Neurosurgery

Fredric B. Meyer, MD
Atlas Of

Basic Approaches to Cranial and Vascular Procedures

Preface
 The purpose of this Atlas is to provide neurosurgeons an educational tool useful when preparing to
perform common intracranial and vascular procedures. Originally, this Atlas was intended to be a
manual of practical information for my neurosurgical residents. Over the years, the conception has
developed from a pen and ink spiralbound manual into the present full-color book. The operations
chosen for review in this Atlas were based on a list created by determining the frequency of each
procedure performed on my service, excluding those required to treat traumatic lesions.
The discussions on anatomy are purposely brief to focus on information that should be immediately
helpful when performing an operation. In fact, the Atlas is organized from the perspective of a surgical
approach. Each of the operations considered deserves a special monograph of its own (and many such
detailed monographs are available); nonetheless, the simplified text of the Atlas is intended to provide a
framework for the surgeon to review ways of accessing a region and performing a particular surgical
procedure.
The operative approaches described here reflect my synthesis of the techniques and surgical skills I
have learned from my teachers and colleagues. I am indebted to my neurosurgical colleagues—past and
present—at the Mayo Clinic: the late Dr. Thoralf M. Sundt, Jr., vascular and brain tumor surgery; Dr.
Burton M. Onofrio, skull base and spine approaches; Dr. David G. Piepgras, vascular, brain tumor, and
skull base surgery; Dr. Edward R. Laws, brain tumor, epilepsy, and pituitary surgery; Dr. Patrick J.
Kelly, brain tumor and stereotactic surgery; Dr. Michael J. Ebersold, brain tumor and skull base surgery;
Dr. W. Richard Marsh, brain tumor and epilepsy surgery; Dr. Dudley H. Davis, brain tumor surgery and
general neurosurgical approaches; Dr. Robert J. Coffey, stereotactic and functional surgery; Dr. William
E. Krauss, spine and skull base surgery; Dr. John L. D. Atkinson, vascular and brain tumor surgery; Dr.
ii
Corey Raffel, pediatric neurosurgery; and Dr. Bruce E. Pollock, stereotactic and functional surgery. Dr.
Pollock, with the assistance of Dr. Atkinson, wrote the chapter on percutaneous procedures for the
treatment of facial pain.
The following general rules of conduct and practice are ones I attempt to follow and encourage
neurosurgical residents to adopt:
1. Always remember the patient who suffered a complication and learn from that
event to avoid making the same mistake in the future.
2. Be humble, especially with a great result.
3. Only recommend an operation that you would choose for yourself if you had the
same condition as the patient. Too often, the choice to proceed with surgery is
based primarily on emotion. When making a surgical decision, your intellect should
control your emotions.
4. Be realistic about your technical skills.
5. Take a minimalist approach. When treating any lesion surgically, consider the
simplest solution. Avoid grandiose, involved options.
6. Use the professional resources available to you, including neuroradiology and
neurology. Be willing to get and to listen to advice.
7. Know the anatomy of the area being operated on.
iii
 8. Do not “pick” (e.g., glioma resection margin, endarterectomy bed, ventricular
ependymal surface after resection of a trigone meningioma). A corollary of this rule
is do not finick or repeatedly reposition an aneurysm clip if the overall placement is
acceptable.
9. Know a few instruments well. Most of the cranial operations on my surgical service
are performed with interchangeable #5 or #7 suction tips, 0.7- and 1.0-mm tip
bipolar cautery, micro ball tip dissector, small black dissecting spatula, straight or
angled Gimmick dissector, Penfield #1 and #3 dissectors, Gerald forceps, straight
and curved microscissors, knot-tying forceps, Castroviejo microneedle holder, and
ring curets.
10. Use one retractor blade. It is exceedingly rare that a second retractor is ever
required. In fact, many operations can be performed without a retraction blade.
11. Always use magnification with illumination. The operating microscope is essential
for most neurosurgical procedures.
12. Be courteous and thankful to the paramedical staff who assist you.

It is a pleasure to acknowledge the great help I received from the following persons. First, Ms.
Gillian Duncan, MS, an expert medical illustrator. Each drawing was meticulously created and then
recreated after she spent many hours in the operating room and anatomy laboratory. In addition to her
iv
great artistic talent, I appreciate the patience she showed in working with me. Mr. Robert Benassi,
Emeritus Head of the Section of Visual Information, assisted with the initial stages of several drawings.
Thanks also to Mr. Steven D. Orwoll, Ms. Sue Mundy, Mr. James J. Tidwell, Mr. Thomas E. Bibby, and
Mr. Fred Graszer, in the Sect ion of Computer Graphics, for their superb computer colorization of the
drawings and to Dr. O. E. Millhouse, Mrs. Roberta J. Schwartz, Mrs. Sharon L. Wadleigh, and Mrs.
Dorothy L. Tienter. in the Section of Publications, for help in editing and preparing the manuscript for
publication. I am also in debted to Mr. Robert E. Anderson for expertly supervising our Thoralf M.
Sundt, Jr. Research Laboratory, Mrs. Wanda L. Windschitl for her outstanding patient care, and to my
long-time dedicated operating room technicians, Mrs. Helen Morrison and Ms. Eileen M. Zirbel. Special
thanks to my secretaries, Ms. Mary M. Soper, for sifting through tedious writing, dictation, and
redictation to produce readable drafts of each chapter, and Ms. Marylin J. Witts, for her remarkable
kindness in helping me take care of neurosurgical patients and for her patience with me. I am also
indebted to our neurosurgical residents whose enthusiasm and intellectual drive continue to teach and
inspire me.
Most importantly, I thank my lovely wife Irene, who is not only a devoted mother of five children
but also a clinical neurologist at the Mayo Clinic. To my children Jenna, Ilana, Benjamin, Jacob, and
Robert, you continue to teach me that in the end the only thing that truly matters is coming home.
FREDRIC B. MEYER, M.D.

