concerns regarding harmful
Herbs and supplements should be treated simi- contamination of raw Chinese herbal medicines
larly to prescription medications and entered into with heavy metals, which may lead to patient
a patient’s chart. The first step in the discussion is complications, as there is no standardized qual-
to assess motivation for use of herbs and supple- ity control for the herbs and supplements [62].
ments. The second step is to educate patients on Even though some of the herbs have been shown
the effects of supplements on their health and the to inhibit cancer cells in preclinical or laboratory
potential interactions of supplements with their studies, further research is needed for safe human
current treatments based on the best available use [63].
evidence. Some products may cause negative
clinical outcomes due to metabolic interactions,
treatment interactions, organ toxicity, cancer pro- Conclusion
motion, or lack of quality control during the man-
ufacturing process. For example, St. John’s wort Patients are increasingly relying on recommen-
(Hypericum perforatum) may decrease the clini- dations from different sources such as media,
cal efficacy of irinotecan or imatinib by induction the Internet, family members, other patients,
of cytochrome p450 enzymes [53, 54]. and healthcare professionals. It is important for
Certain herbs and supplements are also anti- healthcare providers to be open and nonjudg-
oxidants such as green tea extract (GTE) and vita- mental about CIM options being used or con-
mins A, C, and E. These antioxidant supplements sidered. This will enable patients to have open
may interfere with radiation and chemothera- conversations and not fear disclosure of current
peutic agents that depend on oxidative damage CIM use. Integrated oncology providers are an
to exert their cytotoxic effect [55]. In a popula- essential part of modern cancer care as they can
tion of patients with head and neck cancer, use guide patients in their use of natural products and
of beta-carotene and vitamin E during radiation other CIM treatments in order to optimize safety
treatment was associated with increased local and synergy with their current conventional can-
recurrence and incidence of second primary cancer treatments.
cer [56]. We recommend that patients obtain their
antioxidants through whole food sources until Acknowledgments The authors have no financial con-
more evidence is available regarding the safety of flicts of interests to disclose.
antioxidant supplements during treatment.
Certain concentrated natural products may
also lead to organ damage such as hepatotoxic- References
ity or nephrotoxicity. For example, some green
tea extracts (GTE) have been associated with 1. Navo MA, Phan J, Vaughan C, Palmer JL, Michaud
L, Jones KL, et al. An assessment of the utilization of
drug-induced liver injury [57]. Other biosimi- complementary and alternative medication in women
lar compounds like amygdalin, laetrile (purified with gynecologic or breast malignancies. J Clin Oncol
form of amygdalin), and vitamin B17 (extracted Off J Am Soc Clin Oncol. 2004;22(4):671–7.
from apricot kernels) have noted antiproliferative 2. Richardson MA, Sanders T, Palmer JL, Greisinger A,
Singletary SE. Complementary/alternative medicine
activity in vitro but have been associated with use in a comprehensive cancer center and the implica-
cyanide toxicity in some patients [58, 59]. Life- tions for oncology. J Clin Oncol Off J Am Soc Clin
threatening toxicities such as seizures, severe lac- Oncol. 2000;18(13):2505–14.
3. Barnes PM, Bloom B, Nahin RL. Complementary 18. Lai TK, Cheung MC, Lo CK, Ng KL, Fung YH, Tong
and alternative medicine use among adults and M, et al. Effectiveness of aroma massage on advanced
children: United States. Natl Health Stat Rep. cancer patients with constipation: a pilot study.
2007;2008(12):1–23. Complement Ther Clin Pract. 2011;17(1):37–43.
4. Brauer JA, El Sehamy A, Metz JM, Mao 19. Lamas K, Lindholm L, Stenlund H, Engstrom B,

JJ. Complementary and alternative medicine and sup- Jacobsson C. Effects of abdominal massage in man-
portive care at leading cancer centers: a systematic agement of constipation–a randomized controlled
analysis of websites. J Altern Complement Med (New trial. Int J Nurs Stud. 2009;46(6):759–67.
York, NY). 2010;16(2):183–6. 20. Cunningham JE, Kelechi T, Sterba K, Barthelemy N,
5. Lopez G, McQuade J, Cohen L, Williams JT, Spelman Falkowski P, Chin SH. Case report of a patient with
AR, Fellman B, et al. Integrative oncology physician chemotherapy-induced peripheral neuropathy treated
consultations at a Comprehensive Cancer Center: with manual therapy (massage). Support Care Cancer.
analysis of demographic, clinical and patient reported 2011;19(9):1473–6.
outcomes. J Cancer. 2017;8(3):395–402. 21. Mao JJ, Wagner KE, Seluzicki CM, Hugo A, Galindez
6. Witt CM, Balneaves LG, Cardoso MJ, Cohen L, LK, Sheaffer H, et al. Integrating oncology massage
Greenlee H, Johnstone P, et al (2017) A comprehen- into chemoinfusion suites: a program evaluation. J
sive definition for integrative oncology. J Natl Cancer Oncol Pract. 2017;13(3):e207–e16.
Ins Monogr. 2017(52). https://doi.org/10.1093/ 22. Chaoul A, Milbury K, Sood AK, Prinsloo S, Cohen
jncimonographs/lgx012. L. Mind-body practices in cancer care. Curr Oncol
7. Latte-Naor S, Mao JJ. Putting integrative oncology Rep. 2014;16(12):417.
into practice: concepts and approaches. J Oncol Pract. 23. Streeter CC, Whitfield TH, Owen L, Rein T, Karri SK,
2019;15(1):7–14. Yakhkind A, et al. Effects of yoga versus walking on
8. Melchart D, Weidenhammer W, Streng A, Reitmayr mood, anxiety, and brain GABA levels: a randomized
S, Hoppe A, Ernst E, et al. Prospective investigation controlled MRS study. J Altern Complement Med
of adverse effects of acupuncture in 97 733 patients. (New York, NY). 2010;16(11):1145–52.
Arch Intern Med. 2004;164(1):104–5. 24. Rao RM, Telles S, Nagendra HR, Nagarathna R,

9. Zia FZ, Olaku O, Bao T, Berger A, Deng G, Fan AY, Gopinath K, Srinath S, et al. Effects of yoga on natu-
et al (2017) The National Cancer Institute’s confer- ral killer cell counts in early breast cancer patients
ence on acupuncture for symptom management in undergoing conventional treatment. Comment to:
oncology: state of the science, evidence, and research recreational music-making modulates natural killer
gaps. J Natl Cancer Ins Monogr. 2017(52). https://doi. cell activity, cytokines, and mood states in corpo-
org/10.1093/jncimonographs/lgx005. rate employees Masatada Wachi, Masahiro Koyama,
10. Lopez G, Garcia MK, Liu W, Spano M, Underwood Masanori Utsuyama, Barry B. Bittman, Masanobu
S, Dibaj SS, et al. Outpatient acupuncture effects on Kitagawa, Katsuiku Hirokawa. Med Sci Monit.
patient self-reported symptoms in oncology care: a 2008;14(2):LE3–4.
retrospective analysis. J Cancer. 2018;9(19):3613–9. 25. Streeter CC, Gerbarg PL, Saper RB, Ciraulo DA,

11. Millstine D, Chen CY, Bauer B. Complementary and Brown RP. Effects of yoga on the autonomic nervous
integrative medicine in the management of headache. system, gamma-aminobutyric-acid, and allostasis in
BMJ (Clin Res ed). 2017;357:j1805. epilepsy, depression, and post-traumatic stress disor-
12. Linde K, Allais G, Brinkhaus B, Fei Y, Mehring M, der. Med Hypotheses. 2012;78(5):571–9.
Shin BC, et al. Acupuncture for the prevention of 26. Bower JE, Ganz PA, Dickerson SS, Petersen L, Aziz
tension-type headache. Cochrane Database Syst Rev. N, Fahey JL. Diurnal cortisol rhythm and fatigue in
2016(4):Cd007587. breast cancer survivors. Psychoneuroendocrinology.
13. Lee A, Chan SK, Fan LT. Stimulation of the wrist 2005;30(1):92–100.
acupuncture point PC6 for preventing postoperative 27. Bower JE, Garet D, Sternlieb B, Ganz PA, Irwin

nausea and vomiting. Cochrane Database Syst Rev. MR, Olmstead R, et al. Yoga for persistent fatigue in
2015;(11):Cd003281. breast cancer survivors: a randomized controlled trial.
14. Zhang Y, Lin L, Li H, Hu Y, Tian L. Effects of acu- Cancer. 2012;118(15):3766–75.
puncture on cancer-related fatigue: a meta-analysis. 28. Buffart LM, van Uffelen JG, Riphagen II, Brug J,
Support Care Cancer. 2018;26(2):415–25. van Mechelen W, Brown WJ, et al. Physical and
15. Chiu HY, Hsieh YJ, Tsai PS. Systematic review and psychosocial benefits of yoga in cancer patients and
meta-analysis of acupuncture to reduce cancer-related survivors, a systematic review and meta-analysis of
pain. Eur J Cancer Care. 2017;26(2) https://doi. randomized controlled trials. BMC Cancer. 2012;
org/10.1111/ecc.12457. 12:559.
16. Cassileth BR, Vickers AJ. Massage therapy for symp- 29. Carlson LE, Speca M, Faris P, Patel KD. One year
tom control: outcome study at a major cancer center. J pre-post intervention follow-up of psychological,
Pain Symptom Manag. 2004;28(3):244–9. immune, endocrine and blood pressure outcomes of
17. Russell NC, Sumler SS, Beinhorn CM, Frenkel
mindfulness-based stress reduction (MBSR) in breast
MA. Role of massage therapy in cancer care. J Altern and prostate cancer outpatients. Brain Behav Immun.
Complement Med. 2008;14(2):209–14. 2007;21(8):1038–49.
30. Chandwani KD, Thornton B, Perkins GH, Arun B, 42. Erickson KI, Voss MW, Prakash RS, Basak C, Szabo
Raghuram NV, Nagendra HR, et al. Yoga improves A, Chaddock L, et al. Exercise training increases size
quality of life and benefit finding in women undergo- of hippocampus and improves memory. Proc Natl
ing radiotherapy for breast cancer. J Soc Integr Oncol. Acad Sci U S A. 2011;108(7):3017–22.
2010;8(2):43–55. 43. Committee PAGA. Physical activity guidelines advi-
31. Cohen L, Warneke C, Fouladi RT, Rodriguez MA, sory committee report, 2008, vol. 2008. Washington,
Chaoul-Reich A. Psychological adjustment and sleep D.C.: US Department of Health and Human Services;
quality in a randomized trial of the effects of a Tibetan 2008. p. A1–H14.
yoga intervention in patients with lymphoma. Cancer. 44. Lopez G, Eddy C, Liu W, Li Y, Chen M, Bruera E,
2004;100(10):2253–60. et al. Physical therapist-led exercise assessment and
32. Carlson LE, Tamagawa R, Stephen J, Drysdale E, counseling in integrative cancer care: effects on
Zhong L, Speca M. Randomized-controlled trial of patient self-reported symptoms and quality of life.
mindfulness-based cancer recovery versus support- Integr Cancer Ther. 2019;18:1534735419832360.
ive expressive group therapy among distressed breast 45. Wayne PM, Walsh JN, Taylor-Piliae RE, Wells

cancer survivors (MINDSET): long-term follow-up RE, Papp KV, Donovan NJ, et al. Effect of tai chi
results. Psycho-Oncology. 2016;25(7):750–9. on cognitive performance in older adults: system-
33. Derry HM, Jaremka LM, Bennett JM, Peng J,
atic review and meta-analysis. J Am Geriatr Soc.
Andridge R, Shapiro C, et al. Yoga and self-reported 2014;62(1):25–39.
cognitive problems in breast cancer survivors: a 46. Janelsins MC, Peppone LJ, Heckler CE, Kesler SR,
randomized controlled trial. Psycho-Oncology. Sprod LK, Atkins J, et al. YOCAS(c)(R) yoga reduces
2015;24(8):958–66. self-reported memory difficulty in cancer survivors
34. Oh B, Butow PN, Mullan BA, Clarke SJ, Beale PJ, in a nationwide randomized clinical trial: investigat-
Pavlakis N, et al. Effect of medical Qigong on cogni- ing relationships between memory and sleep. Integr
tive function, quality of life, and a biomarker of inflam- Cancer Ther. 2016;15(3):263–71.
mation in cancer patients: a randomized controlled 47. Andersen BL, Thornton LM, Shapiro CL, Farrar WB,
trial. Support Care Cancer. 2012;20(6):1235–42. Mundy BL, Yang HC, et al. Biobehavioral, immune,
35. Milbury K, Mallaiah S, Mahajan A, Armstrong T, and health benefits following recurrence for psycho-
Weathers SP, Moss KE, et al. Yoga program for logical intervention participants. Clin Cancer Res.
high-grade glioma patients undergoing radiother- 2010;16(12):3270–8.
apy and their family caregivers. Integr Cancer Ther. 48. Oken BS, Fonareva I, Wahbeh H. Stress-related cog-
2018;17(2):332–6. nitive dysfunction in dementia caregivers. J Geriatr
36. Maurer GD, Brucker DP, Bahr O, Harter PN,
Psychiatry Neurol. 2011;24(4):191–8.
Hattingen E, Walenta S, et al. Differential utilization 49. Bushnell MC, Ceko M, Low LA. Cognitive and emo-
of ketone bodies by neurons and glioma cell lines: a tional control of pain and its disruption in chronic
rationale for ketogenic diet as experimental glioma pain. Nat Rev Neurosci. 2013;14(7):502–11.
therapy. BMC Cancer. 2011;11:315. 50. Lopez G, Chaoul A, Powers-James C, Eddy CA,

37. Chang HT, Olson LK, Schwartz KA. Ketolytic and Mallaiah S, Gomez TI, et al. Group yoga effects
glycolytic enzymatic expression profiles in malignant on cancer patient and caregiver symptom dis-
gliomas: implication for ketogenic diet therapy. Nutr tress: assessment of self-reported symptoms at a
Metab (Lond). 2013;10(1):47. comprehensive cancer center. Integr Cancer Ther.
38. Nebeling LC, Miraldi F, Shurin SB, Lerner E. Effects 2018;17(4):1087–94.
of a ketogenic diet on tumor metabolism and nutri- 51. Wilckens KA, Tudorascu DL, Snitz BE, Price JC,
tional status in pediatric oncology patients: two case Aizenstein HJ, Lopez OL, et al. Sleep moderates
reports. J Am Coll Nutr. 1995;14(2):202–8. the relationship between amyloid beta and memory
39. Rieger J, Bahr O, Maurer GD, Hattingen E, Franz recall. Neurobiol Aging. 2018;71:142–8.
K, Brucker D, et al. ERGO: a pilot study of keto- 52. Irwin MR, Olmstead R, Carrillo C, Sadeghi N,

genic diet in recurrent glioblastoma. Int J Oncol. Nicassio P, Ganz PA, et al. Tai Chi Chih compared
2014;44(6):1843–52. with cognitive behavioral therapy for the treatment of
40. Artzi M, Liberman G, Vaisman N, Bokstein F,
insomnia in survivors of breast cancer: a randomized,
Vitinshtein F, Aizenstein O, et al. Changes in cerebral partially blinded, noninferiority trial. J Clin Oncol Off
metabolism during ketogenic diet in patients with pri- J Am Soc Clin Oncol. 2017;35(23):2656–65.
mary brain tumors: 1H-MRS study. J Neuro-Oncol. 53. Rahimi R, Abdollahi M. An update on the ability of
2017;132(2):267–75. St. John’s wort to affect the metabolism of other drugs.
41. Kushi LH, Doyle C, McCullough M, Rock CL,
Expert Opin Drug Metab Toxicol. 2012;8(6):691–708.
Demark-Wahnefried W, Bandera EV, et al. American 54. Markert C, Ngui P, Hellwig R, Wirsching T, Kastner
Cancer Society Guidelines on nutrition and physical IM, Riedel KD, et al. Influence of St. John’s wort
activity for cancer prevention: reducing the risk of on the steady-state pharmacokinetics and metab-
cancer with healthy food choices and physical activ- olism of bosentan. Int J Clin Pharmacol Ther.
ity. CA Cancer J Clin. 2012;62(1):30–67. 2014;52(4):328–36.
55. Lawenda BD, Kelly KM, Ladas EJ, Sagar SM,

60. Bromley J, Hughes BG, Leong DC, Buckley

Vickers A, Blumberg JB. Should supplemental anti- NA. Life-threatening interaction between comple-
oxidant administration be avoided during chemo- mentary medicines: cyanide toxicity following inges-
therapy and radiation therapy? J Natl Cancer Inst. tion of amygdalin and vitamin C. Ann Pharmacother.
2008;100(11):773–83. 2005;39(9):1566–9.
56. Bairati I, Meyer F, Gelinas M, Fortin A, Nabid A, 61. Ulbricht C, Chao W, Costa D, Rusie-Seamon E,

Brochet F, et al. Randomized trial of antioxidant Weissner W, Woods J. Clinical evidence of herb-
vitamins to prevent acute adverse effects of radiation drug interactions: a systematic review by the natural
therapy in head and neck cancer patients. J Clin Oncol standard research collaboration. Curr Drug Metab.
Off J Am Soc Clin Oncol. 2005;23(24):5805–13. 2008;9(10):1063–120.
57. Sarma DN, Barrett ML, Chavez ML, Gardiner P, Ko 62. Harris ES, Cao S, Littlefield BA, Craycroft JA,

