(Jurg Hodler) Diseases of The CH - Desconocido PDF
(Jurg Hodler) Diseases of The CH - Desconocido PDF
(Jurg Hodler) Diseases of The CH - Desconocido PDF
Diseases
of the Chest
SYLLABUS and Heart
Courses 2015-2018
Editors
J. Hodler
R.A. Kubik-Huch
G.K. von Schulthess
Ch.L. Zollikofer
123
Diseases of the Chest and Heart 2015-2018
Diagnostic Imaging and Interventional Techniques
J. Hodler R.A. Kubik-Huch G.K. von Schulthess
Ch.L. Zollikofer (Eds)
DISEASES OF THE
CHEST AND HEART
2015-2018
DIAGNOSTIC IMAGING AND INTERVENTIONAL
TECHNIQUES
including the
Nuclear Medicine Satellite Course Diamond
Davos, March 21-22, 2015
Editors
DOI 10.1007/978-88-470-5752-4
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 dis-
similar methodology now known or hereafter developed. Exempted from this legal reservation are
brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for
the purpose of being entered and executed on a computer system, for exclusive use by the purchas-
er of the work. Duplication of this publication or parts thereof is permitted only under the provisions
of the Copyright Law of the Publishers location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the
Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this pub-
lication 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. While the advice and
information in this book are believed to be true and accurate at the date of publication, neither the
authors nor the editors nor the publisher can accept any legal responsibility for any errors or omis-
sions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein
.
IDKD 2015-2018
Preface
Table of Contents
Workshops
Update in the Diagnosis and Staging of Lung Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Jos Vilar, Jeremy J. Erasmus
Missed Lung Lesions: Side by Side Comparison of Chest Radiography with MDCT . . . 80
Nigel Howarth, Denis Tack
CT Evaluation of Chest Pain: Acute Coronary Syndrome and Acute Aortic Syndrome 119
Dominik Fleischmann, Udo Hoffmann
VIII Table of Contents
Magnetic Resonance Imaging of Breast Implants and the Reconstructed Breast . . . 264
Isabelle Thomassin-Naggara, Michael Atlan, Jocelyne Chopier, Emile Darai
IDKD 2015-2018
List of Contributors
Primary lung cancer is the leading cause of cancer mortal- Most patients are in their fifth and sixth decades of life
ity in the world and its incidence is expected to rise in the and approximately three quarters of patients are sympto-
next several decades, especially in more recently industri- matic at presentation [1]. Many patients present with
alized countries such as China. This high mortality is nonspecific systemic manifestations of malignancy, in-
largely explained by the fact that patients with lung cancer cluding anorexia, weight loss, or fatigue. Symptoms also
often present with advanced stage disease. Imaging is im- depend on the local effects of the primary mass, the pres-
portant in the early detection and clinical staging of lung ence of regional or distant metastases, and the coexis-
cancer. Indeed, both the therapeutic options and the man- tence of paraneoplastic syndromes. While solitary pe-
agement of patients with lung cancer are to a considerable ripherally located tumors tend to be asymptomatic, cen-
degree dependent upon disease stage at presentation. De- tral endobronchial tumors can manifest as fever, dyspnea,
tailed knowlegde and the appropriate use of imaging in the hemoptysis, and cough. Symptoms that can occur as a re-
staging evaluation of patients with non-small cell lung can- sult of local growth and the invasion of adjacent nerves,
cer (NSCLC) are required to avoid unnecessary procedures, vessels, and mediastinal structures include chest pain, vo-
excess radiation, and redundant information. This is facil- cal cord paralysis and hoarseness, facial and upper trun-
itated by the use of guidelines as well as the participation cal edema, headaches, neck vein distention, enlarged col-
of multidisciplinary teams in which radiologists, patholo- lateral chest wall vessels (superior vena cava obstruc-
gists, pulmonologists, surgeons, and medical and radiation tion), and dysphagia (esophageal involvement). Clinical
oncologists discuss and reach a consensus on individual- signs and symptoms can also be caused by tumor excre-
ized imaging and treatment. The main objective of this tion of a bioactive substance, or hormone, or as a result
chapter is to review the basic concepts related to the de- of immune-mediated neural tissue destruction caused by
tection, staging, and follow-up of patients with NSCLC. antibody- or cell-mediated immune responses. These
paraneoplastic syndromes occur in 1020% of lung can-
cer patients. Antidiuretic and adrenocorticotropin hor-
Detection of Lung Cancer mones are the more frequently excreted hormones and
can result in hyponatremia and serum hypo-osmolarity
Preclinical Detection and in Cushings syndrome (central obesity, hypertension,
glucose intolerance, plethora, hirsutism), respectively [1].
The most common finding in an asymptomatic patient
with NSCLC is the solitary pulmonary nodule (SPN), de-
fined as a single intraparenchymal lesion 3 cm in di- Role of Imaging in Detecting Lung Cancer
ameter that is not associated with atelectasis or lym-
phadenopathy. When the lesion is 3 cm, it is defined as Imaging has an important role in screening for a lung ma-
a mass. SPNs are detected incidentally on chest radio- lignancy because diagnosis at an early stage, before clin-
graphs or computed tomography (CT) but can also be de- ical presentation, is associated with an improved progno-
tected by screening programs for lung cancer. The chance sis. Lung cancer screening is typically performed with
of a pulmonary nodule being a cancer is directly related low-dose CT (LDCT), as screening with chest radiogra-
to the prevalence of the disease (pretest probability) and phy has been shown to have limited benefit. Randomized
thus is much higher in high-risk groups such as heavy control trials have been or are being conducted to exam-
cigarette smokers. ine the role of screening in lung cancer management. The
National Lung Screening Trial (NLST), a randomized nodules growth can manifest not only as an increase in
study comparing the effectiveness of LDCT vs. chest ra- size but also as an increase in attenuation and/or the de-
diography in more than 50,000 participants, reported sig- velopment or increase in size of a solid component. These
nificant reductions in lung cancer (20%) and all-cause imaging features of growth are suspicious for an in-
mortality (6.7%) [2]. Accordingly, LDCT is currently ad- creased risk of malignancy [7]. The measurement of ser-
vocated as a screening tool for lung cancer. However, op- ial volumes, rather than diameters, together with the com-
timal screening strategies have not been determined. puter-calculated volume doubling time of small nodules
Thus, it is uncertain whether younger patients and/or has been suggested as an accurate and potentially useful
smokers with fewer pack-years of smoking history than method to assess growth [8, 9]. In an analysis of the
stated in the eligibility criteria used for individuals ex- growth rates of stage I lung cancers determined with se-
amined in the NLST (55- to 74-year-old patients with at rial volumetric CT measurements, the median time to a
least 30 pack-years of smoking history and still active volume doubling was 207 days (50% of the tumors dou-
smokers, or former smokers who stopped smoking with- bled in volume at 8 weeks and 75% at 14 weeks) [9]. Re-
in the previous 15 years) will also benefit from screen- cently, nodule mass, defined as the combination of nod-
ing. It is hoped that contentious issues, including over di- ule volume and density, was proposed as a more accurate
agnosis bias, whether detection of pre-clinical disease af- determination of the growth of partly solid nodules [10].
fects survival, and the cost-effectiveness of screening By multiplying nodule volume and density, mass mea-
programs, will be resolved in the coming years [3-5]. surements, which are subject to less variability than vol-
The detection of lung cancer can be incidental, by ume or diameter measurements, may allow earlier detec-
screening, or when the lesion has caused symptoms. Low- tion of the growth of partly solid nodules.
dose CT is more sensitive than chest radiographs for de-
tecting early lung cancer. Morphology
Lung cancers typically have irregular or spiculated mar-
Characterization of a Pulmonary Nodule gins, although this finding can occasionally be seen with
benign nodules as well (Fig. 1). A smooth margin, a fea-
Previous chest radiographs or CT images, if available, ture typical of benign nodules, cannot be used to exclude
may suffice in characterizing a nodule as benign, i.e., if lung cancer because malignant nodules may also have this
there is stability in size for 2 years. However, nodules, appearance. The presence of fat (attenuation 40 to 120
especially subsolid (pure ground glass or partly solid) Hounsfield units, HU) in a pulmonary nodule or mass in-
nodules, can remain stable for 2 years due to their very dicates benignity, usually a hamartoma or, less frequently,
slow growth rates and ultimately prove to be indolent lipoid aspiration pneumonia (Fig. 2). Characteristic pat-
adenocarcinomas. The absence of prior studies or the terns of benign calcification, whether central (involving
presence of signs of radiologic change are indications for 10% of the cross-sectional area of the nodule), diffuse,
CT evaluation to better characterize an SPN. Size, growth or laminated, are usually indicative of prior granuloma-
and morphology are the main parameters of an SPN that tous disease. Popcorn-type calcification is typically due to
should be analyzed with CT. Positron emission tomogra- the chondroid tissue in pulmonary hamartomas but can be
phy (PET) can provide useful information about the
metabolic behavior of the nodule.
Growth
Growth can be estimated by evaluating prior imaging
studies. Lung cancers typically double in volume (a 26%
increase in diameter) between 30 and 400 days (average,
240 days). Very rapid duplication times are generally in-
consistent with malignancy. Rather, volume doubling
times 2030 days are suggestive of an infectious or in-
flammatory etiology but they can also occur with lym-
phoma or rapidly growing metastases. Lung cancers can
have long volume doubling times. In a CT screening
study analyzing the growth rates of small lung cancers,
the volume doubling time ranged from 52 to 1733 days
(mean, 452 days) and approximately 20% of these ma-
lignancies had a volumetric doubling time 2 years [6].
Fig. 1. Poorly differentiated adenocarcinoma. Computed tomogra-
These nodular opacities were typically well-differentiat- phy (CT) shows a nodule with a spiculated margin in the right
ed adenocarcinomas. In contrast to the growth of solid upper lobe. Note that spiculation is typical for primary lung ma-
nodules assessed only on the basis of size, in subsolid lignancy. At resection pleural invasion was present
Update in the Diagnosis and Staging of Lung Cancer 5
a b
Fig. 4 a, b. Adenocarcinoma in situ in an 89-year-old man. CT shows a 3-cm pure ground glass nodu-
lar opacity in the right upper lobe. b Whole-body positron emission tomography (PET) maximum
intensity projection image shows low-grade 18F-2-deoxy-D-glucose (FDG) uptake in the nodule
(arrow). Note that adenocarcinomas manifesting as ground glass nodular opacities are frequently
falsely negative on FDG-PET
Pulmonary nodules can be characterized according to Low-Risk Populations (Little or No History of Smoking, and No
their growth, morphology, and metabolic activity. Stan- Other Risk Factors)
dard imaging protocols based on the pretest probability
(risk) should be used to rule out malignancy. 1. Nodules 4 mm have a very small likelihood of ma-
lignancy such that re-assessment is not necessary.
2. Nodules 4-mm but 6-mm should be re-assessed us-
Guidelines for the Evaluation of an Incidentally ing CT at 12 months; if stable, no further evaluation is re-
Detected Pulmonary Nodule quired. The exception is the non-solid or partially solid
nodule, for which re-assessment may need to be contin-
The Fleischner Society guidelines for the evaluation of ued to exclude the risk of an indolent adenocarcinoma.
an incidentally discovered solid nodule in an adult pa- 3. Nodules 6 mm but 8 mm should be re-assessed us-
tient integrate lesion morphology, growth rate, patient ing CT at 612 months and, if stable, again at 1824
age, and smoking history (see below). In addition, to months.
complement the recommendations for incidentally de- 4. Nodules 8 mm should either be re-assessed using CT
tected solid nodules, the Society recently published rec- at 3, 9, and 24 months, to determine their size stabili-
ommendations specifically aimed at the management of ty, or further evaluated with contrast-enhanced CT, CT-
ground glass and partly solid nodules (see below) [19, PET, biopsy, or resection.
20]. Importantly, for both, risk factors such as smoking
history, family history of lung cancer, or exposure to car- High-Risk Populations (History of Smoking, or Other Exposure
cinogenic agents are not considered in the current guide- or Risk Factor)
lines due to a lack of sufficient data. Other issues to be
aware of are that a slight temporary decrease in size can 1. Nodules 4 mm should be re-assessed at 12 months;
be seen with adenocarcinomas manifesting as ground if stable, no further evaluation is required. The excep-
glass or partly solid nodules, due to fibrosis or atelecta- tion is the non-solid or partially solid nodule, for
sis, and enlargement and/or increasing attenuation with which re-assessment may need to be continued to ex-
or without the new appearance of a solid component dur- clude the risk of an indolent adenocarcinoma.
ing follow up should be managed with a high degree of 2. Nodules 4 mm but 6 mm should be re-assessed us-
suspicion [20]. ing CT at 612 months and, if stable, again at 1824
months.
3. Nodules 6 mm but 8 mm should be re-assessed us-
ing CT at 36 months and, if stable, again at 912
Recommendations for the Management of Incidentally months and 24 months.
Detected Solid Nodules 4. Nodules 8 mm should either be re-assessed using CT
at 3, 9, and 24 months to assess stability or a contrast-
The Fleischner Society guidelines [19] are summarized enhanced CT, CT-PET, biopsy, or resection should be
in the following. performed.
Update in the Diagnosis and Staging of Lung Cancer 7
Recommendations for the Management of Incidentally ing (MRI) and/or PET/CT. A contrast-enhanced CT of the
Detected Solitary Ground Glass and Part Solid Nodules chest is recommended for the evaluation of all patients
with known or suspected primary lung cancer [21]. CT ac-
The Fleischner Society guidelines [20] are summarized curately assesses most characteristics of the primary tu-
in the following. mor (T descriptor), including size and location, but loco-
1. Solitary pure ground glass nodules 5 mm require no regional invasion can be difficult to determine. CT is also
CT follow-up. useful in detecting nodal metastasis and determining the
2. Nodules 5 mm should be initially followed-up at 3 absence or presence of intra- and extrathoracic metastatic
months using CT to confirm persistence, followed by disease, including contralateral lung nodule(s), pleural
annual surveillance CT for a minimum of 3 years. and pericardial nodule(s) and effusions, bone metastases,
3. Solitary, partly solid nodules should be initially fol- and adrenal nodules/masses. CT of the chest alone is suf-
lowed-up at 3 months using CT to confirm persistence. ficient for the staging of patients with pure ground glass
If persistent, with a solid component 5 mm, yearly sur- nodules and an otherwise normal study, and for patients
veillance CT should be performed for a minimum of 3 with peripheral stage IA disease [21]. Otherwise, further
years. If persistent with a solid component 5 mm, then imaging with FDG-PET is recommended for patients eli-
biopsy or surgical resection is recommended. PET/CT gible for curative treatment. FDG-PET/CT has an in-
should be considered for partly solid nodules 10 mm. creasing role in staging, particularly for detecting nodal
(N descriptor) and distant (M descriptor) metastasis.
When PET is unavailable or cannot be performed, a con-
trast-enhanced CT of the abdomen is recommended [21].
Recommendations for the Management of Incidentally
Detected Multiple Ground Glass and Part Solid T Descriptor
Nodules
The T descriptor defines the size, location, and extent of
1. Pure ground glass nodules 5 mm should be fol- the primary tumor and is based on differences in survival
lowed-up using CT at 2 and 4 years. (Fig. 5). However, although a T4 descriptor generally pre-
2. Pure ground glass nodules 5 mm without a dominant cludes resection, tumors with cardiac, tracheal, and ver-
lesion(s) should be initially followed-up using CT at 3 tebral-body invasion are designated in the 7th edition
months to confirm persistence, after which annual sur- staging system as being potentially resectable in the ab-
veillance CT for a minimum of 3 years should be per- sence of N2 and or N3 disease.
formed. The following parameters must be analyzed regarding T
3. A dominant nodule(s) with a partly solid or solid com- descriptor determination: (1) size, (2) location, (3) ex-
ponent should be initially followed-up at 3 months us- tension.
ing CT to confirm persistence. If persistent, biopsy or 1. Tumor size is a determinant of the T descriptor and es-
surgical resection is recommended, especially for le- sential to define surgical and radiotherapy strategies as
sions with a solid component 5 mm. well as to evaluate the response to treatment. In the 7th
edition of the TNM classification, size is a significant
parameter related to survival. T1 is subclassified as T1a
Staging of Lung Cancer (2 cm) or T1b (2 cm to 3 cm); T2 is subclassified
as T2a (3 cm to 5 cm or T2 by other factor and 5
Staging, a standardized anatomic description of disease cm) or T2b (5 cm to 7 cm); tumors 7 cm are clas-
extent, is determined both clinically and pathologically sified as T3. Size is usually measured on CT or MRI.
and guides appropriate treatment. In general, the clinical On CT, lung window settings should be used, as medi-
stage underestimates the extent of disease compared with astinal windowing may underestimate the size.
the pathologic stage. Radiologic imaging is an essential 2. Tumor location is important in that it provides infor-
component of clinical staging and allows the assessment mation to surgeons and radiation oncologists that can
of disease manifestations that are important for surgical, affect therapeutic management. For instance, centrally
oncological, and radiation therapy planning, including located tumors close to the spinal cord impose radia-
size of the primary tumor, its location and relationship to tion dose-volume constraints, and determining tumor
normal anatomic structures in the thorax, and the presence margins is important as they can affect radiotherapy
of nodal and or metastatic disease. The 7th edition of the delivery. This is especially important with the increas-
tumor node metastasis (TNM) staging system allows a ing use of conformal radiation therapy, in which mul-
standardized definition of stage and, consequently, indi- tiple radiation beams are used to generate dose distri-
cates the most appropriate treatment. While useful in as- butions that conform tightly to target volumes. The re-
certaining advanced disease, chest radiography is limited lation of the tumor to a fissure must be reported, as it
in accurately determining TNM descriptors in patients may imply a change in surgical technique (pneu-
with potentially resectable disease, which typically re- monectomy instead of lobectomy) when there is evi-
quires imaging with CT and/or magnetic resonance imag- dence that the lesion crosses the fissure (Fig. 6). The
8 J.Vilar, J.J. Erasmus
proximity to a main pulmonary artery can also involve pulmonary artery involvement may require a pneu-
a change in surgical approach. monectomy rather than a lobectomy in order to obtain
3. Tumor extension can be divided into local and distant. clear surgical margins. Additionally, involvement of the
Local extension is included in the T descriptor, whereas origin of the lobar bronchus or main bronchus may re-
distant metastases relate to the M descriptor. In the quire a sleeve resection or pneumonectomy. Thoracic
TNM staging system, additional pulmonary nodules in wall invasion does not preclude surgery but must be re-
the same lobe are classified as T3, nodule/s in other ip- ported to the surgeon to ensure appropriate surgical re-
silateral lobes are T4 and nodule/s in the contralateral section, typically en bloc resection of the chest wall.
lung are M1a.The determination of the degree of pleur-
al, chest-wall, and mediastinal invasion, as well as central Role of Imaging in T Descriptor Determination
airways, pulmonary veins, and artery involvement is im-
portant not only to radiation oncologists but also to sur- CT and MRI are useful in determining gross chest wall
geons evaluating the tumor for resectability. For instance, or mediastinal invasion (Fig. 7). However, limited loco-
Update in the Diagnosis and Staging of Lung Cancer 9
chemotherapy), while contralateral mediastinal and but their accuracy in detecting metastases to hilar and
scalene or supraclavicular disease (N3) is not. To poten- mediastinal nodes is not optimal because enlarged
tially improve patient management, data are currently be- nodes can be hyperplastic and small nodes can contain
ing collected based on grouping nodal stations together in metastases. A meta-analysis of CT accuracy for nodal
six zones within the current N1 and N2 patient subsets staging in 3438 patients showed a sensitivity of 57%, a
for further evaluation [26]. specificity of 82%, a positive predictive value of 56%,
and a negative predictive value of 83% [27]. MRI has a
Role of Imaging in N Descriptor Determination reported sensitivity of 90.1%, a specificity of 93.1%,
and an accuracy of 92.2% on a per patient basis in de-
In the imaging evaluation of nodal metastasis, size is the tecting nodal metastasis [28, 29]. FDG-PET and FDG-
only diagnostic criterion, with nodes 10 mm in their PET/CT have improved the sensitivity and specificity
short-axis diameter considered abnormal. Chest radio- for detecting mediastinal nodal metastasis compared
graphs are neither sensitive nor specific in evaluating with CT alone (Fig. 10). A meta-analysis showed an
nodal metastases. CT and MRI are better in this regard overall sensitivity of 83% and specificity of 92% for
Update in the Diagnosis and Staging of Lung Cancer 11
a b
c d
Fig. 10 a-d. Nodal metastasis in a 78-year-old man with non-small-cell lung cancer.
a, b CT shows a large mass in the left lower lobe, enlarged subcarinal/azygo-esophageal
(short-axis diameter 1 cm) nodes, and left and right lower paratracheal lymph nodes
(* in b). c Whole-body PET maximum intensity projection image shows increased FDG
uptake within the mass (M) and in the subcarinal/azygo-esophageal nodes (* in c) and
paratracheal nodes (arrows in c). d FDG-PET/CT shows increased FDG uptake in the
paratracheal nodes bilaterally. Mediastinoscopy confirmed nodal metastasis. Because
N3 (contralateral mediastinal) nodal metastasis is nonresectable, the patient was treated
palliatively
FDG-PET in detecting mediastinal nodal metastases when there are no distant metastasis (M0), or to redirect
compared with 59% and 78% on CT [30]. de Langen et nodal sampling by identifying an otherwise undetected
al. showed that nodal sampling may not be necessary for site of metastasis.
nodes measuring 1015-mm on CT, as long as they are
negative on FDG-PET. However, mediastinal lymph The location of nodal metastasis is of major impor-
nodes 16 mm, even if negative on FDG-PET, have a tance in determining management. Ipsilateral hilar (N1)
21% post-test probability of metastatic involvement, ne- nodes are resectable, ipsilateral mediastinal or subcari-
cessitating preoperative pathologic staging [31]. Current nal adenopathy (N2) may be resectable (usually after in-
recommendations for FDG-PET imaging are that it duction chemotherapy); contralateral mediastinal
should be performed in patients with no CT findings of adenopathy and scalene or supraclavicular adenopathy
nodal metastatic disease to corroborate CT findings (N3) are unresectable.
12 J.Vilar, J.J. Erasmus
M Descriptor
Patients with NSCLC commonly have metastases to the
lung, adrenals, liver, brain, bones, and extrathoracic
lymph nodes at presentation. The M1 descriptor describes
these metastases and is divided into two subsets based on
outcome data showing a modest but significant survival
difference [32]. M1a includes nodule(s) in the contralat-
eral lung, pleural effusion and nodule(s), and pericardial
nodule(s) while M1b designates extrathoracic metastasis
(Fig. 5) [32, 33].
Fig. 12 a, b. Squamous cell carcinoma in a 61-year-old woman. a, b CT shows a 1.5-cm nodule in the left lower lobe (arrow in a) and a
large, low-attenuation (10 HU) left adrenal mass (* in b). Note that an attenuation value of 10 HU on non-contrast-enhanced CT is
diagnostic of a benign etiology and no further evaluation is required
a b
a b
12. Kishi K, Homma S, Kurosaki A et al (2004) Small lung tu- mographic imaging in detecting mediastinal lymph node
mors with the size of 1cm or less in diameter: clinical, radio- metastases in nonsmall cell lung cancer. Ann Thorac Surg
logical, and histopathological characteristics. Lung Cancer 79:375-382.
44:43-51. 31. de Langen AJ, Raijmakers P, Riphagen I (2006) The size of
13. Godoy MC, Naidich DP (2009) Subsolid pulmonary nodules mediastinal lymph nodes and its relation with metastatic in-
and the spectrum of peripheral adenocarcinomas of the lung: volvement: a meta-analysis. Eur J Cardiothorac Surg 29:26-29.
recommended interim guidelines for assessment and manage- 32. Postmus PE, Brambilla E, Chansky K et al (2007) The IASLC
ment. Radiology 253:606-622. Lung Cancer Staging Project: proposals for revision of the M
14. Austin JH, Garg K, Aberle D et al (2013) Radiologic implica- descriptors in the forthcoming (seventh) edition of the TNM
tions of the 2011 classification of adenocarcinoma of the lung. classification of lung cancer. J Thorac Oncol 2:686-693.
Radiology 266:62-71. 33. Rami-Porta R, Bolejack V, Goldstraw P (2011) The new tumor,
15. Honda T, Kondo T, Murakami S et al (2013) Radiographic and node, and metastasis staging system. Semin Respir Crit Care
pathological analysis of small lung adenocarcinoma using the Med. 32:44-51.
new IASLC classification. Clin Radiol 68:e21-26. 34. Detterbeck FC, Postmus PE, Tanoue LT (2013) The stage clas-
16. Travis WD, Brambilla E, Noguchi M et al (2011) Internation- sification of lung cancer: Diagnosis and management of lung
al association for the study of lung cancer/american thoracic cancer, 3rd ed: American College of Chest Physicians evi-
society/european respiratory society international multidisci- dence-based clinical practice guidelines. Chest 143(5 Sup-
plinary classification of lung adenocarcinoma. J Thorac Oncol pl):e191S-210S.
6:244-285. 35. Gould MK, Donington J, Lynch WR et al (2013) Evaluation of
17. Hashimoto Y, Tsujikawa T, Kondo C et al (2006) Accuracy of individuals with pulmonary nodules: when is it lung cancer?
PET for diagnosis of solid pulmonary lesions with 18F-FDG Diagnosis and management of lung cancer, 3rd ed: American
uptake below the standardized uptake value of 2.5. J Nucl Med College of Chest Physicians evidence-based clinical practice
47:426-431. guidelines. Chest 143(5 Suppl):e93S-120S.
18. Nomori H, Watanabe K, Ohtsuka T et al (2004) Evaluation of 36. Ohno Y, Koyama H, Onishi Y et al (2008) Non-small cell lung
F-18 fluorodeoxyglucose (FDG) PET scanning for pulmonary cancer: whole-body MR examination for M-stage assess-
nodules less than 3 cm in diameter, with special reference to mentutility for whole-body diffusion-weighted imaging
the CT images. Lung Cancer 45:19-27. compared with integrated FDG PET/CT. Radiology 248:643-
19. MacMahon H, Austin JH, Gamsu G et al (2005) Guidelines for 654.
management of small pulmonary nodules detected on CT 37. Viney RC, Boyer MJ, King MT et al (2004) Randomized con-
scans: a statement from the Fleischner Society. Radiology trolled trial of the role of positron emission tomography in the
237:395-400. management of stage I and II non-small-cell lung cancer. J
20. Naidich DP, Bankier AA, MacMahon H et al (2013) Recom- Clin Oncol 22:2357-2362.
mendations for the management of subsolid pulmonary nod- 38. MacManus MP, Hicks RJ, Matthews JP et al (2001) High rate
ules detected at CT: a statement from the Fleischner Society. of detection of unsuspected distant metastases by PET in ap-
Radiology 266:304-317. parent stage III non-small-cell lung cancer: implications for
21. Silvestri GA, Gonzalez AV, Jantz MA et al (2013) Methods for radical radiation therapy. Int J Radiat Oncol Biol Phys 50:287-
staging non-small cell lung cancer: Diagnosis and manage- 293.
ment of lung cancer, 3rd ed: American College of Chest Physi- 39. van Tinteren H, Hoekstra OS, Smit EF et al (2002) Effective-
cians evidence-based clinical practice guidelines. Chest 143(5 ness of positron emission tomography in the preoperative as-
Suppl):e211S-250S. sessment of patients with suspected non-small-cell lung can-
22. Bruzzi JF, Komaki R, Walsh GL et al (2008) Imaging of non- cer: the PLUS multicentre randomised trial. Lancet 359:1388-
small cell lung cancer of the superior sulcus: part 1: anatomy, 1393.
clinical manifestations, and management. Radiographics 40. Boland GW, Lee MJ, Gazelle GS et al (1998) Characterization
28:551-560; quiz 620. of adrenal masses using unenhanced CT: an analysis of the CT
23. Bruzzi JF, Komaki R, Walsh GL et al (2008) Imaging of non- literature. AJR AM J Roentgenol 171:201-204.
small cell lung cancer of the superior sulcus: part 2: initial 41. Pena CS, Boland GW, Hahn PF et al (2000) Characterization
staging and assessment of resectability and therapeutic re- of indeterminate (lipid-poor) adrenal masses: use of washout
sponse. Radiographics 28:561-572. characteristics at contrast-enhanced CT. Radiology 217:798-
24. Mountain CF, Dresler CM (1997) Regional lymph node clas- 802.
sification for lung cancer staging. Chest 111:1718-1723. 42. Boland GW, Hahn PF, Pena C, Mueller PR (1997) Adrenal
25. Naruke T, Suemasu K, Ishikawa S (1978) Lymph node map- masses: characterization with delayed contrast-enhanced CT.
ping and curability at various levels of metastasis in resected Radiology 202:693-696.
lung cancer. J Thorac Cardiovasc Surg 76:832-839. 43. Boland GW, Lee MJ (1995) Magnetic resonance imaging of
26. Rusch VW, Crowley J, Giroux DJ et al (2007) The IASLC the adrenal gland. Crit Rev Diagn Imaging 36:115-174.
Lung Cancer Staging Project: proposals for the revision of the 44. Outwater EK, Siegelman ES, Huang AB, Birnbaum BA (1996)
N descriptors in the forthcoming seventh edition of the TNM Adrenal masses: correlation between CT attenuation value and
classification for lung cancer. J Thorac Oncol 2:603-612. chemical shift ratio at MR imaging with in-phase and op-
27. Toloza EM, Harpole L, McCrory DC (2003) Noninvasive stag- posed-phase sequences. Radiology 200:749-752.
ing of non-small cell lung cancer: a review of the current evi- 45. Schwartz LH, Ginsberg MS, Burt ME et al (1998) MRI as an
dence. Chest 123(1 Suppl):137S-146S. alternative to CT-guided biopsy of adrenal masses in patients
28. Ohno Y, Hatabu H, Takenaka D et al (2004) Metastases in me- with lung cancer. Ann Thorac Surg 65:193-197.
diastinal and hilar lymph nodes in patients with non-small cell 46. Boland GW, Dwamena BA, Jagtiani Sangwaiya M et al (2011)
lung cancer: quantitative and qualitative assessment with STIR Characterization of adrenal masses by using FDG PET: a sys-
turbo spin-echo MR imaging. Radiology 231:872-879. tematic review and meta-analysis of diagnostic test perfor-
29. Ohno Y, Koyama H, Nogami M et al (2007) STIR turbo SE mance. Radiology 259:117-126.
MR imaging vs. coregistered FDG-PET/CT: quantitative and 47. Kozower BD, Meyers BF, Reed CE (2008) Does positron emis-
qualitative assessment of N-stage in non-small-cell lung can- sion tomography prevent nontherapeutic pulmonary resections
cer patients. J Magn Reson Imaging 26:1071-1080. for clinical stage IA lung cancer? Ann Thorac Surg 85:1166-
30. Birim O, Kappetein AP, Stijnen T, Bogers AJ (2005) Meta- 1169; discussion 1169-1170.
analysis of positron emission tomographic and computed to- 48. Min JW, Um SW, Yim JJ et al (2009) The role of whole-body
16 J.Vilar, J.J. Erasmus
FDG PET/CT, Tc 99m MDP bone scintigraphy, and serum al- phy for diagnosis of bone metastases in patients with lung can-
kaline phosphatase in detecting bone metastasis in patients with cer. Eur J Radiol 81:1007-1015.
newly diagnosed lung cancer. J Korean Med Sci 24:275-280. 50. Liu N, Ma L, Zhou W et al (2010) Bone metastasis in patients
49. Qu X, Huang X, Yan W et al (2012) A meta-analysis of with non-small cell lung cancer: the diagnostic role of F-18
(1)(8)FDG-PET-CT, (1)(8)FDG-PET, MRI and bone scintigra- FDG PET/CT. Eur J Radiol 74:231-235.
IDKD 2015-2018
a b
Fig. 5 a, b. Adenocarcinoma of the lung. Computer aided 3-D volumetric evaluation demonstrating asymmetric growth over an 11-month
period. a Initial evaluation, b follow-up evaluation
situ (AIS), are typically pure ground glass lesions 5 Table 2. Fleischner guidelines for subsolid nodules
mm and 3 cm, respectively, in size. Minimally invasive
Characteristics Imaging/Management
adenocarcinoma (MIA) is typically a partly solid lesion
with an invasive component measuring 5 mm. Larger Solitary pure ground No follow-up
subsolid lesions with more extensive solid invasive tumor glass nodule, 5 mm
are now termed lepidic predominant adenocarcinoma Multiple pure ground 2- and 4-year follow-up
(LPA). Invasive multifocal mucinous adenocarcinoma re- glass nodule, 5 mm
places the entity of multicentric BAC. Solitary pure ground 3 months (then at 1, 2, and 3 years)
Lesions that are exclusively ground glass or have min- glass nodule, 5 mm
imal solid components were observed to progress with a Multiple pure ground 3 months (then at 1, 2, 3 years)
very slow growth rate. In one study, a comparison of glass nodule, one 5 mm
mean volume doubling times in lung cancer patients with Solitary part solid 3 months (then biopsy/resect)
solid, partly solid, or pure ground glass neoplastic lesions 5 mm solid (1, 2, 3 years?)
demonstrated a stepwise increase in the CT-calculated
Multiple part solid 3 months (then biopsy/resect)
volume-doubling times (853 vs. 457 vs. 158 days) [19]. Treat dominant lesion(s)
Therefore, for a subsolid lesion suspected to reflect a
neoplasm in the adenocarcinoma spectrum, the demon-
stration of a 2-year stability may not be enough to ensure
its benign nature. In these cases, an initial follow-up ex- imaging strategies for single and multiple ground glass
amination at 3 months may be performed to determine lesions and indicate potential clinical diagnostic and
whether the lesion has persisted and is inflammatory. Fol- management pathways (Table 2), including more aggres-
low-up may then be performed at longer intervals (12 sive surgical management when a lesion exceeds 1 cm
years). In these irregular, ill-defined lesions, it can be dif- overall or when there is a significant increase in its solid
ficult to measure incremental size; thus, evaluation component. Conversely, according to the recent guide-
should include comparison of comparable contiguous lines, pure ground glass lesions less than 5 mm in size do
thin-section images and a determination of whether the not require follow-up as these almost always reflect be-
ground glass or solid component geographically extends nign entities or atypical adenomatous hyperplasia. The
around more vessels or airways. In this setting, CAD may management of patients with multiple lesions of varying
be of assistance. Although the technology is evolving to density is more complex, especially as radiologically it is
evaluate ground glass lesions, it remains less reliable than not always possible to determine the extent of neoplasia
in solid lesions. It is also important to note that the nat- in multiple ground glass opacities. Current treatment
ural evolution of these lesions also includes transient re- strategies may include resection of the initial lesions fol-
duction in size that may occur when there is histological lowed by localized resection of progressive or recurrent
alveolar collapse. lesions, although there are no standardized long-term da-
The definitive determination that a subsolid lesion is ta to support this approach.
malignant can be problematic. A progressive increment in
the overall size or in the size of the solid component
strongly favors malignancy. PET/CT imaging in these Focal Parenchymal Airspace Disease
cases may be non-contributory because these lesions of-
ten have low SUVmax activity. Additionally, the evalua- A multitude of etiologies underlie the development of fo-
tion of predominantly ground glass lesions by CT-guided cal air-space opacities. The majority of these will be mul-
fine-needle aspiration or biopsy should be considered tifocal and infective in nature. A distinction of pre-
with great caution, as the results are vulnerable to signif- disposing etiologies is impossible unless there are
icant sampling-error effects and likely to vary signifi- uncommon specific features, e.g., the ground-glass halo
cantly from the ultimate excision specimen. The extent of typical of invasive aspergillosis. The differential diagno-
invasive adenocarcinoma in particular can be significant- sis may be narrowed by referring to a combination of ra-
ly underestimated by percutaneous needle sampling, even diological features, chronicity, progression, response to
if directed to the more solid components. treatment, and the immune status of the patient.
There is no consensus on the optimal management of The presence of cavitation in airspace disease can be
patients with single subsolid lesions. Patients with lesions helpful, although the differential can again be broad, in-
characterized by a higher ground glass percentage com- corporating staphylococcal or gram-negative bacterial in-
position are less likely to have nodal disease and have a fections, mycobacterial disease, or fungal infection.
better prognosis than patients with nodules having a Rounded pneumonia is more common in children but it
greater solid component, in which invasive adenocarci- also occurs in adults, usually due to S. pneumoniae in-
noma likely predominates [20, 21]. Expanding on the in- fection. Appearances are mass-like, without air-bron-
terim guideline published in 2009, the Fleischner Soci- chograms resulting from disease propagation through the
etys definitive guidelines, published in 2013, have collateral air-drift mechanisms of the pores of Kohn and
helped to homogenize practice [22, 23]. They address the canals of Lambert. Focal consolidative opacity that is
22 I. Vlahos, G.F. Abbott
chronic may be characterized by minor volume loss and 10. Jennings SG, Winer-Muram HT, Tarver RD, Farber MO (2004)
tractional bronchiectasis; these features are more com- Lung tumor growth: assessment with CTcomparison of di-
ameter and cross-sectional area with volume measurements.
mon in patients with chronic eosinophilic lung disease or Radiology 231:866-871.
organizing pneumonia. While organizing pneumonia may 11. Jennings SG, Winer-Muram HT, Tann M et al (2006) Distrib-
also present with a reverse halo appearance, central ution of stage I lung cancer growth rates determined with ser-
ground glass opacities with peripheral consolidation, this ial volumetric CT measurements. Radiology 241:554-563.
appearance is not specific for the disease. Progressive fo- 12. Revel MP, Merlin A, Peyrard S et al (2006) Software volumet-
ric evaluation of doubling times for differentiating benign ver-
cal airspace disease may reflect neoplastic disease, in- sus malignant pulmonary nodules. AJR Am J Roentgenol
cluding bronchoalveolar cell carcinoma, lymphoma, or, 187:135-142.
occasionally, mucinous metastases from gastrointestinal 13. Honda O, Kawai M, Gyobu, T et al (2009) Reproducibility of
primary tumors. Calcification may be present in myco- temporal volume change in CT of lung cancer: comparison of
bacterial disease and in some cases of amyloidosis. computer software and manual assessment. Br J Radiol
82:742-747.
14. Bolte H, Riedel C, Muller-Hulsbeck S et al (2007) Precision of
computer-aided volumetry of artificial small solid pulmonary
References nodules in ex vivo porcine lungs. Br J Radiol 80:414-421.
15. Henschke CI, Yankelevitz DF, Mirtcheva R et al (2002) CT
1. Woodring JH, Fried AM, Chuang VP (1980) Solitary cavities screening for lung cancer: frequency and significance of part-sol-
of the lung: diagnostic implications of cavity wall thickness. id and nonsolid nodules. AJR Am J Roentgenol 178:1053-1057.
AJR Am J Roentgenol 135:1269-1271. 16. Austin JHM, Garg K, Aberle D et al (2013) Radiologic impli-
2. Honda O, Tsubamoto M, Inoue A et al (2007) Pulmonary cav- cations of the 2011 classification of adenocarcinoma of the
itary nodules on computed tomography: differentiation of ma- lung. Radiology 266:62-71.
lignancy and benignancy. J Comput Assist Tomogr 31:943- 17. Aoki T, Nakata H, Watanabe H et al (2000) Evolution of pe-
949. ripheral lung adenocarcinomas: CT findings correlated with
3. Swensen SJ, Viggiano RW, Midthun DE et al (2000) Lung nod- histology and tumor doubling time. AJR Am J Roentgenol
ule enhancement at CT: multicenter study. Radiology 214: 174:763-768.
73-80. 18. Noguchi M, Morikawa A, Kawasaki M et al (1995) Small ade-
4. Chae EJ, Song JW, Seo JB et al (2008) Clinical utility of dual- nocarcinoma of the lung. Histologic characteristics and prog-
energy CT in the evaluation of solitary pulmonary nodules: nosis. Cancer 75:2844-2852.
initial experience. Radiology 249:671-681. 19. Hasegawa M, Sone S, Takashima S et al (2000) Growth rate of
5. Swensen SJ, Jett JR, Hartman TE et al (2005) CT screening for small lung cancers detected on mass CT screening. Br J Radi-
lung cancer: five-year prospective experience. Radiology ol 73:1252-1259.
235:259-265. 20. Kodama K, Higashiyama M, Yokouchi H et al (2001) Prog-
6. Piyavisetpat N, Aquino SL, Hahn PF et al (2005) Small inci- nostic value of ground-glass opacity found in small lung ade-
dental pulmonary nodules: how useful is short-term interval nocarcinoma on high-resolution CT scanning. Lung Cancer
CT follow-up? J Thorac Imaging 20:5-9. 33:17-25.
7. MacMahon H, Austin JH, Gamsu G et al (2005) Guidelines for 21. Higashiyama M, Kodama K, Yokouchi H et al (1999) Prog-
management of small pulmonary nodules detected on CT nostic value of bronchiolo-alveolar carcinoma component of
scans: a statement from the Fleischner Society. Radiology small lung adenocarcinoma. Ann Thorac Surg 68:2069-2073.
237:395-400. 22. Godoy MC, Naidich DP (2009) Subsolid pulmonary nodules
8. Li F, Aoyama M, Shiraishi J et al (2004) Radiologists perfor- and the spectrum of peripheral adenocarcinomas of the lung:
mance for differentiating benign from malignant lung nodules recommended interim guidelines for assessment and manage-
on high-resolution CT using computer-estimated likelihood of ment. Radiology 253:606-622.
malignancy. AJR Am J Roentgenol 183:1209-1215. 23. Naidich DP, Bankier AA, MacMahon H et al (2013) Recom-
9. Yankelevitz DF, Reeves AP, Kostis WJ et al (2000) Small pul- mendations for the management of subsolid pulmonary lesions
monary nodules: volumetrically determined growth rates detected at CT: a statement from the Fleischner Society. Radi-
based on CT evaluation. Radiology 217:251-256. ology 266:304-317.
IDKD 2015-2018
a b
a b
a b
Fig. 3 a, b. Asymptomatic bronchogenic cyst. The cyst was discovered on the chest radiograph (a, arrow) and confirmed with CT (b, arrow).
Note the high attenuation, similar to that of the soft tissue
a b
Fig. 4 a, b. Large endothoracic goiter. There is with marked tracheal displacement in the chest radiograph (a). Axial CT shows marked
contrast enhancement of the thyroid tissue (b, arrow)
(either bronchogenic or esophageal). These cysts occa- Unlike the anterior mediastinum, fat in a middle me-
sionally are higher in attenuation as a result of infection diastinal lesion cannot be considered benign. Although
or hemorrhage (Fig. 3) and may even contain a fluid- esophageal or tracheal lipomas and esophageal fibrovas-
calcium level from the milk of calcium. The risk of malig- cular polyps contain fat, so may mediastinal liposarco-
nancy in these conditions tends to be low. As with ante- mas. These rare lesions may insinuate through the medi-
rior mediastinal lesions, the ratio of soft tissue to fluid astinum and often have a predilection for the middle me-
needs to be considered. A foregut duplication cyst should diastinum.
have no soft tissue, enhancing element. If soft tissue is Hypervascular lesions in the middle mediastinum are
encountered, one must consider a potentially more sig- most often hypervascular lymph nodes and intrathoracic
nificant process, usually low-attenuating lymphadenopa- extension of a goiter (Fig. 4). These hypervascular nodes
thy. Such low-attenuating lymph nodes may be encoun- (defined as higher in attenuation than skeletal muscle)
tered in lung cancer, mucinous neoplasms, and mycobac- may be seen with melanoma, plasmacytoma, Castlemans
terial disease. disease, Kaposi sarcoma, sarcomas, and thyroid and renal
26 S. Bhalla, J. Caceres
a b
Fig. 5 a, b. Ganglioneuroma. The lesion, visible as a mass in the right apex (a), was confirmed with enhanced CT (b, arrow). Note the low
CT density, which may create confusion with a cystic mass
cell cancers. When the high attenuating structure is tubu- metastases are often considered, one cannot forget about
lar, a vessel must be considered. Aortic arch anomalies diskitis/osteomyelitis. This latter condition can present as
and azygos vein enlargement often present as a middle insidious back pain and can easily be overlooked.
mediastinal mass on radiography. As with the other compartments, attenuation or inten-
Most mediastinal lymphadenopathy will present in the sity can be helpful. On CT, a potential pitfall is that
middle mediastinum. Occasionally, these nodes will be myelin-rich neurogenic lesions may look cystic (Fig. 5).
calcified. These calcified nodes are usually indicative of For this reason, with posterior mediastinal lesions we of-
an old granulomatous process, such as healed tuberculo- ten rely on MR imaging. True posterior mediastinal cys-
sis or histoplasmosis or sarcoidosis, but care must be tak- tic lesions are rare. Although neuroenteric cysts exist,
en to remember that certain tumors also tend to present they are often associated with vertebral anomalies and
with calcified mediastinal lymph nodes, including ovari- rarely encountered de novo in adults. Instead, a cystic le-
an serous adenocarcinomas, mucinous colon neoplasms, sion in the posterior mediastinum is much more likely to
and osteosarcomas. represent a lateral meningocele or post-traumatic nerve
Another potential for a perceived middle mediastinal root avulsion.
mass on radiography will be a dilated esophagus. Al- Fatty lesions are unusual in the posterior mediastinum
though a distal mass may also result in esophageal di- but when encountered may invoke extramedullary
latation, it is usually only achalasia that results in hematopoiesis. While rare, in patients with anemia, ex-
esophageal widening that can be seen on a chest radio- tramedullary hematopoiesis may develop in the posterior
graph. mediastinum. The etiology of this condition remains un-
known. Some authors have postulated that it develops
from extruded marrow while others have suggested that
Posterior Mediastinum it develops from totipotent cells in the paravertebral
space. When the patient is anemic, extramedullary
A vast majority of posterior mediastinal masses will be hematopoiesis will present with bilateral masses that en-
neurogenic in origin. In adults, these tend to be benign hance similar to spleen without a connecting bridge. As
nerve sheath tumors, usually schwannomas and neurofi- the patient returns to normal hematocrit, the yellow mar-
bromas. In children and younger adults, they tend to be row will take over. The net effect is bilateral posterior me-
sympathetic ganglion in origin, such as ganglioneuro- diastinal fatty masses. In the elderly, Bochdalek hernias
blastoma, neuroblastoma, or ganglioneuroma. The key in should be included in the differential diagnosis (Fig. 6).
separating the two groups is to assess the overall shape, Hypervascular lesions in the posterior mediastinum
comparing the z-axis to the xy-axis. Nerve sheath tumors are less helpful than with the other compartments. Most
tend to be spherical (equal in all 3 axes), while the gan- often these are related to an aneurysmal aorta or enlarged
glion lesions are longer in the z-axis and are more cylin- collateral vessels as with aortic coarctation. As described
drical. Osseous lesions represent the second most com- above, extramedullary hematopoiesis may be seen with
mon group of posterior mediastinal disease. Although bilateral hypervascular paravertebral masses.
Approach to Imaging of Mediastinal Conditions in the Adult 27
a b
Conditions That Disregard the Compartment Model Of course, lung cancer may present with metastases to
any compartment and, unfortunately, tends to metastasize
Certain conditions tend to disregard the compartment to more than one region.
model of the mediastinum. Even with these lesions, un-
derstanding the attenuation or intensity can be helpful.
These include infection and hematoma, which will result Conclusion
in fat stranding and soft tissue attenuation throughout the
mediastinum, often in more than one compartment. The mediastinum represents a space that may be impact-
Lymphangiomas and hemangiomas also tend to disre- ed by a large number of lesions. Having an approach
gard the compartment model. The former tend to be flu- based on location and characterization will allow the ra-
id in their attenuation and insinuate throughout, while the diologist the ability to create a useful, targeted differen-
latter will be higher in attenuation. tial diagnosis (Table 1).
28 S. Bhalla, J. Caceres
Suggested Reading Moran CA, Suster S (1997) Primary germ cell tumors of the me-
diastinum: I. Analysis of 322 cases with special emphasis on
Adam A, Hochholzer L (1981) Ganglioneuroblastoma of the pos- teratomatous lesions and a proposal for histopathologic classi-
terior mediastinum: a clinicopathologic review of 80 cases. fication and clinical staging. Cancer 80:681-690.
Cancer 47:373-381. Nakata H, Nakayama C, Kimoto T et al (1982) Computed tomo-
Baron RL, Sagel SS, Baglan RJ (1981) Thymic cysts following ra- graphy of mediastinal bronchogenic cysts. J Comput Assist
diation therapy for Hodgkin disease. Radiology 141:593-597. Tomogr 6:733-738.
Chen JL, Weisbrod GL, Herman SJ (1988) Computed tomography Pombo F, Rodriguez E, Mato J et al (1992) Patterns of contrast en-
and pathologic correlations of thymic lesions. J Thorac Imag- hancement of tuberculous lymph nodes demonstrated by com-
ing 3:61-65. puted tomography. Clin Radiol 46:13-17.
Cohen AJ, Thompson LN, Edwards FH et al (1991) Primary cysts Rosado-de-Christenson ML, Galobardes J, Moran CA (1992) Thy-
and tumours of the mediastinum. Ann Thorac Surg 51:378-386. moma: radiologic-pathologic correlation. RadioGraphics 12:
Do YS, Im JG, Lee BH et al (1995) CT findings in malignant tu- 151-168.
mors of thymic epithelium. J Comput Assist Tomogr 19:192- Rosado-de-Christenson ML, Pugatch RD, Moran CA, Galobardes
197. J (1994) Thymolipoma: analysis of 27 cases. Radiology
Erasmus JJ, McAdams HP, Donnelly LF, Spritzer CE (2000) MR 193:121-126.
imaging of mediastinal masses. Magn Reson Imaging Clin Rosado-de-Christenson ML, Templeton PA, Moran CA (1992)
North Am 8:59-89. From the archives of the AFIP. Mediastinal germ cell tumors:
Faul JL, Berry GJ, Colby TV et al (2000) Thoracic lymphan- radiologic and pathologic correlation. RadioGraphics 12:1013-
giomas, lymphangiectasis, lymphangiomatosis, and lymphatic 1030.
dysplasia syndrome. Am J Respir Crit Care Med 161:1037- Rossi SE, McAdams HP, Rosado-de-Christenson ML et al (2001)
1046. Fibrosing mediastinitis. RadioGraphics 21:737-757.
Gaerte SC, Meyer CA, Winer-Muram HT et al (2002) Fat-contain- Shaffer K, Rosado-de-Christenson ML, Patz EF Jr et al (1994) Tho-
ing lesions of the chest. RadioGraphics 22:615-678. racic lymphangioma in adults: CT and MR imaging features.
Hoffman OA, Gillespie DJ, Aughenbaugh GL, Brown LR (1993) AJR Am J Roentgenol 162:283-289.
Primary mediastinal neoplasms (other than thymoma). Mayo Spizarny DL, Rebner M, Gross BH (1987) CT evaluation of en-
Clin Proc 68:880-891. hancing mediastinal masses. J Comput Assist Tomogr 11:990-
Jeung M-Y, Gasser B, Gangi A et al (2002) Imaging of cystic masses 993.
of the mediastinum. RadioGraphics 22:S79-S93. Strollo DC, Rosado-de-Christenson ML (1999) Tumors of the
Jolles H, Henry DA, Roberson JP et al (1996) Mediastinitis thymus. J Thorac Imaging 14:152-171.
following median sternotomy: CT findings. Radiology 201: Strollo DC, Rosado de Christenson ML, Jett JR (1997) Primary
463-466. mediastinal tumors. Part 1: tumours of the anterior medi-
Knapp RH, Hurt RD, Payne WS et al (1985) Malignant germ cell astinum. Chest 112:511-522.
tumors of the mediastinum. J Thorac Cardiovasc Surg 89: Strollo DC, Rosado-de-Christenson ML, Jett JR (1997) Primary me-
82-89. diastinal tumors: part II. Tumours of the middle and posterior
Long JA Jr, Doppman JL, Nienhuis AW (1980) Computed tomo- mediastinum. Chest 112:1344-1357.
graphic studies of thoracic extramedullary hematopoiesis. J Suwatanapongched T, Gierada DS (2006) CT of thoracic lymph
Comput Assist Tomogr 4:67-70. nodes. Part II: diseases and pitfalls. Br J Radiol 79:999-1006.
Marano R, Liguori C, Savino G et al (2011) Cardiac silhouette Takahashi K, Al-Janabi NJ (2010) Computed tomography and
findings and mediastinal lines and stripes radiograph and CT magnetic resonance imaging of mediastinal tumors. J Magn
scan correlation. Chest 139:1186-1196. Reson Imaging 32:1325-1339.
McAdams HP, Rosado-de-Christenson M, Fishback NF, Templeton Whitten CR, Khan S, Munneke GJ, Grubnic S (2007) A diagnos-
PA (1998) Castleman disease of the thorax: radiologic features tic approach to mediastinal abnormalities. RadioGraphics
with clinical and histopathologic correlation. Radiology 27:657-671.
209:221-228. Woodring JH, Loh FK, Kryscio RJ (1984) Mediastinal hemor-
Miles J, Pennybacker J, Sheldon P (1969) Intrathoracic meningo- rhage: an evaluation of radiographic manifestations. Radiology
cele. Its development and association with neurofibromatosis. 151:15-21.
J Neurol Neurosurg Psychiatry 32:99-110. Zylak CJ, Eyler WR, Spizarny DL, Stone CH (2002) Develop-
Miyake H, Shiga M, Takaki H et al (1996) Mediastinal lymphan- mental lung anomalies in the adult: radiologic-pathologic cor-
giomas in adults: CT findings. J Thorac Imaging 11:83-85. relation. RadioGraphics 22:S25-S43.
Moeller KH, Rosado-de-Christenson ML, Templeton PA (1997) Zylak CJ, Pallie W, Jackson R (1982) Correlative anatomy and
Mediastinal mature teratoma: imaging features. AJR Am J computed tomography: a module on the mediastinum. Radio-
Roentgenol 169:985-990. Graphics 2:555-592.
IDKD 2015-2018
helpful to display airway abnormalities. Minimum inten- Marked homogeneous early contrast enhancement of an
sity projections can highlight airway dilation, diverticu- endobronchial nodule reflects the high vascularity of this
la, and fistulas, as well as the extent of expiratory air tumor.
trapping and emphysema. Maximum intensity projec- Mucoepidermoid carcinoma is a rare tumor that origi-
tions accentuate foci of mucoid impaction and small cen- nates from the minor salivary glands lining the tracheo-
trilobular nodules and tree-in-bud opacities. bronchial tree. It occurs in young patients (40 years-
Volume-rendering techniques (CT bronchography) old) and mainly involves the segmental bronchi, resulting
consist of segmentation of the lumen-wall interface of in airway obstruction. The typical CT appearance is a
the airways. This technique has proven to be of particu- smooth, ovoid, or lobulated endobronchial mass with oc-
lar interest in diagnosing mild changes in airway caliber casional punctuate calcifications and variable contrast
and understanding complex tracheobronchial abnormal- enhancement.
ities. Virtual bronchoscopy provides an internal render- Lymphoma of the trachea is rare and is usually related
ing of the tracheobronchial inner surface. It is used to to mucosa-associated lymphoid tissue. The CT appear-
detect mucosal nodularity indicative of granulomatous ance is non-specific.
or tumoral lesions in proximal airways.
Secondary Tracheobronchial Malignancy
Tracheobronchial Tumors Direct invasion of the central airways by neoplasms of
the thyroid, esophagus, lung, and larynx is much more
MDCT plays a key role in depicting tumors of the tra- common than hematogeneous metastases. CT demon-
cheobronchial tree [3-5] and in assessing their extent strates the primary neoplasm and its extension by conti-
within the lumen, airway wall, and surrounding struc- guity within the main airways (endoluminal mass, de-
tures before treatment planning. Airway lesions 5 mm struction of cartilage, and tracheo-bronchial-esophageal
are usually detected on CT because of the natural con- fistula).
trast between luminal air and soft-tissue attenuation of Many cancers have the potential to metastasize to the
the lesions. trachea and bronchi. Endotracheal or endobronchial
metastases appear as endotracheal nodules or an eccen-
Primary Malignant Tumors tric thickening of the airway wall or as soft-tissue atten-
uation with contrast enhancement.
Compared to laryngeal or bronchial cancers, primary ma-
lignant tumors are rare in the trachea. The most frequent Benign Tracheobronchial Neoplasms
histological types are squamous cell carcinoma and ade-
noid cystic carcinoma, accounting for nearly 80% of all Benign tracheal neoplasms are rare, accounting for 2%
tracheal neoplasms. of lung neoplasms. On CT, they present as endoluminal
Adenoid cystic carcinoma (sialadenoid tumor) is a masses confined within the tracheobronchial lumen with-
low-grade malignancy that is not associated with ciga- out evidence of involvement of surrounding structures.
rette smoking. It occurs in patients in their forties with- Benign neoplasms are typically sharply marginated,
out any sex predilection. In the central airways, adenoid round, and 2 cm in diameter. Because they originate in
cystic carcinoma has a propensity to infiltrate the wall of the mucosa or submucosa, the overlying epithelium is
the airways, with submucosal extension manifesting as a usually intact, resulting in a smooth appearance of the tu-
sessile, polypoid, annular, or diffuse infiltration. The in- mor surface in the airway lumen.
ner surface is smooth and regular. Extraluminal growth, Histology includes mainly hamartomas, lipomas,
visible on CT scan, is a common feature. leiomyomas, fibromas, chondromas, and schwannomas.
Squamous cell carcinoma is the most common prima- Endobronchial harmartomas account for 30% of
ry malignancy of the trachea and primarily occurs in old- intrathoracic hamartomas. The presence of fat with or
er men with a history of cigarette smoking. On CT, the without calcification is diagnostic.
tumor appears as a polypoid intraluminal mass or as an Respiratory papillomatosis (laryngotracheobronchial
eccentric irregular wall thickening with an irregular sur- papillomatosis) is a neoplastic disease caused by human
face. The tumor has a tendency to spread to adjacent me- papillomavirus infection transmitted from mother to
diastinal lymph nodes and to directly invade the medi- child at birth or acquired from orogenital contact. Pa-
astinum. pillomas arise in the larynx and involve the trachea and
Carcinoid tumor is a low-grade malignant neuroen- proximal bronchi in up to 50% of cases. Involvement of
docrine neoplasm representing 12% of primary lung tu- the lungs is rare, occurring in 1% of patients, and
mors. It appears on CT scan as a well-circumscribed manifests on imaging studies as multiple lung nodules
polypoid mass that protrudes into the airway lumen, usu- and cavitary and cystic lesions in the parenchyma. Ma-
ally located in a main or lobar bronchus. Segmental or lignant transformation into squamous cell carcinoma is
lobar atelectasis and obstructive pneumonitis, as well as a rare but serious complication of respiratory papillo-
foci of calcification, are present in 30% of cases. matosis.
Current Approaches to Chronic and Acute Airway Diseases 31
Non-neoplastic Tracheobronchial Disorders [3, 6, 7] become thickened and calcified. Cartilaginous erosions
and ulcerations also may be seen. Stenosis may be pre-
Post-traumatic Stenoses sent in any main, lobar, or segmental bronchus. Develop-
ing nodular or polypoid lesions can protrude into the air-
Post-traumatic strictures of the trachea are usually the way lumen.
result of ischemic injury from a cuffed endotracheal or
tracheostomy tube or extrinsic neck trauma. These in- Relapsing Polychondritis
juries initially heal by the formation of granulation tis-
sue, with subsequent scarring characterized by dense Relapsing polychondritis is a rare systemic autoimmune
mucosal and submucosal fibrosis associated with distor- disease that affects cartilage at various sites, including
tion of cartilage plates. The two principal sites of steno- the ears, nose, joints, and tracheobronchial tree. Histo-
sis following intubation or tracheostomy tube are at the logically, the acute inflammatory infiltrate in both carti-
stoma or at the level of the endotracheal or tracheosto- lage and perichondrial tissue induces progressive disso-
my tube balloon. CT with multiplanar reformations lution and fragmentation of the cartilage followed by fi-
clearly depicts the severity and length of the stricture. brosis. Symmetric subglottic stenosis is the most frequent
Post-intubation stenosis extends for several centimeters manifestation in the chest. As the disease progresses, the
and typically involves the trachea above the level of the distal trachea and bronchi may be involved. CT scans
thoracic inlet. Post-tracheostomy stenosis typically be- show a smooth thickening of the airway wall associated
gins 1.01.5 cm distal to the inferior margin of the tra- with diffuse narrowing. In the early stages of the disease,
cheostomy stoma and extends over 1.52.5 cm. These the posterior wall of the trachea is spared but in advanced
strictures typically have an hourglass configuration with disease, circumferential wall thickening can occur. Tra-
a thickened tracheal wall. Less commonly, tracheal or cheobronchomalacia can develop as a result of weaken-
bronchial stenosis may present as a thin membrane or ing of cartilage, resulting in considerable luminal col-
granulation tissue protruding into the airway lumen. In lapse on expiration. Gross destruction of the cartilagi-
select cases, the degree of stenosis may also be assessed nous rings with fibrosis may cause stenosis.
by virtual bronchoscopy.
Amyloidosis
Infections
Amyloid deposition in the trachea and bronchi may occur
A number of infections, both acute but more often in association with systemic amyloidosis or as an isolat-
chronic, may affect the trachea and proximal bronchi, re- ed process. The deposits can present as single, multifocal,
sulting in both focal and diffuse airway disease. Subse- or diffuse submucosal plaques or masses. The overlying
quent fibrosis can lead to localized airway narrowing. mucosa is usually intact. Dystrophic calcification or os-
The most common causes of infectious tracheobronchitis sification is frequently present. CT scans show focal or,
are acute bacterial tracheitis in immunocompromised pa- more commonly, diffuse thickening of the airway wall
tients, tuberculosis, rhinoscleroma (Klebsiella rhinoscle- and narrowing of the lumen. Calcification may be pre-
romatis), and necrotizing invasive aspergillosis. CT clear- sent. Narrowing of the proximal bronchi can lead to dis-
ly demonstrates the extent of irregular and circumferen- tal atelectasis, bronchiectasis, or both, with or without
tial tracheobronchial narrowing. In some patients an ac- obstructive pneumonia.
companying mediastinitis is evident, manifesting as infil-
tration of mediastinal fat. With active infection, the nar- Tracheobronchopathia Osteochondroplastica
rowed trachea and frequently one or more main bronchi
have an irregularly thickened wall. In the fibrotic or This rare disorder is characterized by the presence of
healed phase, the airway is narrowed but the wall is multiple cartilaginous nodules and bony submucosal
smooth and of normal thickness. Occasionally, because nodules on the luminal surface of the trachea and proxi-
of the presence of chronic fibrous or granulomatous hili- mal airways. Tracheobronchopathia osteochondroplastica
tis/mediastinitis, tuberculosis of the trachea and/or prox- involves males more frequently than females, and most
imal bronchi may mimic airway malignancy on CT. patients are older than 50 years. Histologically, the nod-
ules contain heterotopic bone, cartilage, and calcified
Granulomatosis with Polyangiitis (Wegener Granulomatosis) acellular protein matrix. The overlying bronchial mucosa
is normal, and, because it contains no cartilage, the pos-
Involvement of the large airways is a common manifes- terior wall of the trachea is spared. On CT, tracheal car-
tation of granulomatosis with polyangiitis. Inflammatory tilages are thickened and show irregular calcifications.
lesions may be present with or without subglottic or The nodules may protrude from the anterior and lateral
bronchial stenosis, ulcerations, and pseudotumors. Radio- walls into the lumen; they usually show foci of calcifica-
logic manifestations include thickening of the subglottic tion. The majority of patients are asymptomatic, with on-
and proximal trachea, with a smooth symmetric or asym- ly a small number developing obstructive signs and
metric narrowing over variable length. Tracheal rings can symptoms.
32 P.A. Grenier, J.P. Kanne
shaped, reflecting abnormally dilated bronchioles with Table 7. Causes of and association with obliterative (constrictive)
thickened walls and mucus or exudate filling the lumens. bronchiolitis
Associated peribronchiolar inflammation is often present, Post-infection
contributing to the CT appearance. Tree-in-bud opacities Childhood viral infection (adenovirus, respiratory syncytial
are characteristic of acute or chronic infectious bronchi- virus, influenza, parainfluenza)
Adulthood and childhood (Mycoplasma pneumoniae, Pneu-
olitis. They also occur in diffuse panbronchiolitis and as- mocystis jirovecii in AIDS patients, endobronchial spread of
piration. tuberculosis, bacterial bronchiolar infection)
Post-inhalation (toxic fumes and gases)
Diffuse aspiration bronchiolitis (chronic occult aspiration in the
Poorly Defined Centrilobular Nodules elderly, patients with dysphagia)
Connective tissue disorders (rheumatoid arthritis, Sjgren syn-
Poorly defined centrilobular nodules reflect the presence drome)
of peribronchiolar inflammation in the absence of airway Allograft recipients (bone-marrow transplant, heart-lung, or
filling. When the distribution of nodules is diffuse and lung transplant)
homogeneous, the pattern is suggestive of bronchiolar or Drugs (penicillamine, lomustine)
vascular diseases. Bronchiolar diseases associated with Ulcerative colitis
Other conditions
poorly defined centrilobular nodules include respiratory Bronchiectasis
bronchiolitis, bronchiolitis associated with hypersensitiv- Cystic fibrosis
ity pneumonitis, and follicular bronchiolitis. Hypersensitivity pneumonitis
Diffuse idiopathic pulmonary neuroendocrine cell hyperpla-
sia (DIPNECH)
Decreased Lung Attenuation and Mosaic Perfusion Excessive Sauropus androgynus (katuk, sweet leaf, or star
gooseberry) ingestion
Areas of decreased lung attenuation associated with de- Idiopathic
creased vessel caliber on CT reflect bronchiolar obstruc-
tion and associated reflex vasoconstriction. In acute bron-
chiolar obstruction, decreased perfusion represents a phys-
iologic reflex of hypoxic vasoconstriction whereas irre- minal bronchioles, has a variety of causes and, rarely, is
versible vascular remodeling occurs with chronic bronchi- idiopathic (Table 7). Bronchial wall thickening and
olar obstruction. Areas of decreased lung attenuation re- bronchiectasis, both central and peripheral, are also com-
lated to hypoperfusion can be patchy or widespread. They monly present.
are poorly defined or sharply demarcated, or have a geo- Mosaic attenuation may also be seen in patients with
graphical outline, but in all cases represent a collection of chronic thromboembolic disease; however, in this condi-
affected pulmonary lobules. Redistribution of blood flow tion, frank dilation of the proximal pulmonary arteries
to the normally ventilated areas causes localized increased within the hyperattenuated areas and extensive areas of
attenuation. The patchwork of abnormal areas of low at- hypoattenuation and hypoperfusion are almost always
tenuation and normal lung or less diseased areas results in present.
mosaic attenuation on CT, also termed mosaic perfusion.
Expiratory CT accentuates the pattern of mosaic attenua- Expiratory Air Trapping
tion where areas of air trapping remain low in attenuation
whereas normal lung increases in attenuation. Usually, the Lobular areas of air trapping may be readily apparent on
regional heterogeneity of lung attenuation is apparent on expiratory CT but are occult on inspiratory CT. Foci of
full inspiration on thin-section CT. However, with more lobular air trapping are usually well demarcated, reflecting
extensive air trapping, the lack of regional homogeneity of the geometry of individual or joined lobules. Lobular ar-
lung attenuation can be challenging to detect on inspirato- eas of air trapping may be present on expiratory CT scans
ry scans; as a result, mosaic attenuation becomes visible of normal individuals, especially in the medial and poste-
only on expiratory scans. In patients with particularly se- rior basal segments and apical portions of the superior seg-
vere and widespread involvement of the small airways, the ments of the lower lobes. However, when lobular air trap-
patchy distribution of hypoattenuation and a mosaic pat- ping occurs in the non-dependent portions of the lung or
tern is lost. Inspiratory scans show an apparent uniformi- the overall extent is equal to or greater than one segment,
ty of decreased attenuation in the lungs, and scans ob- the air trapping should be considered as abnormal. Expi-
tained at end expiration may appear normal. In these pa- ratory air trapping occurs in smokers and in patients with
tients, the most striking features are the paucity of pul- asthma, constrictive bronchiolitis, and bronchiolitis asso-
monary vessels and the lack of a change in the cross-sec- ciated with hypersensitivity pneumonitis and sarcoidosis.
tional area of lung between inspiration and expiration.
Mosaic attenuation is seen in patients who have con-
strictive bronchiolitis, bronchiolitis associated with hy- Asthma
persensitivity pneumonitis, asthma, and COPD.
Constrictive bronchiolitis, characterized by submucos- Asthma [3, 14-16] is a chronic inflammatory condition
al circumferential fibrosis along the central axis of ter- resulting from airway hyper-responsiveness to several
Current Approaches to Chronic and Acute Airway Diseases 35
stimuli. It is characterized by episodes of wheezing, centrilobular nodules primarily in the upper lobes reflect
coughing, and dyspnea due to airflow obstruction that re- inflammatory changes in and around the bronchioles (res-
solves spontaneously or following treatment with bron- piratory bronchiolitis), which can be reversible after smok-
chodilators. Over time, airway remodeling (fibrosis, ing cessation and steroids. Mosaic attenuation and expira-
smooth muscle hypertrophy and hyperplasia, and neovas- tory air trapping reflect obstructive bronchiolitis and re-
cularity) leads to persistent airflow obstruction. modeling of the small airways.
The main clinical indication for imaging patients with Large airway disease is also commonly present in
asthma is to identify diseases that can mimic asthma clin- COPD patients [19, 20]. Saber-sheath trachea occurs
ically, particularly hypersensitive pneumonitis, constric- specifically in COPD, particularly males. Bronchial wall
tive bronchiolitis, and tracheal or carinal obstruction by thickening and irregularities, although nonspecific, are
neoplastic or non-neoplastic tracheal disorders. CT scans frequently present. A quantitative assessment of bronchial
of patients with asthma can be normal or show bronchial wall thickening on CT scans can provide an estimate of
abnormalities. Mucoid impaction and linear bands, re- airway remodeling in the absence of extensive emphyse-
flecting subsegmental or segmental atelectasis, are re- ma. Bronchial wall thickness is one of the strongest de-
versible on follow-up. Bronchial wall thickening is com- terminants of FEV1 in patients with COPD. Moderate
monly present and correlates with clinical severity and tubular bronchiectasis can occur, particularly in the lower
the duration of asthma and the degree of airflow obstruc- lobes, following injury to cartilage. Bronchiectasis is of-
tion. It also correlates with pathologic measures of re- ten associated with more severe COPD exacerbations,
modeling from bronchial biopsies. Bronchiectasis may lower airway bacterial colonization, and increased sputum
also be present; its extent is associated with an increased inflammatory markers [21, 22]. However, the presence of
severity of asthma. bilateral varicose and cystic bronchiectasis in patients
In patients with persistent moderate asthma, mosaic at- with panlobular emphysema should raise the diagnosis of
tenuation reflects remodeling in the small airways. The -1-antitrypsin deficiency. Cartilage deficiency in COPD
extent of expiratory air trapping does not change after in- may also induce abnormal collapse of the airway lumen at
halation of salbutamol. By contrast, in patients with mild expiration, contributing to airflow limitation. Tracheoma-
or moderate uncontrolled asthma, therapy with inhaled lacia can also develop.
corticosteroids reduces the degree of air trapping on CT,
suggesting that CT can be used as a surrogate marker for
assessing disease control. References
1. Beigelman-Aubry C, Brillet PY, Grenier PA (2009) MDCT of
the airways: technique and normal results. Radiol Clin North
Airway Disease in COPD Am 47:185-201.
2. Grenier PA, Beigelman-Aubry C, Fetita C et al (2002) New
COPD consists of a group of diseases characterized by frontiers in CT imaging of airway disease. Eur Radiol
slow and progressive airway obstruction as a result of an 12:1022-1044.
3. Naidich DP, Webb WR, Grenier PA et al (2005) Imaging of the
exaggerated inflammatory response, typically from ciga- airways. Lippincott Williams & Wilkins, Philadelphia, PA.
rette smoke [15, 17]. Progressive disease ultimately leads 4. Hansell DM, Armstrong P, Lynch DA et al (2005) Imaging of
to the destruction of lung parenchyma (emphysema) and a diseases of the chest, 4th Ed. Elsevier Mosby, Philadelphia,
permanent reduction in the number and caliber of the PA.
small airways (obstructive bronchiolitis). Both emphyse- 5. Ferretti GR, Bithigoffer C, Righini CA et al (2009) Imaging of
tumors of the trachea and central bronchi. Radiol Clin North
ma, because of the loss of alveolar attachments and de- Am 47:227-241.
creased elastic recoil, and airway wall remodeling, with 6. Kang EY (2011) Large airway diseases. J Thorac Imaging
peribronchiolar fibrosis, are responsible for airflow limita- 26:249-262.
tion. The two types of lesions can coexist in the same pa- 7. Grenier PA, Beigelman-Aubry C, Brillet PY (2009) Nonneo-
tient, and small airway disease may precede the appear- plastic tracheal and bronchial stenoses. Radiol Clin North Am
47:243-260.
ance of emphysema [18]. This explains why individuals 8. Lee EY, Litmanovich D, Boiselle PM (2009). Multidetector
with the same degree of functional impairment may have CT evaluation of tracheobronchomalacia. Radiol Clin North
CT scans showing different morphologic appearances. For Am 47:261-269.
example, one patient with extensive emphysema on CT 9. Ridge CA, ODonnell CR, Lee EY et al (2011) Tracheobron-
can have the same degree of airflow obstruction as anoth- chomalacia: current concepts and controversies. J Thorac
Imaging 26:278-289.
er patient with little or no emphysema, suggesting small 10. Javidan-Nejad C, Bhalla S (2009) Bronchiectasis. Radiol Clin
airway remodeling. These morphologic differences sug- North Am 47:289-306.
gest differences in pathophysiology and the genomic pro- 11. ODonnell AE (2008) Bronchiectasis. Chest 134:815-823.
file. CT can help define the different phenotypes of 12. Feldman C (2011). Bronchiectasis: new approaches to diagno-
sis and management. Clin Chest Med 32:535-546.
COPD, allowing for better stratification of patients in clin- 13. Pipavath SN, Stern EJ (2009) Imaging of Small Airway Dis-
ical trials and, ultimately, for a more personalized treat- ease (SAD). Radiol Clin North Am 47:307-316.
ment. Airway disease phenotypes of COPD include ab- 14. Boiselle P, Lynch D (2008) CT of the airways Humana Press,
normalities of the small and large airways. Poorly defined Totowa, NJ, USA.
36 P.A. Grenier, J.P. Kanne
15. Kauczor HU, Wielptz MO, Owsijewitsch M et al (2011) 19. Brillet PY, Fetita CI, Saragaglia A et al (2008) Investigation of
Computed tomographic imaging of the airways in COPD and airways using MDCT for visual and quantitative assessment in
asthma. J Thorac Imaging 26:290-300. COPD patients. Int J Chronic Obstruct Pulm Dis 3:97-107.
16. Aysola RS, Hoffman EA, Gierada D et al (2008) Airway re- 20. Sverzellati N, Ingegnoli A, Calabr E et al (2010) Bronchial
modeling measured by multidetector CT is increased in se- diverticula in smokers on thin-section CT. Eur Radiol 20:88-
vere asthma and correlates with pathology. Chest 134:1183- 94.
1191. 21. OBrien C, Guest PJ, Hill SL, Stockley RA (2000) Physiolog-
17. Ley-Zaporozhan J, Kauczor HU (2009) Imaging of airways: ical and radiological characterisation of patients diagnosed
chronic obstructive pulmonary disease. Radiol Clin North Am with chronic obstructive pulmonary disease in primary care.
47:331-342. Thorax 55:635-642.
18. McDonough JE, Yuan R, Suzuki M et al (2011) Small-airway 22. Martnez-Garca M, Soler-Catalua JJ, Donat Sanz Y et al
obstruction and emphysema in chronic obstructive pulmonary (2011) Factors associated with bronchiectasis in patients with
disease. N Engl J Med 365:1567-1575. COPD. Chest 140:1130-1137.
IDKD 2015-2018
Non-Specific
isolated
Normal persistent
or cough
expected
cough
Specific
cough
Table 3. Influence of the time of onset on the causes of cough ble of performing the maneuvers (generally those over
Time of onset Neonatal, infancy and childhood the age of 5 years) [1, 2].
Nature Dry or productive In children with a specific cough, further investiga-
Quality Brassy, croupy, honking, paroxysmal, tions may be warranted (Figs. 3, 4), except when asthma
staccato is the main cause [1, 2]. The next sections discuss the var-
Timing Persistent, intermittent, nocturnal, on ious etiologies of chronic cough.
awaking
Triggering factors Cold air, exercise, feeding, seasonal,
starts with a head cold
Alleviating factors Bronchodilators, antibiotics
Chronic Sinusitis
Associated symptoms Wheezing, shortness of breath
Sinusitis is a very common condition in children.
Chronic sinusitis (lasting 3 months) involves symp-
Clinical examination pointers include digital clubbing toms that include nasal obstruction, nasal discharge,
and asymmetrical auscultatory signs. halitosis, and headache [16]. Although some authors re-
Two situations are commonly indicative of significant port a high incidence of incidental sinus opacification
disease. The first is neonatal onset of the cough, suggest- in children [17], those data may be biased by different
ing: (a) a congenital defect leading to feeding problems interpretations of mucosal thickening and by the true
and pulmonary aspiration, (b) cystic fibrosis, (c) primary health condition of the population included in such
ciliary dyskinesia, (d) an anatomical airways abnormali- studies [18]. Tatli et al. [18] found a 52% prevalence of
ty, for example a cyst compressing the airway or tracheo- moderate to severe opacification of the paranasal sinus
malacia, or (e) a chronic viral pneumonia (e.g., cy- on computed tomography (CT) in children with chron-
tomegalovirus or Chlamydia) acquired in utero or during ic cough, with 90% involvement of the maxillary si-
the perinatal period [1, 2]. nuses.
The second situation is persistent productive (moist or Radiographs are used as a screening method for patho-
wet) cough, which requires investigation for the presence logical conditions involving the sinuses but CT remains
of bronchiectasis or any specific suppurative lung condi- the imaging modality of choice for the evaluation of
tion [1, 2]. acute and chronic sinus inflammatory processes [19]. The
association between adenoidal hypertrophy and rhino-
Investigation sinusitis with upper airway inflammation is increasingly
recognized, with one study [20] stating that magnetic
A chest radiograph is indicated for most children with resonance imaging (MRI) can document changes in ade-
chronic cough, unless there is also a minor identified dis- noid size associated with the resolution of rhinosinusitis
order, such as asthma or allergic rhinitis (Table 4). (Fig. 5). Further studies are necessary to define the role
Spirometry should be always attempted in children capa- of MRI in adenoidal hypertrophy.
Reversible airway
obstruction?
Asthma Yes No
If cough does
not settle consider:
Assessment of risk factors for
Fig. 3. Flow chart illustrating the specific pathways for the investigation of children with specific cough, an abnormal chest X-ray, or ab-
normal spirometry. TEF, tracheoesophageal fistula; GERD, gastroesophageal reflux disease; HRCT, high-resolution computed tomography,
MRI, magnetic resonance imaging
a b c
Fig. 5 a-c. Coronal (a), sagittal (b), and axial (c) T2-weighted spin echo magnetic resonance images illustrate paranasal sinus mucosal
thickening and adenoidal hypertrophy in a child with chronic cough related to rhinosinusitis
a b
a b c
Fig. 7 a-c. Chest X-ray and axial contrast enhanced CT scan with volume rendered tomo-
graphy (VRT) in a child with cough and stridor related to a pulmonary artery sling. a Chest
X-ray shows splaying of the trachea (arrows), with an obtuse carinal angle. b Axial CT
shows the narrow trachea (open arrow) with the abnormal LPA passing behind the trachea,
in a sling like fashion. c The VRT image shows long segment tracheal stenosis with a ru-
dimentary R apical tracheal bud (open arrow), with an obtuse carinal angle and narrowing
of the origins of the R and L main bronchi (stovepipe trachea) (white arrow)
Barium swallow and echocardiography can demon- opment of high-speed, high-resolution CT scanning, gen-
strate the presence of a vascular ring. Bronchoscopy in eral anesthesia is usually unnecessary. According to Turn-
this context has to be defined prospectively [15]. Both er et al. [25], the first examination should be a barium
CT and MRI are noninvasive imaging modalities that swallow, followed by a high-resolution chest CT scan in
may help in the planning of surgery [24]. With the devel- the study of vascular rings (Fig. 7).