R, Mahady GB, et al. Safety of green tea extracts: a Scholten R, Kaptchuk T, et al. Heavy metal and pes-
systematic review by the US Pharmacopeia. Drug Saf. ticide content in commonly prescribed individual
2008;31(6):469–84. raw Chinese Herbal Medicines. Sci Total Environ.
58. Mani J, Rutz J, Maxeiner S, Juengel E, Bon D, Roos 2011;409(20):4297–305.
F, et al. Cyanide and lactate levels in patients during 63. Elkady AI, Hussein RA, Abu-Zinadah OA. Effects of
chronic oral amygdalin intake followed by intrave- crude extracts from medicinal herbs Rhazya stricta
nous amygdalin administration. Complement Ther and Zingiber officinale on growth and proliferation of
Med. 2019;43:295–9. human brain cancer cell line in vitro. Biomed Res Int.
59. Milazzo S, Horneber M. Laetrile treatment for cancer. 2014;2014:260210.
Cochrane Database Syst Rev. 2015(4):Cd005476.
The Palliative Care of Patients
with Brain Metastases
54
Rebecca A. Harrison and Eduardo Bruera
Introduction the concept that providing care for cancer patients

should be broader than extending survival.
Central nervous system (CNS) metastasis is a sig- Palliative care was initially delivered to inpa-
nificant and devastating complication of cancer. tients admitted to acute care hospital beds and
Dissemination of disease to the brain and spi- palliative care units [9]. In recent years, multiple
nal cord heralds an aggressive disease trajectory randomized controlled trials have demonstrated
and often imparts a unique and disabling symp- the value of early outpatient palliative care in
tom burden. Within the cancer population, CNS improving multiple physical and psychosocial
metastases present in 10–30% of patients [1, 2]. symptoms and end-of-life quality through reduc-
Brain metastases remain the most common malig- ing emergency room visits, intensive care unit
nant brain tumor in adults [3], with an incidence admissions, and chemotherapy administration
that continues to rise [4, 5]. Intramedullary spine in the last days of life [10]. A number of studies
metastases, while present in only 0.9–5% of can- have found that using the name “supportive care”
cer patients, portend a similarly poor prognosis for their outpatient program increases the likeli-
[6–8]. As such, a fundamental fluency in the care hood of early referral by oncologists [11, 12] and
of these patients among oncologists is essential. results in earlier and higher numbers of patient
referrals [13, 14].
Patient-centered outcomes, such as quality of
The Contribution of Palliative Care life and cognitive performance, are increasingly
incorporated as endpoints in therapeutic clini-
Palliative care is a comprehensive medical and cal trials, acknowledging the impact these factors
interdisciplinary care directed at improving should have on therapeutic decision-making [15,
patient quality of life. Central to its mandate is 16]. In the treatment of brain metastases specifi-
cally, treatment adaptations and interventions are
being evaluated to try to minimize treatment tox-
R. A. Harrison (*) icity and improve patient function [17, 18], and
Department of Neuro-Oncology, The University of
Texas MD Anderson Cancer Center, metrics evaluating cognitive outcomes and quality
Houston, TX, USA of life are routinely incorporated as clinical trial
e-mail: RAHarrison@mdanderson.org endpoints [19, 20].
E. Bruera Patients with CNS metastases are a distinct
Department of Palliative, Rehabilitation, and population (Fig. 54.1). Because of the substrate
Integrative Medicine, The University of Texas MD affected, the disease can cause changes in cog-
Anderson Cancer Center, Houston, TX, USA

https://doi.org/10.1007/978-3-030-42958-4_54
706 R. A. Harrison and E. Bruera
Fig. 54.1 The spectrum Symptoms

of complications in Psychosocial distress Pain
patients with brain Depression Nausea/emesis
metastases Anxiety Fatigue
Brain
Cognitive deficits Motor deficits
metastases
Poor survival
Caregiver distress
prognosis
nition, mobility, language, and independence. cancer, but also they are a therapeutic challenge.
Furthermore, CNS-directed treatments used to Protected by the blood-brain barrier, they are
prolong survival in these patients have the poten- recalcitrant to many systemic chemotherapies,
tial to augment these neurologic deficits. The and CNS-directed therapies such as surgery and
fallout of these changes may include changes in radiation may be limited by the associated neu-
social relationships, decision-making capacity, rologic toxicity. It is often the case that diagnosis
and autonomy. of brain metastases signals likely poor survival,
While possible at any time in the course of ill- but this is not necessarily the case particularly in
ness, CNS metastases most commonly occur in the era of immunotherapy and radiosurgery [24].
the setting of advanced cancer. The vast majority Intramedullary spinal cord metastases have a
of patients with brain metastases have a previ- more precipitous symptom onset than primary
ously identified primary cancer, and most have spinal cord tumors [25, 26]. Surgery is rarely
either primary or secondary lung involvement indicated for intramedullary spinal cord metasta-
before dissemination to the brain [21]. Similarly, ses, given the collateral neurological damage that
most spine metastases are diagnosed just after can result [7]. As such, radiation as stand-alone
3 years after the initial cancer diagnosis [22] and therapy is more typically provided in this popu-
often have disseminated cancer including brain lation. This is contradistinction to patients with
metastases at diagnosis [23]. This renders these vertebral column metastases who are often pal-
patients distinct from patients with primary CNS liated with surgical resection and deformity cor-
tumors, as they have often already accrued end- rection followed by radiation. The mean survival
organ toxicities, fatigue, and psychological and after surgery for intramedullary spine tumors has
social stressors with their cancer prior to their been found to range from 5 to 11.6 months [22,
diagnosis of CNS disease. The nature of their 27]. The majority (80%) of patients with spinal
symptoms and the poor life expectancy associ- metastases die within 3 months of diagnosis of
ated with CNS metastases highlight the value of spine involvement [28]. Most often, it is not the
integrative palliative and antitumor care in the spinal metastasis but the widespread cancer that is
management of patients with CNS metastases. the ultimate cause of mortality in this population.
rognostic Significance of CNS

P Symptom Burden
Metastases
Focal Neurologic Deficits
Involvement of the nervous system in cancer is
associated with poor survival. Not only are CNS Focal neurologic deficits, including motor, sen-
metastases a mark of an aggressive systemic sory, language, and bulbar dysfunction, may
54 The Palliative Care of Patients with Brain Metastases 707
result dependent on anatomic location of the Headache

metastases. These classically obey the structure-
function relationships of the nervous system. Headaches in patients with brain metastases
Motor deficits may arise from involvement of the can arise from mass effect; distortion of pain-
primary motor cortex, spinal cord, supplemen- sensitive intracerebral structures, such as proxi-
tary motor areas, cerebellum, thalamic nuclei, mal vessels, meninges, or venous sinuses; or
and/or deep gray nuclei. Sensory deficits can the development of hydrocephalus. Integral to
similarly occur from involvement of the sensory treating the headache is to understand its patho-
cortex, thalamic nuclei, brainstem, and spinal physiologic contributors. Intracranial pain-sen-
cord. While neuropathic pain is rare in brain sitive structures generate visceral pain, which is
metastases, thalamic lesions have been reported referred to more superficial anatomic structures.
to cause Dejerine-Roussy syndrome, a severe As such, it may not be experienced as originat-
hemibody pain contralateral to the site of the ing from the region of the mass: supratentorial
thalamic lesion [29]. The brainstem carries the lesions commonly generate frontal pain and
neural substrates for fundamental processes such posterior fossa lesions occipital pain. Patients
as respiratory drive, autonomic control, alert- with brain metastases, however, are unlikely to
ness, and oropharyngeal function. Due to the have early morning headache classically associ-
frequent multifocal nature of CNS metastases, a ated with elevated intracranial pressure [33]. The
comprehensive review of neurologic symptoms headaches can take on various semiologies of pri-
is essential in this population. mary headaches, such as tension-type headaches
The nature of neurologic symptoms influences (77%) or, less commonly, migraine headaches or
patient well-being. In patients with primary brain other headache types. A pre-existing headache
tumors, motor deficit, particularly gait impair- history is a risk factor for headache with intracra-
ment, is associated with worsened perceived nial tumor [34].
symptom burden and contributes significantly to
disability at end of life [30]. Motor impairment
has also been identified as the most common rea- Fatigue
son for initiating hospice care [30]. In compar-
ing patients with brain tumors to other systemic Cancer-related fatigue is a pervasive issue in
cancer patients, motor symptoms were a unique patients with brain metastases, as it is in the
contributor to decline in quality of life among cancer population as a whole [35]. Over 95% of
those with brain cancer [31]. These findings are patients receiving chemotherapy or radiotherapy
extrapolated from the primary brain tumor litera- have fatigue, emphasizing its prevalence [36].
ture; further study is warranted to understand the This fatigue is disproportionate to exertion level
precise influence of these symptoms on patient and is not relieved by rest or sleep [37]. Many
quality of life and disease trajectory in those with of the proposed mechanisms of cancer-related
CNS metastases. That said, regardless of pathol- fatigue are central, including alterations in sero-
ogy, substantial motor deficits significantly affect tonin transmission and metabolism [38, 39],
patient’s quality of life and perception of disease hypothalamic-pituitary axis dysfunction [40–42],
burden. and circadian rhythm dysregulation [43]. As
The majority of patients with spinal cord such, it is logical that fatigue would be a promi-
metastases are symptomatic (92%) [32]. Among nent issue in patient’s CNS metastases.
patients with intramedullary spinal cord metas- Similar to other symptoms in this population,
tases, the leading presenting symptoms are dys- baseline fatigue can increase with anticancer ther-
esthesia (77%) followed by paresis (68%) and apies. In evaluating fatigue scores in patients pro-
urinary retention in a minority (23%) [22]. In spectively as they go through whole brain radiation
most patients, neurologic deterioration occurred therapy (WBRT) treatment using several different
in days to weeks from presentation [28]. assessments [35], there was a noted increase in
fatigue scores from baseline to the first month after the potential to be epileptogenic. Seizures are
radiation therapy. In addition, sleep-wake distur- less common in patients with metastases than in
bance occurs at increased frequency in patients those with primary brain tumors, with less than
undergoing brain radiation [44], with reduction a quarter (24%) of patients with brain metasta-
in melatonin secretion and resultant hypothalamic ses being affected [50]. They are most common
dysfunction being proposed mechanisms [45]. in those with melanoma brain metastases (67%)
[50], with hemosiderin irritation of surrounding
brain parenchyma being thought to further lower
Cognitive Dysfunction the seizure threshold. In addition to the structural
lesions themselves, vasogenic edema, medica-
Findings of cognitive impairment have been iden- tions, and intercurrent illness may further lower
tified in one-third of patients with non-CNS cancer the seizure threshold in this population. Aside
even prior to the initiation of anticancer therapy, from the medical implications of seizures, this
a testament to the neuroactive impact of cancer comorbidity can have significant implications for
[46]. This incidence is higher in patients with the patient’s psychological and social well-being.
CNS metastases, with up to 90% of patients hav- They can also contribute markedly to caregiver
ing cognitive impairment at time of diagnosis with stress [51]. As such, they warrant dedicated med-
brain metastases [47]. In addition to the metasta- ical attention, and there should be open discourse
ses themselves, the systemic cancer, cytotoxic and regarding any breakthrough seizures and recom-
hormonal therapies, as well as CNS-directed ther- mended management.
apies such as radiation and radiosurgery may all
contribute to this change. This cognitive change
can have a breadth of implications for the patient, alliative and Supportive Care
P
from navigating treatment decisions, social roles, Interventions
and vocational commitments.
The nature of neurologic deficits in patients
with brain metastases can be loosely predicted by
Supportive/Palliative Interventions for Brain
the neuroanatomy affected, though this popula-
Metastases
tion also has more generalized cognitive changes
• Surgery
than patients with other structural brain diseases.
• Radiation therapy
As a group, patients with brain metastases most
• Chemotherapy
commonly have memory impairment, with defi-
• Steroids
cits in attention, executive function, and language
• Seizure medications
also being present in comparison to healthy con-
• Analgesics
trols [48]. The severity of cognitive impairment
• Rehabilitation
in this population has been found to correlate
• Cognitive exercises and treatment
with total tumor volume [49]. Notably, cognitive
• Caregiver support
deficits have been identified in brain metastases
patients that have no reported functional impair-
ment [48], indicating the importance of aware-
ness of this complication among providers and Surgery
caregivers.
In addition to cytoreduction, surgery can provide
symptomatic relief for patients with CNS metas-
Seizures tases. In retrospective analysis, surgical resection
of brain metastases was found to improve func-
The vast majority of brain metastases occur in tional outcome, reduce neurologic impairment,
the cerebral hemispheres (85%) and thus have and improve quality of life [52]. After surgery,
half of patients regained normal function for a (<6 months from radiation completion), and late
period of time. Similarly, average performance delayed (>6 months from treatment completion)
status was found to improve with surgery in effects [59]. Risk factors for increased toxicity
patients with spinal metastases [22], associated of radiation include age, with very young or old
with improvements in motor and sensory func- populations being more vulnerable, the size of
tion. After surgery, they may be treated with radi- the tumor, and the radiation dose delivered [58].
ation, systemic therapy, or a combination of these The resultant symptoms of brain irradiation
[53]. In appropriately selected patients, surgery result from its impact on cerebral vasculature,
and adjuvant therapies can lead to significant neuroglial cells, and neural progenitor cell popu-
symptom palliation. Patients with poor antici- lations. While whole brain radiation was previ-
pated survival or rapidly progressing systemic ously associated with an acute encephalopathy
disease may not benefit from surgical interven- syndrome [59], this has become less common
tion as the neurologic symptom-free period may given modern fractionation and dosing. Short-
be rapidly eclipsed by symptoms from progres- term memory deficits and verbal fluency deficits,
sive systemic cancer [53]. however, have been identified from 3 months to
1 year following WBRT, with more generalized
cognitive effects occurring and persisting subse-
Radiation Therapy quent to this [60–62]. Increased permeability of
the blood-brain barrier during radiation can lead
Along with surgery, radiation is a frontline treat- to focal vasogenic edema, leading to increased
ment for brain metastases [54, 55]. Whether it be focal neurologic symptoms or seizures. This
stereotactic radiosurgery (SRS) or whole brain increased edema is self-limited, with resolution
radiation therapy (WBRT), most brain metas- in the weeks or months following treatment. In
tases are treated with radiation therapy of some the months and years following radiation, par-
form, regardless of histology. While little study ticularly SRS, radiation necrosis may result;
has focused on the capacity of these adjuvant this needs to be distinguished from tumor pro-
treatments to lead to symptomatic improvement, gression. Whole brain radiation has been found
one study found the period free of neurologic to be associated with greater cognitive impair-
progression after radiation was on the order of ment than SRS and has also been found to impact
weeks, and steroids are often reintroduced to patient-reported quality of life [63, 64]. Attempts
manage progressive symptoms [53]. In addition have been made to narrow the cognitive toxicities
to prolonging patient survival, radiation can con- of radiation including hippocampal sparing tech-
tribute to control of neurologic symptoms [56, niques. In particular, the indications for WBRT
57]. However, toxicities of therapy may also con- have been narrowed in preference of less toxic
tribute significantly to patient symptom burden. strategies [60, 65]. Despite these efforts, current
Fatigue is a common early effect of radiation studies continue to report a sustained decline in
therapy. It often develops during or within a few cognitive function in patients treated with brain
weeks of WBRT completion for brain metasta- radiation.
ses, persisting for weeks after its completion
[58]. More specific neurologic toxicities may
also result from brain radiation. The neurotox- Steroids
icity it imparts is influenced by the radiation
modality used, the dose and fractionation sched- Corticosteroids are recommended to provide
ule, the area of the CNS targeted, and the time symptomatic relief from symptoms on intra-
elapsed since treatment. A temporally based clas- cerebral edema from brain metastases [66], as
sification scheme is frequently used to classify well as reduced tumor-associated pain, nausea,
the neurologic effects of radiation; these include vomiting, and anorexia [67]. The significance of
early acute (during radiation), early delayed steroid response has also been evaluated, with
response to steroids being identified as a posi- worse performance while taking a medicine, and
tive prognostic factor for patient survival [68]. a significantly larger number felt their perfor-
While early symptomatic improvement has been mance was worse with the carbamazepine than
noted with their administration during radiation the levetiracetam.
[69], controversy persists over the precise dosing These agents may also increase fatigue. In
and indications for their use [70]. Individualized particular, those impacting the GABAergic neu-
treatment tailored to patient symptoms and con- rotransmitter system are thought to augment this
dition is recommended over standardized dosing effect, with sodium channel antagonists having
regimens [71]. less of an impact [74]. The mechanisms of their
Despite the relative increased tolerability of influence are not clear, however, as are the rela-
dexamethasone as corticosteroid, there remain tive contributions of medication dose or concur-
significant adverse effects with this agent. On rent medicines on fatigue levels. While generally
evaluating 138 patients retrospectively during thought to have a more benign side effect profile
radiation for primary and metastatic brain cancer, than other anti-seizure medications, levetirace-
the most common adverse effects were elevated tam was found to have a more prominent impact
serum glucose, anxiety, peripheral edema, and on fatigue in one epidemiologic study [75], inde-
Cushing syndrome [71]. A proximal myopathy pendent of its impact on mood. The side effect
may also result from steroid use, contributing profiles of anti-seizure agents are varied, and
to functional disability. These adverse events attention to patient comorbidities and concerns is
increase with prolonged duration of use. Steroid essential in selecting the most appropriate agent
side effects are dose-dependent: in randomizing for the individual.
brain tumor patients to either 4 or 8 mg per day
versus 16 mg per day of dexamethasone, sig-
nificantly more adverse effects were noted in the Rehabilitation
16 mg per day group [72]. While recognizing the
role of steroids in the symptomatic treatment of Rehabilitation in cancer patients possesses
brain metastases patients, minimizing the dose unique challenges. Concerns in regard to frailty
and duration of steroid treatment is integral to and concurrent medical therapies may prevent
minimizing their associated toxicity. full access and use of rehabilitative services [76].
Despite these challenges, rehabilitation is of
demonstrated benefit in patients with CNS can-
Anti-seizure Medicines cer. It leads to tangible functional improvement
in patients with brain [77] and spinal cord tumors
While obtaining control of seizures can contrib- [78]. Functional improvement with rehabilitation
ute greatly to patient quality of life, a breadth has been found to be an independent predictor
of potential toxicities from anti-seizure medi- of overall survival in patients with primary and
cines can occur, and agents are frequently cho- metastatic brain tumors [77], and interventions
sen for the least offensive side effect profile for targeting motor impairment have been found to
a given patient. In patients with cancer, leve- improve both independence and quality of life in
tiracetam and lacosamide are frequently used, brain tumor patients [79].
as they bypass CYP450 metabolism and do not The impact of rehabilitation has been more
interact with other anticancer medicines, and extensively studied in patients with compressive
have a lower incidence of adverse effects than disorders of the spinal cord. In this population,
many other anti-seizure medicines. In particu- patients had sustained improvements in various
lar, anti-seizure medicines may impact cogni- functional measures, mobility, and self-care that
tive function. In a prospective crossover study were maintained 3 months after discharge [80].
comparing levetiracetam and carbamazepine’s Rehabilitation has also been associated with
neuropsychological effects [73], all subjects had improvement in measures of pain, self-care, and
quality of life, as well as reduced depression While there is a lack of studies evaluating
scores [81]. In a population of patients with met- interventions for fatigue in patients with brain
astatic epidural spinal cord compression, rehabil- metastases specifically, pharmacologic and non-
itation had a positive impact on bladder control, pharmacologic therapies have been studied in
with nearly one quarter regaining some control other cancer populations. Psychoeducational
of bladder function with rehabilitation interven- interventions have demonstrated efficacy in
tion [82]. The goal of rehabilitation should be to patients with systemic cancer. Cognitive behav-
improve function, and the role of physical medi- ioral therapy intervention, consisting of eight
cine and rehabilitation in these patients should be weekly structured sessions with a psycho-
advocated. oncologic support group, was found to improve
fatigue in breast cancer patients at its comple-
tion [93]. Energy conservation programs have
Pain Management also demonstrated benefit in reducing fatigue
in patients with systemic cancer [94, 95].
While rare in patients with metastatic brain Exercise holds the strongest evidence of all non-
tumors, neuropathic pain may result from spi- pharmacologic interventions for fatigue, repro-
nothalamic tract involvement of intramedullary ducibly demonstrating benefit in patients with
spine metastases. While there are no evidence- brain tumors and the general cancer population.
based recommendations for treatment, pregaba- Systematic review of studies involving can-
lin has been found to improve central neuropathic cer patients (n = 4881) during or after antican-
pain from spinal cord injury [83–85], with con- cer treatments found that exercise significantly
current positive impacts on sleep and anxiety reduced cancer-related fatigue [96]. It appears to
[83]. Amitriptyline has had variable efficacy in have a palliative effect as patients are undergo-
managing pain from spinal cord injury [86–88], ing cancer treatment and help with restoration of
as has lithium [89]. While opioid medications are energy levels after treatment completion.
not used first line for neuropathic-type pain, they Stimulants are the primary pharmacologic
have been found to be of benefit in central neuro- intervention evaluated for fatigue in cancer
pathic pain [90], including neuropathic pain pre- patients. Methylphenidate, which increases
treated with anti-seizure medicine [91]. This may dopaminergic and noradrenergic transmission,
be of particular use in patients with concurrent has been evaluated with mixed results. While
somatic pain. methylphenidate has been shown to improve
patient-reported fatigue [97, 98], several ran-
domized controlled studies have failed to show
Management of Fatigue significant improvement above placebo [99–
101]. In a systematic review of cancer patients
In the patient with fatigue, a cursory screen of with fatigue (n = 426), the use of stimulants
reversible or treatable contributors is advised. for cancer-related fatigue was supported with
Medications should be optimized, with nones- preliminary evidence [102], and stimulants are
sential medications eliminated and minimum commonly used clinically when pharmacologic
therapeutic doses used. A screen for depression treatment is pursued. Similarly, the evalua-
should occur, with treatment initiated where tion of stimulants in patients with brain tumor
appropriate. Evaluation for nutritional deficien- has yielded mixed results [103–105], and no
cies and metabolic derangements is also war- single agent has been deemed reliably effec-
ranted, particularly in the context of advanced tive. Corticosteroids, frequently used to manage
cancer where end-organ dysfunction is more vasogenic edema in patients with brain metas-
common. A review of sleep patterns is also indi- tases, improve fatigue in cancer patients in the
cated, as management of sleep disorders can short term [106]; however, the effectiveness and
improve cancer-related fatigue [92]. sequelae for long-term use are not known.
Cognitive Interventions At this time, we do not have evidence support-