42 M. do Rosario Matos, G.A. Taylor, C.M. Owens
a b
Fig. 8 a, b. An expiratory chest X-ray (a) and virtual bronchoscopy (b) show gross over-inflation, due to air trapping in a child with obstructive
overinflation and ball valve effect in the right main bronchus due to an inhaled foreign body. This imaging approach increases the diagnostic
accuracy for inhaled FB, whose findings include atelectasis, areas of unilateral or bilateral hyperinflation, or nonspecific opacification
Foreign-Body Aspiration resulting from bronchiectasis are chronic cough and spu-
tum production, although some patients may have a non-
Cough may be an important symptom in acute foreign- productive cough [32].
material inhalation. Foreign-body aspiration (FBA) is Patients without an obvious cause of bronchiectasis
much more frequent in boys than in girls and is more should be diagnostically evaluated for an underlying
common in those under 4 years of age [15]. Although the disorder because the results may lead to treatment that
presentation is usually acute [26], the problem may go can potentially slow or halt progression of the disease
unrecognized for prolonged periods of time if the diag- [32]. Cystic fibrosis (CF) occurs in 1 per 2,0003,000
nosis is missed initially [27]. In that case, the natural his- live births and is probably the most common identifi-
tory may include hemoptysis, cough producing casts in able cause of bronchiectasis in the United States and
the airways, associated wheezing, and associated recur- Europe [32]. CF-related cough may have a neonatal on-
rent pneumonia and infiltrates [1]. set and can be productive or paroxysmal. It may be as-
An expiratory chest X-ray increases the diagnostic ac- sociated with hemoptysis, ill health, recurrent pneumo-
curacy in patients with FBA [7]. Findings include atelec- nia, or pulmonary infiltrates [1]. Non-CF causes of
tasis, areas of unilateral or bilateral hyperinflation, or bronchiectasis include post-infectious disease, primary
nonspecific opacification [28] (Fig. 8). However, a nor- ciliary dyskinesia, immunodeficiency states, retained
mal chest X-ray does not exclude the possibility of FBA. airway foreign body, connective tissue disorders, chron-
If the history is unclear and there are no localized physi- ic aspiration and allergic bronchopulmonary aspergillo-
cal or X-ray findings, flexible bronchoscopy should be sis [33].
performed to locate the inhaled object and rigid bron- In patients with clinically suspected bronchiectasis
choscopy to remove it [7, 29]. without a characteristic chest radiograph, a CT scan can
Virtual bronchoscopy associated with MCDT is a non- confirm the diagnosis. Criteria include an increased
invasive modality for the identification of endobronchial le- bronchial/adjacent pulmonary artery ratio, non-tapering
sions. Helical CT scanning with virtual bronchoscopy may bronchial walls, bronchi visualized at lung periphery,
be performed in selected cases of suspected FBA. When the thickened bronchial walls, mucous plugging and focal
chest radiograph is normal and the clinical diagnosis sug- air-trapping related to small airway disease [33, 34]. CT
gests an aspirated foreign body, helical CT and virtual bron- is also of great value in cases of superadded infection,
choscopy can be considered as a means to avoid rigid bron- such as allergic bronchopulmonary aspergillosis and non-
choscopy if the clinical suspicion is low [30, 31]. tuberculous mycobacterial infection [33] (Fig. 9).
Because CF may be diagnosed early in life due to non-
pulmonary findings (e.g., meconium ileus, malabsorp-
Bronchiectasis tion, and failure to thrive due to pancreatic insufficiency),
CT is used for the diagnosis of its complications rather
Bronchiectasis is defined as abnormal dilatation of thick- than the primary diagnosis. However, in non-CF causes
walled airways. Regardless of the etiology, symptoms of bronchiectasis, CT plays a role in the primary diagno-
Investigating a Child with a Cough: A Pragmatic Approach 43
a b
a b c
Fig. 10 a-c. Axial (a) and (b) sagittal CT scans and (c) virtual bronchoscopy show an endobronchial mass in the right main bronchus of a
child with cough and hemoptysis related to endobronchial Mycobacterium tuberculosis infection.
sis and detection of complications (if chest X-ray is in- chogenic spread, miliary nodules, and segmental or lobar
sufficient) as well as prior to lung surgery [33]. consolidation, which may evolve to caseating necrosis.
CT also defines the extent of disease and its complica-
tions [36] (Fig. 10).
Tuberculosis
Chronic or less common infections that may lead to Considerations on Pediatric CT Radiation Burden
chronic cough are tuberculosis (TB) and non-tuberculous
mycobacteria, fungal, and parasitic infections. TB should The utility of CT scans in the pediatric age needs to be
be considered in any child with a persistent productive balanced with radiation effects. Children are considerably
cough, particularly if there are systemic features such as more sensitive to radiation than adults and have a longer
fever, weight loss, or general malaise. Specific pointers life expectancy, resulting in a higher lifetime cancer mor-
towards this diagnosis in cough investigation are hemo- tality risk [37].
ptysis, associated with severe ill health, recurrent pneu- CT settings in the diagnosis of chest disease can be re-
monia, or pulmonary infiltrates [1]. Since most TB in- duced significantly while maintaining diagnostic image
fections are transmitted by inhalation, primary lesions oc- quality. Jogeesvaran and Owens [33] reported that the
cur in the lungs in over 95% of infected children [34]. dose for a volumetric CT scan is between three and five
Although chest radiography remains the first-line times higher than for a non-contiguous high-resolution
imaging technique in the evaluation of pulmonary TB in CT. Moreover, there is a strong argument that chronic dif-
children, in selected cases CT can provide important in- fuse diseases of the lung parenchyma are adequately im-
formation in the diagnosis and management of the dis- aged with non contiguous CT. The authors recommended
ease. CT typically shows low-attenuation lymph nodes a volumetric scan at the initial diagnostic stage of chron-
with rim enhancement and, eventually, calcification. ic lung disease to exclude co-existent or unsuspected tra-
These findings in HIV-positive patients are considered cheobronchial and vascular anomalies. For children sus-
sufficient to warrant instituting empirical anti-TB thera- pected of having diffuse interstitial lung diseases at diag-
py [35]. Parenchymal lesions include nodules of bron- nosis and at follow-up of either airways or diffuse infil-
44 M. do Rosario Matos, G.A. Taylor, C.M. Owens
trative lung disease, a limited, non-contiguous, conven- 10. Munyard P, Bush A (1996) How much coughing is normal?
tional high-resolution CT should be the preferred exam. Arch Dis Child 74:531-534.
11. Shann F (1996) How often do children cough? Lancet
348:699-700.
12. Chang AB, Phelan PD, Sawyer SM et al (1997) Airway hyper-
Conclusion responsiveness and cough-receptor sensitivity in children with
recurrent cough. Am J Respir Crit Care Med 155:1935-1939.
In the management of illness in children, adult-based da- 13. Chang AB, Phelan PD, Sawyer SM et al (1997) Cough sensi-
tivity in children with asthma, recurrent cough, and cystic fi-
ta should not be extrapolated: children are different from brosis. Arch Dis Child 77:331-334.
adults. 14. Boujaoude ZC, Prat MR (2010) Clinical approach to acute
Most children with cough due to a simple upper respi- cough. Lung 188(Suppl 1):4146.
ratory tract infection will not need further investigations. 15. Jongste JC, Shields MD (2003) Cough 2: Chronic cough in
Chest radiograph should be considered in the presence children. Thorax 58:998-1003.
16. Triulzi F, Zirpoli S (2007) Imaging techniques in the diagno-
of: lower respiratory tract signs, relentlessly progressive sis and management of rhinosinusitis in children. Pediatr Al-
cough, hemoptysis, or features of an undiagnosed chron- lergy Immunol 18:46-49.
ic respiratory disorder. 17. Glasier CM, Ascher DP, Williams KD (1986) Incidental
Children with chronic cough require careful and sys- paranasal sinus abnormalities on CT of children. Am J Neuro-
tematic evaluation for the presence of specific diagnostic radiol 7:861-864.
18. Tatli M, Sanb I, Karaoglanogluc M (2001). Paranasal sinus
indicators and should undergo, as a minimum, a chest ra- computed tomographic findings of children with chronic
diograph and spirometry. cough. Int Journ Pediatr Otorhinolaryngol 60:213-217.
In children with specific cough, further investigations 19. Mafee MF, Tran BH, Chapa AR (2006) Imaging of rhinosi-
may be warranted, except when asthma is the etiologic nusitis and its complications: plain film, CT, and MRI. Clin
Rev Allergy Immunol 30:165-186.
factor. 20. Georgalas C, Thomas K, Owens C et al (2005) Medical treat-
Children with chronic productive purulent cough ment for rhinosinusitis associated with adenoidal hypertrophy
should always be investigated, to document the presence in children: an evaluation of clinical response and changes on
or absence of bronchiectasis and to identify underlying magnetic resonance imaging. Ann Otol Rhinol Laryngol
and treatable causes such as CF and immune deficiency. 114:638-644.
21. Gormley PK, Colreavy MP, Patil N et al (1999) Congenital
Chronic cough starting in the neonatal period usually vascular anomalies and persistent respiratory symptoms in
indicates significant disease, especially if it starts in the children. Int J Pediatr Otorhinolaryngol 51:23-31.
first few days or weeks of life. 22. Johnson JF, Sueoka BL, Mulligan ME, Lugo EJ (1985) Tra-
There are advantages of CT in children but the tech- cheoesophageal fistula: diagnosis with CT. Pediatric Radiolo-
niques and guidelines regarding the use of the different gy 15:134-135.
23. Ou P, Seror E, Layouss W et al (2007) Definitive diagnosis and
modalities in the pediatric population must be kept in surgical planning of H-type tracheoesophageal fistula in a crit-
mind, and the potential risks balanced accordingly. ically ill neonate: First experience using air distension of the
esophagus during high-resolution computed tomography ac-
quisition. Thorac Cardiovasc Surg 133:1116-1117.
24. Chun K, Colombani PM, Dudgeon DL et al (1992) Diagnosis
References and management of congenital vascular rings: a 22-year expe-
rience. Ann Thoraci Surg 53:597-602.
1. Shields MD, Bush A, Everard ML et al (2008) Recommenda- 25. Turner A, Gavel G, Coutts J (2005) Vascular ringspresenta-
tions for the assessment and management of cough in children. tion, investigation and outcome. Eur J Pediatr 164:266-270.
Thorax 63:1-15. 26. Mu L, He P, Sun D (1991) Inhalation of foreign bodies in Chi-
2. Chang AB, Glomb WB (2006) Guidelines for evaluating nese children: a review of 400 cases. Laryngoscope 101:657-
chronic cough in pediatrics: ACCP evidence-based clinical 660.
practice guidelines. Chest 129: 260-283. 27. Raman TS, Mathew S, Ravikumar, Garcha PS (1998) Atelec-
3. Couriel J (1998) Infection in children. In: Ellis ME (ed) In- tasis in children. Indian Pediatr 35:429-435.
fectious diseases of the respiratory tract. Cambridge Universi- 28. Svedstrm E, Puhakka H, Kero P (1989) How accurate is chest
ty Press, Cambridge, UK, pp 406-429. radiography in the diagnosis of tracheobronchial foreign bod-
4. DUrzo A, Jugovic P (2002) Chronic cough. Three most com- ies in children? Pediatr Radiol 19:520-522.
mon causes. Can Fam Physician 48:1311-1316. 29. Barben J, Berkowitz RG, Kemp A, Massie J (2000) Bronchial
5. Chang AB (2005) Cough: are children really different to granuloma wheres the foreign body? Int J Pediatr Otorhino-
adults? Cough 1:7. laryngol 53:215-219.
6. Barr RL, McCrystal DJ, Perry CF, Chang AB (2005) A rare 30. Haliloglu M, Ciftci AO, Oto A et al (2003) CT virtual bron-
cause of specific cough in a child: the importance of follow- choscopy in the evaluation of children with suspected foreign
ing up children with chronic cough. Cough 1:8. body aspiration. Eur J Radiol 48:188-192.
7. Massie J (2006) Cough in children: when does it matter? Pae- 31. Sodhi KS, Saxena AK, Singh M et al (2008) CT virtual bron-
diatric respiratory reviews 7:9-14. choscopy: New non invasive tool in pediatric patients with
8. Sherrill DL, Guerra S, Minervini MC et al (2005) The rela- foreign body aspiration. Indian Journal of Pediatrics 75:511-
tionship of rhinitis to recurrent cough and wheezing: a longi- 513.
tudinal study. Respir Med 99:1377-1385. 32. Rosen MJ (2006) Chronic cough due to bronchiectasis: ACCP
9. Chang AB, Phelan PD, Sawyer SM et al (1997) Airway hy- evidence-based clinical practice guidelines. Chest 129:122-
perresponsiveness and cough receptor sensitivity in children 131.
with recurrent cough. Am J Respir Crit Care Med 155:1935- 33. Jogeesvaran H, Owens CM (2010) Chronic diseases of lung
1939. parenchyma in children: the role. Pediatr Radiol 40:850-858.
Investigating a Child with a Cough: A Pragmatic Approach 45
34. Couriel J (2002) Assessment of the child with recurrent chest 36. Kim W, Moon W, Kim IO et al (1997) Pulmonary tuberculosis
infections. Br Med Bull 61:115-132. in children: evaluation with CT. AJR Am J Roentgenol
35. Pastores SM, Naidich DP, Aranda CR et al (1993) Intrathoracic 68:1005-1009.
adenopathy associated with pulmonary tuberculosis in patients 37. Brenner D, Elliston C, Hall E, Werdon B (2001) Estimated
with human immunodeficiency virus infection. Chest risks of radiation-induced fatal cancer from pediatric CT. AJR
103:1433-1437. Am J Roentgenol 17:289-296.
IDKD 2015-2018
Immunodeficiencies
Defects in one or more of the essential components of the
immune system can result in an increased risk of devel-
oping thoracic abnormalities in pediatric patients [3].
When assessing the chest of children with an immuno-
deficiency, the radiologists role includes: (1) detecting
characteristic imaging features of a specific immuno-
deficiency, (2) evaluating infections, and (3) detecting
malignancies that can occur as a complication of certain
immunodeficiencies. Because of the large number of im-
munodeficiencies, only the disorders that are the most Fig. 3. A 3-year-old boy with X-linked agammaglobulinemia. En-
common or have the most characteristic thoracic mani- hanced coronal reformatted CT shows multi-focal pneumonia from
festations are discussed in the following. Streptococcus infection
48 E.Y. Lee, A.S. Brody
Fig. 14. A 14-year-old boy with granulomatosis with polyangiitis Fig. 16. A 13-year-old boy with granulomatosis with polyangiitis
(Wagner granulomatosis). Axial lung window CT shows pul- who presented with neck pain and shortness of breath. Frontal 3D
monary nodules (arrows) volume-rendered image of the large airway shows a marked nar-
rowing (arrows) of the subglottic airway
ratory tract and the kidney. Typical thoracic manifesta- Miscellaneous Disorders
tions include multiple pulmonary nodules usually rang-
ing in size from 2 mm to several centimeters in diameter Cystic Fibrosis
[1, 3] (Fig. 14). Associated cavitation is seen in larger
pulmonary nodules (2 cm). Consolidation due to un- Cystic fibrosis (CF), the most common cause of pul-
derlying pulmonary hemorrhage and/or ischemic necro- monary insufficiency in childhood, is caused by a muta-
sis may also be present (Fig. 15). In addition, there may tion in the gene encoding the cystic fibrosis transmem-
be large-airway involvement, such as tracheobronchial brance regulator (CFTR) [1, 3]. CF typically affects chil-
wall thickening, tracheobronchial stenosis, or bronchiec- dren of European heritage, with an estimated incidence of
tasis [3] (Fig. 16). 1:2,500 live births. Affected individuals commonly pre-
sent with failure to thrive, meconium ileus syndrome,
malabsorption syndrome, or chronic respiratory infec-
tion. The diagnosis is currently most often made during
newborn screening. An abnormal sweat test or genetic
test can provide a definite diagnosis.
In children with early stage of CF, the thoracic mani-
festations include mild to moderate air trapping (hyper-
inflation) due to an underlying obstruction of the small
airways by abnormally viscid mucus and/or bronchial
wall thickening. Upper lobe predominant bronchiectasis,
peribronchial wall thickening, centrilobular nodular and
tree-in-bud opacities, and mucus plugging with air-trap-
ping are usually present in children with later or ad-
vanced CF [1, 3] (Fig. 17a). Because of the chronic and
recurrent superimposed lung infection, concomitant reac-
tive hilar and mediastinal lymphadenopathy as well as en-
larged bronchial arteries are also often present (Fig. 17b).
Although many patients with CF are evaluated by chest
radiograph, it is not sensitive enough to detect the early
Fig. 15. A 15-year-old girl with granulomatosis with polyangiitis
who presented with dyspnea and a decreased hematocrit. Axial
or subtle lung changes. Currently, high-resolution CT is
lung window CT shows multifocal airspace consolidations in both the most sensitive diagnostic imaging modality for as-
lungs, compatible with pulmonary hemorrhage sessing the morphologic changes of CF lung disease.
Thoracic Manifestations of Pediatric Systemic Diseases 53
a b
Fig. 17 a, b. A 15-year-old girl with cystic fibrosis. a Coronal lung window CT shows extensive upper-lobe-predominant bronchiectatic
changes consistent with lung changes due to cystic fibrosis. b Coronal soft-tissue window CT shows extensive mucus plugging and en-
larged bronchial arteries from chronic infection
A previously published study showed that CT is more hemoglobin can result in anemia and bone marrow in-
sensitive than chest radiographs for detecting CF-related farction. Eventually, sickle cell disease leads to various
lung abnormalities, and that the CT score correlates well acute and chronic complications in multi-organ systems.
with the forced vital capacity (FVC), one-second forced Two main chest complications of sickle cell disease are
expiratory volume (FEV1) ratio on a pulmonary function acute chest syndrome (ACS) and pneumonia [1, 3]. The
test [27]. A major consideration in the imaging of CF pa- former refers to the clinical situation in which patients
tients is the higher risk posed by ionizing radiation expo- with sickle cell disease develop a new opacity on chest
sure. Therefore, in CF patients, MRI with advanced tech- radiographs accompanied by chest pain, fever, and respi-
niques including fast imaging sequences, high-resolution ratory syndrome [29] (Fig. 18). Possible underlying etio-
sequences, and hyperpolarized gas is being actively in- logies for ACS include infection, pulmonary infarction,
vestigated for evaluating CF lung disease [28]. In addi- and fat embolism from vaso-occlusive crises involving
tion to pulmonary disease, pancreatic disease is another the bone marrow. A lack of interval resolution of the pul-
primary abnormality in CF patients. monary opacities on the follow-up study should raise the
Although lung infections are treated with antibiotics, possibility of a superimposed infectious process such as
lung transplantation may be required for survival in pe- pneumonia. The two most common cardiovascular imag-
diatric patients with advanced CF. A broad range of ther- ing findings in pediatric patients with sickle cell disease
apies, including mucolytics and inhaled antibiotics, have are cardiomegaly and pulmonary vascular plethora relat-
improved both the life expectancy and the quality of life ed to chronic anemia. Abnormal osseous changes due to
in CF patients. New medications targeting the specific underlying bone infarction are also frequently present. In
defect in CFTR have recently become available, and the spine, recurrent bone infarction causes indentation of
gene therapy continues to be investigated as a potential the upper and lower vertebral endplates (H-shaped verte-
treatment. bral bodies). Humeral head sclerosis or fragmentation re-
lated to bone infarction can be observed on chest radio-
Sickle Cell Disease graphs. Hand-foot syndrome, which is characterized by a
painful swelling of the hands and feet, may occur in
Sickle cell disease is due to an abnormality in the oxygen- young infants and children. Pediatric patients with hand-
carrying hemoglobin molecule of red blood cells, such foot syndrome may also have soft-tissue swelling and pe-
that the cells have a propensity for to assume an abnor- riosteal new bone formation (Fig. 19). In addition, the
mal, sickle-like shape. Overproduction of abnormal spleen, the function of which is to clear red blood cells,
54 E.Y. Lee, A.S. Brody
Fig. 19. An 8-month-old boy with sickle cell disease who presented
with pain and swelling involving the right foot. Frontal radiograph
of the right foot shows lytic bony changes associated with sclerot-
ic new bone formation involving the first and second metatarsals.
Soft-tissue swelling overlying the first metatarsal is also seen. The Fig. 20. A 17-year-old girl with Langerhans cell histiocytosis in-
constellation of imaging findings is compatible with hand-foot volving the lungs. Axial lung window CT shows randomly distrib-
syndrome in a sickle cell disease patient uted reticulonodular opacities and cysts in both lungs
Thoracic Manifestations of Pediatric Systemic Diseases 55
Tuberous Sclerosis a
a b c
Fig. 1 a-c. Acute pulmonary embolism. Transverse contrast-enhanced CT section (a) allows the diagnosis of a large intraluminal clot (arrow).
Multiplanar reconstructions in the sagittal (b) and coronal (c) planes show the true sagittal and coronal extent of the clot (arrows)
source CT, but without image distortion inside the field to improve the visualization of small pulmonary arteries
of view of the second detector. Tacelli et al. showed that at 80 kV(p) [9].
this scanning mode yields CTA examinations of excellent
quality for thoracic applications in routine clinical prac-
tice, including in patients in pulmonary vascular diseases Perfusion Imaging with Dual-Energy CT
such as acute pulmonary embolism (PE) (Fig. 1) [3].
Lung perfusion with dual-energy CT does not reflect
Improved Morphological Evaluation of the Peripheral blood flow analysis per se, as it provides a measurement
Pulmonary Vasculature at only one time point; rather, it yields an iodine map of
the lung microcirculation at a particular time point. Nu-
With the introduction of MDCT, CTA is now recognized merous parameters are known to influence the distribu-
as the reference standard for diagnosing acute PE [4]. A tion of iodine within pulmonary capillaries; some are
main advantage of MDCT over single-slice CT is the technique-related whereas others are due to the anatomi-
ability to scan the entire volume of the thorax with sub- cal and/or physiological circumstances under which the
millimetric collimation in a very short period of time, data were acquired. Dual-energy CTA is performed using
most often under the duration of a single breath-hold. a scanning protocol similar to the one used in clinical
This is particularly useful when evaluating dyspneic pa- practice. The acquisitions proceed from the top to the bot-
tients. These technological advances have improved both tom of the chest, with an injection protocol similar to that
evaluations of the peripheral pulmonary arteries and the of standard CTA obtained with a single energy on a 64-
accuracy of CT in the workup of acute PE. Simultane- slice scanner. Two categories of images can be recon-
ously, MDCT allows radiologists to scan patients at a structed: Diagnostic scans correspond to contiguous
low kilovoltage, with reductions in the dose of contrast 1-mm-thick transverse CT scans generated from the raw
material and the overall radiation dose. This is of partic- spiral projection data of tube A and tube B (60% from the
ular importance in young female patients who may be acquisition by tube A; 40% from the acquisition by tube
exposed to substantial levels of radiation to breast tissue. B). Lung perfusion scans (i.e., images of the perfused
In the context of acute PE, the latter concern is clinical- blood volume of the lung parenchyma) are generated af-
ly relevant given the lower prevalence of acute PE, which ter determination of the iodine content of every voxel of
has dropped from 33% on angiographic studies to 20% the lung parenchyma on the separate 80- and 140-kVp
on CT/MDCT scans. To date, several studies have inves- images. The images can be rendered in gray-scale or color-
tigated the clinical benefits of low-kilovoltage tech- coded. All images can be displayed as transverse scans,
niques, i.e., 80100 kV(p) vs. 120140 kV(p), the para- complemented as needed by corona land sagittal refor-
meters at which CTA is typically performed. However, mats. Even though a dual-energy acquisition does not
for obvious ethical reasons, these studies were based on correspond to true perfusion imaging, several applica-
the comparative analysis of different populations tions of this pulmonary micro-CTA technique have been
scanned with single-source CT [5-8]. The limitations of investigated [10].
these comparisons include the lack of systematic ad-
justment for individual patient morphology, cardiac he- Acute Pulmonary Embolism
modynamics, and potential underlying respiratory dis-
ease. With the introduction of DSCT, the two tubes can Dual-energy CT can detect endoluminal clots on aver-
be set at different kilovoltages. In addition to the op- aged images obtained with tubes A and B as reliably as
portunity to evaluate lung perfusion, there are benefits can single-source CTA [11]. In a preliminary study, the
for standard CTA as this scanning mode has been shown detectability of perfusion defects beyond obstructive
Modern Diagnosis in the Evaluation of Pulmonary Vascular Disease 59
clots was validated. Perfusion defects in the adjacent Obstructive Airways Diseases
lung parenchyma have the typical perfusion-territorial
triangular shape well known from pulmonary angio- Abnormalities of pulmonary perfusion are a feature of
graphic, scintigraphic, and MRI perfusion studies. Dual- numerous smoking-related respiratory diseases. Peinado
energy CTA can help predict perfusion defects without et al. showed that endothelial dysfunction of the pul-
directly identifying peripheral endoluminal clots that monary arteries is present even in patients with mild
may be located in subsegmental or more distal pul- chronic obstructive pulmonary disease (COPD) [13]. In
monary arterial branches. It can also help differentiate these patients, as well as in smokers with normal lung
lung infarction from less specific peripheral lung con- function, some arteries show a thickened intima, sug-
solidation. gesting that tobacco consumption plays an important role
in the pathogenesis of pulmonary vascular pathologies in
Chronic Pulmonary Embolism COPD. Several structural changes in the early stages of
COPD have been described in experimental and animal
Dual-energy CTA can depict perfusion defects distal to models, including proliferation of smooth-muscle fibers
chronic clots (Fig. 2). Three vascular characteristics of within peribronchiolar arterioles and collagen and elastin
chronic PE may manifest on dual-energy CT imaging. deposition in the thickened intima of vessels [14, 15]. In
First, chronic PE causes a mosaic pattern of lung attenu- preliminary studies, Hoffman et al. [16] showed an in-
ation, characterized by areas of ground-glass attenuation, creased heterogeneity of the local mean transit times of
with enlarged vascular segments intermingled with areas contrast material within the pulmonary microvasculature
of normal lung attenuation and smaller vascular seg- of smokers with normal pulmonary function tests. In
ments. When present, these findings are suggestive of their dual-source, dual-energy CTA study of pulmonary
blood flow redistribution, but they are not consistently lobar perfusion in COPD patients, Pansini et al. [17]
seen on conventional CT in patients with chronic PE. In found that nonsmokers had no alterations in lung struc-
these patients, dual-energy CT can detect ground-glass ture and observed a uniform distribution of iodine con-
attenuation of vascular origin based on the high iodine tent within the upper and lower lobes and between the
content within the areas of ground-glass attenuation, thus right and left lungs. Perfusion scans showed significantly
enabling their distinction from ground-glass attenuation lower iodine content within the lung microcirculation of
secondary to bronchial or alveolar diseases [12]. Second, the upper lobes in emphysematous patients than in smok-
chronic PE can cause calcifications within partially or ers without emphysema and a significantly lower perfu-
completely occlusive chronic clots as well as within pul- sion in the upper than in the lower lung zones, consistent
monary artery walls, when chronic PE is complicated by with the destruction of the lung parenchyma. These struc-
longstanding or severe pulmonary hypertension. The cal- tural abnormalities have important implications given the
cifications are detectable using virtual non-contrast epidemiologic and socioeconomic burden of COPD.
imaging, accessible by dual-energy CT imaging. Third,
the images generated at 80 kV can improve visualization Restrictive Airways Diseases
of the systemic collateral supply characteristic of chron-
ic PE and originating from bronchial and non-bronchial The substantial importance of pulmonary hypertension
systemic arteries. on the clinical course and prognosis of patients with
a b c
Fig. 2 a-c. Chronic thromboembolic disease. Transverse CT section (a) shows a web in the pulmonary artery of the right lower lobe (arrow),
which is confirmed (arrow) by a coronal multiplanar reconstruction (b). Lung window (c) shows a dilated main pulmonary artery (black
arrows), dilated peripheral pulmonary artery (white arrow), and parenchymal areas of hypoperfusion (open arrows)
60 A.A. Bankier, C. Engelke
fibrotic lung disease has been extensively recognized. hold duration [21], (2) a two-phase protocol to scan both
However, the mere measurement of pulmonary artery di- the entire thorax and cardiac cavities with the highest
ameters might not be a reliable parameter in the assess- spatial resolution [22], and (3) a dedicated cardiac MD-
ment of disease severity, due to the potentially confound- CT protocol to assess right ventricular function and myo-
ing role of parenchymal traction on central and peripher- cardial mass [23]. In a comparison of ECG-gated 16-slice
al pulmonary vessels. Moreover, because of the age of the MDCT with equilibrium radionuclide ventriculography,
population in which these diseases usually occur, age- these studies confirmed both the feasibility of the former
related changes have to be taken into account in morpho- in a population of hemodynamically stable patients and
metric-based clinical decision-making and disease classi- the accuracy of the CT-estimated right ventricular ejec-
fication. Overall, and despite promising initial scientific tion fraction. The introduction of 64-slice scanners also
evidence, the roles of MDCT and DSCT in assessing pa- offered further improvement in the integration of mor-
tients with fibrotic lung diseases still need to be deter- phology and cardiac function. Salem et al. reported that
mined. The many ongoing pharmacological trials, notably right and left ventricular function was assessable in 93%
in patients with usual interstitial pneumonits, may pro- of their study patients with various respiratory disorders;
vide an ample study ground in this field. in the remaining patients, an imprecise segmentation of
the right and left ventricular cavities was the limiting fac-
tor for precise calculation of end-systolic and end-dias-
Are There Indications for ECG-Gated CTA Examinations tolic ventricular volumes [24]. The positive results were
When Exploring Pulmonary Vascular Diseases? achievable with dose length-product (DLP) values lower
than those recommended for standard non-ECG-gated
Pulmonary Hypertension examinations. This was possible because of the concur-
rent use of two-dose modulation systems, in particular,
Although a pulmonary trunk diameter 33.2 mm has the adjustment of the milliampere setting for patient size
95% sensitivity for the diagnosis of pulmonary hyperten- and anatomical shape and an ECG-controlled tube cur-
sion (PHT), the specificity of this measurement is only rent.
58%, which is insufficient for the accurate diagnosis of Overall, there is increased evidence and awareness in
PHT, notably in patients with mild disease [18]. More- the radiological community that new scanner technology
over, no correlation has been found between the degree has opened up new opportunities for imaging the car-
of PHT and pulmonary trunk diameter. Electrocardio- diopulmonary system [25, 26]. The techniques have also
graph (ECG)-gated MDCT acquisitions of the entire tho- received increased attention in the field of translational
rax enable the evaluation of novel functional parameters research [27].
in addition to the standard morphology. In patients with
PHT, right pulmonary artery distensibility was recently
shown to be an accurate predictor of PHT on ECG-gated Imaging of Pulmonary Vasculitis
64-slice MDCT scans of the chest [19]. In that study, its
diagnostic value was superior to that of the single mea- Systemic primary vasculitides are idiopathic diseases that
surement of pulmonary trunk diameter. cause an inflammatory injury to the vessel walls. Pul-
monary involvement is frequent, and chest-CT often in
Right Ventricular Function combination with PET/CT is the reference imaging tech-
nique in its assessment. Pulmonary vasculitis occurs in a
Fast rotation speed and dedicated cardiac reconstruction wide variety of systemic and pulmonary vascular disor-
algorithms designed to extend the conventional multislice ders. Most vasculitic entities affecting the lung induce
acquisition data scheme have opened up new opportuni- overlapping disease patterns such as pneumonitis with or
ties for cardiac and thoracic imaging applications. The without capillaritis, diffuse alveolar damage, and acute
first method for ECG-gated examinations of the entire pulmonary hemorrhage, or inflammatory obstruction of
thorax was introduced by Flohr et al., using 4-slice MD- the central pulmonary arteries down to the small vessels,
CT technology [20]. This approach provided greater with chronic secondary pulmonary hypertension with or
anatomical coverage than achieved with standard ECG- without interstitial lung disease. Therefore the clinical
gated spiral scanning. When used with reconstruction ap- symptoms per se or the CT-morphology alone are often
proaches for cardiac applications, precise morphologic nonspecific.
data at the level of the larger mediastinal vessels can be Owing to their complimentary value in the imaging of
obtained. Despite this improvement, major progress in central and peripheral vascular territories and their sec-
fast-scanning multislice CT technology came with the in- ondary parenchymal or interstitial abnormalities, CTA,
troduction of 16-slice MDCT, which allowed the integra- high-resolution CT, and fusion imaging play key roles in
tion of cardiac functional information. Three approaches the noninvasive workup of patients with suspected pul-
were proposed: (1) investigating cardiac global function monary vasculitis. They can indicate the need for further
during a whole-chest multislice CT acquisition with a clinical tests, imaging, or invasive diagnostics and direct
16-1.5-mm collimation to ensure an acceptable breath- medical treatment during follow-up.
Modern Diagnosis in the Evaluation of Pulmonary Vascular Disease 61
The Role of CT in the Differential Diagnosis of Pulmonary monary embolism: improvement of vascular enhancement
Vasculitis with low kilovoltage settings. Radiology 241:899-907.
8. Heyer CM, Mohr PS, Lemburg SP et al (2007) Image quality
and radiation exposure at pulmonary CT angiography with
In large-vessel vasculitis, CT is the method of choice, of- 100- or 120-kVp protocol: prospective randomized study. Ra-
ten in combination with PET, to discriminate macrospcop- diology 245:577-583.
ic vascular abnormalities presenting key pathological fea- 9. Gorgos AB, Remy-Jardin M, Duhamel A et al (2009) Evalua-
tures, such as pulmonary arterial wall thickening with late tion of peripheral pulmonary arteries at 80 kV and at 140 kV:
enhancement, steno-occlusive or thrombo-obliterating dis- dual-energy computed tomography assessment in 51 patients.J
Comput Assist Tomogr 33:981-986.
ease with resulting oligemia, infarction in the dependent 10. Remy-Jardin M, Faivre JB, Pontana F et al (2010) Thoracic ap-
lung, or arterial aneurysms, as a facultative cause of mas- plications of dual energy. Radiol Clin North Amer 48:193-205.
sive pulmonary hemorrhage. CT is the modality of choice 11. Pontana F, Faivre JB, Remy-Jardin M et al (2008) Lung perfu-
to demonstrate the effects of peripheral small-vessel pul- sion with dual-energy multidetector-row CT (MDCT): Feasi-
monary vasculitis on the central pulmonary arteries, such bilityfor the evaluation of acute pulmonary embolism in 117
consecutive patients. Acad Radiol 15:1494-1504.
as secondary chronic pulmonary hypertension due to re- 12. Pontana F, Remy-Jardin M, Duhamel A et al (2010) Lung per-
duction of the total cross-sectional area with arteriolar re- fusion with dual energy multidetector-row CT: can it help rec-
modeling or narrowing of the capillary bed in capillaritis. ognize ground glass opacities of vascular origin? Acad Radiol
Because CT-determined pathologies of the lung 17:587-594.
parenchyma, vessels, and airways are highly variable, di- 13. Peinado VI, Barber JA, Ramirez J et al (1988) Endothelial
dysfunction in pulmonary arteries of patients with mild COPD.
agnosis is a challenging interdisciplinary task. Clinical Am J Physiol 274:L908-L913.
and laboratory data have to be included in close coopera- 14. Yamato Y, Sun JP, Churg A et al (1997) Guinea pig pulmonary
tion with the findings of the referring physician. Howev- hypertension caused by cigarette smoke cannot be explained
er, many vasculitic disorders present characteristic CT by capillary bed destruction. J Appl Physiol 82:1644-1653.
15. Santos S, Peinado VI, Ramirez J et al (2002) Characterization
features. CT-pathologic correlates of the peribronchovas- of pulmonary vascular remodelling in smokers and patients
cular axial interstitium, pulmonary hemorrhage, types of with mild COPD. Eur Respir J 19:632-638.
inflammatory parenchymal infiltration, and secondary 16. Hoffman EA, Simon BA, McLennan G (2006) A structural and
pathologies such as organizing pneumonia, should be con- functional assessment of the lung via multidetector-row com-
sidered with reference to the revised 2012 International puted tomography. Proc Am Thorac Soc 3:519-534.
17. Pansini V, Remy-Jardin M, Faivre JB et al (2009) Assessment
Chapel Hill Consensus Conference on Pulmonary Vas- of lobar pulmonary perfusion in COPD patients: preliminary
culitis and the American College of Rheumatology [28]. experience with dual-energy CT angiography. Eur Radiol
19:2834-2843.
Acknowledgment. We thank Prof. Martine Remy-Jardin, 18. Edwards PD, Bull RK, Coulden R (1998) CT measurement of
main pulmonary artery diameter. Br J Radiol 71:1018-1020.
Lille, France, for her previous work in this field, which 19. Revel MP, Faivre JB, Remy-Jardin M et al (2009) Pulmonary
served as an inspiration and blueprint for this contribu- hypertension: ECG-gated 64-section CT angiographic evalua-
tion. tion of new functional parameters as diagnostic criteria. Radio-
logy 250:558-566.
20. Flohr T, Prokop M, Becker CR et al (2002) A retrospectively
ECG-gated multislice spiral CT scan and reconstruction tech-
References nique with suppression of heart pulsation artifacts for cardio-
thoracic imaging with extended volume coverage. Eur Radiol
1. Rogalla P, Kloeters C, Hein PA (2009) CT technology 12:1497-1503.
overview: 64-slice and beyond. Radiol Clin North Am 47: 21. Coche E, Vlassenbroeck A, Roelants V et al (2005) Evaluation
1-11. of biventricular ejection fraction with ECG-gated 16-slice CT:
2. Petersilka M, Bruder H, Krauss B et al (2008) Technical prin- preliminary findings in acute pulmonary embolism in com-
ciples of dual source CT. Eur J Radiol 68:362-368. parison with radionuclide ventriculography. Eur Radiol
3. Tacelli N, Remy-Jardin M, Flohr T et al (2010) Dual-source 15:1432-1440.
chest CT angiography with high temporal resolution and high 22. Delhaye D, Remy-Jardin M, Teisseire A et al (2006) Estima-
pitch modes: evaluation of image quality in 140 patients. Eur tion of right ventricular ejection fraction by multidetector row-
Radiol 20:1188-1196. CT: comparison with equilibrium radionuclide ventriculogra-
4. Remy-Jardin M, Pistolesi M, Goodman LR et al (2007) Man- phy, Part I. AJR Am J Roentgenol 187:1597-1804.
agement of suspected acute pulmonary embolism in the era of 23. Kim TH, Ryu YH, Hur J et al (2005) Evaluation of right ven-
CT angiography: a statement from the Fleischner Society. Ra- tricular volume and mass using retrospective ECG-gated car-
diology 245:315-329. diac multidetector computed tomography: comparison with
5. Holmquist F, Nyman U (2006) Eighty-peak kilovoltage 16- first-pass radionuclide angiography. Eur Radiol 15:1987-
channel multidetector computed tomography and reduced con- 1993.
trastmedium doses tailored to body weight to diagnose pul- 24. Salem R, Remy-Jardin M, Delhaye D et al (2006) Integrated
monary embolism in azotaemic patients. Eur Radiol 16:1165- cardiothoracic imaging with ECG-gated 64-slice multidetec-
1176. tor-row CT: Initial findings in 133 patients. Eur Radiol
6. Sigal-Cinqualbre AB, Hennequin R, Abada H et al (2004) 16:1973-1981.
Low-kilovoltage multi-detector row chest CT in adults: feasi- 25. Lu GM, Zhao Y, Zhang LJ, Schoepf UJ (2012) Dual-energy
bility and effect on image quality and iodine dose. Radiology CT of the lung. AJR Am J Roentgenol 199:S40-53.
231:169-174. 26. Ko JP, Brandman S, Stember J, Naidich DP (2012) Dual-ener-
7. Schueller-Weidekamm C, Schaefer-Prokop CM, Weber M et al gy computed tomography: concepts, performance, and tho-
(2006) CT angiography of pulmonary arteries to detect pul- racic applications. J Thorac Imaging 27:7-22.
62 A.A. Bankier, C. Engelke
27. Badea CT, Guo X, Clark D et al (2012) Dual-energy micro-CT 28. Jennette JC, Falk RJ, Bacon PA et al (2012) Revised Interna-
of the rodent lung. Am J Physiol Lung Cell Mol Physiol tional Chapel Hill Consensus Conference Nomenclature of
302:1088-1097. Vasculitides. Arthritis & Rheumatism Vol. 65, No. 1, January
2013, 1.11.
IDKD 2015-2018
most frequently isolated pathogens in aspiration pneu- group prospectively compared the clinical and radiologic
monia are gram-negative bacteria. Aspiration pneumo- features of CAP caused by L. pneumophila to those of pa-
nia must be differentiated from aspiration pneumonitis, tients with pneumococcal infections AUHA (Ahuja and
which is a chemical pneumonitis that results from the Kanne, 2014). The authors concluded that Legionella in-
aspiration of noncolonized gastric contents (Mendel- fection may clinically as well as radiologically resemble
sons syndrome). a typical bacterial pneumonia.
The definition of CAP has been challenged over the The chest radiogram is the first-line tool in evaluating
last few years, as it also includes pneumonia in patients patients with suspected CAP. Computed tomography
from nursing homes, rehabilitation hospitals, and outpa- (CT) is reserved for assessing complications or for guid-
tient-based surgical centers who routinely receive inva- ing further diagnostic procedures. It is definitely indicat-
sive medical treatment. The bacteriology and outcome of ed in investigating patients with recurrent or persistent
these patients are more similar to those of NPs. There- pulmonary opacifications.
fore, it has been proposed that pneumonia in outpatients
hospitalized for more than 2 days over the previous 3 Nosocomial Pneumonia
months or who reside in nursing homes or extended-care
facilities should be categorized as health-care-associated By definition, NPs develop in a hospital environment.
pneumonia (HCAP). The incidence of NP ranges from 0.5 to 5 cases per 100
admissions, but in the subgroup of ventilated patients in
Clinical Diagnosis an intensive care setting it may reach 741%. Mortality
rates reported for NPs range from 20% in multihospital
Patients suffering from CAP usually present with fever, studies to 50% in single referral centers and universi-
cough, dyspnea, sputum production, and pleuritic chest ty hospitals.
pain, as well as laboratory signs, such as leukocytosis. Mortality is related to the causative agent. The prog-
Because the clinical symptoms are nonspecific, most nosis associated with aerobic gram-negative pneumonias
people who have fever and cough do not have pneumo- is considerably worse than that associated with gram-
nia; in fact, about 30% of patients, and especially the positive or viral agents.
elderly, are afebrile at presentation. Imaging is one of
the most important tools in the diagnosis of CAP. The Pathogenesis
radiographic identification of a new pulmonary infil-
trate is, in the appropriate clinical setting, indicative of NP develops from bacterial colonization of the orophar-
pneumonia. Conversely, a patient who has fever and ynx followed by the aspiration of oropharyngeal secre-
cough but does not have radiologic proof of pneumonia tions and gastrointestinal contents into the lungs. The
cannot be considered to have pneumonia. In CAP, the majority of NPs are caused by gram-negative bacilli, in-
causative organism is frequently not identified because cluding Pseudomonas aeruginosa, Klebsiella spp., En-
noninvasive tests such as sputum cultures correctly terobacteriaceae spp., Escherichia coli, Serratia marce-
identify the offending organism in only 50% of cases, scens, and Proteus, and to a lesser extent by gram-posi-
and invasive procedures are rarely used in patients with tive cocci, atypical bacteria such as L. pneumophila, and
pneumonia. viruses such as RSV. Microbial contamination of insert-
ed tubes, lines, and catheters is an important source of
Radiographic Diagnosis infection. Less commonly, NP is the result of bacteremia
originating from right-sided endocarditis or septic pelvic
In patients with CAP, the primary role of the radiologist thrombophlebitis. Risk factors are either patient-related
is to detect or to exclude pneumonia. A second task is to (underlying illness, previous surgery, prolonged hospital
aid the clinician in determining the etiologic diagnosis. care) or iatrogenic (intravascular catheters, tracheal
Categorization of the causative organism is sometimes tubes, indwelling catheters, respirator equipment).
possible by integrating clinical and laboratory informa- Sources of infections are hospital personnel and patients
tion with radiographic pattern recognition (see the section with active infections. The inappropriate use of broad-
Radiographic Patterns). A specific etiologic diagnosis, spectrum and prophylactic antibiotics is an additional
however, is difficult to establish, given the increasing and important factor leading to an increased susceptibil-
spectrum of causative organisms and their overlapping ity to NPs.
radiographic features. In a prospective study of 359
adults with CAP, Fang and coauthors compared the radio- Clinical Diagnosis
graphic, clinical, and laboratory features of patients with
bacterial pneumonia (caused by H. influenzae, S. pneu- Compared with CAP, it may be difficult for the clini-
moniae, S. aureus, and aerobic gram-negative bacilli) and cian to diagnose pneumonia in a hospitalized patient.
those with atypical pneumonia (caused by M. pneumoniae The classical findings for pneumonia, such as new
and Chlamydia spp.). The authors found no features that fever, new pulmonary opacification on chest radio-
could reliably differentiate these two groups. Another graphs, cough, sputum production, and elevated
Imaging of Pulmonary Infections 65
leukocyte count may not be present in the hospitalized disease (bronchopneumonia) and bilateral diffuse dis-
patient with NP. Moreover, even if these symptoms are ease.
present, they may not necessarily be caused by pneu- Most commonly, consolidation, confined to a segment
monia. Microbiologic evaluation of the patient with or a lobe of one or both lungs, is caused by typical and
suspected NP (sputum, bronchoalveolar lavage) may or atypical bacteria, whereas bronchopneumonia common-
may not be helpful because of the difficulties in differ- ly relates to Staphylococcus aureus, Haemophilus, My-
entiating contamination from true infection. In addi- coplasma, and tuberculous infection. Diffuse bilateral
tion, pulmonary disease in a hospital environment may lung disease, frequently developing over time, can in
be caused by more than one agent. Therefore, identify- most cases be attributed to viruses and fungi such as
ing a pulmonary infection, the various methods used to Pneumocystis jirovecii. Nodular disease may be attrib-
obtain a specimen, and the value of isolating potential uted to septic emboli (small nodules), larger nodules can
pathogens are matters of constant discussion in the be caused by Nocardia and fungal disorders in immuno-
clinical diagnosis of NP. compromised hosts.
Again, these patterns are nonspecific, they overlap,
Radiographic Diagnosis may be mimicked by non-infectious lung disease and
masked by pre- or coexisting lung conditions. Patterns
Because of the potential difficulties in the clinical diag- can be identified both on chest radiography and CT but
nosis of pneumonia in the hospitalized patient, the radio- CT may help to identify complex or coexisting patterns
logist has an important role in detecting and classifying and aid the novice in establishing a diagnosis.
suspected cases. However, the radiographic diagnosis of
a pulmonary opacity in suspected NP is not as straight-
forward as it is in patients with CAP. The radiographic Opportunistic Infections
diagnosis of a pneumonia may be hampered by preexist-
ing disorders or concomitant lung disease, such as fi- Infectious agents that cause opportunistic pneumonia in
brosing alveolitis, lupus pneumonitis, hemorrhage or humans include representatives from the classifications
contusion, acquired respiratory distress syndrome bacteria, virus, fungus, protozoa, and parasite. In the fol-
(ARDS), tumor, atelectasis, and embolic infarcts. These lowing, we review some of these pathogens and their ap-
disorders may obscure or alter the otherwise characteris- pearance on chest film and CT. Usually, the chest radio-
tic radiographic appearance of a pulmonary opacifica- graph will reveal the abnormality but, occasionally, the
tion and thus render the etiologic approach using pattern increased sensitivity of CT is necessary and even rec-
recognition difficult. ommended to see the pneumonia. Whereas the findings
The difficulties in diagnosing NP can be readily on chest imaging may not be totally pathognomonic of
demonstrated by two examples. Winer-Muram et al. as- the underlying etiology of infection, they may still be
sessed the diagnostic accuracy of bedside chest radio- highly suggestive and will certainly lead to a reasonable
graphy for pneumonia in ARDS patients. The overall di- differential diagnosis.
agnostic accuracy in these patients was only 42% be-
cause of false-negative and false-positive results origi- Human Immunodeficiency Syndrome
nating from diffuse parenchymal areas of increased
opacity that obscured the radiographic features of pneu- Since the first description of HIV/acquired immune de-
monia. Wunderink et al. compared premortem chest ra- ficiency syndrome (AIDS), pneumocystis pneumonia
diographic findings with pulmonary autopsy studies in caused by Pneumocystis jirovecii (PJP) has been one of
ventilated patients with NP. No radiographic sign had a its most common complications. (Previously it was
diagnostic efficiency greater than 68%. The only radio- thought that Pneumocystis carinii was the cause of
graphic sign that correlated with pneumonia, correctly these infections, but this is not the case. The abbrevia-
predicting 60% of the cases, was the presence of an air tion PCP is still used in some circles to refer to pneu-
bronchogram. mocystis pneumonia.) In recent years the incidence of
CT is used more often when an NP is suspected than PJP as a presenting abnormality in patients with AIDS
in patients with CAP, as it can detect early morphologic has decreased. Nevertheless it still accounts for a con-
signs of infection (for example, ground-glass densities). siderable amount of diseases in patients not on highly
CT can also identify a pulmonary opacification in areas active antiretroviral therapy (HAART) or prophylactic
of preexistent disease, detect complications, and guide therapy and should be considered as a diagnosis when
invasive diagnostic procedures, such as bronchoscopy or characteristic radiographic findings are noted. Patients
percutaneous biopsy. with PJP typically present with increasing shortness of
breath; the disease may run a gradual or fulminant
Radiographic Patterns course. Chest films classically reveal a bilateral fine to
medium reticulonodular pattern, generally bilateral but
In general terms, imaging patterns can be grouped into occasionally focal or unilateral. If the patient remains
airspace consolidation (lobar pneumonia), bronchiolar untreated, the radiograph progressively becomes more
66 P. Goodman, H. Prosch, C.J. Herold
opaque and bilateral homogeneous opacities may ulti- and/or superior segment lower-lobe heterogeneous
mately be seen. While upper lobe involvement is more opacities with or without cavitation; on CT scans, imag-
frequent, any lobe may be involved. On CT imaging ing may also demonstrate centrilobular nodules and/or
PJP presents as ground-glass opacification in the areas tree-in-bud opacities. In patients with low CD-4 cell
of involvement, typically perihilar in distribution. The counts and primary infection, homogeneous lobar opac-
chest film occasionally may worsen within a few days ities with adenopathy similar to immune competent
of intravenous trimethoprim-sulfamethoxazole treat- hosts may be seen on chest films, but increased
ment secondary to overhydration and the production of adenopathy may also be present. With post-primary dis-
pulmonary edema, but this can be treated quite effec- ease, the organism disseminates more widely, creating a
tively with diuretics. diffuse, coarse, nodular pattern on chest film similar to
Otherwise, with treatment, the radiographic course is the pattern seen with fungal infections. Cavitation does
one of steady improvement, with complete resolution not develop, as the bodys immune response is weak,
usually occurring by day 11. Pneumatoceles , as seen on such that well-formed granulomas and necrosis are un-
chest films, develops in 10% of patients; the inci- usual. In patients with improving cell-mediated immu-
dence is probably higher in patients evaluated by CT nity secondary to HAART, more typical findings of
scanning. These air-filled cysts are frequently multiple, cavitation might be seen. If the organism is sensitive to
located in the upper lobes, measure 15 cm in diameter, the appropriate therapy, then some resolution of the ab-
and will resolve within 2 months. However, in 35% of normal findings should be observed on chest film with-
patients, pneumatoceles may lead to pneumothorax, in 1 week.
which can be extremely difficult to treat. Overall, pneu- Ordinary bacterial infections now occur with in-
mothorax develops in 5% of patients with PJP and creased frequency and among some patients with HIV
AIDS. In about 10% of patients with PJP, the chest film are the most common type of infection. This increasing
may be normal. In some of these patients with normal percentage of bacterial infections may reflect the larg-
radiographs, a CT scan will show the typical geograph- er number of pulmonary infections in populations in
ic, ground-glass opacities associated with pneumocystis whom antiretroviral and prophylactic therapy is avail-
pneumonia. Lymphadenopathy and pleural effusions are able. Common organisms are S. pneumoniae, H. in-
not part of the PJP picture. fluenzae, S. aureus, and P. aeruginosa. In the immuno-
Cytomegalovirus (CMV) may mimic the appearance compromised, pneumonias caused by these bacteria
of PJP on chest film, with diffuse bilateral fine to medi- usually appear as they do in normal hosts, i.e., as ho-
um reticulonodular opacities. On CT, centrilobular nod- mogeneous, peripheral, lobar opacities. Parapneumonic
ules and ground-glass opacities are reported. In some effusions may be present. S. aureus, and P. aeruginosa
patients with CMV, the presence of discreet nodules, pneumonias may present with cavitations. These should
sometimes several centimeters in size, may help distin- begin to resolve within days of instituting antibacterial
guish between these two entities. Lymphocytic intersti- therapy, with complete resolution of abnormalities usu-
tial pneumonia may also mimic PJP with fine reticular ally occurring in about 2 weeks. Other bacterial etiolo-
opacities on chest radiographs, ground-glass opacities gies, such as Rhodococcus equi and Nocardia aster-
on CT scans, and air-filled cysts seen on both forms of oides, are less common and may present as nodules or
imaging. masses with or without cavitation.
Disseminated fungal infections, such as histoplasmo- With the use of antiretroviral therapy, noninfectious
sis and coccidiomycosis, generally produce bilateral, fair- etiologies of disease have become more prevalent, such
ly symmetric, coarse, nodular opacities on chest radio- as pulmonary hypertension and chronic obstructive pul-
graphs. monary disease (COPD), including emphysema and
The nodules and occasionally reticular opacities are chronic bronchitis. Also, the immune reconstitution in-
larger than those seen with PJP, which may aid in distin- flammatory syndrome (IRIS), occurring in up to 30% of
guishing between the two processes. Discreet larger nod- patients after the initiation of antiretroviral treatment,
ule(s) or disseminated disease may be seen with other may cause confusing clinical and radiological findings.
fungal infections such as aspergillosis and cryptococco- With improving immune status, patients are able to
sis. With angio-invasive fungal infections cavitation may mount a greater inflammatory response to existing or-
occur secondary to ischemia, regardless of the CD-4 lym- ganisms, resulting in worsening clinical disease and in-
phocyte count. creasing lung parenchymal abnormalities and lym-
The imaging appearance of tuberculosis (TB) de- phadenopathy. This usually develops in individuals with
pends on the patients immune status. In patients with low CD-4 counts and high viral loads, typically about
relatively normal CD-4 lymphocyte cell counts, TB will one month after starting therapy (but as early as within
look much like it does in the general population. That days and as late as after several months). It is more com-
is, with primary infection, patients will present with a monly seen in patients with partially treated tuberculo-
homogeneous lobar opacity and ipsilateral hilar and/or sis or non-tuberculous mycobacterial infection. Worsen-
mediastinal adenopathy. With post-primary infection, ing of other infections and the development of sar-
chest films will show apical and posterior upper-lobe coidosis have also been attributed to HAART and IRIS.
Imaging of Pulmonary Infections 67
Two infection-related neoplasms, non-Hodgkins lateral involvement. On chest film, varicella pneumonia
lymphoma (NHL) and Kaposis sarcoma, may also be usually produces bilateral symmetric acinar opacities
seen in patients with AIDS and could cause confusion (poorly marginated nodules 710 mm in diameter) that
in generating a differential diagnosis. NHL will pro- may coalesce as the disease worsens. CT shows similar-
duce well-defined, discete, nodules on chest films. The sized nodules and distribution as well as ground-glass
nodules range in size from about 1 cm to several cen- opacities.
timeters. Solitary or multiple nodules may be noted. Among the fungal organisms seen with some regular-
Lymphadenopathy and pleural effusions are also ob- ity in this group of immunocompromised patients are
served. The nodules have a tendency to grow extreme- PJP, Cryptococcus, and Aspergillus. Other emerging
ly rapidly. agents include Pseudallescheria/Scedosporium species,
Whereas the nodules with NHL are very well defined, Fusarium spp., and Mucorales spp. The appearance of
those associated with Kaposis sarcoma are not. This dis- PJP is similar to that described for patients with
ease produces poorly marginated nodules that tend to co- HIV/AIDS. A recent report suggested that PJP in pa-
alesce and occur in the perihilar lung and lower lobes. tients with HIV/AIDS as opposed to non-HIV/AIDS pa-
On CT, the distribution is along bronchovascular path- tients might have a higher incidence of pneumatoceles,
ways. but less extensive ground-glass opacities (Hardak et al.,
(Other diseases seen in HIV/AIDS patients which oc- 2010).
cur along the bronchovascular bundles include lym- Cryptococcus has numerous types of presentation on
phocitic interstitial pneumonia (LIP), Castleman disease, chest film. Perhaps most common is the appearance of
and sarcoidosis). In almost all cases of Kaposis sarcoma well-defined nodules, usually solitary but sometimes
involving the lungs, cutaneous lesions are also common, multiple. If the nodules become masses, the margins may
as is pleural fluid. become indistinct. The nodules may cavitate. Cryptococ-
cus may also manifest as a lobar pneumonia or diffuse
Other Conditions of the Immune-Compromised heterogeneous reticulonodular opacities.
Aspergillus fumigatus is responsible for many lung
Increasing numbers of transplant procedures, both sol- infections. Up to 10% of pneumonias following allo-
id organ and hematopoietic stem cell, have led to new geneic transplantation are due to Aspergillus. The pat-
populations of immunosuppressed individuals. The un- tern of abnormality seen on chest film depends on the
derlying diseases (e.g. leukemia) or the widespread use patients immune status. In the setting of immunosup-
of induced immunosuppression for treatment purposes pression (neutropenia), the typical appearance is that of
may result in neutropenia or other causes of immune invasive aspergillosis. In this form, the chest film ini-
dysfunction. Steroids are also being used with in- tially demonstrates a poorly marginated area or areas of
creased frequency for a number of medical conditions. homogeneous opacity that may resemble ordinary bac-
Thus, infectious and non-infectious complications in terial pneumonia in appearance and distribution but is
the setting of transplantation or steroid use have be- occasionally rounder and farther from the subpleural
come a major problem. Prophylactic drug treatment for lung than common community infections. In some cas-
pneumocystis, CMV, and occasional fungi may reduce es, the disease is peripheral and wedge-shaped sec-
the number of infections in some, but not all of these ondary to infarction caused by the angio-invasive ob-
patients. struction of pulmonary vessels. In time, the lesions be-
Bacterial pneumonias caused by a variety of organ- come more discreet and rounder, thus resembling lung
isms (e.g., Pseudomonas, Nocardia, Legionella) have a masses.
typical appearance of peripheral homogeneous opacifi- As patients are treated and immune status improves,
cation with or without air-bronchograms. More than one there may be cavitation within the masses with the for-
lobe might be involved. In the case of Legionella the mation of an air crescent. Wall thickness of the cavity is
opacification may simulate a mass. generally moderate. The air crescent is created by the
The appearances on chest film and CT scans of CMV contained necrotic debris within the cavity. On CT, ini-
will be similar to that seen in patients with HIV infec- tially, the areas of homogeneous opacity may have air
tion, as described above. Other viral pathogens are also bronchograms, and commonly, additional regions of in-
seen in this setting, including RSV, parainfluenza virus, volvement are identified. A ground glass opacity that
adenovirus, and influenza virus in lung transplant pa- surrounds (frequently incompletely) a more opaque cen-
tients, and varicella zoster, which may be seen in pa- ter of the lesion results in the halo sign, in which the
tients with lymphoma and those undergoing steroid ground glass portion is an area of hemorrhage and the
therapy. central area is necrotic lung. This sign was thought to be
These same organisms are responsible for infections pathognomonic of invasive aspergillosis but it is also a
in patients who develop graft versus host disease. Many feature of other infections, neoplasms, and inflammato-
of these viral pneumonias have a similar appearance, in- ry diseases. CT may also reveal bronchial wall thicken-
cluding ground-glass and consolidative opacities, cen- ing, peribronchial opacities, and small centrilobular
trilobular and tree-in-bud opacities, and a frequently bi- nodules.
68 P. Goodman, H. Prosch, C.J. Herold
Take-Home Messages: Usefulness of Imaging Methods British Thoracic Society Standards of Care Committee (2001)
BTS guidelines for the management of community acquired
in Pulmonary Infections pneumonia in adults. Thorax Suppl 4: IV1-IV64.
Carmona EM, Limper AH (2010) Update on the diagnosis and
Despite the increasing use of CT imaging for diagnos- treatment of Pneumocystis pneumonia. Ther Adv Respir Dis
ing chest disorders, plain film radiography remains the 5:41-59.
primary imaging modality for patients with suspected Choi MH, Jung JI, Chung WD et al (2014) Acute complications
in patients with hematologic malignancies. RadioGraphics
pneumonia. The presence of an appropriate lung opaci- 34:1755-1768.
ty on a chest radiograph is considered the gold standard Chou S-H S, Prabhu SJ, Crothers K et al (2014) Thoracic diseases
for diagnosing pneumonia. Extensive knowledge of the associated with HIV infection in the era of anti-retroviral ther-
radiographic appearances of pulmonary infections, their apy: clinical and imaging findings. RadioGraphics 34:895-
complications, and their course is essential in aiding the 911.
referring clinician and, ultimately, the patient. CT imag- Conees DJ (1999) Endemic fungal pneumonia in immunocom-
promised patients. J Thorac Imaging 14:1-8.
ing is useful in patients with CAP and NP when there is Connolly JE Jr, McAdams HP, Erasmus JJ et al (1999) Oppor-
a non-resolving or complicated chest film and at times tunistic fungal pneumonia. J Thorac Imaging 14:51-62.
in immunocompromised patients with suspected pul- Coy DL, Ormazabal A, Godwin JD et al (2005) Imaging evalua-
monary infections. CT can help differentiate infectious tion of pulmonary and abdominal complications following
from noninfectious abnormalities. CT may detect hematopoietic stem cell transplantation. RadioGraphics
25:305-318.
empyema, cavitation, and lymphadenopathy when chest Crothers K, Huang L, Goutlet JL et al (2010) HIV infection and
films cannot. CT should be performed in immunocom- risk for incident pulmonary diseases in the combination anti-
promised patients with a clinical suspicion of pneumo- retroviral therapy era. Am J Respir Crit Care Med [Epub
nia when the chest film is normal. This is especially true ahead of print].
when the early diagnosis of pneumonia is critical, as is Donnelly LF (1999) Maximizing the usefulness of imaging in
children with community-acquired pneumonia. AJR Am J
the case with immunocompromised and severely ill pa- Roentgenol 172:505-512.
tients. Ellis SM (2004) The spectrum of tuberculosis and non-tubercu-
To reiterate: No pattern of abnormality seen on chest lous mycobacterial infection. Eur Radiol 14(Suppl 3):E34-
films can be considered pathognomonic of a specific in- E42.
fection. However, the distribution and appearance of Escuissato DL, Gasparetto EL, Marchiori E et al (2005) Pul-
monary infections after bone marrow transplantation: high-
lung opacities, especially in conjunction with clinical in- resolution CT findings in 111 patients. AJR Am J Roentgenol
formation, should enable one to produce a useful, or- 185:608-615.
dered list of most likely possibilities helpful to our clin- Fang GD, Fine M, Orloff J et al (1990) New and emerging etiolo-
ical colleagues and, most importantly, to our patients. gies for community-acquired pneumonia with implications
for therapy. A prospective multicenter study of 359 cases.
Medicine (Baltimore) 69:307-316.
Franquet T, Gimenez A, Hidalgo A (2004) Imaging of oppor-
tunistic fungal infections in immunocompromised patients.
Suggested Reading Eur J Radiol 51:130-138.
French MA (2009) Immune reconstitution inflammatory syn-
Ahuja J, Kanne JP (2014) Thoracic infections in immunocompro- drome: a reappraisal. Clin InfectDis 48:101-107.
mised patients. Radiol Clinics of North Am 52:121-136. Goodman P (2006) Radiographic assessment of HIV-related dis-
Albaum MN, Hill LC, Murphy M (1996) Interobserver reliability eases. In: Peiperl L, Volberding P (eds) HIV InSite Knowl-
of the chest radiograph in community-acquired pneumonia. edge Base. CD-ROM version.
Chest 110:343. Goodman PC (2007) Pulmonary infection in adults. In: Grainger
Aquino SL, Dunagan DP, Chiles C, Haponik EF (1998) Herpes RG, Allison DJ, Dixon AK (eds) Grainger & Allisons diag-
simplex virus 1 pneumonia: patterns on CT scans and con- nostic radiology: a textbook of medical imaging, 5th Ed.
ventional chest radiographs. J Comput Assist Tomogr 22: Churchill Livingstone, London, New York, NY.
795-800. Hansell DM, Armstrong P, Lynch DA, McAdams HP (2005) The
Aviram G, Boiselle PM (2004) Imaging features of bacterial res- immunocompromised patient. In: Rozenshtein A (ed) Imag-
piratory infections in AIDS. Curr Opin Pulm Med 10: ing of diseases of the chest, 4th Ed. Elsevier Mosby, Philadel-
183-188. phia, PA.
Bartlett JG, Dowell SF, Mandell LA et al (2000) Practice guide- Hardak E, Brook O, Yigla M (2010) Radiological features of
lines for the management of community-acquired pneumonia Pneumocystis jirovecii pneumonia in immunocompromised
in adults. Infectious Diseases Society of America. Clin Infect patients with and without AIDS. Lung 188:159-163.
Dis 31:347-382. Herold CJ, Sailer JG (2004) Community-acquired pneumonia and
Bartziokas K, Daenas C, Preau S et al (2010) Vibration Response nosocomial pneumonia. Eur Radiol 14:E2-E20.
Imaging: evaluation of rater agreement in healthy subjects and Heussel CP, Kauczor HU, Ullmann AJ (2004) Pneumonia in neu-
subjects with pneumonia. BMC Medical Imaging 10:6. tropenic patients. Eur Radiol 14:256-271.
Basi SK, Marrie TJ, Huang JQ, Majumdar SR (2004) Patients ad- Huang L, Cattamanchi A, Davis JL et al (2011) HIV-Associated
mitted to hospital with suspected pneumonia and normal pneumocystis pneumonia. Proc Am Thorac Soc 8:294-300.
chest radiographs: epidemiology, microbiology, and out- Huang L, Crothers K (2009) HIV-associated opportunistic pneu-
comes. Am J Med 117:305. monias. Respirology 14:474-485.
Bierry G, Boileau J, Barnig C et al (2009) Thoracic manifestations Ibrahim EH, Traey L, Hill C (2002) The occurrence of ventilator-
of primary humoral immunodeficiency: a comprehensive re- associated pneumonia in a community hospital: risk factors
view. Radiographics 29:1909-1920. and clinical outcomes. Radiology 223:884.
Imaging of Pulmonary Infections 69
Jeong YJ, Lee KS (2008) Pulmonary tuberculosis: up-to-date Metlay JP, Fine MJ (2003) Testing strategies in the initial man-
imaging and management. AJR Am J Roentgenol 191:834- agement of patients with community-acquired pneumonia.
844. Ann Intern Med 138:109-118.
Johnstone J, Majumdar SR, Fox JD, Marrie TJ (2008) Viral infec- Metlay JP, Kapoor WN, Fine MJ (1997) Does this patient have
tion in adults hospitalized with community-acquired pneumo- community-acquired pneumonia? Diagnosing pneumonia
nia: prevalence, pathogens, and presentation. Chest 134:1141- by history and physical examination. JAMA 278:1440-
1148. 1445.
Jung JI, Kim H, Park SH et al (2001) CT differentiation of pneu- Mortensen EM, Copeland LA, Pugh MJ et al (2010) Diagnosis of
monic-type bronchioloalveolar cell carcinoma and infectious pulmonary malignancy after hospitalization for pneumonia.
pneumonia. Br J Radiol 74:490-494. Am J Med 123:66-71.
Kim EA, Lee KS, Primack SL et al (2002) Viral pneumonias in Musher DM, Thorner AR (2014) Community-acquired pneumo-
adults: radiologic and pathologic findings. RadioGraphics nia. N Engl J Med 371:1619-1628.
22:S137. Nambu A, Ozawa K, Kobayashi N, Tago M (2014) Imaging of
Kollef MH, Shorr A, Tabak YP et al (2005) Epidemiology and out- community-acquired pneumonia: Roles of imaging exami-
comes of health-care-associated pneumonia: results from a nations, imaging diagnosis of specific pathogens and dis-
large US database of culture-positive pneumonia. Chest crimination from noninfectious diseases. World J Radiol
128:3854-3862. 6:779-793.
Krishnam MS, Suh RD, Tamasian A et al (2007) Postoperative Niederman MS, Mandell LA, Anzueto A et al (2001) American
complications of lung transplantation: radiologic findings Thoracic Society. Guidelines for the management of adults
along a time continuum. RadioGraphics 27:957-974. with community-acquired pneumonia. Diagnosis, assessment
Lacombe C, Lewiin M, Monnier-Cholley L et al (2007) Imaging of severity, antimicrobial therapy, and prevention. Am J
of thoracic pathology in patients with AIDS. J Radiol Respir Crit Care Med 163:1730-1754.
88:1145-1154. Pupaibool J, Limper AH (2013) Other HIV-associated pneumo-
Leung AN, Brauner MW, Gamsu G et al (1996) Pulmonary tuber- nias. Clin Chest Med 34:243-254.
culosis: comparison of CT findings in HIV-seropositive and Primack SL, Hartmann TE, Lee KS, Mueller NL (1994) Pul-
HIV-seronegative patients. Radiology 198:687. monary nodules and the CT halo sign. Radiology 190:
Leung AN (1999) Pulmonary tuberculosis: the essentials. Radio- 513-515.
logy 210:307-322. Reeders JWAJ, Goodman PC (2001) Differential radiological pat-
Lichtenberger III JP, Sharma A, Zachary KC et al (2012) What terns in AIDS at a glance. In: Reeders JWAJ , Goodman PC
a differential a virus makes: a practical approach to tho- (eds) Radiology of AIDS, Springer-Verlag, Heidelberg, pp
racic imaging findings in the context of HIV infection 293-324.
Part 1, Pulmonary f indings. AJR Am J Roentgenol Reittner P, Mller NL, Heyneman L et al (2000) Mycoplasma
198:1295-1304.
pneumoniae pneumonia: radiographic and high-resolution
Lichtenberger III JP, Sharma A, Zachary KC et al (2012) What a
CT. Features in 28 patients. AJR Am J Roentgenol 174:
differential a virus makes: A practical approach to thoracic
37-41.
imaging findings in the context of HIV infection Part 2, Ex-
Remund KF, Best M, Egan JJ (2009) Infections relevant to lung
trapulmonary findings, chronic lung disease, and immune re-
constitution syndrome. AJR Am J Roentgenol 198:1305- transplantation. Proc Am Thorac Soc 6:94-100.
1312. Reynolds JH, McDonald G, Alton H, Gordon SB (2010) Pneumo-
Lieberman D, Shvartzman P, Korsonsky I, Lieberman D (2003) nia in the immunocompetent patient. Br J Radiol 83:998-
Diagnosis of ambulatory community-acquired pneumonia. 1009.
Comparaison of clinical assessment versus chest X-ray. Scand Shah RM, Wechsler R, Salazar AM, Spirn PW (1997) Early
J Prim Health Care 21:57-60. detection of pneumonia in febrile neutropenic patients: use of
Lim WS, Baudouin SV, George RC et al (2009) British Thoracic thin-section CT. AJR Am J Roentgenol 169:1347-1353.
Society guidelines for the management of community ac- Sharma S, Maycher B, Eschun G (2007) Radiological imaging in
quired pneumonia in adults: update 2009. Thorax 64:1-55. pneumonia: recent innovations. Curr Opin Pulm Med 13:
Little BP, Gilman MD, Humphrey KL et al (2014) Outcome of 159-69.
recommendations for radiographic follow-up of pneumonia Sider L, Gabriel H, Curry DR, Pham MS (1993) Pattern recogni-
on outpatient chest radiography. AJR Am J Roentgenol tion of the pulmonary manifestations of AIDS on CT scans.
202:54-59. RadioGraphics 13:771-784.
Mabie M, Wunderink RG (2003) Use and limitations of clinical Speets AM, Hoes AW, van der Graal Y et al (2006) Chest radio-
and radiologic diagnosis of pneumonia. Semin Respir Infect graphy and pneumonia in primary care: diagnostic yield and
18:72-79. consequences for patient management. Eur Resp J 28:933-
Mandell LA, Bartlett JG, Dowell SF et al (2003) Update of prac- 938.
tice guidelines for the management of community-acquired Syrjala H, Broas M, Suramo I et al (1998) High-resolution com-
pneumonia in immunocompetent adults. Clin Infect Dis puted tomography for the diagnosis of community-acquired
37:1405. pneumonia. Clin Infect Dis 27:358.
Mandell LA, Wunderink RG, Anzueto A et al (2007) Infectious Valles J, Martin Loeches I, Torres A et al (2014) Epidemiology an-
Diseases Society of America/American Thoracic Society con- tibiotic therapy and clinical outcomes of healthcare-associat-
sensus guidelines on the management of community-acquired ed pneumonia in critically ill patients: a Spanish cohort study.
pneumonia in adults. Clin Infect Dis 44 (Suppl 2):S27-S72. Intensive Care Med 40:572-581.
Marchiori E, Zanetti G, Hochhegger B et al (2012) Reversed ha- Vento S, Cainelli F, Temesgen Z (2008) Lung infections after can-
lo sign on computed tomography: State-of-the-art review. cer themotherapy. Lancet Oncol 9:982-992.
Lung 190: 389-394. Wagner AL, Szabunio M, Hazlett KS et al (1998) Radiologic man-
Melbye H (2002) Community pneumonia more help is needed ifestations of round pneumonia in adults. AJR Am J
to diagnose and assess severity. Br J Gen Pract 52:886-888. Roentgenol 170:723.
Melbye H, Dale K (1992) Interobserver variability in the radio- Winer-Muram HT, Rubin SA, Ellis JV et al (1993) Pneumonia and
graphic diagnosis of adult outpatient pneumonia. Acta Radiol ARDS in patients receiving mechanical ventilation: diagnos-
33:79-83. tic accuracy of chest radiograph. Radiology 188:479-485.
70 P. Goodman, H. Prosch, C.J. Herold
Wunderink RG, Waterer GW (2014) Clinical practice. Communi- Yeh JJ, Chen SC, Chen CR et al (2014) A high-resolution com-
ty-acquired pneumonia. N Engl J Med 370:543-551. puted tomography-based scoring system to differentiate the
Wunderink RG, Woldenberg LS, Zeiss J et al (1992) The radio- most infectious active pulmonary tuberculosis from commu-
logic diagnosis of autopsy proven ventilator-associated pneu- nity-acquired pneumonia in elderly and non-elderly patients.
monia. Chest 101:458-463. Eur Radiol 24:2372-2384.
IDKD 2015-2018
a b
Fig. 1 a-c. Minimal aortic injury. A linear filling defect, likely rep-
resenting an intimal flap (arrows), is seen on axial computed to-
mography (CT) images (a) and on coronal reformats (b) in the de-
scending thoracic aorta in a 38-year-old man who was involved in
a motor vehicle accident. Three days after the accident, the intimal
flap appears more rounded (c), possibly due to thrombus forma-
tion. On a follow-up CT prior to hospital discharge 3 weeks later,
the intraluminal findings had resolved without intervention
adventitia. Adventitial tears may involve 360o of the aor- complete rupture of the descending thoracic aorta, medi-
tic wall, so that the pseudoaneurysm appears fusiform or astinal tissues and the periaortic hematoma can produce
barrel-shaped (Fig. 2). When the tear involves only a por- a tamponade effect, preventing rapid exsanguination.
tion of the adventitia, the pseudoaneurysm bulges eccen- This is in contrast to free rupture involving the ascending
trically from the aorta, typically in the region of the aor- aorta, when blood can rapidly fill the pericardial sac,
tic isthmus. causing cardiac tamponade and death. It is important to
The stress placed on the aortic wall during sudden de- recognize free rupture on imaging and to distinguish it
celeration may result in complete rupture, with disruption from contained rupture. Although delayed repair may be
of all three layers of the aortic wall intima, media, and considered in patients with aortic pseudoaneurysm, or
adventitia with extravasation of blood into the adjacent contained rupture, immediate repair is necessary for pa-
anatomic compartment (Fig. 3). In some patients with tients with free aortic rupture.
Imaging of Thoracic Trauma 73
Cardiac Herniation
Cardiac herniation as a result of a traumatic tear of the
pericardium is an uncommon occurrence but may be con-
sidered in the patient with fluctuating blood pressures af-
ter intravenous fluid administration together with tachy-
cardia and dilated jugular veins [7]. Although cardiac-
axis deviation may be apparent on a chest radiograph, more
specific signs of cardiac herniation are better seen on CT
and include the empty pericardial sac, as air fills the
space remaining once the heart is displaced into the
hemithorax, and the collar sign, resulting from the peri-
cardium pinching the heart as it herniates through the
pericardial tear. The tear is typically a longitudinal one
located along the pericardiophrenic nerve and occurs
more commonly on the left side. The mortality of patients
with cardiac herniation is high, with death resulting from
Fig. 2. Contained aortic rupture. A fusiform pseudoaneurysm has torsion along the inferior vena cava and great vessels and
formed in the proximal descending thoracic aorta. This occurs
when the intima and media are torn and the pseudoaneurysm is cardiac strangulation [8].
contained by the intact adventitia
Hemopericardium
Hemopericardium should raise suspicion of injury to the
heart, ascending aorta, or coronary arteries. It is most
commonly associated with injury to the aortic root and
ascending aorta but may also result from cardiac cham-
ber rupture, myocardial contusion, and coronary artery
laceration [8]. Simple fluid should demonstrate attenua-
tion values of 010 Hounsfield units (HU) whereas he-
mopericardium will have attenuation values in the 35-HU
range. The rapid accumulation of hemopericardium may
result in cardiac tamponade. The latter diagnosis in pa-
tients with blunt chest trauma is suggested when CT
demonstrates a triad of high-attenuation pericardial effu-
sion, distention of the inferior vena cava and renal veins,
and periportal low-attenuation fluid [8].
Esophageal Injury
Esophageal injuries caused by external, non-iatrogenic
trauma are rare but potentially catastrophic. Mortality for
undiagnosed rupture after 24 h is 1040%. Chest radio-
graphs may reveal pleural effusions, left sided if the low-
er third of the esophagus is injured, right sided if the up-
Fig. 3. Free rupture. The extravasation of contrast medium from the
transverse thoracic aorta is diagnostic of injury involving all three per two thirds is injured. Pneumomediastinum is also
layers of the aortic wall. Mediastinal hematoma is also present, an commonly seen but is nonspecific and usually not related
indirect sign of aortic injury to esophageal or tracheobronchial injury (see below):
74 L. Ketai, C. Chiles
a b
Fig. 4 a, b. Two patient with acute diffuse lung disease. a Acute respiratory distress syndrome (ARDS) with dependent consolidation and
midlung ground glass opacities. The pneumomediastinum is due to the Macklin effect (arrow). b Trauma with acute aspiration causing
tree in bud opacities and secretions in segmental bronchi (arrow)
Tracheobronchial Rupture
Tracheobronchial injury as a result of blunt-force trauma
Fig. 5. Bronchial rupture. The lung lies in a dependent position
is uncommonly seen, as the majority of patients die be- within the right hemithorax, consistent with the fallen lung sign.
fore reaching the hospital. In a review of 88 surviving pa- Additional evidence of bronchial rupture in this case includes dis-
tients reported in the literature, the rupture involved the ruption of the bronchus, extensive subcutaneous emphysema, and
right main bronchus in 41 (47%), and 76% of the injuries a large pneumothorax despite satisfactory positioning of a thora-
were within 2 cm of the carina [11]. The median time un- cotomy tube
til diagnosis was 9 days. Chest radiography may show in-
direct signs of tracheobronchial rupture, including cervi- Although originally described on chest radiography, this
cal subcutaneous emphysema, pneumomediastinum, and can also be seen on CT, which may demonstrate disrup-
persistent pneumothorax despite satisfactory positioning tion in the wall of the trachea or bronchus; however, this
of a thoracotomy tube [12]. The fallen lung sign de- is more readily identified at bronchoscopy, which should
scribes a collapsed lung lying in a dependent position, be performed in any patient with suspected tracheo-
tethered to the hilum by pulmonary vessels (Fig. 5) [13]. bronchial injury.
Imaging of Thoracic Trauma 75
Hemothorax
Hemothorax occurs in 3050% of blunt injuries but more
commonly requires surgical intervention following pene-
trating trauma. On supine radiographs a posterior layer-
ing hemothorax can cause hazy opacity of the entire Fig. 6. Extrapleural hematoma. Blood has accumulated in the ex-
hemithorax through which the pulmonary vasculature is trapleural space following blunt force trauma to the left chest wall.