ing non-pharmacologic interventions for cogni-
Identification of cognitive impairment should tive impairment in patients with brain metastases.
prompt screening for comorbid conditions, such Cognitive rehabilitation with computer-based
as depression and fatigue [107, 108], as they may attention retraining and compensatory skills
be contributing to cognitive deficits. There is a training has been shown to benefit patients
paucity of data evaluating treatments for cogni- with lower-grade glioma [112], with imme-
tive impairment in patients with brain metastases diate improvement in cognitive performance
specifically. Phase 2 pilot data has supported the and sustained improvement after 6 months.
use of donepezil in patients with primary brain Improvements in cognitive performance have
tumors [109, 110], finding multi-domain improve- also been noted in patients with systemic cancer
ment in neurocognitive function and quality of with compensatory and computer-based cogni-
life measures. A phase 3 study in patients with tive training [113, 114]. Memory and adaptation
primary brain tumors or brain metastases that had training (MAAT), a form of cognitive behavioral
received brain radiation found improvements in therapy developed for cancer patients that builds
both social and emotional well-being as well as adaptive skills to manage cognitive demands, has
overall quality of life in those with more cogni- shown benefit in breast cancer patients without
tive impairment treated with donepezil; however, brain metastases in three clinical trials [115–
there was a negative impact on fatigue and func- 117]. It is not clear whether these results can
tional well-being in those with less impairment be applied to the patients with brain metastases,
[111]. Another phase 3 trial in primary brain however, and as such, this population warrants
tumors showed that there was no impact on com- independent study.
posite cognitive scores, the primary outcome, but
did show improvement in memory and dominant
hand function [111]. Once again, patients with End-of-Life Care
greater pre-treatment impairment were found to
derive the greatest benefit, suggesting cholinergic Patients with CNS metastases may die from the
therapy may be of some benefit to select, very progression of the systemic and/or the CNS dis-
symptomatic patients. Definitive evidence sup- ease [118]. Impaired quality of end of life has
porting a pharmacologic intervention in patients been found in cancer patients receiving more
with brain metastases does not yet exist. aggressive anticancer and medical treatments in
Endeavors to prophylax against treatment- the final weeks of life [119]. Despite the poor
associated cognitive impairment during WBRT prognosis of patients with CNS metastases, evi-
have been evaluated. A placebo-controlled trial dence suggests this population is receiving anti-
of memantine during WBRT found beneficial cancer therapy in very late disease stages. In a
effects on memory function at 4 months, the study of patients with non-small cell lung can-
primary outcome, but this result did not reach cer with brain metastases, nearly a quarter of the
statistical significance [18]. Time to neurocogni- over 5000 patients evaluated died within 30 days
tive decline, however, was lower in patients that of CNS-directed treatments [120]. Death within
received memantine compared to placebo. While 30 days of treatment could be reliably predicted
the primary outcomes of these studies were using the graded prognostic assessment (GPA)
negative, it is important to note the small patient system, incorporating patient- and disease-
numbers and high attrition rates in these stud- related information to prognosticate [121] sug-
ies. As such, incorporating use of these agents, gesting timing of radiation referral could be
particularly memantine, into clinical practice has better timed prior to the more imminent end-of-
been variable. Please see the other chapters in life period. Similarly, in patients evaluated from
this book related to neurocognition and radiation time of whole brain radiation therapy to death,
therapy for more details. the median overall survival after radiation was
80 days [122]. Of this cohort, nearly one-third of coping. In the dementia population, it is the neu-
patients presented to the emergency room two or ropsychiatric symptoms of these patients that are
more times within the last 6 months of life, and most relevant to caregiver burnout and depression
only 68% were referred to palliative care, with [126], suggesting particular attention to the needs
57% of these referrals being during inpatient hos- of caregivers of patients that harbor these mani-
pitalizations. Given the heightened awareness of festations is warranted. A study of caregivers
the value of symptom management in patients of persons with Parkinson’s disease, who often
with brain metastases [17], and the positive value harbor cognitive and motor dysfunction in late
early palliative care can have in this regard [123], disease, shows they have unmet palliative care
attention to the prognosis and the utility of medi- needs, and processes to improve caregiver abil-
cal interventions as patients approach end of life ity to manage the neurologic disability have been
is of central importance. suggested [127]. While we lack understanding of
caregiver needs in patients with brain metastases,
the study of caregivers of other CNS disorders
Caregiver Needs supports this group facing unique challenges and
suggests they may have distinct needs from care-
Caregivers often play a central role in the well- givers of other cancer patients.
being of patients with advanced cancer. The
unique symptom burdens in patients with CNS
metastases imply a distinct role for their caregiv- Conclusions
ers. At this time, there is a paucity of evidence
characterizing the needs and challenges of care- Cancer patients with metastases to the brain and
givers of patients with CNS metastases, and as spine have distinct symptom burdens and disease
such, information can only be derived from the trajectories. Concurrent with the poor prognosis
study of patients with primary brain tumors and imparted by CNS involvement, these patients
other neurologic disease. From the study of care- often suffer from disabling symptom burdens
givers of patients with brain tumors, we know that can influence fundamental motor, sensory,
that they frequently feel untrained, uncompen- and cognitive abilities. Clinicians and family
sated, and unprepared for their role [124]. They members must be aware of the potential influence
find it difficult to adjust to their role at illness these changes may have on patient autonomy and
onset, as well as its increasing demands over the decision-making. The importance of these issues
illness trajectory [124]. This difficulty is likely is emphasized by the incorporation of patient-
compounded by the rapid onset and progression reported and cognitive outcomes as endpoints
of CNS metastases. Distinct from patients with in clinical trials for brain metastases. Persistent
primary brain tumors, most caregivers of those clinical and academic attention to the palliative
with CNS metastases have already been caring care needs of patients with CNS metastases will
for the patient at time of diagnosis of CNS dis- be central to alleviating the suffering imparted by
ease and, as such, may already have fatigue and the devastating complication of cancer.
frustration at onset.
Cognitive dysfunction may have a specific
influence on caregiver resilience and coping. In a References
descriptive cross-sectional study evaluating com-
1. Wen PY, Loeffler JS. Management of brain metasta-
mon coping strategies of caregivers for patients ses. Oncology (Williston Park, NY). 1999;13:941–54,
with cancer, common and effective strategies 957–961; discussion 961–942, 949
included acceptance, planning, positive interpre- 2. Johnson JD, Young B. Demographics of brain metas-
tasis. Neurosurg Clin N Am. 1996;7:337–44.
tation, and growth [125]. In patients with cogni- 3. Gavrilovic IT, Posner JB. Brain metastases: epi-
tive impairment, however, caregivers were more demiology and pathophysiology. J Neuro-Oncol.
likely to use less healthy or effective methods of 2005;75:5–14.
4. Niwinska A, Tacikowska M, Murawska M. The apy: a randomized, double-blind, placebo-controlled

effect of early detection of occult brain metastases trial. Neuro Oncol. 2013;15:1429–37.
in HER2-positive breast cancer patients on survival 19. Lien K, Zeng L, Nguyen J, Cramarossa G, Cella
and cause of death. Int J Radiat Oncol Biol Phys. D, et al. FACT-Br for assessment of quality of life
2010;77:1134–9. in patients receiving treatment for brain metasta-
5. Brufsky AM, Mayer M, Rugo HS, Kaufman PA, ses: a literature review. Expert Rev Pharmacoecon
Tan-Chiu E, et al. Central nervous system metas- Outcomes Res. 2011;11:701–8.
tases in patients with HER2-positive metastatic 20. Pham A, Lo SS, Sahgal A, Chang EL. Neurocognition
breast cancer: incidence, treatment, and survival in and quality-of-life in brain metastasis patients who
patients from registHER. Clin Cancer Res. 2011;17: have been irradiated focally or comprehensively.
4834–43. Expert Rev Qual Life Cancer Care. 2016;1:45–60.
6. Chason JL, Walker FB, Landers JW. Metastatic 21. Tom MI. Metastatic tumours of brain. Can Med
carcinoma in the central nervous system and dorsal Assoc J. 1946;54:265–8.
root ganglia. A prospective autopsy study. Cancer. 22. Payer S, Mende KC, Westphal M, Eicker
1963;16:781–7. SO. Intramedullary spinal cord metastases: an
7. Sung WS, Sung MJ, Chan JH, Manion B, Song J, increasingly common diagnosis. Neurosurg Focus.
et al. Intramedullary spinal cord metastases: a 20-year 2015;39:E15.
institutional experience with a comprehensive litera- 23. Schiff D, O’Neill BP. Intramedullary spinal cord
ture review. World Neurosurg. 2013;79:576–84. metastases: clinical features and treatment outcome.
8. Barron KD, Hirano A, Araki S, Terry Neurology. 1996;47:906–12.
RD. Experiences with metastatic neoplasms involv- 24. Nieder C, Oehlke O, Hintz M, Grosu AL. The chal-
ing the spinal cord. Neurology. 1959;9:91. lenge of durable brain control in patients with brain-
9. Bruera E, Hui D. Conceptual models for integrat- only metastases from breast cancer. Springerplus.
ing palliative care at cancer centers. J Palliat Med. 2015;4:585.
2012;15:1261–9. 25. Sander Connolly E, Winfree CJ, McCormick PC,
10. Hui D, Bruera E. Integrating palliative care into Cruz M, Stein BM. Intramedullary spinal cord
the trajectory of cancer care. Nat Rev Clin Oncol. metastasis: report of three cases and review of the
2015;13:159. literature. Surg Neurol. 1996;46:329–37.
11. Fadul N, Elsayem A, Palmer JL, Del Fabbro E, Swint 26. Potti A, Abdel-Raheem M, Levitt R, Schell DA,
K, et al. Supportive versus palliative care: what's in a Mehdi SA. Intramedullary spinal cord metasta-
name?: a survey of medical oncologists and midlevel ses (ISCM) and non-small cell lung carcinoma
providers at a comprehensive cancer center. Cancer. (NSCLC): clinical patterns, diagnosis and thera-
2009;115:2013–21. peutic considerations. Lung Cancer (Amsterdam,
12. Wong A, Hui D, Epner M, Balankari VR, Cruz VJD, Netherlands). 2001;31:319–23.
et al. Advanced cancer patients’ self-reported per- 27. Kalayci M, Cagavi F, Gul S, Yenidunya S, Acikgoz
ception of timeliness of their referral to outpatient B. Intramedullary spinal cord metastases: diagnosis
supportive/palliative care and their survival data. J and treatment – an illustrated review. Acta Neurochir.
Clin Oncol. 2017;35:–10121. 2004;146:1347–54; discussion 1354
13. Dalal S, Palla S, Hui D, Nguyen L, Chacko R, et al. 28. Grem JL, Burgess J, Trump DL. Clinical features
Association between a name change from palliative and natural history of intramedullary spinal cord
to supportive care and the timing of patient refer- metastasis. Cancer. 1985;56:2305–14.
rals at a comprehensive cancer center. Oncologist. 29. Patel RA, Chandler JP, Jain S, Gopalakrishnan M,
2011;16:105–11. Sachdev S. Dejerine-Roussy syndrome from tha-
14. Dalal S, Bruera S, Hui D, Yennu S, Dev R, et al. lamic metastasis treated with stereotactic radiosur-
Use of palliative care services in a Tertiary Cancer gery. J Clin Neurosci. 2017;44:227–8.
Center. Oncologist. 2016;21:110–8. 30. Amidei C, Kushner DS. Clinical implications of
15. Oliver A, Greenberg CC. Measuring out- motor deficits related to brain tumors. Neuro Oncol
comes in oncology treatment: the importance of Pract. 2015;2:179–84.
patient-centered outcomes. Surg Clin North Am. 31. Osoba D, Brada M, Prados MD, Yung WK. Effect
2009;89:17–vii. of disease burden on health-related quality of life
16. Bottomley A, Aaronson NK. International perspec- in patients with malignant gliomas. Neuro Oncol.
tive on health-related quality-of-life research in 2000;2:221–8.
cancer clinical trials: the European Organisation 32. Rykken JB, Diehn FE, Hunt CH, Schwartz KM,
for Research and Treatment of Cancer experience. J Eckel LJ, et al. Intramedullary spinal cord metas-
Clin Oncol. 2007;25:5082–6. tases: MRI and relevant clinical features from
17. Tsao MN. Brain metastases: advances over the a 13-year institutional case series. AJNR Am J
decades. Ann Palliat Med. 2015;4:225–32. Neuroradiol. 2013;34:2043–9.
18. Brown PD, Pugh S, Laack NN, Wefel JS, Khuntia D, 33. Forsyth PA, Posner JB. Headaches in patients with
et al. Memantine for the prevention of cognitive dys- brain tumors: a study of 111 patients. Neurology.
function in patients receiving whole-brain radiother- 1993;43:1678–83.
34. Valentinis L, Tuniz F, Valent F, Mucchiut M, 49. Meyers CA, Smith JA, Bezjak A, Mehta MP,
Little D, et al. Headache attributed to intracranial Liebmann J, et al. Neurocognitive function and pro-
tumours: a prospective cohort study. Cephalalgia. gression in patients with brain metastases treated
2010;30:389–98. with whole-brain radiation and motexafin gadolin-
35. Pulenzas N, Khan L, Tsao M, Zhang L, Lechner B, ium: results of a randomized phase III trial. J Clin
et al. Fatigue scores in patients with brain metastases Oncol. 2004;22:157–65.
receiving whole brain radiotherapy. Support Care 50. Oberndorfer S, Schmal T, Lahrmann H, Urbanits S,
Cancer. 2014;22:1757–63. Lindner K, Grisold W. The frequency of seizures in
36. Hofman M, Morrow GR, Roscoe JA, Hickok JT, patients with primary brain tumors or cerebral metas-
Mustian KM, et al. Cancer patients’ expectations tases. An evaluation from the Ludwig Boltzmann
of experiencing treatment-related side effects: a Institute of Neuro-Oncology and the Department
University of Rochester Cancer Center–Community of Neurology, Kaiser Franz Josef Hospital, Vienna.
Clinical Oncology Program study of 938 patients Wien Klin Wochenschr. 2002;114:911–6.
from community practices. Cancer. 2004;101:851–7. 51. Karakis I, Cole AJ, Montouris GD, San Luciano
37. Glaus A, Crow R, Hammond S. A qualitative study M, Meador KJ, Piperidou C. Caregiver burden in
to explore the concept of fatigue/tiredness in cancer epilepsy: determinants and impact. Epilepsy Res

patients and in healthy individuals. Support Care Treat. 2014;2014:–808421.
Cancer. 1996;4:82–96. 52. Al-Zabin M, Ullrich WO, Brawanski A, Proescholdt
38. Morrow GR, Andrews PL, Hickok JT, Roscoe JA, MA. Recurrent brain metastases from lung can-
Matteson S. Fatigue associated with cancer and its cer: the impact of reoperation. Acta Neurochir.
treatment. Support Care Cancer. 2002;10:389–98. 2010;152:1887–92.
39. Ryan JL, Carroll JK, Ryan EP, Mustian KM, Fiscella 53. Conill C, Marruecos J, Verger E, Berenguer J,
K, Morrow GR. Mechanisms of cancer-related Lomena F, et al. Clinical outcome in patients with
fatigue. Oncologist. 2007;12(Suppl 1):22–34. intramedullary spinal cord metastases from lung
40. Bower JE, Ganz PA, Aziz N, Fahey JL. Fatigue and cancer. Clin Transl Oncol. 2007;9:172–6.
proinflammatory cytokine activity in breast cancer 54. Gremmer R, Schroder ML, Ten Huinink WW,
survivors. Psychosom Med. 2002;64:604–11. Brandsma D, Boogerd W. Successful manage-
41. Bower JE, Ganz PA, Aziz N. Altered cortisol ment of brain metastasis from malignant germ cell
response to psychologic stress in breast cancer tumours with standard induction chemotherapy. J
survivors with persistent fatigue. Psychosom Med. Neuro-Oncol. 2008;90:335–9.
2005;67:277–80. 55. van den Bent MJ. The role of chemotherapy in brain
42. Bower JE, Ganz PA, Dickerson SS, Petersen L, Aziz metastases. Eur J Cancer (Oxford, England: 1990).
N, Fahey JL. Diurnal cortisol rhythm and fatigue in 2003;39:2114–20.
breast cancer survivors. Psychoneuroendocrinology. 56. Gaspar L, Scott C, Rotman M, Asbell S, Phillips T,
2005;30:92–100. et al. Recursive partitioning analysis (RPA) of prog-
43. Mormont MC, Levi F. Circadian-system alterations nostic factors in three Radiation Therapy Oncology
during cancer processes: a review. Int J Cancer. Group (RTOG) brain metastases trials. Int J Radiat
1997;70:241–7. Oncol Biol Phys. 1997;37:745–51.
44. Miaskowski C, Lee K, Dunn L, Dodd M, Aouizerat 57. Bezjak A, Adam J, Panzarella T, Levin W, Barton
BE, et al. Sleep-wake circadian activity rhythm R, et al. Radiotherapy for brain metastases: defin-
parameters and fatigue in oncology patients before ing palliative response. Radiother Oncol. 2001;61:
the initiation of radiation therapy. Cancer Nurs. 71–6.
2011;34:255–68. 58. Cross NE, Glantz MJ. Neurologic complications of
45. Armstrong TS, Gilbert MR. Practical strategies for radiation therapy. Neurol Clin. 2003;21:249–77.
management of fatigue and sleep disorders in people 59. Sheline GE. Radiation therapy of brain tumors.
with brain tumors. Neuro Oncol. 2012;14 Suppl Cancer. 1977;39:873–81.
4:iv65–iv72. 60. Chang EL, Wefel JS, Hess KR, Allen PK, Lang FF,
46. Wefel JS, Lenzi R, Theriault RL, Davis RN, Meyers et al. Neurocognition in patients with brain metas-
CA. The cognitive sequelae of standard-dose adju- tases treated with radiosurgery or radiosurgery plus
vant chemotherapy in women with breast carcinoma: whole-brain irradiation: a randomised controlled
results of a prospective, randomized, longitudinal trial. Lancet Oncol. 2009;10:1037–44.
trial. Cancer. 2004;100:2292–9. 61. Sun A, Bae K, Gore EM, Movsas B, Wong SJ, et al.
47. Tucha O, Smely C, Preier M, Lange KW. Cognitive Phase III trial of prophylactic cranial irradiation
deficits before treatment among patients with brain compared with observation in patients with locally
tumors. Neurosurgery. 2000;47:324–33; discussion advanced non-small-cell lung cancer: neurocog-
333–324 nitive and quality-of-life analysis. J Clin Oncol.
48. Gerstenecker A, Nabors LB, Meneses K, Fiveash JB, 2011;29:279–86.
Marson DC, et al. Cognition in patients with newly 62. Brown PD, Jaeckle K, Ballman KV, Farace E, Cerhan
diagnosed brain metastasis: profiles and implica- JH, et al. Effect of radiosurgery alone vs radiosur-
tions. J Neuro-Oncol. 2014;120:179–85. gery with whole brain radiation therapy on cognitive
function in patients with 1 to 3 brain metastases: a ologic effects of carbamazepine and levetiracetam.
randomized clinical trial. JAMA. 2016;316:401–9. Neurology. 2007;69:2076–84.
63. Soffietti R, Kocher M, Abacioglu UM, Villa S, 74. Siniscalchi A, Gallelli L, Russo E, De Sarro G. A
Fauchon F, et al. A European Organisation for review on antiepileptic drugs-dependent fatigue:
Research and Treatment of Cancer phase III trial pathophysiological mechanisms and incidence. Eur
of adjuvant whole-brain radiotherapy versus obser- J Pharmacol. 2013;718:10–6.
vation in patients with one to three brain metasta- 75. Mula M, von Oertzen TJ, Cock HR, Yogarajah M,
ses from solid tumors after surgical resection or Lozsadi DA, Agrawal N. Fatigue during treatment
radiosurgery: quality-of-life results. J Clin Oncol. with antiepileptic drugs: a levetiracetam-specific
2013;31:65–72. adverse event? Epilepsy Behav. 2017;72:17–21.
64. Gondi V, Paulus R, Bruner DW, Meyers CA, Gore 76. Palacio A, Calmels P, Genty M, Le-Quang B,
EM, et al. Decline in tested and self-reported cog- Beuret-Blanquart F. Oncology and physical medi-
nitive functioning after prophylactic cranial irradia- cine and rehabilitation. Ann Phys Rehabil Med.
tion for lung cancer: pooled secondary analysis of 2009;52:568–78.
Radiation Therapy Oncology Group randomized 77. Tang V, Rathbone M, Park Dorsay J, Jiang S, Harvey
trials 0212 and 0214. Int J Radiat Oncol Biol Phys. D. Rehabilitation in primary and metastatic brain
2013;86:656–64. tumours: impact of functional outcomes on survival.
65. Brown PD, Asher AL, Ballman KV, Farace E, J Neurol. 2008;255:820–7.
Cerhan JH, et al. NCCTG N0574 (Alliance): a phase 78. Raj VS, Lofton L. Rehabilitation and treat-
III randomized trial of whole brain radiation ther- ment of spinal cord tumors. J Spinal Cord Med.
apy (WBRT) in addition to radiosurgery (SRS) in 2013;36:4–11.
patients with 1 to 3 brain metastases. J Clin Oncol. 79. Kushner DS, Amidei C. Rehabilitation of motor dys-
2015;33:LBA4. function in primary brain tumor patients†. Neuro-
66. Ryken TC, McDermott M, Robinson PD, Ammirati Oncol Pract. 2015;2:185–91.
M, Andrews DW, et al. The role of steroids in the 80. McKinley WO, Conti-Wyneken AR, Vokac CW,
management of brain metastases: a systematic Cifu DX. Rehabilitative functional outcome of
review and evidence-based clinical practice guide- patients with neoplastic spinal cord compressions.
line. J Neuro-Oncol. 2010;96:103–14. Arch Phys Med Rehabil. 1996;77:892–5.
67. Markman M, Sheidler V, Ettinger DS, Quaskey SA, 81. Ruff RL, Ruff SS, Wang X. Persistent benefits of
Mellits ED. Antiemetic efficacy of dexamethasone. rehabilitation on pain and life quality for nonam-
Randomized, double-blind, crossover study with bulatory patients with spinal epidural metastasis. J
prochlorperazine in patients receiving cancer che- Rehabil Res Dev. 2007;44:271–8.
motherapy. N Engl J Med. 1984;311:549–52. 82. Fattal C, Fabbro M, Rouays-Mabit H, Verollet C,
68. Lagerwaard FJ, Levendag PC, Nowak PJ, Bauchet L. Metastatic paraplegia and functional
Eijkenboom WM, Hanssens PE, Schmitz outcomes: perspectives and limitations for reha-
PI. Identification of prognostic factors in patients bilitation care. Part 2. Arch Phys Med Rehabil.
with brain metastases: a review of 1292 patients. Int 2011;92:134–45.
J Radiat Oncol Biol Phys. 1999;43:795–803. 83. Siddall PJ, Cousins MJ, Otte A, Griesing T, Chambers
69. Borgelt B, Gelber R, Kramer S, Brady LW, Chang R, Murphy TK. Pregabalin in central neuropathic
CH, et al. The palliation of brain metastases: final pain associated with spinal cord injury: a placebo-
results of the first two studies by the Radiation controlled trial. Neurology. 2006;67:1792–800.
Therapy Oncology Group. Int J Radiat Oncol Biol 84. Vranken JH, Dijkgraaf MG, Kruis MR, van der
Phys. 1980;6:1–9. Vegt MH, Hollmann MW, Heesen M. Pregabalin in
70. Gaspar LE, Gutin PH, Rogers L, Schneider JF, patients with central neuropathic pain: a random-
Larson D, et al. Pre-irradiation evaluation and man- ized, double-blind, placebo-controlled trial of a
agement of brain metastases. American College of flexible-dose regimen. Pain. 2008;136:150–7.
Radiology. ACR appropriateness criteria. Radiology. 85. Cardenas DD, Nieshoff EC, Suda K, Goto S, Sanin
2000;215 Suppl:1105–10. L, et al. A randomized trial of pregabalin in patients
71. Hempen C, Weiss E, Hess CF. Dexamethasone with neuropathic pain due to spinal cord injury.
treatment in patients with brain metastases and Neurology. 2013;80:533–9.
primary brain tumors: do the benefits outweigh 86. Cardenas DD, Warms CA, Turner JA, Marshall
the side-effects? Support Care Cancer. 2002;10: H, Brooke MM, Loeser JD. Efficacy of amitripty-
322–8. line for relief of pain in spinal cord injury: results
72. Vecht CJ, Hovestadt A, Verbiest HB, van Vliet JJ, of a randomized controlled trial. Pain. 2002;96:
van Putten WL. Dose-effect relationship of dexa- 365–73.
methasone on Karnofsky performance in metastatic 87. Rintala DH, Holmes SA, Courtade D, Fiess RN,
brain tumors: a randomized study of doses of 4, 8, Tastard LV, Loubser PG. Comparison of the effec-
and 16 mg per day. Neurology. 1994;44:675–80. tiveness of amitriptyline and gabapentin on chronic
73. Meador KJ, Gevins A, Loring DW, McEvoy LK, neuropathic pain in persons with spinal cord injury.
Ray PG, et al. Neuropsychological and neurophysi- Arch Phys Med Rehabil. 2007;88:1547–60.
88. Ahn SH, Park HW, Lee BS, Moon HW, Jang SH, placebo- controlled, phase II trial. J Clin Oncol.
et al. Gabapentin effect on neuropathic pain com- 2013;31:2421–7.
pared among patients with spinal cord injury and dif- 101. Moraska AR, Sood A, Dakhil SR, Sloan JA, Barton
ferent durations of symptoms. Spine. 2003;28:341–6; D, et al. Phase III, randomized, double-blind,
discussion 346–347 placebo- controlled study of long-acting methyl-
89. Yang ML, Li JJ, So KF, Chen JY, Cheng WS, et al. phenidate for cancer-related fatigue: North Central
Efficacy and safety of lithium carbonate treatment of Cancer Treatment Group NCCTG-N05C7 trial. J
chronic spinal cord injuries: a double-blind, random- Clin Oncol. 2010;28:3673–9.
ized, placebo-controlled clinical trial. Spinal Cord. 102. Minton O, Richardson A, Sharpe M, Hotopf M,
2012;50:141–6. Stone PC. Psychostimulants for the management of
90. Norrbrink C, Lundeberg T. Tramadol in neuro- cancer-related fatigue: a systematic review and meta-
pathic pain after spinal cord injury: a randomized, analysis. J Pain Symptom Manag. 2011;41:761–7.
double-blind, placebo-controlled trial. Clin J Pain. 103. Butler JM Jr, Case LD, Atkins J, Frizzell B,
2009;25:177–84. Sanders G, et al. A phase III, double-blind, placebo-
91. Barrera-Chacon JM, Mendez-Suarez JL, Jauregui- controlled prospective randomized clinical trial of
Abrisqueta ML, Palazon R, Barbara-Bataller E, d-threo-methylphenidate HCl in brain tumor patients
Garcia-Obrero I. Oxycodone improves pain control receiving radiation therapy. Int J Radiat Oncol Biol
and quality of life in anticonvulsant-pretreated spinal Phys. 2007;69:1496–501.
cord-injured patients with neuropathic pain. Spinal 104. Gehring K, Patwardhan SY, Collins R, Groves MD,
Cord. 2011;49:36–42. Etzel CJ, et al. A randomized trial on the efficacy of
92. Zee PC, Ancoli-Israel S. Does effective management methylphenidate and modafinil for improving cogni-
of sleep disorders reduce cancer-related fatigue? tive functioning and symptoms in patients with a pri-
Drugs. 2009;69(Suppl 2):29–41. mary brain tumor. J Neuro-Oncol. 2012;107:165–74.
93. Eichler C, Pia M, Sibylle M, Sauerwald A, Friedrich 105. Meyers CA, Weitzner MA, Valentine AD, Levin
W, Warm M. Cognitive behavioral therapy in breast VA. Methylphenidate therapy improves cognition,
cancer patients–a feasibility study of an 8 week mood, and function of brain tumor patients. J Clin
intervention for tumor associated fatigue treatment. Oncol Off J Am Soc Clin Oncol. 1998;16:2522–7.
Asian Pac J Cancer Prev. 2015;16:1063–7. 106. Yennurajalingam S, Bruera E. Role of corticoste-
94. Sadeghi E, Gozali N, Moghaddam TF. Effects of roids for fatigue in advanced incurable cancer: is it
energy conservation strategies on cancer related a ‘wonder drug’ or ‘deal with the devil’. Curr Opin
fatigue and health promotion lifestyle in breast can- Support Palliat Care. 2014;8:346–51.
cer survivors: a Randomized Control Trial. Asian 107. Kinsinger SW, Lattie E, Mohr DC. Relationship
Pac J Cancer Prev. 2016;17:4783–90. between depression, fatigue, subjective cogni-
95. Barsevick AM, Dudley W, Beck S, Sweeney C, tive impairment, and objective neuropsychologi-
Whitmer K, Nail L. A randomized clinical trial of cal functioning in patients with multiple sclerosis.
energy conservation for patients with cancer-related Neuropsychology. 2010;24:573–80.
fatigue. Cancer. 2004;100:1302–10. 108. Pendergrass JC, Targum SD, Harrison JE. Cognitive
96. Puetz TW, Herring MP. Differential effects of exer- impairment associated with cancer: a brief review.
cise on cancer-related fatigue during and follow- Innov Clin Neurosci. 2018;15:36–44.
ing treatment: a meta-analysis. Am J Prev Med. 109. Correa DD, Kryza-Lacombe M, Baser RE, Beal
2012;43:e1–24. K, DeAngelis LM. Cognitive effects of donepezil
97. Bruera E, Driver L, Barnes EA, Willey J, Shen L, therapy in patients with brain tumors: a pilot study. J
et al. Patient-controlled methylphenidate for the Neuro-Oncol. 2016;127:313–9.
management of fatigue in patients with advanced 110. Shaw EG, Rosdhal R, D'Agostino RB Jr, Lovato J,
cancer: a preliminary report. J Clin Oncol. Naughton MJ, et al. Phase II study of donepezil in
2003;21:4439–43. irradiated brain tumor patients: effect on cognitive
98. Kerr CW, Drake J, Milch RA, Brazeau DA, Skretny function, mood, and quality of life. J Clin Oncol.
JA, et al. Effects of methylphenidate on fatigue 2006;24:1415–20.
and depression: a randomized, double-blind, 111. Rapp SR, Case LD, Peiffer A, Naughton MM, Chan
placebo-controlled trial. J Pain Symptom Manag. MD, et al. Donepezil for irradiated brain tumor sur-
2012;43:68–77. vivors: a phase III randomized placebo-controlled
99. Bruera E, Valero V, Driver L, Shen L, Willey clinical trial. J Clin. 2015;33:1653–9.
J, et al. Patient-controlled methylphenidate for 112. Gehring K, Sitskoorn MM, Gundy CM, Sikkes SA,
cancer fatigue: a double-blind, randomized, Klein M, et al. Cognitive rehabilitation in patients
placebo-controlled trial. J Clin Oncol. 2006;24: with gliomas: a randomized, controlled trial. J Clin
2073–8. Oncol. 2009;27:3712–22.