Extrapleural fat is displaced inwardly (arrow). The biconvex shape
visible. Hemothorax is more easily detected when it is ac- and the high-attenuation material within the hematoma suggest an
companied by a rim of fluid displacing the lung from the active arterial bleeding source, which may require embolization or
lateral and apical chest wall. Hemothorax is readily visible surgical intervention
76 L. Ketai, C. Chiles
Collections that are biconvex may represent an arterial that indirectly damage the lung. Other insults from trau-
bleeding source and are more likely to require surgical in- ma directly injure the lung parenchyma; these include as-
tervention [20]. piration, fat emboli, inhalation injury, near drowning, and
Chylothorax is an uncommon complication of chest smoke inhalation. When the extent of injury is severe,
trauma. It can result from penetrating trauma to the up- damage caused by most mechanisms has a similar radio-
per mediastinum that injures the thoracic duct as it as- logic appearance. CT often demonstrates consolidation in
cends to the left of the esophagus, resulting in a left sided the most dorsal lung, ground glass opacities in the mid
pleural effusion [21]. Blunt injury to the thoracic duct is lung, and nearly normal appearing lung along the ventral
often related to hyperextension of the spine and occurs surfaces (Fig. 4). Direct pulmonary insults may cause
most commonly at the level of the diaphragm, where the more extensive and asymmetric consolidation than indi-
duct is located in the right hemithorax, thus giving rise to rect injuries.
a right pleural effusion. Pleural fluid from a chylothorax In early stages or in less severe cases of alveolar dam-
may be lower in attenuation than water due to its fat con- age, the clinical history and imaging findings may suggest
tent but in many cases co-existing protein in the fluid a specific cause. For instance, in cases of acute aspiration
raises its attenuation to that of other pleural effusions. pneumonitis, tree-in-bud type centrilobular nodules may
co-exist with CT findings of diffuse alveolar damage [24].
In early or mild cases of fat embolization, CT may demon-
Pulmonary Injury strate ill-defined nodules 1 cm in size in addition to
ground glass opacities and consolidation [25]. The diag-
Contusion nosis of fat emboli, however, is not made radiologically.
The presence of bilateral parenchymal opacities is a ma-
Pulmonary contusions are defined as the leakage of jor diagnostic criteria but confirmation requires the pres-
blood into the alveoli and lung interstitium without the ence of non-radiologic major criteria (e.g., axillary or
presence of a discrete laceration. CT is more sensitive subconjunctival petechiae) or multiple minor criteria
than radiography but many of the contusions seen on CT (e.g., fat within the urine, significant hypoxemia).
alone are not clinically significant. In more extensive in- CT may be more useful in assessing prognosis than in
jury, CT is useful in differentiating contusion from aspi- identifying the etiology of lung injury. For example, in
ration or acute respiratory distress syndrome as the etiol- burn victims, CT imaging can identify those patients
ogy of acute lung injury, by showing lung opacities that most at risk for diffuse lung damage related to smoke in-
have geographic borders and cross pleural fissures. These halation. The addition of bronchoscopic inspection im-
opacities are subpleural but may demonstrate immediate proves prediction, and at least in theory could be replaced
subpleural sparing, particularly in children. If the opaci- by virtual bronchoscopy [26]. More commonly, CT can
ties continue to worsen after 24 h or do not begin to im- be useful in assessing the prognosis of patients with al-
prove after 4872 h, concurrent volume overload, atelec- ready established diffuse alveolar damage. Development
tasis, or super-infection should be suspected. of fibroproliferative changes on CT predicts prolonged
mechanical ventilation, the development of barotrauma
Laceration and ventilator-associated pneumonia [27].
a b
Fig. 8 a, b. Penetrating diaphragm injury. CT sections in a patient stabbed through the anterior chest wall. The trajectory of the knife (arrows)
can be followed inferiorly through the lung and into the liver parenchyma. The course is indicative of a focal laceration of the diaphragm
78 L. Ketai, C. Chiles
Fracture of the chest wall bones can be clinically signif- Sternoclavicular dislocations are clinically relevant due
icant because of their correlation with overall injury to potential direct damage to adjacent structures. These
severity or because of the direct anatomic or physiologic injuries are rare because of the extensive ligamentous
impact of the fracture (Table 2). Fractures of the scapula support, such that forces transmitted along the clavicle
are an example of the former, their presence correlating are far more likely to fracture the clavicle than dislocate
with more severe chest injury. The association is stronger the joint. Anterior dislocations are more common but
if more than two separate regions of the scapula are frac- posterior dislocations are of greater clinical importance
tured (e.g., scapular body and glenoid neck, scapular due to potential impingement on adjacent structures. Di-
spine, and acromion). agnosis of this dislocation without CT is very difficult,
even if special radiographic views (e.g., serendipity view)
are taken [33].
Sternal and Rib Trauma The sternoclavicular joint may also be disrupted in
Sternal fractures are also associated with the severity of scapulothoracic disassociations; a severe injury separates
chest injury, the degree of displacement being the major the scapula and muscular attachments of the upper ex-
determinant. Less than 10% of patients with minimally tremity from the thorax but without disrupting the over-
displaced sternal fractures have spinal fractures or trau- lying skin. Radiographic diagnosis relies on detecting
matic pericardial effusions, but the incidence is much scapular displacement via the scapula index, a measure-
higher in patients with completely displaced sternal frac- ment easily distorted by differences in arm positioning
tures. The location of the fracture within the sternum is and by rotation [34]. In suspected cases CT and magnet-
also significant: those involving the manubrium have the ic resonance imaging are essential for detecting vascular
highest association with spinal fractures [31]. Sterno- and nerve injuries, respectively.
manubrial dislocations share this association with tho-
racic spine trauma, and specifically type 2 dislocations, References
in which hyperflexion displaces the manubrium posteri-
orly with respect to the sternal body. Type 1 sterno- 1. World Health Organization (2010) Injuries and violence: The facts.
manubrial dislocations are those in which the sternal http://whqlibdoc.who.int/publications/2010/9789241599375
body is displaced posteriorly with respect to the manubri- _eng.pdf
um and are usually the result of a direct blow. 2. Chung JH, Cox CW, Mohammed TL et al (2014) ACR appropri-
ateness criteria blunt chest trauma. J Am Coll Radiol 11:345-351.
Rib fractures have prognostic significance and can al- 3. Azizzadeh A, Valdes J, Miller CC 3rd et al (2011) The utility of
so cause direct physiologic impairment. Mortality in- intravascular ultrasound compared to angiography in the diag-
creases slightly in patients with two or more rib fractures nosis of blunt traumatic aortic injury. J Vasc Surg 53:608-614.
visible on CT but is high when nine or more are visible. 4. Paul JS, Neideen T, Tutton S et al (2011) Minimal aortic injury
Flail chests, defined as fractures of three or four contigu- after blunt trauma: selective nonoperative management is safe.
J Trauma 71:1519-1523.
ous ribs in two places, can result in paradoxical motion 5. Gunn ML, Lehnert BE, Lungren RS et al (2014) Minimal
of a segment of the chest wall, resulting in hypoventila- aortic injury of the thoracic aorta: imaging appearances and
tion. Rib fractures now more commonly undergo internal outcome. Emerg Radiol 21:227-233.
fixation than in the past, although the indications and 6. Rojas CA, Restrepo CS (2009) Mediastinal hematomas: aortic
injury and beyond. J Comput Assist Tomogr 33:218-224.
timing of surgery remain controversial. Patients with a 7. Nassiri N, Yu A, Statkus N, Gosselin M (2009) Imaging of
flail chest who require ventilator support but do not have cardiac herniation in traumatic pericardial rupture. J Thorac
a severe contusion are most likely to benefit [32]. Imaging 24:69-72.
Imaging of Thoracic Trauma 79
8. Restrepo CS, Gutierrez FR, Marmol-Velez JA (2012) Imaging 22. Sangster GP, Gonzalez-Beicos A, Carbo AI et al (2007) Blunt
patients with cardiac trauma. Radiographics 32:633-649. traumatic injuries of the lung parenchyma, pleura, thoracic
9. Fadoo F, Ruiz DE, Dawn SK et al (2004) Helical CT esopha- wall, and intrathoracic airways: multidetector computer to-
gography for the evaluation of suspected esophageal perfora- mography imaging findings. Emerg Radiol 14:297-310.
tion or rupture. AJR Am J Roentgenol 182:1177-1179. 23. Chou YP, Kuo LC, Soo KM et al (2014) The role of repairing
10. Suarez-Poveda T, Morales-Uribe CH, Sanabria A et al (2014) lung lacerations during video-assisted thoracoscopic surgery
Diagnostic performance of CT esophagography in patients evacuations for retained haemothorax caused by blunt chest
with suspected esophageal rupture. Emerg Radiol 21:505-510. trauma. Eur J Cardiothorac Surg 46:107-111.
11. Kiser AC, OBrien SM, Detterbeck FC (2001) Blunt tracheo- 24. Prather AD, Smith TR, Poletto DM et al (2014) Aspiration-
bronchial injuries: treatment and outcomes. Ann Thorac Surg related lung diseases. J Thorac Imaging 29:304-309.
71:2059-2065. 25. Gallardo X, Castaner E, Mata JM et al (2006) Nodular pattern
12. Kunisch-Hoppe M, Hoppe M, Rauber K et al (2000) Tracheal at lung computed tomography in fat embolism syndrome: a
rupture caused by blunt chest trauma: radiological and clinical helpful finding. J Comput Assist Tomogr 30:254-257.
features. Eur Radiol 10:480-483. 26. Koljonen V, Maisniemi K, Virtanen K, Koivikko M (2007)
13. Wintermark M, Schnyder P, Wicky S (2001) Blunt traumatic Multi-detector computed tomography demonstrates smoke in-
rupture of a mainstem bronchus: spiral CT demonstration of halation injury at early stage. Emerg Radiol 14:113-116.
the fallen lung sign. Eur Radiol 11:409-411. 27. Ichikado K, Muranaka H, Gushima Y et al (2012) Fibropro-
14. Brar MS, Bains I, Brunet G et al (2010) Occult pneumotho- liferative changes on high-resolution CT in the acute respira-
races truly occult or simply missed: redux. J Trauma 69:1335- tory distress syndrome predict mortality and ventilator depen-
1337. dency: a prospective observational cohort study. BMJ Open
15. Hyacinthe AC, Broux C, Francony G et al (2012) Diagnostic 2:e000545.
accuracy of ultrasonography in the acute assessment of com- 28. Desir A, Ghaye B (2012) CT of blunt diaphragmatic rupture.
mon thoracic lesions after trauma. Chest 141:1177-1183. Radiographics 32:477-498.
16. Yarmus L, Feller-Kopman D (2012) Pneumothorax in the 29. Sliker CW (2006) Imaging of diaphragm injuries. Radiol Clin
critically ill patient. Chest 141:1098-1105. North Am 44:199-211.
17. Bastos R, Baisden CE, Harker L, Calhoon JH (2008) Pene- 30. Dreizin D, Borja MJ, Danton GH et al (2013) Penetrating di-
trating thoracic trauma. Semin Thorac Cardiovasc Surg 20: aphragmatic injury: accuracy of 64-section multidetector CT
19-25. with trajectography. Radiology 268:729-737.
18. Shanmuganathan K, Matsumoto J (2006) Imaging of penetrat- 31. von Garrel T, Ince A, Junge A et al (2004) The sternal fracture:
ing chest trauma. Radiol Clin North Am 244:225-238, viii. radiographic analysis of 200 fractures with special reference to
19. Gignon L, Charbit J, Maury C et al (2014) A simple assess- concomitant injuries. J Trauma 57:837-844.
ment of haemothoraces thickness predicts abundant transfu- 32. Leinicke JA, Elmore L, Freeman BD, Colditz GA (2013) Op-
sion: A series of 525 blunt trauma patients. Injury. doi: erative management of rib fractures in the setting of flail chest:
10.1016/j.injury.2014.08.040 [Epub ahead of print]. a systematic review and meta-analysis. Ann Surg 258:914-921.
20. Chung JH, Carr RB, Stern EJ (2011) Extrapleural hematomas: 33. Gyftopoulos S, Chitkara M, Bencardino JT (2014) Misses and
imaging appearance, classification, and clinical significance. errors in upper extremity trauma radiographs. AJR Am J
J Thorac Imaging 26:218-223. Roentgenol 203:477-491.
21. Seitelman E, Arellano JJ, Takabe K et al (2012) Chylothorax 34. Brucker PU, Gruen GS, Kaufmann RA (2005) Scapulothoracic
after blunt trauma. J Thorac Dis 4:327-330. dissociation: evaluation and management. Injury 36:1147-1155.
IDKD 2015-2018
The detection of carcinoma on a chest radiograph re- standing of the overlooked lesion. There is no harm done
mains difficult, with implications for patient manage- by learning from ones mistakes!
ment. Nowadays, missed detection of a lung carcinoma is
by far the most frequent cause of malpractice suits (42%
of cases) [4]. Nonetheless, whereas overlooking chronic Missed Nodules
airways disease, pneumonia, and interstitial lung disease
may not have the same potential medico-legal implica- Nodular Lesions: Tumors
tions, the consequences for patient care could be critical.
By seeking a correlation between MDCT and plain Nodular lesions are frequently due to lung cancer, which
chest radiograph with respect to missed lung lesions, in- may be primary or secondary. Lung cancer is probably one
terpretation of the latter imaging modality can be im- of the most common lung diseases that radiologists en-
proved. During the course of clinical work, when report- counter in practice. Berbaum formulated the concept that
ing chest CT, whenever available, every effort should be perception is better if you know where to look and what to
made to review previous chest radiographs and their re- look for [33]. Our first example is that of a 53-year-old
ports, as this provides one of the best learning tools for man who complained of pain in the right axilla for 4
chest radiograph interpretation. months and underwent chest radiography. The postero-
A CT scan can be performed in patients with a nega- anterior and lateral radiographs were interpreted as showing
tive chest radiograph when there is a high clinical suspi- normal findings (Fig. 1a, 1b). Subsequent MDCT showed
cion of chest disease. CT scan, especially high-resolution a right superior sulcus mass with rib destruction (Fig. 1c,
CT (HRCT), is more sensitive than plain films for the d). Needle biopsy established a diagnosis of bronchogenic
evaluation of interstitial disease, bilateral disease, cavita- carcinoma (adenocarcinoma). Hindsight bias [34] with the
tion, empyema, and hilar adenopathy. CT is not general- information available from the MDCT makes the initial le-
ly recommended for routine use because the data it pro- sion extremely obvious. Careful scrutiny of both apices is
vides in chronic airways disease and pneumonia are lim- essential when reporting a frontal chest radiograph.
ited, the cost of imaging is high, and there is no evidence Radiological errors can be divided into two types
that outcome is improved. Thus, a chest radiograph is the [35]: cognitive, in which an abnormality is seen but its
preferred method for initial imaging, with CT scan re- nature is misinterpreted, and perceptual, or the miss, in
served for further characterization (e.g., evaluation of which a radiological abnormality is not seen by the radio-
pattern and distribution, detecting of cavitation, adenopa- logist on initial interpretation. The perceptual type is es-
thy, mass lesions, or collections). timated to account for approximately 80% of radiologi-
Many methods have been suggested for the correct in- cal errors [36].
terpretation of the chest radiograph, but there is no pre- Our second patient illustrates the complexity of the de-
ferred scheme or recommended system. The clinical ques- tection of a lung nodule located close to the hilum. A 77-
tion should always be addressed. An inquisitive approach is year-old man with known prostate cancer underwent
always helpful and being aware of the areas where mistakes chest radiography for right upper quadrant abdominal
are typically made is essential. Hidden abnormalities can pain (Fig. 2a, b). The radiographs were reported as nor-
thus be looked for. The difficult hidden areas that must mal. The coronal and sagittal reformats demonstrated the
be checked are the lung apex, areas superimposed over the position of the nodule (Fig. 2c, d), which, with the bene-
heart, around each hilum, and below the diaphragm. In the fit of hindsight, can be seen clearly on the posteroanterior
following, we concentrate on difficult areas, such as lesions and lateral chest radiographs.
at the lung apices or bases or adjacent to or obscured by the
hila or heart. For a systematic approach, we divide this re- Nodular Lesions: Infections
view into three sections addressing specific problems:
missed nodules, missed consolidations, and missed intersti- Nodular lesions attributed to pulmonary infections are
tial lung diseases. Finally, we examine some of the common most often seen in nosocomial pneumonias and in im-
signs that may help in the detection of lesions located in dif- munocompromised patients. They may be caused by bac-
ficult anatomic areas of the chest. teria such as Nocardia asteroides and Mycobacterium tu-
berculosis, septic emboli, and fungi. Nocardia asteroides
causes single or nodular infiltrates with or without cavi-
Specific Problems tation. Invasive pulmonary aspergillosis (IPA), mucor,
and Cryptococcus neoformans may present with single or
Specific problems of missed lung lesions can be divided multiple nodular infiltrates, which often progress to
into missed nodules, missed consolidations, and missed wedge-shaped areas of consolidation. Cavitation is com-
interstitial lung diseases. In the first two, the overlooked mon later in the course of the infiltrate. In the appropri-
pathology may have been detected if special attention had ate clinical setting, CT may aid in the diagnosis of IPA by
been paid to known difficult hidden areas. The follow- demonstrating the halo sign. Figure 3 shows the imag-
ing examples show how a side-by-side comparison of the ing results from a 43-year-old woman with fever after a
chest radiograph and CT images improves our under- bone marrow transplant. The posteroranterior radiograph
82 N. Howarth, D. Tack
a b c
Fig. 1 a-d. A 53-year-old man who underwent chest radiography for pain in the right axilla.
Posteroranterior (a) and lateral (b) radiographs were interpreted as normal. With hindsight
bias from MDCT, the right apical mass is obvious. MDCT coronal and sagittal images with
soft tissue (c) and bone (d) windows show a right apical mass with bone destruction
a b c
Fig. 2 a-d. A 77-year-old man with right upper quadrant pain. Posteroranterior (a) and lateral
(b) radiographs were interpreted as normal. With hindsight, the 13-mm nodule in the supe-
rior segment of the lingula can be seen. Coronal (c) and sagittal (d) reformats (lung window)
show the position of the lingular nodule, close to the hilum
Missed Lung Lesions: Side by Side Comparison of Chest Radiography with MDCT 83
a b
a b c
Fig. 4 a-c. A 46-year-old woman with cough and right-sided chest pain. The posteroranterior radiograph was interpreted as normal (a). Coronal
(b) and sagittal (c) reformats showing consolidation in the anterior segment of the right lower lobe
was interpreted as normal (Fig. 3a). With hindsight, a lar carcinoma [37]. Acute airspace pneumonia is charac-
subtle infiltrate can be seen at the left apex. Conspicuity terized by a mostly homogeneous consolidation of lung
is lessened by the overlying clavicle and 1st rib. The axi- parenchyma and well-defined borders; it does not typi-
al CT image (Fig. 3b) shows nodular consolidation with cally respect segmental boundaries. An air bronchogram
crescentic cavitation (the crescent sign) and a sur- is very common. Progression to lobar consolidation may
rounding ground-glass infiltrate (the halo sign). These occur. As with lung nodules, whether consolidation is de-
characteristic findings of IPA are best identified on CT. tected or missed on the plain chest radiograph may de-
pend on a combination of the same factors of size, den-
sity, location, and overlying structures. Location is a sig-
Missed Consolidations nificant factor for missed consolidations [38, 39]. Con-
solidations in the middle lobe and both lower lobes can
Airspace Disease be difficult to diagnose, especially when only the poste-
rioranterior view is obtained [40]. Figure 4 shows the ra-
Airspace disease is usually caused by bacterial infections. diographs from a 46-year-old woman with cough and
However, it can be seen in viral, protozoal, and fungal in- right-sided chest pain. The posteroanterior radiograph
fections as well as malignancy, typically brochioloalveo- was interpreted as normal (Fig. 4a). Due to a clinical
84 N. Howarth, D. Tack
a b c
d e
suspicion of pulmonary embolism, MDCT was request- of pneumonia when the chest radiograph is normal. This
ed, which showed consolidation in the anterior segment is especially true when the early diagnosis of pneumonia
of the right lower lobe. The coronal and sagittal reformats is critical, as is the case with immunocompromised and
demonstrated the extent of the consolidation (Fig. 4b, c). severely ill patients [43].
There were no signs of pulmonary embolism on the con- Figure 5 shows the imaging studies from a 38-year-old
trast medium study. A diagnosis of right lower lobe pneu- immunocompromised man with fever. The posteroanterior
monia was established and the patient was treated suc- and lateral radiographs (Fig. 5a, b) were interpreted as
cessfully with antibiotics. showing a peri-hilar reticular infiltrate with right upper
Chest radiography is the most frequently performed di- lobe consolidation. MDCT was requested to further char-
agnostic investigation requested by general practitioners acterize the infiltrate and revealed ground-glass opacifi-
in Europe [41]. Chest radiography is considered the gold cation with bilateral lung cysts (Fig. 5 c-e ). Pneumocys-
standard for the diagnosis of pneumonia. It can be used tis pneumonia was confirmed by bronchoalveolar lavage.
to diagnose pneumonia when an infiltrate is present and
to differentiate pneumonia from other conditions that
may present with similar symptoms, such as acute bron- Missed Interstitial Lung Disease
chitis. The results of the chest radiograph may in some
cases suggest a specific etiology (for example, a lung ab- Diffuse (Interstitial or Mixed Alveolar-Interstitial) Lung Disease
scess) or reveal a complication (empyema) or coexisting
abnormalities (bronchiectasis, bronchial obstruction, in- Diffuse lung disease presenting with widely distributed
terstitial lung disease). Chest radiography remains a valu- patchy infiltrates or interstitial reticular or nodular ab-
able diagnostic tool in primary-care patients with a clin- normalities can be produced by a number of disease en-
ical suspicion of pneumonia as it substantially reduces tities. An attempt is usually made to separate the group
the number of misdiagnosed patients [42]. MDCT can of idiopathic interstitial pneumonias from known causes,
help in differentiating infectious from non-infectious ab- such as infections, associated systemic disease, or drug-
normalities and may detect empyema, cavitation, and related. The most common infectious organisms are
lymphadenopathy when the chest radiograph cannot. viruses and protozoa. In general, the etiology of an un-
MDCT imaging is useful in patients with community- derlying pneumonia cannot be specifically diagnosed
acquired pneumonia when there is an unresolving or because the patterns overlap. It is beyond the aim of this
complicated chest radiograph and should be performed in chapter to discuss in detail the contribution of MDCT to
immunocompromised patients with a clinical suspicion the diagnosis of diffuse infiltrative lung disease. The
Missed Lung Lesions: Side by Side Comparison of Chest Radiography with MDCT 85
development of HRCT has resulted in markedly improved ened lateral costophrenic angle may have a sharp, angu-
diagnostic accuracy in acute and chronic diffuse infiltra- lar appearance. In the supine patient, air in the pleural
tive lung disease [44-47]. The chest radiograph remains space (pneumothorax) collects anteriorly and basally
the preliminary radiological investigation of patients within the nondependent portions of the pleural space;
with diffuse lung disease but is often non-specific. Pat- when the patient is upright, the air collects in the apico-
tern recognition in diffuse lung disease has been the sub- lateral location. If air collects laterally rather than medi-
ject of controversy for many years. Extensive disease ally, it deepens the lateral costophrenic angle and pro-
may be required before an appreciable change in radio- duces the deep sulcus sign. In Fig. 6, a deep sulcus sign
graphic density or an abnormal radiographic pattern is seen on the left, in addition to a continuous diaphrag-
can be detected on the plain chest radiograph. At least matic sign, present when there is air between the di-
10% of patients who are ultimately found to have biop- aphragm and the heart.
sy-proven diffuse lung disease have an apparently nor-
mal chest radiograph [48]. HRCT, and now MDCT, have Spine Sign
become integral components of the clinical investigation
of patients with suspected or established interstitial lung On the normal lateral chest radiograph, the attenuation
disease. These techniques have had a major impact on decreases (the lucency increases) as one progresses down
clinical practice [49, 50]. the thoracic vertebral bodies [52]. If the attenuation in-
creases, locally or diffusely, there must be a posteriorly
located lesion (Fig. 7). This lesion might not be seen on
Key Signs for Reducing the Risk of Errors on Chest the frontal view, as it might be hidden by the heart or the
X-rays hila.
Take-Home Messages
Despite the increasing use of CT imaging in the diagno-
sis of patients with chest disorders, chest radiography is
still the primary imaging method in patients with sus-
pected chest disease. The presence of an infiltrate on a
chest radiograph is considered the gold standard for di-
agnosing pneumonia. Extensive knowledge of the radio-
graphic appearance of pulmonary disorders is essential
when diagnosing pulmonary disease. Chest radiography
is also the imaging tool of choice in the assessment of
complications and in the follow-up of patients with pul-
monary diseases.
MDCT plays an increasing role in the diagnosis of
chest diseases, especially in patients with unresolving
symptoms. CT will aid in the differentiation of infection
and non-infectious disorders. The role of CT in suspect-
ed or proven chest disease can be summarized as follows:
1. CT is valuable in the early diagnosis of chest disease,
Fig. 6. A 78-year-old man with acute left chest pain and a previous especially in patients in whom an early diagnosis is
history of pneumoconiosis. Bedside chest radiograph shows a thin important (immunocompromised patients, critically ill
white line near the left chest wall (white arrows), corresponding to patients)
the left lung visceral pleura, and indicating a pneumothorax. The 2. CT may help with the characterization of pulmonary
deep lucency of the left lateral costophrenic angle extending to-
wards the abdomen is an indirect sign of pneumothorax (black ar- disorders
row). The continuous diaphragmatic sign is also seen as air sepa- 3. CT is an excellent tool in assessing the complications
rating the diaphragm from the heart (white hollow arrow) of chest disease
86 N. Howarth, D. Tack
a b c
Fig. 7 a-d. A 69-year-old woman with chronic obstructive pulmonary disease and hemoptysis.
The posteroanterior chest radiograph (a) shows an opacity next to the right border of the
heart (arrow) and obliterating the right side of the spine. This silhouette sign of the right
posterior mediastinal border indicates that the lesion is in a posterior location in the right
lower lobe. Lateral view (b) shows an increased density (arrow) of the lower spine
compared with the upper and middle thoracic spine (spine sign). This increased density
is due to a large mass in the right lower lobe. Coronal CT image (c) shows the right lower
lobe mass obliterating the border of the mediastinum. The sagittal CT image (d) shows the
posterior location of the mass
4. CT is required in the investigation of patients with a 8. Aideyan UO, Berbaum K, Smith WL (1995) Influence of pri-
persistent or recurrent pulmonary infiltrate. or radiologic information on the interpretation of radiographic
examinations. Acad Radiol 2:205-208.
A side-by-side comparison between the chest radio- 9. Carmody DP, Nodine CF, Kundel HL (1980) An analysis of
graph and MDCT images when confronted with a missed perceptual and cognitive factors in radiographic interpretation.
lung lesion is very instructive. The radiologist should be Perception 9:339-344.
able to understand the reasons for missing certain lesions. 10. Austin JH, Romney BM, Goldsmisth LS (1992) Missed bron-
By adopting this inquisitive approach, both cognitive and chogenic carcinoma: radiographic findings in 27 patients with
a potentially resectable lesion evident in retrospect. Radiology
perceptual errors can be reduced. 182:115-122.
11. Turkington PM, Kennan N, Greenstone MA (2002) Minsinter-
pretation of the chest x-ray as a factor in the delayed diagno-
sis of lung cancer. Postgrad Med J 78:158-160.
References 12. Woodring JH (1990) Pitfalls in the radiologic diagnosis of lung
cancer. AJR Am J Roentgenol 154:1165-1175.
1. Berlin L (1986) Malpractice and radiologists: an 11.5-year 13. Monnier-Cholley L, Arrive L, Porcel A et al (2001) Charac-
perspective. AJR Am J Roentgenol 147:1291-1298. teristics of missed lung cancer on chest radiographs: a French
2. Berlin L (1995) Malpractice and radiologists in Cook County, experience. Eur Radiol 11:597-605.
IL: trends in 20 years of litigation. AJR Am J Roentgenol 14. Krupinski EA, Berger WG, Dallas WJ et al (2003) Searching
165:781-788. for nodules: what features attract attention and influence de-
3. Potchen EJ, Bisesi MA (1990) When is it malpractice to miss tection? Acad Radiol 10:861-868.
lung cancer on chest radiographs? Radiology 175:29-32. 15. Samei E, Flynn MJ, Eyler WR (1999) Detection of subtle lung
4. Baker, SR, Patel RH, Yang L et al (2013). Malpractice suits in nodules: relative influence of quantum and anatomic noise on
chest radiology: an evaluation of the histories of 8265 radio- chest radiographs. Radiology 213:727-734.
logists. Journal of Thoracic Imaging 28:388-391. 16. Kundel HL, Nodine CF, Krupinski EA (1989) Searching for
5. Quekel LG, Kessels AG, Goei R et al (1999) Miss rate of lung lung nodules. Visual dwell indicates locations of false-positive
cancer on the chest radiograph in clinical practice. Chest and false-negative decisions. Invest Radiol 24:472-478.
115:720-4. 17. Samuel S, Kundel HL, Nodine CF et al (1995) Mechanism of
6. Melamed MR (2000) Lung cancer screening results in the Na- satisfaction of search: eye position recordings in the reading of
tional Cancer Institute New York study. chest radiographs. Radiology 194:895-902.
7. Marcus PM (2001) Lung cancer screening: an update. J Clin 18. Quekel LG, Goei R, Kessels AG et al (2001) Detection of lung
Oncol 19:83S-86S. Cancer 89:2356-2362. cancer on the chest radiograph: impact of previous films,
Missed Lung Lesions: Side by Side Comparison of Chest Radiography with MDCT 87
clinical information, double reading, and dual reading. J Clin 36. Berlin L, Hendrix RW (1998) Perceptual errors and negli-
Epidemiol 54:1146-1150. gence. AJR Am J Roentgenol 170:863-867.
19. Tsubamoto M, Kuriyama K, Kido S et al (2002) Detection of 37. Jung JI, Kim H, Park SH et al (2001) Differentiation of pneu-
lung cancer on chest radiographs: analysis on the basis of size monic-type bronchioloalveolar cell carcinoma and infectious
and extent of ground-glass opacity at thin-section CT. Radio- pneumonia. Br J Radiol 74:490-494.
logy 224:139-144. 38. Melbye H, Dale K (1992) Interobserver variability in the
20. Shah PK, Austin JH, White CS et al (2003) Missed non-small radiographic diagnosis of adult outpatient pneumonia. Acta
cell lung cancer: radiographic findings of potentially re- Radiol 33:79-83.
sectable lesions evident only in retrospect. Radiology 226:235- 39. Albaum MN, Hill LC, Murphy M (1996) Interobserver relia-
241. bility of the chest radiograph in community-acquired pneumo-
21. Samei E, Flynn MJ, Peterson E, Eyler WR (2003) Subtle lung nia. Chest 110 :343.
nodules: influence of local anatomic variations on detection. 40. Chotas HG, Ravin CE (1994) Chest radiography: estimated
Radiology 228:76-84. lung volume and projected area obscured by the heart, medi-
22. Kelsey CA, Moseley RD, Brogdon BG et al (1977) Effect of astinum, and diaphragm. Radiology 193:403-404.
size and position on chest lesion detection. AJR Am J 41. Woodhead M, Gialdroni Grassi G, Huchon GJ et al (1996) Use
Roentgenol 129:205-208. of investigations in lower respiratory tract infection in the
23. Kimme-Smith C, Hart EM, Goldin JG et al (1996) Detection community: a European survey. Eur Respir J 9:1596-1600.
of simulated lung nodules with computed radiography: effects 42. Speets AM, Hoes AW, van der Graaf Y et al (2006) Chest ra-
of nodule size, local optical density, global object thickness, diography and pneumonia in primary care: diagnostic yield
and exposure. Acad Radiol 3:735-741. and consequences for patient management. Eur Respir J
24. Kundel HL, Revesz G, Toto L (1979) Contrast gradient and the 28:933-938.
detection of lung nodules. Invest Radiol 14:18-22. 43. Heussel CP, Kauczor HU, Ullmann AJ (2004) Pneumonia in
25. Kundel HL (1975) Peripheral vision, structured noise and film neutropenic patients. Eur Radiol 14:256-271.
reader error. Radiology 114:269-273. 44. Mathieson JR, Mayo JR, Staples CA et al (1989) Chronic dif-
26. Revesz G, Kundel HL (1977) Psychophysical studies of de- fuse inflitrative lung disease: comparison of diagnostic accu-
tection errors in chest radiology. Radiology 123:559-562. racy of CT and chest radiography. Radiology 171:111-116.
27. Kundel HL, Revesz G (1976) Lesion conspicuity, structured 45. Grenier P, Valeyre D, Cluzel P et al (1991) Chronic diffuse in-
noise, and film reader error. AJR Am J Roentgenol 126:233- terstitial lung disease: diagnostic value of chest radiography
238. and high-resolution CT. Radiology 179:123-132.
28. Wu M-H, Gotway MB, Lee TJ et al (2008) Features of non- 46. Padley SPG, Hansell DM, Flower CDR et al (1991) Compar-
small cell lung carcinomas overlooked at digital chest radio- ative accuracy of high resolution computed tomography and
graphy. Clin Radiol 63:518-528. chest radiography in the diagnosis of chronic diffuse inflitra-
29. Herman PG, Hessel SJ (1975) Accuracy and its relationship to tive lung disease. Clin Radiol 44:222-226.
experience in the interpretation of chest radiographs. Invest 47. Webb WR, Mller NL, Naidich DP (2001) High-resolution CT
Radiol 10:62-67. of the lung, 3rd Ed. Lippincott Williams & Wilkins, Philadel-
30. Bass JC, Chiles C (1990) Visual Skill. Correlation with detec- phia, PA.
tion of solitary pulmonary nodules. Invest Radiol 25:994-998. 48. Epler GR, McLoud TC, Gaensler EA et al (1978) Normal
31. Sone S, Li F, Yang ZG et al (2000) Characteristics of small chest roentgenograms in chronic diffuse inflitrative lung dis-
lung cancers invisible on conventional chest radiography and ease. N Engl J Med 298:934-939.
detected by population based screening using spiral CT. Br J 49. Aziz ZA, Wells AU, Bateman ED et al (2005) Interstitial lung
Radiol 73:137-145. disease: effects of thin-section CT on clinical decision making
32. Yang ZG, Sone S, Li F et al (2001) Visibility of small periph- Radiology 238:725-733.
eral lung cancers on chest radiographs: influence of densito- 50. Flaherty KR, King TE Jr, Raghu G et al (2004) Idiopathic in-
metric parameters, CT values and tumour type. Br J Radiol terstitial pneumonia: what is the effect of a multi-disciplinary
74:32-41. approach to diagnosis? Am J Respir Crit Care Med 170:904-
33. Berbaum KS (1995) Difficulty of judging retrospectively 910.
whether a diagnosis has been missed . Radiology 194:582- 51. Kong A (2003) The deep sulcus sign. Radiology 228:415-416.
583. 52. Ely JW, Berbaum KS, Bergus GR et al (1996) Diagnosing left
34. Berlin L (2000) Hindsight bias. AJR Am J Roentgenol lower lobe pneumonia: usefulness of the spine sign on later-
175:597-601. al chest radiographs. J Fam Pract 43:242-248.
35. Berlin L (2001) Defending the missed radiographic diag- 53. Algn O, Gkalp G, Topal U (2011) Signs in chest imaging.
nosis. AJR Am J Roentgenol 176:317-322. Diagnostic and Interventional Radiology, 17:18-29.
IDKD 2015-2018
of chest radiography. However, a more important issue manageably short) list of differential diagnoses. To this
than sensitivity is the confidence and accuracy with end, a proposed schema, presented in the form of ques-
which the diagnosis can be made. In the oft-quoted land- tions that the observer should ask (in roughly the order
mark study by Mathieson and colleagues, experienced given), is as follows.
chest radiologists were asked to independently indicate
up to three diagnoses based on evaluations of the chest Is There a Real Abnormality?
radiographs and CT scans of 118 patients with a variety
of biopsy-confirmed ILDs [18]. Importantly, for the first- This is a crucial first question: the radiologist must first
choice diagnosis, the readers were asked to assign a lev- determine whether what is shown on HRCT represents
el of confidence. The first important finding of this study real disease. CT features attributable to technical fac-
(which effectively put HRCT on the map) was that, com- tors/normal variation (for instance, caused by a poor in-
pared with chest radiography, a confident diagnosis was spiratory effort, inadequate mAs, regions of physiologi-
made nearly twice as often using HRCT. The second, and cally dependent atelectasis) must not be overinterpreted
perhaps more striking, message was that when experi- and reported as disease. Making the distinction between
enced radiologists were confident of the diagnosis on normality and abnormality can also be difficult when
HRCT, they were almost always correct [18]. However, a there is apparently minimal disease or, conversely, when
confident diagnosis on chest X-ray (which, incidentally, there is diffuse abnormality, as may occur with subtle but
was offered in only 25% of the cases), was associated widespread decreased attenuation (mosaicism) or in-
with a significantly lower rate of correct diagnoses. creased (ground-glass opacity) attenuation.
The results of subsequent studies have not always mir-
rored those of the initial study by Mathieson [18, 19]. If There Is an Abnormality, What Is/Are the Predominant HRCT
However, because of study design, the majority of the Pattern(s)?
comparative studies in HRCT probably undervalued its
true utility [1]. Firstly, there was no recourse to pre-test Having decided that there is a definite abnormality on
probabilities for observers in early series and, therefore, HRCT, the observer should attempt to identify the domi-
these do not accurately reflect clinical practice. Secondly, nant pattern(s) using only the standard radiological terms
radiologists (and specifically those with an interest in tho- [15]. The use of non-standard, terminology (e.g., patchy
racic disease) have become increasingly familiar with the opacification, parenchymal opacities) or descriptive terms
spectrum of HRCT patterns and disease. This, almost cer- in which there is an implied pathology (e.g., interstitial
tainly, would be associated with a proportionate increase pattern or alveolitis) is misleading and best avoided.
in the confidence of experienced observers in making
HRCT diagnoses, were such a study to be repeated today. What Is the Distribution of the Disease?
Some justification for this last statement comes from a
study that addressed the clinically vexing issue of end- Since many DILDs have a predilection for certain zones,
stage lung disease [20], in which two experienced tho- an evaluation of the dominant distribution is of diagnos-
racic radiologists independently made correct first-choice tic value. For instance, it is known that in the majority of
diagnoses in just under 90% of cases, with nearly two- patients with idiopathic pulmonary fibrosis (IPF) the dis-
thirds of those diagnoses made with high confidence. On ease tends to be most obvious in the mid- to lower zones.
first inspection, these data seem less than impressive. This contrasts with fibrosis in patients with sarcoidosis,
However, the results of open lung biopsy (the supposed which typically has a predilection for the upper lobes. In
gold-standard for a diagnosis of DILD) are not infre- addition, the radiologist should take note of the axial dis-
quently inconclusive, no doubt in part relating to the de- tribution (central vs. peripheral) which, in contrast with
gree of observer variability between pathologists [21]. chest X-ray, can readily be made on HRCT. The potential
value can be demonstrated by again using the example of
IPF and sarcoidosis: the former is commonly peripheral
An Approach to HRCT Diagnoses (subpleural) whereas in the latter the distribution tends to
be central (and bronchocentric). A final example is seen
It must be acknowledged that HRCT interpretation can in patients with organizing pneumonia, in which consol-
be difficult, even for trained thoracic radiologists! This is idations may have a striking perilobular predilection [22].
not surprising given the sheer numbers of documented
DILDs. These broad-ranging disorders manifest with a Are There Any Ancillary Findings?
relatively small number of histopathological patterns
(e.g., fibrosis, consolidation, intra-alveolar hemorrhage) Ancillary HRCT features may suggest or, indeed, exclude
which, in turn, are reflected by a similarly select group of certain diagnoses. Thus, the presence or absence of the
HRCT features (reticulation, ground-glass opacification, following may be of diagnostic value in specific cases:
nodularity, thickening of interlobular septa). However, a. Pleural thickening/effusions /plaques ( calcification)
with a systematic approach to HRCT interpretation the may suggest asbestos-related lung disease as opposed
observer should, in time, be able to offer a sensible (and to IPF as a possible cause of lung fibrosis.
90 S.R. Desai, J.R. Galvin
b. Reactive intrathoracic nodal enlargement (hilar/medi- microscopy, there is temporally heterogeneous fibrosis
astinal) is normal in fibrotic DILDs. However, sym- admixed with areas of unaffected lung [2]. In areas of fi-
metrical hilar nodal enlargement may suggest a diag- brosis, there will be characteristic honeycombing. The
nosis of sarcoidosis or occupational lung disease. In- disease has a striking basal and subpleural predilection.
trathoracic nodal enlargement is uncommon in pul- To make a diagnosis of a UIP pattern on HRCT, the
monary vasculitides (e.g., Wegeners granulomatosis). radiologist must therefore look for the following: (1) a
c. Bronchiectasis with co-existent suppurative airways reticular pattern, (2) honeycombing (with or without trac-
disease in a patient with established pulmonary fibro- tion bronchiectasis), and (3) a subpleural, basal, disease
sis may point to a diagnosis of an underlying connec- distribution [23] while ensuring that features inconsistent
tive tissue disease, such as rheumatoid arthritis. with a UIP pattern (i.e., mid/upper zone predominance, a
d. The demonstration of intrapulmonary cysts in the ap- bronchovascular distribution, ground-glass opacification
propriate clinical context suggests a range of diag- more extensive than reticulation, widespread micronodu-
noses, including Langerhans cell histiocytosis, lym- larity, multiple cysts away from areas of honeycombing,
phangioleiomyomatosis, and lymphocytic interstitial a mosaic attenuation pattern in three or more lobes, and
pneumonia. consolidation) are absent. In the absence of honeycomb-
e. Consolidation indicative of organizing pneumonia may ing (but with the other features in the list), an HRCT di-
co-exist with signs of fibrosis and parenchymal distor- agnosis of possible UIP can be made.
tion in patients with connective-tissue-disease-related The presence of a UIP pattern on HRCT is accurate
lung disease. and obviates histologic confirmation [3, 24]. However,
typical appearances may not be present in over half of pa-
What Is the Likely Pathology? tients with biopsy-proven disease [24, 25].
Knowledge of the relationships between the HRCT ap- (Cryptogenic) Organizing Pneumonia
pearances and the possible histopathological correlates is
crucial. Thus, in a patient with predominant consolidation Organization is a common response to lung injury and is
it is reasonable to conclude that the dominant pathology part of the normal process of lung repair. It is represented
involves the air spaces, whereas in a patient with reticula- on histology by plugs of fibroblastic tissue that fill the
tion the likely pathological process affects the interstitium. alveolar spaces. This same fibroblastic tissue may be iden-
tified in respiratory and terminal bronchioles, explaining
What Is the Clinical Background? the use of the older term bronchiolitis obliterans orga-
nizing pneumonia, but which has been replaced by or-
Clinical data must always be integrated when formulating ganizing pneumonia (OP) [12]. OP was first recognized
a radiological opinion. However, it is often advisable to in 1923 as a response to unresolved pneumonia [26]. Most
review the clinical information after evaluation of the ra- cases are likely to be post-infectious; however, the organ-
diological features. This is particularly true at the very ism is rarely recovered. The OP pattern is also a common
start of HRCT interpretation, when the radiologist is de- feature in a wide range of other diseases, including colla-
ciding whether or not there is a real abnormality (see gen-vascular disease, hypersensitivity pneumonitis,
above). Specific clinical features that may be of impor- chronic eosinophilic pneumonia, drug reaction, and radi-
tance in HRCT interpretation include basic demographic ation-induced lung injury. The appearance of OP on imag-
data (age, gender, ethnicity), smoking history, time ing is highly variable, depending on the prior injury and
course of the illness (i.e., have symptoms developed over the stage at which it is imaged [27]. A minority of patients
hours and days or weeks and months?), and any relevant present with solitary pulmonary nodules or focal areas of
past medical history. consolidation. However, the dominant finding in OP is bi-
lateral consolidation that is peripheral, often with sparing
of the subpleural portion of the lung. Opacities are often
HRCT Appearances in Select DILDs perilobular and may be associated with septal lines. In
some patients, OP may progress to an non-specific inter-
A working knowledge of the relationship between stitial pneumonia pattern of fibrosis (see below). Al-
histopathological changes and HRCT patterns and the though many cases of OP will clear with steroids, a sub-
typical appearance of common DILDs is of value in day- stantial minority of patients are left with significant dis-
to-day practice. The following sections briefly consider ability due to pulmonary fibrosis [27-29].
the HRCT appearances of a few DILDs.
Non-specific Interstitial Pneumonia
Idiopathic Pulmonary Fibrosis/Usual Interstitial Pneumonia
After UIP, non-specific interstitial pneumonia (NSIP) is
IPF is a chronic progressive fibrosing interstitial lung dis- the most common pattern of idiopathic interstitial pneu-
ease associated with a histologic and/or HRCT pattern of monias (IIP) and linked with a better survival [13, 24,
usual interstitial pneumonia (UIP) [23]. At low-power 30]. This pattern of IIP may be idiopathic but more
Plain Film and HRCT Diagnosis of Interstitial Lung Disease 91
commonly it is seen in a variety of clinical contexts, in- In the final stages, the appearance may be indistinguish-
cluding connective tissue disorders (especially systemic able from severe bullous emphysema.
sclerosis) and as a consequence of drug-related toxicity. It is important to recognize that findings of RB, DIP,
At the histologic level there are varying amounts of in- PLCH and the NSIP pattern of fibrosis commonly coex-
terstitial inflammation and fibrosis, which, in stark con- ist in biopsies of dypneic smokers. Some of the histolog-
trast to what is seen in UIP, has a temporally and spatial- ic changes are reflected on imaging, while others are be-
ly uniform appearance. low the resolution of chest CT.
On HRCT, one of the key findings is ground-glass
opacification which is typically bilateral and symmetri- Cigarette Smoking-Related Fibrosis
cally distributed in the lower zones [30, 31]. In some pa-
tients, the extent of ground-glass may decrease over time The relationship between cigarette smoke and fibrosis re-
and become replaced by reticulation (i.e., with UIP-like mains contentious [36, 4550]. Niewoehners original de-
features) [32]. Reticulation (generally without significant scription of RB did not include fibrosis of the alveolar
honeycombing) is usually also present. wall [33].However, there is substantial support for a rela-
tionship between cigarette smoke exposure and a pattern
Respiratory Bronchiolitis/Respiratory Bronchiolitis Interstitial of alveolar wall fibrosis other than UIP [46, 51-55]. In
Lung Disease/Desquamative Interstitial Pneumonia and our experience there is a group of dyspneic cigarette
Pulmonary Langerhans Cell Histiocytosis smokers who, as seen on CT, present with a combination
of well-formed cystic spaces that follow the typical up-
Respiratory bronchiolitis (RB), of variable severity, is an per-lobe-predominant distribution of smoking-related
almost invariable pathologic finding in all smokers [33]. emphysema, with variable surrounding ground glass
Importantly, this lesion is asymptomatic and not associ- opacity and reticulation that may extend into the lower
ated with physiologic impairment in the vast majority of lung zones [48]. These patients commonly present with
individuals. However, in a small minority, there will be strikingly normal flows and volumes on pulmonary func-
the clinical manifestations of an interstitial lung disease. tion testing and a low diffusing capacity. The unexpect-
It is this clinico-pathologic/radiologic entity that has been edly normal flows and volumes are the result of the op-
called respiratory bronchiolitis interstitial lung disease posing effects of emphysema and fibrosis [56].
(RBILD). The cardinal HRCT signs of RB/RBILD in-
clude soft centrilobular nodules, ground-glass opacifi- Sarcoidosis
cation, smooth thickening of interlobular septa, and lob-
ular foci of decreased attenuation [34-36]. Non-caseating epithelioid-cell granulomata are the
Desquamative interstitial pneumonia (DIP) was first histopathologic hallmark of sarcoidosis [57]. Since the
described by Leibow in 1965. Dypsneic patients with DIP granulomata have a tendency to distribute along the lym-
were found to have numerous inflammatory cells in alve- phatics, the lymphatic pathways that surround the axial
olar spaces [37]. The cells were thought to be desqua- interstitium that invests bronchovascular structures and
mated pneumocytes but are now recognized as the same those located subpleurally (including the subpleural lym-
macrophages identified in patients with RB and RBILD. phatics along the fissures) are typically involved. Not sur-
The majority of patients with DIP are heavy smokers and prisingly, a nodular infiltrate (presumably reflecting con-
the disease is now considered part of the spectrum of in- glomerate granulomata) with a propensity to involve the
flammatory lung disease related to the inhalation of cig- bronchovascular elements is a characteristic CT finding
arette smoke [34]. Patients with DIP suffer an increased [58, 59]. Subpleural nodularity is also commonly seen. In
incidence of pulmonary fibrosis that fits the histologic the later stages of the disease there may be obvious signs
pattern of NSIP [38, 39]. Imaging in patients with DIP is of established lung fibrosis with upper zone volume loss,
typified by homogeneous or patchy areas of ground glass parenchymal distortion, and traction bronchiectasis. Be-
opacity in the mid and lower lung zones [39, 40]. cause of the bronchocentric nature of the disease, signs
Langerhans cell histiocytosis (LCH) is confined to the of small-airways disease are seen at CT in some patients
lungs in approximately 90% of patients. Although the with sarcoidosis [60].
pathogenesis is unknown, most if not all of the patients
are cigarette smokers. Pulmonary LCH (PLCH) is char- Hypersensitivity Pneumonitis
acterized by bronchiolocentric collections of Langerhans
cells admixed with a variety of other inflammatory cells Exposure to a range of organic antigens will cause lung
forming a stellate nodule [41]. Over time infiltration of disease in some patients, probably due to an immuno-
the airway wall results in its damage and subsequent di- logically mediated response [57]. In the subacute stage,
latation [42]. In the late stage of PLCH, small stellate there is an interstitial infiltrate comprising lymphocytes
scars are surrounded by emphysematous spaces. Imaging and plasma cells with a propensity to involve the small
reflects the histologic progression, with early bronchiolo- airways (bronchioles). Scattered non-caseating granulo-
centric nodules in the upper lobes progressing to a com- mata may be seen. Predictably, at CT, there is diffuse
bination of bizarrely shaped cysts and nodules [43, 44]. ground-glass opacification, ill-defined centrilobular
92 S.R. Desai, J.R. Galvin
nodules, and lobular areas of decreased attenuation on 17. Mller NL, Miller RR, Webb WR et al (1986) Fibrosing alve-
images performed at end-expiration [61, 62]. The CT olitis: CT-pathologic correlation. Radiology 160:585-588.
18. Mathieson JR, Mayo JR, Staples CA, Mller NL (1989) Chron-
appearances of subacute hypersensitivity pneumonitis ic diffuse infiltrative lung disease: comparison of diagnostic ac-
may be identical to those of RBILD [36]. Consideration curacy of CT and chest radiography. Radiology 171:111-116.
of the smoking history may help in the differentiation: a 19. Padley SPG, Hansell DM, Flower CDR, Jennings P (1991)
history of smoking is the norm in the vast majority of pa- Comparative accuracy of high resolution computed tomogra-
tients with RBILD, whereas cigarette smoke appears to phy and chest radiography in the diagnosis of chronic diffuse
infiltrative lung disease. Clin Radiol 44:222-226.
protect against the development of hypersensitivity 20. Primack SL, Hartman TE, Hansell DM, Mller NL (1993)
pneumonitis. End-stage lung disease: CT findings in 61 patients. Radiology
189:681-686.
21. Nicholson AG, Addis BJ, Bharucha H et al (2004) Inter-ob-
server variation between pathologists in diffuse parenchymal
References lung disease. Thorax 59:500-505.
22. Ujita M, Renzoni EA, Veeraraghavan S et al (2004) Organiz-
1. Wells AU (1998) Clinical usefulness of high resolution com- ing pneumonia: perilobular pattern at thin-section CT. Radiol-
puted tomography in cryptogenic fibrosing alveolitis. Thorax ogy 232:757-761.
53:1080-1087. 23. Raghu G, Collard HR, Egan J et al (2011) An official
2. King TE Jr., Costabel U, Cordier JF et al (2000) Idiopathic ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis
pulmonary fibrosis: diagnosis and treatment. International - evidence-based guidelines for diagnosis and management.
consensus statement. Am J Respir Crit Care Med 161:646- Am J Respir Crit Care Med 183:788-824.
664. 24. Flaherty KR, Thwaite EL, Kazerooni EA et al (2003) Radio-
3. Hunninghake GW, Zimmerman MB, Schwartz DA et al (2001) logical versus histological diagnosis in UIP and NSIP: survival
Utility of a lung biopsy for the diagnosis of idiopathic pul- implications. Thorax 58:143-148.
monary fibrosis. Am J Respir Crit Care Med 164:193-196. 25. Sverzellati N, Wells AU, Tomassetti S et al (2010) Biopsy-
4. Goh NSL, Desai SR, Veeraraghavan S et al (2008) Interstitial proved idiopathic pulmonary fibrosis: spectrum of nondiag-
lung disease in systemic sclerosis: a simple staging system. nostic thin-section CT diagnoses. Radiology 254:957-964.
Am J Respir Crit Care Med 177:1248-1254. 26. Symmers D, Hoffman AM (1923) The increased incidence of
5. Wells AU, Antoniou KM (2014) The prognostic value of the organizing pneumonia. JAMA 81:297-298.
gap model in chronic interstitial lung disease: the quest for a 27. Kligerman SJ, Franks TJ, Galvin JR (2013) Organization and
staging system. Chest 145:672-674. fibrosis as a response to lung injury in diffuse alveolar dam-
6. Walsh SLF, Wells AU, Sverzellati N et al (2014) An integrated age, organizing pneumonia, and acute fibrinous amd organiz-
clinico-radiological staging system for pulmonary sarcoidosis: ing pneumonia. Radiographics 33:1951-1975.
a case-control study. Lancet Respiratory Medicine 2:123-130. 28. Akira M, Inoue Y, Arai T et al (2011) Long-term follow-up of
7. Ryerson CJ, Vittinghoff E, Ley B et al (2014) Predicting sur- high-resolution CT findings in non-specific interstitial pneu-
vival across chronic interstitial lung disease: the ILD-GAP monia. Thorax 66:61-65.
model. Chest 145:723-728. 29. Lee JW, Lee KS, Lee HY et al (2010) Cryptogenic organizing
8. Walsh SLF, Sverzellati N, Devaraj A et al (2014) Connective pneumonia: serial high-resolution CT findings in 22 patients.
tissue disease related fibrotic lung disease: high-resolution AJR Am J Roentgenol 195:916-922.
computed tomographic and pulmonary function indices as 30. Flaherty KR, Martinez FJ, Travis W, Lynch JP, III (2001) Non-
prognostic determinants. Thorax 2014:216-222. specific interstitial pneumonia (NSIP). Semin Respir Crit Care
9. Edey AJ, Devaraj A, Barker RP et al (2011) Fibrotic idiopath- Med 22:423-434.
ic interstitial pneumonias: HRCT findings that predict mortal- 31. MacDonald SLS, Rubens MB, Hansell DM et al (2001) Non-
ity. Eur Radiol 21:1586-1593. specific interstitial pneumonia and usual interstitial pneumo-
10. Mayo JR, Webb WR, Gould R et al (1987) High-resolution CT nia: comparative appearances at and diagnostic accuracy of
of the lungs: an optimal approach. Radiology 163:507-510. thin-section CT. Radiology 221:600-605.
11. Murata K, Khan A, Rojas KA, Herman PG (1988) Optimiza- 32. Silva CIS, Mller NL, Hansell DM et al (2008) Nonspecific
tion of computed tomography technique to demonstrate the interstitial pneumonia and idiopathic pulmonary fibrosis:
fine structure of the lung. Invest Radiol 23:170-175. changes in pattern and distribution of disease over time. Radio-
12. Travis WD, King TE, Jr., and the Multidisciplinary Core Pan- logy 247:251-259.
el (2002) American Thoracic Society/European Respiratory 33. Niewoehner DE, Kleinerman J, Rice DB (1974) Pathologic
Society international multidiscplinary consensus classification changes in the peripheral airways of young cigarette smokers.
of idiopathic interstitial pneumonias. Am J Respir Crit Care N Engl J Med 291:755-758.
Med 165:277-304. 34. Heyneman LE, Ward S, Lynch DA et al (1999) Respiratory
13. Travis WD, Costabel U, Hansell DM et al (2013) An official bronchiolitis, respiratory bronchiolitis-associated interstitial
American Thoracic Society/European Respiratory Society lung disease, and desquamative interstitial pneumonia: differ-
statement: update of the international multidisciplinary classi- ent entities or part of the spectrum of the same disease
fication of the idiopathic interstitial pneumonias. Am J Respir process? AJR Am J Roentgenol 173:1617-1622.
Crit Care Med 188:733-748. 35. Moon J, du Bois RM, Colby TV et al (1999) Clinical signifi-
14. Remy-Jardin M, Remy J, Deffontaines C, Duhamel A (1991) As- cance of respiratory bronchiolitis on open lung biopsy and its
sessment of diffuse infiltrative lung disease: comparison of con- relationship to smoking related interstitial lung disease. Tho-
ventional CT and high-resolution CT. Radiology 181:157-162. rax 54:1009-1014.
15. Hansell DM, Bankier AA, MacMahon H et al (2008) Fleis- 36. Desai SR, Ryan SM, Colby TV (2003) Smoking related inter-
chner society: glossary of terms for thoracic imaging. Radiol- stitial lung diseases: histopathological and imaging perspec-
ogy 246:697-722. tives. Clin Radiol 58:259-268.
16. Hansell DM, Lynch DA, McAdams HP, Bankier AA (2010) 37. Liebow AA, Steer A, Billingsley JG (1965) Desquamative in-
Basic patterns in lung disease. In: Hansell DM, Lynch DA, terstitial pneumonia. Am J Med 39:369-404.
McAdams HP, Bankier AA, Eds. Imaging of Diseases of the 38. Wells AU, Nicholson AG, Hansell DM (2007) Smoking-in-
Chest, 5th Ed. Mosby Elsevier, pp 83-151. duced diffuse interstitial lung diseases. Thorax 62:904-910.
Plain Film and HRCT Diagnosis of Interstitial Lung Disease 93
39. Craig PJ, Wells AU, Doffman S et al (2004) Desquamative in- 51. Lang MR, Fiaux GW, Gillooly M (1994) Collagen content of
terstitial pneumonia, respiratory bronchiolitis and their rela- alveolar wall tissue in emphysematous and non-emphysema-
tionship to smoking. Histopathology 45:275-282. tous lungs. Thorax 49:319-326.
40. Hartman TE, Primack SL, Kang E-Y et al (1996) Disease pro- 52. Adesina AM, Vallyathan V, McQuillen EN et al (1991) Bron-
gression in usual interstitial pneumonia compared with chiolar inflammation and fibrosis associated with smoking: a
desquamative interstitial pneumonia: assessment with serial morphologic cross-sectional population analysis. Am Rev
CT. Chest 110:378-382. Respir Dis 143:144-149.
41. Vassallo R, Ryu JH, Colby TV et al (2000) Pulmonary Langer- 53. Frasca JM, Auerbach O, Carter HW, Parks VR (1983) Mor-
hans-cell histiocytosis. N Engl J Med 342:1969-1978. phologic alterations induced by short-term cigarette smoking.
42. Kambouchner M, Basset F, Marchal J et al (2002) Three-di- Am J Pathol 111:11-20.
mensional characterisation of pathologic lesions in pulmonary 54. Niewoehner DE, Hoidal JR (1982) Lung fibrosis and emphy-
langerhans cell histiocytosis. Am J Respir Crit Care Med sema: divergent responses to a common injury? Science
166:1483-1490. 217:359-360.
43. Leatherwood DL, Heitkamp DE, Emerson RE (2007) Pul- 55. Auerbach O, Garfinkel L, Hammond EC (2014) Relation of
monary Langerhans cell histiocytosis. Radiographics 27:265- smoking and age to findings in lung parenchyma: a micro-
268. scopic study. Chest 65:29-35.
44. Abbott GF, Rosado-de-Christenson ML, Franks TJ et al (2004) 56. Washko GR, Hunninghake GW, Fernandez IE et al (2011)
Pulmonary Langerhans cell histiocytosis. Radiographics Lung volumes and emphysema in smokers with interstitial
24:821-841. lung abnormalities. N Engl J Med 364:897-906.
45. Wright JL, Tazelaar HD, Churg A (2011) Fibrosis with em- 57. Colby TV, Carrington CB (1995) Interstitial lung disease. In:
physema. Histopathology 58:517-524. Thurlbeck WM, Churg AM (Eds) Pathology of the lung, 2nd
46. Katzenstein AL, Mukhopadhyay S, Zanardi C, Dexter E Ed. Thieme Medical Publishers, New York, pp 589-737.
(2010) Clinically occult interstitial fibrosis in smokers: classi- 58. Lynch DA, Webb WR, Gamsu G et al (1989) Computed to-
fication and significance of a surprisingly common finding in mography in pulmonary sarcoidosis. J Comput Assist Tomogr
lobectomy specimens. Hum Pathol 41:316-325. 1989;13(3):405-10.
47. Churg A, Mller NL, Wright JL (2010) Respiratory bronchi- 59. Brauner MW, Grenier P, Mompoint D et al (1989) Pulmonary
olitis/interstitial lung disease: fibrosis, pulmonary function, sarcoidosis: evaluation with high-resolution CT. Radiology
and evolving concepts. Arch Pathol Lab Med 134:27-32. 172:467-471.
48. Galvin JR, Franks TJ (2009) Smoking-related lung disease. J 60. Hansell DM, Milne DG, Wilsher ML, Wells AU (1998) Pul-
Thorac Imaging 24:274-284. monary sarcoidosis: morphologic associations of airflow ob-
49. Kawabata Y, Hoshi E, Murai K et al (2008) Smoking-related struction at thin-section CT. Radiology 209:697-704.
changes in the background lung of specimens resected for lung 61. Remy-Jardin M, Remy J, Wallaert B, Mller NL (1993) Suba-
cancer: a semiquantitative study with correlation to postoper- cute and chronic bird breeder hypersensitivity pneumonitis: se-
ative course. Histopathology 53:707-714. quential evaluation with CT and correlation with lung function
50. Yousem SA, Colby TV, Gaensler EA (1989) Respiratory bron- tests and bronchoalveolar lavage. Radiology 189:111-118.
chiolitis-associated interstitial lung disease and its relationship 62. Hansell DM, Wells AU, Padley SPG, Mller NL (1996) Hy-
to desquamative interstitial pneumonia. Mayo Clin Proc persensitivity pneumonitis: correlation of individual CT pat-
64:1373-1380. terns with functional abnormalities. Radiology 199:123-128.
IDKD 2015-2018
Introduction Lungs
Chest radiographs are frequently obtained for the assess- Lung Volume
ment of patients who present with a variety of thoracic
complaints. In addition, asymptomatic patients undergo- In normal subjects, postero-anterior (PA) chest radio-
ing a physical exam, a hospital admission, or a preopera- graphs obtained at full inspiration or total lung capacity
tive assessment may also undergo chest radiography. (TLC) result in the visualization of the ten posterior ribs
Accurate interpretation of chest radiographs requires above the diaphragm. Lung volume may be increased in
knowledge of imaging anatomy, as obscuration of nor- patients with obstructive diseases such as emphysema
mally visualized structures may be the only clue to the and reduced in those with restrictive diseases such as pul-
presence of an abnormality. Radiography allows the vi- monary fibrosis.
sualization and assessment of the lungs, central airways, Atelectasis may involve the entire lung, a lung lobe, or
pulmonary vasculature, mediastinum, heart, great ves- a pulmonary segment [1]. Obstructive (resorption)
sels, and chest wall (including skeletal and soft-tissue atelectasis is characterized by the absence of intrinsic air
structures). The superimposition of complex structures bronchograms. As it may result from endoluminal ob-
of various radiographic densities (air, water, calcium, struction, a centrally obstructing neoplasm such as lung
metal, fat) makes radiographic interpretation challeng- cancer must be excluded (Fig. 1). Relaxation (passive)
ing. Radiologists must work with technologists to atelectasis often results from mass effect, commonly
achieve proper patient positioning and optimal radi- from pleural effusion. Cicatricial atelectasis is related to
ographic technique. Unusual opacities must be assessed pulmonary fibrosis. Rounded atelectasis occurs adjacent
to determine whether they are intrinsic or extrinsic to to pleural thickening in which subpleural atelectatic lung
the patient and to exclude artifacts and foreign bodies. folds upon itself.
In addition, the radiologist must carefully evaluate all Direct signs of atelectasis include fissural displace-
visualized medical devices for appropriate course and ment, bronchovascular crowding, and shift of calcified
position. granulomas. Indirect signs include increased pulmonary
opacity, ipsilateral mediastinal shift, hilar displacement,
ipsilateral hemidiaphragm elevation, and compensatory
A Systematic Assessment hyperinflation of the adjacent lung [2].
a b
Fig. 1 a, b. Left upper lobe atelectasis. Postero-anterior (PA) chest radiograph (a) shows left upper lobe atelectasis manifesting as a vague left
upper lung zone opacity with elevation of the left hemidiaphragm. The lateral chest radiograph (b) shows the opaque atelectatic left upper
lobe and anterosuperior displacement of the major fissure (arrows). The etiology of the atelectasis was a left upper lobe lung cancer
Table 1. Differential diagnosis of air space opacification (ASO) A pulmonary mass is a round or ovoid pulmonary
Finding(s) Disease opacity 3 cm in diameter and is highly suspicious for
malignancy, typically lung cancer. The radiologist should
Focal/segmental ASO Pneumonia, contusion, infarct,
lung cancer (adenocarcinoma)
look for pertinent ancillary findings of malignancy in-
cluding other lung nodules, lymphadenopathy, pleural ef-
Lobar ASO Pneumonia, endogenous lipoid
pneumonia, adenocarcinoma fusion, and chest wall involvement.
Patchy ASO Infection, cryptogenic organiz-
Interstitial opacities may manifest with reticular, lin-
ing pneumonia, adenocarcino- ear, and/or small nodular opacities. As the normal inter-
ma, metastases, emboli stitium is not visible on radiography, visualization of pe-
Diffuse ASO Edema, hemorrhage, ripheral subpleural reticular opacities is always abnormal.
pneumonia A reticulonodular pattern occurs when abnormal linear
Perihilar ASO Edema, hemorrhage opacities are superimposed on micronodular opacities.
Peripheral ASO Eosinophilic pneumonia, acute
respiratory distress syndrome,
contusion Table 3. Ancillary findings in patients with interstitial lung disease
and differential diagnostic considerations
Rapidly changing/resolving ASO Edema, eosinophilic pneumo-
nia, hemorrhage Finding(s) Disease
Hilar lymph node enlargement Sarcoidosis, lymphangitic
carcinomatosis, viral pneumonia
Table 2. Differential diagnosis of the solitary pulmonary nodule Clavicular/osseous erosions Rheumatoid associated ILD
Finding(s) Disease (UIP, NSIP)
Pleural effusion Infection, edema
Granuloma
Pleural plaques Asbestosis
Primary malignancy: Lung cancer, carcinoid tumor
Solitary metastasis Hyperinflation PLCH, end-stage sarcoidosis,
lymphangioleiomyomatosis,
Hamartoma emphysema with UIP
Focal organizing pneumonia
Esophageal dilatation Scleroderma associated ILD
Hematoma (UIP, NSIP), recurrent aspiration
Conglomerate masses Silicosis, sarcoidosis, talcosis
granuloma, for which further evaluation is not required. Basilar sparing PLCH, sarcoidosis
However, many pulmonary nodules are indeterminate on Basilar predominance UIP, aspiration
radiography and require further assessment and charac- ILD, interstitial lung disease; UIP, usual interstitial pneumonia;
terization with computed tomography (CT) to exclude NSIP, nonspecific interstitial pneumonia; PLCH, pulmonary
malignancy (Fig. 2 and Table 2). Langerhans cell histiocytosis
96 M.L. Rosado-de-Christenson, J.S. Klein
a b
Fig. 2 a, b. Solitary pulmonary nodule. a PA chest radiograph shows a nodule in the left mid lung zone (arrow). b Unenhanced chest com-
puted tomography (CT) (lung window) shows a lobular soft-tissue nodule (arrow) in the left upper lobe. Typical carcinoid tumor was di-
agnosed by CT-guided needle biopsy
a b
Fig. 3 a, b. Interstitial lung disease. a PA chest radiograph of a 62-year-old man with chronic progressive dyspnea shows diffuse bilateral pe-
ripheral reticular interstitial opacities and low lung volumes. b Coronal contrast-enhanced chest CT shows subpleural reticulation, traction
bronchiectasis (arrow), and honeycomb cysts (arrowhead), consistent with usual interstitial pneumonia
Interstitial opacities frequently result from interstitial stitial opacities in diseases such as sarcoidosis, siliciosis,
edema characterized by perihilar haze, peribronchial and lymphangitic carcinomatosis.
thickening, septal thickening, and subpleural edema often The idiopathic interstitial pneumonias are a distinct
associated with cardiomegaly and pleural effusion [4]. group of disorders often characterized by basilar pre-
Associated radiographic findings can help limit the dif- dominant pulmonary fibrosis associated with volume
ferential diagnosis (Table 3). loss. The diagnosis usually requires further imaging with
Cells and fibrosis may also infiltrate the pulmonary in- high-resolution chest CT to assess for the presence or ab-
terstitium, producing reticular and reticulonodular inter- sence of honeycombing (Fig. 3).
A Systematic Approach to Chest Radiographic Diagnosis 97
a b
Fig. 4 a, b. Lymphadenopathy. a PA chest radiograph shows bilateral hilar, right paratracheal, and aortopulmonary window lymph node
enlargement associated with reticulonodular opacities in the upper and mid lung zones. b Contrast-enhanced chest CT shows bilateral hilar
and subcarinal lymphadenopathy and multifocal perilymphatic pulmonary nodules, consistent with suspected sarcoidosis
and a small portion of the left atrial appendage superior- Lines, Stripes, and Interfaces
ly. The right ventricle projects anteriorly and inferiorly on
the lateral chest radiograph. The posterior cardiac border The anterior and posterior junction lines represent the in-
is formed by the left ventricle inferiorly and the left atri- terfaces between the right and left upper lobes, and may
um superiorly. be thickened by fat, lymphadenopathy, or mediastinal
The heart must be assessed for its shape and size. The masses. The paravertebral stripes may be thickened by
normal pericardium is not visible on radiography. En- lymphadenopathy or fat. An abnormal convex contour of
largement of the cardiac silhouette may result from car- the upper azygoesophageal recess may result from sub-
diac enlargement and/or pericardial effusion. When the carinal lymphadenopathy or a bronchogenic cyst, while
volume of the latter is large, a water bottle heart mor- hiatus hernia often produces convexity of the lower one-
phology may be seen on frontal chest radiographs or an third of the azygoesophageal recess. Convexity of the
epicardial fat pad sign on lateral radiography. In the aortopulmonary reflection may be caused by mediastinal
epicardial fat pad sign, the pericardial effusion appears fat, lymphadenopathy, anterior mediastinal masses, or
as a curvilinear band of soft tissue 2 mm thick and out- anomalous vasculature [7].
lined by mediastinal fat anteriorly and by subepicardial
fat posteriorly. Constrictive pericarditis may manifest Mediastinal Masses
with linear pericardial calcification. Cardiac calcifica-
tions may correspond to coronary artery, valvular, or an- Mediastinal masses include primary and secondary neo-
nular calcifications or curvilinear calcification in a left plasms, mediastinal cysts, vascular lesions, glandular en-
ventricular aneurysm secondary to myocardial infarction. largement (thyroid and thymus), and herniations (hiatus
and Morgagni). Since 10% of mediastinal masses are vas-
Systemic Arteries cular in their etiology, a vascular lesion should always be
considered in a patient with a mediastinal contour abnor-
The normal aortic arch is readily visible on radiography mality.
and characteristically produces an indentation on the left The first step in the assessment of a mediastinal mass
inferior tracheal wall. With increasing aortic atheroscle- is determining that there is indeed a mediastinal abnor-
rosis and ectasia, a larger portion of the aorta is visible mality. Focal unilateral mediastinal masses are typically
and may exhibit intimal atherosclerotic calcification. The primary neoplasms, enlarged lymph nodes, cysts, and
left para-aortic interface projects through the left heart vascular aneurysms or anomalous vessels. While diffuse
and courses vertically towards the abdomen. The left sub- symmetric mediastinal widening without mass effect can
clavian artery is seen as a concave left superior mediasti- be seen in mediastinal lipomatosis, when lobulated or
nal interface on frontal chest radiography. A right aortic asymmetric it should suggest lymphadenopathy in a pa-
arch is usually associated with a right descending tho- tient with advanced lung cancer, metastatic disease, or
racic aorta. In the absence of associated congenital heart lymphoma (Fig. 5).
disease, it is also frequently associated with non-mirror The mediastinal mass should then be localized within
image branching characterized by an aberrant left sub- a radiographic mediastinal compartment (anterior, mid-
clavian artery, which may be seen as an indentation on dle or posterior) based on the lateral chest radiograph.
the posterior trachea on lateral chest radiography [7]. Ancillary findings should be noted, such as benign pres-
sure erosion in patients with paravertebral masses (typi-
Systemic Veins cal of neurogenic neoplasms). The cervicothoracic sign
can be used to localize superior mediastinal and thoracic
The azygos arch is visible at the right tracheobronchial inlet masses. Clinical factors such as age, gender, and
angle and normally measures 1 cm in diameter when presence or absence of symptoms allow the radiologist to
the patient is in the upright position. The azygos arch may provide a focused differential diagnosis prior to proceed-
be contained within an accessory azygos fissure, as an ing to cross-sectional imaging. Mediastinal widening in
anatomic variant. Enlargement of the azygos arch may the setting of trauma may represent hemorrhage from
occur in azygos continuation of the inferior vena cava, in traumatic vascular injury (Table 5) [8].
which the vertical portion of the azygos vein manifests as
a vertically oriented right-sided mediastinal interface [7].
Pleura
Pulmonary Arteries
Pleural abnormalities are those that involve the pleural
Enlargement of the pulmonary arteries may represent space, such as gas (pneumothorax) or fluid (pleural effu-
pulmonary hypertension and is typically associated with sion), or the pleural surfaces, including thickening (pleur-
enlargement of the pulmonary trunk. The pulmonary al plaques, neoplasms) with or without pleural calcifica-
trunk is visible as a left mediastinal interface located tion (pleural plaques, fibrothorax).
above the heart and below the aortopulmonary window Pneumothorax manifests with visualization of a pleur-
on frontal radiography. al line and may be traumatic, iatrogenic, or spontaneous.
A Systematic Approach to Chest Radiographic Diagnosis 99
a b
Table 5. Differential diagnosis of mediastinal masses Spontaneous pneumothorax is categorized as primary (no
Anterior Middle Posterior visible underlying lung disease) and secondary (underly-
ing lung disease).
Lymphoma Lung cancer Neurogenic
neoplasm
Pleural effusion is categorized as transudative or ex-
Thymic neoplasm Lymphadenopathy Aneurysm udative, and manifests as blunting of the costophrenic an-
(lymphoma, metastases) gle, producing a meniscus. Transudative pleural effusions
Germ cell neoplasm Foregut cyst Extramedullary are usually secondary to heart failure. Exudative effu-
hematopoiesis sions are often secondary to infection, malignancy, or
Thyroid goiter Hiatus hernia pulmonary infarction. Pleural effusions may exhibit loc-
ulation, which should raise the suspicion of empyema in
a b
a patient with signs and symptoms of infection. An air- wall could indicate an air leak in the setting of trauma,
fluid level in the pleural space in the absence of prior in- pneumomediastinum or pneumothorax.
tervention is diagnostic of a bronchopleural fistula. Mas-
sive pleural effusions and pleural effusions with associat-
ed pleural nodules should suggest malignancy. Circum- References
ferential nodular pleural thickening is virtually diagnos-
tic of malignancy. The differential diagnosis includes 1. Ashizawa K, Hayashi K, Aso N et al (2001) Lobar atelectasis:
metastatic disease, malignant pleural mesothelioma (Fig. diagnostic pitfalls on chest radiography. Br J Radiol 74:89-97.
2. Abbott GA (2012) Approach to atelectasis and volume loss. In:
6), and lymphoma [9, 10]. Rosado-de-Christenson ML (ed) Diagnostic imaging Chest.
Diffuse pleural calcification may reflect a fibrothorax Amirsys, Salt Lake City, UH, pp 1-56.
and can result from prior hemothorax or tuberculous 3. Hodnett PA, Ko JP (2012) Evaluation and management of in-
empyema. Multifocal bilateral discontinuous pleural cal- determinate pulmonary nodules. Radiol Clin North Am
cifications are typical of pleural plaques in asbestos- 50:895-914.
4. Cardinale L, Priola AM, Moretti F et al (2014) Effectiveness
related pleural disease [10]. of chest radiography, lung ultrasound and thoracic computed
tomography in the diagnosis of congestive heart failure. World
J Radiol 6:230-237.
Chest Wall 5. Boiselle P (2008) Imaging of the large airways. Clin Chest
Med 41:617-626.
6. Felson B (1973) Localization of intrathoracic lesions. In: Fel-
The symmetry of normal chest wall structures such as son B (ed) Chest roentgenology. Saunders, Philadelphia, PA,
the breast shadows in women and the ribs and scapulae pp 39-41.
should be analyzed to detect chest wall abnormalities. 7. Gibbs JM, Chandrasekhar CA, Ferguson EC (2007) Lines and
Nonsurgical absence of a portion of a rib or vertebral stripes: Where did they go? From conventional radiography
to CT. Radiographics 27:33-48.
body may be instrumental in making the diagnosis of ma- 8. Whitten CR, Khan S, Munneke GJ et al (2007) A diagnostic
lignancy. Rib destruction adjacent to a thoracic mass is approach to mediastinal abnormalities. Radiographics 27:657-
indicative of chest wall involvement [11]. Benign pres- 671.
sure erosion is characteristic of neurogenic neoplasms 9. Cardinale L, Ardissone F, Asteggiano F et al (2013) Diffuse
and chest wall vascular abnormalities. Soft-tissue calcifi- neoplasms of the pleural serosa. Radiol Med 118:366-378.
10. Gallardo X, Castaner E, Mata JM (2000) Benign pleural dis-
cation may indicate prior trauma (myositis ossificans), eases. Eur J Radiol 34:87-97.
collagen vascular disease (dermatomyositis), or the pres- 11. Tateishi U, Gladish GW, Kusumoto M et al (2003) Chest wall
ence of a vascular lesion (hemangioma) or a bone-form- tumors: Radiologic findings and pathologic correlation. Radio-
ing malignancy (chondrosarcoma). Gas within the chest graphics 23:1491-1508.
IDKD 2015-2018
Soft-Tissue Masses
These must first be differentiated from asymmetries aris-
ing from patient positioning, contralateral surgery (mas-
tectomy), and atrophy (poliomyelitis and stroke). Prima-
ry tumors are rare. Malignancies include fibrosarcoma
and malignant fibrohistiocytoma, with lipomas as the
most common benign primary tumor. Neurogenic tu-
mors, such as neurofibromas, arise from intercostal
Fig. 1. Extraparenchymal mass forms an obtuse angle with the
chest wall in a patient with a solitary plasmacytoma. Note the bone nerves or the paraspinal ganglia and can be found at mul-
destruction indicating the origin of the tumor outside the lung and tiple sites in neurofibromatosis type 1 (von Reckling-
pleura hausen disease).
Inflammatory Diseases of the Chest Wall and cardiac injury. Sternal fractures are very difficult to
visualize on frontal chest radiographs and may be seen on
Bacterial infection of the chest wall is rare and typically the lateral view, but nearly all are visible by CT, which
involves the ribs and sternum. Osteomyelitis can spread will demonstrate any adjacent hematoma. Significantly
hematogenously or directly from an adjacent infectious displaced sternal fractures should lead one to look for as-
process, typically Staphylococcus aureus or Pseudo- sociated cardiac injury. Lower rib fractures can be asso-
monas aeruginosa. Radiography may demonstrate soft- ciated with hepatic, renal, or splenic injury. Abdominal
tissue swelling overlying rib destruction and a periosteal radiographs are helpful to depict fractures of the lower
reaction in chronic cases but computed tomography (CT) ribs, as they are better visualized.
and MRI are much more sensitive and specific. CT is A flail chest (or segment) occurs with five or more
limited in its detection of bone marrow abnormalities, consecutive rib fractures (or three or more ribs fractured
while MRI allows the early detection of osteomyelitis by in two places). The flail segment moves paradoxically
showing bone marrow edema on T2 and marrow hypo- during respiration, with resultant ventilatory compro-
intensity on T1 images. However, CT is more sensitive mise. Fractures of the thoracic spine account for 1530%
than MRI in showing cortical destruction and a periosteal of all spinal fractures. The most vulnerable segment is at
reaction. the thoracolumbar junction (T9T12). Most (~70%) tho-
Chest wall involvement with tuberculosis (TB) is un- racic spine fractures are visible on radiographs, but CT
common, occurring via hematogenous spread or, more shows almost all of them and, more importantly, will
rarely, by direct extension from the underlying lung or demonstrate the displacement or retropulsion of bone
pleura; the latter is referred to as empyema necessitans. fragments into the spinal canal and cord.