100. Bruera E, Yennurajalingam S, Palmer JL, Perez- 113. Park J-H, Jung YS, Kim KS, Bae SH. Effects of
Cruz PE, Frisbee-Hume S, et al. Methylphenidate compensatory cognitive training intervention for
and/or a nursing telephone intervention for fatigue breast cancer patients undergoing chemotherapy: a
in patients with advanced cancer: a randomized, pilot study. Support Care Cancer. 2017;25:1887–96.
114. Bail J, Meneses K. Computer-based cognitive train- 120. Ryoo JJ, Batech M, Zheng C, Kim RW, Gould MK,
ing for chemotherapy-related cognitive impair- et al. Radiotherapy for brain metastases near the
ment in breast cancer survivors. Clin J Oncol Nurs. end of life in an integrated health care system. Ann
2016;20:504–9. Palliat Med. 2017;6:S28–s38.
115. Ferguson RJ, Sigmon ST, Pritchard AJ, LaBrie 121. Sperduto PW, Kased N, Roberge D, Xu Z, Shanley
SL, Goetze RE, et al. A randomized trial of R, et al. Summary report on the graded prognostic
videoconference- delivered cognitive behavioral assessment: an accurate and facile diagnosis-specific
therapy for survivors of breast cancer with self- tool to estimate survival for patients with brain
reported cognitive dysfunction. Cancer. 2016;122: metastases. J Clin Oncol. 2012;30:419–25.
1782–91. 122. Stavas M, Arneson K, Friedman J, Misra S. From
116. Ferguson RJ, McDonald BC, Rocque MA, whole brain to hospice: patterns of care in radiation
Furstenberg CT, Horrigan S, et al. Development of oncology. J Palliat Med. 2014;17:662–6.
CBT for chemotherapy-related cognitive change: 123. Temel JS, Greer JA, Muzikansky A, Gallagher ER,
results of a waitlist control trial. Psycho-Oncology. Admane S, et al. Early palliative care for patients
2012;21:176–86. with metastatic non-small-cell lung cancer. N Engl
117. Ferguson RJ, Ahles TA, Saykin AJ, McDonald BC, J Med. 2010;363:733–42.
Furstenberg CT, et al. Cognitive-behavioral man- 124. Schubart JR, Kinzie MB, Farace E. Caring for the
agement of chemotherapy-related cognitive change. brain tumor patient: family caregiver burden and
Psycho-Oncology. 2007;16:772–7. unmet needs. Neuro Oncol. 2008;10:61–72.
118. Pesce GA, Klingbiel D, Ribi K, Zouhair A, von 125. Saria MG, Courchesne N, Evangelista L, Carter
Moos R, et al. Outcome, quality of life and cogni- J, MacManus DA, et al. Cognitive dysfunction in
tive function of patients with brain metastases from patients with brain metastases: influences on care-
non-small cell lung cancer treated with whole brain giver resilience and coping. Support Care Cancer.
radiotherapy combined with gefitinib or temozolo- 2017;25:1247–56.
mide. A randomised phase II trial of the Swiss Group 126. Cheng ST. Dementia caregiver burden: a research
for Clinical Cancer Research (SAKK 70/03). Eur J update and critical analysis. Curr Psychiatry Rep.
Cancer (Oxford, England: 1990). 2012;48:377–84. 2017;19:64.
119. Hui D, Kim SH, Roquemore J, Dev R, Chisholm G, 127. Goy ER, Carter J, Ganzini L. Neurologic disease at
Bruera E. Impact of timing and setting of palliative the end of life: caregiver descriptions of Parkinson
care referral on quality of end-of-life care in cancer disease and amyotrophic lateral sclerosis. J Palliat
patients. Cancer. 2014;120:1743–9. Med. 2008;11:548–54.
Index
A Atezolizumab, 211
Aantiepileptic drug class (AEDs), 156 Attention-deficit hyperactivity disorder, 152
Abemaciclib, 223 Atypical hemangiomas, 527
Abiraterone acetate (AA), 516
Accelerated repopulation, 273
Acupuncture, 697, 698 B
Acute encephalopathy, 416 Basic Score for Brain Metastases (BSBM), 84
Acute lymphoblastic leukemia (ALL), 515 Bateson’s plexus, 170
Adrenocorticotropic hormone (ACTH), 202 Bauer modified score, 497
Adverse radiation effect (ARE), 393 Behavioral tests, 40
Aerobic exercise, 699 Bevacizumab, 144, 221, 402, 478
Alectinib, 210 Bilsky epidural disease grading system, 537
Alliance trial N0574, 285 Biologic pain, 536
Allodynia, 675, 683 Biologically effective dose (BED), 330, 331
αβ T cells, 355 Biomechanical model
Amino acid tracers, 103 Denis’ model, 585
Anaplastic lymphoma kinase (ALK), 8, 72, 210, 211 focal kyphosis, 583, 584
gene mutations, 29 Biomechanics
Androgen-directed therapies (ADT), 516 cervical fixation
Anterior column reconstruction, 545 cervical spine fixation, 547, 548, 550
Anti-angiogenic therapies, 517 cervicothoracic junction, 550
Anti-CRMP5 antibodies, 201 occipitocervical fixation, 547
Anti-CTLA-4 antibody, 211 surgical stabilization
Antiepileptic drugs (AEDs), 178 aggressive surgical decompression, 545
Anti-gamma-butyric acid receptors, 201 anterior column reconstruction, 545
Antigen-presenting cells (APCs), 356 anterior decompression, 545
activation, of T cells, 356, 357 interbody grafts, 545–546
Anti-NMDAR encephalitis, 201 pedicle screws, 546, 547
Antiresorptive therapy, 505, 507, 509 vertebral body resection, 545
Antiseizure drug therapy (AEDs), 129, 710 thoracolumbar fixation
gabapentin, 131 anterior fixation, 550–551
levetiracetam, 129 posterior fixation, 551
oxcarbazepine and carbamazepine, 131 Bipolar disorders, 154, 155
phenytoin, 130 Bisphosphonates
topiramate, 131 bone affinity and antiresorptive potency, 506
zonisamide, 131 breast cancer, 507
Anxiety disorders, 155 metastatic prostate cancer, 507
Apparent diffusion coefficient (ADC), 99, 398 multiple myeloma, 507, 508
Aromatase inhibitor therapy, 509 pharmacokinetic properties, 506
Aromatase inhibitors (AIs), 515, 516 safety, 508
Arterial spin labeling (ASL), 96 Bladder cancer, 251
ASCO TAPUR (Targeted Agent and Profiling Utilization Bleomycin, 478
Registry (TAPUR) Study), 228 Blood brain barrier (BBB), 25, 26, 37, 38, 70, 208, 357, 399

https://doi.org/10.1007/978-3-030-42958-4
720 Index
Blood oxygenation level dependent (BOLD) QUARTZ trial, 87

imaging, 434 tumor subtype, 90
Blood-tumor barrier (BTB), 26, 40 clinical presentation
Bone metastases cognitive impairment, 119
bisphosphonates focal neurologic deficits, 118, 119
bone affinity and antiresorptive potency, 506 headaches, 119
breast cancer, 507 LMD, 120
metastatic prostate cancer, 507 pituitary metastases, 120
multiple myeloma, 507, 508 seizures, 119, 120
pharmacokinetic properties, 506 CNS proliferation, 70
safety, 508 CSCs, 22, 23
bone formation and resorption, 503 DWI, 95, 99–101
denosumab, 508, 509 EMT/MET, 21–22
optimization bone health, 509 epithelioid glioblastoma, 57
pathophysiology, 504 extracranial disease receive systemic therapy, 69
prevalence, 504 FACT-BR score, 91
treatment approach, 504, 505 FACT-Brain questionnaire, 91
Bone morphogenetic proteins (BMPs), 546 final metastatic phase
Bone quality, 547 BBB, 25, 26
Bone remodeling process, 503 brain metastasis microenvironment, 27
Bone-forming osteoblasts, 503 BTB, 26
Bone-resorbing osteoclasts, 503 inflammatory and immunosuppressive
Bone-targeted therapies, 517, 518 microenvironment, 26, 27
Brachytherapy, 659 fine needle aspiration, 90
adjuvant radiotherapy modality genetic heterogeneity, 70–71
practical considerations, 443 genomic profiling, 71, 72
radiobiological considerations, 443, 444 GPA Worksheet, 86–87
resection cavity volume, 442, 443 Graded Prognostic Assessment Score, 86
timing, 443 H&E-stained slides, 58
Cs-131 brachytherapy immunohistochemical stains, 58, 59, 61
I-125 versus, 452, 453 immunohistochemical workup, 58, 60
outcomes in brain metastases, 450–452 interferon and vinblastine, 90
I-125 brachytherapy intracarotid/intracranial approaches, 37
high-activity I-125 brachytherapy, 448 intracranial and extracranial tumor, 70
limitations and complications, 449 intraoperative consultation, 53, 54, 56, 57
permanent implant, 448 intraoperative squash preparation, 55
permanent low activity I-125 brachytherapy, intraoperativemanagement
448, 449 DTI, 434, 435
temporary implant, 447 electrocorticography, 435
perioperative brachytherapy, 444, 445, 447 fMRI, 434
BRAF inhibitors, 236, 237, 240 stereotactic navigation, 433, 434
BRAF V600 mutations, 236 TMS, 434, 435
Bragg peak, 277, 307 limitations, 418, 419
Brain, immunity in, 357, 358 “liquid” phase of
Brain metastases (BMs), 315, 371 CTCs, 23, 24
biomarkers in, 65 ctDNA, 25
brain metastatic cascade, steps of, 20 DCCs, 23, 24
BrM cells, 39 exosomes, 24, 25
Clark’s Level IV, 90 management of, 429, 430
classification systems preoperative risk assessment, 431, 432
age, 90 pre-surgical evaluation, 430, 431
controlled primary tumor, 83 prognosis, 430, 431
diagnosis-specific GPA indices, 87 resection method, 436
extracranial metastases, 83, 87 surgery vs. radiation, 432, 433
GI-GPA, 90 management of seizures (see Seizures
KPS, 83 management)
median survival time, 85 melanoma-molGPA, 91
melanoma-molGPA, 84 mice modelling
nomogram, 84 CRISPR/Cas9 technology, 47
non-lung malignancies, 90 PDXs, 46, 47
Index 721
spontaneous models of, 46 Braking radiation, 276

syngeneic mouse cell, 46 Breast cancer, 10, 11
mitigate neurotoxicity, strategies to, 417, 418 actionable mutations, 76
molecular / cellular MRI, 105 HER2 antibodies and TKIs, 75, 76
molecular assessment, 62, 63, 65 TNBC, 75
molecular heterogeneity and selection, 28 Breast cancer brain metastases (BCBM)
molecular workup, 64 complications, 228
MR perfusion, 95 HER2-positive therapy
ASL, 96 capecitabine and neratinib, 225
DSC-MR, 96 lapatinib, 225
post-treatment period, 98 paclitaxel and neratinib, 225
pre-treatment imaging, 96–98 T-DMI cohort, 226
MR spectroscopy, 95, 98, 99 tucatinib, 226
NCF incidence, 219
chemotherapy, 409, 410 initial therapy for, 220
immunotherapy and targeted agents, 410, 411 principles, 220, 221
PCI, 415, 416 screening, 220
SRS, 412, 415 systemic therapy
WBRT, 411, 412 bevacizumab, 221
neurocognitive effects, 407 capecitabine, 221
chemotherapeutic agents, 408 cisplatin plus etoposide, 221
multidisciplinary management, 407 clinical trials, 224
NCF, characteristics, 408, 409 etirinotecan pegol, 222
NCF, surgery impact, 409 high dose methotrexate, 221
neuropathological assessment, 53 iniparib and irinotecan, 223
novel PET agents, 106 ixabepilone with capecitabine, 222
NSCLC nanoliposomal irinotecan, 222
ALK tyrosine kinase inhibitors, 73, 74 phase II trials, 223
EGFR tyrosine kinase inhibitor, 72–73 radiation, 226, 227
immune checkpoint inhibitors, 74 temozolomide, 222
parenchymal brain metastases, 90 Breast-GPA index, 90
parental cell line, 37 Bremsstrahlung, 276
PDX model, 39 Breslow depth, 90
PET, 101–103 Brief Pain Inventory (BPI), 681
pleural fluids or lymph node metastases, 37 Brigatinib, 210
post-contrast T1 mapping, 105, 106 Brivaracetam, 129
postoperative management, 436, 437 Brownian motion, 100
preclinical models, 37, 39 Brown-Sequard syndrome, 121
primary cancer, 19, 20 Buparlisib (BKM120), 223
radiation-induced neuroxoticity, 416, 417 Bupropion, 161
secondary brain tumors, 19, 20
"seed and soil" hypothesis, 20, 21
SEER based incidence, 19 C
small cell and non-small cell lung cancer, 56, 117 Calreticulin, 360
solid-appearing lesion, 91 Canadian Clinical Trials Group (CCTG), 290
spinal metastasis, 29 Cancer pain
SRS, 69 brain procedure, 691, 692
SWI, 103, 104 intrathecal delivery of medication, 689, 690
systemic therapy, 40–42 mechanisms of, 687, 688
therapeutic implications, 23 resective and ablative spinal procedures, 688, 689
unbiased screens spinal cord stimulation, 690, 691
in situ transcriptomics, 44, 45 Cancer stem cells (CSCs), 22, 23
in vitro transcriptomics, 42–44 Cancer-mediated hypercoagulability, 141
non-coding RNAs, 45, 46 Cancer-related hypercoagulability, 141, 142
WBRT, 39, 40, 69, 83 Capecitabine, 177, 221
whole brain radiation therapy, 90 Carbamazepine, 131
Brain metastasis velocity (BMV), 290, 291 Carboplatin, 247
Brain radiotherapy, 413–414 Carotid body tumors (CBTs), 249
Brain tumors, 393, 395, 403 Carotid paraganglioma, 249
Brainstem lesions, management of, 386, 387 Castration-resistant prostate cancer (CRPC), 13, 514
722 Index
CD8+T cells, 195 epidemiology, 143–144

Cell plasma onconeural proteins, 196 management, 145, 146
Central nervous system (CNS) malignancy, 69 mechanisms and pathophysiology, 144
Central nervous system (CNS) metastases prognosis, 146
adjuvant analgesics, 680 safety of anticoagulation, 146
anticonvulsant drugs ischemic stroke
carbamazepine, 681 acute treatment management, 142
gabapentin, 680, 681 clinical presentation, 140
oxcarbazepine, 681, 682 diagnosis, 141
pregabalin, 680, 681 epidemiology, 139–140
antidepressants pathophysiology, 140, 141
amitriptyline, 682 prognosis, 143
duloxetine, 682 secondary prevention management, 142–143
cognitive impairment, 677 Ceritinib, 210
diversion, 679, 680 Cervical cancer, 250
epidemiologic studies Cervical spine fixation, 547, 548, 550
autopsy studies, 3, 4 anterior fixation, 547
population and hospital series, 4, 7 cervical spine fixation, 547, 548, 550
epidemiologic trends cervicothoracic junction, 550
breast cancer, 10, 11 Cervicothoracic junction, 550
CRC, 13 Cesium-131 (Cs-131) brachytherapy
lung cancer, 7–10 I-125 brachytherapy versus, 452, 453
melanoma, 11, 12 outcomes in brain metastases, 450–452
NSCLC, 7–9 Chemotherapy, 83, 176–178, 394, 407
rare CNS metastases, 13, 14 NCF, 409, 410
RCC, 12, 13 Choriocarcinoma, 249
SCLC brain metastases, 9, 10 Circulating tumor cells (CTCs), 23, 24
incidence Circulating tumor DNA (ctDNA), 25, 71
and characteristics of, 8 Cisplatin, 477
and median OS, 10 Cisplatin plus etoposide, 221
proportions of brain metastases, 5–6 CK M6 system, 306
ketamine, 676, 677 Clear cell sarcomas of the kidney (CCSK), 259
methadone, 676 Clinical target volume (CTV), 320
neurosurgical/radiotherapeutic options, 3 Coagulopathy, 144
opioid analgesics Cobalt-60, 276
fentanyl, 676 Cobimetinib, 237, 240
hydromorphone, 676 Cognitive behavioral therapy (CBT), 158, 159, 699
morphine sulfate, 675, 676 Cognitive dysfunction, 708
tramadol, 676 Colorectal cancer (CRC), 13, 28, 245, 246
opioid overdose, 677–679 Complementary and alternative medicine (CAM)
opioid safety considerations, 677 classification, 695
socioeconomic burden, 14, 15 definition, 695
topical agents modality, 695
capsaicin, 683 Complementary and integrative medicine (CIM)
lidocaine, 682, 683 acupuncture, 697, 698
Central nervous system (CNS) metastases brain clinical consultation, 696
metastases, see Brain metastases herbs and supplements, 699, 700
Central nervous system (CNS) metastases spinal massage, 698
metastases, 121 mind-body practices, 698
CEREBEL study, 225, 226 physical well-being
Cerebral blood volume (CBV), 96 exercise, 699
Cerebral metastasis, surgical resection for, 430 nutrition, 698
Cerebral spinal fluid (CSF), 120 psychological well-being, 699
Cerebral venous sinus thrombosis (CVST), 146, 147 symptom management, 696–697
Cerebrospinal fluid (CSF), 45 Compton effect, 276
Cerebrovascular disease complications Conical collimators, 299
CVST, 146, 147 Controlled Oral Word Association tests, 288, 289
ICH Conventional external beam radiation therapy (cEBRT),
clinical presentation, 145 560, 624, 645, 649
diagnosis, 145 Conventional magnetic resonance imaging, 523
Index 723
Cordotomy, 688 Dynamic contrast-enhanced magnetic resonance imaging