Chest wall involvement in TB manifests as osseous and
cartilaginous destruction and soft-tissue masses with cal- Bone Masses
cifications and rim enhancement [2]. Fungal infections of
the chest wall occur in immunocompromised patients, The most common non-neoplastic tumor of the thoracic
with Aspergillus accounting for 8090% of cases [3]. In skeleton is fibrous dysplasia, which accounts for 30%
15% of the cases of thoracic actinomycosis, the fungus of the benign bone tumors of the chest wall [7]. These
invades the chest wall, creating fistulas and empyema ne- slow-growing tumors often occur in the posterolateral
cessitans. Postoperative patients are at risk of Nocardia ribs and are usually asymptomatic unless pressure
infection. symptoms or pathologic fractures occur. Fibrous dys-
plasia manifests as an expansile lytic lesion with a hazy
Congenital Diseases of the Chest Wall or ground glass appearance. Plasmacytoma and multiple
myeloma are the most common malignant neoplasms of
Poland syndrome is characterized by aplasia of the the thoracic skeleton and can present as extraparenchy-
pectoralis major muscles, and occasionally by hypopla- mal masses with bone destruction, similar in appear-
sia of the ipsilateral chest wall, scapula, and ribs. Imag- ances to other metastases [8]. Bone lesions are well de-
ing reveals a smaller and more lucent hemithorax, fined; they have a punched out appearance. In advanced
while cross sectional imaging demonstrates the absence cases there is marked erosion as well as areas of expan-
of the pectoral muscles and a smaller, hypoplastic sion and destruction of bone cortex, sometimes with
hemithorax [4]. thick ridging around the periphery, giving a soap bub-
ble appearance.
Pectus Excavatum
Thalassemia
In this variant in chest wall development (incidence
1/4001/1000), the sternum is depressed relative to the With extramedullary hematopoiesis, hematopoietic tis-
anterior chest (ribs) and often tilts rightward, with the sues outside the bone marrow produce blood cells. This
mediastinum leftward [3, 5]. In most cases, it is a cos- occurs mostly in the chronic hemolytic anemias, with
metic issue only, but in severe cases, it can lead to pain, thalassemia being the most common. Normal marrow
dyspnea, and restrictive lung disease. On radiographs, the tissue can expand outside the medulla through perme-
right side of the mediastinum is indistinct, simulating ative erosions or by the reactivation of previously dor-
middle lobe airspace disease. The depressed sternum is mant hematopoietic tissues [9]. Ineffective erythro-
visible on the lateral projection [3]. poiesis leads to expansion of the bone marrow space in
the ribs, long bones, vertebral column, skull, and, char-
Injuries to the Thoracic Skeleton, Rib Fractures, and Trauma acteristically, facial bones, and may lead to osteoporo-
sis. Paravertebral and rib lesions are most often seen in-
Upper rib fractures suggest severe trauma, and a search cidentally but can occasionally cause mass effect on the
should be made for associated injury of the aorta, great spinal canal and cord compression. On radiography, bi-
vessels, and brachial plexus [6]. Fractures of the medial lateral lobulated paravertebral masses and rib expansion
clavicle and sternum are also associated with vascular are seen.
Diseases of the Chest Wall, Pleura, and Diaphragm 103
Morgagni Hernia
Less than 10% of congenital hernias are Morgagni hernias.
They arise anteromedially due to the failed fusion of the
sternal and costal portions of the hemidiaphragm [18, 19].
They are more common in females and occur more fre-
Fig. 4. Nodular thickening of the left pleura, including the mediasti- quently on the right, possibly due to greater support pro-
nal aspect, in a 43-year-old patient with malignant mesothelioma
vided by the pericardial attachments on the left. In adults,
protrusion of omentum is the most common form, and on-
ly rarely, bowel, stomach or liver. On imaging, Morgagni
hernia will appear as a fatty mass in the right cardiophrenic
from the abdominal structures on radiographs. Multipla- angle that can be difficult to differentiate from a prominent
nar CT is ideally suited to visualize the diaphragm due to epicardial fat pad. The differential diagnosis will include
its differential attenuation from adjacent intraperitoneal lipoma, teratoma, thymoma, thymolipoma, or liposarcoma.
fat and aerated lung. CT shows the diaphragm as a smooth Detection of displaced curvilinear omental vessels within
soft-tissue attenuation band-like structure that is thicker the mass or coursing across the diaphragmatic defect,
at the edges and thinner (sometimes imperceptibly) to- characteristic of Morgagni hernia, is best appreciated on
wards the central tendon [15]. Folds or bundles of muscles coronal CT or MRI.
(slips) run obliquely along the inferior surface, are often
asymmetric, and can have a mass like appearance if they Diaphragmatic Eventration
are seen in cross-section, especially in males, where they
are more prominent. Multiple slips can give the dia- An eventration is a congenital circumscribed area of di-
phragm a scalloped appearance where the muscle bulges aphragmatic muscular aplasia or thinning that allows
upwards on either side of them. Defects and hernias may bulging of the abdominal contents towards the thorax,
be congenital (Bochdalek, Morgagni or eventration) or ac- due to the relative pressure gradient between the peri-
quired (hiatus hernia, trauma, or paralysis). Predisposing toneum and the thorax [17]. Unlike a true hernia, the ab-
conditions for all abdominal hernias include pregnancy, dominal contents are still confined by a layer of pleura,
trauma, obesity, chronic constipation, and chronic cough. diaphragmatic tendon and peritoneum [15]. Symptoms
are rare and include tachypnea, dyspnea, recurrent pneu-
Bochdalek Hernia monia, and failure to thrive. These cases can be treated
with surgical plication (folding or tucking and suturing).
In adults, 90% of congenital hernias are Bochdalek her- Eventration occurs most commonly on the right side an-
nias. These hernias are more common on the left because teromedially and can become more pronounced with in-
of the protective effect of the liver on the right posterior- creasing intra-abdominal pressure, usually as a result of
ly [16]. Bochdalek hernias occur posterolaterally due to obesity. Imaging shows an anterior elevation of the
the persistence of a gap between the costal and vertebral hemidiaphragm, since the thinning and weakening in-
portions of the diaphragm, through which peritoneal fat volve this portion, with a normal posterior position. With
and, less commonly, kidney and other solid organs or diaphragmatic paralysis, the posterior portion will also be
bowel protrude [17]. Neonates and young children under elevated. On CT, there may be relative thickening of the
a year may present with respiratory symptoms, while old- muscle at the edge of the eventration. In some cases the
er children and adults more likely present with acute or edges are undercut, with upwards ballooning of the even-
chronic gastrointestinal symptoms, if any. Rarely, patients trated portion [17].
can present with bowel incarceration, strangulation, per-
foration, or shock. However, nowadays, most Bochdalek Esophageal Hiatus Hernia
hernias are discovered incidentally on CT imaging, as
small defects, particularly in patients with longstanding The esophageal hiatus is at the T10 level and contains the
obstructive lung disease. For incidentally discovered her- esophagus, vagus nerve, and sympathetic nerve branches
nias, no further action is needed. Imaging appearances of [16]. This ring functions as an anatomic sphincter by
Diseases of the Chest Wall, Pleura, and Diaphragm 107
Table 1. Typical and atypical features of pulmonary sarcoidosis at The latter results in a mosaic pattern on expiration CT
high-resolution computed tomography that may even precede the appearance of other parenchy-
Typical features mal abnormalities and is considered an early sign of ex-
Lymphadenopathy: hilar, mediastinal, bilateral symmetric, well de- tralymphatic disease spread.
fined Sarcoidosis may present with very characteristic high-
Nodules: micronodules (2- to 4-mm), well defined, larger coalesc-
ing nodules
resolution CT (HRCT) features that secure the diagnosis
Perilymphatic distribution: subpleural, interlobular septa, peri- and obviate further invasive procedures. However, sar-
bronchovascular coidosis is also known to present with less typical fea-
Fibrosis: reticular densities, architectural distortion, traction tures (Table 1).
bronchiectasis, volume loss Especially in advanced disease stages with fibrosis,
Distribution: predominance of parenchymal abnormalities in upper
lobes and perihilar areas the differential diagnosis between sarcoidosis and other
fibrosing diseases can be very challenging. Table 2 sum-
Atypical features marizes the imaging findings that are helpful in differen-
Lymphadenopathy: unilateral, isolated, in atypical mediastinal lo-
cations tiating several diseases.
Airspace consolidations: conglomerate masses, confluent alveolar Complications include the development of aspergillo-
opacities (alveolar sarcoidosis) mas and pulmonary hypertension. Mycetomas are a typ-
Linear opacities: interlobular septal thickening ical complication of stage IV sarcoidosis: they present as
Fibrocystic changes: cysts, bullae, honeycomb like opacities, up- a soft-tissue mass located in a preexisting cavitation in
per lobe predominance
Airway involvement: atelectasis, mosaic pattern patients with fibrotic sarcoidosis. Life-threatening
Pleural disease: effusion, chylo/hemothorax, pneumothorax, pleur- hemoptysis may require immediate interventional thera-
al plaques py (embolization of the bronchial arteries). Pulmonary
hypertension occurs in patients with end-stage fibrosis
but may be also caused by mediastinal fibrosis, extrinsic
bronchiectasis, due to the surrounding fibrosis. Granulomas compression of the pulmonary arteries by lym-
within the bronchial walls lead to irregular bronchial wall phadenopathy, or intrinsic sarcoid vasculopathy, includ-
thickening with narrowing of the large and small airways. ing features of pulmonary veno-occlusive disease.
Collagen Vascular Diseases (NSIP). Of all connective tissue diseases, RA is the most
common to present with a UIP pattern. The CT findings
Collagen vascular diseases are characterized primary on in interstitial pneumonia associated with RA are often in-
clinical grounds, namely, typical clinical complaints and distinguishable from those of the idiopathic varieties;
physical examination findings. The presence of specific however, other findings, such as nodules, pulmonary ar-
autoantibodies may greatly assist in the correct diagnosis. terial enlargement, or pleural abnormality, may provide a
Involvement of the respiratory system is often seen in clue to the underlying diagnosis. UIP and NSIP both typ-
collagen vascular diseases and results in significant mor- ically demonstrate a subpleural and basilar distribution of
bidity and mortality. It is important to note that lung ab- findings. A confident diagnosis of UIP may be made
normalities may precede the other clinical manifesta- when honeycombing and reticulation is also present. In
tions, sometimes by more than 5 years. NSIP there is an absence of honeycombing and sub-
Lung injury from collagen vascular disease can affect pleural sparing is present.
each portion of the lung, the pleura, alveoli, interstitium, The introduction of a new generation of biologic
vasculature, lymphatic tissue, and both the large and agents used to treat RA has resulted in a new array of po-
small airways. Commonly, more than one compartment is tential pulmonary side effects. The most important of
involved. Many of the parenchymal manifestations of these is the impaired immunity related to the use of anti-
collagen vascular disease are similar to those of the idio- tumor necrosis factor (TNF) antibodies (etanercept, in-
pathic interstitial pneumonias and can be classified using fliximab, and adalimumab), which has resulted in a sub-
the same system. stantially increased incidence of tuberculosis (sometimes
Lung biopsy is rarely obtained in patients with a de- disseminated or extraarticular) and of nontuberculous
fined connective tissue disease. For this reason, HRCT of- mycobacterial infection.
ten determines the predominant pattern of injury, which in Low-dose methotrexate may be associated with suba-
turn is important in determining treatment and prognosis. cute hypersensitivity pneumonitis in 25% of patients.
While many patterns of injury are associated with the Pre-existing radiographic evidence of interstitial lung dis-
particular collagen vascular diseases, certain patterns are ease in patients with RA suggests a predisposition to the
more representative than others. For instance, nonspecif- development of methotrexate pneumonitis. Infections,
ic interstitial pneumonia is the most common pattern in such as pneumocystis pneumonia, are another potential
patients with scleroderma. It is also important to note that complication of therapy.
more than one pattern of injury may be present in the
same patient. Scleroderma (Progressive Systemic Sclerosis)
Pulmonary arterial hypertension usually causes en- superimposed on a background of reticular abnormality
largement of the main and proximal pulmonary arteries, with traction bronchiectasis. This pattern reflects the
as seen on chest radiograph or CT, but normal-sized pul- characteristic histologic combination of OP and fibrotic
monary arteries do not exclude the diagnosis. The pres- NSIP. On serial evaluation, the changes of consolidation,
ence of pericardial thickening or fluid in patients with ground glass abnormality, reticular abnormality, and trac-
scleroderma is also a strong predictor of echocardio- tion bronchiectasis may all be partially reversible with
graphic pulmonary hypertension. treatment. Consolidation may also progress to reticular
There is an increased prevalence of lung cancer in pa- abnormality.
tients with scleroderma, especially in those with lung fibro- PM/DM can be associated with other collagen vascular
sis. The relative risk of malignancy ranges from 1.8 to 6.5. diseases (SLE, scleroderma, RA, Sjgren syndrome).
There is an increased risk for malignancy (especially breast
Systemic Lupus Erythematosus lung, ovary, and stomach malignancies), which in up to
20% of patients is concurrently diagnosed within 1 year of
Pleuritis is the most commonly seen pleuropulmonary follow-up.
manifestation of systemic lupus erythematosus (SLE),
found in 4060% of patients, and may or may not be as- Sjgren Syndrome (SS)
sociated with pleural effusion.
Fibrotic interstitial lung disease is less frequent in SLE CT provides substantial information regarding the pat-
than in the other collagen vascular diseases. While pul- terns of pulmonary involvement by Sjgren syndrome
monary infection is said to be the most common pul- (SS). These may be divided into airway abnormality, in-
monary complication of lupus, acute pulmonary hemor- terstitial fibrosis, pulmonary hypertension, and those
rhage is also an important pulmonary complication of suggestive of lymphoid interstitial pneumonia.
this condition, characterized radiologically by diffuse or Airway-related abnormalities consist of bronchial wall
patchy consolidation and ground glass abnormality. thickening, bronchiectasis, bronchiolectasis, and the tree-
In patients with lupus, acute lupus pneumonitis is a in-bud pattern. Small-airway disease may manifest as a
poorly defined entity, characterized by a variable degree mosaic attenuation pattern and expiratory air trapping.
of respiratory impairment accompanied by focal or dif- NSIP and lymphoid interstitial pneumonia (LIP) are
fuse pulmonary consolidation. It is now believed that the most common patterns of parenchymal lung disease.
most cases previously identified as lupus pneumonitis NSIP resembles that seen in scleroderma. LIP is charac-
probably represented acute interstitial pneumonia with or terized by ground glass abnormality due to the homoge-
without pulmonary hemorrhage. It is mostly associated neous lymphocytic infiltration. Peribronchovascular, cen-
with renal infiltration and multi-organ failure. trilobular, and subpleural nodules may also be seen, and
Other complications of lupus may include diaphrag- cysts measuring 530 mm are often present. These cysts
matic dysfunction, pulmonary hypertension, and pul- frequently contain thin septa; they may be associated with
monary thromboembolism, which may be related to the soft-tissue nodules sitting either close to them or in their
presence of antiphospholipid antibodies. Diaphragmatic walls. Similar cysts may be found in follicular bronchio-
dysfunction, thought to be due to a diaphragmatic myo- litis. These changes are ascribed to bronchiolar obstruc-
pathy, is manifested by reduced lung volumes (shrinking tion on the basis of lymphocytic wall infiltration. Cysts
lungs with plate-like atelectasis). are helpful in distinguishing LIP from lymphoma.
Lymphoma should be suspected if consolidation, large
Polymyositis/Dermatomyositis nodules (1 cm), mediastinal lymphadenopathy or effu-
sions are present. However, similar large, pseudo-alveo-
The presence of interstitial lung disease (ILD) in poly- lar, poorly defined nodules were found in four patients
myosistis/dermatomyositis (PM/DM) correlates strongly with combined amyloidosis and LIP. In contrast to other
with the presence of anti-Jo-1 antibodies. Between 50% cystic lung diseases, such as lymphangioleiomyomatosis,
and 70% of patients who are anti-Jo-1 positive have ILD the cysts of LIP show a peribronchovascular and lower
whereas the frequency of ILD falls to about 10% if anti- lung predominance.
bodies are absent. ILD may antedate myositis in patients
with anti-Jo-1 antibodies. Mixed Connective Tissue Disease
The most common pathological findings are NSIP and
organizing pneumonia (OP), often occurring in combina- Mixed connective tissue disease (MCTD) is an overlap
tion. As with other collagen vascular diseases, the occur- syndrome that is a distinct clinicopathological entity. The
rence of interstitial pneumonia may precede the develop- principal characteristics are the presence of: (1) features
ment of clinical myositis. of SLE, scleroderma, and PM/DM, occurring together or
Lung disease associated with PM/DM or with the evolving sequentially during observation; and (2) anti-
antisynthetase syndrome, a closely related entity, often bodies to an extractable nuclear antigen (RNP).
has a typical CT appearance consisting of confluent Pulmonary involvement is common in MCTD. A study
ground glass opacity and consolidation in the lower lobes of 144 unselected patients found CT evidence of infiltrative
114 C. Schaefer-Prokop, B.M. Elicker
Table 3. The most recent classification of systemic vasculitis: from the 2012 Chapel Hill Consensus conference
Vasculitis Pulmonary-renal Pulmonary Pulmonary
syndrome hemorrhage hypertension
Small-vessel vasculitis, GPA Pulmonary involvement 90% In about 10%
ANCA-associated (formerly Wegeners) Renal involvement 80%
Eosinophilic granulomatous up to 25% Rare
vasculitis with polyangiitis
GPA(Church Strauss)
Microscopic polyangiitis Frequently associated Frequent
with necrotizing
glomerulonephritis
Small vessel vasculitis, Cryoglobulinemic IgA
immune-complex type (Henoch Schoenlein)
HUV
Anti-GBM
(formerly Goodpastures)
Medium-vessel vasculitis PAN Very rare idiopathic,
associated with hepatitis B;
part of RA, SLE
Large-vessel vasculitis Kawasaki syndrome
GCA Rare, focal Rare
Takayasu arteritis Rare, focal Rare
Collagen vascular diseases RA Very rare, diffuse Up to 60%
SS Rare, diffuse with necrotizing 1060%
vasculitis
MCTD Rare, diffuse Up to 45%
SLE Up to 60%, late In at least 4%, diffuse 545%
ANCA, anti-neutrophil cytoplasmic antibody; GPA, granulomatous vasculitis with polyangiitis; HUV, hypocomplementemic vasculitis;
GBM, glomerular basement membrane; PAN, polyarteritis nodosa; GCA, giant cell arteritis; RA, rheumatoid arthritis; SS, Sjgren syn-
drome; MCTD, mixed connective tissue disease; SLE, systemic lupus erythematosus
lung disease in 67%. Many affected patients are asympto- established collagen vascular diseases is not yet known.
matic. The pulmonary abnormalities resemble those seen in The term lung-dominant CTD has been proposed for
SLE, SS, and PM/DM. Thus, pleural thickening and pleur- the subgroup of patients with: (1) NSIP, UIP, LIP, OP, or
al and pericardial effusions are common. Ground glass at- diffuse alveolar damage in the lung determined by his-
tenuation is the most frequent parenchymal abnormality. tology or HRCT; (2) insufficient extrathoracic features of
The CT pattern corresponds most closely to that of NSIP. a definite collagen vascular disease but a combination of
Less frequent findings include honeycombing, consolida- serologic features suggesting an autoimmune process;
tion, and poorly defined centrilobular nodules. and (3) no identifiable alternative cause for interstitial
Other important complications of MCTD are pul- pneumonia.
monary arterial hypertension, and esophageal dysmotili-
ty, with sequelae of recurrent aspiration.
Pulmonary Vasculitis/Diffuse Alveolar Hemorrhage
Undifferentiated Connective Tissue Disease (or Interstitial
Lung Disease with Autoimmune Features) The pulmonary vasculitides encompass a clinically, radio-
logically, and histopathologically heterogeneous group of
Not all patients with interstitial lung disease meet the cri- diseases that are usually associated with a systemic vas-
teria of having a collagen vascular disease. Among patients culitis. The clinical symptoms and radiologic signs sug-
with histologically proven UIP or NSIP whose disease was gestive of pulmonary vasculitis include diffuse alveolar
originally diagnosed as idiopathic, because it did not ful- hemorrhage (DAH), acute glomerulonephritis, upper-air-
fill the criteria of one of the established collagen vascular ways disease, lung or cavitary nodules, mononeuritis
diseases, subgroups of patients with one or more serolog- multiplex, and palpable purpura.
ical features of an autoimmune process have been subse- The classification of systemic vasculitis remains con-
quently identified. Several terms have been suggested to troversial. In the most recent classification, the Chapel Hill
describe the condition detected in these subgroups, includ- Consensus Conference (CHCC) of 2012 (Table 3), the
ing undifferentiated connective tissue disease (UCTD), in- main consideration remained the predominant vascular
terstitial lung disease with autoimmune features, and lung- size involved, but the presence of ANCA (anti-neutrophil
dominant CTD. Whether these patients have a better prog- cytoplasmic antibody) was newly added. Small vessel vas-
nosis than patients without these autoimmune features culitis (SVV) is divided into ANCA-associated vasculitis
or even a prognosis comparable to that of patients with (AAV) and immune complex SVV. AAV represents necro-
Pulmonary Manifestations of Systemic Diseases 115
tizing vasculitis with only few or no immune deposits pre- tem (up to 50%). Involvement of the peripheral nerves is
dominantly affecting the small vessels; patients are MPO- more frequent while glomerulonephritis (GN) and pul-
ANCA or PR3 ANCA positive but in a minority of cases monary hemorrhage are less common than in GPA.
ANCA-negative. ANCA specificity should be indicated Up to 30% of patients are ANCA-positive, mostly
because it appears to identify distinct categories of disease. MPO-ANCA. Histology demonstrates small vessel (ar-
The combined involvement of the lungs and kidneys teries and veins) vasculitis with eosinophilic infiltrates,
by some of these diseases accounts for their description and vascular and extravascular granulomas. Five-year
as pulmonary-renal syndrome. Note that capillaritis/ survival is reported in up to 80% of patients; 50% of the
vasculitis can also manifest in the setting of the various deaths are related to cardiac involvement.
collagen vascular diseases. Upper-airways disease and pulmonary abnormalities
are seen in up to 70% of patients with CSS. The most
ANCA-Associated Granulomatous Vasculitis with Polyangiitis common pulmonary findings include transient, multifo-
(GPA, Formerly Wegeners Disease) cal, and non-segmental consolidations without zonal
predilection. The presence of (non-cavitating) small nod-
Granulomatous vasculitis with polyangiitis is the most ules or diffuse reticular densities has also been reported.
common of the AAV. It affects the sinuses, kidneys, and If located subpleurally, the consolidations mimic
lungs, resulting in the classic triad of symptoms com- eosinophilic pneumonia. Diffuse consolidations or
prising sinusitis and/or tracheobronchitis, pathological ground glass attenuation reflect hemorrhage.
chest X-ray (with or without hemoptysis), and micro- Airway abnormalities are also an important thoracic
hematuria. However, any part of the body may be in- manifestation and include wall thickening, dilatation, small
volved. Females and males are affected equally and at nodules, and mosaic perfusion. Increased interlobular septa
any age (mostly age 4055 years). Airway involvement may reflect edema caused by cardiac and/or renal involve-
is more frequent in males. The most common cause of ment or eosinophilic infiltration of the septa. Eosinophilic
death is renal failure. With treatment, the 24-month sur- pleural effusion is seen in up to 50% of patients.
vival is 80%. The factors initiating disease are unknown, Airway changes consisting of tree-in-bud, bronchial
but current data support the involvement of (recurrent) wall thickening, and bronchial dilatation are likely relat-
infection for GPA. Up to 80% of patients are (in the ed to asthma and develop in almost patients. Wall thick-
course of disease) ANCA-positive (mostly PR3-AN- ening can also be caused by eosinophilic involvement.
CA). Histologic findings include necrotizing granulo-
matous vasculitis of the small to medium vessels with- Microscopic Polyangiitis/Diffuse Alveolar Hemorrhage
out associated infection.
The most frequently seen pulmonary abnormalities are Microscopic polyangiitis is a systemic necrotizing small-
multiple nodules with or without a CT halo sign (peri- vessel vasculitis with granulomatous inflammation. Clin-
nodular hemorrhage). They tend to involve the subpleur- ically, it is characterized by a long prodromal phase with
al regions, but there is no predilection for upper or lower weight loss and fever followed by a rapidly progressive
lung zones. Although the nodules can be as large as 10 GN. Microscopic polyangiitis is the most common cause
cm, most are smaller. Cavitations have thick walls. Pe- of pulmonary-renal syndrome: rapidly progressive GN is
ripheral wedge-shaped consolidations resemble infarcts seen in up to 90%, pulmonary involvement in up to 50%.
and may also cavitate. Diffuse consolidations or ground More than 75% of patients are ANCA-positive in the
glass opacities represent pulmonary hemorrhage. Fibrot- course of disease, mostly MPO ANCA.
ic changes reflect preexisting disease. Microscopic polyangiitis is characterized by the com-
In addition there may be a concentric thickening of the bination of GN and diffuse alveolar hemorrhage (DAH).
tracheal or bronchial walls with diameter reduction, po- Imaging findings include patchy, bilateral, or diffuse air-
tentially causing atelectasis. Bronchial abnormalities space opacities. The opacity can show features of consol-
mainly involve the segmental and subsegmental bronchi. idations and ground glass, depending on the amount of
Involvement of the subglottic trachea is most typical. alveolar filling by blood. The opacifications may be dif-
fuse or more pronounced in the dependent lower parts of
Churg-Strauss Syndrome the lungs. A halo surrounding a consolidation or nodule
underlines the character of the hemorrhage. During the
Eosinophilic granulomatous disease with polyangiitis phase of resorption, interlobular lines (crazy paving)
(Churg-Strauss syndrome, CSS) is caused by a small-ves- are increasingly seen.
sel systemic vasculitis that almost exclusively occurs in Repeated hemorrhage leads to fibrosis, with honey-
patients with asthma and is characterized by a marked combing, reticulation, and traction bronchiectasis.
serum eosinophilia. Clinically, radiologically, and patho-
logically it combines features of GPA and eosinophilic Goodpasture Syndrome/DAH
pneumonia (allergic granulomatosis and angiitis). Many
other organs may be involved including the heart (up to Goodpasture syndrome describes the clinical triad of
47%), the skin (up to 40%), and the musculoskeletal sys- circulating anti-glomerular basement membrane anti-
116 C. Schaefer-Prokop, B.M. Elicker
body, DAH, and (necrotizing) GN. Presenting symp- Three radiographic manifestations of amyloidosis are
toms comprise an acute onset of dyspnea and hemo- described: nodular parenchymal, diffuse alveolar septal,
ptysis. and tracheobronchial. The nodular parenchymal form
Histologically, there is a linear deposition of IgG along presents radiographically as one or more solid lung nod-
the glomerular basement membranes, such that renal ules, often found incidentally. These may be confused
biopsy is used to establish the diagnosis. The serology with malignancy, particularly when spiculated borders
comprises c-ANCA or p-ANCA positivity in 30% and are present. A slow growth rate is typical, and calcifica-
anti-basement membrane antibodies in 90% of patients. tion may be present. The diffuse alveolar septal form is
Young males are more often affected than females characterized by the presence of widespread deposits
(M:F9:1), although the condition is also seen in elder- throughout the lungs. The most common findings are
ly females. With treatment, the prognosis is good. Recur- small nodules (typically in a perilymphatic distribution),
rent episodes cause pulmonary fibrosis. consolidation, ground glass opacity, and reticulation. The
Imaging features consists of diffuse or patchy ground presence of calcification associated with these abnormal-
glass or consolidations based on alveolar hemorrhage ities may be particularly suggestive. The tracheobronchial
that typically resolve within days. The subpleural space is form shows extensive thickening of the trachea and/or
usually spared, with predominance instead of the peri- bronchi with or without calcification.
hilar areas in the mid and lower lung zones. Pleural effu- The imaging findings of light-chain deposition disease
sion is uncommon. After recurrent episodes, traction are less well described. There are two types: nodular and
bronchiectasis, a reticular pattern, and honeycombing diffuse. The nodular type presents with scattered solid
may evolve. nodules, often associated with cysts. The nodules are typ-
DAH caused by pulmonary capillaritis (Table 3) needs ically located within or adjacent to the wall of the cysts.
to be differentiated from bland pulmonary hemorrhage The diffuse type may show diffuse small nodules, resem-
(e.g., coagulation disorders, mitral stenosis, drug- bling the diffuse alveolar septal type of amyloidosis.
induced, etc.) and hemorrhage associated with diffuse
alveolar damage (drug-induced, acute respiratory distress Erdheim-Chester Disease
syndrome, bone marrow transplantation, or crack cocaine
inhalation). Erdheim-Chester disease is an infiltrative disorder in
which non-Langerhans cell histiocytes are found in one
or more organ systems. The primary organs affected in-
Miscellaneous Systemic Disorders clude bones, brain, kidneys, and the cardiovascular sys-
tem. Middle-aged males are most commonly affected.
Inflammatory Bowel Disease While pathology is helpful in excluding lymphoprolifer-
ative malignancies, the pathological findings may be non-
Ulcerative colitis and Crohn disease are associated with specific; thus typical radiographic features are a key to
a wide variety of pulmonary complications, which may the correct diagnosis. Characteristic imaging findings in
precede the diagnosis of inflammatory bowel disease the chest include peri-adventitial aortic soft-tissue thick-
(IBD) or may occur years after the initial diagnosis, and ening that may involve the aorta and its main branches
even after complete colectomy for ulcerative colitis. In- diffusely, giving rise to the term coating of the aorta.
deed there is some suggestion that pulmonary complica- The most common lung finding is smooth interlobular
tions may be more common after surgical treatment, per- septal thickening, resembling pulmonary edema. The
haps because anti-inflammatory treatment is withdrawn. most characteristic finding outside of the chest is cir-
Both Crohn disease and ulcerative colitis can be associ- cumferential perinephric soft tissue, resembling lym-
ated with tracheobronchitis and airway stenosis. phoma.
Bronchiectasis and bronchial wall thickening are also
common. Small-airway involvement can have a pattern of IgG4-Related Sclerosing Disease
panbronchiolitis. Parenchymal abnormalities associated
with IBD include OP, pulmonary hemorrhage, and, in This disorder was originally thought to be isolated to the
Crohn disease, granulomatous infiltration. pancreas and thus was previously called lymphoplasmo-
cytic sclerosing pancreatitis. However, it is now realized
Amyloidosis and Light-Chain Deposition Disease that IgG4-related sclerosing disease is a systemic order
associated with IgG4 plasma cells and with fibrosclero-
These rare disorders are characterized by the extracellu- sis in multiple organ systems. The primary organs in-
lar deposition of proteins in one or more organs. They volved include the pancreas, hepatobiliary system, sali-
share clinical, radiographic, and pathological features. vary glands, and lymph nodes. The findings in the chest
The main difference between these two entities is the are variable and may include nodules, lymphadenopathy,
Congo red staining pattern and the fibrillar structure seen airway thickening, and pleural abnormalities. IgG4-related
on electron microscopy of amyloidosis. Both entities may sclerosing disease may account for a proportion of the
be associated with a plasma cell dyscrasia. idiopathic interstitial pneumonias.
Pulmonary Manifestations of Systemic Diseases 117
Acknowledgment. D.A. Lynch, Dept. of Radiology, NJH Remy-Jardin M, Remy J, Cortet B et al (1994) Lung changes in
Denver, CO, US, for his contribution to the preparation of rheumatoid arthritis: CT findings. Radiology 193:375-382.
this manuscript. Tanaka N, Kim JS, Newell JD et al (2004) Rheumatoid arthritis-re-
lated lung diseases: CT findings. Radiology 232:81-91.
Johkoh T, Mller NL, Pickford HA et al (1999) Lymphocytic in- Susanto I, Peters JI (1997) Acute lupus pneumonitis with normal
terstitial pneumonia: thin-section CT findings in 22 patients. chest radiograph. Chest 111:1781.
Radiology 212:567-572. Worthy SA, Muller NL, Hansell DM, Flower CD (1998) Churg
Taouli B, Brauner MW, Mourey I et al (2002) Thin-section chest strauss syndrome: The spectrum of pulmonary CT findings in
CT findings of primary Sjogrens syndrome: correlation with 17 Patients. AJR Am J Roentgenol 170:297-300.
pulmonary function. Eur Radiol 12:1504-1511. Chung MP, Yi CA, Lee HY et al (2010) Imaging of pulmonary vas-
culitis. Radiology 255:322-341.
Mixed connective tissue disease Khan I, Watts RA (2013) Classification of ANCA associated vas-
Bodolay E, Szekanecz Z, Devenyi K et al (2005) Evaluation of in- culitis. Curr Rheumat Rep 15:383.
terstitial lung disease in mixed connective tissue disease Castaner E, Alguersuari A, Andreu M et al (2013) Imaging find-
(MCTD). Rheumatology (Oxford) 44:656-661. ings in pulmonary vasculitis. Seminars in Ultrasound, CT and
Fagan KA, Badesch DB (2002) Pulmonary hypertension associated MRI 33:567-579.
with connective tissue disease. Prog Cardiovasc Dis 45:225-234.
Kozuka T, Johkoh T, Honda O et al (2001) Pulmonary involvement Inflammatory bowel disease
in mixed connective tissue disease: high-resolution CT find- Camus P, Piard F, Ashcroft T et al (1993) The lung in inflammato-
ings in 41 patients. J Thorac Imaging 16:94-98. ry bowel disease. Medicine (Baltimore) 72:151-183.
Garg K, Lynch DA, Newell JD (1993) Inflammatory airways dis-
Unclassifiable connective tissue disease, lung-dominant CTD ease in ulcerative colitis: CT and high-resolution CT features.
Fischer A, Brown KK (2014) Interstitial lung diseases in undiffer- J Thorac Imaging 8:159-163.
entiated forms of connective tissue disease. Arthritis Care and Kelly MG, Frizelle FA, Thornley PT et al (2006) Inflammatory
Research doi: 10.1002/acr.22394 [Epub ahead of print]. bowel disease and the lung: is there a link between surgery and
Fischer A, du Bois R (2012) Interstitial lung diseases in connective bronchiectasis? Int J Colorectal Dis 21:754-757.
tissue disorders. Lancet 380:689.
Amyloidosis and light chain deposition disease
Vasculitis Bhargava P, Rushin JM, Rusnock EJ et al (2007) Pulmonary light
Attali P, Begum R, Romdhane HB et al (1998) Pulmonary Wegen- chain deposition disease: report of five cases and review of the
ers granulomatosios: changes at follow up CT. Eur Radiol literature. Am J Surg Pathol 31:267-276.
8:1009-1113. Urban BA, Fishman EK, Goldman SM et al (1993) CT evaluation
Cordier JF, Valeyre D, Guillevin L et al (1990) Pulmonary Wegen- of amyloidosis: spectrum of disease. RadioGraphics 13:
ers granulomatosis. A clinical and imaging study of 77 cases. 1295-1308.
Chest 97:906-912.
Haworth SJ, Savage COS, Carr D et al (1985) Pulmonary hemor- Erdheim-Chester disease
rhage complicating Wegeners granulomatosis and microscop- Brun A-L, Touitou-Gottenberg D, Haroche J et al (2010) Erdheim-
ic polyarteriitis. BMJ 290:1175. Chester disease: CT findings of thoracic involvement. Eur Ra-
Jennette JC, Falk RJ (2007) Nosology of primary vasculitis. Cur- diol 20:2579-2587.
rent opinion in rheumatology 19:10.
Marten K, Schnyder P, Schirg E et al (2005) Patternbased differ- IgG4-related sclerosing disease
ential diagnosis in pulmonary vasculitis using volumetric CT. Ryu JH, Sekiguchi H, Yi ES (2012) Pulmonary manifestations of
AJR Am J Roentgenol 184:1843. immunoglobulin G4-related sclerosing disease. Eur Respir J
Specks U, Deremee RA (1990) Granulomatous vasculitis. Wegen- 39:180-186.
ers granulomatosis and Churg Strauss syndrome. Rheum Dis
Clin North Am 16: 377-397.
IDKD 2015-2018
levels of cardiac troponin-I (0.09 ng/ml); (2) NSTEMI Table 2. Stanford conceptual classification of acute aortic syn-
(non-ST-segment elevation myocardial infarction): a new dromes
finding of ST-segment depression of 1 mm or T-wave in- 1. Aortic dissection and dissection variants (diseased media)
version of at least 3 mm in at least two anatomically con- 1.a Classic aortic dissectiona
tiguous leads and elevated serial levels of troponin-I 1.b Intramural hematoma variant
1.c Limited intimal tear (limited dissection)a
(0.09 ng/mL); (3) UAP (unstable angina pectoris): clini- 2. Penetrating atherosclerotic ulcer (diseased intima)a
cal symptoms suggestive of ACS, such as typical chest dis- 3. Rupturing thoracic aortic aneurysma
comfort or the equivalent, with an unstable pattern of chest a Classic aortic dissection, limited intimal tear, penetrating athero-
pain (at rest, new onset, or crescendo angina), optimally sclerotic ulcer and rupturing aneurysm are lesions that can occur
with a markedly positive stress test (SPECT, Echo, or with or without associated intramural hematoma.
treadmill testing) and/or an invasive coronary angiogram
demonstrating a 50% epicardial coronary stenosis.
derlying pathology and allows for a broader spectrum of
manifestations, particularly within the dissection category.
Acute Aortic Syndrome: Definition and Classification We also consider symptomatic thoracic aortic aneurysms
as an acute aortic syndrome, and treat IMH as a dissection
Acute aortic syndrome is a contemporary clinical term, variant or as an unspecific imaging finding rather than as
analogous to acute coronary syndrome and referring to a a separate disease entity (Table 2).
spectrum of acute life-threatening abnormalities of the aor-
ta associated with intense chest or back pain [25]. Acute
aortic syndromes are much less common than acute car- ECG-Gated CT: Improved Chest Pain Triage in Patients
diac events, with an estimated annual incidence of approx- with Acute Coronary and Acute Aortic Syndromes
imately 6 per 100,000 for acute aortic dissection [3] (the
most common acute aortic condition requiring emergency Nowadays, fast scanner technology combined with
operative repair) versus more than 400 per 100,000 for heart-rate-reducing medication has made it possible to
acute MI. While the presumptive diagnosis of acute MI is image the coronary arteries without motion artifacts in
usually considered first, excluding acute thoracic aortic most patients. Indeed, state-of-the-art scanners acquire
disease is imperative to expedite appropriate treatment, 64320 cross-sections per rotation, depicting vascular
avoid deleterious thrombolysis, and circumvent unneces- details with a spatial resolution of 0.5 mm. ECG-syn-
sary delays associated with emergency coronary angiogra- chronized, contrast-enhanced images of the heart and
phy. The clinical picture of an acute aortic syndrome can coronary arteries can be acquired in one to five heart cy-
be caused by a wide range of aortic lesions and predispos- cles. The diagnostic performance of coronary CTA has
ing conditions. Traditionally, acute aortic syndromes have been investigated extensively in patients with stable
been ascribed to three main abnormalities [25]: (1) aortic coronary artery disease (CAD). Using invasive angio-
dissection, (2) intramural hematoma (IMH), and (3) pene- graphy as a reference, coronary CTA is more sensitive
trating atherosclerotic ulcer (PAU). Although poorly re- (98100%) than any other noninvasive technique [30].
flecting the underlying pathology and without including Because of its high negative predictive value (99100%),
the full spectrum of acute aortic diseases, these categories coronary CTA is recommended in patients with a low to
are almost universally used in the literature [26-28]. They intermediate probability of CAD and in patients with an
are also at the core of the current classifications of acute inconclusive functional test [31]. A normal cardiac CT
aortic diseases by the European and American Heart As- examination is associated with a low adverse cardiac
sociations (Table 1), with added subcategories for limited event rate in the following years [32] The reported per-
tears and iatrogenic/traumatic dissections [28, 29]. patient specificity (85%) is lower because stenosis
At our institution we use a conceptually different clas- severity is overestimated, often due to the presence of
sification of acute aortic disorders that is based on the un- calcifications, but it is not inferior to that of other non-
invasive techniques. In addition, radiation exposure has
decreased dramatically over the past years, such that dos-
Table 1. ESC Task force, and ACCF/AHA/AATS/ACR/ASA/SCA/
SCAI/SIR/STS/SVM guidelines for the diagnosis and management es of 5 mSv are now common practice using state-of-
of aortic diseases the-art technology; moreover, very recent innovations
permit doses under 1 mSv in selected patients [33]. Giv-
ESC/AHA/ACC/AATS classification
en the practical limitations of functional testing in the
Class 1 Aortic dissection ED setting and the relatively low prevalence of CAD in
Class 2 Intramural hematoma
Class 3 Discrete/Subtle dissection/Limited tear
patients visiting EDs, direct coronary visualization by
Class 4 Penetrating atherosclerotic ulcer CTA offers an attractive diagnostic alternative for the
Class 5 Iatrogenic and traumatic dissection early triage of ACS patients.
ESC, European Society of Cardiology; AHA, American Heart As-
CTA of the thorax and abdomen is a well-established
sociation; ACC, American College of Cardiology; AATS, Ameri- imaging technique for many acute abnormalities of the
can Association for Thoracic Surgery aorta. Even without the use of ECG gating, CTA of the
CT Evaluation of Chest Pain: Acute Coronary Syndrome and Acute Aortic Syndrome 121
aorta has high sensitivity and specificity for the detection mentation of ECG gating for acute coronary CTA pro-
of acute thoracic aortic diseases and it has largely re- grams.
placed conventional diagnostic angiography [34]. How-
ever, over the last decade it has become clear that the Combined Triple-Rule-Out Protocol
spectrum of acute aortic abnormalities also includes sub-
tle yet important aortic lesions that evade nearly all cross- While it is technically possible to build CT protocols that
sectional imaging techniques, including non-gated CTA. allow imaging and interpretation of the entire aorta, the
Conversely, using ECG-gated CTA, the aortic root, coro- coronary arteries, and the pulmonary vasculature, such
nary arteries, and valve apparatus can be accurately as- studies are not performed at our institutions. A complete
sessed preoperatively. Finally, modern endovascular triple-rule-out protocol would require a longer injection
treatment options require much higher degrees of accu- of contrast agent (to opacify systemic and pulmonary ar-
racy in the delineation of pathology and treatment plan- teries simultaneously), resulting in artifacts from the su-
ning: both the exact size and location of a primary inti- perior vena cava and the right heart. To perform a guide-
mal tear and the branching pattern of the aortic arch are line-conforming work-up for aortic dissection, the scan-
affected by cardiac pulsation artifacts, and their suppres- ning range would have to include not only the entire chest
sion is highly desirable. but also the entire abdominal aorta and iliofemoral access
vessels, which implies greater radiation exposure. While
it is possible to detect an aortic dissection using coronary
CT Protocols in Patients with Acute Chest Pain CTA, and acute coronary abnormalities or abnormal myo-
cardial enhancement can be identified on a gated CTA for
The CT techniques used for the assessment of CAD are evaluation of acute aortic syndrome, these situations tend
different from those used in patients with acute aortic dis- to be rare. Some studies have suggested the limited effi-
eases; thus, selection of the correct protocol is the first cacy of these protocols since they are likely to be ordered
step in any CT imaging procedure and is based on clini- in patients with undifferentiated chest pain, who have a
cal judgment by ED physicians. very low event rate for any of the syndromes [35]. Con-
trary to common belief, there is not a wide clinical over-
Coronary CTA lap between ACS and acute aortic syndromes.
Table 3. Randomized controlled trials comparing coronary CTA with the standard of care in the evaluation of acute chest pain
Study CT-STAT ACRIN ROMICAT II
(2011) (2012) (2012)
Population 699 1370 985
TIMI risk score 04 TIMI risk score 02 Lowintermediate risk
MI 0.9% MI 1% MI 2.5%
Randomization 1:1 2:1 1:1
Control group SPECT MPI Usual care Usual care
accurately predict adverse cardiac events over the next 6 incorporation of morphologic plaque assessment and by
months to 2 years. While patients without CAD remain applications to assess the functional relevance of CAD
virtually event-free, those with non-obstructive CAD on CTA. Additionally, high-sensitive-troponin assays
have a slightly increased risk, and those with obstructive may allow the safe exclusion of ACS without the need
CAD are at the highest risk [32]. for further diagnostic testing in a proportion of patients.
Hence, early triage by cardiac CT may re-focus efforts
towards those patients with (conflicting) low elevations
Current Recommendations and Future Expectations in high-sensitive-troponins.
CT Evaluation of Acute Aortic Syndromes simply due to the much faster helical acquisition with
short rotation times and detector bank widths of at least
The sensitivity and specificity of non-gated CT for acute 4 cm.
aortic abnormalities have been reported to approach The following guiding principles may help in the deci-
100%. CTA can reliably assess complications of aortic sion to implement ECG gating for acute aortic syndromes:
diseases and guide treatment decisions. However, there Suppression of pulsation artifacts is highly desirable in
are subtle lesions of the aorta that can be missed with the ascending aorta
non-gated CT [40]. Cardiac pulsation artifacts may also Dynamic (time resolved) visualization is intriguing but
mimic disease, particularly in younger individuals. While may not provide critical information beyond an arti-
ECG gating clearly improves both assessment of the aor- fact-free static dataset
tic root, including the coronary artery origins and the Assuming a prevalence of 5% of limited tears (dissec-
valve apparatus, and visualization of subtle aortic lesions, tion variant) that might be missed without gating and
the superiority of gated versus non-gated CT acquisitions those cases in which pulsation artifacts mimic aortic
in acute aortic disease has yet to be assessed in a prospec- lesions or do not allow their exclusion, a reasonable es-
tive trial. The possibility of reconstructing time-resolved timate for the proportion of patients benefitting from
datasets to display dynamic changes in the position of a gating is probably in the order of 510%
dissection flap or the size of a true or false lumen pro- If ECG-gating can be achieved at no or minimal addi-
vides intriguing insights into the hemodynamic conse- tional dose and without an unreasonably complicated
quences of a given pathology. The best CT protocol for workflow, it is the opinion of the authors that gating
the evaluation of acute aortic syndromes depends on sev- should be implemented
eral factors, including the scanner generation, patient If ECG-gating is not routinely used, e.g., in a setting
population, and expected prevalence of acute aortic dis- with a low prevalence of patients with acute aortic dis-
eases, the comfort and training level of night- and week- ease, a dedicated gated protocol should still be avail-
end technologists, as well as the expertise of readers. able for problem solving (e.g., in the 510% of cases
ECG-gating technologies include retrospective gating, in which it is deemed necessary) following an incon-
which allows 4D reconstructions but comes with a slight clusive non-gated study.
radiation-dose penalty. Prospective triggering involves a
lower dose of radiation but requires larger detectors if the
entire thoracic aorta is to be covered. High-pitch helical Aortic Dissection and Its Variants
scans (so-called flash mode) are an intriguing alternative,
but require high-power X-ray tubes or small patient size. Our conceptual approach to acute aortic syndromes
It is also worth mentioning that improved scanner tech- groups all the manifestations of a diseased aortic media
nology reduces pulsation artifacts even without gating, into the dissection group. The common pathologic
126 D. Fleischmann, U. Hoffmann
denominator of these diseases is the presence of an ab- penetrating ulcer, or even trauma blood can extend in-
normal aortic medial layer, traditionally (albeit not com- to the neighboring aortic wall and result in IMH. These
pletely accurately) described as cystic medial necrosis. local hematomas typically do not extend far along the
Cystic medial necrosis is the common pathologic end- aorta or its branches when associated with a PAU, but are
point of several underlying acquired and inherited condi- typically local.
tions, ranging from severe hypertension and normal ag- While IMH can occur in underlying cystic medial
ing to familial aortic diseases, vasculitis, and connective necrosis as well as in PAUs, this does not make these le-
tissue diseases such as Ehlers-Danlos, Loeys-Dietz, or sions the same entity, and the notion that PAUs can be-
Marfan disease. come dissections and vice versa does not reflect our ex-
Classic dissection (class I) is characterized by a pri- perience. The one patient group in which overlapping and
mary intimal tear (in which blood enters the false lumen), indistinguishable features of PAU, IMH, and elements of
a false channel, or a lumen within the aortic media that dissection may occur in the same individual are typically
is separated from the true lumen by a dissection flap older individuals with moderate atherosclerotic burden
made of intima and a considerable portion of the media. and presumed coexisting cystic medial necrosis, which is
A common variant of aortic dissection manifests as an a known manifestation of normal aging as well. In this
IMH (class II). The abnormal space within the aortic me- particular situation the overlapping imaging features may
dia enabled by the underlying cystic medial necrosis is reflect coexisting intimal (atherosclerotic) and medial
filled by thrombus rather than by flowing blood. It is im- (cystic degeneration) disease in the same patient.
portant to realize that elements of classic dissection and
its IMH variant may coexist in the same individual. For Penetrating Atherosclerotic Ulcer (Class IV)
example, it is not uncommon to find areas of intramural
blood in patients who otherwise have features of classic The pathologic definition of a PAU is a deep ulcerated
dissection. At the same time, in most patients with IMH plaque that penetrates the internal elastic lamina into the
it is now recognized that small primary intimal tears as medial layer of the aorta [42]. PAU is therefore a mani-
well as other small communications between the true lu- festation of atherosclerosis (and not of cystic medial dis-
men flow channel of the aorta and portions of the throm- ease). The risk profile of patients with PAU is different
bosed false channel do exist [41]. from that of patients with aortic dissection. Patients are
A so-called limited tear of the aorta, or limited dissec- typically older, with several other atherosclerotic comor-
tion (class III), is a rare lesion characterized by a partial- bidities, including a history of stroke, peripheral vascular
thickness tear of the aortic wall, extending from the inti- disease, CAD, and aortic aneurysms. CT images fre-
ma into part or all of the media [40]. This lesion could be quently show extensive atherosclerotic changes, often
regarded as a very large primary intimal tear but without with more than one ulcer-like lesion.
development of a separate flow channel in the media Given the multiplicity of lesions and the common ob-
(false lumen) as in classic dissection. Limited tears occur servation of atherosclerotic ulcers in non-symptomatic pa-
more often in the ascending aorta than more distally and tients, it is important to try to distinguish between non-
are associated with dilated or aneurysmal aortic diame- penetrating ulcers (i.e. ulcerated plaque confined to the
ters. The underlying pathology, typically with cystic me- thickened intima), chronic healed penetrating ulcers
dial degeneration, and risk profiles of patients with lim- (which are re-endothelialized and not an acute threat), and
ited tears are similar to those with classic dissection. those lesions that acutely penetrate the aortic wall with a
high risk of complications such as perforation and rupture.
Intramural Hematoma (Class II) The key distinguishing features of acute lesions are intra-
mural blood and periaortic stranding, which can help iden-
While the term intramural hematoma simply means tify a culprit lesion. Acute PAUs without IMH do exist,
blood in the aortic wall, it has been used misleadingly though, and in the setting of acute aortic pain a CT scan
(class II lesion) to describe an entity on the same level as without IMH or periaortic stranding near an identifiable
an aortic dissection or penetrating atherosclerotic ulcer, ulcer-like lesion does not exclude an acute aortic condi-
which has resulted in considerable confusion and incon- tion. In these cases, follow up imaging is recommended.
sistent classification of aortic diseases in the literature.
We do not consider IMH as a specific disease entity, but Iatrogenic and Traumatic Dissection (Class V)
instead as an important imaging finding, because it is un-
specific to the underlying pathology or etiology. Intra- Aortic and other arterial dissections can occur as a com-
mural blood can be seen to variable degrees in any acute plication of diagnostic and interventional catheter-based
aortic condition: if fresh clot fills the entirety of the dis- procedures, as well as a complication of cardiac or vas-
section plane in patients with cystic medial necrosis, it is cular surgery. While patients with iatrogenic dissections
best considered a variant manifestation of aortic dissec- may have underlying vascular diseases that predispose
tion. Intramural blood is also typically associated with them to aortic dissection, even the non-diseased aortic
PAUs: any communication between the aortic lumen and wall can be mechanically dissected by subintimal ad-
the media be it through a primary intimal tear, a focal vancement of endovascular wires and catheters.
CT Evaluation of Chest Pain: Acute Coronary Syndrome and Acute Aortic Syndrome 127
18. Tosteson AN, Goldman L, Udvarhelyi IS, Lee TH (1996) Cost- agnosis and Treatment of Aortic Diseases of the European So-
effectiveness of a coronary care unit versus an intermediate ciety of Cardiology (ESC). Eur Heart J 35:2873-2926.
care unit for emergency department patients with chest pain. 30. Yang L, Zhou T, Zhang R et al (2014) Meta-analysis: diagnos-
Circulation 94:143-150. tic accuracy of coronary CT angiography with prospective
19. Conti A, Paladini B, Magazzini S et al (2002) Chest pain unit ECG gating based on step-and-shoot, flash and volume modes
management of patients at low and not low-risk for coronary for detection of coronary artery disease. European Radiology
artery disease in the emergency department. A 5-year experi- 24:2345-2352.
ence in the Florence area. Eur J Emerg Med 9:31-36. 31. Task Force Members, Montalescot G, Sechtem U, Achenbach
20. Solinas L, Raucci R, Terrazzino S et al (2003) Prevalence, S et al (2013) 2013 ESC guidelines on the management of sta-
clinical characteristics, resource utilization and outcome of pa- ble coronary artery disease: The task force on the management
tients with acute chest pain in the emergency department. A of stable coronary artery disease of the European Society of
multicenter, prospective, observational study in north-eastern Cardiology. Eur Heart J 34:2949-3003.
Italy. Ital Heart J 4:318-324. 32. Hulten EA, Carbonaro S, Petrillo SP et al (2011) Prognostic
21. Roos JE, Willmann JK, Weishaupt D et al (2002) Thoracic aor- value of cardiac computed tomography angiography: a sys-
ta: motion artifact reduction with retrospective and prospective tematic review and meta-analysis. J Am Coll Cardiol 57:1237-
electrocardiography-assisted multi-detector row CT. Radiolo- 1247.
gy 222:271-277. 33. Fuchs TA, Stehli J, Bull S et al (2014) Coronary computed to-
22. Braunwald E, Antman EM, Beasley JW et al (2002) mography angiography with model-based iterative reconstruc-
ACC/AHA guideline update for the management of patients tion using a radiation exposure similar to chest x-ray exami-
with unstable angina and non-st-segment elevation myocar- nation. Eur Heart J 35:1131-1136.
dial infarction-2002: Summary article: a report of the Ameri- 34. Shiga T, Wajima Z, Apfel CC et al (2006) Diagnostic accura-
can College of Cardiology/American Heart Association task cy of transesophageal echocardiography, helical computed to-
force on practice guidelines (committee on the management mography, and magnetic resonance imaging for suspected tho-
of patients with unstable angina). Circulation 106:1893-1900. racic aortic dissection: systematic review and meta-analysis.
23. Luepker RV, Apple FS, Christenson RH et al (2003) Case de- Arch Intern Med 166:1350-1356.
finitions for acute coronary heart disease in epidemiology and 35. Rogers IS, Banerji D, Siegel EL et al (2011) Usefulness of
clinical research studies: a statement from the AHA Council comprehensive cardiothoracic computed tomography in the
on Epidemiology and Prevention; AHA Statistics Committee; evaluation of acute undifferentiated chest discomfort in the
World Heart Federation Council on Epidemiology and Preven- emergency department (capture). Am J Cardiol 107:643-650.
tion; The European Society of Cardiology Working Group on 36. Raff GL, Chinnaiyan KM, Cury RC et al (2014) SCCT guide-
Epidemiology and Prevention; Centers for Disease Control lines on the use of coronary computed tomographic angiogra-
and Prevention; and the National Heart, Lung, and Blood in- phy for patients presenting with acute chest pain to the emer-
stitute. Circulation 108:2543-2549. gency department: a report of the society of cardiovascular
24. Ornato JP (2003) Management of patients with unstable angi- computed tomography guidelines committee. J Cardiovasc
na and non-st-segment elevation myocardial infarction: update Comput Tomog 8:254-271.
ACC/AHA guidelines. Am J Emerg Med 21:346-351. 37. Hoffmann U, Truong QA, Schoenfeld DA et al; Investigators
25. Vilacosta I, Roman JAS (2001) Acute aortic syndrome. Heart R-I (2012) Coronary CT angiography versus standard evalua-
85:365-368. tion in acute chest pain. N Engl J Med 367:299-308.
26. Erbel R, Alfonso F, Boileau C et al (2001) Diagnosis and man- 38. Litt HI, Gatsonis C, Snyder B et al (2012) CT angiography for
agement of aortic dissection. Eur Heart J 22:1642-1681. safe discharge of patients with possible acute coronary syn-
27. Evangelista A, Mukherjee D, Mehta RH et al (2005) Acute in- dromes. N Engl J Med 366:1393-1403.
tramural hematoma of the aorta: a mystery in evolution. Cir- 39. Goldstein JA, Chinnaiyan KM, Abidov A et al; Investigators
culation 111:1063-1070. C-S (2011) The CT-STAT (coronary computed tomographic
28. Hiratzka LF, Bakris GL, Beckman JA et al (2010) 2010 AC- angiography for systematic triage of acute chest pain patients
CF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM to treatment) trial. J Am Coll Cardiol 58:1414-1422.
guidelines for the diagnosis and management of patients with 40. Svensson LG, Labib SB, Eisenhauer AC, Butterly JR (1999)
thoracic aortic disease: a report of the American College of Intimal tear without hematoma: an important variant of aortic
Cardiology Foundation/American Heart Association task dissection that can elude current imaging techniques. Circula-
force on practice guidelines, American Association for Tho- tion 99:1331-1336.
racic Surgery, American College of Radiology, American 41. Williams DM, Cronin P, Dasika N et al (2006) Aortic branch
Stroke Association, Society of Cardiovascular Anesthesiolo- artery pseudoaneurysms accompanying aortic dissection. Part
gists, Society for Cardiovascular Angiography and Interven- I. Pseudoaneurysm anatomy. J Vasc Interven Radiol JVIR
tions, Society of Interventional Radiology, Society of Tho- 17:765-771.
racic Surgeons, and Society for Vascular Medicine. Circula- 42 Nathan DP, Boonn W, Lai E et al (2012) Presentation, compli-
tion 121:e266-369. cations, and natural history of penetrating atherosclerotic ulcer
29. Erbel RV, Aboyans V, Boileau C et al (2014) 2014 ESC Guide- disease. J Vasc Surg 55:10-15.
lines on the diagnosis and treatment of aortic diseases: Docu- 43. Johansson G, Markstrom U, Swedenborg J (1995) Ruptured
ment covering acute and chronic aortic diseases of the thoracic thoracic aortic aneurysms: a study of incidence and mortality
and abdominal aorta of the adult. The Task Force for the Di- rates. J Vasc Surg 21:985-988.
IDKD 2015-2018
Introduction
With the ever-increasing use of computed tomography
(CT) for imaging, there is always the possibility of find-
ing something on the exam that was not anticipated and is
not related to the primary indication for obtaining the ex-
am. These unanticipated findings have been labeled inci-
dentalomas and, depending on their significance, a cas-
cade of additional imaging exams can result. Care must be
taken to limit unnecessary downstream imaging due to
these unanticipated findings. In a patient with an inciden-
tal finding, the first task of the radiologist is to determine
its stability. An active search of the patients prior imaging
exams in the picture archiving and communication sys-
tems is critical. Appropriate comparison exams may in- Fig. 1. Axial image shows a right axillary lymph node with a fatty
hilum (arrow) indicating a benign lymph node
clude prior abdominal, renal, or virtual colonoscopy CT
exams, magnetic resonance (MR) or ultrasound exams of
the liver or biliary tree, or a thoracic or cardiac CT in intravenous contrast. As with other incidental lymph
which the organs of the upper abdomen were incidentally nodes in the mediastinum and hilum, assessment for clin-
included. For lesions of the thoracic inlet, a prior neck or ical significance should include a short-axis measure-
cervical spine CT may be helpful. If the patient is new to ment as well as evaluation of the lymph node density. In
your institution, comparison exams from other institutions general a short-axis measurement of 1.0 cm suggests a
can often be electronically transferred and thus obtained benign lymph node [1]. However, a lymph node may be
with relatively little effort, and are of great value. Obtain- enlarged secondary to reaction to local infection or to a
ing prior exams can save the patient the significant cost, neoplastic process. The density of the lymph node is also
radiation exposure, and anxiety associated with repeat ex- important since lymph nodes containing a fatty hilum
ams. If the incidentaloma is new or there are no com- (Fig. 1) are considered benign regardless of size. Diffuse,
parisons to determine stability, a decision must be made central, or peripheral eggshell calcifications also indicate
regarding the overall significance of the finding, which a benign etiology. However, eccentric calcification
often has yet to be fully assessed. We have included some should not be considered a benign feature.
of the incidental findings that may be encountered on rou- The thyroid gland is usually at least partially included
tine thoracic and cardiac CT imaging, with a focus on on a chest CT exam. Thyroid nodules are extremely com-
their management. mon in the adult population and the vast majority are be-
nign. Recent guidelines have been published by Hoang et
al. [2] to diminish the number of unnecessary follow-up
Thoracic Inlet exams. (Fig. 2). For adults 35 years of age, a follow-up
ultrasound is recommended if the nodule is 1.5 cm. In
Coronal reconstructed images can be helpful in detecting adults 35 years of age, a follow-up ultrasound is rec-
supraclavicular lymph nodes, which may be more diffi- ommended if the nodule is 1 cm. In addition, any nod-
cult to differentiate from adjacent vascular structures on ule in the pediatric age group or any nodule regardless of
axial images, especially if the exam is performed without patient age that is fluorodeoxyglucose-avid on positron
Chest Wall
Coronal and sagittal reformatted images are useful in Fig. 3. Sagittal image demonstrating a compression fracture of the
evaluating skeletal structures. The ribs are problematic T9 vertebral body; the fracture was not present on the prior exam
because a single rib is not visualized on a single axial
image. The use of bone windows improves visualization
of the lateral aspects of the ribs on sagittal images; pos- Upper Abdomen
terior aspects are best visualized on coronal reformatted
images. Rib lesions will be more obvious while Liver
scrolling through the data set than if each image is in-
dependently reviewed. Sternal fractures can be missed Although commonly encountered, most liver lesions are
on axial images because of the plane of imaging but are benign, even in patients with known malignancy. Berland
very well depicted on sagittal views. Likewise, the ver- et al. [3] published guidelines for the management of in-
tebral column is ideally suited for review on sagittal cidental liver findings on CT (Fig. 4). They categorized
views, where compression fractures (Fig. 3) and lytic liver lesions based on the size and appearance of the le-
and sclerotic lesions may be more obvious than on axi- sion and the relative risk of the patient. Hepatic cysts are
al images. very common and present as sharply marginated, low-
CT is not ideal for breast imaging. However, assess- attenuation lesions of 20 Hounsfield units (HU). Typical
ment of breast tissue should be included in the search hemangiomas show peripheral, nodular enhancement that
pattern. Breast cancer may be visible as an asymmetric fills in centrally on delayed images (Fig. 5). Findings that
soft-tissue mass that sometimes shows enhancement with are more concerning include low-attenuation lesions with
the administration of intravenous contrast. Skin thicken-
ing from tumor or radiation can also be visualized, as
well as evidence of prior surgery such as lymph node dis-
section. The presence of collateral vessels in the chest 0.5 cm in low or Any size lesion with
wall may alert the clinician to the presence of vascular average risk patient benign imaging features
pathology elsewhere, such as a stenosis or occlusion at
the thoracic inlet.
Subcutaneous nodules such as metastases occur more No follow-up required
commonly with lung cancer, breast cancer, or melanoma.
However, soft-tissue nodules of the chest wall are more
typically related to benign etiologies such as sebaceous
cysts, or, when involving the anterior upper abdominal Robust enhancement in low- or
wall, injection sites from subcutaneous administration of average-risk patient
medication. Occasionally, abnormalities such as calcifi- Fig. 4. Liver lesions do not always require follow-up imaging;
cations in patients with scleroderma or neurofibromas in rather, the decision depends on the imaging features and size of the
patients with neurofibromatosis will be present. lesion and the risk to the patient
Incidental Findings on Thoracic and Cardiac CT 131
10 HU?
Yes No
Benign
14 cm 4 cm
Conclusion
In the routine practice of interpreting thoracic and cardiac
CT exams, there will be frequent incidental findings that
are not related to the patients indication for the exam and
may not be symptomatic [9]. The majority of these inci-
dental lesions will be benign in etiology and not require
additional attention. However, in patients with incide-
nalomas, it is key that we, as radiologists, make efforts
to compare current and prior examinations to determine
Fig. 9. Axial image shows a pure ground glass opacity nodule the chronicity of the abnormality, in order to avoid un-
(arrow) in the right lower lobe
necessary additional work-up. Conversely, when a new
finding is significant, it is important that we recommend
appropriate follow-up management, including additional
evaluation to ensure that patients are adequately treated imaging or specialty consultation. Appropriate manage-
[8]. The follow-up of solid nodules includes determination ment is critical for patient safety as well as for the over-
of their average of the length and width and of the pres- all resource utilization of the medical system.
ence of known risk factors for lung carcinoma. It should
be noted that these guidelines do not apply to patients
who are suspected of having, or who have, a known ma- References
lignancy, patients 35 years of age, and patients with un- 1. Onuma Y, Tanabe K, Nakazawa G et al (2006) Noncardiac
explained fever. The guidelines for the follow-up of sub- findings in cardiac imaging with multidetector computed to-
solid (ground glass or part solid/part ground glass) nod- mography. J Am Coll Cardiol 48:402-406.
ules (Fig. 9) are based on the size of the lesion [8]. Only 2. Hoang JK, Langer JE, Middleton WD et al (2014) Managing
subsolid nodules 5 mm in diameter require follow-up. incidental thyroid nodules detected on imaging: white paper of
the ACR incidental thyroid findings committee. J Am Coll Ra-
Solid nodules are followed for a period of 2 years, at diol pii: S1546-1440 (14) 00627-9 [E_pub a head to print].
which time a lack of change is considered to be evidence 3. Berland LL, Silverman SG, Gore RM et al (2010) Managing
of a benign process; subsolid nodules require at least 3 incidental findings on abdominal CT: white paper of the ACR
years of surveillance, as the malignancies that demon- incidental findings committee. J Am Coll Radiol 7:754-773.
strate this pattern are typically slow-growing. When a 4. Elefteriades JA, Farkas EA (2010) Thoracic aortic aneurysm:
clinically pertinent controversies and uncertainties. J Am Coll
lung nodule is identified, it should also be assessed on Cardiol 55:841-857.
sagittal or coronal reformatted images. Nodules that ap- 5. Hiratzka LF, Bakris GL, Beckman JA et al (2010) ACCF/
pear discoid are likely benign, regardless of their con- AHA/AATS/ACR/ASA/SCA/SCAI/SIR/STS/SVM Guide-
spicuity on axial images. Likewise, triangular or polygo- lines for the diagnosis and management of patients with tho-
racic aortic disease. J Am Coll Cardiol 55:e27-e129.
nal nodules that are intimately associated with the lung 6. MacMahon H, Austin JH, Gamsu G et al (2005) Guidelines for
fissures most likely represent intrapulmonary lymph nodes. management of small pulmonary nodules detected on CT
Emphysema is usually related to smoking and is commonly scans: a statement from the Fleischner Society. Radiol
found in the same patient population undergoing CT exams 237:395-400.
for smoking-related pathologies. In emphysema, there is 7. Naidich DP, Bankier AA, MacMahon H et al (2013) Recom-
mendations for the management of subsolid pulmonary nod-
destruction of lung parenchymal tissue resulting in focal ules detected at CT: a statement from the Fleischner Society.
lucencies. It is important to distinguish emphysema from Radiol 266:304-317.
cystic lung disease, which is often managed differently. 8. Truong MT, Ko JP, Rossi SE et al (2014) Update in the evalu-
The key finding to distinguish emphysema from cystic ation of the solitary pulmonary nodule. RadioGraph 34:1658-
lung disease is the ability to identify a discernible wall 1679.
9. Teague SD, Rissing S, Mahenthiran J, Achenbach S (2012)
around the focal lucency in the lung parenchyma, which Learning to interpret the extracardiac findings on coronary CT
indicates the latter (i.e., pulmonary Langerhans cell his- angiography examinations. J Cardiovasc Comput Tomogr
tiocytosis or lymphangioleiomyomatosis) rather than 6:232-245.
IDKD 2015-2018
k-space acquisition. SSFP images provide very rapid MRI is the method of choice for longitudinal follow-
imaging, with blood having very high signal intensity up in patients who are undergoing therapeutic interven-
compared to the ventricular wall. Because there is a large tions. For example, in cardiac stem cell trials MRI was
amount of signal available, the SSFP technique is often the test of choice to assess changes in LV size, function,
combined with parallel imaging, which improves imag- and mass [2, 3]. From a research perspective, the sample
ing speed by two- to three-fold. A single cine slice of the size needed to detect LV parameter changes in a clinical
heart with a 40-ms temporal resolution is acquired in trial is far smaller, in the range of one order of magnitude,
about 6 s on a modern MR scanner. when MRI rather than 2D echocardiography is used, thus
markedly reducing the time and cost of patient care and
pharmaceutical trials [4, 5].
T1 and T2 Mapping A unique MR technique, myocardial tagging (spatial
modulation of magnetization, SPAMM technique), has
The detection of myocardial fibrosis using contrast-en- been developed, in which heart muscle is labeled with a
hanced cardiac MR depends on differences in signal in- dark grid and cardiac rotation, strain, and the displace-
tensity between scarred regions and adjacent normal my- ment and deformation of different myocardial layers dur-
ocardium. However, in diffuse myocardial fibrosis these ing the cardiac cycle are analyzed in 3D (Fig. 1).
differences in signal intensity are lacking. Measurement SPAMM technique helps in assessing regional myocar-
of myocardial T1 times (T1 mapping) using non-contrast dial wall motion but it requires sophisticated software to
or gadolinium-enhanced inversion recovery prepared se- extract local function parameters such as 2D or 3D myo-
quences has the potential to identify diffuse myocardial cardial strains [6, 7].
fibrosis that correlates well with ex vivo fibrosis content.
T1 mapping calculates myocardial T1 relaxation times
using image-based signal intensities and can be per- Assessment of Cardiomyopathies
formed using standard cardiac MR scanners and radiolo-
gy workstations. Current studies show that a myocardium In the visualization of left and right ventricular morphol-
with diffuse fibrosis will have greater retention of con- ogy and function, MRI is a noninvasive tool that has a
trast and thus shorter T1 times than normal myocardium. high degree of accuracy and reproducibility. It is also su-
Amyloidosis is a deposition disease in which the T1 time perior to echocardiography in the determination of ven-
of the myocardium is likewise altered. These observations tricular mass and volumes [8] and has become the gold
suggest the utility of T1 mapping in the detection of dif- standard for the in vivo identification of the phenotypes
fuse myocardial fibrosis and amyloidosis. of cardiomyopathies [9].
T2 and T2* maps of the myocardium may also be cal-
culated. T2* maps have clinical utility in the assessment Dilated Cardiomyopathy
of iron content of the myocardium (e.g., hemochromato-
sis). T2 mapping has been used to evaluate diseases that In dilated cardiomyopathy, MRI is useful to study ventric-
have associated myocardial edema, such as myocarditis ular morphology and function (Fig. 2), by analyzing wall
and acute myocardial infarction. thickening, impaired fiber shortening, and end-systolic
The different clinical and emerging applications of wall stress, which is a very sensitive parameter of a change
cardiac MRI are described below. in LV systolic function. It can also accurately assess the
morphology and function of the right ventricle (RV),
which is frequently affected in dilated cardiomyopathy.
Assessment of Ventricular Function Late contrast-enhanced T1-weighted images are helpful in
detecting the changes that occur in acute myocarditis. In
MRI is a very accurate and highly reproducible technique this setting, increased gadolinium accumulation is thought
for measuring ejection fraction and ventricular volumes to be due to inflammatory/hyperemia-related increased
noninvasively in three dimensions. For this reason, it has flow, slow wash-in/wash-out kinetics, and diffusion into
become the gold standard to which other modalities are necrotic cells [10]. There is evidence of similar changes in
compared [1]. Simpsons rule is applied to determine chronic dilated cardiomyopathy [11]. Contrast-enhanced
ejection fraction and volumes. In assessments of volumes MRI may also increase the sensitivity of endomyocardial
and mass, bright-blood SSFP sequences are typically biopsy by revealing inflamed areas, which aids in deter-
used, with 3040 phases per cardiac cycle. Breath-hold- mining the appropriate biopsy site. Moreover, the presence
ing techniques with acquisition times of 612 s reduce and extent of myocardial late enhancement has been
blurring of the endocardial border. Generally, for accurate shown to predict clinical outcome [12, 13].
measurement of volume and mass, entire coverage of the
left ventricle (LV) with short-axis views from the mitral Hypertrophic Cardiomyopathy
plane are recommended. Slice thickness is typically 810
mm, but in the evaluation of subtle changes the thickness Due to its high accuracy, MRI is increasingly used to as-
should be reduced appropriately. sess morphology, function, tissue status, and degree of
136 D. Revel, D.A. Bluemke
a b
Fig. 1 a, b. Cardiac magnetic resonance (MR) tagging. Magnetic strips are placed in the heart before contraction (diastole, a) and then
followed throughout the cardiac cycle (systole, b) to measure the regional contraction of different portions of the heart. Note that in the left
image the tag lines are curved, representing deformation of the myocardium. Cardiac MR is considered the reference standard for the
measurement of regional function of the heart compared to other imaging modalities
LV outflow tract (LVOT) obstruction in patients with hy- bulent jet during systolic LVOT obstruction is easily de-
pertrophic cardiomyopathy (HCM) (Fig. 3). In addition, tected using suitable echo times (about 4 ms). MRI also
it is very accurate in assessing a LV mass, regional hy- detects the systolic anterior motion of the mitral valve, in
pertrophy patterns, and the different phenotypes of HCM a four-chamber or a short-axis view on the valvular plane
(e.g., apical HCM) [7]. Post-surgical changes after myo- [15], and can be used to document quantify mitral regur-
mectomy can also be reliably monitored [14]. The tur- gitation. A newer technique is to measure the effective
LVOT area by MR planimetry during systole, which over-
comes the problem of interstudy variability of the LVOT
gradient due to its independence from the hemodynamic
status. There are preliminary data showing that assess-
ment of diastolic function using MRI may be superior to
the determination of conventional parameters by echocar-
diography. Analysis of the early untwisting motion of the
myocardium could be helpful in assessing diastolic func-
tion [16]. Other functional changes that make use of myo-
cardial tagging include a reduction in posterior rotation,
reduced radial displacement of the inferoseptal my-
ocardium, reduced 3-D myocardial shortening, and het-
erogeneity of regional function. With late enhancement
imaging after the injection of gadolinium chelates (Fig.
4), MRI may help to detect areas of fibrosis, the prog-
nostic value of which has been demonstrated [17, 18].
MRI also easily detects the acute and chronic changes af-
ter septal artery ablation [14, 19].
cavities and walls, with excellent depiction of the my- Restrictive Cardiomyopathy
ocardial anatomy. Although T1-weighted spin-echo im-
ages may reveal fatty infiltration, thinned walls, and dys- Primary infiltration of the myocardium by fibrosis or
plastic trabecular structures, these particular findings are other tissues leads to the development of restrictive car-
actually quite infrequent. Moreover, the subjective as- diomyopathy, which is characterized by normal LV size
sessment of RV wall thinning, wall motion abnormalities, and systolic function, severe diastolic dysfunction, and
and fatty infiltration of the myocardium by cardiac MR biatrial enlargement. Restrictive cardiomyopathy needs
may be problematic. Thus, the international Task Force to be differentiated from constrictive pericarditis, which
proposed revised criteria for the clinical diagnosis of is a primary disease of the pericardium rather than the
ARVC/D. Major and minor criteria defined by cardiac myocardium. LV size and thickness are quantified us-
MR contribute to the final diagnosis. Regional RV dys- ing gradient-echo sequences. Atrial enlargement is as-
function associated with an increase of RV end-diastolic sessed in a four-chamber view. Mitral regurgitation
volume (110 mL/m2 for males and 100 mL/m2 for fe- should be assessed as well. The restrictive diseases that
males) and RV ejection fraction 40% are considered to can be effectively assessed using MRI are described be-
be major criteria for the diagnosis of ARVC/D [22]. low [23, 24].
138 D. Revel, D.A. Bluemke
a b
Fig. 5 a, b. Axial black-blood image of the heart. The T1 image (a) shows replacement of the myocardium by fat signal (arrows) in the left
and right ventricles. The fat suppression T1-weighted image (b) shows dark signal at the corresponding sites of fat deposition
Cardiac Magnetic Resonance 139
a b
Fig. 6 a, b. Black-blood image of the heart in long (a) and short (b) axis reveals a thickened pericardium (arrows) and fluid (b, arrows).
Pericardial effusion and thickening 4 mm is abnormal
140 D. Revel, D.A. Bluemke
Myocarditis
In patients in whom myocarditis is suspected, cardiac MR
has become the primary tool for the noninvasive diagno-
sis of myocardial inflammation. Myocarditis is defined as
the inflammation of myocardial tissue and is an important
underlying etiology of other myocardial diseases, such as
dilated cardiomyopathy. The definitive diagnosis is fre-
quently confirmed based on the clinical history, the clin-
ical assessment, and non-invasive test results, in which
case cardiac MR is nowadays crucial. Pericardial effusion
has been reported in 3257% of patients with myocardi-
tis although it is not specific for the diagnosis. T2- Fig. 7. Black-blood image of the heart in a long-axis view shows a
mass (arrow) in the right atrium that was subsequently diagnosed
weighted or T2 mapping images sensitively detect tissue as lipoma
edema because of the long T2 values of edematous tissue.
However, the definitive diagnosis is usually based on the
observation on late enhanced images of focal signal in-
crease typically localized to the subepicardial regions of MRI is generally reserved for use when other modalities
the LV and possibly extension through the ventricular fail or provide suboptimal information. Double inversion
wall. Late enhancement may be multifocal or diffuse in recovery sequences can show valve morphology as well
distribution but with no coronary artery disease distribu- as evidence of associated secondary changes (chamber
tion. Consensual diagnostic criteria using cardiac MR dilatation, myocardial hypertrophy, post-stenotic changes
have been proposed (Lake Louise Consensus Criteria) in the great vessels, or thrombus in any of the chambers).
for the diagnosis of acute myocarditis [33]. Semi-quantitative assessment of valvular stenosis or re-
gurgitation can be obtained by measuring the area of sig-
nal void on gradient-echo images. The duration or extent
Evaluation of Cardiac and Paracardiac Masses of the signal void correlates with the severity of the aor-
tic stenosis, and the total area of signal loss with the
Primary cardiac tumors are rare (0.0020.3% incidence) severity of mitral regurgitation. This technique has a very
and the majority (75%) are benign. Metastatic tumors are high sensitivity (98%), specificity (95%), and accuracy
20- to 40-fold more common than primary tumors. MRI (97%) for diagnosing aortic and mitral regurgitation. The
is ideal for delineating the morphologic details of a mass signal void, however, is dependent upon certain scan pa-
(including extent, origin, hemorrhage, vascularity, calci- rameters, such as echo time, voxel size, and image orien-
fication, and effects on adjacent structures). Protocols in- tation relative to the flow jet. Phase-contrast MRI can be
clude the combined use of axial black-blood sequences used to assess the severity of valvular stenosis (by mea-
and axial bright-blood cine images. Functional MRI is suring the peak jet velocity) by calculating the valve ori-
useful to study the pathophysiologic consequences of a fice area and the transvalvular pressure gradient.
cardiac mass. Specifically, benign myxomas (the most
common cardiac tumor) appear brighter than myocardi-
um on T2 weighting; cine images may reveal the charac- Evaluation of Myocardial Perfusion and Ischemia
teristic mobility of the pedunculated tumor. Lipomas
(Fig. 7) appear brighter on spin-echo T1-weighted im- MRI has been validated as a reliable and useful tool in
ages. The diagnosis is verified by a decrease in signal in- the assessment of regional left ventricular perfusion [35].
tensity using a fat suppression technique [34]. In this setting, MRI relies upon monitoring the first pass
of a contrast agent. After a rapid intravenous contrast in-
jection, there is marked signal enhancement first in the
Evaluation of Valvular Diseases right ventricular cavity, then in the left ventricular cavity,
and finally in the left ventricular myocardium. This is
Echocardiography with color Doppler is usually the first- completed within 2030 s. Peak signal intensity is relat-
line imaging modality for diagnosing valvular diseases. ed to the concentration of the contrast agent in the local
Cardiac Magnetic Resonance 141
tissue and is directly proportional to coronary blood flow. and the wide variety of tools that the imaging physician
A comparison between perfusion MRI at rest and after must thoroughly understand in order to appropriately de-
the infusion of pharmacologic agents (adenosine, re- ploy them.
gadenoson, and persantine) and standard methods (an-
giography or radionuclide scintigraphy) has shown that
the sensitivity of MRI is equal or superior to stress myo- References
cardial radionuclide scintigraphy, and the specificity is
similar [36, 37]. Assessing wall motion further improves 1. Hundley WG, Bluemke DA, Finn JP et al (2010) ACCF/
ACR/AHA/NASCI/SCMR 2010 expert consensus docu-
the performance of MRI [38]. ment on cardiovascular magnetic resonance: a report of the
American College of Cardiology Foundation Task Force on
Expert Consensus Documents. J Am Coll Cardiol 55:2614-
Myocardial Viability 2662.