Coronal/bitemporal incision, 479 (DCE-MRI)
Corticosteroids, 478, 709, 710 atypical hemangiomas, 527
Cranioplasty, 482, 483 chordoma, 528
Crizotinib, 9, 73, 210 Gd-DTPA contrast agent, 524
Cryoablation, 580 healthy bone marrow and tumor vascularity, 526
CT/fluoroscopic-guided thermal ablation, 580 limitations, 528
CTLA-4 inhibitor ipilimumab, 237 malignant and benign vertebral compression
Cyberknife fractures, 526
arrangement and shaping, 306 MR acquisition, 525
beam properties, 305, 306 pharmacokinetic two-compartment model, 524–525
cost, 306 radiation therapy, 528
immobilization and image-guidance, 306 T1-weighted perfusion MRI technique, 524
modern model, 306, 307 Dynamic FET-PET, 103
treatment planning and delivery, 306 Dynamic susceptibility contrast (DSC) MR perfusion,
Cyclin-dependent kinase (CDK) pathway, 70 96, 97
Cyclophosphamide, 477 Dysesthesias, 675
Cytarabine, 176 Dysthymia, 153
Cytochrome P-450 (CYP) system, 132
Cytokeratin (CK), 58
Cytotoxic T lymphocyte-associated antigen-4 (CTLA-4), E
12, 361, 362 Eastern Cooperative Oncology Group, 240
Echinoderm microtubule like protein 4 (EML4) gene, 73
Echinoderm microtubule-associated protein-like 4
D (EMLA4), 210
Dabrafenib, 12, 77 Electrocorticography
Danger Model, 355, 356 brain metastases, 435
Decelerating radiation, 276 End-of-life care, 712, 713
Decompressive laminectomy, 545, 569 Endometrial carcinoma, 250
Deep vein thrombosis (DVT), 140 Endothelial growth factor receptor (EGFR) mutations, 498
Dejerine-Roussy Syndrome, 707 Enteral nutrition, 476
Delayed radiation-induced vasculitic EpCAM-negative CTCs, 24
leukoencephalopathy (DRIVL), 395 Epidermal growth factor receptor (EGFR), 25, 72,
Demineralized bone matrix (DBM), 546 208–210, 246
Denis’ model, 585 EGFR-directed therapies, 516, 517
2-deoxy-[18F]fluoro-D-glucose (FDG), 398 EGFR-mutant non-small cell lung cancer (NSCLC), 41
Depression, 153 EGFR-TKI-sensitizing mutation (EGFRm), 42
Depth-dose curve, 276, 277 tyrosine kinase inhibitors, 41, 177
Dialectical behavior therapy (DBT), 159 Epidural compression, 489, 490
Diffusion tensor imaging (DTI), 99, 531 Epidural hemorrhage (EDH), 143
brain metastases, 434, 435 Epidural spinal cord compression (ESCC), 490, 535, 576
Diffusion-weighted imaging (DWI) Epithelial membrane antigen (EMA), 59
diagnostic imaging, 530 Epithelial mesenchymal transition (EMT), 21–22
limitations, 531 Epithelioid glioblastoma, 54
monitoring treatment response, 530, 531 Erlotinib, 8, 72, 177, 293, 294
MR acquisition, 530 Esophageal cancer, 246, 247
Diffusion-weighted magnetic resonance imaging (DWI), Etirinotecan pegol (pegylated irinotecan), 222
99–101 European Organization for Research and Treatment of
Digitally reconstructed radiographs (DRRs), 301 Cancer (EORTC 22852-26001), 285
Dignity therapy, 160 EuroQol 5-D (EQ-5D), 497
Diode lasers, 458 Exosomes, 24, 25
Disseminated intravascular coagulation (DIC), 144 Expanded access programs (EAPs), 212
DNA-damage response, 271, 272 Expressive supportive counseling, 699
Donepezil, 417 External beam radiotherapy (EBRT), 659
Dormant cancer cells (DCCs), 23, 24
Dorsal root entry zone (DREZ), 689
Double-strand DNA break, 271, 276 F
Doxorubicin, 40, 478 Faces Pain Scale (FPS), 676
Drop metastases, 488 FACT-Brain QOL score, 91
Dual energy computed tomography (DECT), 531 Familial adenomatous polyposis (FAP), 245
Dynamic contrast enhanced (DCE), 408 Fatigue, 707
724 Index
Flap delay, 480 Gynecology

Flattening filter-free (FFF) treatment, 304 cervical cancer, 250
Flexion abduction external rotation (FABER) choriocarcinoma, 249
maneuver, 491 endometrial carcinoma, 250
Flt-3 ligand, 361 ovarian/fallopian tube cancer, 249, 250
Fluorescent in situ hybridization (FISH), 60
F-18 Fluorothymidine, 106
5-Fluorouracil, 478 H
Focal neurologic deficits, 118, 119, 706, 707 Head and neck cancers
Focal seizures, 126 paragangliomas, 248, 249
Fotemustine, 236 squamous cell carcinoma, 248
Functional MRI (fMRI), brain metastases, 434 Headaches, 119, 707
Health-related quality of life (HRQoL) tools, 497
Hemilaminotomy, 689
G Hepatocellular carcinoma (HCC), 247
Gabapentin, 131 HER2-directed therapies, 75, 516
Galea aponeurotica, 473 Hereditary nonpolyposis colorectal cancer (HNPCC),
Gallbladder cancer, 248 245, 272
Gamma Knife High-activity I-125 brachytherapy, 448
arrangement and shaping, 302 High-volume hormone-sensitive prostate cancer
beam properties, 302 (HSPC), 514
cost, 303 Hippocampal avoidance, 281, 282, 284, 287–289
immobilization and image-guidance, 303 Hippocampal avoidance whole-brain radiation therapy
modern model, 303 (HA-WBRT), 417
radiosurgery, 394 Hippocampal sparing (HSI), 40
intracranial lesion consistent with metastatic HLA associations with PNDs, 197
melanoma, 389 HMGB2, 360
for post-operative resection cavity, 385 Hopkins Verbal Learning Test-Revised (HVLT-R)
treatment planning and delivery, 302, 303 scores, 285
Gamma Knife stereotactic radiosurgery (GKRS), 379, Hormonal therapy
380, 401 androgen-directed therapies, 516
for brainstem metastatic lesion, 387 aromatase inhibitors, 515, 516
multifocal metastatic disease rather than WBRT, 384 HER2-directed therapies, 516
patient with multifocal metastatic breast cancer, 383 Hospital Anxiety and Depression Scale (HADS), 158
after WBRT, 381 Human EGFR 2 (HER2), 70
Gamma rays, 276 Huntington’s disease, 152
Gastric cancer, 247 Hybrid therapy, 559
Gastrointestinal cancers Hydroxyapatite, 482
colorectal cancer, 245, 246 Hyperbaric oxygen therapy, 480
esophageal cancer, 246, 247 Hypoactive deliriumoften masquerades, 154
gallbladder cancer, 248 Hypofractionated stereotactic radiosurgery (HF-SRS),
gastric cancer, 247 322, 329
hepatocellular carcinoma, 247 clinical outcomes, 333–335
pancreatic cancer, 247, 248 clinical trials, 337
GATA binding protein 3 (GATA3), 58 modeling tumor kill and normal tissue damage,
Gefitinib, 8, 72 330–333
Gemcitabine monotherapy, 248 ongoing clinical trials, 336, 337
Genetic heterogeneity, 70–71 studies, 335–336
Genetically engineered mouse models (GEMM), 38, 46 Hypofractionation, 322
Genitourinary cancers Hypothalamic-pituitary-adrenal (HPA) axis, 152
bladder cancer, 251 Hypoxia-inducible factor-1 alpha (HIF-1α), 395
prostate cancer, 251
renal cell carcinoma, 251
testicular cancer, 252 I
Germline antigens, 358 I-125 Brachytherapy
Gestational trophoblastic disease (GTD), 249 high-activity I-125 brachytherapy, 448
Glial acidic fibrillary protein (GFAP), 57 limitations and complications, 449
Glucocorticoids, 156 permanent implant, 448
Graded prognostic assessment (GPA), 84 permanent low activity I-125 brachytherapy, 448, 449
Guillain-Barre syndrome, 202, 203 temporary implant, 447
Index 725
IDO, 359 International Response Assessment in Neuro-Oncology

IFN-β:radiosurgery, 360, 361 (RANO) Working Group, 240
Immune checkpoint inhibitors (ICIs), 203, 361, 362 Intracranial hemorrhage (ICH), 143
Immune regulation, radiation, 361 clinical presentation, 145
Immune suppression, by tumors, 359 diagnosis, 145
Immunogenic cell death (ICD), radiosurgery, 359, 360 epidemiology, 143–144
Immunologically privileged site, 399 management, 145, 146
Immunotherapy, 83, 363, 394, 401 mechanisms and pathophysiology, 144
brain, immunity in, 357, 358 prognosis, 146
NCF, 410, 411 safety of anticoagulation, 146
optimal radiation parameters for, 364, 365 Intradural metastatic tumors, 488
and radiation necrosis, 399, 400 Intramedullary spinal cord metastasis (ISCM), 29
T cells Intraoperative radiotherapy (IORT)
APC activation of, 356, 357 advantages, 659
in immunity, 355, 356 brachytherapy, 659
tumor catheter-directed interstitial HDR approach, 667
immunosurveillance, 358, 359 common spinal brachytherapy sources and physical
immune suppression by, 359 properties, 660
Immunotherapy response assessment for neuro-oncology external beam radiotherapy, 659, 662
(iRANO) guidelines, 96 history, 660
InCise MLC system, 307 image-guided intensity modulated radiation therapy, 659
Increased intracranial pressure (ICP), 127 iodine-125, 662, 663
Infratentorial disease, 118 Ir-192, 666
Integrative medicine, see Complementary and Integrative LDR brachytherapy, 663, 665
medicine (CIM) P-32 plaque, 666, 667
Integrative Medicine Center model, 696 painful or unstable spinal metastases, 668
Intensity modulated proton therapy (IMPT), 307 phosphorous-32, 666
Intensity modulated radiation therapy (IMRT), 659 proton therapy, 659
Interbody grafts, 545–546 short-range HDR plaques, 666
Intercranial neurosurgical tumors small linear accelerator, 659
chemotherapy spinal brachytherapy sources and physical
alkylating agents, 477 properties, 661
anti-angiogenesis, 478 therapeutic doses of radiation, 659
antimetabolites, 478 ultrahypofractionated EBRT techniques, 665
chemotherapeutic antibiotics, 477 Yttrium-90, 666
corticosteroids, 478 Intraparenchymal hemorrhage (IPH), 143
incision planning, 478–480 Intrathecal chemotherapy, 265
management of complication Intratumoral hemorrhage (ITH), 144
cranioplasty materials, 482 Intravenous immunoglobulin (IVIG), 202
drains, 483 Iodine-125 (I-125), 662, 663, 665
scalp reconstruction algorithm, 481, 482 Iohexol, 489
timing of cranioplasty, 482, 483 Ionizing radiation, 275, 276
medical history, 476 IPASS trial, 208
prevention of complication Ipilimumab, 95, 211, 212, 294, 364, 400
flap delay, 480 iRANO criteria, 240
hyperbaric oxygen therapy, 480 Ischemia, 395
laser angiography, 480 Isotropic diffusion maps, 99
nitroglycerin ointment, 480, 481
tissue expansion, 480
radiation, 477 J
scalp Japanese Radiation Oncology Study Group
anatomy, 473, 474 (JROSG 99-1), 285
blood supply, 475 Juvenile pilocytic astrocytomas (JPAs), 264
innervation, 474
tension, 476, 477
tissue perfusion, 476 K
Interferon-alfa, 156 Kaplan-Meier curves, 87, 89
Interictal epileptiform discharges (IEDs), 128 Karnofsky Performance Score (KPS), 74, 83, 293
International Germ Cell Cancer Collaboration Group, 252 Kempke’s “T-bar” incision, 479
International League Against Epilepsy (ILAE), 126 Ketogenic diet (KD), 698
726 Index
KEYNOTE 024 study, 212 CSF

Kristen Rat Sarcoma 2 Viral Oncogene Homolog leukocyte count, 173
(KRAS), 211 markers, 174
Kyphoplasty, 488, 489, 491, 579, 613, 688 protein, 173
cytology, 173
diagnostic scheme
L cerebral, 172
Lacosamide, 129 clinical signs and symptoms, 171
Lambert Eaton myasthenic syndrome (LEMS), 202 cranial nerves, 172
Lambert-Eaton myasthenic syndrome, 196 neuroimaging, 172–173
Lamotrigine, 129 spinal, 172
LANDSCAPE trial, 225 elevated intracranial pressure, 173
Lapatinib, 75, 76, 225, 294 glucose concentration, 173, 174
Laser ablation after stereotactic radiosurgery incidence of, 170
(LAASR), 466 leptomeninx, 169
Laser angiography, 480 metastatic cancer cells, 171
Laser interstitial thermal therapy (LITT), 580 mutational analysis, 174–175
applications, 458 pachymeninx, 169
brain metastases prognosis of, 170–171
cancer survival, 462 treatment
clinical studies, 463–465 cytarabine, 176
conventional treatment, 462 fixed neurologic deficits, 175
FLAIR MRI sequences, 463 intrathecal therapy, 176
lesion size, 463 MTX, 176
multicenter phase II trial, 466 new treatment agents, 178
posterior fossa, 463 radiotherapy, 175, 176
prevalence, 462 response measurement, 175
QOL, 466 supportive therapy, 178
tumor ablation, 463 targeted therapies, 177, 178
Visualase system, 463 thiotepa, 176
volumetric analysis, 463 untargeted therapies, 176–177
cerebral radiation necrosis tumor markers, 174
advantages, 469 Leptomeninx, 169
biopsy, 467 Levetiracetam, 129
biopsy-proven CRN, 469 Limbic encephalitis (LE), 187, 200, 201
clinical studies, 468 Linac
lesion location, 467 arrangement and shaping, 304
lesion regrowth, 467 beam properties, 304
molecular mechanism, 466 cost, 305
pathophysiology, 466 immobilization and image-guidance, 305
phase II trial, 468 modern model, 305
treatment, 467 technological developments, 304
tumor recurrence, 467 treatment planning and delivery, 304, 305
diode lasers, 458 Linear accelerator, 394
history, 457, 458 Linear energy transfer (LET), 276
mechanism, 458 Linear-Quadratic (LQ) Model, 332
Nd:YAG lasers, 458 Lissauer’s tract, 689
NeuroBlate System, 458, 459 Longitudinal MRS, 99
operative technique, 461, 462 Lorlatinib, 210
post-LITT, 462 Low-molecular-weight heparin (LMWH), 142
pre-LITT, 460, 461 Lumbar puncture, 129
preoperative preparation, 460 Lung cancer, 7–10
Visualase Thermal Therapy System, 458, 459 anaplastic lymphoma kinase gene, 210, 211
Laser interstitial thermotherapy (LITT), 402 epidermal growth factor receptor, 208–210
Leptomeningeal carcinomatosis, 169 immunotherapy
Leptomeningeal disease (LMD), 120, blood brain barrier, 212
241, 436, 488, 489 expanded access programs, 212
Leptomeningeal metastasis KEYNOTE 024 study, 212
brain parenchyma, 169 PD-L1 antibody expression, 211
cerebrospinal fluid markers, 174 pembrolizumab, 212
choroid plexus epithelium, 169 tumor mutational burden, 212
Index 727
Kristen Rat Sarcoma 2 Viral Oncogene Homolog, 211 targeted therapy

radiation therapy and immunotherapy, 213 BRAF V600 mutations, 236
Lung metastatic derivative (LM), 44 BRAF/MEK inhibition, 237
Lung-molGPA, 84 dabrafenib, 236
Luteinizing hormone releasing hormone (LHRH), 516 radiation therapy, 240
Lymph nodes, as OAR, 365, 366 vemurafenib, 236
Lymphoma, 488 Melanoma brain metastasis, 42
Lynch syndrome, 245, 272 Melanoma-molGPA, 84
Melanotic schwannoma, 54
Memantine, 384
M Memorial Sloan Kettering Cancer Center
Magnetic resonance guided focused ultrasound (MSKCC), 260
(MRgFUS), 692 Memory and adaptation training (MAAT), 712
Magnetic resonance spectroscopy, 398 Meningeal carcinomatosis, 169
Magnetic resonance venography (MRV), 147 Mesenchymal-epithelial transition (MET), 21–22
M.D. Anderson Symptom Inventory Brain Tumor Metastatic epidural spinal cord compression (MESCC),
Module (MDASI-BT), 289 623, 624
Major depressive disorder (MDD), 153 Metastatic melanoma, 54
Major histocompatibility complex (MHC), radiosurgery, Metastatic vertebral lesions, 570
360, 361 Methotrexate (MTX), 176, 478
Major histocompatibility complex class I (MHC-I) Methyl methacrylate, 482
molecules, 195 METIS trial, 292
Malnutrition, 476 Metropolitan Detroit Cancer Surveillance System
Massage, 698 (MDCSS), 4
MATCH trial, 228 Metropolitan Detroit SEER database, 9
Matrix metalloproteinases (MMPs), 25–26 Microglia, 357
MD Anderson Cancer Center, 285 Midline myelotomy, 689
MD Anderson trial, 285 Mind-body practices, 698
Meaning-centered psychotherapy, 160 Mindfulness-based stress reduction (MBSR), 159
Mechanical instability Minimal access surgery (MAS)
bisphosphonates, 595 decompression and stabilization, 577
denosumab, 595 mechanical radiculopathy, 578
kyphoplasty, 593, 594 Minimally invasive surgical (MIS) approaches
multiple myeloma, 592 contraindications, 576
radiation, 594 NOMS decision framework, 576
stabilization, 591, 592 SINS criteria, 576
vertebroplasty, 593, 594 Mirtazapine, 161
Mechanical pain, 536, 644 Mitogen activated protein kinase (MAPK) pathway, 77
Mechanical radiculopathy, 578 Mitogen-activated protein (MEK) kinase, 77
Mechanical stability Mitogen-activated protein kinase (MEK) inhibitors, 12
animal and cadaveric work, 589, 590 Mitomycin C, 478
osteoblastic v. osteolytic lesions and CT imaging, 590 Monoclonal antibody immunotherapies, 399
Mechlorethamine, 477 Mood disorders
Melanoma, 11, 12 antidepressants, 161
immune checkpoint inhibitors anxiety disorders, 155
ipilimumab and nivolumab, 238 bipolar disorders, 154, 155
pembrolizumab and bevacizumab, 238 cancer-related outcomes
stereotactic radiosurgery, 238 caregivers, 158
immune therapy with SRS, 239 engagement in treatment, 157
immunotherapy, 77, 78 healthcare utilization and costs, 158
MAPK pathway, 77 morbidity, 157
Pi3K/AKT/mTOR pathway, 77 mortality, 157
risk factors, 235 definition, 151
SRS, 346–348 depressive disorders
systemic therapy differential diagnosis, 153–154
fotemustine, 236 epidemiology, 152–153
leptomeningeal disease, 241 etiology, 152
radiation necrosis, 240, 241 emotional well-being, 151
response criteria for brain metastasis, 240 physical well-being, 151
stereotactic radiosurgery, 239 role of screening, 158–159
temozolomide, 236 suicidality, 155
728 Index
Mood disorders (cont.) N-methyl-D-sspartate (NDMA) antagonists, 676–677

symptoms, cancer treatments N-methyl-D-aspartate receptors, 287, 288
AEDs, 156 Nociceptive pain, 674
hormonal agents, 156 Nociceptive system, 673
immunotherapies, 156 NOMS framework, 490
radiation therapy, 156, 157 Nonbacterial thrombotic endocarditis (NBTE), 140
steroids, 156 Nonconvulsive status epilepticus (NCSE), 128, 133
treatment strategies Non-small cell lung cancer (NSCLC), 7–9, 19, 515
depression and anxiety symptoms, 160 ALK tyrosine kinase inhibitors, 73–74
drug-drug interactions, 162 EGFR tyrosine kinase inhibitor, 72–73
mechanism of delivery, 162 immune checkpoint inhibitors, 74
medications, 160–162 Non-steroidal anti-inflammatory drugs (NSAIDs), 162
mood stabilizers, 162 Novalis TX system, 305
potential targets, 160 NRAS mutations, 236
primary symptom of interest, 160 Nuclear medicine imaging, 398
problematic side effects, 160, 161 Numb chin syndrome, 172
psychotherapy, 159–160 Nutrition, 476
Motexafin gadolinium (MGd), 293
Motor deficits, 707
MTX-induced leukoencephalopathy, 176 O
Multifocal disease, stereotactic radiosurgery SRS, O-(2-[18F]fluoroethyl)-l-tyrosine (FET), 101
383, 384 Obsessive Compulsive Disorder, 152
Multifocal radio-resistant disease, SRS for, 388–390 Occipitocervical fixation, 547
Multi-leaf collimators (MLCs), 300 Oligodendrocytes, 395
Multiple endocrine neoplasia type 2 (MEN 2), 248 Onconeural antibodies, 194
Multiple melanoma metastases, SRS, 348–350, 382, Onconeural proteins, 194
388–390 Opsoclonus-myoclonus syndrome (OMS), 201, 202
Multi-voxel MRS techniques, 98 Organ at risk (OAR), lymph nodes as, 365, 366
M-Vision software, 459 Osimertinib, 42, 209, 210, 294
Osteoblastic lesion formation, 585, 587
Osteolytic lesion formation, 585, 587, 588
N Osteonecrosis of the Jaw (ONJ), 508
N-acetyl aspartate (NAA), 98, 398 Osteoprotegerin (OPG), 504
Nanoliposomal irinotecan, 222 Oswestry Disability Index (ODI), 497
National Center for Complementary and Alternative Ovarian/fallopian tube cancer, 249, 250
Medicine (NCCAM), 695 Oxcarbazepine, 131
National Comprehensive Cancer Network
(NCCN), 151, 208
NCI MATCH (Molecular Analysis for Therapy Choice) P
trials, 228 Pachymeninx, 169
Nd:YAG lasers, 458 Paclitaxel, 40
NEfERT-T randomized clinical trial, 225 Pain
Neurocognitive function (NCF) affective component of pain, 674
characteristics, 408, 409 central sensitization, 674
chemotherapy, 409, 410 definition, 673
immunotherapy and targeted agents, 410, 411 descending supraspinal systems, 674
PCI, 415, 416 nociceptors, 673
SRS, 412, 415 peripheral sensitization, 674
surgery impact, 409 types, 674, 675
WBRT, 411, 412 wind-up phenomenon, 674
Neurocognitive sparing radiotherapy, 384 Palliative care, 371
Neurocognitive tests, clinical trial battery of, 415 anti-seizure medicines, 710
Neurofibromatosis (NF-1), 248 brain metastases, 705
Neurological immune related adverse events caregivers needs, 713
(nirAEs), 203 CNS metastases, 705, 706
Neuronavigation software suites (nTMS), 435 prognostic significance, 706
Neuropathic pain, 674, 675 cognitive impairment, 712
Nickel-60, 276 corticosteroids, 709, 710
Nitroglycerin ointment, 480, 481 early outpatient, 705
Nivolumab, 74, 95, 211 end-of-life care, 712, 713
Index 729
fatigue, management, 711 incidence, 259

pain management, 711 multimodal therapy, 264, 265
radiation therapy, 709 pathophysiology, 260, 261
rehabilitation, 710, 711 prediction occurrence, 265
and supportive care interventions, 708–711 prevalence, 259
surgery, 708, 709 prophylaxis, 265
symptoms proton therapy, 264
cognitive dysfunction, 708 radiation treatment, 261, 264
fatigue, 707 surgical treatment, 261
focal neurologic deficits, 706, 707 tumor characteristics, 260, 261
headaches, 707 Pembrolizumab, 74, 95, 211, 212, 410
seizures, 708 Permanent low activity I-125 brachytherapy, 448, 449
PALOMA-2 trial, 223 Pertuzumab, 75
Pancreatic cancer, 247, 248 Pharmacokinetic two-compartment model, 524–525
Pan-hypopituitarism, 120 Phase III BEACON trial, 222
Paragangliomas, 248, 249 Phenytoin, 130
Paraneoplastic encephalitis, 125 Phosphatase and tensin homolog (PTEN), 71
Paraneoplastic neurological disorders (PNDs) Pituitary adenoma versus metastatic carcinoma, 56
autoimmune encephalitis, 200, 201 Pleomorphic xanthoastrocytoma, 54
cancers, 188 Pleuropulmonary blastoma (PPB), 259
cell plasma onconeural proteins, 196 Poly ADP-ribose polymerase (PARP), 75
classical PNDs, 189 Polymethylmethacrylate (PMMA) injection, 613, 614
definite PNDs, 189 Polyneuropathy, organomegaly, endocrinopathy,
distinct HLA associations with PNDs, 196–197 monoclonal gammopathy, and skin changes
immune checkpoint inhibitors therapy, 203 (POEMS) syndrome, 188
immune response against onconeural Positron emission tomography (PET), 95, 101–103, 129
proteins, 194, 195 Posterior reversible encephalopathy syndrome (PRES),
intracellular onconeural proteins, 195, 196 144
Lambert Eaton myasthenic syndrome, 202 Postoperative complications
non-classical PNDs, 189 CSF leaks, 606, 608
onconeural antibodies, 194 hardware failure, 608–610
onconeural antigens, 194 intraoperative and post-operative hemorrhage, 604,
opsoclonus-myoclonus syndrome, 201–202 605
pathogenesis, 187, 188 medical, 603
prevalence, 187 neurologic deterioration, 610
probable PNDs, 194 risk factors, 603, 604
subacute cerebellar degeneration, 197 surgical-related complications, 603
subacute sensory neuronopathy, 201 venous thromboembolism, 610
symptoms, 187 wound infection and dehiscence, 605–607
treatment, 202, 203 Pregabalin, 129
Paresthesias, 675 Prescription Drug Monitoring Program (PDMP), 680
Patent foramen ovale (PFO), 140 Programmed death-1 (PD-1), 211, 362, 410
Pathologic vertebral compression fractures (VCF) antibody, 211–213
anesthesia care, 616 inhibitors, 399
contraindications, 617, 618 Progressive disease, 372
diagnosis, 617 Prophylactic cranial irradiation (PCI), 286, 287, 291,
kyphoplasty, 613, 618 415, 416
pain quality and characteristics, 616–617 Prophylaxis, 132, 265
patient selection, 616 Prostate cancer, 251
PMMA injection, 613, 614 Proton magnetic resonance spectroscopy (MRS), 98
post-procedural care, 614, 615 Proton radiation, 277, 278
potential complications, 618 Proton therapy
radiofrequency ablation, 619 arrangement and shaping, 307
treatable vertebral levels, 618 beam properties, 307
vertebral augmentation procedures, 613, 614 cost, 308
vertebroplasty, 614, 618 efficacy and appropriate indications, 308
Patient derived xenografts (PDX), 38, 39, 46, 47 immobilization and image-guidance, 308
Patient Health Questionnaire-9 (PHQ-9), 158 treatment planning, 307, 308
Patient-centered outcomes, 705 Pseudoprogression, 101
Pediatric solid tumors Pulmonary embolism (PE), 140
730 Index
Q Reoxygenation, 273
Quality of Life after Treatment for Brain Metastases Resection cavity, SRS, 317, 320, 322
(QUARTZ) trial, 284 Resection method, brain metastases, 436
Quantitative susceptibility mapping (QSM), 102 Response Assessment in Neuro-Oncology Brain
“Question mark” incision, 479 Metastases (RANO-BM), 228, 372
R S
Radiation injury, SRS, 349, 352 Salvage irradiation, 371
Radiation myelitis, 659 RANO-BM CNS disease response criteria, 372
Radiation necrosis, 235, 240, 241, 324 recurrent brain metastases, 371, 372
development, 393, 394 therapeutic options, 372, 373
diagnosis, 396–398 repeat irradiation, rationale for, 373
immunotherapy and, 399, 400 SRS after SRS, 373, 374
pathophysiology, 395, 396 SRS after WBRT, 374, 375
surveillance for, 398, 399 WBRT after WBRT, 375, 376
treatment of, 400, 402, 403 SANDPIPER trial, 223
Radiation physics SATURN trial, 293
ionizing radiation, 275, 276 Scalp reconstruction algorithm, 481, 482
photon radiation, 276, 277 Schizophrenia, 152
proton radiation, 277, 278 Score Index for Radiosurgery (SIR), 83–84
Radiation therapy (RT), 175, 379, 407 Secondary malignancy, 275, 277, 278
in cancer, 359 Seizures, 708
Radiation Therapy Oncology Group (RTOG), 83, 84 antiepileptic drugs, 130, 133
Radiation-induced cavernous malformations, 104 gabapentin, 131
Radiation-induced neurotoxicity, 408 levetiracetam, 129
brain metastases, 416, 417 oxcarbazepine and carbamazepine, 131
Radicular pain, 644 phenytoin, 130
Radioactive decay, 276 topiramate, 131
Radiobiology zonisamide, 131
accelerated repopulation, 273 anti-epileptic drugs, 130
cell cycle and redistribution, 273 clinical manifestations, 126–127
DNA-damage response, 271, 272 convulsive status epilepticus, 134
normal tissue side effects, 275 diagnostic evaluation, 128–129
radiosensitizers, 273–275 end of life, 135
tumor hypoxia, 272, 273 epidemiology, incidence, and etiology, 125
Radiofrequency ablation (RFA), 580, 619 etiologies of, 126
Radionecrosis, 333, 334 NCSE, 133
Radiosensitizers, 273–275 placing driving restrictions, 134
Radiosensitizing drugs, 273–275 status epilepticus, 133
Radiosurgery, 393 treatment
combination of checkpoint inhibitors with, 362–364 AEDs, 129
ICD, 359, 360 drug interactions, 132
immune regulation, 361 prophylaxis, 132
lymph nodes as OAR, 365, 366 side effects, 132, 133
MHC and IFN-β, 360, 361 tumor-directed therapy, 131–132
radiation therapy, in cancer, 359 tumor-associated epilepsy, 127
RECIST criteria, 222, 240 Selective serotonin reuptake inhibitor (SSRI), 161, 162
Recurrent brain metastases Self-Nonself Model (SNS), 355
salvage irradiation, 371, 372 Sensory deficits, 707
salvageirradiation, therapeutic options, 372, 373 Separation surgery
Recurrent ependymomas, 264 delayed complications, 566
Recursive partitioning analysis (RPA), 84, 316 immediate complications, 565, 566
Referred pain, 536 intraoperative adjuvants
Rehabilitation, 710, 711 fenestrated screws/cement augmentation, 564
Re-irradiation GKRS, SRS P32 brachytherapy, 564
after whole brain radiotherapy, 381, 382 ultrasound guidance, 563
dosage and treatment considerations, 381 vertebroplasty, 563
Relative biological effectiveness (RBE), 276 mechanical instability, 560
Relative cerebral blood volume (rCBV), 398 neurologic assessment, 560
Renal cell carcinoma (RCC), 12, 13, 251, 499, 500 NOMS decision framework, 560
Index 731
oncologic consideration, 560 biologic pain, 490, 491

postoperative management biomechanical model
SBRT, 565 Denis’ model, 585
simulation, 565 focal kyphosis, 583, 584
rationale, 560, 561 bone-targeted therapies, 517, 518
surgical approach, 562, 563 chemotherapy
surgical considerations, 561, 562 breast cancer, 514
systemic disease, 560 non-small cell lung cancers, 515
Serotonin-norepinephrine reuptake inhibitors prostate cancer, 514, 515
(SNRIs), 161 clinical evaluation, 490–491
Short Form 36 (SF-36), 497 clinical presentation and diagnosis, 496
Single-drug therapy, 129 incidentally-found asymptomatic lesions, 645
Single-fraction stereotactic radiosurgery spinal cord compression, 644
(SF-SRS), 329, 330 vertebral metastases, 644
Single-isocenter, multi-target (SIMT) SRS, 336 without neurologic abnormality, 644
Single-photon emission computerized tomography CNS metastases, 518
(SPECT), 98 complete/incomplete neurologic deficits, 569
Single-voxel MRS methods, 98 computed tomography, 489, 497
Skeletal Oncology Research Group (SORG) conventional MRI, 523
nomogram, 497 decompressive laminectomy, 569
Skull base tumors, radiosurgery for, 380 diagnosis and decision-making, 544
Skull-base metastasis, SRS, 379 diagnostic evaluation, 497
dosage and treatment considerations, 380 diffusion tensor imaging, 531
skull base tumors, 380 diffusion-weighted imaging, 497
Small cell lung cancer (SCLC), 7, 9, 10, 46, 201, 202 diagnostic imaging, 530
brain metastases, 291, 292 limitations, 531
Smearing technique, 308 monitoring treatment response, 530, 531
Sodium glycididazole, 293 MR acquisition, 530
Somatic nociceptive pain, 674 dual energy CT, 531
Sorafenib, 247 dynamic contrast-enhanced MRI perfusion
Spinal cord compression imaging, 497
clinical presentation and diagnosis, 644 dynamic contrast-enhanced magnetic resonance
conventional external beam radiation therapy, 649 imaging
decision making algorithms, 652 atypical hemangiomas, 527
grading systems, 647 chordoma, 528
neurologic symptoms, 646 Gd-DTPA contrast agent, 524
oncologic assessment, 645 healthy bone marrow and tumor vascularity, 526
by radiation, 649, 650 limitations, 528
spinal involvement, 646, 647 malignant and benign vertebral compression
spinal stability, 646 fractures, 526
stereotactic radiosurgery, 650, 651 MR acquisition, 525
surgical management, 649 pharmacokinetic two-compartment model,
treatment complications 524–525
acute exacerbation of pain, 654 radiation therapy, 528
long-term neurologic complications, 654 T1-weighted perfusion MRI technique, 524
long-term non-neurologic complications, 654–655 EGFR-directed therapies, 516, 517
treatment failures, 654 epidemiology, 513
treatment outcomes epidural spinal cord compression treatment, 535
control of metastatic spine pain, 652 health-related quality of life (HRQoL) tools, 497
epidural tumor control and neurologic hormonal therapy
compromise, 652, 653 androgen-directed therapies, 516
re-irradiation outcomes, 653 aromatase inhibitors, 515, 516
Spinal cord stimulation, 690, 691 HER2-directed therapies, 516
Spinal Instability Neoplastic Score, 490, 544, imaging, 537
560, 576, 646 initial manifestation, 495
Spinal metastases, 121 immunotherapy, 518
ablation, 580 incidence, 495, 569
anatomical location, 569 kyphoplasty, 579
androgen-directed therapies, 516 leptomeningeal disease, 488, 489
anti-angiogenic therapies, 517 magnetic resonance imaging, 497
732 Index
Spinal metastases (cont.) symptomatic lesions, 569