2. Makkar RR, Smith RR, Cheng K et al (2012) Intracoronary
cardiosphere-derived cells for heart regeneration after myocar-
Viability has been defined as dysfunctional myocardium dial infarction (CADUCEUS): a prospective, randomised
sustained by coronary artery disease with limited or ab- phase 1 trial. Lancet 379:895-904.
sent scarring and the potential of functional recovery af- 3. Rokosh G, Ghafghazi S, Bolli R (2013) Mechanisms of stem
ter revascularization. Cardiac MRI may help in predict- cell effects: insights from MRI. Minerva Cardioangiol 61:605-
616.
ing potential recovery based on the degree of high en- 4. Grothues F, Smith GC, Moon JC et al (2002) Comparison of
hancement through the myocardium [39]. Alternatively, interstudy reproducibility of cardiovascular magnetic reso-
as in dobutamine echocardiography, inotropic stimulation nance with two-dimensional echocardiography in normal sub-
of LV contraction at a low dose of dobutamine is a pre- jects and in patients with heart failure or left ventricular
dictor of recovery [39, 40]. hypertrophy. Am J Cardiol 90:29-34.
5. Bellenger NG, Burgess MI, Ray SG et al (2000) Comparison
of left ventricular ejection fraction and volumes in heart fail-
ure by echocardiography, radionuclide ventriculography and
Coronary Artery Imaging cardiovascular magnetic resonance; are they interchangeable?
Eur Heart J 21:1387-1396.
MRI of the coronary arteries has had significant progress 6. Reeder SB, Du YP, Lima JA, Bluemke DA (2001) Advanced
cardiac MR imaging of ischemic heart disease. Radiographics
in recent years. Whole-heart MRI technique involves ax- 21:1047-1074.
ial electrocardiogram-gated SSFP acquisitions with navi- 7. Shehata ML, Cheng S, Osman NF et al (2009) Myocardial
gator pulses to track the motion of the diaphragm [41- tissue tagging with cardiovascular magnetic resonance. J Car-
44]. Fat suppression is used to suppress mediastinal and diovasc Magn Reson 11:55.
8. Bottini PB, Carr AA, Prisant LM et al (1995) Magnetic reso-
pericardial fat. Acquisitions take between 5 and 15 min nance imaging compared to echocardiography to assess left
during free breathing. At 1.5T in particular, no contrast ventricular mass in the hypertensive patient. Am J Hypertens
agent is used. The spatial resolution of whole-heart ac- 8:221-228.
quisitions is approximately 1 mm in the z-axis direction, 9. Hundley WG, Bluemke DA, Finn JP et al (2010) ACCF/ACR/
which is greater than the resolution of CT scanning AHA/NASCI/SCMR 2010 expert consensus document on car-
diovascular magnetic resonance: a report of the American Col-
(0.50.6 mm) [43, 45]. Nevertheless, image quality is ex- lege of Cardiology Foundation Task Force on Expert Consen-
cellent for the exclusion of coronary artery anomalies. sus Documents. Circulation 121:2462-2508.
Unlike CT scanning, coronary MRI is not sensitive to 10. Davies MJ (2000) The cardiomyopathies: an overview. Heart
highly calcified coronary artery calcification. Given the 83:469-474.
relative ease of application, coronary imaging is often 11. Nazarian S, Bluemke DA, Lardo AC et al (2005) Magnetic
resonance assessment of the substrate for inducible ventricular
added as a routine part of the protocol for young patients tachycardia in nonischemic cardiomyopathy. Circulation
with unexplained chest pain. The detection rate of coro- 112:2821-2825.
nary artery stenosis in selected patient populations ap- 12. Piers SR, Tao Q, van Huls van Taxis CF et al (2013) Contrast-
pears to be similar to that achieved with CT scanning, ex- enhanced MRI-derived scar patterns and associated ventricu-
cept that distal arteries less well evaluated [43, 46]. In ad- lar tachycardias in nonischemic cardiomyopathy: implications
for the ablation strategy. Circ Arrhythm Electrophysiol 6:875-
dition, the technique has a lower technical success rate 883.
than CT scanning: non-diagnostic examinations occur in 13. Doltra A, Amundsen BH, Gebker R (2013) Emerging concepts
about 20% of patients. for myocardial late gadolinium enhancement MRI. Curr Car-
diol Rev 9:185-190.
14. Wu KC (2006) Utility of cardiac MRI in the diagnosis of
hypertrophic cardiomyopathy. Curr Cardiol Rep 8:41.
Conclusion 15. Maron MS, Olivotto I, Harrigan C et al (2011) Mitral valve ab-
normalities identified by cardiovascular magnetic resonance
MRI has emerged as a definitive examination for a wide represent a primary phenotypic expression of hypertrophic
array of complex cardiovascular problems. Its is a promi- cardiomyopathy. Circulation 124:40-47.
16. Ho CY, Sweitzer NK, McDonough B et al (2002) Assessment
nent imaging modality in multiple facets of nonischemic of diastolic function with Doppler tissue imaging to predict
and ischemic cardiomyopathy. Continuing challenges are genotype in preclinical hypertrophic cardiomyopathy. Circula-
the complexity of the modality compared to CT scanning tion 105:2992-2997.
142 D. Revel, D.A. Bluemke
17. Maron MS (2012) Clinical utility of cardiovascular magnetic of the American College of Cardiology Foundation Appropri-
resonance in hypertrophic cardiomyopathy. J Cardiovasc Magn ate Use Criteria Task Force, American Heart Association,
Reson 14:13. American Society of Echocardiography, American Society of
18. Noureldin RA, Liu S, Nacif MS et al (2012) The diagnosis of Nuclear Cardiology, Heart Failure Society of America, Heart
hypertrophic cardiomyopathy by cardiovascular magnetic res- Rhythm Society, Society for Cardiovascular Angiography and
onance. J Cardiovasc Magn Reson 14:17. Interventions, Society of Cardiovascular Computed Tomogra-
19. Wu KC, Heldman AW, Brinker JA et al (2001) Microvascular phy, Society for Cardiovascular Magnetic Resonance, and So-
obstruction after nonsurgical septal reduction for the treatment ciety of Thoracic Surgeons. J Am Coll Cardiol 63:380-406.
of hypertrophic cardiomyopathy. Circulation 104:1868. 36. Nandalur KR, Dwamena BA, Choudhri AF et al (2007) Diag-
20. te Riele AS, Tandri H, Bluemke DA (2014) Arrhythmogenic nostic performance of stress cardiac magnetic resonance imag-
right ventricular cardiomyopathy (ARVC): cardiovascular ing in the detection of coronary artery disease: a meta-analy-
magnetic resonance update. J Cardiovasc Magn Reson 16:50. sis. J Am Coll Cardiol 50:1343-1353.
21. Jain A, Tandri H, Calkins H, Bluemke DA (2008) Role of car- 37. Lipinski MJ, McVey CM, Berger JS et al (2013) Prognostic
diovascular magnetic resonance imaging in arrhythmogenic value of stress cardiac magnetic resonance imaging in patients
right ventricular dysplasia. J Cardiovasc Magn Reson 10:32. with known or suspected coronary artery disease: a systemat-
22. Marcus FI, McKenna WJ, Sherrill D et al (2010) Diagnosis of ic review and meta-analysis. J Am Coll Cardiol 62:826-838.
arrhythmogenic right ventricular cardiomyopathy/dysplasia: 38. Schwitter J, Wacker CM, Wilke N et al (2013) MR-IMPACT II:
proposed modification of the task force criteria. Circulation Magnetic Resonance Imaging for Myocardial Perfusion As-
121:1533-1541. sessment in Coronary artery disease Trial: perfusion-cardiac
23. Friedrich MG (2000) Magnetic resonance imaging in cardio- magnetic resonance vs. single-photon emission computed to-
myopathies. J Cardiovasc Magn Reson 2:67-82. mography for the detection of coronary artery disease: a com-
24. Schulz-Menger J, Friedrich MG (2000) Magnetic resonance parative multicentre, multivendor trial. Eur Heart J 34:775-781.
imaging in patients with cardiomyopathies: when and why. 39. Shan K, Constantine G, Sivananthan M, Flamm SD (2004)
Herz 25:384-391. Role of cardiac magnetic resonance imaging in the assessment
25. Vignaux O (2005) Cardiac sarcoidosis: spectrum of MRI fea- of myocardial viability. Circulation 109:1328-1334.
tures. AJR Am J Roentgenol 184:249-254. 40. Gutberlet M, Frohlich M, Mehl S et al (2005) Myocardial
26. Tadamura E, Yamamuro M, Kubo S et al (2005) Effectiveness viability assessment in patients with highly impaired left ven-
of delayed enhanced MRI for identification of cardiac sar- tricular function: comparison of delayed enhancement, dobut-
coidosis: comparison with radionuclide imaging. AJR Am J amine stress MRI, end-diastolic wall thickness, and TI201-
Roentgenol 185:110-115. SPECT with functional recovery after revascularization. Eur
27. Pennell DJ, Berdoukas V, Karagiorga M et al (2006) Random- Radiol 15:872-880.
ized controlled trial of deferiprone or deferoxamine in beta- 41. Yonezawa M, Nagata M, Kitagawa K et al (2014) Quantitative
thalassemia major patients with asymptomatic myocardial analysis of 1.5-T whole-heart coronary MR angiograms ob-
siderosis. Blood 107:3738-3744. tained with 32-channel cardiac coils: a comparison with con-
28. Pennell DJ (2005) T2* magnetic resonance and myocardial ventional quantitative coronary angiography. Radiology
iron in thalassemia. Ann N Y Acad Sci 1054:373-378. 271:356-364.
29. Carpenter JP, Grasso AE, Porter JB (2013) On myocardial 42. Yoon YE, Kitagawa K, Kato S et al (2012) Prognostic value of
siderosis and left ventricular dysfunction in hemochromatosis. coronary magnetic resonance angiography for prediction of
J Cardiovasc Magn Reson 15:24. cardiac events in patients with suspected coronary artery
30. Maceira AM, Prasad SK, Hawkins PN (2008) Cardiovascular disease. J Am Coll Cardiol 60:2316-2322.
magnetic resonance and prognosis in cardiac amyloidosis. J 43. Sakuma H (2011) Coronary CT versus MR angiography: the
Cardiovasc Magn Reson 10:54. role of MR angiography. Radiology 258:340-349.
31. Karamitsos TD, Piechnik SK, Banypersad SM et al (2013) 44. Nagata M, Kato S, Kitagawa K et al (2011) Diagnostic accu-
Noncontrast T1 mapping for the diagnosis of cardiac amyloi- racy of 1.5-T unenhanced whole-heart coronary MR angiogra-
dosis. JACC Cardiovasc Imaging 6:488-497. phy performed with 32-channel cardiac coils: initial single-
32. Rajiah P (2011) Cardiac MRI: Part 2, pericardial diseases. AJR center experience. Radiology 259:384-392.
Am J Roentgenol 197:W621-634. 45. Kefer J, Coche E, Legros G et al (2005) Head-to-head com-
33. Friedrich MG, Sechtem U, Schulz-Menger J et al (2009) Car- parison of three-dimensional navigator-gated magnetic reso-
diovascular magnetic resonance in myocarditis: A JACC White nance imaging and 16-slice computed tomography to detect
Paper. J Am Coll Cardiol 53:1475-1487. coronary artery stenosis in patients. J Am Coll Cardiol 46:92-
34. ODonnell DH, Abbara S, Chaithiraphan V et al (2009) Car- 100.
diac tumors: optimal cardiac MR sequences and spectrum of 46. Pouleur AC, le Polain de Waroux JB, Kefer J et al (2008) Di-
imaging appearances. AJR Am J Roentgenol 193:377-387. rect comparison of whole-heart navigator-gated magnetic res-
35. Wolk MJ, Bailey SR, Doherty JU et al (2014) onance coronary angiography and 40- and 64-slice multide-
ACCF/AHA/ASE/ASNC/HFSA/HRS/SCAI/SCCT/SCMR/ST tector row computed tomography to detect the coronary artery
S 2013 multimodality appropriate use criteria for the detection stenosis in patients scheduled for conventional coronary an-
and risk assessment of stable ischemic heart disease: a report giography. Circ Cardiovasc Imaging 1:114-121.
IDKD 2015-2018
stratify future cardiovascular risk. Only patients at high- have no significant stenoses [3, 6]. Conversely, non-inva-
er risk will be referred to interventional procedures, with sive testing does not always guide clinical management
the final goal of improving their outcome [2]. since 27% of patients with high-risk imaging results do
The management of patients with stable chest pain not undergo ICA [7].
starts with an estimation of the pre-test probability (PTP)
of CAD, based on age, sex, and type of symptoms ac-
cording to an updated predictive model [3]. Patients with Nuclear Myocardial Perfusion Imaging (MPI) and
a PTP score 15% should not undergo further testing Hybrid CCTA-MPI Imaging for the Diagnosis of Stable CAD
while in patients with a PTP 85% non invasive testing
is not indicated for diagnosis but only for risk stratifica- Stress MPI by SPECT and PET is an established and
tion [2]. According to international guidelines, patients widely used approach in the diagnostic and risk assess-
with an intermediate PTP should preferably undergo ment of patients with suspected CAD. These techniques
stress MPI rather than exercise echocardiography [2, 4]. provide information on the presence, extent, and severity
CCTA is considered a reasonable alternative in interme- of myocardial perfusion abnormalities, which define the
diate-low likelihood patients or in case of doubtful stress functional significance of coronary disease and are relat-
tests results. ed to event risk. Since CCTA is able to depict anatomical
Only in the presence of high-risk findings at non inva- abnormalities of the vessel wall and has a high predictive
sive imaging should patients be referred to invasive in- value in excluding significant coronary stenoses and in
vestigation and possible coronary intervention [2, 5]. identifying high-risk coronary anatomy, a combination of
High-risk findings are defined by the presence of left functional and anatomical information, obtained either
main or proximal left anterior descending arterial disease by side-by-side interpretation of stress MPI and CCTA
and/or three-vessel disease as determined by CTCA images or by using a hybrid approach, has recently been
and/or by the documentation of significant inducible my- proposed as a more accurate strategy in patients with
ocardial ischemia (10% of the left ventricular myo- known or suspected CAD [8] (Fig. 1).
cardium) at stress testing. ICA in patients with stable Large observational studies have demonstrated that a
CAD is not recommended as a first testing method, al- normal MPI yields a favorable prognosis, with an annu-
though in current practice as many as 30% of patients alized event-rate of 0.6%, while an abnormal scan im-
with no symptoms (including no angina) and 16% of pa- plies a 3- to 7-fold increase in annual cardiac events, es-
tients without non-invasive testing, and 15% of those pecially those related to the extent and severity of perfu-
with prior normal non-invasive testing (resting electro- sion defects [9-11]. Additional prognostic data on left
cardiography, echocardiography, CT, or a stress test) un- ventricular volumes and function (e.g., ejection fraction)
dergo ICA [3]. Perhaps as a consequence, 62% of stable at baseline and possible evidence of transient left ven-
patients without known CAD who undergo elective coro- tricular dysfunction after stress are provided by the gated
nary angiography in the USA and 42% of those in Europe acquisition of SPECT data [12].
An Integrative Approach to the Imaging of Ischemic Heart Disease 145
The demonstration by SPECT imaging of significant Non Invasive Imaging of the Heart and Large Vessels
inducible myocardial ischemia not only defines a high
cardiovascular risk for a specific patient but also guides Cross-sectional CT and MR imaging are well suited to
treatment to modify that risk and improve outcome. assess the gross morphology of the heart and large ves-
Hachamovitch et al. [13] demonstrated that revascular- sels. CT is currently routinely applied to evaluate acute
ization procedures have a beneficial impact on outcome chest pain (life-threatening coronary occlusion, aortic
only in the presence of ischemia involving 10% of the dissection, and pulmonary embolism should initially be
left ventricular myocardium at SPECT imaging. In a excluded). However, it should be noted that the differen-
prospective nuclear substudy of the COURAGE trial, a tial diagnosis of acute chest pain is quite extensive and
better prognosis was related to the extent of ischemia re- many other causes may be incidentally diagnosed by CT
duction evidenced by SPECT [14]. These data justify the (e.g., pneumonia, lung disease, chest wall disease, peri-
use of SPECT as one of the techniques recommended by cardial disease). In many hospitals excluding pul-
the European Society of Cardiology (ESC) guidelines to monary embolism is a very common and sometimes
diagnose CAD, stratify risk in patients with suspected misused indication for urgent CT of the chest. In the
disease, and guide further management and treatment of emergency setting, CT is a preferred technique due to its
the patient to improve outcome. speed, availability, robust image quality, and ease of use
In analogy to SPECT, the presence and extent of myo- in acutely ill patients who may require assisted ventila-
cardial perfusion defects demonstrated by PET provide tion and direct supervision. The complications of is-
strong prognostic information [15]. PET allows measure- chemic heart disease may be well shown by CT and/or
ment of absolute myocardial blood flow and flow reserve MR imaging (e.g., scar, aneurysm, contained rupture,
(MFR) and has incremental prognostic value compared to thrombus).
the evaluation of perfusion defects alone. A blunted MFR MR imaging is a versatile technique for evaluating
is an independent predictor of risk compared to more many aspects of ischemic and non-ischemic heart dis-
common prognostic indicators, such as transient regional ease. In stable patients with suspected CAD, imaging
perfusion defects, previous myocardial infarction, and will mostly focus on the detection of coronary stenosis
left ventricular ejection fraction [16]. Interestingly, high- and defining the functional significance of the stenosis.
risk, diffuse anatomical CAD may manifest as apparent- The focus of imaging will be different in patients with
ly normal regional perfusion images with homogeneous suspected acute coronary syndromes (acute infarction,
tracer distribution; however, it may be recognized by unstable angina pectoris). In the acute setting it will be
global reductions of myocardial blood flow and MFR at important to use imaging tools to diagnose or exclude
quantitative PET imaging [17]. coronary occlusion, especially when the clinical signs
Improved risk stratification is obtained by combining and symptoms are inconclusive. CCTA is an excellent
the anatomical information of CCTA with the functional gatekeeper for excluding CAD in the emergency room
data of MPI. The likelihood of diagnosing obstructive setting and thereby decrease the length of stay of pa-
CAD by ICA is highest when both CCTA and MPI are tients in the hospital. MR MPI has also successfully
abnormal [18]. The hybrid approach is also useful to been employed as a gatekeeper in the emergency room
stratify risk in patients with doubtful results at either setting for the exclusion of ischemia due to coronary
functional or anatomical evaluation. Accordingly, a com- artery stenosis. After coronary artery occlusion, the de-
bined anatomical and functional assessment provides pendent myocardium distal from the occluded coronary
complementary rather than overlapping diagnostic and artery is at risk of necrosis, thus mandating early and
prognostic information. timely intervention. The area at risk may be visualized
on T2-weighted MR sequences, which will show a re-
gion of high signal intensity due to developing edema in
CT and CMR Imaging in the Diagnosis of Stable CAD the area at risk. Early intervention may prevent myocar-
dial necrosis in the area at risk (aborted infarction).
CT and MRI techniques have also been developed to The necrosis develops over several hours, starting as a
evaluate myocardial perfusion. Currently, in MPI, MR is wavefront at the endocardial site within the confines of
more widely used than CT. Both are performed with the the area at risk that, when not halted by intervention,
patient at rest and the results are compared to perfusion then progresses over time and may fill the entire area at
under adenosine stress. MR perfusion imaging provides risk. There is some discussion as to whether T2-weighted
dynamic information on the wash-in of gadolinium-based techniques are reliable enough for quantification of the
contrast agents into the myocardium. Ischemia will be vi- risk area, because of potential imaging artifacts.
sualized during stress MR perfusion as a perfusion defect Nonetheless, MR-based estimates of the area at risk
(cold spot). CT perfusion is more static and provides have been incorporated as end-points in several cardiol-
perfusion information mostly at a fixed time point in a ogy trials.
non-dynamic manner. The difference between ischemic The wave front of progressing necrosis can be defined
and normal myocardium is far greater on MR than on CT very accurately by late gadolinium enhancement (LGE).
imaging. Combining LGE and T2-weighted MR images is an
146 A. de Roos, D. Neglia
option to assess the infarct:risk ratio as an important clin- sive differential diagnosis of its own. Primary (idiopath-
ical parameter. LGE has been used for over 25 years to ic), secondary (toxic, infiltration, etc.), and complex ge-
characterize myocardial necrosis, in both the acute setting netic forms of cardiomyopathy may present with the di-
and chronic setting. T2-weighted MR techniques may be lated phenotype. The idiopathic form of dilated car-
useful to distinguish acute from chronic infarcts. Acute diomyopathy is diagnosed after excluding secondary
infarcts will show LGE in conjunction with high my- forms. Idiopathic dilated cardiomyopathy may be further
ocardial signal (acute edema) in the area at risk, whereas characterized by LGE. Among the LGE patterns ob-
chronic infarcts will show LGE in the absence (no edema served in dilated cardiomyopathy are the so-called mid-
present) of high signal intensity in the myocardium at wall stripe as well as patchy, diffuse, and subepicardial
risk. enhancement. Interestingly, LGE has prognostic implica-
tions in idiopathic dilated cardiomyopathy, as the mid-
wall stripe pattern may herald arrhythmias and sudden
CT and CMR Imaging in Patients with Heart Failure death. LGE patterns may be helpful for risk stratification
and guiding treatment options (e.g., implantable car-
The epidemic growth of the number of patients suffer- dioverter-defibrillator).
ing from heart failure constitutes a diagnostic and ther- The differential diagnosis of hypertrophic phenotypes
apeutic challenge. CT and MR imaging are playing an starts with the exclusion of arterial hypertension and aor-
increasingly important role in the work-up and follow- tic stenosis as the underlying causes of pressure overload
up of patients with heart failure. The diagnostic chal- on the left ventricle. It is also important to distinguish
lenge in heart failure is to identify its most likely cause. global or diffuse hypertrophy from local hypertrophy
First and foremost, it is important to exclude CAD, e.g., (lumps and bumps, as seen in genetic forms of hyper-
by using CCTA to exclude coronary stenosis and trophic cardiomyopathy). The diffuse hypertrophic phe-
plaques. However, it may also be important to evaluate notype has a differential diagnosis that should be consid-
the myocardium directly, to assess the presence of myo- ered from the imaging perspective. Genetically deter-
cardial scar and infarction, because in a small number mined hypertrophic (obstructive) cardiomyopathy may
of patients infarction may be present after recanalization present as diffuse global hypertrophy, although localized
of a previously occluded coronary artery. In this respect forms are more common (e.g., classic asymmetric septal
LGE by MR imaging has a central role in the identifi- hypertrophy). Many common, but also uncommon, infil-
cation of ischemic myocardial scar (i.e., subendocardial trative myocardial diseases may present with the hyper-
LGE). trophic phenotype (e.g., amyloid, storage diseases, Fab-
The MR protocol in the evaluation of patients with rys disease, inflammatory processes such as sarcoid,
heart failure includes functional imaging (global and re- medication such as amiodarone and chloroquine). Genet-
gional function), velocity-encoded MR to assess mitral ically determined hypertrophic cardiomyopathy has many
flow, T2-weighted sequences to assess edema, T1-weighted additional morphological features that may be a clue for
sequences (T1 mapping with techniques such as Look- the correct diagnosis, such as the appearance and length
Locker technique or modified Look-Locker inversion re- of the mitral leaflets, the systolic anterior motion of the
covery sequences) to assess diffuse fibrosis and other dif- anterior leaflet of the mitral valve, secondary mitral in-
fuse myocardial infiltration unseen with LGE, and LGE sufficiency and atrial enlargement, crypts in the myocar-
sequences to assess scar location, size, and distribution. dial wall as an early sign of hypertrophy in a mutation
The combined assessment of the location and extent of is- carrier, and various patterns of LGE. Hypertrophic car-
chemic scar, the presence and severity of secondary mi- diomyopathy is a common cause of sudden death in ath-
tral regurgitation, and precise estimates of left ventricular letes and apparently healthy young people, although the
volumes provide a surgical road-map for planning scar death rate may be lower than previously suggested. The
resection in conjunction with mitral valve repair (e.g., risk of arrhythmias and sudden death may be related to
Dor procedure). the presence and extent of scar tissue as demonstrated by
These combined techniques constitute a comprehen- LGE. Typical LGE enhancement occurs at the right ven-
sive set of tools with which to obtain a specific diagno- tricular attachment points, although the scar can be wide-
sis of underlying ischemic cardiomyopathy. They also of- ly distributed in various patterns. A transmural infarct-
fer a wide array of diagnostic features that may be help- like pattern in an aneurysmatical apical scar may imply a
ful in differentiating ischemic from non-ischemic car- worse prognosis.
diomyopathy and in further differentiating the multitude Finally, amyloid cardiomyopathy may present as a hy-
of common and uncommon causes of non-ischemic car- pertrophic phenotype. The LGE pattern may demonstrate
diomyopathies. In the differential diagnosis of non- a typical diffuse subendocardial pattern throughout the
ischemic cardiomyopathy it may be useful to make an left ventricle, possibly also involving the right ventricle
initial distinction between dilated (i.e., globally dilated and atrial walls. T1 mapping techniques are currently
left ventricle, thin walls) and hypertrophic (thick walls, used to characterize diffuse myocardial involvement to
locally or diffuse) phenotypes. The dilated phenotype is better advantage, such as in amyloid cardiomyopathy. T1
a quite common cause of heart failure and has an exten- mapping is a promising tool for characterizing diffuse
An Integrative Approach to the Imaging of Ischemic Heart Disease 147
19. McMurray JJ, Adamopoulos S, Anker SD et al (2012) ESC 20. Bonow RO, Maurer G, Lee KL et al (2011) Myocardial via-
Guidelines for the diagnosis and treatment of acute and chron- bility and survival in ischemic left ventricular dysfunction. The
ic heart failure 2012: The Task Force for the Diagnosis and New England Journal of Medicine 364:1617-1625.
Treatment of Acute and Chronic Heart Failure 2012 of the Eu- 21. Ling LF, Marwick TH, Flores DR et al (2013) Identification of
ropean Society of Cardiology. Developed in collaboration with therapeutic benefit from revascularization in patients with left
the Heart Failure Association (HFA) of the ESC. Eur Heart J ventricular systolic dysfunction: inducible ischemia versus hi-
33:1787-1847. bernating myocardium. Circ Cardiovasc Imaging 6:363-372.
IDKD 2015-2018
by signs or symptoms of hypoperfusion or ischemia to the cause of IMH has never been definitively proven. IMH is
distal organs, extremities, or brain. treated similar to dissections in terms of initial diagnosis
Traditionally, acute aortic syndromes have been cate- as well as clinical management. Most IMHs occur in the
gorized as aortic dissection, intramural hematoma, and descending thoracic aorta and can be associated with se-
penetrating atherosclerotic ulcer. vere pain. The imaging features vary depending on the
Aortic dissection (AD) is characterized by a separation amount of blood accumulated in the wall of the aorta, but
of the aortic media, creating an intimal-medial complex typically the normal wall measures <7 mm in thickness.
that separates from the remaining aortic wall. Blood The natural history of IMH can be quite variable. Rough-
flowing between the intimal-medial complex (intimal ly one-third of them will progress to AD, one third will
flap), and the remaining wall is considered to be within a be stable, and one third will resolve [3]. The mortality for
false lumen, whereas blood flow bounded by the intima patients with IMH involving the ascending aorta is simi-
is considered to be within a true lumen. Multiple com- lar to that of patients with classic dissection; therefore
munication points can be observed between the true and these patients are treated as though they have a classic
false lumens. The most proximal communication is de- AD, with emergent surgery [4]. Patients with IMH in-
fined as the entry tear and the remaining points of com- volving the descending thoracic aorta can initially be giv-
munication as the exit tears, implying flow directional- en appropriate therapy for hypertension and then fol-
ity from true to false lumen and from false to true lumen, lowed.
respectively. The actual incidence of AD is difficult to de- Penetrating atherosclerotic ulcer (PAU) is a condition
fine, since AD involving the ascending aorta is often fa- that originates with atherosclerotic plaque involvement of
tal and patients frequently die prior to hospitalization. the aorta, primarily the descending thoracic aorta (Fig. 1).
Likewise, AD is sometimes misdiagnosed on initial pre- As the plaque evolves, the ulceration penetrates the in-
sentation; consequently, these patients are also at risk for ternal elastic lamina into the media of the aortic wall.
death outside of the hospital. Nevertheless, various pop- Over time, the PAU may extend through all three layers
ulation-based studies suggest that the incidence of AD of the aortic wall to form a false or pseudoaneurysm. A
ranges from 2 to 4 case per 100,000 patients [2]. finding of PAU does not necessarily imply the existence
The temporal definition of acute AD is a dissection of an acute aortic syndrome. Signs of IMH or extravasa-
that is identified less than 2 weeks after the onset of tion indicate acuity.
symptoms, while sub-acute ranges from 2 to 6 weeks fol- There are two limitations to this traditional classifica-
lowing an initial painful episode. Chronic AD is defined tion. The first concerns the omission of a rupturing true
as a dissection identified >6 weeks after the onset of pain. aortic aneurysm, as the nature of the presentation and the
Intramural hematoma (IMH) is an entity that was first severity of the event are similar in PAU and the other
described approximately 25 years ago as a stagnant col- acute aortic syndromes. The second limitation is that
lection of blood within the aortic wall. The common as- IMH, defined as a stagnant intramural collection of
sociation of IMH with pathologically detected intimal de- blood, can be observed in the setting of AD, PAU, and
fects led to the hypothesis that most are sequelae of pen- rupturing aortic aneurysm. As such, it is a feature or char-
etrating atherosclerotic ulcer. An alternative cause of acteristic associated with any of the acute aortic syn-
IMH, typically invoked in the absence of an intimal de- dromes, reflecting degradation of the aortic wall as a har-
fect, is rupture of the vasa vasorum. This hypothetical binger of impending aortic rupture.
a b c
Fig. 1 a-c. Rupturing thoracic aortic aneurysm. a Unenhanced computed tomography (CT) demonstrates a high-attenuation hematoma in the
right pleural space and middle mediastinum. There is an aneurysm of the descending aorta. b Unenhanced CT section 5 cm inferior to (a),
reveals an intramural hematoma (IMH) at the inferior margin of the aneurysm, with extravasation of blood into the middle mediastinum.
Note the displacement of intimal calcium along the inner wall of the IMH. c Following the administration of intravenous contrast materi-
al, the IMH is harder to visualize because of the wider window used to display the CT angiogram.
Acute Aortic Syndrome: State-of-the-Art Diagnostic Imaging 151
a b
Fig. 2 a, b. Type B aortic dissection. a The true lumen is completely collapsed posteri-
orly and only the true lumen fills with contrast material. The entry tear was (not shown)
in the proximal descending aorta. b Because of their supply from the aortic true lumen,
the renal arteries do not opacity and the kidneys are not perfused
In consideration of these two points, a new classifica- 5. The presence of IMH, PAU, or calcification.
tion scheme has been proposed based upon the primary 6. Extension of the aortic abnormality to include inter-
location of the lesion within the aortic wall. In this new branch vessels, including dissection and aneurysm,
classification scheme, there are three pathological enti- and secondary evidence of end-organ injury (e.g., re-
ties: AD, PAU, and rupturing aortic aneurysm [5]. These nal or bowel hypoperfusion).
three entities are differentiated by the fact that AD prin- 7. Evidence of aortic rupture, including periaortic and me-
cipally involves the aortic media, PAU originates within diastinal hematoma, pericardial and pleural fluid, and
the aortic intima, and aortic aneurysm is a disease of all contrast extravasation from the aortic lumen (Fig. 2).
three layers. The presence of IMH is an observation or 8. When a prior examination is available, direct image-
epi-phenomenon to be applied to any of these three fun- to-image comparison to determine whether there has
damental pathologies. In the setting of an isolated IMH been any increase in lesion diameter.
without PAU, non-communicating dissection has been
proposed as a descriptor, although the term IMH is
more commonly associated with this lesion. CT-Based Imaging Approaches to Acute Aortic
Syndromes
Essential Elements of Aortic Imaging Reports High-quality and comprehensive aortic and end-organ as-
sessment is performed using multi-detector row CT with
In 2010, a group of medical organizations representing at least 16 detector rows. This scanner configuration al-
the disciplines of cardiology, radiology, thoracic surgery, lows for imaging from the neck through the pelvis, ac-
and anesthesia published guidelines for the diagnosis and quiring 1.5-mm-thick transverse sections during the ar-
management of patients with thoracic aortic diseases. In terial phase of enhancement from an intravenous contrast
this report, the authors identified eight essential elements administration. It also allows for the use of electrocar-
that should be addressed in aortic imaging reports [2]. diographic (ECG) gating of the scan when appropriate, as
While these guidelines are not comprehensive, nor do described below.
they imply the necessity of reporting every element in
every case, they are a useful construct from which to Unenhanced CT
build any formalized description of the imaging findings
of an acute aortic syndrome. They are the following. An unenhanced scan prior to the administration of intra-
1. The location at which the aorta is abnormal. venous contrast can be valuable for the detection of intra-
2. The maximum diameter of any dilation, measured mural and periaortic blood, which is often subtle. It can al-
from the external wall of the aorta, perpendicular to so be useful for mapping the specific regions of the aorta
the axis of flow, and the length of the aorta that is ab- that are abnormal and thus guide the mode of subsequent
normal. CT angiographic acquisition. While an associated increase
3. For patients with genetic syndromes who are at risk for in radiation exposure results from this approach, the po-
aortic root disease, measurements of the aortic valve, tential value of the information almost always outweighs
sinuses of Valsalva, the sinotubular junction, and the the risk. Using dual-energy scanning, a virtual unenhanced
ascending aorta. scan might obviate the need for a separate unenhanced ac-
4. The presence of internal filling defects consistent with quisition. However, this approach has not been compre-
thrombus or atheroma. hensively validated in acute aortic syndromes. Moreover,
152 T. Grist, G.D. Rubin
it eliminates the possibility of using the preliminary unen- the presence of unrelated but important abdominal aorto-
hanced acquisition to guide the decision to gate the CT an- iliac pathology, the value of assessing the caliber of a
giogram. transfemoral delivery route to intra-aortic repair devices,
and the possibility of abdominal visceral ischemia, scan
CT Angiography ranges that extend through the abdomen and pelvis are
highly recommended as a routine approach to imaging
While unenhanced imaging can reveal aortic dilation, in- acute aortic syndromes. By beginning the scan in the
tramural and extra-aortic hemorrhage, and in uncommon neck and extending inferiorly below the lesser trochanters
circumstances directly visualize an intimal flap, the use of of the femurs, the scan range will comfortably include
intravenously administered contrast medium is required for several centimeters of the cervical carotid arteries
a complete assessment in patients in whom acute aortic through the bifurcation of the femoral artery. Scan ranges
syndrome is suspected (Fig. 3). The volume and flow rate that include less anatomy risk the possibility that impor-
of the contrast material should be adjusted based on patient tant observations will be missed, such that additional CT
size. A concentrated iodine solution containing 350 mg scans with further injections of iodinated contrast mater-
of iodine/mL will assure adequate intravenous delivery of ial may be required.
iodine with a safe and reliable flow rate of the contrast
material into the peripheral vein. Typical volumes and Electrocardiographic Gating
injector flow rates for iodinated contrast range between 60
and 115 mL at flow rates between 3.5 and 6 mL/s. When the ascending aorta is involved by an acute aortic
To assure diagnostic aortic enhancement throughout syndrome, electrocardiographic (ECG) gating can be
the CT acquisition, the duration of the contrast injection valuable. ECG gating allows for clear delineation of the
should exceed the scan duration by 510 s, and initiation position of the intimal flap across the cardiac period,
of the CT angiographic acquisition should be based on distinction of the involvement of the structures of the aor-
the active monitoring of the arrival of iodine within the tic root including the coronary artery ostia and the aortic
descending thoracic aorta. annulus, elimination of pulsation-related artifacts that can
blur the aortic wall, and subtle regions of extravasation.
Imaging the Abdominal Aorta and Iliac Arteries Unlike the use of ECG gating in the setting of coronary
artery disease assessment, the strategy for using ECG gat-
Because of the likelihood of direct extension of thoracic ing in acute aortic syndromes does not rely upon the ma-
aortic disease into the abdominal aorta and iliac arteries, nipulation of heart rate or coronary artery dimension us-
ing -blockers or nitrates. Regardless of the basal heart
rate, the placement of ECG leads and the acquisition of a
retrospectively gated CT scan (with judicious use of ECG
directed X-ray tube current pulsing to minimize radiation
exposure) allow for a four-dimensional assessment of the
aortic root, aortic valve, coronary arteries, and ascending
aorta. It is sufficient to reconstruct 10 phases every 10%
of the R-R interval. Gating is only beneficial through the
thoracic aorta. The abdomen and pelvis are imaged after
the chest acquisitions, using a non-gated acquisition with
a minimization of the delay between the two scans so that
only one contrast injection is required.
associated with dissection [6]. These scans are performed pression of branch vessels, especially the coronary arter-
with minimal MRI table time and by using fast acquisi- ies. Likewise, cine MRI can be used to further character-
tion techniques. In patients who are stable, contrast-en- ize the relationship of a type A dissection flap to the aor-
hanced magnetic resonance angiography (MRA) may be tic valve and any resultant aortic valve insufficiency. T1-
used to further diagnose and characterize the configura- and T2-weighted imaging is typically performed with fast
tion of the intimal flap and the involvement of branch spin-echo technique (Table 1).
vessels in acute AD. Finally, gadolinium contrast-enhanced MRA is used to
In the setting of acute dissection with thrombosis lu- characterize the luminal pathology associated with acute
men and intramural hematoma, SSFP MRI alone may be aortic syndrome [7]. Contrast-enhanced MRA is usually
inadequate for detecting hemorrhage in the wall of the acquired during the arterial phase of contrast as well as
aorta. In this clinical scenario, T1- and T2- weighted during the delayed steady state phase of contrast en-
spin-echo technique is typically necessary for diagnosis. hancement. Some authors describe the use of time-re-
MRI using these techniques may be useful in detecting solved 3D MRA during a small initial bolus of gadolinium-
and characterizing the age of the intramural hematoma, based contrast agent, thus allowing the dynamic de-
with typical characteristics such as a transition of the he- lineation of the filling pathways of the true and false lu-
morrhage from deoxyhemoglobin in the early stage of men and branch vessels [8]. These images are typically
IMH to extracellular methemoglobin in its late stage. followed by high-resolution 3D images for more precise
In the setting of aortic aneurysm with suspected rup- anatomic characterization.
ture, patients typically are unstable and therefore far more Extracellular contrast agents are typically the primary
complicated to image with MRI; in these cases, CT is the diagnostic enhancement agents. Contrast agents with
preferred imaging modality. MRI has been used, howev- higher relaxivities are preferred because of their greater
er, to evaluate stable patients with suspected complica- signal at a lower dose as well as their protein binding,
tions of aneurysm and can be particularly helpful in pa- which facilitates delayed imaging in the steady state. For
tients with suspected inflammatory or infectious causes delayed imaging, a fat-suppressed post-contrast T1-
of the aneurysm, due to the exquisite sensitivity of MRI weighted gradient echo image, with spoiling of the trans-
to inflammation in the tissues surrounding the aorta. In verse magnetization, is typically acquired. The delayed
this setting, the use of gadolinium-based contrast agents, images are particularly helpful in evaluating extraluminal
especially those with prolonged blood-pool retention, can pathology affecting the aorta, including aortic leaks, ar-
be particularly informative. teritis, and infection, and in characterizing the size and
extent of hematoma outside the aortic wall (Fig. 3). The
MRI Technique arterial phase of contrast-enhanced images is used to de-
lineate branch vessel involvement of the intimal flap,
Evaluation of thoracic pathology with MRI begins with characterizing the size of aortic enlargement in dissection
rapid SSFP imaging of the aorta and its major branches. and/or dissecting aneurysm, and determining the nature
This technique, in which high-performance gradients are of the intraluminal pathology associated with acute aor-
applied to acquire images using a short repetition time tic syndrome, such as an intimal flap, associated throm-
(TR) and short echo time (TE), are especially useful for bosis, or an atherosclerotic aortic plaque causing in-
imaging aortic vascular pathology in the absence of creased risk of distal atheroembolization.
gadolinium contrast medium (Table 1) [6]. The images
are typically initially acquired in axial and oblique sagit- MRI Findings in Acute Aortic Syndrome
tal projections, and gated to the diastolic phase of the car-
diac waveform. For a more complete characterization of MRI findings in classic AD are similar to those identified
intimal flap motion in the setting of AD, cine SSFP MRI on CT, including a displaced intimal flap, thrombosed lu-
can be used to further delineate the entry and exit zones men, and demonstration of entry and exit tears [9]. One
of the intimal flap, as well as potential dynamic com- key finding on CT that is not reliable on MRI is the
Table 1. Magnetic resonance imaging techniques for the diagnosis and monitoring of acute aortic syndromes
Sequence Plane TR/TE (MS) Matrix Acceleration factor Gating
SSFP 2D axial/sagittal 3/1.5 256192 12 Yes
Time-resolved CE-MRA 3D TRICKS sagittal 3/1 19212832 3
High-resolution CE-MRA 3D sagittal 3/1 320256128 4 No
T1 fast spin echo 2D axial 600/10 256192 n/a No
T2 fast spin echo 2D axial 2500/60 256192 n/a No
4D flow phase contrast 3D sagittal 10/4 25612864 610 Yes
TR, repetition time; TE, echo time; SSFP, steady-state free precession; CE, contrast-enhanced; MRA, magnetic resonance angiography;
TRICKS, time-resolved imaging of contrast kinetics; n/a, not applicable
154 T. Grist, G.D. Rubin
a b
a b
Fig. 5 a, b. Penetrating aortic ulcer. a Early arterial phase and b delayed steady state phase magnetic resonance angiogram (MRA) demon-
strating penetrating aortic ulceration involving the superior surface of the aortic arch (arrows). Note the excellent delineation of the
aortic adventitia on the delayed images due to contrast enhancement during the steady-state imaging phase
displaced intimal calcifications associated with a dis- T1- and T2-weighted images in the chronic phase (Fig. 4).
placed intimal flap. Contrast-enhanced MRA can demon- The exquisite soft-tissue contrast obtained on delayed
strate filling patterns in AD and may be helpful in delin- contrast-enhanced MRI may be helpful in delineating
eating branch-vessel and end-organ perfusion. penetrating aortic ulceration. The immediate arterial-
Spin-echo MRI may allow characterization of the age phase images on contrast-enhanced MRI may show the
of the intramural hematoma associated with an AD or size and extent of ulceration, whereas delayed images en-
penetrating aortic ulceration. Classic features of aortic in- hance visualization of the aortic adventitia and the sur-
tramural hematoma include intermediate to low signal in- rounding soft tissues, thus allowing a more definitive
tensity in the acute phase on both T1- and T2-weighted se- characterization of the aortic enlargement and associated
quences, with gradual transition to high signal intensity on structures in penetrating aortic ulceration (Fig. 5).
Acute Aortic Syndrome: State-of-the-Art Diagnostic Imaging 155
a b
8. Finn JP, Baskaran V, Carr JC et al (2002) Thorax: Low-dose 9. Gebker R, Gomaa O, Schnackenburg B et al (2007) Compari-
contrast-enhanced three-dimensional MR angiography with son of different MRI techniques for the assessment of thoracic
subsecond temporal resolution - Initial results. Radiology aortic pathology: 3D contrast enhanced MR angiography, tur-
224:896-904. bo spin echo and balanced steady state free precession. Int J
Cardiovasc Imaging 23:747-756.
IDKD 2015-2018
different among tumors, depending on whether 50% or artery, which should not be covered without performing
50% of the tumor abuts the pleura, or whether the tumor a carotid-carotid bypass; and the celiac trunk. Elongation
does not abut the pleura at all. A tumor surrounded by of the aortic arch can make an exact and easy placement
lung parenchyma is highly electrically and thermally in- of the stent graft difficult. Paraplegia is a rare but serious
sulated by the air-filled lung parenchyma compared to a side effect in case the descending aorta is treated but is a
tumor with pleural contact and will require less energy de- less common occurrence than in open surgery.
position. Matsumura et al. [9] evaluated the safety and effec-
tiveness of thoracic endovascular aortic repair (TEVAR)
with a contemporary endograft system compared with
Vascular Interventions open surgical repair of descending thoracic aortic
aneurysms and large ulcers. They included 230 patients
Vascular interventions can be divided into arterial and ve- and compared 16 TEVAR subjects treated with a single
nous interventions. Among the former are balloon angio- type of stent graft with 70 patients undergoing open
plasty of the supra-aortic arteries, such as the subclavian surgery. The 30-day survival rate was non-inferior
artery, implantation of thoracic endografts, and em- (p0.01) for the TEVAR group compared with the open
bolization of bronchial arteries. Rarely performed inter- group (98.1% vs. 94.3%). Cumulative major morbidity
ventions include transarterial techniques for tumor treat- scores were significantly lower at 30 days in the TEVAR
ment, such as chemoperfusion of the lateral thoracic, group than in the open group (1.33.0 vs. 2.3.6,
mammarian, and bronchial arteries in the treatment of p0.01). TEVAR patients had fewer cardiovascular, pul-
bronchial or breast cancer. Vascular interventions involv- monary, and vascular adverse events. No ruptures or
ing the pulmonary artery include the occlusion of arteri- conversions occurred in the first year after the proce-
oportal fistulas, particularly in patients with hereditary dure. Reintervention rates were similar in both groups.
hemorrhagic telangiectasia. Local thrombolysis or throm- At 12 months, aneurysm growth was identified in 7.1%
bodestruction of pulmonary emboli is another rare inter- (8/112), endoleak in 3.9% (4/103), and migration (10
vention but it is a promising alternative in emergency mm) in 2.8% (3/107); other device-related problems
cases involving pulmonary embolism (PE). were infrequent. At 1 year of follow-up, they concluded
Venous interventions in the thorax include central ve- that thoracic endovascular aortic repair with the Zenith
nous stents, either to treat malignancies or to recanalize TX2 endovascular graft (William Cook Europe, ApS,
central venous stenoses in order to allow successful Bjaeverskov, Denmark) is a safe and effective alternative
drainage in dialysis patients, the placement and mainte- to open surgical repair for the treatment of anatomically
nance of central venous catheters, fibrin sheath stripping, suitable aneurysms and ulcers of the descending tho-
and the removal of foreign bodies. racic aorta.
Only some of these interventions merit further discussion In a prospective trial, Nienaber et al. [10] investigated
in the following. However, embolization of the bronchial 140 patients with stable type B aortic dissection who
arteries and treatment of malignant venous stenoses, were randomly assigned to elective stent-graft placement
while uncommon, might be helpful in treating patients in addition to optimal medical therapy (n72) or to opti-
with acute symptoms and are considered below. mal medical therapy (n68) with surveillance. There was
no difference in all-cause mortality: cumulative survival
Arterial Interventions was 97.0%3.4% with optimal medical therapy vs.
91.3%2.1% with thoracic endovascular aortic repair
Stent Graft Implantation (p0.16). Aorta-related mortality was not different
(p0.42), and the risk for the combined end point of aor-
The routine commercial availability of stent grafts for the ta-related death (rupture) and progression (including con-
thoracic aorta has drastically changed the treatment of version or additional endovascular or open surgical inter-
traumatic or iatrogenic pseudoaneurysms, true vention) was similar (p0.86). Three neurologic adverse
aneurysms, intramural hematomas, and symptomatic events occurred in the thoracic endovascular aortic repair
type B aortic dissections. There has been a change of group (one case each of paraplegia, stroke, and transient
strategy for some types of thoracoabdominal aneurysms paraparesis) vs. one episode of paraparesis with medical
but they still present a challenging but luckily small treatment. It was concluded that in survivors of uncom-
group of cases. There is currently no good endoluminal plicated type B aortic dissection, elective stent-graft
alternative for type A aortic dissections or other placement does not improve 1-year survival and adverse
aneurysms of the ascending thoracic aorta. events.
Technically, stent graft implantation into the thoracic The same group recently published their long-term re-
aorta is straightforward, as it involves the insertion of a sults [11], reporting that patients with best medical treat-
simple tube graft or several tube grafts. Important land- ment but no TEVAR had an aortic-related mortality of
marks are the left subclavian artery, which can be usual- 3.6% while those with TEVAR did not show late aortic
ly covered by a stent graft in case the right vertebral events between 2 and 5 years. This difference was signif-
artery is patent; sometimes the left common carotid icant although the overall mortality was not different
160 D. Vorwerk
between the two treatment approaches. However, the au- group of 51 patients with isolated subclavian artery oc-
thors did find stabilization of the dissection in patients clusive disease treated with CSBG using polytetrafluo-
with TEVAR beyond 1 year of follow-up, while patients roethylene grafts. The mean follow-up for the PTA/stent
with optimal medical therapy had a constant progression group was 3.4 years vs. 7.7 years for the CSBG group.
over time. Paraplegia after TEVAR is a significant risk, The technical success rate for the CSBG group was 100%
requiring immediate treatment with corticosteroid and vs. 98% (119/121) for the PTA/stent group. The overall
spinal drainage. perioperative complication rate in the stent group was
15.1% (18/119: 11 minor and 7 major complications) vs.
Percutaneous Transluminal Angioplasty of Supra-aortic 5.9% (3/51: 2 phrenic nerve palsies and 1 myocardial in-
Vessel Origins farction) in the bypass group (p0.093). There was no
perioperative stroke or mortality in the CSBG group. The
Supra-aortic percutaneous transluminal angioplasty major perioperative complications in the stent group in-
(PTA) is mainly performed on the left subclavian artery, cluded four thromboembolic events, one case of conges-
but in some cases on the brachiocephalic trunk, the prox- tive heart failure, one case of reperfusion arm edema, and
imal carotid arteries, and the right subclavian artery. one of pseudoaneurysm. The 30-day patency rate was
Stenosis or occlusions may be caused mainly by athero- 100% for the bypass group and 97% (118/121) for the
sclerosis and less by inflammatory processes such as PTA/stent group. The primary patency rates at 1, 3, and
Takayasu arteritis. Indications for interventions are em- 5 years were 100%, 98%, and 96% for the CSBG group
bolic events, neurological symptoms due to steal or vs. 93%, 78%, and 70% for the stent group, respectively
malperfusion, or brachial claudication. Dilatation of the (p0.0001). Freedom from symptom recurrence was al-
left subclavian artery is relatively safe, with a low rate of so statistically superior in the bypass group vs. the stent
vertebral embolization. However stent placement is fre- group (p0.0001). The authors concluded that
quently necessary due to an insufficient post-PTA result. PTA/stenting of the subclavian artery should be the pro-
Sixt el al. [12] retrospectively analyzed 108 interven- cedure of choice for high-risk patients, but CSBG should
tions of atherosclerotic lesions in subclavian arteries or be offered to good-risk surgical candidates who may be
the brachiocephalic trunk, representing 92% of the pa- seeking a more durable procedure.
tients treated for subclavian artery obstructive disease dur-
ing a 10-year period. The primary success rate was 97%: Bronchial Artery Embolization
100% for stenoses (78/78) and 87% for total occlusions
(26/30). Treatment modalities included PTA alone (13%; Bronchial artery embolization (BAE) is not a simple inter-
n14) or stenting (87%; n90) with balloon-expandable vention for several reasons. First, the anatomy of the
(n61) or self-expanding (n17) devices, or both bronchial artery is variable and the arteries are frequently
(n12). The 1-year primary patency rate of the 97 pa- very small, which often makes their cannulation difficult.
tients eligible for follow-up was 88%: 79% for the PTA Feeders to bronchial artery bleeding sources mainly ar-
subgroup and 89% for the stenting subgroup (p0.2). teriopulmonary fistulae due to tumoral or inflammatory
Babic et al. [13] reported a reasonably high technical changes may originate not only from the bronchial artery
success of 82% and a complication rate of 7% in 56 pa- but also from many other arteries in the thorax, such as the
tients treated for chronic total occlusions of the subcla- subclavian artery, the thyrocervical trunk, or the mammar-
vian artery. Patency after 3 years was 83%. ian or costal arteries. Moreover, there is variation in col-
Van Hattum et al. [14] evaluated the results of PTA and laterals in the region, including the spinal arteries.
stent placement in isolated brachiocephalic trunk lesions Particles, glue and coils are used as embolization ma-
in 30 patients with isolated clinically significant stenoses terials but a golden rule requires that the artery is sealed
(n25) or occlusions (n5) of the brachiocephalic as close as possible to the bleeding source, to avoid re-
artery. The initial technical success rate was 83% (occlu- currence via collateral feeders shortly thereafter.
sions, 60%; stenoses, 88%), and the clinical success rate Benign reasons for bronchial bleeding include
81%. Two patients had major complications, and four ex- bronchiectasis, chronic bronchitis, aspergillosis, tubercu-
perienced minor complications. At a median follow-up of losis, pneumonia, and abscesses. Malignant sources are
24 months (4 weeks to 92 months), the primary clinical predominantly bronchial cancers.
patency rate was 79% (95% confidence interval (CI): The results of percutaneous intervention are relatively
57%, 104%), with 83% (95% CI: 60%, 105%) for arter- good. Kato et al. [16] treated 101 patients and achieved a
ies with stents and 67% (95% CI: 13%, 120%) for those technical success of 100%, a 1-year success of 77.7%, and
without stents (p0.11). The primary technical patency a 5-year success of 62.5%. A better outcome was observed
rate was 50% (95% CI: 24%, 76%). in patients with tuberculosis and bronchitis and a less
AbuRahma et al. [15] compared the results of a large favorable one in those with pneumonia and abscesses.
series of PTA/stenting procedures in the subclavian artery Lee et al. [17] described their experience with BAE
with the results of a series of carotid-subclavian bypass in 54 patients that were treated for massive hemoptysis
grafts (CSBG). Subclavian artery PTA/stenting was in a 5-year period. The underlying pathologies included
performed in 121 patients, who were compared with a bronchiectasis (n31), active tuberculosis (n9),
Interventional Techniques in the Thorax of Adults 161
pneumoconiosis (n3), lung cancer (n2), and pul- Recently, an Italian group described a a single cohort
monary angiodysplasia (n1). Surgery was considered if study in which a combination of segmental pulmonary
the patient had acceptable pulmonary reserve and a bleed- artery embolization and RFA was used in 17 patients
ing source was clearly identified. If the patient was not with 20 nodules. The technical success rate was good
considered fit for surgery, BAE was attempted. Hemopty- and a clinical complete response was achieved in 65%
sis ceased with conservative management only in seven [22].
patients (13%). In the 27 (50%) patients who underwent
surgical resection, the procedures included lobectomy Pulmonary Artery Embolectomy
(n21), bilobectomy (n4), and pneumonectomy (n2).
In-hospital mortality after surgery was 15%. Postoperative Treatment of acute symptomatic PE by catheter-directed
morbidity occurred in eight patients, including prolonged methods has long been a goal of interventional radiolo-
ventilatory support, bronchopleural fistulae, empyema, gists but its realization has been hampered by two prob-
and myocardial infarction. BAE was also used in 21 pa- lems in particular. First, massive PE may lead to relative-
tients not eligible for surgery. The procedure was success- ly mild symptoms not necessitating aggressive treatment.
ful in 17 patients, without any complications. Second, intravenous thrombolysis is usually effective in
Woo et al. [18] retrospectively compared BAE with most patients. Moreover, those patients who are candi-
particles vs. bucrylate glue. In a considerably large co- dates for a percutaneous approach die early or are in very
hort (293 procedures with particles, 113 with glue), they poor clinical condition, as most are hemodynamically un-
achieved a technical success of 94% vs. 97%. The com- stable and require emergency treatment but are not trans-
plication rate was similar but freedom from rebleeding portable. This logistical aspect has prevented the wide-
was better: at 1, 3, and 5 years, 88%, 85%, and 83% with spread use of catheter-based interventions. Nonetheless,
glue vs. with 77%, 68%, and 66% particles. among the small cohorts of patients treated with this ap-
Witt et al. [19] performed BAE using platinum coils proach, the results have been encouraging.
with Dacron fibers in 30 consecutive patients with bleed- Kuo et al. [23] retrospectively analyzed 70 consecutive
ing from bronchial carcinoma. The aim of that study was patients over a 10-year period with suspected acute PE
to compare immediate results with respect to bleeding who had been referred for pulmonary angiography and/or
cessation as well as recurrence, and survival rates with intervention. The criteria for study inclusion were pa-
BAE vs. conservative management. Active bleeding tients who received catheter directed intervention (CDI)
stopped immediately in all patients. A comparison of the due to angiographically confirmed massive pulmonary
BAE group with the controls showed that the cessation of embolism and hemodynamic shock (shock index 0.9).
first-time hemoptysis was 100% in the BAE group vs. CDI involved suction embolectomy and fragmentation
93% in the non-BAE group. The rates of bleeding recur- with or without catheter thrombolysis. Twelve patients
rence were similar in the two groups (BAE 50% vs. non- were treated with CDI. Seven patients (58%) were re-
BAE 47%). In case of recurrent bleeding, repeated BAE ferred for CDI after failing to respond to systemic infu-
led to a definite cessation of pulmonary hemorrhage in sion with 100 mg of tissue plasminogen activator, and
every case. By contrast, all patients with recurrent he- five patients (42%) had contraindications to systemic
moptysis without a repeated BAE (8 patients, 27%) and thrombolysis. Catheter-directed fragmentation and em-
all patients with bleeding recurrence who were in the bolectomy were performed in all patients (100%) and
non-BAE group died from pulmonary hemorrhage (8 pa- catheter-guided thrombolysis in eight patients (67%).
tients, 53%). The mean survival time of the BAE group Technical success was achieved in all 12 cases (100%).
was significantly longer than that of the non-BAE group There were no major procedural complications. Signifi-
The authors therefore concluded that BAE was beneficial cant hemodynamic improvement (shock index: 0.9)
in patients tumoral pulmonary bleeding. was reported in 10 of the 12 patients (83%). The remain-
ing two patients (17%) died secondary to cardiac arrest
Venous Interventions within 24 h. The authors concluded that CDI is poten-
tially a life-saving treatment for patients who have not re-
Pulmonary Artery Embolization sponded to or cannot tolerate systemic thrombolysis.
Lin et al. [24] evaluated the treatment outcome in pa-
Pulmonary artery embolization is an infrequently used tients with massive PE who were treated with either ultra-
tool to treat traumatic damage to the pulmonary artery in- sound-accelerated thrombolysis using the EkoSonic En-
duced by catheters or trauma and to treat pulmonary dovascular System (EKOS; EKOS; Bothell, WA) or
artery-vein shunts. In single institutions it has been mod- catheter-directed thrombolysis (CDT). Twenty-five pa-
ified as chemoembolization and used in the palliative tients underwent 33 CDI for massive pulmonary embolism
treatment of patients with lung cancer [20]. There are no during the study period. Among them, EKOS was per-
clinical randomized data on pulmonary artery emboliza- formed in 15 (45%) and CDT in 18 (55%). In the EKOS
tion, only an experimental study in rats [21] that showed group, complete thrombus removal was achieved in 100%
a significant reduction of tumor size compared to intra- cases. In the CDT cohort, complete or partial thrombus
venous chemotherapy. removal was accomplished in 7 (50%) and 2 (14%) cases,
162 D. Vorwerk
respectively. Comparing treatment success based on Benign Venous Stenoses and Occlusions
thrombus removal, EKOS resulted in an improved treat-
ment outcome compared with the CDT group (p0.02). Partial venous obstruction usually does not become clin-
The mean time of thrombolysis was 17.45.23 h and ically evident as long as there is a collateral network. In
25.37.35 h in the EKOS and CDT groups, respectively patients with dialysis fistulas, however, high venous flow
(p0.03). The mortality rate was 9.1% and 14.2% (not sig- may result in severe venous congestions in case the ipsi-
nificant). Treatment-related hemorrhagic complication lateral venous outflow is obstructed. These obstructions
rates in the EKOS and CDT group were 0% and 21.4%, re- are frequently sequelae of an earlier placement of venous
spectively (p0.02). A significant reduction in Miller dialysis catheters into the subclavian vein. However,
scores was noted in both groups following catheter-based since over time the policy has changed to a jugular access
interventions. for those catheters, there are fewer central venous outflow
problems.
Malignant Venous Obstruction (Superior Vena Cava Syndrome) Treatment of benign venous stenoses is not simple, as
balloon angioplasty usually fails to gain long-term im-
Another percutaneous procedure that has become in- provement. Stent placement yields excellent short-term
creasingly popular is the placement of metallic stents to results but is burdened by a high rate of recurrent
treat superior vena cava syndrome (SVCS). Unlike emer- stenoses due to massive neointimal tissue growth within
gency radiation, stenting gives rapid relief of symptoms the stents, thus requiring frequent reinterventions. Poly-
within a few hours or even immediately. The technique is mer-covered stent grafts may be more promising in pre-
relatively simple and can be achieved from either a venting recurrent stenoses but the results of larger stud-
brachial or a transfemoral approach. If the obstruction is ies are still pending.
associated with thrombus, stenting can be combined with Haage et al. [27] analyzed the effectiveness of stent
thrombolysis. placement as the primary treatment for central venous
Lanciego et al. [25] used stent placement as the first- obstruction in 50 patients undergoing hemodialysis. Stent
choice treatment for the relief of symptoms in 52 can- deployment (n57, Wallstent stents) was successful in all
cer patients who were diagnosed with SVCS confirmed patients, with early rethrombosis (within 1 week) noted
by cavography or phlebography. Wallstent prostheses in only one patient (2%). Among the 73 episodes of re-
(n73) were inserted in all patients. A single stent was obstruction, 54 (74%) were treated percutaneously with
sufficient in 37 patients, two stents were required in 11, angioplasty alone and 19 (26%) necessitated additional
three stents in 2, and four stents in another 2 patients. stent placement. The 3-, 6-, 12-, and 24-month primary
Contraindications for the procedure were severe car- patency rates were 92%, 84%, 56%, and 28%, respec-
diopathy or coagulopathy. Resolution of symptoms was tively. Cumulative overall stent patency was 97% after 6
achieved in all patients within 72 h. At follow-up, six and 12 months, 89% after 24 months, and 81% after 36
obstructions, one partial migration to the right atrium, and 48 months.
two incorrect placements, and four stent shortenings Bakken et al. [28] compared the outcomes of primary
were noted. All were successfully resolved by repeat angioplasty (PTA) vs. primary stenting (PTS) in a dialy-
stenting. Symptom-free survival ranged from 2 days to sis access population. PTS was used to treat 26 patients
17 months (mean: 6.4 months). The authors concluded (26 stenoses) and PTA in 47 patients with 49 central ve-
that the Wallstent vascular endoprosthesis is an effective nous stenoses. The PTS group underwent 71 percutaneous
initial treatment in patients with SVCS of neoplastic interventions (average, 2.72.4 interventions per steno-
origin. sis), and the PTA group 98 interventions (average,
Nagata et al. [26] retrospectively investigated the util- 2.01.6 interventions per stenosis). Primary patency was
ity of metallic stent placement, mainly the spiral Z-stent equivalent between the two groups by Kaplan-Meier
(S-Z-stent), for the treatment of malignant obstruction of analysis, with 30-day rates of 76% for both and 12-month
the SVC in 71 patients with SVCS. The 71 patients who rates of 29% for the PTA group and 21% for the PTS
underwent stent placement were followed until death. group (p0.48). Assisted primary patency was also equiv-
The technical success rate was 100%, the initial clinical alent (p0.08), with a 30-day patency rate of 81% and a
success rate was 87% (62/71), the primary clinical pa- 12-month rate of 73% for the PTA group, vs. 84% at 30
tency rate was 88% (57/65), and the secondary clinical days and 46% at 12 months in the PTS group. Ipsilateral
patency rate was 95% (62/65). Stent obstruction oc- hemodialysis access survival was equivalent in the two
curred at a rate of 12% (8/65), requiring secondary stent- groups. The authors concluded that PTS does not improve
ing. The survival times of the 57 patients in whom there the patency rates compared to PTA and does not add to
was no SVCS recurrence ranged from 1 week to 29 the longevity of ipsilateral hemodialysis access sites.
months (mean: 5.4 months). The S-Z-stent was consid- Jones et al. [29] used a stent graft (Viabahn) to treat 30
ered to be suitable for the treatment of malignant ob- patients with central venous stenosis. Surveillance was
struction of the SVC. Unilateral stent placement was ef- carried out at 3, 6, 9, 12, 18, and 24 months using diag-
fective for relief of SVCS in patients with bilateral bra- nostic fistulography. The technical success rate was 100%.
chiocephalic vein obstruction. Primary patency rates were 97%, 81%, 67%, and 45%;
Interventional Techniques in the Thorax of Adults 163
primary assisted patency rates were 100%, 100%, 80%, and automated cutting needles using a coaxial technique. Car-
and 75% at 3, 6, 12, and 24 months, respectively. Twelve diovasc Intervent Radiol 23:266-272.
2. Richardson CM, Pointon KS, Manhire AR, Macfarlane JT
patients required further stent-grafts to maintain patency. (2002) Percutaneous lung biopsies: a survey of UK practice
Ferguson et al. [30] used stents in six patients with fi- based on 5444 biopsies. Br. J Radiol 75:731-735.
brosing mediastinitis with venous obstruction. Four pa- 3. Chojniak R, Isberner RK, Viana LM et al (2006) Computed to-
tients were treated with intravascular stents placed percu- mography guided needle biopsy: experience from 1,300 pro-
taneously. One patient underwent surgical intravascular cedures. Sao Paulo Med J 124:10-14
4. Poretti FP, Brunner E, Vorwerk D (2002) Simple localization
stent placement and one patient declined surgical therapy. of peripheral pulmonary nodules - CT-guided percutaneous
The right pulmonary artery was treated in three patients, hook-wire localization. Rofo Fortschr Geb Rontgenstr Neuen
the SVC in one patient, and three pulmonary veins in the Bildgeb Verfahr 174:202-207.
fifth patient. Each intervention resulted in hemodynamic 5. de Bare T., Farouil G, Deschamps F (2013) Lung Cancer Ab-
improvement with subsequent clinical improvement. Rein- lation: What is the evidence? Semin Intervent Radiol 30:
151-156.
tervention was required within 12 months in two of the 6. Gillams A, Khan Z, Osborn P, Lees W (2013) Survival after
four percutaneously treated patients. The authors conclud- radiofrequency ablation in 122 patients with inoperable colo-
ed that stents are an effective option in improving short- rectal lung metastases. Cardiovasc Intervent Radiol 36:
term vascular patency. In-stent stenosis was a frequent 724-730.
complication in patients with fibrosing mediastinitis. 7. Schlijper RC, Grutters JP, Houben R et al (2014) What to
choose as radical local treatment for lung metastases from co-
Riley et al. [31] investigated the treatment of SVCS as lo-rectal cancer surgery or radiofrequency ablation? Cancer
a complication of pacemaker implantation, carrying out a Treat Rev 40:60-67.
Pubmed search to identify cases of symptomatic SVCS 8. Tramontano AC, Cipriano LE, Kong CY (2013) Microsimula-
that developed following the implantation of permanent tion model predicts survival benefit of radiofrequency ablation
and stereotactic body radiotherapy vs. radiotherapy for treating
pacemakers or implantable cardioverter defibrillators. Pa- inoperable stage I non-small cell lung cancer. AJR Am J
tients in the study had been treated with one of five differ- Roentgenol 200:1020-1027.
ent modalities: anticoagulation, thrombolysis, venoplasty, 9. Matsumura JS, Cambria RP, Dake MD; TX2 Clinical Trial In-
stenting, and surgical reconstruction. The literature search vestigators (2008) International controlled clinical trial of tho-
identified 74 publications reporting 104 eligible cases in racic endovascular aneurysm repair with the Zenith TX2 en-
dovascular graft: 1-year results. J Vasc Surg 47:247-257.
which SVCS presented at a median of 48 months after de- 10. Nienaber CA, Kische S, Akin I et al (2010) Strategies for sub-
vice implantation. Anticoagulation, thrombolysis, and acute/chronic type B aortic dissection: the Investigation Of
venoplasty alone were all associated with high recurrence Stent Grafts in Patients with type B Aortic Dissection (IN-
rates. Surgery and stenting were more successful, with re- STEAD) trial 1-year outcome. J Thorac Cardiovasc Surg 140(6
currence rates of 12% and 5% over a median follow-up of Suppl):S101-108.
11. Nienaber CA, Kische S, Rousseau H et al; INSTEAD-XL tri-
16 (range: 2179) and 9.5 (range: 260) months, respec- al (2013) Endovascular repair of type B aortic dissection: long-
tively. Based on our own observations, over the last 40 term results of the randomized investigation of stent grafts in
years conservative treatments have been replaced by surgi- aortic dissection trial. Circ Cardiovasc Interv 6:407-416.
cal reconstruction, and most recently by stenting, as the 12. Sixt S, Rastan A, Schwarzwlder U et al (2009) Results after
most common therapeutic modality employed. balloon angioplasty or stenting of atherosclerotic subclavian
artery obstruction. Catheter Cardiovasc Interv 73:395-403.
There is some evidence [32] that covered stents with 13. Babic S, Sagic D, Radak D et al (2012) Initial and long-term
an expanded polytetrafluoroethylene coating are more ef- results of endovascular therapy for chronic total occlusion of
fective in preventing restenosis than PTA alone, but bare the subclavian artery. Cardiovasc Intervent Radiol 35:
stents have yet to be compared with stent grafts. Limita- 255-262.
14. van Hattum ES, de Vries JP, Lalezari F et al (2007) Angio-
tions in size and diameter and the problem of overriding plasty with or without stent placement in the brachiocephalic
collateral veins restrict the wide-spread use of stent grafts artery: feasible and durable? A retrospective cohort study. J
in the thoracic venous circulation. Vasc Interv Radiol 18:1088-1093.
15. AbuRahma AF, Bates MC, Stone PA et al (2007) Angioplasty
and stenting vs. carotid-subclavian bypass for the treatment of
isolated subclavian artery disease. J Endovasc Ther 14:
Conclusion 698-704.
16. Kato A, Kudo S, Matsumoto K et al (2000) Bronchial artery
The great variation of vascular and nonvascular proce- embolization for hemoptysis due to benign diseases: immedi-
dures in the thorax offers a vast opportunity for the use of ate and long-term results. Cardiovasc Intervent Radiol 23:
351-357.
many different therapeutic techniques, some rarely per- 17. Lee TW, Wan S, Choy DK (2000) Management of massive he-
formed or restricted to just a few centers, and others with moptysis: a single institution experience. Ann Thorac Cardio-
a relevance and general importance for many patients. vasc Surg 6:232-235.
18. Woo S, Yoon CJ, Chung JW et al (2013) Bronchial artery em-
bolization to control hemoptysis: comparison of N-butyl-2-
cyanoacrylate and polyvinyl alcohol particles. Radiology
References 269:594-602.
19. Witt Ch, Schmidt B, Geisler A et al (2000) Value of bronchial
1. Laurent F, Latrabe V, Vergier B, Michel P (2000) Percutaneous artery embolisation with platinum coils in tumorous pul-
CT-guided biopsy of the lung: comparison between aspiration monary bleeding. Eur J Cancer 36:1949-1954.
164 D. Vorwerk
20. Vogl TJ, Wetter A, Lindemayr S, Zangos S (2005) Treatment of treatment of a malignant obstruction of the superior vena ca-
unresectable lung metastases with transpulmonary chemoem- va. Cardiovasc Intervent Radiol 30:959-967.
bolization: preliminary experience. Radiology 234:917-922. 27. Haage P, Vorwerk D, Piroth W et al (1999) Treatment of he-
21. Schneider P, Kampfer S, Loddenkemper C et al (2002) modialysis-related central venous stenosis or occlusion: results
Chemoembolization of the lung improves tumor control in a of primary Wallstent placement and follow-up in 50 patients.
rat model. Clin Cancer Res 8:2463-2468. Radiology 212:175-180.
22. Gadaleta CD, Solbiati L, Mattioli V et al (2013) Unresectable 28. Bakken AM, Protack CD, Saad WE et al (2007) Long-term
lung malignancy: combination therapy with segmental pul- outcomes of primary angioplasty and primary stenting of cen-
monary arterial chemoembolization with drug-eluting micros- tral venous stenosis in hemodialysis patients. J Vasc Surg
pheres and radiofrequency ablation in 17 patients. Radiology 45:776-783.
267:627-637. 29. Jones RG, Willis AP, Jones C et al (2011) Long-term results of
23. Kuo WT, van den Bosch MA, Hofmann LV et al (2008) stent-graft placement to treat central venous stenosis and oc-
Catheter-directed embolectomy, fragmentation, and thrombol- clusion in hemodialysis patients with arteriovenous fistulas. J
ysis for the treatment of massive pulmonary embolism after Vasc Interv Radiol 22:1240-1245.
failure of systemic thrombolysis. Chest 134:250-254. 30. Ferguson ME, Cabalka AK, Cetta F, Hagler DJ (2010) Results
24. Lin PH, Annambhotla S, Bechara CF et al (2009) Comparison of intravascular stent placement for fibrosing mediastinitis.
of percutaneous ultrasound-accelerated thrombolysis vs. Congenit Heart Dis 5:124-133.
catheter-directed thrombolysis in patients with acute massive 31. Riley RF, Petersen SE, Ferguson JD, Bashir Y (2010) Manag-
pulmonary embolism. Vascular 17 Suppl 3:S137-147. ing superior vena cava syndrome as a complication of pace-
25. Lanciego C, Chacon JL, Julian A et al (2001) Stenting as first maker implantation: a pooled analysis of clinical practice. Pac-
option for endovascular treatment of malignant superior vena ing Clin Electrophysiol 33:420-425.
Cavasyndrome. AJR Am J Roentgenol 177:585-593. 32. Haskal ZJ, Trerotola S, Dolmatch B (2010) Stent graft vs.
26. Nagata T, Makutani S, Uchida H et al (2007) Follow-up results balloon angioplasty for failing dialysis-access grafts. N Engl J
of 71 patients undergoing metallic stent placement for the Med 362:494-503.
IDKD 2015-2018
Percutaneous Transthoracic Needle Biopsy thoracentesis, an aliquot of collected fluid is sent to labora-
tory analysis to try to more specifically identify the cause
Transthoracic biopsy of pulmonary and mediastinal le- of pleural fluid accumulation. In therapeutic thoracentesis,
sions is a commonly performed diagnostic procedure. It fluid is withdrawn to relieve symptoms such as dyspnea
is used to confirm malignancy in anticipation of appro- (most often) or hemodynamic compromise.
priate therapy or to spare patients with benign disease The benefits of real-time ultrasonography in improv-
from unnecessary surgery. In the following, space limita- ing the success and safety of thoracentesis (as well as
tions prohibit a detailed discussion of techniques and chest tube placement) have been clearly documented,
complications; instead, the reader is referred to several such that its use is now strongly recommended [13, 14].
reviews of this topic for details [1-5]. Current areas of Ultrasound evaluation confirms the presence and location
interest include the use of alternate image-guidance of fluid, improves the characterization of fluid collec-
methods, such as real-time ultrasound for peripheral lung tions, and helps to avoid puncture of the lungs or other
lesions and cone-beam computed tomography (CT) organs during the procedure.
[6, 7]; integrated navigational systems for guidance [8]; Hemorrhagic complications may occur as a result of
methods to decrease the rate of post-procedural pneu- thoracentesis, although the degree of risk relative to var-
mothorax [9]: and the procurement of tissue for genomic ious coagulation parameters is controversial [13]. Since
analysis [10]. the intercostal neurovascular bundle is typically on the
undersurface of the rib, it is prudent to try to enter the
pleural space above the rib. However, there is consider-
Lung Tumor Localization able variability in the position of the intercostal artery
and its branches based on age and relative thoracic posi-
Pulmonary nodule localization may be requested in tion. The British Thoracic Society therefore recommends
patients scheduled for thoracoscopic resection of small, that unless fluid is loculated posteriorly, thoracentesis
peripheral pulmonary nodules, in order to confidently lo- should be performed in the triangle of safety, bordered
calize these lesions [11]. Several methods of localization anteriorly by the lateral edge of the pectoralis major mus-
have been described, including hook-wire localization, cle, laterally by the lateral edge of the latissimus dorsi, in-
image-guided dye injection, the placement of fiducials or feriorly by the fifth intercostal space, and superiorly by
clips, intraoperative ultrasound of the deflated lung, and the base of the axilla [14].
radionuclide injection. A method using microcoil deploy- Pneumothorax may also develop during thoracentesis,
ment through the nodule and trailing to the pleural surface with lower rates generally associated with the use of ul-
has been described that appears to have several advantages trasound guidance. In pneumothorax resulting simply
compared to previously described methods [12]. from the leakage of air via the needle or catheter there is
no clinical significance. However, pneumothorax may al-
so develop due to direct injury to the visceral pleural sur-
Drainage of Thoracic Fluid Collections face, although this should be largely avoidable with good
technique and imaging guidance. Finally, air may leak in-
Thoracentesis to the lung as an ex-vacuo phenomenon in cases of
trapped lung. In these instances, as fluid is aspirated the
One of the most common image-guided interventions in lung fails to expand, and air may leak from the lung as
the thorax is thoracentesis. Thoracentesis is performed for the intrapleural pressures decrease. This has been report-
two main, sometimes overlapping indications: In diagnostic ed as the primary cause of pneumothorax following
therapeutic thoracentesis [15], but patients with these Pleuroperitoneal shunts, placed surgically or percuta-
types of pneumothorax rarely require therapy. neously, allow patients to actively treat their dyspnea on
Another potential complication of therapeutic thora- their own. The disadvantages include obstruction or in-
centesis is re-expansion pulmonary edema. This compli- fection of the device and the need to pump the device
cation is rare, generally mild, and responds to conserva- multiple times per day.
tive measures including oxygen administration and di- Another option, unavailable as a percutaneous tech-
uresis. It seems more common in younger patients, those nique, is surgical stripping of the pleura. Although very
with more chronic effusions, and those in whom large effective, this method carries high morbidity and an ap-
amounts of fluid are removed [16]. The latter considera- proximately 10% mortality rate.
tion has led to recommendations of maximal fluid
amounts during thoracentesis (usually 10001500 mL). Empyema and Parapneumonic Effusion
However, it has also been argued that larger volumes of
pleural fluid can be safely withdrawn as long as the pa- Pleural effusions associated with infection (parapneu-
tient does not develop vague chest discomfort or exces- monic effusion and empyema) tend to evolve in stages. In
sively negative (20 cm H2O) intrapleural pressures, as the initial, exudative stage, the fluid is thin and sterile and
measured by pleural manometry [17]. associated with a low white blood cell (WBC) count. In
the fibropurulent stage, bacteria invade the pleural space,
Malignant Pleural Effusions the WBC count increases (an empyema is defined as
grossly purulent fluid), and fibrin begins to be deposited
Malignant pleural effusions are the second most common on pleural surfaces and as septae within the fluid. Final-
cause of exudative pleural effusions and the most com- ly, in the organization stage, a thick peel of fibrous tissue
mon cause of such effusions in patients over 60 years of encapsulates the pleural space. Chest-tube drainage is
age. While malignant pleural effusions can occur with most effective in the earlier stages but ineffective in the
virtually any primary tumor, they are most often the re- organization stage. Intuition would suggest that elimina-
sult of lung cancer, breast cancer, or lymphoma. Patients tion of the fibrin while fluid is being drained would fa-
tend to present with dyspnea (over 50% of patients), chest cilitate drainage. Indeed, encouraging studies in this re-
pain (usually dull in nature), cough, and generalized sys- gard appeared as early as the 1940s [21] although inter-
temic symptoms that reflect advanced metastatic disease. est waned partly due to allergic reactions to streptokinase,
There are many techniques for the management of ma- the primary fibrinolytic agent used at that time. A recent
lignant effusions [18-20], depending on the type and multicenter trial [22] compared patients with pleural in-
stage of malignancy. In advanced malignancy, the goal is fection (purulent pleural fluid, pleural fluid with pH 7.2,
palliative and is directed primarily toward relieving the and signs of infection, or proven bacterial infection of the
dyspnea, with the exception of the pharmacologic man- pleural space) in a double-blind trial. The 454 patients
agement of pain. Indeed, opiate therapy, in combination were randomly assigned to treatment with intrapleural
with oxygen, is often underutilized in patients with a streptokinase or placebo. There were no significant dif-
short life expectancy but may be quite helpful in reliev- ferences between the groups with regard to endpoints of
ing dyspnea. Therapeutic thoracentesis is a first step in death or requirement for surgery. More recent studies
more definitive therapy as it is important to determine have also looked at deoxyribonucleases in combination
whether fluid removal does, in fact, result in improve- with fibrinolytic agents [23]. In one such study, 210 pa-
ment in dyspnea. However, thoracentesis rarely provided tients with pleural infection were randomly assigned to
lasting control of dyspnea, with fluid reaccumulating in receive one of four treatments: (1) double placebo; (2) in-
98100% of patients within 30 days. trapleural tissue plasminogen activator (t-PA) and a
Indwelling tunneled catheters, placed with imaging DNase; (3) t-PA and placebo; (4) DNase and placebo.
guidance, are another management option in patients with The combined treatment group was associated with im-
malignant pleural effusions. These catheters are well-ac- proved fluid drainage, reduced frequency of surgical re-
cepted by patients, provide good palliation of dyspnea, and ferral, and a shorter hospital stay compared to the place-
result in auto-pleurodesis in over 50% and as high as bo group. The use of either agent alone did not differ sig-
70% of patients, thus allowing eventual catheter removal. nificantly from placebo. Thus, the efficacy of these
Tube thoracostomy can be combined with chemical agents remains controversial [22-24].
pleurodesis, with the aim of causing inflammation and
the adherence of the visceral and parietal pleural sur-
faces. Talc is a commonly used agent, although its opti- Thoracic Ablation Procedures
mal dosage and formulation remain uncertain; in addi-
tion, patients may require prolonged hospitalization dur- While surgical resection remains the standard of care for
ing this form of therapy. Pleurodesis can also be per- early-stage lung cancer, a variety of ablative techniques
formed thoracoscopically; while more invasive, this ap- have shown promise in the treatment of select primary and
proach can provide excellent distribution of the scleros- metastatic pulmonary malignancies [25, 26]. The largest
ing agents and be combined with mechanical abrasion. reported experience is with radiofrequency ablation,
Nonvascular Interventional Radiology of the Thorax 167
although experience with microwave ablation, cryoabla- 11. Tamara M, Oda M, Fujimori H et al (2010) New indication for
tion, and other techniques is accumulating as well. Le- preoperative marking of small peripheral pulmonary nodules
in thoracoscopic surgery. Interact Cardiovasc Thorac Surg
sions most amenable to successful ablation are small 11:590-593.
(3 cm), solitary, and remote from critical structures. 12. Mayo JR, Clifton JC, Powell TI et al (2009) Lung nodules: CT-
Complications are similar to, although greater in frequen- guided placement of microcoils to direct video-assisted thora-
cy than, percutaneous lung biopsy. Few studies have, as coscopic surgical resection. Radiology 250:576-578.
yet, compared either the various ablative techniques or 13. Sachdeva A, Shepherd RW, Lee HJ (2013) Thoracentesis and
thoracic ultrasound: state of the art. Clin Chest Med 34:
ablation with surgery vs. stereotactic radiation therapy. 1-9.
14. Havelock T, Teoh R, Laws D et al (2010) Pleural procedures
and thoracic ultrasound: British Thoracic Society pleural dis-
References ease guideline 2010. Thorax 66:828-829.
15. Heidecker J, Huggins JT, Sahn SA, Doelken P (2006) Patho-
1. Winer-Muram HT (2006) The solitary pulmonary nodule. physiology of pneumothorax following ultrasound-guided tho-
Radiology 239:34-49. racentesis. Chest 130:1173-1184.
2. Manhire A, Charig M, Clelland C et al (2003) Guidelines for 16. Echevarria C, Thowmey D, Dunning J, Chanda B (2008) Does
radiologically guided lung biopsy. Thorax 58:920-936. re-expansion pulmonary oedema exist? Interact Cardiovasc
3. Ahrar K, Wallace M, Javadi S, Gupta S (2008) Mediastinal, Thorac Surg 7:485-489.
hilar, and pleural image-guided biopsy: current practice and 17. Feller-Kopman D, Berkowitz D, Boiselle P, Ernst A (2007)
techniques. Semin Respir Crit Care Med 29:350-360. Large-volume thoracentesis and the risk of reexpansion pul-
4. Moore EH (2008) Percutaneous lung biopsy: an ordering clin- monary edema. Ann Thorac Surg 8:1656-1661.
icians guide to current practice. Semin Respir Crit Care Med 18. Thomas JM, Musani AI (2013). Malignant pleural effusions: a
29:323-334. review. Clin Chest Med 34:459-471.
5. Wu CC, Maher MM, Shepard JA (2011) Complications of CT- 19. Davies HE, Lee YCG (2013) Management of malignant pleur-
guided percutaneous needle biopsy of the chest: prevention al effusions: questions that need answers. Curr Opin Pulm
and management. AJR Am J Roentgenol 196: W678-682. Med 19:374-379.
6. Moreira BL, Guimaraes MD, de Oliveira AD et al (2014) Val- 20. Kastelik JA (2013) Management of malignant pleural effusion.
ue of ultrasound in the imaging-guided transthoracic biopsy of Lung 191:165-175.
lung lesions. Ann Thorac Surg 97:1795-1797. 21. Tillett WS, Sherry S (1949) The effect in patients with strep-
7. Lee SM, Park CM, Lee KH et al (2014) C-arm cone-beam CT- tococcal fibrinolysis (streptokinase) and streptococcal des-
guided percutaneous transthoracic needle biopsy of lung nod- oxyribonuclease on fibrinous, purulent, and sanguinous pleur-
ules: clinical experience in 1108 patients. Radiology 271:291- al exudations. J Clin Invest 28:173-190.
300. 22. Maskell NA, Davies CWH, Nunn AJ et al (2005) U.K. con-
8. Grasso RF, Faiella E, Luppi G et al (2013) Percutaneous lung trolled trial of intrapleural streptokinase for pleural infection.
biopsy: comparison between an augmented reality CT naviga- N Engl J Med 352:865-874.
tion system and standard CT-guided technique. In J Comput 23. Rahman NM, Maskell NA, West A et al (2011) Intrapleural use
Assist Radiol & Surg 8:837-848. of tissue plasminogen activator and Dnase in pleural infection.
9. Zaetta JM, Licht MO, Fisher JS et al (2010). A lung biopsy N Engl J Med 365:518-526.
tract plug for reduction of postbiopsy pneumothorax and oth- 24. Rahman NM (2012) Intrapleural agents for pleural infection:
er complications: results of a prospective, multicenter, ran- fibrinolytics and beyond. Curr Opin Pulm Med 18:326-332
domized, controlled clinical study. J Vasc Interv Radiol 25. De Baere T (2011) Lung tumor radiofrequency ablation: where
21:1235-1243. do we stand? Cardovasc Intervent Radiol 34:241-251.
10. Marshall D, LaBerge JM, Firetag B et al (2013) The changing 26. Charmarthy MR, Gupta M, Hughes TW et al (2014) Image-
face of percutaneous image-guided biopsy: molecular profil- guided percutaneous ablation of lung malignancies: a mini-
ing and genomic analysis in current practice. J Vasc Interv Ra- mally invasive alternative for nonsurgical patients or unre-
diol 24:1094-1103. sectable tumors. J Bronchol Intervent Pulmonol 21:68-81.
IDKD 2015-2018
Positioning of the Arms ners. These filters modify beam strength by strongly at-
tenuating the peripheral projections and weakly attenuat-
In thoracic CT, the patients arms should be raised over ing central ones. This evens the X-ray fluence at the de-
his or her head to achieve a dose-efficient thoracic CT tector and reduces the radiation dose to less-attenuating
scan with acceptable image quality. The increased atten- peripheral structures. Incorrect centering defeats the pur-
uation by the arms in the x-axis forces the automatic tube pose of the bowtie filter, leading to an increased radiation
current modulation to increase the radiation dose to dose to peripheral structures and a decreased image qual-
maintain constant image quality, thus increasing the radi- ity [10]. Clinical studies have found that up to 95% of pa-
ation dose to most organs. Furthermore, with the arms tients receiving a thoracic or abdominal CT scan are not
next to the thorax, beam-hardening artifacts are in- centered properly in the vertical plane [10], with an aver-
creased, degrading image quality (Fig. 2). Compared to age error of 2.6 cm. Li et al. quantified this effect, there-
arm-down positioning, placing the arms above the head by demonstrating that a vertical mis-centering of 3 cm re-
increases subjective and objective image quality [8] and sults in an increase in the peripheral radiation dose of up
leads to a dose reduction of up to 45% in CT of the tho- to 18%, with a 6% increase in image noise [10]. Thus,
rax and abdomen [9]. If a patient is unable to lift one arm, technicians should be trained to meticulously center pa-
the technician should still position the other arm above tients such that the bowtie filters are utilized optimally.
the head.
Patient Positioning in the Isocenter of the CT Scanner Technical Advances for Dose Reduction
The detrimental radiation dose effect of mis-centering pa- In the last 10 years, CT manufacturers have introduced a
tients within the scanner gantry is often not appreciated by number of effective tools that reduce the radiation dose
technicians and radiologists. The increased radiation dose while maintaining or improving image quality. The clinical
associated with mis-centering arises from the pre-patient consequence of impaired image quality associated with
X-ray beam bowtie filters installed in all modern CT scan- dose reduction has been demonstrated in experimental
170 A. Euler, Z. Szucs-Farkas, J.R. Mayo, S.T. Schindera
a b
Fig. 2 a, b. Localizer radiograph and transverse CT slice of the lung base in a patient who underwent two CT examinations, one in which
the arms were raised over the head (a) and the other with the arms placed next to the thorax due to immobility (b). The second CT scan
resulted in an increase of beam hardening artifacts and reduced image quality in the lung bases
studies using computer-simulated low-dose scans [11]. In cal task specified on the clinical requisition. Therefore,
the experimental design, intra-observer agreement was the effectiveness of ATCM dose reduction is dependent
lower when noisy, low-dose rather than higher dose ex- on the radiologist. Each manufacturer uses a proprietary
ams were interpreted [12]. The authors concluded that op- method to define the desired image quality (e.g., refer-
timal dose reduction should be achieved without sacrific- ence mA by Siemens, noise index by GE, standard devi-
ing image quality. Since CT manufacturers use diverse ation by Toshiba, reference images by Philips). In chest
technical approaches for dose reduction, there are some CT, a dose reduction of 2026% can be achieved in aver-
differences in the technical implementations and their re- age-size patients using ATCM rather than fixed tube cur-
sultant effectiveness. rent [16]. However, in very large patients, ATCM can re-
sult in increased radiation doses [17] because the system
Automatic Tube Current Modulation attempts to maintain uniform image quality almost inde-
pendently of body size. Thus, the operator needs to be
Since radiation dose is linearly related to tube current at aware of this important caveat and to adjust the image
constant kVp, a decrease in the tube current results in quality level in obese patients.
dose reduction. Although image noise is impacted by A few years ago, organ-based tube current modulation
dose reduction, there is no impact on the CT numbers was introduced by one manufacturer with the goal of re-
(Hounsfield units) of the body tissues scanned. The aim ducing the tube current (xy plane) when the X-ray tube
of automatic tube current modulation (ATCM) is to main- was positioned over radiosensitive organs, including the
tain operator-selected image quality throughout the scan female breast, the thyroid gland, and the lens of the eye.
while reducing the radiation dose to the patient. State-of- Long term follow-up data from the Life Span Study have
the-art ATCM modulates the tube current in three dimen- shown that the radiation sensitivity of female breast tis-
sions (x-, y-, and z-axes) for every slice of the scan, based sue is higher than previously thought. Consequently, in
on the X-ray attenuation within the slice. Tube current the 2007 ICRP (International Commission on Radiologi-
modulation is implemented based on either: cal Protection) publication [18], the tissue weighting fac-
1. attenuation measurements obtained from the topogram tor of the female breast was increased. Because in chest
or scout views: This approach is used by all manufac- CT the breasts are included in the scan range but are not
turers for z-axis (longitudinal) ATCM modulation, but the target organ of the imaging procedure, it is important
is used by GE and Toshiba for the x- and y-axes (trans- to decrease the radiation dose to these radiosensitive tis-
verse or angular) [13, 14]; sues. Organ-based tube current modulation reduces the
2. tube current adjustment on the fly using the mea- tube current in the anterior 120 arc of the rotation by up
sured attenuation of the previous one-half gantry rota- to 25% of the tube current, while increasing the tube cur-
tion: This approach is used by Siemens for x, y (trans- rent to 125% over the remaining 240 of the rotation to
verse or angular) ATCM modulation [15]. maintain constant image quality in the center of the pa-
The noise level or desired image quality is user-pro- tient [19] (Fig. 3). In phantom studies, an average reduc-
grammable and typically adjusted according to the clini- tion of the breast dose up to 37% was achieved [20].
Dose-Lowering Strategies in Computed Tomography Imaging of the Lung and Heart 171
a b
Fig. 4 a, b. Examples of lowering the tube voltage in pulmonary CT angiography to rule out pulmonary embolism. a A transverse slice in a
male patient (95 kg body weight) who was scanned with 80 kVp [150 mA; volume CT dose index (CTDIvol) of 4.2 mGy]. b An obese
male patient (150 kg body weight) scanned with 100 kVp (100 mA, CTDIvol of 9.3 mGy). Both images were acquired on a 16-slice MDCT
scanner and reconstructed with filtered back projection. The images clearly demonstrate that, even with dated CT technology, lowering of
the tube voltage in pulmonary angiography is possible without impairing diagnostic accuracy
172 A. Euler, Z. Szucs-Farkas, J.R. Mayo, S.T. Schindera
measurements; specifically, an 80-kVp protocol is ap- the use of IR techniques for dose reduction in general, ra-
plied for patients with a body weight up to 100 kg and a diologists are advised to pay extra attention to the diag-
100-kVp protocol for patients with a body weight 100 nostic accuracy with respect to the detection of small pul-
kg (Fig. 4). At least two manufacturers offer automatic monary nodules on the IR images. This advice is given
tube voltage selection, which selects an optimal tube volt- against the background of the substantially decreased im-
age based on patient size, selected CT protocol, and op- age noise achieved with IR but no impact of improved le-
erator-defined image quality [25]. In a prospective study, sion detectability [28].
Niemann et al. demonstrated a dose reduction of 39% in
pulmonary CT angiography using automatic tube voltage
selection [25]. Instead of the standard tube voltage of 120 Special Indications
kVp, in 28% of the cases a tube voltage of 100 kVp and
in 69% of the cases a tube voltage of 80 kVp were auto- Pulmonary CT Angiography in Pregnancy
matically selected.
Low-dose pulmonary CT angiography is a valuable al-
Iterative Reconstruction Technique ternative for a ventilation/perfusion scan to exclude pul-
monary embolism in a pregnant woman with elevated D-
While the commonly used filtered back-projection image dimer level and negative compression sonography of the
reconstruction technique is computationally efficient, it veins of the lower extremities. The use of tube voltages
suffers from noise and artifact limitations. Despite their of 80 or 100 kVp and limiting the scan range from the
availability over 20 years ago, iterative reconstruction aortic arch to the diaphragm will effectively reduce the
(IR) techniques never gained acceptance due to their radiation dose. The increased radiosensitivity of the glan-
computational intensity and slow image reconstruction dular breast tissue during pregnancy deserves particular
speed. The advent of faster computer platforms and the attention. Although the impact of using breast shields in
demand for a reduction of the radiation dose without an preventing higher absorbed breast dose is controversial
increase in image noise has spurred research into IR. The [29, 30], a reduction of tube current in the anterior part
nomenclature of IR techniques is not yet standard. Man- of the thorax with organ-based tube current modulation
ufacturers have developed a number of proprietary tech- might offer a partial solution, as mentioned earlier.
niques involving different strengths and varying in re-
construction speed, spatial resolution, and image noise. In Lung Cancer Screening
general, in IR an initial image is formed that is then used
to generate simulated raw scan data, which are compared Recent US and Japanese lung cancer screening trial data
to the acquired scan data. A new image is then generated have shown a statistically significant (p0.05) reduction
based on the difference. This process is repeated, or iter- in lung cancer mortality when lung cancers are diagnosed
ated, to further improve the image. IR can operate either at an early stage [31, 32]. Ultra low dose lung cancer
with image and raw scan data space (Adaptive Statistical screening is possible because the large contrast differ-
Iterative Reconstruction, ASIR, GE Healthcare; Adaptive ences between the pulmonary nodules and surrounding
Iterative Dose Reduction 3D, AIDR 3D, Toshiba, Sino- air-filled lung allows the use of images with very high
gram Affirmed Iterative Reconstruction, SAFIRE, noise levels; however, mediastinal images will have re-
Siemens Healthcare; and iDose, Philips Healthcare) or duced diagnostic accuracy with these techniques. The im-
raw scan data space alone (MBIR or Veo, GE Healthcare; age quality and diagnostic accuracy for the lung
Advanced Modeled Iterative Reconstruction, ADMIRE, parenchyma using low mA values have been investigated
Siemens Healthcare). IR is a powerful tool to either im- [12, 33]. Results from phantom studies showed that 80 or
prove image quality in large patients or to maintain diag- 100 kVp can be combined with 25 mA with no deterio-
nostically adequate noise while decreasing radiation dose ration in the detection of small lung nodules [33]. A re-
in small or intermediate sized patients. cent phantom study achieved high image quality and sen-
There are numerous reports of substantial dose reduc- sitivity in pulmonary nodule detection at an effective
tions using IR rather than standard filtered back projec- dose of 0.06 mSv using a novel IR technique and tin fil-
tion while maintaining image quality [26, 27]. However, tration for spectral shaping [34].
radiologist applying IR techniques have to be aware of
the fact that images thus reconstructed tend have a Cardiac CT Dose Reduction
blotchier and more pixilated appearance, impacting the
evaluation of normal anatomic lung structure such as the In cardiac CT, the requirement for scan acquisition to be
interlobular fissures, subsegmental bronchial walls, and synchronized to the phase of the cardiac cycle leads to
small peripheral blood vessels [27]. This occurs particu- specialized dose reduction strategies [35]. The initial ap-
larly with very aggressive noise reduction in order to proach to cardiac gating was retrospectively gated helical
achieve sub-mSv chest CT scans. Moreover, the image acquisitions that used low pitch values (approximately
quality of sub-mSv chest CT is suboptimal for diagnos- 0.20.5 depending on heart rate and single vs. dual tube
ing mediastinal structures such as lymph nodes [27]. In gantry configuration) to provide raw scan data at all
Dose-Lowering Strategies in Computed Tomography Imaging of the Lung and Heart 173
spatial locations in all phases of the cardiac cycle. Initial for the most motion-free time point for the coronary ar-
retrospective helical acquisition allowed the reconstruc- teries. In some cases, motion-free visualization of differ-
tion of low-noise images at multiple cardiac phases. The ent parts of the vessels (e.g., proximal vs. distal portion
images were typically reconstructed at 5 or 10% incre- of the right coronary artery) occurs at different time
ments of the R to R interval. These image sets provided points. However, these additional projections are associ-
visualization of the coronary arteries and the motion of ated with an increased radiation dose such that some au-
the ventricular walls and valves but at the cost of a high thors recommend against the routine use of padding [36].
radiation dose (932 mSv). Moreover, this acquisition
technique was radiation dose wasteful for pure coronary
artery imaging, as the arteries are only motion free at Conclusion
6080% of the R to R interval for heart rates below 80
beats per minute. This review has outlined a number of practical methods
The dose efficiency of retrospective helical acquisition for dose reduction that are independent of the CT scan-
was improved by 50% with the implementation of EKG ner make and model. Other very effective scanner-spe-
tube current modulation, which reduces tube current by cific technical innovations are also available on current-
5090% during high-motion portions of the cardiac cy- generation scanners. They require that radiologists are
cle. This provides low-noise coronary artery images in di- aware of the benefits and limitations of their use. When
astole and noisier wall and valve motion images at se- all current dose reduction strategies are utilized, sub-mSv
lected increments through the remaining cardiac cycle. cardiothoracic CT is achievable in non-obese patients.
The use of tube current modulation is recommended for
all retrospective cardiac-gated acquisitions.
Prospective cardiac gating provides further dose re- References
ductions for coronary artery imaging by acquiring raw 1. Mettler FA, Jr., Bhargavan M, Faulkner K et al (2009) Radio-
scan data only at 6080% of the cardiac cycle. This scan logic and nuclear medicine studies in the United States and
acquisition mode is a modification of step and shoot, worldwide: frequency, radiation dose, and comparison with
with the scanner table stationary while a full detector other radiation sources 1950-2007. Radiology 253:520-531.
width (e.g., 40 mm) of raw scan data are acquired. The 2. Aroua A, Samara ET, Bochud FO et al (2013) Exposure of the
Swiss population to computed tomography. BMC Medical
scan table is then moved (e.g., 35 mm) and raw data are Imaging 13:22.
acquired at another position of the heart during the next 3. Preston DL, Ron E, Tokuoka S et al (2007) Solid cancer inci-
heart beat. Thus, the detector steps through the entire car- dence in atomic bomb survivors: 1958-1998. Radiation Res
diac volume, while the 5-mm overlap ensures there are no 168:1-64.
4. Pearce MS, Salotti JA, Little MP et al (2012) Radiation expo-
gaps in the coronary artery anatomy due to minor varia- sure from CT scans in childhood and subsequent risk of
tions in heart positions between acquisitions. This results leukaemia and brain tumours: a retrospective cohort study.
in a pitch factor of 0.875. Depending on the size of the Lancet 380:499-505.
heart and the width of the detector, three to five table 5. Mathews JD, Forsythe AV, Brady Z et al (2013) Cancer risk in
steps are used. The result is a 77% dose reduction com- 680,000 people exposed to computed tomography scans in
childhood or adolescence: data linkage study of 11 million
pared to retrospective helical acquisition without tube Australians. BMJ 346:f2360.
current modulation. Wide-area detectors capable of 6. Flors L, Leiva-Salinas C, Norton PT et al (2013) Endoleak de-
covering the entire z-axis length of the heart (e.g., 16 cm) tection after endovascular repair of thoracic aortic aneurysm
can perform this acquisition without table movement. using dual-source dual-energy CT: suitable scanning protocols
This allows a pitch factor of 1 but requires cone beam and potential radiation dose reduction. AJR Am J Roentgenol
200:451-460.
corrections in the image reconstruction. 7. Prokop M (2008) [Radiation dose in computed tomography.
A further modification of prospective gating is provid- Risks and challenges]. Der Radiologe 48:229-242.
ed in dual tube CT scanners operating in helical mode. 8. Karlo C, Gnannt R, Frauenfelder T et al (2011) Whole-body
This technique is known as high-pitch or flash imaging. CT in polytrauma patients: effect of arm positioning on tho-
Using two CT tubes, complete raw scan data sets can be racic and abdominal image quality. Emerg Radiol 18:285-293.
9. Brink M, de Lange F, Oostveen LJ et al (2008) Arm raising at
obtained using a pitch factor of 3.2. The high table speed exposure-controlled multidetector trauma CT of thoracoab-
allows coverage of the entire heart during diastole in pa- dominal region: higher image quality, lower radiation dose.
tients with a heart rate below 65 beats per minute. This Radiology 249:661-670.
mode is not useful secondary to motion artifact in pa- 10. Li J, Udayasankar UK, Toth TL et al (2007) Automatic patient
centering for MDCT: effect on radiation dose. AJR Am J
tients with higher heart rates. Roentgenol 188:547-552.
In all prospective acquisitions, transient elevation in 11. Mayo JR, Whittall KP, Leung AN et al (1997) Simulated dose
heart rate or arrhythmias can seriously degrade image reduction in conventional chest CT: validation study. Radiolo-
quality. With the exception of high-pitch acquisitions, gy 202:453-457.
this source of image degradation can be partially mitigat- 12. Mayo JR, Kim KI, MacDonald SL et al (2004) Reduced radi-
ation dose helical chest CT: effect on reader evaluation of
ed by acquiring redundant projections. This modification structures and lung findings. Radiology 232:749-756.
is referred to as padding and allows movement of the ac- 13. Kalra MK, Maher MM, Toth TL (2004) Comparison of Z-ax-
quisition window within the diastolic interval to search is automatic tube current modulation technique with fixed tube
174 A. Euler, Z. Szucs-Farkas, J.R. Mayo, S.T. Schindera
current CT scanning of abdomen and pelvis. Radiology 26. Kalra MK, Woisetschlager M, Dahlstrom N et al (2013) Sino-
232:347-353. gram-affirmed iterative reconstruction of low-dose chest CT:
14. Kalra MK, Maher MM, Kamath RS et al (2004) Sixteen-de- effect on image quality and radiation dose. AJR Am J
tector row CT of abdomen and pelvis: study for optimization Roentgenol 201:W235-244.
of Z-axis modulation technique performed in 153 patients. Ra- 27. Padole A, Singh S, Ackman JB et al (2014) Submillisievert
diology 233:241-249. chest CT with filtered back projection and iterative recon-
15. Greess H, Lutze J, Nomayr A et al (2004) Dose reduction in struction techniques. AJR Am J Roentgenol 203:772-781.
subsecond multislice spiral CT examination of children by on- 28. Schindera ST, Odedra D, Raza SA et al (2013) Iterative re-
line tube current modulation. Eur Radiol 14:995-999. construction algorithm for CT: can radiation dose be decreased
16. Kalra MK, Rizzo S, Maher MM et al (2005) Chest CT per- while low-contrast detectability is preserved? Radiology
formed with z-axis modulation: scanning protocol and radia- 269:511-518.
tion dose. Radiology 237:303-308. 29. Vollmar SV, Kalender WA (2008) Reduction of dose to the fe-
17. Schindera ST, Nelson RC, Toth TL et al (2008) Effect of pa- male breast in thoracic CT: a comparison of standard-protocol,
tient size on radiation dose for abdominal MDCT with auto- bismuth-shielded, partial and tube-current-modulated CT ex-
matic tube current modulation: phantom study. AJR Am J aminations. Eur Radiol 18:1674-1682.
Roentgenol 190:W100-W105. 30. Hurwitz LM, Yoshizumi TT, Goodman PC et al (2009) Radia-
18. The 2007 Recommendations of the International Commission tion dose savings for adult pulmonary embolus 64-MDCT us-
on Radiological Protection (2007) ICRP publication 103. An- ing bismuth breast shields, lower peak kilovoltage, and auto-
nals of the ICRP 37:1-332. matic tube current modulation. AJR Am J Roentgenol 192:244-
19. Duan X, Wang J, Christner JA et al (2011) Dose reduction to 253.
anterior surfaces with organ-based tube-current modulation: 31. Aberle DR, DeMello S, Berg CD et al (2013) Results of the
evaluation of performance in a phantom study. AJR Am J two incidence screenings in the National Lung Screening Tri-
Roentgenol 197:689-695. al. New England J Med 369:920-931.
20. Lungren MP, Yoshizumi TT, Brady SM et al (2012) Radiation 32. Nawa T, Nakagawa T, Mizoue T et al (2012) A decrease in
dose estimations to the thorax using organ-based dose modu- lung cancer mortality following the introduction of low-dose
lation. AJR Am J Roentgenol 199:W65-73. chest CT screening in Hitachi, Japan. Lung Cancer 78:225-
21. Taylor S, Litmanovich DE, Shahrzad M (2014) Organ-based 228.
tube current modulation: are womens breasts positioned in the 33. Christe A, Charimo-Torrente J, Roychoudhury K et al (2013)
reduced-dose zone? Radiology 140694 [Epub ahead of print]. Accuracy of low-dose computed tomography (CT) for detect-
22. Heyer CM, Mohr PS, Lemburg SP et al (2007) Image quality ing and characterizing the most common CT-patterns of pul-
and radiation exposure at pulmonary CT angiography with monary disease. Eur J Radiol 82:e142-150.
100- or 120-kVp protocol: prospective randomized study. Ra- 34. Gordic S, Morsbach F, Schmidt B et al (2014) Ultralow-dose
diology 245:577-583. chest computed tomography for pulmonary nodule detection:
23. Schueller-Weidekamm C, Schaefer-Prokop CM, Weber M et first performance evaluation of single energy scanning with
al (2006) CT angiography of pulmonary arteries to detect pul- spectral shaping. Invest Radiol 49:465-473.
monary embolism: improvement of vascular enhancement 35. Mayo JR, Leipsic JA (2009) Radiation dose in cardiac CT. AJR
with low kilovoltage settings. Radiology 241:899-907. Am J Roentgenol 192:646-653.
24. Szucs-Farkas Z, Christe A, Megyeri B et al (2014) Diagnostic ac- 36. Leipsic J, LaBounty TM, Ajlan AM et al (2013) A prospective
curacy of computed tomography pulmonary angiography with randomized trial comparing image quality, study interpretabil-
reduced radiation and contrast material dose: a prospective ran- ity, and radiation dose of narrow acquisition window with
domized clinical trial. Invest Radiol 49:201-208. widened acquisition window protocols in prospectively ECG-
25. Niemann T, Henry S, Faivre JB et al (2013) Clinical evaluation triggered coronary computed tomography angiography. J Car-
of automatic tube voltage selection in chest CT angiography. diovasc Comput Tomogr 7:18-24.
Eur Radiol 23:2643-2651.
NUCLEAR MEDICINE SATELLITE COURSE
DIAMOND
IDKD 2015-2018
FDG-PET Imaging in NSCLC FDG-PET staging [10]. After systematic lymph node dis-
section in the majority of the patients, 14.3% had occult
FDG-PET for Staging in NSCLC nodal (N1 or N2) metastasis. Total N1 and N2 involve-
ment was detected in 9.5% and 4.8%, respectively. Mul-
Selection of the appropriate patients for radical local tivariate analysis demonstrated that a primary tumor with
treatment, whether surgery, radiotherapy, or multimodal a SUVmax 7.3 was an independent predictor of occult
treatment, is essential. Today, FDG-PET for the staging nodal metastasis. In the retrospective study by Stiles et al.
of nodal and distant metastases is the standard of care, of 266 patients with stage IA disease according to CT-
based on international guidelines. In a prospective study and FDG-PET-based staging, mediastinal lymph node
of NSCLC patients with mostly locally advanced disease, dissection detected N1 and N2 disease in 6.8% and 4.9%
initial staging was performed using CT and with FDG- of the patients, respectively [11]. Tumor size 2 cm and
PET thereafter. FDG-PET led to a change from curative FDG-PET positivity were risk factors for understaging by
to palliative therapy by upstaging the disease extent in FDG-PET.
25% of the patients [6]. A similar study was performed Overall mediastinal staging accuracy was evaluated in
in a patient cohort referred for radical radiotherapy based a meta-analysis. For CT staging, the median sensitivity
on CT staging [7]. The 76 patients were mostly cN. Af- and specificity of mediastinal staging were 61% and
ter FDG-PET staging, radical radiochemotherapy was 79%, respectively. For FDG-PET, the corresponding val-
performed in only 66% of the patients, with palliative ues were 85% and 90. However, the specificity of FDG-
treatment in the remaining 34% mostly because of the de- PET staging was lower when CT showed enlarged lymph
tection by FDG-PET imaging of distant metastases or of nodes [12]. Consequently, all positive nodal findings in
more extensive nodal disease considered too extensive FDG-PET need to be confirmed pathologically, using en-
for successful radical radiochemotherapy. At 4 years af- dobronchial ultrasound or mediastinoscopy.
ter treatment with curative and palliative intent, OS was
35.6% and 4.1%, respectively, thus confirming the accu- Consequences of Improved Staging Accuracy Using FDG-PET
rate selection of a high-risk population using FDG-PET
staging. In patients with stage I NSCLC based on FDG-PET stag-
The timely performance of FDG-PET staging, before ing, stereotactic body radiotherapy (SBRT) of the prima-
the start of radical treatment, is crucial. If the interval be- ry tumor only, without elective treatment of the hilar or
tween staging and the start of radical treatment is too mediastinal lymph node regions, is the guideline-recom-
long, FDG-PET needs to be repeated. Everitt et al. eval- mended standard of care in all patients who are medical-
uated two sequential FDG-PET/CT images acquired from ly inoperable [13]. The omission of elective nodal irradi-
82 patients with a median interval of 24 days prior to the ation combined with other high-precision technologies
start of treatment [8]. Interscan disease progression (respiration correlated imaging, image guidance, intra-
(TNM stage) was detected in 11 (39%) patients. Treat- fractional motion management) allows for irradiation
ment intent changed from curative to palliative in 8 with escalated irradiation doses, which are delivered in a
(29%) because of the detection of newly developed dis- hypo-fractionated manner: biological equivalent doses
tant metastases in the second FDG-PET/CT; in 7 of these (BED) 100 Gy are delivered in 18 fractions.
patients the change in treatment was based on the PET Local tumor control in SBRT is significantly improved
image component. The average standardized uptake val- compared to conventionally fractionated radiotherapy
ue (SUV) increased by 16%. and reaches 90% in the majority of studies. This im-
In patients with NSCLC, FDG-PET is especially use- proved local control translates into significantly im-
ful for nodal staging, which is of fundamental importance proved OS [14], which is mainly limited by the comor-
in therapeutic decision-making (curative vs. palliative; bidities of the patients as well as systemic progression of
surgical vs. non-surgical; conventionally fractionated ra- the disease. Regional failures are rare after FDG-PET
diotherapy vs. stereotactic body radiotherapy) and for tar- staging. Senthi et al. reported a 7.8% regional failure rate
get-volume definition in radical radiotherapy. It is well at 2 years in a large cohort of 676 patients [15].
documented that nodal staging using CT imaging is of In locally advanced NSCLC, involved-field irradiation
low sensitivity and specificity. This is especially true in without elective nodal irradiation is currently practiced in
patients with cN0 disease. DCunha et al. reported the re- most studies and centers, if FDG-PET had been per-
sults of the CALGB 9761 study of 502 patients with clin- formed for staging purposes. Involved-field irradiation
ical stage I disease as determined by CT staging [9]. Af- reduces the irradiated volume substantially, thus allowing
ter surgical resection and mediastinal lymph node dissec- either a reduction of toxicity or iso-toxic escalation of the
tion, 14% of the patients were shown to have had patho- irradiation dose [16]. Early studies confirmed low rates
logic stage II disease and another 13.5% stage III disease. of isolated regional failures in un-irradiated hilar and me-
Two studies evaluated the value of FDG-PET staging diastinal regions [17], with only 1 out of 44 patients treat-
in patients with clinical stage IA disease based on CT ed with involved-field irradiation developing an isolated
staging. Park et al. reported a retrospective study of 147 nodal failure. The value of involved-field irradiation af-
patients who had clinical stage IA disease according to ter FDG-PET staging is currently being evaluated in a
FDG-PET Imaging for Advanced Radiotherapy Treatment of Non-Small-Cell Lung Cancer 179
prospective randomized multi-center trial (PET-Plan, tumor had a significantly worse survival than patients
NCT00697333). with a complete metabolic response. Most importantly,
the location of residual metabolic-active areas within the
FDG-PET as a Prognostic Marker in NSCLC primary tumor after therapy correlated with the initially
high FDG uptake areas determined pre-radiotherapy.
The correlation of pre-treatment, intra-treatment, or ear- Consequently, pre-treatment FDG-PET imaging could be
ly post-treatment FDG-PET characteristics with outcome used to identify subvolumes of the primary tumor at risk
could be used for patient stratification and subsequent for incomplete response after definitive radiotherapy.
treatment adaptation. However, whether FDG-uptake is a A similar study was performed by a group from the
prognostic maker in NSCLC is discussed controversially University of Michigan [22]. In 15 stage IIII NSCLC
in the literature. The uncertainty is at least partially ex- patients treated with a definitive dose of fractionated ra-
plained by differences in the methodology of outcome diotherapy, pre-treatment (2 weeks), intra-treatment (after
modeling. Firstly, different endpoints have been used to approximately 45 Gy), and post-treatment (34 months)
determine a correlation with FDG-PET characteristics: FDG-PET/CT images were acquired. A significant corre-
local tumor control, progression-free survival, and OS. lation between metabolic tumor response during radio-
Secondly, different metrics of FDG-PET images have therapy and metabolic tumor response 3 months post-
been used for outcome correlation: SUVmax, metabolic radiotherapy was determined.
tumor volume, and, more recently, texture characteristics
such as coarseness, contrast, and busyness. Additionally, Consequences of FDG-PET as a Prognostic Marker
tumor volume is a well known independent prognostic
marker, which could confound analysis using FDG-PET If either the pre-treatment or the intra-treatment FDG-
as a prognostic marker [18]. Another matter of contro- PET characteristics of the tumor correlate with local tu-
versy is whether or to what extent inflammatory reactions mor control and/or survival, the appropriate strategy
during or shortly after radio(chemo)therapy influence the would then be to adapt radiotherapy to these tumor-indi-
value of FDG-PET for outcome modeling. vidual functional and biological characteristics. This con-
The prognostic value of pre-treatment FDG-PET was cept of adaptive radiotherapy was first proposed more
evaluated in a meta-analysis of 21 studies by Paesmans et than 10 years ago [23] and is currently under clinical in-
al. [19]. The endpoint was OS. Data from individual da- vestigation.
ta patients were not available; instead, the median SUV Key questions in adaptive radiotherapy remain unan-
value of each study was used as a threshold. The study swered, which has prevented its board adaptation. A dis-
detected a poor prognosis for patients with a high vs. a cussion of these issues in detail is beyond the scope of
low SUV, with an overall combined hazard ratio of 2.08. this chapter but the questions include: Which functional
However, it did not find an optimal SUV threshold but imaging modality, PET tracer, and image characteristics
only that higher SUVs resulted in worse outcome. This are most appropriate for tumor characterization? What
may have been a consequence of the limitations of the are the relevant time points for image acquisition and
meta-analysis or that rather than two groups of patients subsequent treatment adaptation? What are the dynamics
there was a continuous increase in the hazard with in- of functional and biological tumor characteristics and is
creasing SUV. multiple-step adaptation required? Are biological charac-
Matchay et al. reported one of the largest prospective teristics representative for one patient and one tumor as a
studies, in which the pre- and post-treatment FDG-PET whole, or can tumors be subdivided into areas of differ-
characteristics of 250 patients with locally advanced ent biology and function? How can biology, treatment,
NSCLC treated with conventional concurrent platinum- and outcome be correlated with one another? How can
based radiochemotherapy without surgery were analyzed biological information be translated into a robust radio-
[20]. The 2-year survival rate for the entire population therapy plan? How can the outcome of adaptive radio-
was 42.5%. Neither pre-treatment SUVpeak nor SUVmax therapy be evaluated? And how can functional and bio-
correlated with OS, which is in contrast to the meta- logical markers other than imaging, e.g., genomic mark-
analysis described above. In contrast, SUVpeak and ers and RNA expression, be integrated into treatment?
SUVmax in FDG-PET images acquired approximately 14 Despite these uncertainties, the most advanced studies
weeks after treatment correlated with OS. Patients with on adaptive radiotherapy for locally advanced NSCLC
higher residual FDG-PET uptake had a significantly are summarized below.
worse OS. The University of Michigan group reported a pilot
In a further analysis, Aerts et al. analyzed the spatial study in which radiotherapy was adapted after the deliv-
correlation between pre- and post-treatment SUV in 55 ery of 4050 Gy based on FDG-PET/CT [24]. Among the
patients treated with chemoradiation for locally advanced 14 patients with stages IIII NSCLC, the mid-treatment
NSCLC [21]. Pre- and post-treatment FDG-PET images FDT-PET/CT showed a metabolic complete response in
were acquired about 2 weeks and 12 weeks prior to and two and increased FDG uptake in the adjacent lung tissue
after radiotherapy, respectively. The authors reported that in another two. In the remaining 10 patients, CT-mor-
patients with residual metabolic-active areas within the phological and FDG-PET metabolic volumes decreased
180 M. Guckenberger, L. Rudofsky, N. Andratschke
by 26% (range, 15% to 75%) and 44% (range, 10% sis was that pre-treatment inflammation in the lungs
to 100%), respectively. Re-planning of radiotherapy makes pulmonary tissue more susceptible to radiation
based on a reduced metabolic volume was then per- damage. The authors reported that the 95th percentile of
formed in six patients and allowed a substantial intensi- FDG uptake in the lungs, excluding clinical tumor vol-
fication of radiotherapy of 30102 Gy (mean 58 Gy) ume, correlated significantly with the risk of developing
without an increased risk of toxicity compared to a sce- radiation-induced pneumonitis. Pre-treatment FDG-PET
nario without mid-treatment plan adaptation (iso-toxic could therefore be used for risk stratification and in the
dose escalation approach). Even this small study by Feng selection of patients for closer follow-up. Even more im-
et al. [24] clearly demonstrated that: (1) adaptive radio- portant from a radiotherapy perspective is the finding that
therapy is not feasible in all patients; (2) not all patients the fraction of the 5%, 10%, and 20% highest SUV vox-
will benefit from adaptive radiotherapy; but that (3) the els that received irradiation doses 25Gy improved the
potential benefit of normal tissue sparing or treatment in- correlation with radiation-induced lung toxicity. Sparing
tensification is considerable in some patients. This con- these lung areas from low-dose irradiation could there-
cept of adaptive radiotherapy using mid-treatment FDG- fore decrease the risk of lung toxicity.
PET/CT is currently being investigated in a randomized
phase II trial (NCT01507428). Gross Tumor Definition Using FDG-PET Imaging
Another study based on the experiences of Aerts et al.
[21] is underway. In this randomized phase II study of pa- Target-volume definition of peripheral early-stage
tients with inoperable stage IB to stage III NSCLC NSCLC is associated with small inter- and intra-observer
(NCT01024829) [25], patients are randomized to con- variability. However, the accurate analysis of breath-
ventional radiotherapy with a total dose of 66 Gy given ing-induced tumor motion and its integration into target-
in 24 fractions of 2.75 Gy with an integrated boost to the volume definition are essential steps in the treatment
primary tumor as a whole (Arm A) or with an integrated process. Respiration-correlated 4D-CT is the current
boost to the 50% SUVmax area of the primary tumor method of choice, despite its limitations due to residual
(from the pre-treatment FDG-PET scan) (Arm B). The motion artifacts and the short image acquisition time.
primary endpoint of this study is progression-free sur- FDG-PET imaging, with its long image acquisition time,
vival. No results have been published so far. may provide a tool similar to slow CT scanning to evalu-
ate tumor motion. Although this concept appears straight-
Normal Tissue Characterization Using FDG-PET Imaging forward, studies have shown only a poor correlation be-
tween PET and the gold-standard of 4D-CT [28]. FDG-
FDG uptake in inflammatory processes may negatively PET for the assessment of tumor-motion amplitude and
influence tumor imaging and characterization but at the motion patterns is therefore not recommended.
same time may help in the visualization, quantification, Several 4D technologies have been developed to ac-
or prediction of inflammatory radiation-induced toxicity. quire and/or reconstruct respiration-correlated 4D PET
Pneumonitis is the most relevant toxicity in patients un- images [29]. Using these technologies should allow for
dergoing radiotherapy for lung cancer. Its development more precise tumor characterization and delineation even
correlates with the radiotherapy dose, i.e., the mean lung in the presence of large tumor motion.
dose and the volume of the lung exposed to 20 Gy, but Definition of the target volume is associated with sub-
the correlation is rather weak. stantial uncertainties especially in locally advanced-stage
A group from Melbourne correlated post-treatment NSCLC. These uncertainties have been quantified in sev-
(median 70 days) FDG uptake in the lungs with the de- eral studies that evaluated inter-observer variability in tar-
velopment of radiation-induced pneumonitis [26]. The get-volume definition and the variability in the target con-
authors reported a significant association between the tours between multiple experts in lung cancer treatment. In
worst Radiation Therapy Oncology Group (RTOG) grade the study of Steenbackers et al. based on 22 patients with
of pneumonitis occurring at any time after radiotherapy early and locally advanced stage NSCLC [30], 11 experts
and the severity of FDG uptake in the lung. No associat- delineated the gross tumor volume using CT images only
ed between FDG uptake and the duration of pneumonitis and then, 1 year later, on matched FDG-PET/CT images.
was found. The authors concluded that FDG-PET may be The variability between the experts was reduced from 1
useful in the prediction, diagnosis, and therapeutic mon- cm (one standard deviation) to 0.4 cm, clearly showing the
itoring of radiation pneumonitis. added value of the FDG-PET imaging component. Two
However, the post-treatment prediction of pneumonitis clinical situations have been identified in which FDG-PET
or lung toxicity has no value in terms of modifying ra- makes a relevant difference compared to CT imaging on-
diotherapy to reduce the risk of pneumonitis, e.g., by re- ly: The difficulties of nodal staging were described above,
distribution of the radiotherapy dose or dose de-escala- and FDG-PET is essential for defining the nodal target
tion. Petit et al. [27] reported promising results from a volume, especially in cases of involved-node irradiation.
study in which pre-treatment FDG uptake in the lungs The other situation in which FDG-PET can reduce uncer-
was shown to correlate with the post-radiotherapy devel- tainties in target-volume definition is the differentiation
opment of radiation-induced lung toxicity. The hypothe- between macroscopic tumor and atelectasis.
FDG-PET Imaging for Advanced Radiotherapy Treatment of Non-Small-Cell Lung Cancer 181
17. De Ruysscher D, Wanders S, van Haren E et al (2005) Selec- volume during radiotherapy for non-small-cell lung cancer and
tive mediastinal node irradiation based on FDG-PET scan da- its potential impact on adaptive dose escalation and normal tis-
ta in patients with non-small-cell lung cancer: a prospective sue sparing. Int J Radiat Oncol Biol Phys 73:1228-1234.
clinical study. Int J Radiat Oncol Biol Phys 62:988-994. 25. van Elmpt W, De Ruysscher D, van der Salm A et al (2012)
18. Reymen B, Van Loon J, van Baardwijk A et al (2013) Total The PET-boost randomised phase II dose-escalation trial in
gross tumor volume is an independent prognostic factor in pa- non-small cell lung cancer. Radiother Oncol 104:67-71.
tients treated with selective nodal irradiation for stage I to III 26. Mac Manus MP, Ding Z, Hogg A et al (2011) Association be-
small cell lung cancer. Int J Radiat Oncol Biol Phys 85:1319- tween pulmonary uptake of fluorodeoxyglucose detected by
1324. positron emission tomography scanning after radiation therapy
19. Paesmans M, Berghmans T, Dusart M et al (2010) Primary tu- for non-small-cell lung cancer and radiation pneumonitis. Int
mor standardized uptake value measured on fluorodeoxyglu- J Radiat Oncol Biol Phys 80:1365-1371.
cose positron emission tomography is of prognostic value for 27. Petit SF, van Elmpt WJ, Oberije CJ et al (2011) [(1)(8)F]fluo-
survival in non-small cell lung cancer: update of a systematic rodeoxyglucose uptake patterns in lung before radiotherapy
review and meta-analysis by the European Lung Cancer Work- identify areas more susceptible to radiation-induced lung tox-
ing Party for the International Association for the Study of icity in non-small-cell lung cancer patients. Int J Radiat Oncol
Lung Cancer Staging Project. J Thorac Oncol 5:612-619. Biol Phys 81:698-705.
20. Machtay M, Duan F, Siegel BA et al (2013) Prediction of sur- 28. Hanna GG, van Sornsen de Koste JR, Dahele MR et al (2012)
vival by [18F]fluorodeoxyglucose positron emission tomogra- Defining target volumes for stereotactic ablative radiotherapy
phy in patients with locally advanced non-small-cell lung can- of early-stage lung tumours: a comparison of three-dimen-
cer undergoing definitive chemoradiation therapy: Results of sional 18F-fluorodeoxyglucose positron emission tomography
the ACRIN 6668/RTOG 0235 Trial. J Clin Oncol 31:3823- and four-dimensional computed tomography. Clinical oncolo-
3830. gy (Royal College of Radiologists (Great Britain)) 24:e71-80.
21. Aerts HJ, van Baardwijk AA, Petit SF et al (2009) Identifica- 29. Kruis MF, van de Kamer JB, Houweling AC et al (2013) PET
tion of residual metabolic-active areas within individual motion compensation for radiation therapy using a CT-based
NSCLC tumours using a pre-radiotherapy (18)Fluo- mid-position motion model: methodology and clinical evalua-
rodeoxyglucose-PET-CT scan. Radiother Oncol 91:386-392. tion. Int J Radiat Oncol Biol Phys 87:394-400.
22. Kong FM, Frey KA, Quint LE (2007) A pilot study of 30. Steenbakkers RJ, Duppen JC, Fitton I et al (2006) Reduction
[18F]fluorodeoxyglucose positron emission tomography scans of observer variation using matched CT-PET for lung cancer
during and after radiation-based therapy in patients with non delineation: a three-dimensional analysis. Int J Radiat Oncol
small-cell lung cancer. J Clin Oncol 25:3116-3123. Biol Phys 64:435-448.
23. Ling CC, Humm J, Larson S et al (2000) Towards multidi- 31. Sridhar P, Mercier G, Tan J et al (2014) FDG PET metabolic
mensional radiotherapy (MD-CRT): biological imaging and tumor volume segmentation and pathologic volume of prima-
biological conformality. Int J Radiat Oncol Biol Phys 47:551- ry human solid tumors. AJR Am J Roentgenol 202:1114-1119.
560. 32. Huang K, Senthi S, Palma DA (2013) High-risk CT features
24. Feng M, Kong FM, Gross M et al (2009) Using fluo- for detection of local recurrence after stereotactic ablative ra-
rodeoxyglucose positron emission tomography to assess tumor diotherapy for lung cancer. Radiother Oncol 109:51-57.
IDKD 2015-2018
b c
d e
Fig. 1 a-e. FDG-PET/CT in a 76-year-old man. Images display an FDG-avid squamous cell carcinoma in the right upper lobe (a, b) as well
as pretracheal (c), right paratracheal (d), and subcarinal (e) lymph node metastases. The classification pT1B pN2 cM0 was proven by his-
tology. Notably, the left lateral border of the trachea differentiates left from right paratracheal lymph nodes. Thus, the pretracheal lymph
node metastases (c) would not mean N2 but N3 disease if the lung tumor was in the left lung. Subcarinal lymph node metastases (e) are
always considered N2 disease regardless of the location of the primary lung tumor
but N3 disease if the primary is located in the left lung is no longer in line with current treatment guidelines. For
(Fig. 1). instance, stage IIIA with N2 lymph node metastasis may
The M descriptor has been divided into M1a and M1b, be considered inoperable if lymph node metastases are
for intrathoracic and distant spread, respectively. Patients bulky or in multiple lymph node levels that lie in the ip-
with malignant pleural effusions have been up-staged silateral mediastinum. A curative operation may be con-
from T4 to M1 (Table 1). sidered if down-staging of N3 disease with neoadjuvant
The labeling of stages also was revised between the 6th chemotherapy is successful. In stage IV disease, curative
and 7th editions (Table 2). For example, T2bN0 has treatment approaches may be considered today even if
moved from stage IB to stage IIA, and T2aN1 from stage oligometastatic disease is found with involvement of the
IIB to stage IIA. Patients with tumors 7 cm move from adrenal glands or the brain. Last but not least, standard-
IB to IIB if there is no lymph node metastasis, and from ized treatment strategies are to be defined in adenocarci-
IIB to IIIA if they have N1 lymph nodes. Patients with- nomas with EGFR mutations or ALK rearrangements.
out N3 lymph nodes but with an additional nodule in the This brief summary demonstrates the complexity of
same lobe have been moved from stage IIIB to stage IIIA. tumor staging in lung cancer and thus shows the need for
The new TNM staging system for lung cancer directly a multidisciplinary approach.
impacts treatment algorithms. Treatment approaches for
NSCLC are being constantly adapted and may be shifted T Staging
from standardized to personalized strategies. Formerly,
stages I to IIIA were considered operable, while stages CT and FDG-PET/CT are important imaging modalities
IIIB and VI were considered not to be operable. This in the evaluation of the primary tumor (T staging). CT is
186 L. Husmann, P. Stolzmann
known to have a low accuracy in the evaluation of inva- FDG-positive lymph nodes, suspicious for metastases,
sion of the pleura, the chest wall, and/or the medi- should generally be confirmed by biopsy, if relevant for
astinum, and the correct differentiation between tumor patient management.
and peritumoral atelectasis is often difficult [6, 7]. FDG-
PET/CT may improve accuracy in the detection of tu- M Staging
mor invasion into the chest wall [8-10]. This is important
particularly in patients with a poor cardio-pulmonary re- Distant metastatic disease is already present in 1836%
serve, since the preoperative determination of chest wall of patients with newly diagnosed NSCLC [17]. The abil-
infiltration is desirable in order to avoid extended en ity to detect distant metastases with a high sensitivity,
bloc resection. specificity, and accuracy (94%, 97%, and 96%, respec-
To evaluate infiltration of the mediastinum (T4), con- tively) [18] is the main advantage of FDG-PET/CT and
trast-enhanced CT may be employed but it has a relative- renders PET/CT superior to other imaging modalities.
ly low sensitivity, specificity, and accuracy (68%, 72%, The high sensitivity of FDG-PET/CT to detect distant
and 70%, respectively) [7]. According to our experience, metastases changes patient management in a consider-
PET/CT is not accurate in clearly identifying direct inva- able number of patients, mostly by the up-staging of dis-
sion of the mediastinal vasculature and airways even if ease. Thus, FDG-PET/CT becomes a cost effective imag-
the CT component of FDG-PET/CT is performed with ing modality despite its high operational cost due to a
contrast medium. significant reduction of morbidity though undesired
FDG-PET/CT is helpful for the differentiation be- surgeries [19-21].
tween tumor and peritumoral atelectasis. This allows for The American National Comprehensive Cancer Net-
a more precise determination of the T stage and thereby work guidelines advocate FDG-PET/CT for all patients
impacts surgical planning as sleeve lobectomy is pre- with newly diagnosed NSCLC provided that there is no
ferred over pneumonectomy and radiotherapy plan- evidence of distant metastases. An additional magnetic
ning as the accurate information of tumor extent pro- resonance imaging (MRI) study of the brain is recom-
vided by FDG-PET/CT contributes to a change in the ra- mended in patients with stage II-IV disease to complete
diation field in 3040% of patients [11]. accurate whole-body tumor staging. This is mandatory
because the ability to detect brain metastases is low in
N Staging FDG-PET/CT (see below).
b c
d e
a
Fig. 2 a-e. FDG-PET/CT in a 69-year-old woman with an FDG-avid adenocarcinoma of the left lower lobe (a-c). FDG-PET/CT shows not
only N2 disease (a, e) but also an FDG-avid pneumonia in the right upper lobe (d, e), which may be misinterpreted as metastasis, result-
ing in upstaging to T4 disease (i.e., nodule in another lobe on the same side) and thus in total pneumonectomy. Final tumor classification
after resection of the left lower lobe and mediastinal lymphadenectomy was pT1B pN2 cM0
a b c
Fig. 3 a-c. FDG-PET/CT in a 58-year-old woman with an FDG-avid adenocarcinoma in the left lung. Initial staging (a) shows extensive lymph
node metastases (N3); TNM stage was cT1 pN3 cM0. First follow-up (b) 2 months after the scan in (a) shows the complete metabolic re-
sponse of both the lung cancer and all lymph nodes metastases after first line chemotherapy with cisplatin and alimta. Second PET/CT fol-
low-up (c) 5 months after the scan in (b) and while the asymptomtic patient was on maintenance therapy with alimta. The recurrence of hi-
lar and mediastinal lymph nodes metastases and new abdominal lymph nodes metastases are seen. Chemotherapy with taxotere was started
188 L. Husmann, P. Stolzmann
Hence, FDG-PET/CT has the potential to reduce side ef- 13. Steinert HC, Hauser M, Allemann F et al (1997) Non-small
fects and costs when therapy is ineffective [27]. cell lung cancer: nodal staging with FDG PET versus CT with
correlative lymph node mapping and sampling. Radiology
202:441-446.
14. Vansteenkiste JF, Stroobants SG, De Leyn PR et al (1998)
Recurrent Lung Cancer Lymph node staging in non-small-cell lung cancer with FDG-
PET scan: a prospective study on 690 lymph node stations
Patients with NSCLC who have been treated with curative from 68 patients. J Clin Oncol 16:2142-2149.
15. Pieterman RM, van Putten JW, Meuzelaar JJ et al (2000) Pre-
intent are usually followed up with CT every 612 months. operative staging of non-small-cell lung cancer with positron-
FDG-PET/CT is used to evaluate equivocal CT findings, emission tomography. N Engl J Med 343:254-261.
due to its very high accuracy in distinguishing recurrent 16. Chao F, Zhang H (2012) PET/CT in the staging of the non-
disease from benign treatment effects. If tumor recurrence small-cell lung cancer. J Biomed Biotechnol 2012:783739.
is suspected or confirmed, a FDG-PET/CT for restaging is 17. Quint LE (2007) Staging non-small cell lung cancer. Cancer
Imaging 7:148-159.
indicated to differentiate local from distant recurrence, and 18. Hellwig D, Ukena D, Paulsen F et al (2001) [Meta-analysis of
thus to plan local or systemic treatment [28, 29] (Fig. 3). the efficacy of positron emission tomography with F-18-fluo-
Moreover, initial data suggests that FDG-PET/CT is useful rodeoxyglucose in lung tumors. Basis for discussion of the
in the detection of recurrences in asymptomatic NSCLC German Consensus Conference on PET in Oncology 2000].
patients after potentially curative operations [28-30]. Pneumologie 55:367-377.
19. van Tinteren H, Hoekstra OS, Smit EF et al (2002) Effective-
ness of positron emission tomography in the preoperative as-
sessment of patients with suspected non-small-cell lung can-
References cer: the PLUS multicentre randomised trial. Lancet 359:1388-
1393.
1. Charloux A, Quoix E, Wolkove N et al (1997) The increasing 20. Kelly RF, Tran T, Holmstrom A et al (2004) Accuracy and cost-
incidence of lung adenocarcinoma: reality or artefact? A re- effectiveness of [18F]-2-fluoro-deoxy-D-glucose-positron
view of the epidemiology of lung adenocarcinoma. Int J Epi- emission tomography scan in potentially resectable non-small
demiol 26:14-23. cell lung cancer. Chest 125:1413-1423.
2. Travis WD, Brambilla E, Noguchi M et al (2011) Internation- 21. Fischer B, Lassen U, Mortensen J et al (2009) Preoperative
al association for the study of lung cancer/american thoracic staging of lung cancer with combined PET-CT. N Engl J Med
society/european respiratory society international multidisci- 361:32-39.
plinary classification of lung adenocarcinoma. J Thorac Oncol 22. Higashi K, Ueda Y, Seki H et al (1998) Fluorine-18-FDG PET
6:244-285. imaging is negative in bronchioloalveolar lung carcinoma. J
3. Janne PA, Meyerson M (2012) ROS1 rearrangements in lung Nucl Med 39:1016-1020.
cancer: a new genomic subset of lung adenocarcinoma. J Clin 23. Erasmus JJ, McAdams HP, Patz EF, Jr. et al (1998) Evaluation
Oncol 30:878-879. of primary pulmonary carcinoid tumors using FDG PET. AJR
4. Sobin L, Gospodarowisz M, Wittekind C (eds) (2010) TNM Am J Roentgenol 170:1369-1373.
classification of malignant tumours, 7th Ed. Blackwell Pub- 24. Lardinois D, Weder W, Roudas M et al (2005) Etiology of soli-
lishing, Oxford, UK. tary extrapulmonary positron emission tomography and com-
5. Mountain CF, Dresler CM (1997) Regional lymph node clas- puted tomography findings in patients with lung cancer. J Clin
sification for lung cancer staging. Chest 111:1718-1723. Oncol 23:6846-6853.
6. Webb WR, Gatsonis C, Zerhouni EA et al (1991) CT and MR 25. Soussan M, Augier A, Brillet PY et al (2014) Functional imag-
imaging in staging non-small cell bronchogenic carcinoma: re- ing in extrapulmonary sarcoidosis: FDG-PET/CT and MR
port of the Radiologic Diagnostic Oncology Group. Radiology features. Clin Nucl Med 39:e146-159.
178:705-713. 26. Tiseo M, Ippolito M, Scarlattei M et al (2014) Predictive and
7. Rendina EA, Bognolo DA, Mineo TC et al (1987) Computed prognostic value of early response assessment using 18FDG-
tomography for the evaluation of intrathoracic invasion by PET in advanced non-small cell lung cancer patients treated
lung cancer. J Thorac Cardiovasc Surg 94:57-63. with erlotinib. Cancer Chemother Pharmacol 73:299-307.
8. Antoch G, Stattaus J, Nemat AT et al (2003) Non-small cell 27. Skoura E, Datseris IE, Platis I et al (2012) Role of positron
lung cancer: dual-modality PET/CT in preoperative staging. emission tomography in the early prediction of response to
Radiology 229:526-533. chemotherapy in patients with non-small-cell lung cancer. Clin
9. Lardinois D, Weder W, Hany TF et al (2003) Staging of non- Lung Cancer 13:181-187.
small-cell lung cancer with integrated positron-emission to- 28. Jimenez-Bonilla JF, Quirce R, Martinez-Rodriguez I et al
mography and computed tomography. N Engl J Med (2013) Diagnosis of recurrence and assessment of post-recur-
348:2500-2507. rence survival in patients with extracranial non-small cell lung
10. Shim SS, Lee KS, Kim BT et al (2005) Non-small cell lung cancer evaluated by 18F-FDG PET/CT. Lung Cancer 81:71-76.
cancer: prospective comparison of integrated FDG PET/CT and 29. Kanzaki R, Higashiyama M, Maeda J et al (2010) Clinical val-
CT alone for preoperative staging. Radiology 236:1011-1019. ue of F18-fluorodeoxyglucose positron emission tomography-
11. Nestle U, Walter K, Schmidt S et al (1999) 18F-deoxyglucose computed tomography in patients with non-small cell lung
positron emission tomography (FDG-PET) for the planning of cancer after potentially curative surgery: experience with 241
radiotherapy in lung cancer: high impact in patients with at- patients. Interact Cardiovasc Thorac Surg 10:1009-1014.
electasis. Int J Radiat Oncol Biol Phys 44:593-597. 30. Toba H, Sakiyama S, Otsuka H et al (2012) 18F-fluo-
12. McLoud TC, Bourgouin PM, Greenberg RW et al (1999) Bron- rodeoxyglucose positron emission tomography/computed to-
chogenic carcinoma: analysis of staging in the mediastinum mography is useful in postoperative follow-up of asympto-
with CT by correlative lymph node mapping and sampling. matic non-small-cell lung cancer patients. Interact Cardiovasc
Radiology 182:319-323. Thorac Surg 15:859-864.
IDKD 2015-2018
Fat-Attenuation Nodules
Nodules with macroscopic fat apparent on chest CT of-
ten represent pulmonary hamartomas (mesenchymomas).
These nodules can be purely fat, primarily soft tissue, or Panlobular Lower lobe, diffuse
contain coarse popcorn calcification. Hamartomas typ-
ically show no increased metabolic activity on nuclear
imaging, although they can grow slowly over several
years.
Diffuse Micronodules
The differential diagnosis of diffuse micronodules (10
mm diameter) depends on their distribution in the lungs. Paraseptal Upper and mid lungs,
The spatial resolution of CT technique used in hybrid subpleural
imaging is typically not sufficient enough to fully char-
acterize the morphology and distribution of diffuse mi-
cronodules. However, occasionally, small branching
(Y- or V-shaped) opacities (tree-in-bud) may be apparent
on hybrid imaging studies. Tree-in-bud opacities are a CT
finding of bronchiolitis and nearly always reflect infec-
tion. Occasionally, tree-in-bud opacities reflect bronchi-
olitis from aspiration and may be encountered in patients Fig. 2. Patterns of emphysema
How To Approach Incidental Findings on Computed Tomography Images of the Lungs Obtained in Hybrid Imaging 191
CT, OP manifests as peripheral and peribronchial foci of ic bronchial obstruction, bronchial wall damage, or
consolidation. Air bronchograms are usually visible within parenchymal fibrosis. With parenchymal fibrosis, the di-
the foci of consolidation. In approximately 20% of pa- lation of bronchi, termed traction bronchiectasis, occurs
tients, the reversed halo or atoll sign is present, de- by the maturation and retraction of fibrous tissue formed
fined as a focus of peripheral consolidation with central adjacent to an airway. The CT findings of bronchial dila-
ground-glass opacity. In a small number of patients under- tion include lack of tapering of bronchial lumena (the pri-
going chest wall radiation for breast carcinoma, OP can de- mary sign of bronchiectasis), internal diameter of bronchi
velop outside the small portion of irradiated lung, even af- greater than that of the adjacent pulmonary artery (signet
fecting the contralateral lung. The CT appearance of OP in ring sign), bronchi visible within 1 cm of the costal pleu-
this setting is similar to that of other causes. ra or abutting the mediastinal pleura, and mucus-filled di-
lated bronchi. Bronchiectasis can be described as cylin-
Eosinophilic Pneumonia drical, varicose, or cystic, depending on the morphology.
Pleural lipomas are benign neoplasms containing fat and Esophageal Duplication Cysts
a thin capsule. Occasionally, they are large enough to be
visible on chest radiographs, and most are easily de- Esophageal duplication cysts usually occur in the lower
tectable on cross-sectional imaging. On CT, pleural lipo- mediastinum adjacent to the esophagus. Esophageal du-
mas have fat attenuation and lie in contact with the chest plication cysts can also be lined with pancreatic or gas-
wall. A thin capsule may be visible. tric mucosa, and intracystic hemorrhage can occur.
Esophageal Diverticula
Mediastinum
Esophageal diverticula are described as either pulsion or
Normal Thymus traction, depending on the underlying cause. Pulsion di-
verticula usually occur in the superior or inferior medi-
The normal thymus is large at birth and gradually atro- astinum whereas traction diverticula typically form in the
phies, usually replaced primarily with fat in early adult- central mediastinum, usually from adjacent inflammation
hood. However, normal thymic tissue may be apparent in such as tuberculous lymphadenitis.
middle-age adults as well. On CT, the normal thymus con-
sists of two triangular or ellipsoid lobes, one typically Neurogenic Neoplasm
larger than the other. A trapezoidal configuration can also
occur, particularly in young women. Strands of fat cours- Neurogenic neoplasms are the most common posterior me-
ing through normal thymic tissue are often apparent. diastinal masses. The vast majority of mediastinal neuro-
How To Approach Incidental Findings on Computed Tomography Images of the Lungs Obtained in Hybrid Imaging 195
genic neoplasms are benign and include schwannoma and MAC can be exuberant, with milk of calcium deposition
neurofibromas. Neurofibromas commonly course along and a tumefactive appearance. MAC is typically an inci-
the intercostal nerves and can lead to bone remodeling. dental finding and is only rarely associated with dys-
Schwannomas may be seen extending into and widening rhythmias. However, studies have correlated the presence
the neural foramen, taking on a dumbbell appearance. of MAC with an increased risk of atherosclerotic disease
These lesions are best characterized with dedicated MRI. of the coronary arteries.
The chest wall consists of the bones of the thorax, the Soft Tissues
supporting and surrounding musculature, subcutaneous
fat, and the skin. Incidental findings, particularly those of Lipoma
the skeleton, are common and the majority of these find-
ings require neither mentioning nor further workup. Chest wall lipomas are commonly identified on cross-
sectional CT. Most are small and are clinically silent.
Bones They are commonly seen within a chest wall muscle as
an encapsulated, homogeneous fat-attenuation mass.
The ribs, thoracic vertebrae, clavicles, sternum (includ-
ing manubrium and xyphoid) constitute the bones of Sebaceous Cyst
the chest wall. The bones of the shoulder girdle (scapu-
la and proximal humerus) are typically included, as Sebaceous cysts are occasionally seen on cross-sectional
well. imaging of the chest. They may result from inflammation
How To Approach Incidental Findings on Computed Tomography Images of the Lungs Obtained in Hybrid Imaging 197
or infection of a sebaceous gland, resulting in obstructing Einstein AJ, Johnson LL, Bokhari S et al (2010) Agreement of vi-
of the normal drainage route. On CT, they have a defined sual estimation of coronary artery calcium from low-dose CT
attenuation correction scans in hybrid PET/CT and SPECT/CT
capsule, central low attenuation, and occur immediately with standard Agatston score. Journal of the American College
deep to the skin. They can become infected or inflamed of Cardiology 56:1914-1921.
and may have increased FDG uptake. Targeted ultrasound Elicker B, Pereira CA, Webb R, Leslie KO (2008) High-resolution
can be used if the diagnosis is uncertain. computed tomography patterns of diffuse interstitial lung dis-
ease with clinical and pathological correlation. Jornal
brasileiro de pneumologia: publicacao oficial da Sociedade
Breast Tissue Brasileira de Pneumologia e Tisilogia 34:715-744.
Holloway BJ, Rosewarne D, Jones RG (2011) Imaging of thoracic
CT does not perform well in evaluating breast tissue. aortic disease. Br J Radiol 84 Spec No 3:S338-354.
Abnormalities encountered on CT are best evaluated Jeong YJ, Kim KI, Seo IJ et al (2007) Eosinophilic lung diseases:
with mammography with or without additional imaging a clinical, radiologic, and pathologic overview. Radiographics
27:617-637; discussion 637-619.
with breast ultrasound or MRI. Patients with abnormal MacMahon H, Austin JH, Gamsu G et al (2005) Guidelines for
hybrid imaging findings in the breast, such as soft-tis- management of small pulmonary nodules detected on CT
sue nodules or fluid collections, should be referred for scans: a statement from the Fleischner Society. Radiology
dedicated breast imaging if it has not been recently per- 237:395-400.
formed. Naidich DP, Bankier AA, MacMahon H et al (2013) Recommen-
dations for the management of subsolid pulmonary nodules
detected at CT: a statement from the Fleischner Society. Radio-
logy 266:304-317.
Suggested Reading Roberton BJ, Hansell DM (2011) Organizing pneumonia: a kalei-
doscope of concepts and morphologies. European Radiology
Andreu J, Mauleon S, Pallisa E et al (2002) Miliary lung disease 21:2244-2254.
revisited. Current problems in diagnostic radiology 31:189- Sayyouh M, Vummidi DR, Kazerooni EA (2013) Evaluation and
197. management of pulmonary nodules: state-of-the-art and future
Bogaert J, Francone M (2013) Pericardial disease: value of CT and perspectives. Expert Opin Med Diagn 7:629-644.
MR imaging. Radiology 267:340-356. Seaman DM, Meyer CA, Gilman MD, McCormack FX (2011) Dif-
Cox CW, Rose CS, Lynch DA (2014) State of the art: Imaging of fuse cystic lung disease at high-resolution CT. AJR Am J
occupational lung disease. Radiology 270:681-696. Roentgenol 196:1305-1311.
Date H (2009) Diagnostic strategies for mediastinal tumors and Shahrzad M, Le TS, Silva M (2014) Anterior mediastinal masses.
cysts. Thoracic surgery clinics 19:29-35, vi. AJR Am J Roentgenol 203:W128-138.
IDKD 2015-2018
and its high accuracy compared to invasive coronary an- individual myocardial-wall territories with impaired myo-
giography has been demonstrated [12]. With the increas- cardial perfusion by the subtending coronary artery was
ing acceptance of CCTA, the associated radiation expo- initially described by Namdar and colleagues [22]. The
sure has been the focus of growing attention. However, added diagnostic clinical value beyond that of either tech-
the introduction of prospective ECG-triggering, including nique alone or that of side-by-side analysis has been
high-pitch spiral acquisition, has led to a substantial dose demonstrated [23, 24]. In low-risk populations, hybrid
reduction [13, 14]. Recently, iterative reconstruction al- imaging may increase the confidence to rule out CAD; for
gorithms for CCTA have demonstrated significant noise example, in the stepwise evaluation of CAD, when equiv-
reduction, allowing tube current and tube voltage reduc- ocal results were obtained with the first study and a sec-
tion without degradation of image quality such that in ond modality is needed to finally rule out disease. Many
some settings CCTA can be performed using a radiation of these patients would end up having invasive coronary
exposure similar to a chest X-ray [15]. angiography, whereas hybrid imaging increases diagnostic
confidence by avoiding equivocal findings, which helps to
Functional Assessment: MPI with SPECT reduce the number of patients unnecessarily exposed to
the risk associated with invasive coronary procedures. It
MPI with SPECT represents the most widely available, was recently shown that cardiac hybrid imaging (by
robust, and by far best established noninvasive method SPECT and CCTA) in CAD evaluation has a profound
for the evaluation of ischemic heart disease [16]. A per- impact on patient management and may contribute to op-
fect agreement of MPI SPECT and coronary angiogra- timal downstream resource utilization [25]. At the other
phy, however, is inherently precluded because the main end of the spectrum, i.e., in older patients, who often suf-
role of SPECT is not to correctly predict or exclude epi- fer from multivessel disease with more jeopardized myo-
cardial coronary lesions but rather to evaluate the physio- cardium, hybrid imaging provides important comprehen-
logical relevance of a given lesion. In fact, only 40% of sive information to allow timely and appropriate treat-
coronary lesions with a luminal narrowing of 5070% in- ment. In these settings, the value of hybrid imaging lies
duce ischemia, as documented in the FAME sub study by far beyond the simple addition of a further diagnostic test,
Tonino et al. [17]. Despite this wisdom, SPECT MPI has as it allows the accurate spatial association of perfusion
often been compared with findings of coronary angio- defects with their subtending coronary stenosis (Fig. 1).
graphy, as this is the generally accepted standard of refer- The CT part of hybrid imaging has an excellent abili-
ence for coronary lesions. One of the largest reports, the ty to rule out anatomic CAD, but an abnormal CCTA
British Royal Brompton and UCL study of Thallium (or an abnormal conventional angiography) is a poor pre-
and Technetium (ROBUST), assessed 2,560 patients who dictor of ischemia. Therefore, MPI testing is recom-
were randomized to one of the commonly used tracers mended to identify patients who might benefit from a
(thallium, sestamibi, or tetrofosmin) during stress, main- revascularization procedure, i.e., those with an ischemic
ly adenosine-induced. This study documented a sensitiv-
ity of 91% and a specificity of 87% without differences
among the tracers [18].
Several attempts to improve MPI by using iterative re-
construction algorithms, early-imaging protocols, or dif-
ferent tracers have provided valuable benefits. However,
the novel CZT detector systems have the potential to rep-
resent the much awaited milestone in the technical im-
provement of MPI. The CZT technique has enabled a
substantial miniaturization of the detectors, a develop-
ment exploited by new cameras in that a stationary posi-
tioning of numerous CZT detectors can be chosen, with
pinhole geometry around the heart. As a result, the scan
on the CZT camera covers the entire heart, allowing very
fast acquisition times, comparable to those of cardiac
PET imaging. The first clinical studies reported very en-
couraging results [19], a high accuracy for the detection
of angiographically documented CAD [20], and the fea-
sibility of fast acquisition protocols [21].
burden 10% [26, 27]. The technologic refinements im- significant coronary artery disease in patients undergoing
plemented in the latest generation of CT scanners have coronary computed tomographic angiography: results from the
multinational coronary CT angiography evaluation for clinical
reduced the number of non-evaluable coronary segments, outcomes: an international multicenter registry (CONFIRM).
and further improvements may be expected. However, the Circulation 124:2423-2432, 2421-2428.
two pieces of information obtained with perfusion imag- 10. Bartunek J, Sys SU, Heyndrickx GR et al (1995) Quantitative
ing versus morphology are difficult to compare and will coronary angiography in predicting functional significance of
likely remain complementary. By contrast, the receiver stenoses in an unselected patient cohort. J Am Coll Cardiol
26:328-334.
operator characteristic analysis for detecting perfusion 11. Windecker S, Kolh P, Alfonso F et al (2014) 2014 ESC/EACTS
defects (by SPECT) has been shown to result in a similar Guidelines on myocardial revascularization: The Task Force on
area under the curve for the reference standard (conven- Myocardial Revascularization of the European Society of Car-
tional angiography) as for CCTA, documenting the com- diology (ESC) and the European Association for Cardio-Tho-
racic Surgery (EACTS) Developed with the special contribution
parable performance and limitations of both anatomic of the European Association of Percutaneous Cardiovascular In-
and morphologic techniques [28]. terventions (EAPCI). EuroIntervention [Epub ahead of print].
Results from a multicenter study emphasize the value 12. Budoff MJ, Dowe D, Jollis JG et al (2008) Diagnostic perfor-
of a combined functional and anatomical approach even mance of 64-multidetector row coronary computed tomo-
without image fusion, i.e., without creating a hybrid im- graphic angiography for evaluation of coronary artery stenosis
in individuals without known coronary artery disease: results
age, showing that this combination allows improved risk from the prospective multicenter ACCURACY (Assessment
stratification [29]. The added value of hybrid imaging by Coronary Computed Tomographic Angiography of Individ-
seems most pronounced for functionally relevant lesions uals Undergoing Invasive Coronary Angiography) trial. J Am
in distal segments and diagonal branches and in vessels Coll Cardiol 52:1724-1732.
with extensive coronary lesions of heavy calcification on 13. Buechel RR, Husmann L, Herzog BA et al (2011) Low-dose
computed tomography coronary angiography with prospective
CCTA. The prognostic value of cardiac hybrid imaging electrocardiogram triggering: feasibility in a large population.
has been confirmed, and matched defects on hybrid J Am Coll Cardiol 57:332-336.
imaging have been shown to be a strong predictor of ma- 14. Achenbach S, Marwan M, Ropers D et al (2010) Coronary
jor adverse cardiovascular events [30]. Cardiac hybrid computed tomography angiography with a consistent dose be-
imaging in CAD evaluation may have the potential to op- low 1 mSv using prospectively electrocardiogram-triggered
high-pitch spiral acquisition. Eur Heart J 31:340-346.
timize the downstream resource utilization [25]. 15. Fuchs TA, Stehli J, Bull S et al (2014) Coronary computed to-
mography angiography with model-based iterative reconstruc-
tion using a radiation exposure similar to chest X-ray exami-
References nation. Eur Heart J 35:1131-1136.
16. Underwood SR, Anagnostopoulos C, Cerqueira M et al (2004)
1. Flotats A, Knuuti J, Gutberlet M et al (2011) Hybrid cardiac Myocardial perfusion scintigraphy: the evidence. Eur J Nucl
imaging: SPECT/CT and PET/CT. A joint position statement Med Mol Imaging 31:261-291.
by the European Association of Nuclear Medicine (EANM), 17. Tonino PA, Fearon WF, De Bruyne B et al (2010) Angio-
the European Society of Cardiac Radiology (ESCR) and the graphic versus functional severity of coronary artery stenoses
European Council of Nuclear Cardiology (ECNC). Eur J Nucl in the FAME study fractional flow reserve versus angiography
Med Mol Imaging 38:201-212. in multivessel evaluation. J Am Coll Cardiol 55:2816-2821.
2. Dorbala S, Di Carli MF, Delbeke D et al (2013) SNMMI/AS- 18. Kapur A, Latus KA, Davies G et al (2002), A comparison of
NC/SCCT guideline for cardiac SPECT/CT and PET/CT 1.0. three radionuclide myocardial perfusion tracers in clinical
J Nucl Med 54:1485-1507. practice: the ROBUST study. Eur J Nucl Med Mol Imaging
3. Hodler J, Schulthess GK, Zollikofer CL (2011) Diseases of the 29:1608-1616.
Heart, Chest & Breast. Diagnostic Imaging and Interventional 19. Buechel RR, Pazhenkottil AP, Herzog BA et al (2010) Real-
Techniques. Springer, pp 199-202. time breath-hold triggering of myocardial perfusion imaging
4. Achenbach S, Daniel WG (2001) Noninvasive coronary an- with a novel cadmium-zinc-telluride detector gamma camera.
giographyan acceptable alternative? N Engl J Med Eur J Nucl Med Mol Imaging 37:1903-1908.
345:1909-1910. 20. Fiechter M, Ghadri JR, Kuest SM et al (2011) Nuclear my-
5. Shaw LJ, Shaw RE, Merz CN et al (2008) Impact of ethnicity ocardial perfusion imaging with a novel cadmium-zinc-tel-
and gender differences on angiographic coronary artery dis- luride detector SPECT/CT device: first validation versus inva-
ease prevalence and in-hospital mortality in the American Col- sive coronary angiography. Eur J Nucl Med Mol Imaging
lege of Cardiology-National Cardiovascular Data Registry. 38:2025-2030.
Circulation 117:1787-1801. 21. Herzog BA, Buechel RR, Katz R et al (2010) Nuclear my-
6. White CW, Wright CB, Doty DB et al (1984) Does visual in- ocardial perfusion imaging with a cadmium-zinc-telluride de-
terpretation of the coronary arteriogram predict the physiolog- tector technique: optimized protocol for scan time reduction. J
ic importance of a coronary stenosis? N Engl J Med 310:819- Nucl Med 51:46-51.