MAS approaches systemic assessment, 539
decompression and stabilization, 577 systemic anticancer therapies, 513
mechanical radiculopathy, 577–578 targeted therapies/immunotherapy, 496, 518
percutaneous pedicle screw fixation, 579 targeted treatment paradigms
mechanical assessment, 538, 539 breast cancer, 498
mechanical instability lung cancer, 498, 499
bisphosphonates, 595 prostate cancer, 499
denosumab, 595 renal cell carcinoma, 499, 500
kyphoplasty, 593, 594 treatment, 491, 492
multiple myeloma, 592 tumor location, 488, 489
radiation, 594 tumor-specific molecular signatures, 496
stabilization, 591, 592 vertebral metastases and immunotherapy, 518
vertebroplasty, 593, 594 vertebroplasty, 579
mechanical pain, 490, 491 Spinal stereotactic radiosurgery (SSRS), 624, 626, 630
mechanical stability Spine Instability Neoplastic Score (SINS), 538, 539
animal and cadaveric work, 589, 590 Spine laser interstitial thermal therapy (sLITT)
osteoblastic v. osteolytic lesions and clinical outcomes and results, 630–632
CT imaging, 590 high-resolution T2WI, 627
MIS approaches intraoperative MRI, 626, 627
contraindications, 576 metastatic epidural spinal cord
NOMS decision framework, 576 compression, 623, 624
SINS criteria, 576 patient selection, 626
mechanical pain, 644 percutaneous instrumentation with cement
metastatic spine, 589 augmentation, 629
minimally invasive surgery techniques, 570 rationale, 625
motor weakness and/or myelopathy, 536, 537 SSRS, 624
MRI, 489, 490 sterile reference array, 627, 628
neurologic assessment, 536 symptomatic and neurologic improvement, 624
neurological evaluation, 491 transpedicular or transforaminal trajectory, 627
NOMS framework, 513, 535, 536 tumor histology, 624
numbness/tingling, 537 Spine Oncology Study Group Outcome Questionnaire
oncologic assessment, 537, 538 (SOSGOQ), 497
osteoblastic lesions, 587 Spread out Bragg peak (SOBP), 277
osteolytic lesions, 587, 588 Squamous cell carcinoma, 248
osteoporotic spine, 588, 589 Squamous cell lung cancers (SQCLC), 28
pain, 536 Standardized uptake value (SUV), 101
biologic, 536 Status epilepticus, 133
mechanical, 536 Stereotactic ablative radiotherapy (SABR), 665
prevalence, 536 Stereotactic body radiotherapy (SBRT),
referred, 536 559, 565, 624, 665
parameters, 513 Stereotactic navigation, brain metastases, 433, 434
pathophysiology, 496 Stereotactic radiosurgery (SRS), 12, 40, 69, 83, 95, 207,
patient’s native immune system, 513 213, 316, 379, 650, 651
patient-reported outcome (PRO) tools, 497 advantages and disadvantages, 309, 311
percutaneous pedicle screw fixation, 579 beam shaping, 299, 300
positron emission tomography, 497 brainstem lesions, management of, 386, 387
postoperative complications (see Postoperative clinical outcomes, 310
complications) Cyberknife
plain X-rays, 497 arrangement and shaping, 306
primary tumor pathology, 487 beam properties, 305, 306
prognostic factors, 497 cost, 306
radicular pain, 644 goal, 305
radiographic evaluation, 490 immobilization and image-guidance, 306
separation surgery (see Separation surgery) modern model, 306, 307
short tau inversion recovery, 497 treatment planning and delivery, 306
SINS framework, 590, 591 dosimetry, 310, 311
spinal instability, 543 eloquent/critical structures, 325
supportive care features, 309
analgesia, 519 frame-based SRS, 302
corticosteroids, 519 frameless SRS, 302
Index 733
Gamma Knife Stereotactic radiotherapy (SRT), 337

arrangement and shaping, 302 Steroids, 400, 478
beam properties, 302 Stevens-Johnson Syndrome, 162
cost, 303 Stress, 699
immobilization and image-guidance, 303 Subacute cerebellar degeneration, 197
modern model, 303 Subacute sensory neuronopathy, 201
treatment planning and delivery, 302, 303 Subarachnoid hemorrhage (SAH), 143
image-guidance systems, 301 Subdural hemorrhage (SDH), 143
immobilization, 301 Suicidality, 155
large metastases, 343, 346 Superficial musculoaponeurotic system (SMAS), 473
Linac Superparamagnetic iron oxide (SPIO) nanoparticles, 105
arrangement and shaping, 304 Surgery
beam properties, 304 brain metastases, 432, 433
cost, 305 SRS vs., 323
immobilization and image-guidance, 305 Surveillance Epidemiology and End Results (SEER)
modern model, 305 database, 4
technological developments, 304 Susceptibility-weighted imaging (SWI), 103, 104
treatment planning and delivery, 304, 305 Systemic anticancer therapies, 513
management Systemic therapy, 560
hypofractionation and postoperative resection
cavity, 322
patients with more than four brain T
metastases, 324 “T-bar” incision, 480
postoperative irradiation, 317 T cell receptor (TCR) dimer, 355
postoperative resection cavity, 317, 320, 322 T cells
randomized controlled clinical trials, APC activation of, 356, 357
318–319, 321 in immunity, 355, 356
surgical resection, 317 T1-weighted dynamic contrast-enhanced (DCE) MR
melanoma, 346–348 perfusion, 96
multiple melanoma metastases, elderly patient with T790M mutations, 209
comorbidities, 388–390 Tachyphylaxis, 678
multiple metastases, 348–350, 382 Tamoxifen, 162
NCF, 412, 415 Targeted Agent and Profiling Utilization Registry
neurocognitive side effects of, 413–414 Study, 228
neurocognitive-sparing treatment, multifocal disease, Targeted drug therapies, 83
383, 384 Taselisib, 223
post-resection cavities, frameless to, 385, 386 Temozolomide (TMZ), 177, 222, 236, 274, 293, 410
preoperative, 322, 323 Temporal lobe seizures, 126
principles, 299 Tension, 476, 477
prognostic scoring systems and patient Testicular cancer, 252
selection, 316, 317 Testicular germ-cell tumor (TGCT), 252
proton therapy Thecal sac deformation, 490
arrangement and shaping, 307 Thiotepa, 176, 477
beam properties, 307 Thoracolumbar fixation
cost, 308 anterior fixation, 550–551
efficacy and appropriate indications, 308 posterior fixation, 551
immobilization and image-guidance, 308 Tissue expansion, 480
treatment planning, 307, 308 Tissue perfusion, 476
radiation injury, 349, 352 Tissue plasminogen activator (tPA), 142
re-irradiation GKRS after WBRT, 381, 382 Tokuhashi revised score, 497
in reirradiation setting, 324 Toll like receptors (TLRs), 356
skull-base metastasis, 379 Tomita score, 497
dosage and treatment considerations, 380 Topiramate, 129, 131
skull base tumors, 380 Total hysterectomy with bilateral salpingo-
treatment planning, 300 oophorectomy, 250
tumor bed radiosurgery, 352, 353 Tourette Syndrome, 152
versus surgery, 323 Trail-Making Test, 288, 289
after WBRT, salvage irradiation, 374, 375 Trametinib, 12
with/without WBRT, 323, 324 Transcranial magnetic stimulation (TMS), brain
workflow, 301, 302 metastases, 434, 435
734 Index
Transesophageal echocardiography (TEE), 141 Whole brain radiotherapy (WBRT), 19, 39, 69,
Translational Breast Cancer Research Consortium 83, 90, 95, 207, 209, 213, 315, 329,
(TBCRC) 022 trial, 225 379, 707
Transthoracic echocardiography (TTE), 141 brain metastasis velocity, 290, 291
Transverse myelitis, 203 concomitant systemic agents, 292–294
Trastuzumab, 11, 75, 226 conventional WBRT
Tricyclic Antidepressants (TCAs), 161 acute, early-delayed, or late-delayed
Triple negative breast cancer (TNBC), 39, 75 complications, 284
Tucatinib, 226 approach, 283
Tumor, immune suppression by, 359 lateral portal, 283
Tumor bed radiosurgery, SRS, 352, 353 omission of WBRT, 284, 285
Tumor emboli, 140 prophylactic cranial irradiation, 286, 287
Tumor hypoxia, 272, 273 stereotactic radiosurgery, 285, 286
Tumor immunosurveillance, 357–359 surgical resection, 286
Tumor infiltrating lymphocytes (TILs), 26 efficacy, 282
Tumor location, 408 modern WBRT
Tumor treating fields (TTFields), 292 hippocampal avoidance, 288, 289
Tumor-directed therapy, 131–132 newly diagnosed brain metastases, 290
Tumor-induced spinal instability, 579 NMDA receptor antagonists, 287, 288
Tyrosine kinase inhibitors (TKI), 132 NCF, 411, 412
optimal patient selection, 294, 295
salvage irradiation
V SRS after, 374, 375
Vacuum bite-block systems, 301 WBRT after, 375, 376
Vascular endothelial growth factor (VEGF), 26, 144, 395 small cell lung cancer brain
Vascular endothelial growth factor pathway metastases, 291, 292
inhibitors, 517 TTFields, 292
Vascular endothelial growth factor receptor (VEGFR), Wound healing
12, 208, 251 collagen maturation, 636
Vasogenic edema, 430 complications
Vemurafenib, 77, 236, 237, 240 CSF leak, 641
Venous thromboembolism (VTE), 139, 146, 610 dehiscence/infections, 641
Vertebral augmentation procedures (VAP), 613, 614 seroma, 641
Vertebral metastases, 644 multidisciplinary approach, 637
Vertebrectomy platelet cytokine release, 636
indications, 570 postoperative optimization, 640, 641
minimal invasive lateral retropleural/transthoracic preoperative optimization, 637
approach, 571–573 reconstructive options, 637, 638
preoperative planning, 570, 571 revision spine surgery, 640
surgical approaches, 570 spinal metastases patient, 636
Vertebroplasty, 579, 614, 688
Visceral nociceptive pain, 674
Visual Analog Scale (VAS), 497 X
Visualase Thermal Therapy System, 458, 459 Xeroderma pigmentosum (XP), 271
Vitamin K deficiency, 144
Volumetric modulated arc therapy (VMAT), 304
von Hippel-Lindau syndrome (VHL), 54, 248 Y
Yoga, 698
W
Weinstein-Boriani-Biagini tumor classification, 627 Z
Weschler Memory Scale-III Word Lists test, 288 Zonisamide, 131

Stereotactic Radiosurgery Technology

Uploaded by

Copyright:

Available Formats

Stereotactic Radiosurgery Technology

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stereotactic Radiosurgery Technology

Uploaded by

Copyright:

Available Formats

Central Nervous

Central Nervous System

ISBN 978-3-030-42957-7 ISBN 978-3-030-42958-4 (eBook)

© Springer Nature Switzerland AG 2020

My heartfelt gratitude goes to my family for their support and

To my beautiful daughter, Hannah, who has filled our hearts

To my parents, Samuel Emerson and Kristine Sandberg Knisely,

I strongly recommend Central Nervous System Metastases: Diagnosis and

We live in an interesting time as we treat patients with central nervous system

Please contact me directly (ror9068@med.cornell.edu) should you have

New York, NY, USA Rohan Ramakrishna, MD

Part II Evaluation of CNS Metastasis

12 Leptomeningeal Disease and the Role of

Part III Radiation Therapy of CNS Metastasis

26 Salvage Irradiation for Patients with Recurrent

Part IV Surgical Treatment of Brain Metastasis

Part V Spinal Metastases: Foundations

37 Systemic Therapies for Patients with Metastatic

Part VI Surgical Treatment of Spinal Metastases

Manmeet S. Ahluwalia, MD Burkhardt Brain Tumor and Neuro-Oncology

Mariana E. Bradshaw, PhD, ABPP Section of Neuropsychology,

Atman M. Desai, MD, FAANS Department of Neurosurgery, Stanford

Lorenz C. Hofbauer, MD Department of Medicine III, Division for

Ilya Laufer, MD Department of Neurosurgery, Memorial Sloan Kettering

David M. Otterburn, MD Department of Plastic Surgery, NewYork-­

Robert C. Rostomily, MD Department of Neurosurgery, Houston Methodist

Mark M. Souweidane, MD, FACS, FAAP Department of Neurosurgery,

Panagiotis J. Vlachostergios, MD, PhD Department of Medicine, Division

Introduction secondary to many factors, including longer

© Springer Nature Switzerland AG 2020 3

lowing brain metastasis diagnosis with 10.7-day References

Table 2.1 SEER-based incidence of brain metastases

© Springer Nature Switzerland AG 2020 19

limit the effectiveness of systemic chemother- Seed and Soil

Endothelial cell Fibroblast/mesenchymal cell Monocytes/eosinophil

Red blood cell Macrophages Circulating free DNA

Extracellular matrix Lymph node

Fig. 2.1 Steps of brain metastatic cascade

Primary tumor site Brain tumor site

Epithelial markers Transcription factors miRNA Stem-cell markers Mesenchymal markers

Localized tumor Residual disease Micrometastasis Macrometastasis Brain tumor relapse

First-line therapy Asymptomatic progression Second-line therapy

- Molecular profiling & clonal evolution

Although CTCs are not routinely identified Exosomes

M1 to M2 immunosuppressive phenotype [71,  he Final Metastatic Phase: Brain

Molecular Heterogeneity shared by paired primary and BM specimens,

 odels for Brain Metastasis

© Springer Nature Switzerland AG 2020 37

GFP+/ Luc+ BrM1 BrM2 BrM3

Braf CA;Cdkn2alox/lox;Ptenlox/lox;myrAkt1 Rb1lox/lox;Trp53lox/lox

Table 3.1 Use of preclinical models to test targeted therapies

iting extracranial disease thus increasing overall Melanoma brain metastasis patients also ben-

Genetically engineered mouse models (GEMMs)

Neuropathological assessment of brain metasta- the differential diagnosis. Metastases tend to

© Springer Nature Switzerland AG 2020 53

As an example from our own practice, a pituitary Permanent Histology

Thyroid Adrenocortical Renal cell

routinely assess for neuroendocrine differentia- susceptibility to immune checkpoint inhibitors is

 olecular Assessment of Brain

Variant Type VAF CN

New York, NY, USA Rohan Ramakrishna, MD

Part II Evaluation of CNS Metastasis

12 Leptomeningeal Disease and the Role of

Part III Radiation Therapy of CNS Metastasis

26 Salvage Irradiation for Patients with Recurrent

Part IV Surgical Treatment of Brain Metastasis

Part V Spinal Metastases: Foundations

37 Systemic Therapies for Patients with Metastatic

Part VI Surgical Treatment of Spinal Metastases

David M. Otterburn, MD Department of Plastic Surgery, NewYork-

Introduction secondary to many factors, including longer

limit the effectiveness of systemic chemother- Seed and Soil

Although CTCs are not routinely identified Exosomes

M1 to M2 immunosuppressive phenotype [71, he Final Metastatic Phase: Brain

Molecular Heterogeneity shared by paired primary and BM specimens,

odels for Brain Metastasis

As an example from our own practice, a pituitary Permanent Histology

olecular Assessment of Brain

Introduction depending on the primary malignancy [1].

monly the threonine-to-methionine substitution SCLC: Anaplastic Lymphoma Kinase

Melanoma: Mitogen-Activated was 7.5 months [69]. Interestingly, genomic

Introduction radiation therapy (WBRT), chemotherapy, tar-

Introduction In the postradiation period, the primary role of

Dynamic susceptibility contrast (DSC) MR per- retreatment Imaging for Differential

Posttreatment MR Spectroscopy

In the posttreatment period, MR perfusion is used Technique

Pretreatment The role of SWI in posttreatment imaging is also

Introduction in part due to improved imaging techniques as

Epidemiology, Incidence, monly cortical lesions in the frontal lobe, parietal

Diagnostic Evaluation also not routinely needed for those without

Phenytoin seizures in patients with cancer, however, due to

of primary brain tumors with chemotherapy, Untoward Effects of Anticonvulsants

Driving quality of life. Driving restrictions can signifi-

Introduction Ischemic Stroke

The hypercoagulable state associated with cancer, Epidemiology

Management worthwhile considering endovascular therapy

Clinical Presentation in areas outside the hemorrhage is also sugges-

Depressive Disorders development of mood symptoms [12]. Depressive

ties, and involvement in activities that are likely to Suicidality

When a medication suspected of contributing Morbidity

Healthcare Utilization and Costs patients with earlier-stage disease. Studies specif-

Introduction most surface of the brain parenchyma lies the glia

Cerebral due to oculomotor, trochlear, or abducens nerve

[110–119]. In one study of parenchymal brain Response Measurement

General Considerations Radiation Therapy

Introduction inciting cancer often asymptomatic or occult. The

Encephalitis tosis [43]. Thus, the typical workup includes