824. 22. Namdar M, Hany TF, Koepfli P et al (2005) Integrated
7. Uren NG, Melin JA, De Bruyne B et al (1994) Relation be- PET/CT for the assessment of coronary artery disease: a fea-
tween myocardial blood flow and the severity of coronary- sibility study. J Nucl Med 46:930-935.
artery stenosis. N Engl J Med 330:1782-1788. 23. Gaemperli O, Kaufmann PA. (2008), Hybrid cardiac imaging:
8. Patel MR, Peterson ED, Dai D et al (2010) Low diagnostic more than the sum of its parts? J Nucl Cardiol 15:123-126.
yield of elective coronary angiography. N Engl J Med 24. Gaemperli O, Schepis T, Valenta I et al (2007) Cardiac image
362:886-895. fusion from stand-alone SPECT and CT: clinical experience. J
9. Cheng VY, Berman DS, Rozanski A et al (2011) Performance Nucl Med 48:696-703.
of the traditional age, sex, and angina typicality-based ap- 25. Fiechter M, Ghadri JR, Wolfrum M et al (2012) Downstream
proach for estimating pretest probability of angiographically resource utilization following hybrid cardiac imaging with an
Integrated Cardiac Imaging 201
integrated cadmium-zinc-telluride/64-slice CT device. Eur J 28. Gaemperli O, Schepis T, Valenta I et al (2008) Functionally rel-
Nucl Med Mol Imaging 39:430-436. evant coronary artery disease: comparison of 64-section CT
26. Hachamovitch R, Hayes SW, Friedman JD et al (2003) Com- angiography with myocardial perfusion SPECT. Radiology
parison of the short-term survival benefit associated with 248:414-423.
revascularization compared with medical therapy in patients 29. van Werkhoven JM, Schuijf JD, Gaemperli O et al (2009)
with no prior coronary artery disease undergoing stress my- Prognostic value of multislice computed tomography and gat-
ocardial perfusion single photon emission computed tomogra- ed single-photon emission computed tomography in patients
phy. Circulation 107:2900-2907. with suspected coronary artery disease. J Am Coll Cardiol
27. Shaw LJ, Berman DS, Maron DJ et al (2008) Optimal medical 53:623-632.
therapy with or without percutaneous coronary intervention to 30. Pazhenkottil AP, Nkoulou RN, Ghadri JR et al (2011) Prog-
reduce ischemic burden: results from the Clinical Outcomes Uti- nostic value of cardiac hybrid imaging integrating single-pho-
lizing Revascularization and Aggressive Drug Evaluation ton emission computed tomography with coronary computed
(COURAGE) trial nuclear substudy. Circulation 117:1283-1291. tomography angiography. Eur Heart J 32:1465-1471.
IDKD 2015-2018
Anterior mediastinal tumors that may be encountered on thymoma, and thymic carcinoma. Of these, lymphomas that
18F-fluorodeoxyglucose-positron emission tomography/ involve the mediastinum are the most common.
computed tomography (FDG-PET/CT) scans include ma- Lymphomas that involve the anterior mediastinum
lignant lymphoma, benign and malignant germ cell tumors, may be Hodgkin lymphoma (HL) (Figs. 1, 2) or non-
a b c d
e f g
a b c
Fig. 2 a-c. A 31-year-old male with nodular sclerosing Hodgkin lymphoma. a Staging PET-CT was performed. MIP image (b) shows a hy-
permetabolic mediastinal mass with lower cervical, mediastinal, and diaphragmatic lymphadenopathy (SUVmax11.7). He was treated
with ABVD (adriamycin, bleomycin, vinblastine and darcarbacine) 8 and radiotherapy. Subsequent recurrent disease in the left lower
chest wall was treated with DHAP (dexamethasone, cytarabine, cisplatin) and autologous bone marrow transplantation (BMT). PET-CT
was performed post-BMT. MIP image (b) and fused axial PET-CT image (c) show a residual mediastinal mass with minimal residual up-
take (Deauville 5-point score2). Punch biopsy showed fibrofatty tissue only, with no malignancy
a b c
d e
Fig. 3 a-e. A 37-year-old female with primary mediastinal B-cell lymphoma. Staging PET/CT was performed. MIP image (a), contrast-en-
hanced CT (b) and fused axial PET-CT image (c) show a heterogeneously contrast-enhancing mass in the anterior and middle mediastinum
that extends from the thoracic inlet to the lower end of the sternum. The lesion shows central necrosis and peripheral contrast enhancement
(SUVmax10.6). The end-chemotherapy PET scan shows a complete metabolic response with residual mass. The mediastinal mass was
treated with radiotherapy (RT). PET-CT was performed 3 months after end of RT. MIP image (d) and fused axial PET-CT image (e) show
suspicious focal uptake in the mass. Follow-up PET-CT 3 months later shows resolution of the focal uptake suggesting false positive find-
ings in the previous scan (d, e)
Hodgkin lymphoma subtypes, with the latter including zone lymphoma (B-cell lymphoma, unclassifiable, with
diffuse large B-cell lymphoma (DLBCL), primary me- features intermediate between DLBCL and classical
diastinal B-cell lymphoma (PMBCL) (Fig. 3), gray- HL), and T-cell lymphoblastic lymphoma [1] (Fig. 4).
204 P.L. Khong
a b
b
Fig. 4 a-c. A 10-year-old male with T-lymphoblastic lymphoma. Staging PET/CT scan was performed. MIP image (a), fused axial PET-CT
image of 2 sections (b, c) show a hypermetabolic anterior mediastinal mass (SUVmax6.4). The patient responded poorly to chemother-
apy and died from progressive disease 2 years later
In about 5% of these cases, the anterior mediastinum is before consideration of radiotherapy (RT), because the
the only site of disease and may manifest as isolated PET/CT findings may have an impact on RT planning
thymic involvement, as isolated nodal involvement, or and design of the RT field. A single nodal mass/medi-
as a combination of both [2]. HL accounts for the ma- astinal mass should be measured on the CT scan across
jority of mediastinal lymphomas. Lymphoblastic lym- the longest measurement. In HL, tumor bulk 10 cm is
phomas mainly occur in children and adolescents, while defined as bulky disease and has prognostic and thera-
DLBCL is usually seen in young to middle-aged adults. peutic implications. Moreover, contrast-enhanced CT, if
Diagnosis is made based on the pathological examina- not already performed, should ideally be done as part of
tion, with mandatory immunohistochemical staining, a one-stop approach during a single imaging session
and a typical clinical presentation. PMBCL belongs to combined with PET/CT [6]. Baseline findings should be
the group of aggressive DLBCLs but was distinguished used to determine whether contrast-enhanced CT or low-
as a separate entity in the WHO 2008 classification due er dose and non-enhanced PET/CT will suffice for addi-
to its specific clinical and pathological features, name- tional imaging examinations. A bone marrow biopsy is
ly, gene expression profile studies showing common no longer indicated for routine staging of HL or most
features with classical HL and thus having an impact on DLBCLs [7].
the choice of therapy [1]. Since all the above-mentioned For response assessment, conventional morphologic
lymphomas that involve the mediastinum fall into the imaging using CT or magnetic resonance imaging was
subtypes are FDG-avid [3], FDG-PET/CT is the imag- shown to lack specificity in the characterization of resid-
ing modality of choice for staging and treatment re- ual masses that often remain after treatment. However,
sponse assessment. FDG-PET allows better discrimination between viable
For disease staging, compared to conventional CT, tissue and fibrotic residual masses, by showing the al-
FDG-PET/CT may lead to a change in disease stage, tered metabolism of the former (Figs. 13). Thus, FDG-
more often upstaging [4, 5]. This is particularly important PET/CT was incorporated into the 2007 International
FDG PET/CT in the Imaging of Mediastinal Masses 205
Working Group guidelines for response assessment of Early treatment assessment is potentially of impor-
these malignant lymphomas [8]. In the previous guide- tance for response-adapted treatment, both to optimize
lines, visual interpretation of positive or negative treatment and to avoid unnecessary toxicity. For HL, it
PET/CT scans was based on comparison with the medi- has been found that early metabolic changes are highly
astinal blood pool for end-of-treatment response assess- predictive of final treatment response and progression-
ment [8]. For HL, end-of-treatment scans were shown to free survival [9]. Currently, based on interim PET find-
carry a high negative predictive value (80100%), al- ings after two courses of ABVD (adriamycin, bleomycin,
though the positive predictive value is more variable vinblastine, and darcarbacine) chemotherapy, early treat-
(25100%). In the recent updated recommendation, a 5- ment intensification with BEACOPPesc (bleomycin,
point scale (5-PS) based on visual assessment is used to etoposide, adriamycin, cyclophosphamide, vincristine,
assess response to PET/CT [7]. The 5-PS score assess- procarbazine, and prednisone escalated regimen) should
es uptake at the site of initial disease as follows: 1, no be instituted in patients with a positive PET scan. Multi-
uptake; 2, uptake mediastinum, 3, uptake medi- ple clinical trials are on-going to determine the value of
astinum but liver; 4, uptake moderately higher than early PET in response-adapted therapy for the identifica-
liver; 5, uptake markedly higher than liver (2-to 3-fold tion of patients with poor treatment response who may
higher than a large region of normal liver) and/or new benefit from escalation to more aggressive therapy, and
lesions. In the recommendation, a score of 1 or 2 repre- for patients with good treatment response who can be
sents a complete metabolic response (CMR); a score 3 cured with less than standard therapy, including omission
of represents a probable CMR in patients receiving stan- of RT to the mediastinum.
dard treatment (Fig. 1); and a score of 4 or 5 with re- In children, FDG accumulation in the thymus is a com-
duced uptake from baseline is considered a partial meta- mon, normal finding and thymic rebound hyperplasia can
bolic response, but if the assessment is at the end of be observed after chemotherapy (Fig. 5), which may be a
treatment it represents residual metabolic disease. An potential pitfall in the interpretation of FDG-PET/CT in
increase in FDG uptake to a score of 5, or a score of 5 pediatric patients. Thymic rebound may reflect a hemato-
with no decrease in uptake, and new FDG-avid foci con- logical rebound phenomenon, characterized by lymph
sistent with lymphoma represent treatment failure follicles with large nuclear centers and plasma cell infil-
and/or progression. tration. Although thymic rebound usually appears within
For interim assessment, patients should undergo scan- 612 months after cessation of chemotherapy, it can
ning as long after the previous chemotherapy administra- develop over a period as short as 1 week. The time course
tion as possible, since non-specific FDG uptake may oc- of FDG uptake in the thymus is such that it is usually not
cur with treatment-related inflammation. A minimum of apparent at cessation of therapy, with uptake reaching a
3 weeks but preferably 68 weeks after the completion of peak within 12 months after treatment (10 months in a
the last chemotherapy cycle, 2 weeks after granulocyte few reported series), with a slow decline thereafter [10,
colony-stimulating factor treatment, or 3 months after RT 11]. Moreover, there is often a concurrent increase in
is recommended (Fig. 3). bone marrow uptake due to increased bone marrow
a c
a b c
d e f
Fig. 7 a-f. PET-CT scans from patients with thymoma. a A 61-year-old female with a history of ovarian cancer. Fused axial PET-CT image
shows an anterior mediastinal nodule with low grade uptake (SUVmax1.6) which was detected as an incidental finding. Biopsy showed
a type AB thymoma with microscopic foci of capsular invasion. b A 51-year-old female with an incidental finding of an anterior medi-
astinal mass (SUVmax2.0). Biopsy showed a type AB thymoma. c-f A 45-year-old female with invasive thymoma type B2. MIP image
(c) and fused axial PET-CT images at 3 sections of the mediastinum (d-f) show a moderately hypermetabolic anterior mediastinal mass
with mediastinal invasion (d) and extension to the left posterior pleura and hemidiaphragm (e, f) (SUVmax5.2)
FDG PET/CT in the Imaging of Mediastinal Masses 207
a b
c
Fig. 8 a-c. PET-CT scan of a 46-
year-old male with metastatic
thymic carcinoma who presented
with hoarseness and dyspnea. MIP
image (a), contrast-enhanced CT
image (b) and fused axial PET-CT
image (c) show the invasive tumor
has a necrotic center (SUVmax
12.7). Additional nodules were
detected in the lower anterior me-
diastinum; skeletal metastases
were also present (a)
a b
is very rare. Thymic carcinomas are aggressive in ap- for up front surgery, and those with high-risk tumors,
pearance, can be associated with vessel invasion and/or who will need to have an open biopsy for histological
the invasion of mediastinal fat, and may present with diagnosis before surgery and for neoadjuvant therapy, if
nodal and distant metastases [13] (Fig. 8). FDG uptake appropriate. Cystic changes may be found in thymomas,
is significantly higher in thymic carcinoma than in thy- thymic carcinoma, and germ-cell tumors (Fig. 9) and
moma and can be used to differentiate between high- rarely in lymphoma before treatment. Calcifications are
grade and low-grade thymomas [14, 15]. This can aid in sometimes seen in lymphoma post-therapy and in ter-
selecting patients with low-risk tumors, who are eligible atomas, which may also contain fat. Clinical history,
208 P.L. Khong
e.g., B-cell symptoms in malignant lymphoma, the as- 7. Cheson BD, Fisher RL, Barrington SF et al (2014) Recommen-
sociation of myasthenia gravis in thymic epithelial tu- dations for initial evaluation, staging and response assessment
of Hodgkin and Non-Hodgkin lymphoma: The Lugano Classi-
mors, and biochemical tests are helpful, e.g., elevated fication. J Clin Oncol 32:3059-3067.
serum -fetoprotein levels in patients with malignant 8. Cheson BD, Pfistner B, Juweid ME et al (2007) Revised re-
germ-cell tumors. sponse criteria for malignant lymphoma. J Clin Oncol 25:579-
586.
9. Biggi A, Gallamini A, Chauvie S et al (2013) International val-
idation study for interim PET in ABVD-treated advanced-
References stage Hodgkin lymphoma: interpretation criteria and concor-
dance rate among reviewers. J Nucl Med 54: 683-690.
1. Dabrowska-Iwanicka A, Walewski JA (2014) Primary medi- 10. Goethals I, Hoste P, Vriendt CD et al (2010) Time-dependent
astinal large B-cell lymphoma. Curr Hematol Malig Rep changes in 18F-FDG activity in the thymus and bone marrow
9:273-283. following combination chemotherapy in paediatric patients
2. Priola AM, Galetto G, Priola SM (2014) Diagnostic and func- with lymphoma. Eur J Nucl Med Mol Imaging 37:462-467.
tional imaging of thymic and mediastinal involvement in lym- 11. Gawande RS, Khurana A, Messing S et al (2012) Differentia-
phoproliferative disorders. Clin Imaging 38:771-784. tion of normal thymus from anterior mediastinal lymphoma
3. Weiler-Sagie M, Bushelev O, Epelbaum R et al (2010) F-18- and lymphoma recurrence at pediatric PET/CT. Radiology
FDG avidity in lymphoma readdressed: a study of 766 pa- 262:613-622.
tients. J Nucl Med 51:25-30. 12. Quint LE (2007) Imaging of anterior mediastinal masses. Can-
4. Schaefer NG, Hany TF, Taverna C et al (2004) Non-Hodgkin cer Imaging 7:S56-S62.
lymphoma and Hodgkin disease: coregistered FDG PET and 13. Lococo F, Cesario A, Okami J et al (2013) Role of combined
CT at staging and restagingdo we need contrast-enhanced 18F-FDG-PET/CT for predicting the WHO malignancy grade
CT? Radiology 232:823-829. of thymic epithelial tumors: a multicenter analysis. Lung Can-
5. Isasi CR, Lu P, Blaufox MD (2005) A metaanalysis of 18F-2- cer 82:245-251.
deoxy-2-fluoro-D-glucose positron emission tomography in 14. Luzzi L, Campione A, Gorla A (2009) Role of fluorine-
the staging and restaging of patients with lymphoma. Cancer flurodeoxyglucose positron emission tomography/computed
104:1066-1074. tomography in preoperative assessment of anterior mediastinal
6. Barrington SF, Mikhaeel NG, Kostakoglu L et al (2014) Role masses. Eur J Cardiothorac Surg 36:475-479.
of imaging in the staging and response assessment of lym- 15. Treglia G, Sadeghi R, Giovanella L et al (2014) Is 18F-FDG
phoma: consensus of the International conference on Malig- PET useful in predicting the WHO grade of malignancy in
nant lymphomas imaging working group. J Clin Oncol thymic epithelial tumors? A meta-analysis. Lung Cancer 86:
32:3048-3058. 5-13.
IDKD 2015-2018
Fig. 1. Attenuation correction in PET/MRI. The attenuation map is generated using fat and water images from a DIXON magnetic reso-
nance sequence. The -map only differentiates between air, lung, fat, and soft tissue. Using this -map NAC (non-attenuation-corrected)
PET images are then corrected according to the respective tissue-specific attenuation coefficients
ways to generate a -map had to be found [6], resulting The first application of this approach was in imaging
in the development of different approaches. of the head, where a rigid model is sufficient [17, 18].
1. The segmentation-based approach using individual For whole-body imaging, however, an elastic model is
attenuation coefficients for specific tissue classes; necessary, which remains a problem regarding the
this approach was already described in 1981 [10]. It translation into clinical use. The combined use of a
was adopted by our group, who proposed using it for template- and segmentation-based approach has been
the generation of a -map for whole-body PET/MRI described [19].
[11]. By making use of water- and fat-weighted im- 3. It is also possible to generate -maps using PET
ages from a Dixon magnetic resonance sequence, emission data [20, 21]. The underlying principle of
each voxel can be assigned to one of the following tis- this approach is that any voxel exhibiting at least some
sue classes: air, lung, fat, or soft tissue [12] (Fig. 1). tracer uptake has to be part of the body and must thus
Cortical bone is ignored because it is virtually impos- add to the attenuation. An undesirable property of
sible to segment this tissue class from these data. Da- many tracers, namely, their unspecific uptake in non-
ta acquisition usually requires about 18 s per bed po- targeted tissues, is thus a prerequisite for this tech-
sition during one breath-hold. The obvious major dis- nique. This approach is particularly useful in combi-
advantage of this approach is the limited number of nation with the segmentation-based approach, to ad-
tissue classes with fixed attenuation coefficients. This dress motion and truncation problems when body
is particularly relevant in cardiac imaging, as there is parts are outside the field of view (FOV) [7, 22]. Ac-
significant inter-patient variability in lung structure tually, this combined approach is used in the Siemens
and location. The neglecting of bone may result in dif- Biograph mMR, where the FOV of the magnetic res-
ferences in tumor uptake between PET/CT and onance device is only 45 cm [22, 23]. The extension
PET/MRI that can be as high as 23% [11, 13, 14]. of the field of view is the most obvious solution to
Other disadvantages of this approach are that contrast this problem [24].
agents may cause inaccuracies due to reduced T1 val- Another source of photon scatter and attenuation in
ues, and metal implants do not contribute to the at- PET/MRI are artificial components outside the body but
tenuation map [15]. This segmentation-based ap- within the FOV, such as coils, patient monitoring devices,
proach is used in currently commercially available positioning aids, earphones, cables, and the patient bed it-
PET/MRI scanners, such as the Biograph mMR [11] self. For the MRI system these objects are invisible and
or the Ingenuity TF [13]. Our group demonstrated a thus do not contribute to the -map. Only the patient bed
good correlation between PET/MRI and PET/CT da- and the fixed coils are implanted in the precalculated at-
ta in oncologic studies [16]. tenuation models [25]. Consequently, vendors have re-
2. The second route to -map generation is the template- designed their hardware, although the effects are still de-
(or atlas-) based approach. Here, the model-based tectable. Whether there is a relevant impact on diagnostic
known -map is adjusted to the patients real anatomy. scans is not yet known [26].
Integrated Cardiovascular PET/MR: Lessons Learned 211
Cardiac Devices volume data have to be coregistered with the MRI data in
various (partially overlapping) slices of various positions
Cardiac devices, such as pacemakers, implantable car- (e.g., short axes; two-, three-, and four-chamber view).
dioverter defibrillators (ICDs), or cardiac resynchroniza- However, researchers are aware of this problem and tech-
tion therapy (CRT), cause artifacts on CT that result in an niques such as motion-triggered acquisition [32] and
overestimation of the attenuation. The effect on the quan- software corrections [33] are being developed that allow
tification of tracer uptake seems to be negligible in free-breathing during imaging.
PET/CT [27]. Analogously, non-magnetic metals cause a Still, parallel PET/MRI is much more favorable than
signal void in MRI that exceeds the actual size of the ob- sequential PET/MRI. The sequential approach (either on
ject, possibly resulting in an underestimation of the at- separate scanners or on scanners connected via a com-
tenuation. More importantly, these devices may interact mon rail system) is not only inconvenient for the patient
with the radiofrequency, which may result in their mal- and personnel; it is also logistically demanding. Parallel
function or heating. Therefore, many subjects cannot be PET/MRI, on the other hand, can not only improve pa-
examined using MRI (and thus by PET/MRI); this is a tient compliance and comfort, it can also increase patient
growing problem because the number of patients with throughput and thus cost-efficiency. Two potential work
diseases qualifying for the implantation of such cardiac flows are depicted in Fig. 2.
devices is increasing [28, 29]. Phase analysis using nu- An application that might benefit from the truly si-
clear medicine studies (such as PET or SPECT) has been multaneous acquisition of PET and MRI data is myocar-
proposed to better identify patients who might respond to dial perfusion imaging (MPI). The initial benefit would
CRT implantation [30, 31]; however, larger clinical trials be the capability to cross-validate one modality with the
are still warranted. other (which is of research interest). Taking into consider-
ation PET, with its vast variety of specific tracers and its
good volume-coverage, and MRI, with its high in-plane
Workflows for Cardiac PET/MRI resolution, a synergistic effect would, for example, be the
separate investigation of epicardial and endocardial perfu-
As mentioned above, attenuation correction is a crucial sion. However, some hurdles still need to be overcome:
aspect of imaging and so is the exact alignment of the - (1) MRI usually only acquires a few slices of the heart,
map and the acquired PET data. One major advantage of (2) Gd-chelate based contrast medium, used in most ap-
MRI over CT is that the attenuation correction scan may plications, has properties that are unfavorable for truly
be repeated as often as necessary without any radiation. quantitative perfusion assessment, such as a low extrac-
Unfortunately, the subsequent necessary choice of the tion fraction and a non-specific (only partially perfusion-
ideal -map is a time-consuming, inconvenient matter. dependent) uptake by the myocardium. The major advan-
According to our experience, however, in the vast major- tage is, however, that PET and MRI data can be acquired
ity of cases there is satisfactory alignment between PET under truly identical conditions, even though the injection
data and the -map. speed of the imaging agent (PET: 30 s, MRI: 5 s) as
Another important consideration regarding PET/MRI well as the acquisition time (MRI: 1 min, PET: 10
is that most data are acquired in parallel and not truly min) differ significantly.
simultaneously. PET acquires a volume with frame
lengths ranging from a few seconds up to 2030 min (or
more) depending on the chosen protocol (tracer, half-life, Cardiovascular Applications Using Hybrid PET/MRI
injected activity, endurance of the patient, etc.). Conse-
quently, the acquired PET data comprise motion that took Myocardial Perfusion Imaging
place during the scan (e.g., heart beat, ventilation and
true patient motion) whereas in MR data are usually ac- The diagnosis of flow-limiting coronary artery disease
quired sequentially (even in 3D acquisitions), with acqui- (CAD) and the investigation of the hemodynamic signif-
sition times ranging from about 50 ms for dynamic scans icance of known CAD are the most regularly conducted
to several breath-holds for high-resolution images. A ma- studies in nuclear cardiology, due to its high sensitivity
jor problem with sequential imaging is that the data are and specificity [34] and its high value in patient man-
acquired over several breath-holds; since the repro- agement [35-37]. PET MPI helps in clinical decision-
ducibility of breath-holding is limited, this results in par- making as the extent of ischemic tissue determined by
tially overlapping (or partially lacking) data. However, this imaging modality allows patients to be assigned to
the imaging of moving objects with MRI would result in optimal therapy, either revascularization or medical ther-
enormous, unacceptable artifacts. With cardiac apy [38]. Furthermore, there is increasing evidence that
PET/MRI, this raises the issue, that ungated (cardiac cy- PET MPI offers advantages over SPECT MPI due to the
cle and ventilation) PET images are compared to or fused absolute quantification of myocardial blood flow (MBF),
with, for example, end-diastolic magnetic resonance im- attenuation correction, and superior image quality, espe-
ages that are acquired during breath-hold. Another in- cially in obese patients [39]. So far, several PET perfu-
convenient and time-consuming matter is that the PET sion tracers have been established: N-13 ammonia, O-15
212 C. Rischpler, S.G. Nekolla, M. Schwaiger
water, rubidium-82 (Rb-82), and F-18 flurpiridaz. The mentioned advantages of PET MPI and its use in many
relatively short half-lives of N-13 ammonia (10 min), centers, MRI has also been evaluated recently for the pur-
O-15 water (122 s) and Rb-82 (76 s) allow rapid PET pose of MPI. The underlying principle that first-pass
MPI with only short lag times between the stress and the MRI with gadolinium-based contrast agents (e.g., Gd-
at-rest studies. The disadvantages of these tracers are, DTPA) can be used to detect CAD was shown about 20
however, the high positron energies, resulting in de- years ago [42]; since then, a multitude of studies have in-
creased image quality (Rb-82), the need of an on-site cy- vestigated its value in flow-limiting CAD [43, 44]. While
clotron (N-13 ammonia, O-15 water) or of a generator, these studies consisted almost exclusively of visual
which is only cost-effective when a certain patient- analysis, the absolute quantification of MBF is known to
throughput can be guaranteed (Rb-82), or the fact that be of great value in patients suffering from extensive
stress imaging is only feasible when the patient is ad- CAD and balanced ischemia [45, 46]. The feasibility of
ministered the pharmacological agent and the tracer quantifying MBF using MRI has been investigated. One
while inside the scanner (N-13 ammonia, O-15 water, study evaluated a rather simple approach, namely, the up-
Rb-82). F-18 flurpiridaz offers some advantages, e.g. a slope ratio as an index of coronary flow reserve [47].
low positron energy, a half-life that is long enough to en- However, in a subsequent study, this approach was defin-
able tracer distribution even to hospitals and clinics in itively shown to result in an underestimation of the flow
rural areas, the option of ergometer exercise with tracer reserve when compared to N-13 ammonia PET MPI [48].
injection outside the scanner, and possibly a simplified Subsequently, a more complex model, using the central
MBF estimation approach [40, 41]. Despite the above- volume principle, was investigated for absolute MBF
Integrated Cardiovascular PET/MR: Lessons Learned 213
a b c
Fig. 3 a-c. FDG PET/MRI viability imaging shortly after acute myocardial infarction (MI). Example from a patient who was imaged a few
days after acute MI by FDG PET/MRI (hyperinsulinemic/euglycemic clamp). Four-chamber views of late gadolinium enhancement (LGE)
MRI (a), FDG PET (c), and fusion of LGE MRI and FDG PET (b) are depicted. Early after acute MI, different patterns of FDG uptake
and LGE transmurality can be observed: (1) normal FDG uptake and no LGE (white arrows), (2) reduced/absent FDG uptake and trans-
mural LGE (black arrows), and (3) reduced FDG uptake and non-transmural LGE (red arrows)
quantification [49]. This approach was recently studied in alternative. Notably, however, unlike FDG PET, this ap-
41 patients who underwent PET and MRI MPI separate- proach images non-viable, scarred myocardium, informa-
ly [50]. Good agreement for the coronary flow reserve tion used to draw conclusions regarding the potential of
between these two modalities was obtained; the correla- non-scarred myocardium to recover. The enrichment of
tion was weak for absolute MBF values, though. It should Gd-chelates in increased extracellular space (such as
be noted, however, that a rare quantification method for scarred myocardium) results in a reduced wash-out of the
N-13 ammonia was applied [51] and that differences in contrast agent compared to remote myocardium which
absolute MBF may also be due to the fact that Gd- may be visualized in cardiac MRI using inversion recov-
chelates do not enter the cells and thus show a distribu- ery sequences, in which the signal of the remote my-
tion only in the interstitial space (also in increased inter- ocardium is nulled [57]. Despite the fundamental differ-
stitial spaces such as scarred myocardium) [52]. ence in these approaches (imaging viable vs. non-viable
There are patients in whom even extremely well-toler- tissue) there is good agreement between the two modali-
ated MRI contrast agents cannot be used and alternative ties [58]. One essential property of MRI is its high in-
approaches, without the use of any contrast medium, are plane resolution (13 mm), which enables the differenti-
being developed. For example, arterial spin labeling ation between transmural (50%) and non-transmural
(ASL), which originates from techniques to determine (50%) scarred myocardium [59]. Another interesting
cerebral blood flow, has been investigated in a preclinical finding is that in patients with suspected CAD but with-
study, where it demonstrated a good correlation with the out known myocardial infarction, even small areas of
results of O-15 water PET [53]. In patients, ASL was able scarred myocardium carry prognostic significance [60].
to measure the increase in blood flow induced by adeno- A similar observation was described in a study compar-
sine [54]. The obvious advantages of ASL are the possi- ing areas of non-transmural LGE with SPECT MPI [61].
bility of repetitive and continuous measurements of MBF Since small areas of infarction may be overlooked by
without need for any contrast agents PET because of partial volume effects, this fact is rele-
In summary, cardiac MPI using simultaneous vant in hybrid PET/MRI, as an improved tissue classifi-
PET/MRI might be valuable to detect subendocardial is- cation can be expected. Our group made similar observa-
chemia by MRI and to quantify absolute MBF using PET. tions in patients shortly after myocardial infarction, in
whom severely reduced FDG uptake was accompanied by
Myocardial Viability Imaging non-transmural LGE signal only (Fig. 3). Accordingly,
hybrid PET/MRI might ultimately result in an improved
Another potential application of PET/MR in cardiology is prediction of wall motion recovery after revasculariza-
myocardial viability imaging. Hypoperfused myocardium tion, when information such as wall motion and thicken-
exhibits a shift of its metabolism from fatty acids towards ing, FDG uptake, perfusion, and LGE transmurality are
glucose, a state called hibernation [55]. Several studies integrated.
have shown that myocardium in this state is prone to re- As mentioned above, the assessment of global and re-
cover after revascularization [56]. Usually, FDG PET is gional wall motion is a relevant surrogate marker for myo-
the method of choice for non-invasive viability imaging cardial vitality and therapy response after revasculariza-
[55]. Recently, cardiac MRI using the late gadolinium en- tion. Several studies have shown that left ventricular pa-
hancement (LGE) technique has emerged as a possible rameters, such as ejection fraction, end-systolic volume,
214 C. Rischpler, S.G. Nekolla, M. Schwaiger
12. Coombs BD, Szumowski J, Coshow W (1997) Two-point 29. Cohen JD, Costa HS, Russo RJ (2012) Determining the risks
Dixon technique for water-fat signal decomposition with B0 of magnetic resonance imaging at 1.5 tesla for patients with
inhomogeneity correction. Magn Reson Med 38:884-889. pacemakers and implantable cardioverter defibrillators. Am J
13. Schulz V, Torres-Espallardo I, Renisch S et al (2011) Auto- Cardiol 110:1631-1636.
matic, three-segment, MR-based attenuation correction for 30. Henneman MM, van der Wall EE, Ypenburg C et al (2007)
whole-body PET/MR data. Eur J Nucl Med Mol Imaging Nuclear imaging in cardiac resynchronization therapy. J Nucl
38:138-152. Med 48:2001-2010.
14. Samarin A, Burger C, Wollenweber SD et al (2012) PET/MR 31. Uebleis C, Ulbrich M, Tegtmeyer R et al (2011) Electrocar-
imaging of bone lesionsimplications for PET quantification diogram-gated 18F-FDG PET/CT hybrid imaging in patients
from imperfect attenuation correction. Eur J Nucl Med Mol with unsatisfactory response to cardiac resynchronization ther-
Imaging 39:1154-1160. apy: initial clinical results. J Nucl Med 52:67-71.
15. Frst S, Souvatzoglu M, Rischpler C (2012) Effects of MR 32. Adluru G, Chen L, Kim SE et al (2011) Three-dimensional
contrast agents on attenuation map generation and cardiac PET late gadolinium enhancement imaging of the left atrium with
quantification in PET/MR. J Nucl Med 53 (Suppl 1):139. a hybrid radial acquisition and compressed sensing. J Magn
16. Drzezga A, Souvatzoglou M, Eiber M et al (2012) First clini- Reson Imaging 34:1465-1471.
cal experience with integrated whole-body PET/MR: compar- 33. Xue H, Zuehlsdorff S, Kellman P et al (2009) Unsupervised
ison to PET/CT in patients with oncologic diagnoses. J J Nu- inline analysis of cardiac perfusion MRI. Medical image com-
cl Med 53:845-855. puting and computer-assisted intervention: MICCAI Interna-
17. Hofmann M, Steinke F, Scheel V et al (2008) MRI-based at- tional Conference on Medical Image Computing and Comput-
tenuation correction for PET/MRI: a novel approach combin- er-Assisted Intervention 12:741-749.
ing pattern recognition and atlas registration. J Nucl Med 34. Klocke FJ, Baird MG, Lorell BH et al (2003) ACC/AHA/AS-
49:1875-1883. NC guidelines for the clinical use of cardiac radionuclide
18. Schreibmann E, Nye JA, Schuster DM et al (2010) MR-based imagingexecutive summary: a report of the American College
attenuation correction for hybrid PET-MR brain imaging sys- of Cardiology/American Heart Association Task Force on
tems using deformable image registration. Medical physics Practice Guidelines (ACC/AHA/ASNC Committee to Revise
37:2101-2109. the 1995 Guidelines for the Clinical Use of Cardiac Radionu-
19. Hofmann M, Bezrukov I, Mantlik F et al (2011) MRI-based at- clide Imaging). J Am Coll Cardiol 42:1318-1333.
tenuation correction for whole-body PET/MRI: quantitative 35. Schwaiger M, Melin J (1999) Cardiological applications of nu-
evaluation of segmentation- and atlas-based methods. J Nucl clear medicine. Lancet 354:661-666.
Med 52:1392-1399. 36. Yoshinaga K, Chow BJ, Williams K et al (2006) What is the
20. Nuyts J, Dupont P, Stroobants S et al (1999) Simultaneous prognostic value of myocardial perfusion imaging using ru-
maximum a posteriori reconstruction of attenuation and activ- bidium-82 positron emission tomography? J Am Coll Cardiol
ity distributions from emission sinograms. IEEE Trans Med 48:1029-1039.
Imaging 18(393-403. 37. Merhige ME, Breen WJ, Shelton V et al (2007) Impact of
21. Censor Y, Gustafson D, Lent A, Tuy H (1979) A new approach myocardial perfusion imaging with PET and (82)Rb on
to the emission computerized tomography problem: Simulta- downstream invasive procedure utilization, costs, and out-
neous calculation of attenuation and activity coefficients. comes in coronary disease management. J Nucl Med
IEEE Trans Nucl Sci 26:2275-2279. 48:1069-1076.
22. Nuyts J, Michel C, Fenchel M et al (2010) Completion of a 38. Hachamovitch R, Hayes SW, Friedman JD et al (2003) Com-
truncated attenuation image from the attenuated PET emission parison of the short-term survival benefit associated with
data. IEEE Nucl Sci Symp Conf Record (NSS/MIC) 2123- revascularization compared with medical therapy in patients
2127. with no prior coronary artery disease undergoing stress my-
23. Delso G, Martinez-Moller A, Bundschuh RA et al (2010) The ocardial perfusion single photon emission computed tomogra-
effect of limited MR field of view in MR/PET attenuation cor- phy. Circulation 107:2900-2907.
rection. Medical physics 37:2804-2812. 39. Flotats A, Bravo PE, Fukushima K et al (2012) (82)Rb PET
24. Blumhagen JO, Ladebeck R, Fenchel M, Scheffler K (2012) myocardial perfusion imaging is superior to (99m)Tc-labelled
MR-based field-of-view extension in MR/PET: B(0) homoge- agent SPECT in patients with known or suspected coronary
nization using gradient enhancement (HUGE). Magn Reson artery disease. Eur J Nucl Med Mol Imaging 39:1233-1239.
Med 70:1047-1057. 40. Rischpler C, Park MJ, Fung GS et al (2012) Advances in PET
25. Delso G, Martinez-Moller A, Bundschuh RA et al (2010) myocardial perfusion imaging: F-18 labeled tracers. Annals of
Evaluation of the attenuation properties of MR equipment for nuclear medicine 26:1-6.
its use in a whole-body PET/MR scanner. Physics in medicine 41. Sherif HM, Nekolla SG, Saraste A et al (2011) Simplified
and biology 55:4361-4374. quantification of myocardial flow reserve with flurpiridaz F
26. MacDonald LR, Kohlmyer S, Liu C et al (2011) Effects of MR 18: validation with microspheres in a pig model. J Nucl Med
surface coils on PET quantification. Medical physics 38:2948- 52:617-624.
2956. 42. Manning WJ, Atkinson DJ, Grossman W et al (1991) First-pass
27. DiFilippo FP, Brunken RC (2005) Do implanted pacemaker nuclear magnetic resonance imaging studies using gadolini-
leads and ICD leads cause metal-related artifact in cardiac um-DTPA in patients with coronary artery disease. J Am Coll
PET/CT? J Nucl Med 46:436-443. Cardiol 18:959-965.
28. Marinskis G, Bongiorni MG, Dagres N et al (2012) Perform- 43. de Jong MC, Genders TS, van Geuns RJ et al (2012) Diag-
ing magnetic resonance imaging in patients with implantable nostic performance of stress myocardial perfusion imaging for
pacemakers and defibrillators: results of a European Heart coronary artery disease: a systematic review and meta-analy-
Rhythm Association survey. Europace 14:1807-1809. sis. Eur Radiol 22:1881-1895.
216 C. Rischpler, S.G. Nekolla, M. Schwaiger
44. Nandalur KR, Dwamena BA, Choudhri AF et al (2007) Diag- with signs or symptoms of coronary artery disease. Circulation
nostic performance of stress cardiac magnetic resonance imag- 113:2733-2743.
ing in the detection of coronary artery disease: a meta-analy- 61. Wagner A, Mahrholdt H, Holly TA et al (2003) Contrast-en-
sis. J Am Coll Cardiol 50:1343-1353. hanced MRI and routine single photon emission computed to-
45. Kajander SA, Joutsiniemi E, Saraste M et al (2011) Clinical mography (SPECT) perfusion imaging for detection of suben-
value of absolute quantification of myocardial perfusion with docardial myocardial infarcts: an imaging study. Lancet
(15)O-water in coronary artery disease. Circ Cardiovasc Imag- 361:374-379.
ing 4:678-684. 62. Bellenger NG, Davies LC, Francis JM (2000) Reduction in
46. Parkash R, deKemp RA, Ruddy TD et al (2004) Potential util- sample size for studies of remodeling in heart failure by the
ity of rubidium 82 PET quantification in patients with 3-ves- use of cardiovascular magnetic resonance. J Cardiovasc Magn
sel coronary artery disease. J Nucl Cardiol 11:440-449. Res 2:271-278.
47. Schwitter J, Nanz D, Kneifel S et al (2001) Assessment of my- 63. Rajappan K, Livieratos L, Camici PG, Pennell DJ (2002) Mea-
ocardial perfusion in coronary artery disease by magnetic res- surement of ventricular volumes and function: A comparison
onance: a comparison with positron emission tomography and of gated PET and cardiovascular magnetic resonance. J Nucl
coronary angiography. Circulation 103:2230-2235. Med 43:806-810.
48. Ibrahim T, Nekolla SG, Schreiber K et al (2002) Assessment 64. Boyd HL, Gunn RN, Marinho NVS et al (1996) Non-invasive
of coronary flow reserve: comparison between contrast-en- measurement of left ventricular volumes and function by gat-
hanced magnetic resonance imaging and positron emission to- ed positron emission tomography. Eur J Nucl Med 23:1594-
mography. J Am Coll Cardiol 39:864-870. 1602.
49. Jerosch-Herold M (2010) Quantification of myocardial perfu- 65. Fink C, Krissak R, Henzler T et al (2011) Radiation dose at
sion by cardiovascular magnetic resonance. J Cardiovasc Magn coronary CT angiography: second-generation dual-source CT
Res 12:57. versus single-source 64-MDCT and first-generation dual-
50. Morton G, Chiribiri A, Ishida M et al (2012) Quantification of source CT. AJR Am J Roentgenol 196:W550-557.
absolute myocardial perfusion in patients with coronary artery 66. Hausleiter J, Meyer T, Hermann F et al (2009) Estimated radi-
disease: comparison between cardiovascular magnetic reso- ation dose associated with cardiac CT angiography. JAMA
nance and positron emission tomography. J Am Coll Cardiol 301:500-507.
60:1546-1555. 67. Fayad ZA, Fuster V, Fallon JT et al (2000) Noninvasive in vi-
51. Choi Y, Huang SC, Hawkins RA et al (1993) A simplified vo human coronary artery lumen and wall imaging using
method for quantification of myocardial blood flow using nitro- black-blood magnetic resonance imaging. Circulation
gen-13-ammonia and dynamic PET. J Nucl Med 34:488-497. 102:506-510.
52. Zierler K (2000) Indicator dilution methods for measuring 68. Kim WY, Stuber M, Bornert P (2002) Three-dimensional black-
blood flow, volume, and other properties of biological systems: blood cardiac magnetic resonance coronary vessel wall imaging
a brief history and memoir. Annals of biomedical engineering detects positive arterial remodeling in patients with nonsignifi-
28:836-848. cant coronary artery disease. Circulation 106:296-299.
53. McCommis KS, Zhang H, Herrero P et al (2008) Feasibility 69. Ibrahim T, Makowski MR, Jankauskas A et al (2009) Serial
study of myocardial perfusion and oxygenation by noncontrast contrast-enhanced cardiac magnetic resonance imaging
MRI: comparison with PET study in a canine model. Magn demonstrates regression of hyperenhancement within the coro-
Reson Imaging 26:11-19. nary artery wall in patients after acute myocardial infarction.
54. Zun Z, Varadarajan P, Pai RG et al (2011) Arterial spin labeled JACC Cardiovasc Imaging 2:580-588.
CMR detects clinically relevant increase in myocardial blood 70. Rudd JH, Warburton EA, Fryer TD et al (2002) Imaging ath-
flow with vasodilation. JACC Cardiovasc Imaging 4:1253-1261. erosclerotic plaque inflammation with [18F]-fluorodeoxyglu-
55. Ghosh N, Rimoldi OE, Beanlands RS, Camici PG (2010) As- cose positron emission tomography. Circulation 105:2708-
sessment of myocardial ischaemia and viability: role of 2711.
positron emission tomography. Eur Heart J 31:2984-2995. 71. Davies JR, Rudd JH, Weissberg PL, Narula J (2006) Radionu-
56. Di Carli MF (2008) Predicting improved function after my- clide imaging for the detection of inflammation in vulnerable
ocardial revascularization. Current opinion in cardiology plaques. J Am Coll Cardiol 47:C57-68.
13:415-424. 72. Joshi NV, Vesey AT, Williams MC et al (2014) 18F-fluoride
57. Klein C, Schmal TR, Nekolla SG et al (2007) Mechanism of positron emission tomography for identification of ruptured
late gadolinium enhancement in patients with acute myocar- and high-risk coronary atherosclerotic plaques: a prospective
dial infarction. J Cardiovasc Magn Res 9:653-658. clinical trial. Lancet 383:705-713.
58. Klein C, Nekolla SG, Bengel FM et al (2002) Assessment of 73. Saraste A, Laitinen I, Weidl E et al (2012) Diet intervention re-
myocardial viability with contrast-enhanced magnetic reso- duces uptake of alphavbeta3 integrin-targeted PET tracer 18F-
nance imaging: comparison with positron emission tomogra- galacto-RGD in mouse atherosclerotic plaques. J Nucl Cardi-
phy. Circulation 105:162-167. ol 19:775-784.
59. Kim RJ, Wu E, Rafael A et al (2000) The use of contrast-en- 74. Laitinen I, Notni J, Pohle K et al (2013) Comparison of cyclic
hanced magnetic resonance imaging to identify reversible my- RGD peptides for alphavbeta3 integrin detection in a rat mod-
ocardial dysfunction. N Engl J Med 343:1445-1453. el of myocardial infarction. EJNMMI research 3:38.
60. Kwong RY, Chan AK, Brown KA et al (2006) Impact of un- 75. Beer AJ, Pelisek J, Heider P et al (2014) PET/CT imaging of
recognized myocardial scar detected by cardiac magnetic res- integrin alphavbeta3 expression in human carotid atheroscle-
onance imaging on event-free survival in patients presenting rosis. JACC Cardiovasc Imaging 7:178-187.
PEDIATRIC RADIOLOGY SATELLITE COURSE
KANGAROO
IDKD 2015-2018
Introduction Ultrasound
Pediatric chest neoplasms are heterogeneous in etiology Ultrasound can often be of great utility in evaluating neo-
and presentation and their diagnosis is often challenging. plasms that arise from or abut the chest wall. It is rela-
Radiography, ultrasonography, computed tomography tively inexpensive, widely available, and does not require
(CT), and magnetic resonance imaging (MRI) are all of the use of ionizing radiation. It allows for the assessment
utility in the diagnosis of pediatric chest neoplasms. In of tissue characteristics and vascularity in many superfi-
this chapter we review the imaging appearance of various cial chest lesions. However, evaluating the mediastinal
common and rare pediatric chest neoplasms. structures through a trans-thoracic approach, especially
In the following, chest neoplasms are categorized as structures in the middle and posterior mediastinum, is of-
chest wall malignancies, mediastinal masses, metastatic ten difficult because of difficulties in finding acoustic
disease, and primary lung parenchymal neoplasms. We windows between ribs and issues with depth of penetra-
emphasize the imaging features of neoplasms, which can tion [2]. Ultrasound is also often problematic for intra-
be used to obtain an accurate diagnosis of visualized pulmonary lesions. Lung parenchymal processes can be
masses prior to pathologic evaluation. difficult to evaluation due to acoustic reflections from the
We also discuss the distinguishing characteristics that air-containing lungs. Also, any process that does not
differentiate neoplastic processes from congenital mal- abut/reach the chest wall is usually difficult to evaluate.
formation or other non-neoplastic mass lesions that occur
in the chest. Differentiation between these categories is Computed Tomography
often challenging, but a consideration of clinical infor-
mation may facilitate this process. The mainstay of imaging lesions originating from or
We begin with a discussion of the various methods of within the lung parenchyma is chest CT. CT offers a high
imaging pediatric chest neoplasms. spatial resolution, excellent contrast to noise ratio, and a
high speed of acquisition such that imaging can usually
be done within a single, short, patient breath hold. It per-
mits a higher level of anatomic resolution and tissue char-
Imaging of Chest Neoplasms acterization than radiography and, unlike MRI, is not im-
peded by respiratory motion. However, CT necessitates
the use of ionizing radiation, and radiation dose issues of-
Radiography ten arise in pediatric patients who require serial follow-
up evaluations. New, low dose CT dose techniques have
Radiography is a common initial method for imaging the reduced the radiation dose required for diagnostic imag-
pediatric chest and for evaluating pediatric chest neo- ing, but concerns remain regarding repeated radiation ex-
plasms. It is widely available in a variety of settings and posure in diagnostic imaging [1].
inexpensive. However, it is limited in its inherent contrast
resolution between different soft-tissue structures and for Magnetic Resonance Imaging
precise anatomic localization of the extent of a patholog-
ic process. Nonetheless, radiography remains useful in MRI allows high levels of contrast and anatomic resolu-
both the initial evaluation and follow-up of pediatric tion in the pediatric chest without the use of ionizing ra-
chest neoplasms [1]. diation. By applying various magnetic resonance pulse
sequences and imaging techniques, a broad spectrum of On radiography, lymphoma frequently presents as a
tissue characteristics can be visualized. In particular, large anterior mediastinal mass. CT is critical in the ini-
hypercellularity in malignant neoplastic structures often tial diagnosis, to evaluate for potential tracheobronchial
manifests as restricted diffusion on diffusion-weighted compromise. Life threatening airway obstruction is seen
imaging (DWI). Thus, DWI can be useful in the initial in up to 10% of patients with mediastinal lymphoma. En-
evaluation and staging of pediatric malignancies. Howev- dotracheal anesthesia is reported to be safe in patients
er, even though DWI sequences are typically relatively with lymphoma in the anterior mediastinum when the ex-
rapid, issues with motion artifacts during respiration and pected cross-sectional area of the trachea is reduced by
susceptibility artifact from air in the lungs remain [3]. no more than 50% [7].
Another matter of concern is that MRI in the pediatric The role of MRI in the initial diagnosis of lymphoma
population often requires the use of general anesthesia in is generally considered to be limited. However, some re-
order to prevent patient motion during longer imaging se- ports indicate that whole body DWI may be useful in the
quences. In these instances, MRI thus brings with it the staging of lymphoma. MRI may also be of value in de-
inherent risks of general anesthesia. This is countered, termining the amount of fibrous stroma in a lymphoma-
however, by the lack of ionizing radiation in MRI, an ad- tous mass, which can indicate the expected residual mass
vantage over CT. after treatment [6].
Germ-Cell Tumors
Neoplasms of the Pediatric Mediastinum
Germ cell tumors are less common masses of the anterior
Anterior Mediastinum mediastinum, accounting for 618% of all mediastinal
masses [6]. The majority are benign teratomas (80% of all
Hodgkins and Non-Hodgkins Lymphoma mediastinal germ cell tumors) [4, 5]. Malignant germ cell
tumors in the pediatric population include both semino-
Hodgkins lymphoma is the most common mediastinal
matous and non-seminomatous lesions. Seminomatous le-
mass in the pediatric population. Malignant lymphoma
sions typically lack serologic markers and usually do not
overall accounts for 23% of all mediastinal masses in
contain internal calcifications. For this reason, seminomas
children [4, 5]. Presentation with an anterior mediasti-
typically require histologic diagnosis. Non-seminomatous
nal mass occurs with lymphomatous infiltration of the
germ cell tumors can usually be identified through a com-
thymus and lymph nodes (Fig. 1). Hodgkins disease is
bination of imaging appearance, serum markers such as
typically nodal and spreads in direct continuity, where-
-human chorionic gonadotropin and -fetoprotein, and
as non-Hodgkins lymphoma tends to be extranodal and
clinical symptoms such as precocious puberty [8].
to spread hematogenously. A mediastinal mass occurs
in 50% of patients with non-Hodgkins lymphoma
[4-6]. In both processes, invasion of the pericardium Thymic Neoplasms
can result in pericardial effusion. Pleural effusion, on Primary thymic neoplasms are typically benign in the pe-
the other hand, can result from venous or lymphatic diatric population, with thymolipomas as the most com-
compromise and does not necessarily indicate direct mon lesions. Thymolipomas usually present as fat- and
tumor extension [4]. soft-tissue-containing lesions of the anterior medi-
astinum. They typically do not have internal calcifica-
tions [9, 10].
Most of the other mass-type lesions arising from the
thymus are non neoplastic. Of these, perhaps the most
important to consider is benign thymic hyperplasia. In in-
fants, the mean thickness of the thymus is 1.50 cm, with
a standard deviation of 0.46 cm. In children over the age
of 10 years, the mean thickness of the thymus is 1.05 cm,
with a standard deviation of 0.36. An abnormal process
should be considered when thymic thickness on cross-
sectional imaging is over two standard deviations above
normal (upper limits of normal thymic thickness are 2.42
cm from ages 010 and 1.77 cm respectively) [4, 11].
The thymus normally atrophies as a result of cytotox-
ic injury, but may undergo secondary, benign hyperplasia
thereafter. This period is referred to as thymic rebound.
Fig. 1. An 18-year-old-female presenting with Hodgkins lym-
phoma as a mediastinal mass. The mass is irregular in shape, hav- The smooth contours of the hypertrophied thymus and a
ing lost the normal morphology of the thymus, and is heteroge- homogeneous soft-tissue density distinguish thymic re-
neous in appearance bound from tumor recurrence [6] (Fig. 2).
Pediatric Chest Tumors Including Lymphoma 221
Lymphadenopathy
Lymphadenopathy, either in association with lymphoma
or with an infectious or inflammatory process, is often
encountered in the middle mediastinum. Identifiable me-
diastinal lymph nodes in infants are typically abnormal.
Fig. 2. A 16-year-old female with thymic rebound after undergoing
In adolescents, lymph nodes 1 cm in their short axis are
chemotherapy for a sarcoma of the lower extremity. The margins of the typically considered abnormal [4, 15].
thymus are preserved and the thymus is homogenous in appearance
Posterior Mediastinal Neoplasms
Thymic cysts are another potential non-neoplastic le- Neurogenic Tumors
sion that involve the anterior mediastinum. These are flu-
id density, smooth, well-demarcated lesions found within Neurogenic tumors make up approximately 90% of pos-
otherwise normal thymic tissue [6, 12]. terior mediastinal masses in the pediatric population [6].
Tumors of ganglion-cell origin include neuroblastoma,
Middle Mediastinal Neoplasms ganglioneuroblastoma, and ganglioneuroma. Neuroblas-
toma is most common amongst infants and young chil-
The most common primary pediatric middle mediastinal dren with a median patient age at presentation of 2 years
neoplasms are lymphomatous or leukemic [4, 5]. Other- (Figs. 3, 4). Ganglioneuroma, typically a benign lesion,
wise, primary middle mediastinal neoplasms are unusual occurs in older children with a median age at presenta-
in the pediatric population. Other neoplasms of the mid- tion of 10 years. For ganglioneuroblastoma, the median
dle mediastinum include cardiac tumors, which are rare, age at presentation is 5.5 years [6, 16].
and metastatic disease, which is more common. The ma- All three tumors of ganglion cell origin typically ap-
jority of middle mediastinal masses in the pediatric pop- pear as a vertically elongated mass with tapered superior
ulation are either infectious, inflammatory, or congenital and inferior margins. Internal calcifications are seen in
(such as bronchopulmonary foregut cysts; discussed be- 30% of ganglion cell origin tumors and adjacent bony
low) in origin. changes are common [17].
Cardiac Tumors
Cardiac tumors are rare in the pediatric population. The
most common pediatric tumor is a cardiac rhabdomyoma
which is usually seen in association with tuberous scle-
rosis. In one series, 91% of cardiac rhabdomyomas were
found in association with tuberous sclerosis. In general,
any intracavitary mass found in infants should be consid-
ered a rhabomyoma until proven otherwise [13, 14].
Other tumors include cardiac fibroma, which is
thought to be a hamartomatous lesion, and cardiac myx-
oma. The latter are typically found in association with
myxoma syndrome, a familial disorder [13, 14].
Foregut Cysts
Fig. 3. A 3-month-old male with a partially calcified posterior me-
The most frequently encountered developmental anomaly diastinal mass found to be a neuroblastoma. Normal thymic tissue
of the pediatric middle mediastinum is the foregut cyst, is seen anteriorly
222 A. Oshmyansky, T.A.G.M. Huisman
Chest Wall Neoplasms Vascular malformations of the chest wall are a het-
erogeneous group of congenital lesions that include ve-
Benign Soft-Tissue Lesions nous sascular malformations, hemangiomas, arteriove-
nous malformations/fistula, and lymphatic malforma-
Lipomas are benign masses consisting of adipose tissue tions (Fig. 7). A complete discussion of vascular malfor-
that commonly occurs in the chest wall. Lipoblastoma is mations is beyond the scope of this review. However,
a tumor of fetal embryonic fat; despite its rapid growth it vascular malformations can typically be diagnosed based
is typically benign. After resection, a lipoblastoma can re- on their avidly bright signal on T2 weighted MRI se-
cur years later as a mature lipoma. The two lesions are quences as well as their prominent vascular flow.
typically differentiated histologically and are usually eas- Pheloboliths can often be seen with venous malforma-
ily identified on CT or MRI by the presence of macro- tions. In most cases, the dynamic contrast enhancement
scopic fat within the lesion [24, 25]. features of vascular malformations readily allow their
Other benign neoplastic lesions include neurofibro- differentiation [28].
mas, discussed above, and fibromas. Fibromas associated
with Gardners syndrome are nodular subcutaneous le- Malignant Soft-Tissue Lesions
sions that contain hyaline or collagen. They are soft-
tissue-density structures on CT and slightly T1 and T2 Rhabdomyosarcoma is the most common pediatric sarco-
bright on MRI. However, histologic diagnosis is normal- ma, although it is generally uncommon in the chest wall
ly required [24, 26]. (Fig. 8). Rhabdomyosarcoma accounts for 24% of pedi-
Mesenchymal hamartomas are benign lesions resulting atric malignancies. On imaging, it presents as an enhanc-
from the overgrowth of normal skeletal elements of the ing, aggressive mass with imaging characteristics similar
rib (Fig. 6). They are well defined, extra-pleural, and often to that of skeletal muscle [24, 28].
cystic. Hemorrhagic and cystic findings with associated Neuroblastoma can also occur in the chest wall in ad-
fluid-fluid levels and partial calcification are said to be dition to the posterior mediastinum and has similar imag-
diagnostic [24, 27]. ing characteristics.
a b c
Fig. 6 a-c. A 16-year-old male with a chest wall hamartoma of the right lower chest. a Axial T2 fat-saturated image shows a heterogeneous,
partially cystic mass, initially mistaken for a neuroblastoma; b axial post-contrast MRI; c axial post-contrast CT through the lesion
a b
Fig. 7 a, b. A large lymphangioma extending from the mediastinum into the left chest on axial T2 fat-saturated sequence (a) and axial post-
contrast MRI sequence (b)
224 A. Oshmyansky, T.A.G.M. Huisman
a b
Fig. 8 a, b. A 16-year-old male with an alveolar rhabdomyosarcoma of the left posterior chest wall. a Axial CT scan through the chest shows
a soft-tissue mass arising from the soft tissues of the right back. b Axial post-contrast MRI of the lesion shows avid contrast enhancement
a b c
Fig. 9 a-c. A 15-year old female with a mass arising from right first rib, found to be Ewings sarcoma. Axial CT (a), axial post-contrast MRI
(b), and focused ultrasound (c) examinations
Small Round Blue Cell Tumors Langerhans cell histiocytosis is a multisystem disease
characterized by the proliferation of bone-marrow-
Small round blue cell tumors, including Ewings sarcoma derived Langerhans cells and eosinophils. Imaging features
and primitive neuroectodermal tumors (PNETs), can oc- include lytic, expansive bone lesions that are typically T2
cur in the chest wall in both osseous and extraosseous lo- bright and T1 dark [30].
cations (Fig. 9). These are collectively referred to as
Askin tumors of the chest wall [24]. One-third of extra-
osseous Ewings tumors and half of PNETs present in the Conclusions
chest wall. Both lesions are typically heterogeneous, en-
hancing lesions on CT and MRI [28]. However, the imag- Pediatric chest neoplasms are a heterogeneous group of
ing appearance of either lesion is non-specific and histo- diseases. They can arise from any structure within the pe-
logic correlation is required for a definitive diagnosis. diatric chest. Differentiation between malignant and be-
Cross-sectional imaging is useful for demarcating the ex- nign processes such as congenital abnormalities is of es-
tent of disease [29]. sential importance. Overall, primary pediatric pulmonary
malignancies are uncommon. However, mediastinal
Osseous Neoplasms masses and chest wall masses are not and should prompt
a thorough investigation.
Osseous malignancies such as osteosarcoma can occur in
the chest wall and the spine, just as they do in other por-
tions of the body. In particular, in the pediatric popula- References
tion, leukemia should be considered when multiple os-
seous lesions are seen. Osseous lymphoma is also a con- 1. Kuhn JP, Slovis TL, Haller JO et al (2004) Caffeys pediatric
diagnostic imaging, 10th ed. Mosby, Philadelphia, PA.
sideration when an osseous lesion is seen in children. 2. Haller JO, Schneider M, Kassner EG et al (1980) Sonograph-
However, it should be noted that primary osseous lym- ic evaluation of the chest in infants and children. AJR Am J
phoma is relatively uncommon in children [29]. Roentgenol 134:1019-1027.
Pediatric Chest Tumors Including Lymphoma 225
3. Siegel MJ, Acharyya S, Hoffer FA et al (2013) Whole-body 17. Sofka CM, Semelka RC, Kelekis NL et al (1999) Magnetic
MR imaging for staging of malignant tumors in pediatric pa- resonance imaging of neuroblastoma using current techniques.
tients: results of the American College of Radiology Imaging Magn Reson Imaging 17:193-198.
Network 6660 Trial. Radiology 266:599-609. 18. Daldrup HE, Link TM, Wortler K et al (1998) MR imaging of
4. Merten DF (1992) Diagnostic imaging of mediastinal masses thoracic tumors in pediatric patients. AJR Am J Roentgenol
in children. AJR Am J Roentgenol 158:825-32. 170:1639-1644.
5. King RM, Telander RL, Smithson WA et al (1982) Primary 19. Senac MO, Jr., Wood BP, Isaacs H et al (1991) Pulmonary blas-
mediastinal tumors in children. J Pediatr Surg 17:512-520. toma: a rare childhood malignancy. Radiology 179:743-746.
6. Franco A, Mody NS, and Meza MP (2005) Imaging evaluation 20. McCarville MB (2010) Malignant pulmonary and mediastinal
of pediatric mediastinal masses. Radiol Clin North Am tumors in children: differential diagnoses. Cancer Imaging 10
43:325-353. Spec no A:S35-41.
7. Shamberger RC, Holzman RS, Griscom NT et al (1991) CT 21. Neville HL, Hogan AR, Zhuge Y et al (2009) Incidence and
quantitation of tracheal cross-sectional area as a guide to the outcomes of malignant pediatric lung neoplasms. J Surg Res
surgical and anesthetic management of children with anterior 156:224-230.
mediastinal masses. J Pediatr Surg 26:138-142. 22. Chong S, Lee KS, Chung MJ et al (2006) Neuroendocrine tu-
8. Shin MS, Ho KJ (1983) Computed tomography of primary mors of the lung: clinical, pathologic, and imaging findings.
mediastinal seminomas. J Comput Assist Tomogr 7:990-994. Radiographics 26:41-57.
9. Juanpere S, Canete N, Ortuno P et al (2013) A diagnostic ap- 23. McCarville MB, Lederman HM, Santana VM et al (2006) Dis-
proach to the mediastinal masses. Insights Imaging 4:29-52. tinguishing benign from malignant pulmonary nodules with
10. Gaerte SC, Meyer CA, Winer-Muram HT et al (2002) Fat-con- helical chest CT in children with malignant solid tumors. Ra-
taining lesions of the chest. Radiographics 22 Spec No:S61- diology 239:514-520.
78. 24. Watt AJ (2002) Chest wall lesions. Paediatr Respir Rev 3:328-
11. St Amour TE, Siegel MJ, Glazer HS et al (1987) CT appear- 338.
ances of the normal and abnormal thymus in childhood. J 25. Whyte AM and Powell N (1990) Mediastinal lipoblastoma of
Comput Assist Tomogr 11:645-650. infancy. Clin Radiol 42:205-206.
12. Mishra SK, Melinkeri SR, Dabadghao S (2001) Benign thymic 26. Jabra AA, Taylor GA (1993) MRI evaluation of superficial soft
hyperplasia after chemotherapy for acute myeloid leukemia. tissue lesions in children. Pediatr Radiol 23:425-428.
Eur J Haematol 67:252-254. 27. Groom KR, Murphey MD, Howard LM et al (2002) Mes-
13. Becker AE (2000) Primary heart tumors in the pediatric age enchymal hamartoma of the chest wall: radiologic manifesta-
group: a review of salient pathologic features relevant for clin- tions with emphasis on cross-sectional imaging and
icians. Pediatr Cardiol 21:317-323. histopathologic comparison. Radiology 222:205-211.
14. Mader MT, Poulton TB, White RD (1997) Malignant tumors 28. David EA, Marshall MB (2011) Review of chest wall tumors:
of the heart and great vessels: MR imaging appearance. Radio- a diagnostic, therapeutic, and reconstructive challenge. Semin
graphics 17:145-153. Plast Surg 25:16-24.
15. Lee JKT (2006) Computed body tomography with MRI corre- 29. Faro SH, Mahboubi S, Ortega W (1993) CT diagnosis of rib
lation, 4th Ed. Lippincott Williams & Wilkins, Philadelphia, anomalies, tumors, and infection in children. Clin Imaging
PA. 17:1-7.
16. Lonergan GJ, Schwab CM, Suarez ES et al (2002) Neuroblas- 30. Hashmi MA, Haque N, Chatterjee A et al (2012) Langerhans
toma, ganglioneuroblastoma, and ganglioneuroma: radiologic- cell histiocytosis of long bones: MR imaging and complete
pathologic correlation. Radiographics 22:911-934. follow up study. J Cancer Res Ther 8:286-288.
IDKD 2015-2018
Magnetic Resonance Imaging cular imaging. These diseases include coarctation of the
aorta, cardiac septal defects, TGA, Ebsteins anomaly,
Both anatomic and functional information on cardiovas- TOF, vascular rings and sling, and partial anomalous pul-
cular diseases can be obtained with MRI. It is a particu- monary venous return.
larly useful imaging technique due to the lack of ionizing
radiation exposure, unlike CT. Static spin echo black Coarctation of the Aorta
blood sequences (T1/T2) or cine gradient echo bright
blood sequences (2D steady state free precession) are Coarctation of the aorta is the most common congenital
useful for evaluating cardiovascular morphology [1, 5]. anomaly of the thoracic aorta, occurring in 0.04% of
Extracardiac thoracic vasculature can be assessed with newborns [57]. It is characterized by the presence of
spin echo black blood techniques. If there is a need for a congenital narrowing of the aorta near the region where
dynamic assessment of the thoracic vessels, cine gradient the ductus or ligamentum (after regression) arteriosus in-
echo imaging can be acquired throughout the cardiac cy- serts (Fig. 1). Coarctation of the aorta is seen in approx-
cle [1, 5]. Contrast-enhanced 3D magnetic resonance an- imately 7% of patients with congenital heart disease and
giography (MRA), which can provide superb morpholog- has a strong male predominance (male-to-female ratio of
ical evaluation of the thoracic vasculature with time re- 4:1) [57]. Although it has been historically categorized
solved techniques, is useful in pediatric patients with un- into two types: infantile (i.e., preductal) and adult (i.e.,
derlying cardiovascular diseases [1, 5]. In fact, compared postductal) [57], this categorization is currently not fa-
to CTA, MRA has the advantage of superior tissue char- vored because all aortic coarctations are juxtaductal in
acterization and direct multiplanar imaging. Furthermore, location, with various degrees and extension of aortic
newer 3T MRA scanners can deliver improved signal-to- narrowing. Affected neonates typically present with con-
noise ratio, which can be traded for higher resolution and gestive heart failure whereas affected older children usu-
for decreased overall scanning times when a parallel ally come to medical attention during the evaluation of
imaging technique is employed. Once the MRA dataset is arterial hypertension in the upper extremities and nor-
obtained, maximum intensity projection and 3D magnet- mal-to-decreased blood pressure in the lower extremities
ic resonance images can enhance the visualization of car- [57].
diovascular anomalies and abnormalities. In neonates with coarctation of the aorta, increased pul-
monary vascularity with cardiomegaly is often seen on
chest radiographs. In older children with delayed diagno-
Spectrum of Imaging Findings sis of aortic coarctation, the figure of 3 sign, which re-
flects pre-stenotic dilatation of the aortic arch and left sub-
This review focuses on congenital cardiovascular dis- clavian artery, indentation at the coarctation site, and post-
eases that are considered to be of widespread clinical rel- stenotic dilatation of the descending aorta, may be present
evance and of current importance in pediatric cardiovas- [5-7]. Other typical concomitant findings are collateral
a b
a b
Transposition of the Great Arteries in series rather than in parallel as in D-TGA. Systemic ve-
nous return flows to the right atrium, enters the morpho-
TGA refers to a congenital malformation of the great ar- logic left ventricle, and is eventually delivered to the pul-
teries in which they are reversed in their origins from the monary circulation via the main pulmonary artery. Pul-
heart (Fig. 4). There are two different types of TGA: monary venous return subsequently flows to the left atri-
D-transposition of the great arteries (D-TGA) and L-trans- um, enters the morphologic right ventricle, and is delivered
position of the great arteries (L-TGA) [13]. In D-TGA, to the systemic circulation via the aorta. While patients
the aorta arises from the morphologic systemic right ven- with D-TGA usually present with cyanosis during the first
tricle and the pulmonary artery arises from the morpho- day of life, those with L-TGA are typically asymptomatic.
logic systemic right ventricle, resulting in ventricular arte- On chest radiographs, increased pulmonary vasculari-
rial discordance. The pulmonary and systemic circulations ty and mild to moderate cardiomegaly are usually seen in
exist in parallel, requiring bidirectional shunting between patients with D-TGA [13]. Additionally, the superior
the two sides of the heart via an ASD, VSD, patent ductus mediastinum is narrowed because of the abnormal posi-
arteriosus (PDA), or patent foramen ovale for survival. By tion of the great vessels and a concave main pulmonary
contrast, congenitally corrected TGA, or L-TGA, is char- artery. The constellation of these findings results in the
acterized by both ventriculoarterial and atrioventricular radiographic appearance of the cardiomediastinal silhou-
discordance. The pulmonary and systemic circulations are ette, referred to as the egg-on-a-string sign [2]. CTA or
MRA is useful for the assessment of postoperative com-
plications including narrowing of the anastomosis of the
aorta or pulmonary artery in addition to branch pul-
monary artery stenosis and coronary artery narrowing.
Corrective surgery via a Jatene arterial switch pro-
cedure is required for managing patients with D-TGA in
the first few days of life [13] (Fig. 5). By contrast, pa-
tients with congenitally corrected L-TGA may require
pacemaker placement for progressive right ventricular
dysfunction and conduction abnormalities in later life.
Ebsteins Anomaly
Ebsteins anomaly is the most common cause of marked
cardiomegaly in the newborn period. It is characterized by
the displacement of the septal and posterior leaflets of the
tricuspid valve towards the apex of the right ventricle of
Fig. 4. D-transposition of the great vessels (D-TGA). Axial CT the heart [10]. This results in atrialization of a portion
shows the aorta (A) located anterior to the pulmonary artery (PA) of the morphologic right ventricle and subsequent
230 E.Y. Lee
Tetralogy of Fallot
Four anomalies of the cardiac structures comprise TOF:
VSD, infundibular pulmonary stenosis, an overriding aor-
ta, and right ventricular hypertrophy. It is the most com-
mon congenital heart defect resulting in cyanosis. A right
aortic arch with a mirror-image branching, peripheral
pulmonary arterial stenosis, and an anomalous anterior
Fig. 5. Jatene procedure. Axial CT after an arterial switch proce- descending coronary artery that courses over the right
dure for TGA shows characteristic draping of the main pulmonary ventricular outflow tract are also often associated with
artery (MP) around the aorta (A) TOF. Affected pediatric patients typically present with
cyanosis in the first 6 months of life.
The radiographic findings of TOF vary according to
enlargement of the right atrium and a reduction in the size the severity of the obstruction of the pulmonary outflow
of the anatomic right ventricle [10]. This condition may be tract. The classic appearance of TOF is a boot-shaped
an isolated finding or associated with other congenital heart due to an underlying right-sided aortic arch, a small
heart defects, such as a patent foramen ovale or an atrial or concave main pulmonary artery, and right ventricular
septal defect, allow the passage of a right-to-left shunt. Al- hypertrophy [2] (Fig. 7). Pulmonary vascularity can be
though patients with mild Ebsteins anomaly may be either normal or diminished. The heart size is usually
asymptomatic, in those with severer forms of the disease normal. However, increased pulmonary vascularity and
cyanosis and systolic murmur are often present [10]. cardiomegaly similar to that in patients with VSD can be
On radiographs, decreased pulmonary vascularity, a seen in affected patients with very mild pulmonary steno-
concave pulmonary artery segment, and marked car- sis, which is sometimes referred to as a pink tetralogy.
diomegaly due to underlying right atrial enlargement are MRI can be useful for evaluating the morphology and
characteristically seen [2, 10] (Fig. 6a). In affected pa- function of the cardiac chambers, calculating blood flow
tients with concomitant PDA, the pulmonary vasculari- and velocities through the vessels to determine the pres-
ty may be increased. CT and MRI can better demon- ence of regurgitation, stenoses, and relative flow, and
strate a small right ventricle and atrialized right ventri- mapping collateral vessels in affected patients with pul-
cle (Fig. 6b), and MRI is better than echocardiography monary atresia [1, 2].
a b
Fig. 6 a, b. Ebsteins anomaly in a 4-year-old female. a Chest radiograph shows decreased pulmonary vascularity and cardiomegaly with a
prominent right heart (arrows). b Axial MRI demonstrates a large atrialized superior right ventricle (ARV). RA, right atrium
Pediatric Cardiovascular Diseases 231
a b
Fig. 8 a, b. Double aortic arch in a 4-month old male who presented with stridor and repeated apnea. a Axial maximum intensity projection
image shows a double aortic arch (R, right arch; L, left arch; S, superior vena cava). b 3D volume-rendered image demonstrates a double
aortic arch (R, right arch; L, left arch) surrounding the trachea (T)
232 E.Y. Lee
[46, 8, 12, 13] (Fig. 8b). In particular, paired inspirato- tion of the esophagus is seen on frontal esophagogram,
ry and expiratory multi-detector CT should be considered and oblique compression of the posterior esophagus on
as part of the routine preoperative evaluation of tracheo- lateral esophagogram. The entire course of the right aor-
malacia (TM) in pediatric patients with double aortic tic arch, the aberrant left subclavian artery, and tracheal
arch because 33.3% of these patients have underlying compression are best evaluated with CT and MRI with
concomitant TM [14]. multiplanar and 3D reconstruction [46, 12, 15] (Fig.
The double aortic arch in symptomatic pediatric pa- 9b). Surgical division of underlying vascular ring is the
tients is currently treated with surgical division of the current management procedure of choice in symptomatic
smaller of the two aortic arches. When concomitant TM pediatric patients with right aortic arch with an aberrant
is present, surgical repair of the trachea is also required left subclavian artery [46, 12].
[12, 13, 15].
Pulmonary Artery Sling
Right Aortic Arch with an Aberrant Left Subclavian Artery
In pulmonary artery sling, also known as anomalous left
The right aortic arch with an aberrant subclavian artery pulmonary artery, the left pulmonary artery arises from
is the second most common vascular ring in the pedi- the right pulmonary artery [16, 17] (Fig. 10). Embryo-
atric population. Embryologically, it is due to persis- logically, it is due to the regression or failure of devel-
tence of the right dorsal aorta, regression of the left dor-
sal aorta, and regression at the left fourth aortic arch
[46, 8, 12]. In this condition, the aorta is located on the
right side of the spine and the left subclavian artery
comes off the descending aorta, crosses the midline, and
enters the left hemithorax (Fig. 9a). The ligamentum ar-
teriosum (i.e., the remnant of the ductus arteriosus),
which originates from a bulbous dilatation at the base of
the left subclavian artery (the diverticulum of Kom-
merell) and attaches to the left pulmonary artery, is al-
so present. The trachea and esophagus are encircled by
the aortic arch on the right, the ligamentum arteriosum
and the left pulmonary artery on the left, the ascending
aorta anteriorly, and the descending aorta posteriorly
[46, 12].
On the frontal chest radiograph, a right paratracheal
opacity representing the right-sided aortic arch and an ab-
sent left-sided aortic arch shadow are characteristic. The
Fig. 10. Pulmonary artery sling in a 2-day-old female who present-
right-sided compression or left-sided deviation of the tra- ed with severe respiratory distress. Axial CT shows an anomalous
chea at the level of the right aortic arch and an aberrant left pulmonary artery (arrow) arising from the right main pul-
left subclavian artery are also seen. Right-sided indenta- monary artery (RP). T, Trachea
a b
Fig. 9 a, b. Right aortic arch with an aberrant left subclavian artery in a 2-year-old female who presented with respiratory distress. a Axial
CT shows a right aortic arch (RA) with an aberrant left subclavian artery (arrow). Note the nasogastric tube in the esophagus. T, trachea.
b 3D volume rendered image demonstrates the exact location, degree, and extent of the tracheal compression (arrow) caused by a right
aortic arch and an aberrant left subclavian artery
Pediatric Cardiovascular Diseases 233
opment of the left pulmonary artery [4, 16, 17]. The vein drains into the left brachiocephalic or innominate
anomalous left pulmonary artery courses between the tra- vein, creating a vertical vein positioned lateral to the left
chea and esophagus on its way to the left side after aris- superior mediastinum (Fig. 11). On chest radiographs,
ing from the right pulmonary artery. This can lead to ex- left superior mediastinal widening is seen. In case of the
trinsic compression of the distal trachea and/or right main scimitar, or hypogenetic lung syndrome, which is a spe-
stem bronchus, and thus in respiratory distress. Intrinsic cial form of anomalous pulmonary venous return from
large airway anomalies such as TM and tracheal stenosis the lung that is associated with a hypoplastic lung, a ver-
as well as right lung agenesis may also be present con- tically oriented, curvilinear opacity representing the
comitantly [16, 17]. scimitar vein, projecting over the lower hemithorax and
On frontal chest radiograph, either a hyperinflated or ipsilateral hypoplastic lung, is seen on chest radiographs
atelectatic right lung, depending on the degree of right [6, 16] (Figs. 12, 13). In case of right superior PAPVR,
main stem bronchial narrowing, is seen. On lateral chest CT and MRI can show an anomalous vein that usually
radiograph or esophagogram, anterior displacement of drains into the SVC and an associated sinus venosus
the trachea and posterior displacement of the esophagus
due to an anomalous left pulmonary artery are character- a
istically seen. CT and MRI with multiplanar and 3D re-
construction of the mediastinal vessels and large airway
can demonstrate the entire course of an anomalous left
pulmonary artery and an associated large airway narrow-
ing or stenosis [4, 13, 1517].
Surgical reimplantation of the left pulmonary artery
and tracheal reconstruction in case of tracheal stenosis or
TM are current treatment options in symptomatic patients
[12, 15, 17].
Fig. 11. Left upper lobe partial anomalous pulmonary venous return Fig. 12 a, b. Scimitar syndrome in a 3-year-old female. a Chest ra-
in a 17-year-old male. Axial (right) and coronal (left) CT show an diograph shows a curvilinear opacity (arrow) in the right lower
anomalous left upper lobe partial anomalous pulmonary vein (ar- lobe, representing the anomalous draining vein. b Angiography
rows). A, aorta; S, superior vena cava demonstrates an anomalous draining vein (arrow)
234 E.Y. Lee
Conclusion
A variety of congenital cardiovascular diseases can occur
in the pediatric population. Clear knowledge of the un-
derlying etiologies and their characteristic imaging ap-
pearances can help radiologists avoid potential pitfalls
and render a more timely and accurate diagnosis in pedi-
atric patients with cardiovascular diseases. This, in turn,
will contribute to optimal pediatric patient care.
References
1. Browne LP, Krishnamurthy R, Chung T (2011) Preoperative
and postoperative MR evaluation of congenital heart disease in
children. Radiol Clin North Am 49:1011-1024.
2. Ferguson EC, Krishnamurthy R, Oldham SA (2007) Classic
imaging sings of congenital cardiovascular anomalies. Radio-
graphics 27:1323-1334.
3. Goo HW (2011) Cardiac MDCT in children: CT technology
overview and interpretation. Radiol Clin North Am 49:997-1010.
Fig. 13. Scimitar syndrome in a 2-year-old male. 3D volume ren- 4. Kondrachuk O, Yalynska T, Tammo R, Lee EY (2012) Multi-
dered image shows the anomalous vein from the right lower lobe detector computed tomography evaluation of congenital medi-
(arrow) draining into the inferior vena cava astinal vascular anomalies in children. Semin Roentgenol
47:127-134.
5. Lee EY, Browne LP, Lam W (2012) Noninvasive magnetic res-
onance imaging of thoracic large vessels in children. Semin
Roentgenol 47:45-55.
6. Hellinger JC, Daubert M, Lee EY, Epelman M (2011) Con-
genital thoracic vascular anomalies: evaluation with state-of-
the-art MR imaging and MDCT. Radiol Clin North Am
49:969-996.
7. Tawes RL Jr, Aberdeen E, Waterston DJ et al (1969) Coarcta-
tion of the aorta in infants and children. A review of 333 op-
erative cases, including 179 infants. Circulation 39 (5 Suppl
1):l173-184.
8. Lee EY, Siegel MJ, Hildebolt CF et al (2004) MDCT evalua-
tion of thoracic aortic anomalies in pediatric patients and
young adults: comparison of axial, multiplanar, and 3D im-
ages. AJR Am J Roentgenol 182:777-784.
9. Lee EY, Siegel MJ, Chu CM et al (2004) Amplatzer atrial sep-
tal defect occlude for pediatric patients: radiographic appear-
ance. Radiology 233:471-476.
10. Galea J, Ellul S, Schembri A et al (2014) Ebstein anomaly: a
review. Neonatal Netw 33:268-274.
11. Woods RK, Sharp RJ, Holcom GW 3rd et al (2001) Vascular
anomalies and tracheoesophageal compression: a single insti-
tutions 25-year experience. Ann Thorac Surg 72:434-438.
Fig. 14. Right upper lobe partial anomalous pulmonary venous re- 12. Lee EY, Boiselle PM, Shamberger RC (2010) Multidetector
turn in a 7-year-old female who presented with exertional shortness computed tomography and 3-dimensional imaging: preopera-
of breath and mild hypoxia. Bright-blood MRA shows right upper tive evaluation of thoracic vascular and tracheobronchial
lobe anomalous pulmonary venous return (arrow) into the superi- anomalies and abnormalities in pediatric patients. J Pediatr
or vena cava (SVC) Surg 45:811-821.
13. Lee EY, Siegel MJ (2007) MDCT of tracheobronchial nar-
rowing in pediatric patients. J Thorac Imaging 22:300-309.
14. Lee EY, Zurakowski D, Waltz DA et al (2008) MDCT evalua-
ASD, which is often associated with right superior tion of the prevalence of tracheomalacia in children with medi-
PAPVR [6] (Fig. 14). astinal aortic vascular anomalies. J Thorac Imaging 23:258-265.
15. Lee EY, Greenberg SB, Boiselle PM (2011) Multidetector
The current management procedure of choice of computed tomography of pediatric large airway diseases: state-
PAPVR in symptomatic pediatric patients is surgical of-the-art. Radiol Clin North Am 49:869-893.
reconnection of the anomalous pulmonary vein with the 16. Lee EY, Boiselle PM, Cleveland RH (2008) Multidetector CT
atrium. Imaging plays an important role in the assess- evaluation of congenital lung anomalies. Radiology 247:632-648.
ment of postoperative complications, including throm- 17. Lee EY (2007) MDCT and 3D evaluation of type 2 hypoplas-
tic pulmonary artery sling associated with right lung agenesis,
bosis or stenosis of the reimplanted anomalous vein hypoplastic aortic arch, and long segment tracheal stenosis. J
[6, 15, 16]. Thorac Imaging 22:346-350.
IDKD 2015-2018
lung disease at a dose equivalent to approximately ten the most frequent feature of interstitial pneumonitis/fibro-
chest radiographs. Images are reconstructed based on a sis [20]. There are no discriminating features for the radio-
high spatial resolution algorithm and displayed with a logical appearance of DIP, which includes widespread
wide window setting, at a width of 1,500 Hounsfield ground glass opacification; thus, it has a wide differential
units (HU) and at a level of 500 HU. diagnosis, including subacute extrinsic allergic alveolitis,
If a child is unable to breath-hold, the scans can be per- opportunistic infections such as Pneumocystis jirovecii
formed during quiet breathing, and decubitus scans can pneumonia and, although it is rare in children, sarcoido-
replace expiratory scans (the dependent lung behaving as sis. On HRCT, ground glass opacification is the dominant
the expiratory lung). pattern and often has a lower zone predominance; long-
The role of HRCT in paediatric ILD is evolving. In term follow-up usually shows complete resolution
adults, the diagnostic accuracy of HRCT has led to a de- Based on the few reported cases of childhood NSIP,
crease in the number of lung biopsies. The histospecific there is a distinct pattern of involvement. In a review of
accuracy of HRCT compared with chest radiographs in CTs of six cases of biopsy-proven NSIP [21], a distinct
making a correct first-choice diagnosis in an adult popu- upper zone predominance of honeycomb pattern and
lation with diffuse lung disease ranges from 46% to 75% parenchymal distortion on a background of diffuse
for HRCT and from 38% to 63% for chest radiography ground glass opacification was seen in three children.
[1517]. HRCT had a diagnostic accuracy of 56% for a The honeycomb pattern resembled emphysematous de-
confident first-choice diagnosis in one series of 20 chil- struction with some cystic change.
dren with ILD [18]. This is comparable to a more recent Seely et al. described a similar HRCT appearance in
series of 20 children with biopsy-proven ILD from a sin- their review of paediatric idiopathic interstitial fibrosis [20],
gle institution [19]. A correct first-choice diagnosis was but did not provide pathological correlation. Ground glass
made in 61% of the cases on HRCT compared with 34% opacification, honeycombing and peripheral consolidation
on chest radiographs. were features seen in three other cases of childhood NSIP.
The diseases that were correctly diagnosed on HRCT In the study by Copley et al. [19], one case of DIP resem-
with a high degree of confidence were alveolar pro- bled NSIP with predominant upper lobe involvement.
teinosis, pulmonary lymphangiectasia and idiopathic pul- The existence of UIP (usual interstitial pneumonitis) in
monary haemosiderosis. Differentiation between non- childhood is controversial and said to be exceedingly
specific interstitial pneumonitis (NSIP), desquamative in- rare. The adult form of UIP has a poor prognosis. The
terstitial pneumonitis (DIP) and lymphocytic interstitial HRCT findings of adult UIP include a subpleural reticu-
pneumonitis (LIP) was, however, less reliable. lar pattern of honeycombing that may result eventually in
There are several pitfalls in the interpretation of HRCT thick-walled cystic spaces and irregularity of the pleural
in children. One of the most important is in distinguish- surface with underlying traction bronchiectasis and bron-
ing diffuse ground glass infiltration from increased lung chiolectasis.
attenuation resulting from a suboptimal inspiration. In the LIP is characterised histologically by a diffuse inter-
upper zones, the position of the posterior tracheal mem- stitial infiltrate of polyclonal lymphocytes, plasma cells
brane is helpful in distinguishing between the two. The and histocytes. LIP is associated with viral infections in-
posterior tracheal membrane is convex outwards in inspi- cluding Ebstein Barr virus (EBV) and the acquired im-
ration, and appears horizontal or slightly concave on ex- munodeficiency syndrome. LIP is present in up to 30%
piration. The other difficulty that may be encountered is of children with human immunodeficiency virus (HIV)
in determining which areas of the lung are abnormal disease, especially those from the sub-Saharan African
when an investigation reveals a widespread mosaic pat- continent, where EBV is endemic. This is unlike the sit-
tern of lung attenuation. Deciding on whether the areas uation in adult HIV disease, in which LIP is rare. Other
of diminished attenuation represent, for example, small associations include autoimmune diseases and lympho-
airways disease or whether areas of increased attenuation proliferative disease secondary to underlying congenital
(ground glass opacity) represent diffuse infiltration can immunodeficiency.
be challenging. Although expiratory images may be help- Chest radiographs show reticulonodular change with
ful, obtaining images at known phases of respiration is or without areas of consolidation. HRCT shows intersti-
usually not achievable in children. Thus, the above-men- tial reticulonodular change of varying degrees There is
tioned use of lateral decubitus imaging CT, with the de- usually marked resolution of the striking radiological ap-
pendent lung simulating expiration, is an important tool. pearances with therapy for HIV disease; however, if there
has been chronic interstitial disease, fibrosis and ensuing
traction bronchiectasis may result.
HRCT Features of Diffuse Interstitial Lung Disease Follicular bronchiolitis is a term coined to described a
form of LIP in which the lymphoid aggregates congre-
Idiopathic Interstitial Pneumonitis gate around the small airways.
Chronic pneumonitis of infancy is a recently described
NSIP, DIP and LIP in childhood appear to share common pathological entity. Its HRCT appearance was reported
CT appearances. Widespread ground glass attenuation is based on a single biopsy-proven case, in which wide-
The Radiology of Diffuse Interstitial Pulmonary Disease in Children: Pearls, Pitfalls and Newly Recognised Disorders 237
spread ground glass opacification without focal areas of fat with bilateral pleural effusions and pleural thickening
consolidation or cystic change was seen. [23] is highly suggestive of the diagnosis. The condition
is related to congenital lymphangiectasia, in which there
is abnormal dilatation of the lymphatics. The CT findings
Connective Tissue Disorders of these two conditions are indistinguishable.
involvement, but this is not always associated with an Newer Disorders and Mimics of ILDs of Childhood
adverse outcome [29, 30]. Isolated pulmonary LCH is ex-
tremely rare in children. Pulmonary Veno-occlusive Disease
In children under 10 years of age, pulmonary involve-
ment tends to regress spontaneously (V. Nandhuri, person- Patients with this rare disorder present with the clinical
al communication), but in older children the acute features picture of pulmonary hypertension. Chest X-ray shows
resemble those seen in adults. HRCT shows multiple thin- enlarged pulmonary arteries, Kerley B lines, pleural effu-
walled cysts usually less than 1 cm in size and predomi- sions and mediastinal adenopathy. CT demonstrates pul-
nantly within the upper and mid-zones of the lungs. monary arterial enlargement, mediastinal lymphadenopa-
Nodules show diffuse distribution throughout the lung, thy and normal main pulmonary veins but shows smooth
with relative sparing of the costophrenic angles. Multiple interlobular septal thickening due to peripheral pul-
nodules measuring 13 mm in size occur and may be monary venous congestion. There is a mosaic pattern of
peribronchial in distribution. Nodules may cavitate and lung attenuation.
are the precursors of thin-walled cystic lesions. The dis- Aetiological factors include the idiopathic type, inher-
tribution and profusion of the nodules correlates with dis- ited factors, post viral infections, autoimmune disorders
ease activity [31]. and post bone marrow transplantation [32].
Pleuroparenchymal Fibroelastosis
Depositional Lung Diseases
Pleuroparenchymal fibroelastosis is a rare disorder, ob-
Lipid Storage Diseases served most often after allogenetic bone marrow trans-
plantation, although it has previously been reported as an
Niemann-Pick disease is an inherited defect in sphin- idiopathic, and related to dyskeratosis congenita. Fibrosis
gomyelinase production and has five clinical variants re- of the peripheral lung parenchyma with pleural thicken-
sulting in sphingomyelin deposition within the lung, liv- ing occurs often with pneumothorax. The lung apices are
er, bone marrow and brain. HRCT shows a diffuse retic- often more severely affected [33].
ulonodular pattern with nodules 12 mm in size due to
the accumulation of aggregates of large multi-vaculolated Filamin A Protein Deficiency Related Progressive Multilobar
foam cells deposited in the lungs. Emphysema
Gauchers disease is due to a deficiency of beta-glu-
cosidase (the enzyme catabolizer of glucosylceramide), The role of the filamin proteins in lung growth was not
resulting in the accumulation of the latter in reticulo- previously recognized, but their importance for normal
endothelial cells of the liver, spleen, lymph nodes, bones lung development and growth is crucial. Filamin protein
and the brain (especially in the infantile form of the dis- A (FLNA) encodes actin-binding cytoskeletal scaffold-
ease). The Gaucher cells then accumulate in the alveolar ing protein, which is involved in neuronal migration,
interstitial and adjacent air spaces, leading to diffuse mil- cardiovascular development and connective tissue in-
iary or reticulo-nodular patterns that may be associated tegrity.
with lytic rib lesions. Mutations of FLNA are associated with multisystemic
disorders, including periventricular cerebral nodular het-
Aspiration Pneumonitis erotopia, cardiac valve dysplasia, aneurysms and Ehlers-
Danlos variants. Infants normally present with progres-
Children with chronic gastro-oesophageal reflux with sive respiratory deterioration, sometimes complicated by
silent or clinically overt aspiration of gastric contents pulmonary hypertension.
can develop very severe pulmonary dysfunction with as- Chest X-ray and CT demonstrate hyperlucent lung
sociated radiological changes. Chest X-ray initially parenchyma with peripheral pulmonary vascular attenua-
shows air space disease, with subsequent interstitial fi- tion, especially in the upper and middle lobes [34].
brosis. In the infant, recurrent episodes of unexplained
respiratory distress related to feeding, together with peri-
hilar consolidations especially in the dependent right up- Conclusion
per and superior segments lower lobe should suggest as-
piration pneumonitis. Diffuse ILD in children comprises a rare but heteroge-
In the more chronic untreated forms, diffuse airway neous group of conditions, many of which have no known
and interstitial changes with peripheral pulmonary fibro- cause. There is a clear need for definitive histological
sis ensue. In these more advanced cases lung biopsy classification, which, along with the increasing experi-
shows numerous fat-laden macrophages, confirming the ence in HRCT, will aid us in our ability to diagnose and
diagnosis of aspiration-induced pulmonary fibrosis. follow these unusual, fascinating conditions.
The Radiology of Diffuse Interstitial Pulmonary Disease in Children: Pearls, Pitfalls and Newly Recognised Disorders 239
References 18. Fan L (1998) Pediatric interstitial lung disease. In: Schawrz
ML, King TE (eds) Interstitial lung disease. BC Decker,
1. Koh DM, Hansell DM (2000) Computed tomography of diffuse Hamilton, ON, pp 103-118.
interstitial lung disease in children. Clin Radiol 55:659-667. 19. Copley SJ, Coren M, Nicholson AG et al (2000) Diagnostic
2. Bush A, du Bois R (1996) Congenital and paediatric intersti- accuracy of thin-section CT and chest radiography of paedi-
tial disease. Curr Opin Pulm Med 2:347-356. atric interstitial lung disease. AJR Am J Roentgenol 172:549-
3. Reid L (1984) Lung growth in health and disease. BR J Dis 554.
Chest 74:113-139. 20. Seely JM, Effmann EC, Muller NL (1997) High resolution CT
4. McGowan SE (1992) Extracellular matrix and the regulation of the paediatric chest: imaging findings. AJR Am J
of lung development and repair. FASEB J 6:2895-2904. Roentgenol 168:1269-1275.
5. Minoo O, King RJ (1994) Epithelial-mesenchymal interactions 21. Fishback N, Koss M (1996) Update on lymphoid interstitial
in lung development. Annu Rev Physiol 56:13-45. pneumonitis. Curr Opin Pulm Med 2:429-433.
6. Hislop A, Wigglesworth JS, Desai R (1986) Alveolar develop- 22. Albafouille V, Sayegh N, Coudenhove SD et al (1999) CT scan
ment in the human fetus and infant. Early Hum Dev 13:1-11. patterns of pulmonary alveolar proteinosis in children. Pediatr
7. Thurlbeck WM (1982) Postnatal human lung growth. Thorax Radiol 28:147-152.
37:564-571. 23. Swensen SJ, Tashijian JH, Myers JL et al(1996) Pulmonary
8. Dunnill MS (1982) The problem of lung growth. Thorax veno-occlusive disease. CT findings in eight patients. AJR Am
37:561-563. J Roentgenol 167:937-940.
9. Thibeault DW, Heimes B, Rezaiekhaligh M, MabryS (1993) 24. Cheah FK, Sheppard MN, Hansell DM (1993) Computed to-
Chronic modifications of lung and heart development in glu- mography of diffuse pulmonary haemorrhage with pathologi-
cocorticoid treated newborn rats exposed to hyperoxia or room cal correlation. Clin Radiol 48:89-93.
air. Pediatr Pulmonal 6:81-88. 25. Fink CW, Cimaz R (1997) Early onset sarcoidosis: not a be-
10. Chan KN, Noble-Jamieson CM, Elliman A et al (1989) Lung nign disease. J Rheum 24:174-177.
function in children of low birth-weight. Arch Dis Child 26. Tsagris VA, Liapi-Adamidou G (1999) Sarcoidosis in infancy:
64:1284-1293. a case with pulmonary involvement as a cardinal manifesta-
11. Lawler SD, Ukaejiojo, Reeves BR (1975) Interaction of mater- tion. Eur J Paediatr 158:258-260.
nal and neonatal cells in mixed-lymphocyte cultures. Lancet 27. Gurney JW (1992) Hypersensitivity pneumonitis. Radiol Clin
2:1185-187. North Am 30:1219-1230.
12. Miyawaki TH, Seki H, Taga K et al (1985) Dissociated pro- 28. Hansell DM, Wells AU, Padley SP, Muller NL (1996) Hyper-
duction of interleukin-2 and immune gamma interferon by sensitivity pneumonitis: correlation of individual CT patterns
phytohaemagglutinen stimulated lymphocytes in healthy in- with functional abnormalities. Radiology 199:123-128.
fants. Clin Exp Immunol 59:505-511. 29. Adler BD, Padley SP, Muller NL et al (1992) Chronic Hyper-
13. Lewis DB, Yu CC, Meyers J et al (1991) Cellular and molec- sensitivity pneumonitis: high resolution CT and radiographic
ular mechanisms for reduced interleukin 4 and interferon gam- features in 16 patients. Radiology 185:91-95.
ma production by neonatal T cells. J Clin Invest 87:194-202. 30. Ha SY, Helms P, Fletcher M et al (1992) Lung involvement in
14. McCarthy KM, Gong JL, Telford, Schneeberger EE (1992) Langerhans cell histiocytosis: prevalence, clinical features
Ontogeny of 1a accessory cells in fetal and newborn rat and outcome. Paediatrics 89:466-469.
lung. Am J Respir Cell Mol Biol 6:349-356. 31. Egeler RM, DAngio GJ(1995) Langerhancellhistiocytosis. J
15. Grenier P, Valeyre D, Cluzel P et al (1991) Chronic diffuse in- Paediatr 127:1-11.
terstitial lung disease: diagnostic value of chest radiography 32 Heath D, Segel N, Bishop J (1966) Pulmonary veno occlusive
and high resolution CT. Radiology 179:123-132. disease: clinical manifestations. Circulation 34:242-248.
16. Padley SPG, Hansell DM, Flower CDR, Jennings P (1991) 33 von der Thsen JH, Hansell DM, Tominaga M et al (2011)
Comparative accuracy of high resolution CT and chest radio- Pleuroparenchymal fibroelastosis in patients with pulmonary
graphy in the diagnosis of diffuse infiltrative lung disease. Clin disease secondary to bone marrow transplantation. Mod Pathol
Radiol 44:222-226. 24:1633-1639.
17. Copley SJ, Coren M, Nicholson AJ et al (2000) Diagnostic ac- 34. Guillerman RP, Metwalli ZA, Burrage LC et al (2013) Con-
curacy of thin-section computed tomography and chest radi- genital multilobar pseudo-emphysema: a severe progressive
ograph in paediatric interstitial lung disease. AJR Am J lung growth disorder associated with filamin a gene mutations.
Roentgenol 174:549-554. Pediatr Radiol 43 (Suppl 2):S304.
IDKD 2015-2018
Introduction
Respiratory distress is a common problem in the neonate,
and imaging plays an important part in the initial diag-
nosis and management of these infants. This chapter dis-
cusses the most common medical and surgical conditions
of the lung encountered in the newborn, with emphasis
on the role of a multi-modal evaluation for surgical
planning.
Medical Diseases
Approximately half of all premature infants suffer from
some form of respiratory distress. The most common
causes include surfactant deficiency (50%), transient
tachypnea (4%), and sepsis (2%). In the full-term infant,
transient tachypnea, meconium aspiration syndrome, and
persistent pulmonary hypertension are the most frequent-
ly encountered etiologies of acute respiratory failure. Be-
cause there is a great deal of overlap in the radiographic Fig. 1. Surfactant deficiency. A 27-week-gestation infant on the first
appearance of several conditions, knowledge of the pa- day of life. Portable chest radiograph shows hypoinflated lungs
with scattered air-bronchograms
tients gestational age and post-partum age, the clinical
setting, and the degree of respiratory support is crucial in
accurately assessing the underlying pathological process
and its progression or resolution. Standard treatment regimens include positive pressure
ventilation and endotracheal administration of exoge-
Surfactant Deficiency (Hyaline Membrane Disease) nous surfactant. Treatment with surfactant typically re-
sults in symmetric central clearing of the lungs. Howev-
Alveoli that are deficient in surfactant tend to collapse, er, the uneven distribution of surfactant may result in
resulting in generalized hypoinflation of the lungs, venti- over-aeration of one lung or lung segment (Fig. 2). Un-
lation-perfusion mismatch, and poor gas exchange. complicated surfactant deficiency typically resolves
Breathing causes overdistention of open alveoli and shear within the first week of life. However, high levels of in-
stresses on the fragile lung. As a result, there is leakage spired oxygen and inspiratory pressures may lead to pro-
of proteinaceous fluid into the alveolar lumen (hyaline longed respiratory difficulties, with eventual fibrosis
membranes), which further impairs gas exchange and and emphysema [1].
worsens the respiratory distress.
The typical radiographic findings are hypoinflated Pneumonia
lungs with a diffuse, reticulo-nodular appearance to the
lungs and air bronchograms, caused by fluid-filled alve- The incidence of pneumonia in the newborn appears to be
oli interspersed with air-filled sections of lung (Fig. 1). inversely proportional to gestational age, occurring in up
Meconium Aspiration
Meconium aspiration syndrome consists of fetal acidosis
and subsequent hyperperistalsis, passage of meconium,
fetal gasping, and aspiration of meconium into the tra-
chea and lungs [2]. Thick meconium may cause physical
obstruction of the distal airways, resulting in patchy, al-
ternating areas of pulmonary overinflation and atelecta-
sis. Meconium also inactivates surfactant and causes sec-
ondary surfactant deficiency and poor lung compliance.
The clinical course of many patients with meconium as-
piration syndrome is complicated by pulmonary hyper-
tension, which increases illness severity and worsens
clinical outcomes. The radiographic features of meconi-
um aspiration vary widely, depending on the severity of
aspiration. The mildest cases manifest as pulmonary
hyperinflation with scattered interstitial pulmonary
opacities, while in more severe cases patchy alveolar and
Fig. 3. Neonatal pneumonia. A 26-week-gestation infant at 2 weeks
of life. Portable chest radiograph shows a focal right upper lobe
interstitial infiltrates and more pronounced hyperinflation
pneumonia with air bronchograms superimposed on early chronic are seen, often together with pneumothorax and pneumo-
lung disease mediastinum as complications (Fig. 5).
242 G.A. Taylor
Fig. 5. Meconium aspiration. A 39-week-gestation infant on the Fig. 6. Pulmonary air leak. A 27-week-gestation infant with surfac-
first day of life. Heavy meconium staining and low APGAR scores tant deficiency requiring positive pressure ventilation. Chest X-ray
at birth were recorded. Portable chest X-ray shows hyperinflated shows tubular and flame-shaped lucencies throughout the left lung
lungs with coarse bilateral interstitial opacities and focal right up- due to pulmonary interstitial emphysema. A large left pneumotho-
per lobe atelectasis rax is also present
a b
Fig. 7 a, b. Bronchial atresia and congenital pulmonary airway malformation. A 36-week-gestation infant at 2 weeks of life. Prenatal sonography
had depicted a focal right lower lobe lesions (not shown). Chest radiograph (a) shows an ill-defined opacity in the right lower lobe (arrows). Coro-
nal reconstruction of a contrast-enhanced chest CT (b) shows a hyperlucent segment of the right lower lobe with a lobulated, mucus-filled bron-
chocele (arrow) due to segmental bronchial atresia
Scintimammography
Einat Even-Sapir
Sourasky Medical Center, Tel Aviv, Israel
Nuclear medicine has been i nvolved in the field of func- and the breast hangs freely through a small ring of de-
tional tumor imaging for decades. Its hallmark is the de- tectors. The first system to become commercially avail-
tection of viable tumor tissue based on functional and bi- able was a stationary flat detector-based PEM scanner
ological characteristics rather than on an altered tumor that used limited-angle tomosynthetic reconstruction
morphology. Scintigraphy has an important role in mon- with a spatial resolution of 2.4 mm [5]. A newer design
itoring the response to therapy, by distinguishing active is mammography with molecular imaging (MAMMI), a
tumor mass, where the tumor-seeking tracer still accu- dedicated breast PET system that uses a complete ring of
mulates, from residual mass composed of necrosis or detectors for full tomographic image reconstruction with
scar, where tracer no longer accumulates. The use of a spatial resolution of 1.6 mm [6]. Dedicated breast PET
scintigraphy in tumor detection depends on the mecha- systems were shown to be significantly more sensitive for
nism of tracer uptake in the tumor, the pharmacokinetic the detection of breast cancer than either whole-body
and normal biodistribution of the tracer, and on the tech- PET or PET-CT [7].
nology of the detecting system. The tracer used for breast imaging with a gamma-cam-
The detection of breast cancer using single-photon- era is 99mtechnetium (Tc)-sestaMIBI. The mechanisms of
emitting tracers with a gamma camera and positron-emit- the enhanced cellular uptake of 99mTc-sestaMIBI in can-
ting tracers with PET technology has been an ongoing cer cells is the subject of ongoing investigation. 99mTc-
challenge of nuclear medicine [1]. The whole-body sys- sestaMIBI is a small lipophilic cation whose uptake is
tems used originally in scintimammography were sub- nine times higher in tumor tissue than in normal tissue,
optimal for the detection of small lesions in the breast, as reflecting the high vascularity and mitochondrial activity
the distance from the collimator to the breast is approxi- of the former [8, 9]. The major advantage of 99mTc-
mately 15 cm and to the region of interest up to 25 cm. sestaMIBI in breast cancer imaging is its wide availabil-
Instead, dedicated systems for breast imaging were need- ity in the routine practice of nuclear medicine depart-
ed, which are now commercially available. ments, as it is also used in cardiac perfusion studies.
Gamma-cameras dedicated to breast imaging may be Moreover, unlike PET tracers, it does not require cy-
single- or dual-headed cameras; in the latter, each breast clotron facilities for its production.
is positioned between the two detectors in two views sim- The most commonly used tracer for scintimammogra-
ilar to the acquisition mode on mammography, with no phy with positron-emission technology is 18F-fludeoxyglu-
pressure. Detectors can be the NaI (Tl) detectors rou- cose (FDG). As a rule of thumb, 18F-FDG uptake is high-
tinely used in nuclear medicine, a technology referred to er in breast tumors with prognostically unfavorable char-
as breast-specific gamma imaging (BSGI), or the new acteristics [10]. Primary tumor 18F-FDG uptake was found
generation of pixilated semiconductor [cadmium zinc tel- to correlate with tumor size, histological type and grade,
luride (CZT)] detectors with improved spatial resolution. pleomorphism, lymphatic invasion, high Ki-67 level, and
A dual-headed system composed of CZT detectors, the triple negativity (i.e., negative for the estrogen receptor, the
molecular breast imaging system (MBI), was reported to progesterone receptor, and the human epidermal growth
have a high sensitivity in the detection of clinically rele- factor receptor 2) [11]. Yet, it should be borne in mind that
vant breast lesions as small as 3 mm [2-4]. 18F-FDG avidity is a characteristic of the individual tumor
Several dedicated breast imaging systems for positron- and some tumor types, such as ductal carcinoma in situ and
emitting tracers have been designed. The earliest one, lobular carcinoma, may show only low-intensity uptake [7,
positron emission mammography (PEM), uses two oppo- 12]. Moreover, when dedicated PET systems are used for
sitely placed detectors, as in mammography. With anoth- breast imaging, heterogeneous uptake in a large tumor
er prototype, dedicated breast PET, the patient is prone probably reflects tumor heterogeneity within the mass [13].
Publications on the clinical use of scintimammography patients response to therapy, and to the antiangiogenic ef-
have explored its role in tumor detection in patients with fects of the chemotherapeutic agent, which can alter the
no known malignancies and as a diagnostic tool in pa- contrast kinetics thus leading to an overestimate of the re-
tients with diagnosed breast cancer. The reported sensi- sponse [1]. Assessing response to therapy by means of
tivity of BSGI for the detection of breast cancer is functional rather than anatomic changes is another indi-
78100% for all tumor types, including lobular carcino- cation for scintimammography. 99mTc-sestaMIBI is a sub-
ma [2, 14]. Using the MBI system, researchers at the strate of the transmembrane P-glycoprotein (Pgp-170)
Mayo Clinic were able to detect small malignant lesions present in cells overexpressing the multidrug resistance
of 3 mm and calculated a sensitivity of 90% for abnor- gene (MDR1). Pgp acts as a protective pump by extrud-
malities 520 mm in size. Scintigraphy was found to be ing out of the tumor cells a wide range of molecules, in-
a valuable procedure when routinely used screening cluding 99mTc-sestaMIBI. Studies performed in humans
modalities, mainly mammography, were suboptimal, such demonstrated that in tumors with high levels of Pgp,
as in the case of patients with dense breasts. Based on the 99mTc-sestaMIBI efflux was significantly faster than in
screening of 936 at-risk women, Rhodes et al., of the the control group or in tumors with no Pgp, suggesting a
Mayo Clinic, reported a sensitivity of 27% for mammo- role for 99mTc-sestaMIBI scintigraphy in monitoring the
graphy alone and 91% for combined mammography and response to chemotherapy [8, 9]. Serial 99mTc-sestaMIBI
MBI [15]. This group reported the good performance of scintimammography was shown to accurately predict tu-
MBI with a low-dose of 8 mCi 99mTc-sestaMIBI, with mor response in patients with locally advanced breast car-
ongoing efforts aimed at dose-reduction to perform MBI cinoma undergoing neoadjuvant chemotherapy. Patients
with 4 mCi, with an effective dose twice that of a screen- with residual high uptake had poorer disease-free survival
ing mammogram. and overall survival, and uptake at completion of treat-
A recently published meta-analysis evaluated eight stud- ment was found to be an independent prognostic factor
ies on the use of 18F-FDG-PEM, comprising 873 women [18]. In a recent study, changes in the tumor to back-
with breast lesions. The pooled sensitivity and specificity ground ratio on MBI images performed at 35 weeks
values of PEM were 85% and 79%, respectively, accord- following the initiation of neoadjuvant chemotherapy
ing to a per-lesion-based analysis [16]. Currently, however, were found to accurately predict the presence or absence
scintigraphy is not recommended as a routine screening of residual disease at the completion of treatment [19].
tool because of the higher total body radiation compared to An MBI system was installed at our center in 2009. We
routinely used breast imaging modalities. approached our clinicians, breast surgeons, oncologists,
Accurate assessment of disease extent in the breast is and breast radiologists in conducting a registry research
essential to optimize the treatment approach in patients aimed at understanding the potential complementary role
with newly diagnosed breast cancer. As in the case of MRI, of MBI in the imaging algorithm of breast imaging. In
scintimammography can identify additional, unexpected patients with unknown cancer, indications for MBI in-
sites of disease, resulting in a change in diagnosis from lo- cluded genetic and familial high-risk; equivocal findings
calized to multifocal, multicentric, or even bilateral dis- on mammography, ultrasound (US), and/or MRI; nipple
ease. In a retrospective study of 159 women with one sus- discharge with no suggested breast abnormality; con-
picious breast lesion on physical exam and/or mammogra- traindication for MRI or claustrophobia in patients with
phy, BSGI detected additional suspicious lesions in 29% equivocal mammography and US assessment; and further
and occult cancer in 9%, both in the same breast as the in- assessment of the contralateral breast. In patients with di-
dex lesion (6%) and in the contralateral breast (3%) [12]. agnosed breast cancer, the indications for MBI were: as-
PEM and MRI were performed and compared in 388 sessment of the disease extent in the breasts of patients
patients with diagnosed cancer who were scheduled for with newly diagnosed solitary breast cancer planned for
surgery. Additional cancers were found in 21% of the lumpectomy, in which the surgeon was uncertain that dis-
women. PEM and MRI had comparable breast-level sen- ease was localized; in patients with locally advanced dis-
sitivity, although MRI had greater lesion-level sensitivity ease before and after neoadjuvant therapy; in assessing
and more accurately depicted the need for mastectomy. the presence of residual disease after surgery or of sus-
PEM had greater specificity at the breast and lesion lev- pected recurrence; to search for a primary occult tumor
els. In 3.6% of the women, tumors were seen only on in patients with metastatic axillary lymph nodes of breast
PEM [16]. In another study of 367 patients, 4% had con- cancer; and in follow-up.
tralateral cancer. The sensitivity and specificity of MRI in Scintimammography technology and clinical cases re-
detecting these lesions was 93% and 89.5% respectively, flecting the complementary role of scintimammography
compared to 73% and 95%, respectively, for PEM [17]. are discussed in the workshop for this chapter.
In patients with locally advanced disease receiving
neoadjuvant therapy, monitoring response to therapy by
morphologic imaging modalities reduces potential errors References
in interpretation due to findings of fibrosis or scar for-
mation and tumor fragmentation rather than concentric 1. Fowler AM (2014) A molecular approach to breast imaging. J
shrinkage, which can lead to an underestimation of the Nucl Med 55:177-180.
Scintimammography 249
2. Sun Y, Wei W, Yang HW, Liu JL (2013) Specific gamma imag- prognostic factors in breast carcinoma. Nucl Med Commun
ing as an adjunct modality to mammography for diagnosis of 34:1055-1067.
breast cancer: a systemic review and meta-analysis. Eur J Nu- 12. Buck A, Schirrmeister H, Khn T et al (2002) FDG uptake in
cl Med Mol Imaging 40:450-463. breast cancer: correlation with biological and clinical prog-
3. Brem RF, Shahan C, Rapleyea JA et al (2010) Detection of oc- nostic parameters. Eur J Nucl Med Mol Imaging 29:1317-
cult foci of breast cancer using breast-specific gamma imag- 1323.
ing in women with one mammographic or clinically suspicious 13. Koolen BB, Vidal-Sicart S, Benlloch Baviera JM, Valds Ol-
breast lesion. Acad Radiol 17:735-743. mos RA (2014) Evaluating heterogeneity of primary tumor
4. Hruska CB, Phillips SW, Whaley DH et al (2008) Molecular (18)F-FDG uptake in breast cancer with a dedicated breast
breast imaging: use of a dual-head dedicated gamma camera PET (MAMMI): a feasibility study based on correlation with
to detect small breast tumors. AJR Am J Roentgenol PET/CT. Nucl Med Commun 35:446-452.
191:1805-1815. 14. Brem RF, Ioffe M, Rapelyea JA et al (2009) Invasive lobular
5. MacDonald L, Edwards J, Lewellen T et al (2009) Clinical carcinoma: detection with mammography sonography MRI,
imaging characteristics of the positron emission mammogra- and breast-specific gamma imaging. AJR Am J Roentgenol
phy camera: PEM Flex Solo II. J Nucl Med 50:1666-1675. 192:379-383.
6. Moliner L, Gonzalez AJ, Soriano A et al (2012) Design and 15. Rhodes DJ, Hruska CB, Phillips SW et al (2011) Dedicated
evaluation of the MAMMI dedicated breast PET. Med Phys dual-head gamma imaging for breast cancer screening in
39:5393-5404. women with mammographically dense breasts. Radiology
7. Kalinya JE, Berg WA, Schilling K et al (2014) Breast cancer 258:106-118.
detection using high-resolution breast PET compared to whole- 16. Berg WA, Madsen KS, Schilling K et al (2011) Breast cancer:
body PET or PET/CT. Eur J Nucl Med Mol Imaging 41:260-275. comparative effectiveness of positron emission mammography
8. van Leeuwen FW, Buckle T, Kersbergen A et al (2009) Nonin- and MR imaging in presurgical planning for the ipsilateral
vasive functional imaging of P-glycoprotein-mediated doxoru- breast. Radiology 258:59-72.
bicin resistance in a mouse model of hereditary breast cancer 17. Berg WA1, Madsen KS, Schilling K et al (2012) Comparative
to predict response, and assign P-gp inhibitor sensitivity. Eur J effectiveness of positron emission mammography and MRI in
Nucl Med Mol Imaging 36:406-412 the contralateral breast of women with newly diagnosed breast
9. Vecchio SD, Zannetti A, Salvatore B et al (2006) Functional cancer. AJR Am J Roentgenol 198:219-232.
imaging of multidrug resistance in breast cancer. Phys Med 18. Dunnwald LK, Gralow JR, Ellis GK et al (2005) Residual tu-
21(Suppl 1):24-27. mor uptake of [99mTc]-sestaMIBI after neoadjuvant chemo-
10. Koolen BB, Vrancken Peeters MJ, Wesseling J et al (2012) As- therapy for locally advanced breast carcinoma predicts sur-
sociation of primary tumour FDG uptake with clinical, vival. Cancer 103:680-688.
histopathological and molecular characteristics in breast can- 19. Mitchell D1, Hruska CB, Boughey JC et al (2013) 99mTc-
cer patients scheduled for neoadjuvant chemotherapy. Eur J sestaMIBI using a direct conversion molecular breast imaging
Nucl Med Mol Imaging 39:1830-1838. system to assess tumor response to neoadjuvant chemotherapy
11. Ekmekcioglu O, Aliyev A, Yilmaz S et al (2013) Correlation in women with locally advanced breast cancer. Clin Nucl Med
of 18F-fluorodeoxyglucose uptake with histopathological 38:949-956.
IDKD 2015-2018
Since the first edition of the magnetic resonance imaging geneous enhancement, and not very high in signal on
(MRI) section of the illustrated BI-RADS lexicon, the bright fluid imaging.
field of breast MRI has grown enormously. Understand-
ably, certain terms have been added or deleted as better
terms have been identified. Additionally, concepts such Mass
as background parenchymal enhancement (BPE) have
been proposed and included. A section on nonenhancing A mass is an area of enhancement with an epicenter and
findings has been added. A section devoted to implant convex borders, existing as a three-dimensional (3D)
description and assessment is new. structure. The committee decided not to assign a size re-
quirement, recognizing that suspicious masses can be of
all sizes. In general, as the mass size increases so does the
Focus likelihood of malignancy.
Mass descriptors for shape and margins have been
An enhancing focus is a tiny round pin-point dot of en- adopted from the mammography BI-RADS lexicon. In
hancement that is seen only on the postcontrast images. general, as with mammography, circumscribed oval or
It is a round, homogeneously enhancing area with cir- round masses are seen more with benign lesions, where-
cumscribed margin. In general, foci are too small to ex- as irregular, spiculated masses are more likely to be ma-
hibit internal enhancement characteristics. Foci can be lignant. Care should be taken with respect to morpholo-
found in women of any age and menopausal status. A fo- gy, as round circumscribed masses on MRI represent can-
cus can be benign or malignant. cer more frequently than at mammography. There are sev-
In general, foci are a few millimeters in size; however, eral possible explanations for the presence of morpho-
applying strict size criteria is not favored. It must be not- logically benign lesions representing cancer on MRI.
ed that cancers 5 mm can be identified in the breast on First of all, MRI does not have the spatial resolution of
MRI. If margins and internal enhancement can be as- mammography with current techniques and field strength
sessed, then the lesion would be considered a mass. As so that margin analysis internal enhancement may suffer.
techniques improve, with higher resolution capabilities, Second, the cancers with benign morphology on MRI are
fewer lesions will be described as a focus and more will usually small, and smaller than we might be used to de-
be classified as a mass. tecting on mammography. As with most imaging tech-
If a focus is not unique, it is likely to represent a com- niques, the ability to resolve margins depends on the size
ponent of the patients background parenchymal en- of the lesion. Kinetic evaluation is important when con-
hancement (BPE). If it is unique and separate from the sidering these morphologically benign appearing lesions.
BPE, it may warrant evaluation. Margin analysis, kinetic As with other imaging techniques, the worst feature of
analysis, internal enhancement and T2-weighted charac- the lesion under evaluation should be used to determine
teristics can be evaluated to determine whether a focus is the need for biopsy.
likely to be benign or suspicious. Imaging features that A mass has internal enhancement that can be charac-
favor benign etiology are circumscribed, persistent, ho- terized. In general, homogeneous enhancement and
mogeneous and very high signal intensity on bright fluid nonenhancing internal septations indicate a possible be-
imaging. Imaging characteristics that favor malignancy nign process. While it is certainly possible to see classic
are irregular, spiculated, wash out kinetics, rim or hetero- appearances of certain lesions, morphologic overlap can
occur between benign and malignant lesions; if there is
This chapter was previously published in: Hodler J, von Schulthess any doubt, biopsy should be performed. The committee
GK, Zollikofer ChL (eds) Musculoskeletal Diseases 2013-2016. recognizes that when masses become large and ill-de-
Springer-Verlag Italia, Milano 2013, pp. 275-281. fined they might be described as regional enhancement.
Diseases of the Chest and Heart 2015-2018,
DOI: 10.1007/978-88-470-5752-4_34 Springer-Verlag Italia 2015 250
Breast MRI BI-RADS: Second Edition Highlights 251
Mass analysis can benefit from bright fluid sequences As in the fifth edition of BI-RADS Mammography,
[i.e., T2 weighted or short-tau inversion recovery (STIR)] in the amount of fibroglandular parenchyma is not de-
addition to the postcontrast sequences obtained. In general, scribed using percentages. Unlike mammography, where
benign mass lesions can be increased in signal relative to fi- noncalcified breast lesions can be obscured by dense tis-
broglandular parenchyma on bright fluid imaging, particu- sue, breast MRI is able to easily reveal an enhancing sus-
larly cysts, lymph nodes and fibroadenomas. Cancers may picious lesion independent of breast composition. There-
or may not exhibit increased signal on bright fluid imaging. fore, the amount of fibroglandular parenchyma does not
Cancer can be heterogeneously high in signal on bright flu- adversely impact lesion detectability.
id imaging if the tumors are necrotic, cellular or mucinous.
Mucinous carcinoma and liposarcoma classically demon- Background Parenchymal Enhancement
strate very high signal on bright fluid sequences; however,
there are usually other suspicious features, such as irregu- As MRI is performed with intravenous contrast, the fi-
lar shape or noncircumscribed margins, that warrant biopsy. broglandular breast parenchyma can demonstrate contrast
enhancement. BPE refers to the normal enhancement of
the patients fibroglandular tissue on the first postcontrast
Non-Mass Enhancement image. BPE refers to both the volume of enhancement as
well as the intensity of enhancement, and an evaluation
In the first edition of the MRI section of the illustrated of background enhancement should take both into con-
BI-RADS lexicon, non-mass enhancement (NME) was sideration.
used to describe BPE as well as areas that are still con- The background enhancement is described as one of
sidered to be NME. With greater experience and under- the following:
standing of BPE, some terms have been removed and the Minimal
NME descriptors have been refined. NME describes en- Mild
hancement in a pattern that does not have convex borders Moderate
and may have intervening fat or normal fibroglandular Marked
tissue contained within the extent of the enhancement. Although these categories are roughly quartiles, as-
Clumped enhancement refers to enhancement that has signing strict percentages to indicate the degree of en-
the appearance of cobble-stones where there are small hancement is likely artificial, difficult to assess without
aggregates of enhancement that are variable in size and automation, and should be avoided. In general, BPE might
morphology. The term clumped refers to enhancement in not be evenly distributed throughout the entire breast. Due
a focal, linear or linear-branching, segmental or regional to preferential blood supply, there is the probability of
distribution. The term clumped on MRI is similar to the greater enhancement in the upper outer quadrant of the
term pleomorphic on mammography, as it indicates en- breast and along the inferior aspect of the breast (former-
hancement in varying shapes and sizes. As ductal carci- ly described as sheet-like enhancement). BPE might be
noma in situ (DCIS) can present with this morphologic more prominent in the luteal phase of the cycle if the pa-
pattern, a description of clumped usually indicates a need tient is premenopausal. Therefore, for elective examina-
for biopsy. The diagnosis of DCIS is usually made solely tions (e.g., high-risk screening), effort should be made to
on lesion morphology, as many times the kinetic appear- schedule the patient in the second week of her cycle (days
ance does not meet minimal threshold and the time in- 7-14) to minimize the issue of background enhancement.
tensity curves are not typical for malignancy. Despite scheduling the patient at the optimal time of her
cycle, BPE will still occur and the BPE terms should be
applied. Women in whom cancer has been diagnosed and
Report Organization MRI is performed for staging (i.e., diagnostic) should be
imaged with MRI regardless of the timing of the men-
Amount of Fibroglandular Tissue strual cycle or menstrual status.
The pattern of BPE can be variable from patient to pa-
MRI is unique in that 3D volumetric data can be acquired tient, though in general the pattern of BPE for an individ-
from the image, and separation of fat and fibroglandular ual is fairly constant. It is uncertain what the patterns of en-
parenchyma is performed relatively easily. There are no hancement mean, therefore description beyond the recom-
data comparing mammographic density (breast composi- mended descriptors is optional. There is some evidence that
tion) with MRI assessment of amount of fibroglandular BPE might indicate a level of risk for the development of
tissue. Density is a term that should be applied only to breast cancer, as therapeutic measures such as anti-estrogen
mammography. The amount of fibroglandular tissue therapy can decrease the level of BPE. However, BPE does
should be described as one of the following: not appear to affect the ability to detect breast cancer.
Almost entirely fatty BPE can occur regardless of the menstrual cycle or
Scattered fibroglandular tissue menopausal status of the patient. BPE might not be di-
Heterogeneous fibroglandular tissue rectly related to the amount of fibroglandular parenchy-
Extreme fibroglandular tissue ma present. Patients with extremely dense breasts at
252 E.A. Morris
mammography might demonstrate little or no BPE, examination is technically unacceptable (e.g., poor fat
whereas patients with mildly dense breasts might demon- suppression, poor positioning) and would not be suffi-
strate marked BPE. Nevertheless, most of the time, cient for interpretation, a meaningful report could not be
younger patients with dense breasts are more likely to issued and a category 0 may be issued. The MRI exami-
demonstrate BPE. nation has characteristics that make it unique in compar-
In general, BPE is progressive over time; however, sig- ison to mammography and ultrasound. The first and most
nificant and fast enhancement can occur on the first post- obvious difference is the use of a contrast agent. This
contrast image despite fast imaging techniques. BPE on adds the parameter of blood flow to morphology with the
MRI is unique to a patient as is breast density at mam- associated flow metrics that may be calculated. The sec-
mography. A description of background enhancement ond is the acquisition of the exact same number and se-
should be included in the breast MRI report. quences whether the examination is for screening or di-
Patterns of BPE are under investigation, as there is wide agnostic. As with mammography, BI-RADS 0 should be
variation in the appearance from woman to woman. BPE used in the screening setting only. In interpreting breast
may present as multiple foci either uniformly scattered or MRI, there is enough information on the properly per-
more focal in one area, described previously as stippled formed examination to decide to biopsy or recommend
enhancement. Stippled enhancement is a pattern of BPE; it short-term follow-up of a specific finding. MRI, like
is usually diffuse and symmetric, however it can present as mammography, can give a category 0 for prior MRIs be-
a focal finding (particularly in an area where cysts are fore a report is issued that for auditing purposes will be
found) suggesting focal fibrocystic disease. replaced by the final assessment rendered in the addend-
ed report once prior examinations do become available
Non-Enhancing Findings similar to a category 0 for technical reasons. This recom-
mendation may change in the future when MRI screening
Non-enhancing findings seen on the precontrast or bright becomes more commonplace.
fluid images are benign. Examples include cysts, duct ec- A final assessment of 0 is helpful when a finding on
tasia and some fibroadenomas and postoperative collec- MRI is suspicious but a benign corresponding finding
tions. Assessment of the absence of enhancement is best on an additional study would prevent a biopsy. If a cat-
made on the subtraction image. Follow-up or biopsy of egory 0 is given on MRI, then an explanatory note in the
areas of non-enhancement is not necessary unless there MRI report clarifying why this suspicious morpholo-
are suspicious findings on another imaging modality, gy is not immediately given a 4 or 5 is called for. For
such as mammography or ultrasound. example, if a mass is suspicious on MRI but there is a
possibility that it might represent a benign finding such
as a lymph node, a targeted ultrasound that would prove
Assessment Categories that the lesion is benign would prevent a biopsy. In the
case of an ultrasound recommendation following the
The final assessment should be based on the most suspi- MRI examination, the terms MRI directed or MRI
cious finding present in each breast. A separate BI- targeted ultrasound are preferable to second look ul-
RADS assessment for each breast should be stated after trasound, as it is not always certain that a first look ul-
the impression text. If the interpretation is straightfor- trasound has been performed. Another example where
ward and the same for both breasts, an overall impression category 0 would be useful is for a finding on MRI that
that includes both breasts may be used. The overall as- is most likely fat necrosis, but the reader would like to
sessment should be based on the most worrisome find- confirm and correlate the findings to a mammogram
ings present in each breast. For example, if benign find- that is not available.
ings, such as lymph nodes or cysts, are noted along with When additional studies are compared or completed, a
a more suspicious finding, such as a spiculated mass, the final assessment category attached to those additional
final assessment code should be reported category 4 or 5. studies would close out the MRI 0. When interpreting
Similarly, if immediate additional evaluation is needed MRI it is extremely helpful to have all available imaging
for one breast for a suspicious finding (with targeted ul- studies in order to give a complete report. If the addi-
trasound, for example), and there is a probably benign tional studies can be reported in the same report, separate
finding in the breast as well, the final assessment code paragraphs indicating the pertinent findings from each
for that breast would be category 4. If a breast with a imaging study can contribute to the final integrated as-
known cancer has an additional suspicious finding war- sessment that takes into consideration the findings of all
ranting biopsy, then the final assessment code for that imaging studies.
breast is category 4, not category 6.
Category 1
Category 0
This is a normal examination. A description of the fi-
Every effort should be made not to use category 0 in broglandular tissue and background parenchymal en-
reading breast MRI. However, in the event that the hancement should be included.
Breast MRI BI-RADS: Second Edition Highlights 253
assessment is benign (BI-RADS category 2). On the mined that tissue diagnosis is not appropriate, then a cat-
other hand, if there are, for example, residual suspi- egory 6 assessment should be rendered accompanied by
cious lesions, the appropriate assessment is category 4 the recommendation that subsequent management now
or 5. should be directed to the cancer. As for any examination
There is one other potentially confusing situation in- in which there is more than one finding, the management
volving the use of assessment category 6. This occurs section of the report might include a second sentence that
when, prior to complete surgical excision of a biopsy- describes the appropriate management for the finding(s)
proven malignancy, breast imaging demonstrates one or not covered by the overall assessment.
more possibly suspicious findings other than the known
cancer. Because subsequent management should first
evaluate them as yet undetermined finding(s), involving Suggested Reading
additional imaging, imaging-guided tissue diagnosis or
American College of Radiology (ACR) (2013) ACR BI-RADS,
both, it must be made clear that in addition to the known 5th Ed.
malignancy there is at least one more finding requiring DOrsi CJ, Mendelson EB, Morris EA et al (2012) Breast imaging
specific prompt action. The single overall assessment reporting and data system: ACR BI-RADS. American College
should be based on the most immediate action needed. If of Radiology, Reston, VA.
a finding or findings are identified for which tissue di- American College of Radiology (ACR) (2013) ACR BI-RADS
Magnetic Resonance Imaging, 2nd Ed.
agnosis is recommended, then a category 4 or 5 assess- Morris EA, Ikeda DM, Lehman C et al (2012) Breast Imaging Re-
ment should be rendered. If at additional imaging for porting and Data System. American College of Radiology, Re-
finding(s) other than the known malignancy, it is deter- ston, VA.
IDKD 2015-2018
a b
c d
fast-growing cancers with higher vascularity. Moreover, larity, higher vascularity in the periphery, and peripheral,
they are not reproducible when the examination condi- marginal vessels connecting with internal vessels, signif-
tions differ (device, choice of flow parameter, location of icantly differed from malignant patterns. Typical of ma-
measurement). Additionally, flow measurements are lignant lesions were radially aligned external vessels, ra-
time-consuming and require vast experience. By contrast, dial internal vessels connected with radial external ves-
morphological parameters are easier to assess, more re- sels, and a higher degree of internal than external vessels.
producible, and of a higher diagnostic yield [15]. Thus, It is worth mentioning that when using the Doppler func-
Madjar recommended simply observing the presence and tion, it is essential to apply only a slight pressure on the
number of vessels. Setting the cut-off for the depiction of probe and to keep the field of view (FOV) very small,
vessels at 2, he was able to reach a sensitivity and thus obtaining better Doppler signals (Fig. 3).
specificity for the detection of malignancy of 90% and To summarize, the essential features to be applied
93%, respectively [15]. Moreover, vascular pattern helps when evaluating a lesion with color Doppler are the pres-
to predict the probability of a lesions malignancy if used ence of vascularity, the number of vessels, and the vessel
together with other criteria [14]. In this context, it can be pattern. However, vascular categorization is not always
helpful to consider the exact course of the vessels adja- applicable due to poor vessel visualization and, even
cent to the lesion, as performed by Svennson [16] more importantly, vascular characterization can only be
(Fig. 2). Thus, benign vascular patterns, such as avascu- regarded as an adjunct to other b-mode features.
a b
Benign Malignant
a b
Fig. 3 a, b. Demonstration of the use of the color Doppler function in a strongly vascularized cancer lesion. A small region of interest, just
around the lesions, should be chosen to achieve better Doppler signals. Only very slight pressure should be applied on the probe (a). If
higher pressure is applied the Doppler signals, and thus vascularization, cannot be depicted (b)
Ultrasound elastography is a recently introduced exami- The manual compression applied on the probe can be de-
nation technique. In response to an external force, breast scribed as rapid and sinusoidal movements of the hand.
tissue is displaced, with stiff tissue being less displaced The induced tissue movements between the frames are
than soft tissue. Carcinomas exhibit an increased stiffness calculated by a dedicated software, and the resulting sig-
and thus less tissue displacement than normal breast tis- nals registered before and after tissue displacement are
sue, and this mechanical characteristic is used to differ- transformed into b-modes images. Strain data are dis-
entiate benign from malignant lesions. Thus, elastogra- played as either a black/white elastogram, with black rep-
phy is primarily a characterization rather than a detection resenting a stiff area and white a soft area, or a color map
tool, due to following physical properties: elasticity can [19]. Color-coding is not uniform among the manufac-
generally be seen as the relation of necessary stress (pres- turers, thus partially aggravating the interpretation. The
sure) to the relative change in length obtained (strain). It most commonly applied color map, primarily used by the
describes how much pressure has to be applied to elasti- Hitachi ultrasound device, is the one developed by Itoh et
cally deform a tissue, which is dependent on its intrinsic al. [20]. This elasticity score (the Tsukuba score, named
elasticity (Young`s modulus of elasticity) [17]. Young`s after the clinic where the work was conducted) has been
modulus of elasticity can therefore be considered as a de- validated in large series of histologically proven breast le-
scription of tissue elasticity according to the formula: sions and is based upon a 5-point scale (Fig. 4), with the
E / (elasticitypressure applied to the breast/ breast risk of malignancy increasing from a score of 1 (be-
tissue deformation under pressure). nign) to a score of 5 (malignant). Numerous studies
This differentiation criterion of absent strain (in- have shown that the additional use of an elasticity score
creased stiffness) has been evaluated in terms of its addi- increases specificity from 5683% to 6887% compared
tional benefit to normal b-mode features in a series of to normal b-mode images [21-23]. Another helpful tool is
studies, which demonstrated a general increase in speci- to compare the b-mode image with the elastographic im-
ficity [18]. Nevertheless, the conditions of application age: benign lesions on the elastographic display are typi-
and interpretation considerably vary between the differ- cally identical in size or smaller than on b-mode display,
ent elastography software and devices. Essentially, two whereas malignant lesions appear to be larger on elastog-
main elastography techniques can be distinguished: free- raphy than on b-mode display.
hand strain elastography and shear-wave elastography. Nevertheless, the main disadvantage of strain elastog-
Strain elastography can be further subdivided into real- raphy is that it is strongly operator-dependent such that
time-elastography (e.g., Hitachi) and tissue-Doppler/ tis- large images may vary widely [23, 24]. Attempts have
sue strain imaging (e.g., Toshiba), whereas shear-wave been made to reduce the interobserver variability by im-
elastography can be subdivided into genuine shear-wave plementing semi-quantitative region of interest (ROI)
elastography (e.g., Supersonic) and acoustic radiation measurements, such as the strain-ratio (strain of sur-
force impulse (AFRI) technology (e.g., Siemens). In the rounding fat tissue/ strain of the lesion to be investigat-
following, only real-time elastography, as the most com- ed). Lesion assessment was further improved by the ad-
monly applied form of strain elastography, and genuine ditional use of a cut-off value, which can vary between
shear-wave elastography are discussed in detail. 2.27 and 4.3 [21, 25, 26]. Values below the cut-off value
258 C. Kurtz
Tsukuba Score
1
benign
Tsukuba Score
2
benign
Tsukuba Score
3
probably benign
indicate a benign lesion whereas values above the cut-off pression/decompression, and angulation of the probe.
indicate a suspicious lesion. Although a certain improve- Hence, misinterpretations are possible by improper com-
ment in interobserver agreement [21] and accuracy [25] pression, deeply located lesions (3 cm), large lesions
has been obtained by the strain-ratio compared to the ba- (especially lesions larger than the FOV), non-perpendic-
sic Tsukuba-elastography scoring, the difficulty in accu- ular orientation of the probe, and if the lesions slides out
rately assigning a lesion to a certain category remains, of the FOV. When drawing the FOV, it is essential that the
since strain-ratio measurements are based upon the indi- lesion is surrounded by sufficient adjacent tissue to avoid
vidually generated elastography image. images of poor quality and misleading color coding; thus,
Moreover, free-hand elastography requires a certain it is advisable to draw the FOV as large as possible (from
training phase to learn the adequate manual skills to be the subcutaneous fat layer to the superficial fascia of pec-
applied [19]: pressure on the probe, frequency of com- toral muscle) (Fig. 5).
a b
Fig. 5 a, b. Demonstration of improper field of view (FOV) placement for strain elasticity measurements (a, b) of a hypoechoic lesion with
posterior shadowing on b-mode (a). The originally placed FOV covers only the lesion; thus, its estimation using the obtained color map re-
mains difficult (b). For correct calculations, the lesion has to be surrounded by enough adjacent tissue. Therefore, the size of the FOV
should reach from the subcutaneous fat layer to the superficial fascia of pectoral muscle (red rectangle in b)
Recent Developments in Breast Ultrasound with a Special Focus on Shear-Wave Elastography 259
a b
a b
Fig. 7 a, b. Visualization of shear waves in cystic lesions. Viscous cysts (a) support shear waves, thus demonstrating a blue color. Non-
viscous cysts (b) do not support shear waves, resulting in a black/absent shear-wave signal. As no increased stiffness is observed at the
margin of the lesion, it can still be classified as benign
Although shear-wave images can be interpreted more Stronger (pre-)compression: Overly strong pressure
reliably than strain elastography, difficulties remain in applied on the probe will cause the peak in the generated
image generation and assessment. shear waves to occur at an earlier time, thus simulating a
higher shear-wave speed, as observed in cancers (Fig. 8).
False-Positives and Strategies to Avoid Them Therefore, it is crucial to use only slight pressure, except
the lesion is small (5 mm ), since small lesions might
False positives can be generated by: (1) poorly de- be missed if insufficient pressure is used.
formable lesions (e.g. scar, larger calcifications, fibrous Insufficient FOV: As with strain imaging, the FOV
fibroadenoma, focal strong fibrosis); (2) too-strong com- should be as large as possible, with sufficient adjacent
pression; and (3) insufficient normal tissue surround- normal tissue surrounding the lesion. This is important
ing the lesion (too-small FOV). because the propagation of shear waves in the surround-
Poorly deformable lesions: This feature is peculiar to ing tissue is more visible such that misinterpretation will
the lesion and thus other imaging modalities, including be unlikely. In particular, artifacts (from too-strong pre-
mammography, b-mode, Doppler function, and compar- compression) arising between the subcutaneous fat layer
isons to a preceding imaging study of the same lesion and the lesion are easier to perceive (vertical-cord arti-
have to be taken into account. fact). The 4-pattern visual evaluation system described by
a b c d
Fig. 8 a-d. False positive findings by applying high pressure on the probe. Left picture (a) indicates correct pressure, since no vertical-cord
artifacts within the subcutaneous fat layer are seen. The subcutaneous fat layer is blue (white arrow). b-d With increasing pressure on the
probe, higher shear wave speed is simulated thus producing a color coding typical of cancer. The green-yellow background noise within
the subcutaneous fat layer indicates the artifact induced by too strong (pre-)compression (white arrow)
Recent Developments in Breast Ultrasound with a Special Focus on Shear-Wave Elastography 261
Tozaki [29] is very helpful in better differentiating be- Lesions larger than the FOV: Although the reasons are
tween real positive findings and artifacts (Fig. 9). still unclear, it seems to be related to the fact that the le-
If a pattern-2 vertical- cord artifact (Fig. 8 b-d) is sion is not surrounded by sufficient normal tissue such
observed during real-time elastography, then less pressure that only differences in the elasticity the lesion and the
should be applied until the artifact disappears (Fig. 8a). surroundings are too small [18].
Only if a colored area at the lesion`s margin persists can Small lesions: Normal (pre-)compression can be insuf-
the lesion be assumed to have increased stiffness. ficient to generate abnormal shear waves, given that the
area of stiffness is too small. Applying a higher (pre-)com-
False-Negatives and Strategies to Avoid Them pression on the probe often allows the visualization of
shear waves [19], even if the degree of stiffness is subtle.
False negatives occur due to the following: (1) deeply lo-
cated lesions (2.53 cm); (2) as a reflection of the can- Diagnostic Value of Shear-Wave Elastography
cer type (high stiffness in scirrhous cancer, less stiffness
in mucinous cancer or ductal carcinoma in situ); (3) le- A previous large multicenter trial [30] evaluated diverse
sions larger than the FOV (large single lesion, diffusely parameters of shear-wave elastography in terms of highest
infiltrating cancer); (4) small lesions (5 mm) diagnostic value if used in addition to the normal
Deep lesions: At a depth of approximately 2.53 cm, BIRADS classification. The results showed that the quali-
shear waves are hardly visible. Better visibility can some- tative parameter Ecol (observed color in or around the le-
times be gained by rotating the patient, so that the lesion sion) had the highest accuracy (AUCBIRADSEcol
under investigation is possibly displaced to a more su- 0.971), followed by the quantitative parameter of maxi-
perficial layer. mum elasticity Emax (kPa) (AUCBIRADSEcol 0.962).
Cancer type: If shear waves are absent on the standard For this reason, only the observation of the displayed col-
mode, they are more likely to be visualized using the pen- or is of major significance and might be sufficient for le-
etration mode (increased amplitude of the shear waves) sion assessment. The same study found that masses with
(Fig. 6). Nevertheless, in certain cancer types abnormal Emax3080 kPa had a malignancy rate of 8.9% and those
shear waves cannot be depicted; in such cases, other with Emax80 to 160 had a malignancy rate of 39.9%.
imaging modalities or further ultrasound techniques/fea- At an Emax 60 kPa, 74.1% of the masses were malignant.
tures have to be considered. In a preceding multinational study, shear-wave elastography
262 C. Kurtz
Table 1. Shear wave parameters for upgrading BIRADS 3 respectively downgrading BIRADS 4 a lesions whith an increase in specificity
and no change in sensitivity according to Berg [30]
BIRADS 3 parameters for BIRADS 4 parameters for Increase in
BIRADS upgrading BIRADS downgrading specificity
Ecol red Ecol black to light blue 61.178.5
Irregular shape on SW elastography oval shape on SW elastography 61.169.4
Emax160 kPa Emax 80 kPa 61.177.4
was shown to be particularly helpful in borderline lesions lesion (THI, compound imaging, color Doppler, 3D imag-
(between benign and malignant) according to b-mode cri- ing, elastography). But depending on the chosen ultra-
teria. By using these elastographic features as an adjunct to sound device, only a certain number of technical features
b-mode features, these lesions could be more accurately will be available with different diagnostic value. Irrespec-
assigned to either BIRADS 3 or 4 [30]. Initially classified tive of the technical component is utilized, ultrasound al-
BIRADS 4a lesions with a lack of stiffness were more fre- ways has to be considered in the context of further imag-
quently benign, and initially classified BIRADS 3 lesions ing modalities or technical components. This chapter has
with increased stiffness were more often malignant. With focused on THI, color Doppler, and elastography, as fast
the use of certain parameters for upgrading BIRADS 3 le- and easy methods that are also very helpful in lesion de-
sions and downgrading BIRADS 4a lesions, specificity tection and differentiation. In particular, shear-wave elas-
could be increased while maintaining sensitivity (Table 1). tography is considered a promising method because it is
The conclusion would be that unclear findings, e.g., a less operator-dependent than strain elastography and has a
complicated cysts planned for biopsy, could be more fre- significant increase in specificity. The major benefit is
quently sent to follow-up, and initially classified BIRADS seen in lesions primarily categorized as BIRADS 3 or 4a,
3 lesions with an increased stiffness but in fact malignant which can be more accurately assessed based on their
could be sent earlier to biopsy. However, shear-wave elas- elastographic features. In this regard, patients with suspi-
tography should not be used to change the strategy in typ- cious lesions (BIRADS 4a characterization without shear-
ical BIRADS 2 , BIRADS 4c and BIRADS 5 lesions, as it wave elastography) with benign features on elastography
might lead to misguided decisions. (dark blue color within the lesion and its surroundings)
Another study [31] was performed to exactly deter- could be spared from biopsy and instead be sent to follow-
mine the benefit of shear-wave elastography in ultra- up. Conversely, unclear lesions (BIRADS 3 characteriza-
sound-characterized BIRADS 3 lesions (primarily char- tion without shear-wave elastography) with suspicious
acterized without elastographic features). The aim was to elastographic features could be sent for biopsy with high-
search for features that would help to not send the patient er confidence. To reduce misinterpretations and avoid
to follow-up but rather to reclassify the lesion to a new misguided decisions, it is essential to be familiar with the
BIRADS 2 without downgrading a malignant lesion to reasons for false-positives and false-negatives. It is also
new BIRADS 2. Only the color (black to dark blue) and important that every imaging finding be evaluated in a
Emax 20 kPa were sufficiently safe and helpful criteria larger context, such as clinical complaints, findings in
to downgrade a benign-appearing BIRADS 3 lesions to a previous examinations, age of the patient, individual risk
new BIRADS 2, thus sparing patients from short-term situation, and comparison with other imaging methods.
follow-up.
Further advantages of shear-wave elastography cur-
rently being investigated or still to be evaluated are its ap- References
parent ability to more precisely depict the extent of inva-
sive lobular carcinomas compared to b-mode ultrasound, 1. Woo J (2004) A short history of the development. Radiol Clin
and to better visualize small satellite lesions in the sur- N Am 42:845-851.
2. Berg WA, Blume JD, Cormack JB et al (2008) Combined
roundings of the index cancer, irrespective of the tumor screening with ultrasound and mammography vs mammogra-
entity (ductal or lobular cancer). phy alone in women at elevated risk of breast cancer. JAMA
299:2151-2163.
3. Mundinger A, Wilson ARM, Weisman C et al (2008) Breast ul-
trasound- update. EJC Supplements 9:11-14.
Summary 4. ACR BI-RADS Ultrasound (2003). In: ACR Breast Imaging
Reporting and Data System, Breast Imaging Atlas. American
Nowadays, breast diagnostics have to be considered as a college of Radiology, Reston, VA.
complementary multimodality approach, one that is main- 5. Candelaria RP, Hwang L, Bouchard RR et al (2013) Breast ultra-
ly based on mammography and ultrasound and supple- sound: current concepts. Semin Ultrasound CT MR 34:213-225.
6. Mendelson EB, Bhm-Vlez M, Berg WA et al (2013) ACR
mented by MRI, whenever required. However, ultrasound BI-RADS Ultrasound. In: ACR BI-RADS Atlas, Breast
on its own has emerged as a complex examination modal- Imaging Reporting and Data System. American College of Ra-
ity, offering diverse technical features to characterize a diology, Reston, VA,
Recent Developments in Breast Ultrasound with a Special Focus on Shear-Wave Elastography 263
7. Sehgal CM, Weinstein SP, Arger PH et al (2006) A review of 20. Itoh A, Ueno E, Tohno E et al (2006) Breast disease: clinical
breast ultrasound. J Mammary Gland Biol Neoplasia 11:113- application of US elastography for diagnosis. Radiology
123. 239:341-350.
8. Rosen EL, Soo MS (2001) Tissue harmonic imaging sonogra- 21. Fischer T, Peisker U, Fiedor S et al (2012) Significant differ-
phy of breast lesions: improved margin analysis, conspicuity, entiation of focal breast lesions: raw data-based calculation of
and image quality compared to conventional ultrasound. Clin strain ratio. Ultraschall Med 33:372-379.
Imaging 25:379-384. 22. Thomas A, Degenhardt F, Farrokh A et al (2010) Significant
9. Szopinski KT, Pajk AM, Wysocki M et al (2003) Tissue har- differentiation of focal breast lesions: calculation of strain ra-
monic imaging: utility in breast sonography. J Ultrasound Med tio in breast sonoelastography. Acad Radiol 17:558-563.
22:479-487. 23. Thomas A, Kmmel S, Fritzsche F et al (2006) Real-time so-
10. Hartmann A, Kunz M, Kstlin S et al (1999) Hypoxia-induced noelastography performed in addition to B-mode ultrasound
up-regulation of angiogenin in human malignant melanoma. and mammography: improved differentiation of breast lesions?
Cancer Res 59:1578-1583. Acad Radiol 13:1496-504.
11. Milz P, Lienemann A, Kessler M et al (2001) Evaluation of 24. Yoon JH, Kim MH, Kim EK et al (2011) Interobserver vari-
breast lesions by power Doppler sonography. Eur Radiol ability of ultrasound elastography: how it affects the diagnosis
11:547-554. of breast lesions. AJR Am J Roentgenol 196:730-736.
12. Gokalp G, Topal U, Kizilkaya E (2009) Power Doppler sonog- 25. Zhi H, Xiao XY, Yang HY et al (2010) Ultrasonic elastography
raphy: anything to add to BI-RADS US in solid breast mass- in breast cancer diagnosis: strain ratio vs 5-point scale. Acad
es? Eur J Radiol 70:77-85. Radiol 17:1227-1233.
13. Carpentier GL, Roubidoux MA, Fowlkes JB et al (2008) Sus- 26. Mansour SM, Omar OS (2012) Elastography ultrasound and
picious breast lesions: assessment of 3D Doppler US indexes questionable breast lesions: does it count? Eur J Radiol
for classification in a test population and fourfold cross-vali- 81:3234-3244.
dation scheme. Radiology 249:463-470. 27. Cosgrove DO, Berg WA, Dor CJ et al (2012) Shear wave elas-
14. Raza S, Baum JK (1997) Solid breast lesions: evaluation with tography for breast massesis highly reproducible. Eur Radiol
power Doppler US. Radiology 203:164-168. 22:1023-1032.
15. Madjar H, Sauerbrei W, Hansen L (2011) Multivariate analy- 28. Cosgrove D, Piscaglia F, Bamber et al (2013) EFSUMB guide-
sis of flow data in breast lesions and validation in a normal lines and recommendations on the clinical use of ultrasound
clinical setting. Ultraschall Med 32:511-517. elastography. Part 2: Clinical applications. Ultraschall Med
16. Svensson WE, Pandian AJ, Hashimoto H (2010) The use of 34:238-253.
breast ultrasound color Doppler vascular pattern morphology 29. Tozaki M1, Fukuma E (2011) Pattern classification of Shear-
improves diagnostic sensitivity with minimal change in speci- Wave Elastography images for differential diagnosis be-
ficity. Ultraschall Med 31:466-474. tween benign and malignant solid breast masses. Acta Radiol
17. Krouskop TA, Wheeler TM, Kallel F et al (1998) Elastic mod- 52:1069-1075.
uli of breast and prostate tissues under compression. Ultrason 30. Berg WA, Cosgrove DO, Dor CJ et al (2012) Shear-wave
Imaging 20:260-274. elastography improves the specificity of breast US: the BE1
18. Fischer T, Sack I, Thomas A (2013) Characterization of focal multinational study of 939 masses. Radiology 262: 435-449.
breast lesions by means of elastography. Rofo 185:816-823. 31. Schfer FK, Hooley RJ, Ohlinger R et al (2013) ShearWave
19. Barr RG, Destounis S, Lackey LB 2nd et al (2012) Evaluation Elastography BE1 multinational breast study: additional
of breast lesions using sonographic elasticity imaging: a mul- SWE features support potential to downgrade BI-RADS-
ticenter trial. J Ultrasound Med 31:281-287. 3 lesions. Ultraschall Med 34:254-259.
IDKD 2015-2018
T2-weighted sequence
(water-specific images)
T1-weighted sequence
(fat-specific images)
Optional: T1 fast-spin
gadolinium-weighted sequence
Fig. 1. MRI standard proto-
col and findings according
to tissue characteristics
266 I. Thomassin-Naggara, M. Atlan, J. Chopier, E. Darai
a b
enhanced cosmetic results by conserving most of the skin The potential complications of autologous reconstruc-
envelope. The need for systematic imaging follow-up of tion include partial or total necrosis, pulmonary em-
these patients is unclear, including for patients with on- bolism, edema, hematoma, and seroma. After 12 years,
cological indications [11, 12]. Normal mammary gland is steatonecrosis (including potential sclerosing retraction
found in about 60% of patients after a skin-sparing mas- and calcification) may occur, with the risk depending on
tectomy, according to Tokin et al. [13]. In skin- and nip- the quality of the anastomosing pedicle. According to
ple-sparing mastectomy performed in oncology patients, Garvey et al., the rates of fat necrosis are higher for the
it should be noted that the risk of local recurrence corre- TRAM flap (58.5%) than for the DIEP flap (17.7%) [16].
lates with the margin status and the type of mastectomy. Moreover, abdominal hernias occur more frequently in
According to Kroll et al. [14], regardless of the recon- patients receiving a TRAM flap (16%) than a DIEP flap
struction technique, this risk is 46% at 5 years, similar (1%). Together, these considerations account for the pref-
to that in the group of patients not undergoing recon- erence for and further refinement of DIEP rather than
struction. Depending on the location of the recurrence, TRAM reconstructions [18, 19].
the patient may or may not be symptomatic. In 5070%
of cases, the recurrence develops in the anterolateral part MR Findings
of the reconstructed breast and is easily detected by the
clinician, but in 30% of cases the recurrence may be in In the follow-up of patients with a reconstructed
the axillary region or within the thoracic wall, which in breast(s), MRI is not systematically performed except in
either case is very difficult to diagnose clinically. When patients who are at high risk of breast and ovarian cancer.
the patient is symptomatic, the radiologist must be able to The radiologist may have to analyze the autologous re-
distinguish normal imaging features from features of ma- construction if the patient undergoes MRI for the con-
lignancy, which as a prerequisite requires an awareness of tralateral breast or if there is a suspicion of recurrence in
the different types of breast reconstruction options. the reconstructed breast. At MRI, the volume of the flap
decreases over time such that eventually it is composed
Autologous Flaps only of fat or muscle or of a fatty component, depending
on the type of autologous reconstruction. The radiologist
General Considerations should examine the remaining mammary gland and its lo-
cation to help the clinician to orient the sites of normal
The two most common types of conventional pedicled tissue and/or correlate normal tissue with a potential ab-
flaps are the latissimus dorsi myocutaneous (LDM) flap normality at clinical examination.
and the transverse rectus abdominis myocutaneous
(TRAM) flap; the latter is harvested from excess abdom- Autologous Fat Grafting
inal tissue. These autologous breast reconstruction meth-
ods allow the insertion of a muscle with a skin island and This technique of breast augmentation consists of har-
are mainly indicated when there is a lack of skin after vesting the fat by classic liposuction followed by treat-
conservative mastectomy. The LDM flap is vascularized ment of the harvested tissue (centrifugation, washing,
by the thoracodorsal pedicle whereas the epigastric supe- and/or filtration) to enhance the viability of the
rior vessels supply the TRAM flap. These flaps include adipocytes and mesenchymal stem cells, before their in-
the muscle to insure good vascularization of the skin and jection into the breast. Autologous fat grafting can be
fat. By contrast, free flaps, such as the deep inferior epi- used as the sole technique in a reconstruction or it may
gastric perforator (DIEP) flap, are becoming increasing- be combined with autologous-tissue- or implant-based
ly popular because they are muscle-sparing. Autologous breast reconstructions. If the autologous fat graft tech-
tissue transfer involves complete separation of the donor nique is used by itself, e.g., in breast augmentation, all
tissue. In the case of the abdominal excess transferred in the layers of the breast can be injected: pectoralis major
the DIEP flap, a pedicle is dissected in the muscle, and muscle, the retropectoral space, above the gland, and
there is absolutely no muscle within the flap (hence the within the gland. When it is combined with another tech-
name perforator flap). The harvested tissue is then con- nique of breast augmentation, the lipoaspirate is mainly
nected to the recipient site by an anastomosis, with mi- injected into the subcutaneous space above the gland.
crosurgical sutures performed under the microscope: the
deep inferior vessels are ligated to the internal mammary Imaging Features
vessels lying under the cartilage of the third or fourth rib
[15]. The techniques required for perforator flaps are The classic MRI findings of a fat-injection site include
more complex, with a significantly longer operating time, hyperintensity on non-fat-saturated T1-weighted images,
than those involved in the placement of a conventional hypointensity on fat-saturated images, and hypointensity
pedicled flap [16, 17]. Other flaps that may be used are on T2-weighted images [15]. Three types of abnormali-
the superior inferior epigastric perforator (SIEP) flap, the ties as seen on MRI have been described:
superior gluteal artery perforator (SGAP) flap, and the 1. Oil cyst without any enhancement after gadolinium in-
inferior gluteal artery perforator (IGAP) flap. jection (BI-RADS 2) (Fig. 3)
268 I. Thomassin-Naggara, M. Atlan, J. Chopier, E. Darai
a b e
c d f
Fig. 3 a-f. Autologous fat grafting, oil cyst. Primary breast augmentation by transverse rectus abdominis myocutaneous (TRAM) flap com-
bined with autologous fat grafting. a Axial TSE T2-weighted sequence. b Axial EG T1-weighted sequence. c Axial Native EG T1 DCE
MR weighted sequence after the 24 after injection. d Axial Substracted EG T1 DCE MR weighted sequence after the 24 after injection.
e, f Sagittal Native EG T1-weighted sequence (upper and lower right). Magnetic resonance imaging shows the presence of a high T1 sig-
nal intensity without any wall enhancement in the posterior part of the breast corresponding to the location of autologous fat grafting
2. Complex cyst of fat necrosis that displays rim en- 7. Adrada BE, Miranda RN, Rauch GM et al (2014) Breast im-
hancement (BI-RADS 3) plant-associated anaplastic large cell lymphoma: sensitivity,
specificity, and findings of imaging studies in 44 patients.
3. Granuloma appearing as a heterogeneous mass con- Breast Cancer Res Treat 147:1-14.
taining inflammatory and fibrous tissue (BI-RADS 4). 8. Tolhurst DE (1978) Nutcracker technique for compression
In the third case, the detection of a small fatty compo- rupture of capsules around breast implants. Plast Reconstr
nent inside may help in the diagnosis. Surg 61:795.
9. Baker JL, Bartels RJ, Douglas WM (1976) Closed compres-
sion technique for rupturing a contracted capsule around a
Cancer Recurrence Risk breast implant. Plast Reconstr Surg 58:137-141.
10. Ikeda DM, Borofsky HB, Herfkens RJ et al (1999) Silicone
According to the Society of Plastic Surgeons, in a report breast implant rupture: pitfalls of magnetic resonance imaging
based on a multicenter European study, the use of autolo- and relative efficacies of magnetic resonance, mammography,
gous fat grafting does not affect radiologic follow-up and and ultrasound. Plast Reconstr Surg 104:2054-2062.
11. Yoo H, Kim BH, Kim HH et al (2014) Local recurrence of
is not associated with an increased risk of breast cancer breast cancer in reconstructed breasts using TRAM flap after
[20]. Long-term follow-up of these patients is ongoing. skin-sparing mastectomy: clinical and imaging features. Eur
Radiol 24:2220-2226.
12. Helvie MA, Bailey JE, Roubidoux MA et al (2002) Mammo-
References graphic screening of TRAM flap breast reconstructions for de-
tection of nonpalpable recurrent cancer. Radiology 224:211-216.
1. Calobrace MB, Capizzi PJ (2014) The biology and evolution 13. Tokin C, Weiss A, Wang-Rodriguez J, Blair SL (2012) Onco-
of cohesive gel and shaped implants. Plast Reconstr Surg logic safety of skin-sparing and nipple-sparing mastectomy: a
134:S6-11. discussion and review of the literature. Int J Surg Oncol
2. Beretta G, Panseri S, Manzo A et al (2014) Analytical investi- 2012:921821.
gations on elastomeric shells of new Poly Implant Prothse 14. Kroll SS, Schusterman MA, Tadjalli HE et al (1997) Risk of
(PIP) breast and from sixteen cases of surgical explantation. J recurrence after treatment of early breast cancer with skin-
Pharm Biomed Anal 98:144-152. sparing mastectomy. Ann Surg Oncol 4:193-197.
3. Del Pozo JL, Tran NV, Petty PM et al (2009) Pilot study of 15. Margolis NE, Morley C, Lotfi P et al (2014) Update on imag-
association of bacteria on breast implants with capsular con- ing of the postsurgical breast. Radiogr Rev Publ Radiol Soc N
tracture. J Clin Microbiol 47:1333-1337. Am Inc 34:642-660.
4. Oulharj S, Pauchot J, Tropet Y (2014) PIP breast implant re- 16. Garvey PB, Buchel EW, Pockaj BA et al (2006) DIEP and
moval: a study of 828 cases. J Plast Reconstr Aesthetic Surg pedicled TRAM flaps: a comparison of outcomes. Plast Re-
JPRAS 67:302-307. constr Surg 117:1711-1719.
5. Spear SL, Sher SR, Al-Attar A, Pittman T (2014) Applications 17. Granzow JW, Levine JL, Chiu ES, Allen RJ (2006) Breast re-
of a cellular dermal matrix in revision breast reconstruction construction with the deep inferior epigastric perforator flap:
surgery. Plast Reconstr Surg 133:1-10. history and an update on current technique. J Plast Reconstr
6. Cordeiro PG, McCarthy CM (2006) A single surgeons 12-year Aesthetic Surg JPRAS 59:571-579.
experience with tissue expander/implant breast reconstruction: 18. Yueh JH, Slavin SA, Adesiyun T et al (2010) Patient satisfac-
part II. An analysis of long-term complications, aesthetic out- tion in postmastectomy breast reconstruction: a comparative
comes, and patient satisfaction. Plast Reconstr Surg 118:832- evaluation of DIEP, TRAM, latissimus flap, and implant tech-
839. niques. Plast Reconstr Surg 125:1585-1595.
Magnetic Resonance Imaging of Breast Implants and the Reconstructed Breast 269
19. Chun YS, Sinha I, Turko A et al (2010) Comparison of mor- 20. Petit JY, Lohsiriwat V, Clough KB et al (2011) The oncologic
bidity, functional outcome, and satisfaction following bilateral outcome and immediate surgical complications of lipofilling
TRAM versus bilateral DIEP flap breast reconstruction. Plast in breast cancer patients: a multicenter study Milan-Paris-Ly-
Reconstr Surg 126:1133-1141. on experience of 646 lipofilling procedures. Plast Reconstr
Surg 128:341-346.