v
Chapter 1

Pterional
Approach
Chapter 1: Pterional Approach

PROCEDURE

Although Dandy noted the importance of the keyhole bur hole to approach the optic chiasm and
neurosurgeons have long used the subfrontal approach, Yasargil deserves credit for developing the
pterional approach. He emphasized that combining the frontotemporal approach with resection of the
outer sphenoid wing provides access to the circle of Willis, with minimal retraction of the brain. In the
classic pterional approach, the temporalis muscle and its investing fascia are cut off the bone at the
junction between the posterior orbital rim and the zygomatic arch, and through an incision in the
temporal line, the muscle can be dissected posteriorly without injury to its nerve or blood supply,
thereby minimizing muscle atrophy. However, with this intrafascial approach, there is risk of injury to
the frontalis branch of the facial nerve from retraction or dissection. With the muscle-splitting flap
described below (and advocated by many surgeons), approximately the front half of the temporal is
muscle is dissected anteriorly in conjunction with the skin flap. In this way, there is minimal risk of
injury to the facial nerve. One disadvantage of this is that the angle of the approach is altered ever so
slightly, and this sometimes requires more retraction of the brain to achieve exposure.
Head Position

7
 For the muscle-splitting flap, the head is extended slightly; this allows the frontal lobe to fall away
from the skull base, thus decreasing the need for brain retraction. The head is also rotated approximately
25 to 30 degrees opposite to the side of the craniotomy. Occasionally, it is necessary to put a towel under
the patient’s shoulder on the side of the craniotomy. Minimal shaving of the head is required. The
incision is made 1.0 to 1.5 cm behind the hairline, so that the scar is not visible.
Skin Flap
The incision is started in front of the ear and carried to the midline, avoiding the anterior limb of the
superficial temporal artery. The incision is made through both the skin and the temporal is muscle, and
together, they are swept forward. The muscle fibers can be dissected off the cranium with a periosteal
elevator and with bipolar coagulation of arterial feeders. This reduces the degree of atrophy of the
temporal is muscle (atrophy can occur when a monopolar cutting current is used to separate the muscle
from the skull). The skin flap and temporal is muscle are retracted with 3 to 4 fishhooks placed under
tension.
Bone Flap

With current craniotomes, it is not necessary to place multiple bur holes. Still, it is helpful to mark
on the skull surface the four bur holes that would be required if Gigli saws were used, When a
craniotome is used, there is a tendency to limit the craniotomy, thereby compromising exposure, The
critical keyhole bur hole is placed just behind the orbital rim, immediately superior to the frontal
zygomatic suture. This keyhole provides access to both the anterior and middle cranial fossae. The
second bur hole is placed just above the orbital rim, between the inner canthus and the mid-pupillary

8
line, By placing this bur hole medial to the mid-pupillary line, the surgeon can use a more subfrontal
approach, if required. The lower this bur hole, the less brain retraction required. Although it is preferable
to stay out of the frontal sinus, exposure should never be sacrificed only to preserve the integrity of the
sinus. If the frontal sinus is entered and, the mucosa has not been violated, a piece of temporal is muscle
should be placed in the opening to obliterate the communication. Alternatively, if the mucosa has been
violated, it should be removed. A piece of temporalis muscle is used to plug the deep ostium. A third bur
hole is placed in the parietal bone 4 cm posterior to the second bur hole in the linea temporalis near the
coronal suture, A fourth bur hole is placed low in the squamous temporal bone behind the sphenoid
temporal suture, When making the cut between this fourth bur hole and the keyhole bur hole, it is
important that the incision swing low and anterior along the floor of the middle cranial fossa, A diamond
bur can be used instead of a craniotome to create this fourth cut. This has the advantage of ensuring that
the cut is low and anterior, to better expose the temporal lobe. A low and anterior bone cut becomes
important when a lateral approach or line of site along the sphenoid wing is desirable for approaching
difficult posterior communicating artery aneurysms.
After elevation of the bone flap , it is helpful to remove the outer one-third to one-half of the
sphenoid wing down close to the superior orbital fissure, First, the dura mater is stripped off the
sphenoid wing with a Penfield dissector or periosteal elevator. Next, a diamond bur is used to drill off
the outer sphenoid wing for a depth of approximately 3 cm. Approximately a 1.0- to 1.5-cm width of
outer sphenoid wing can be removed to improve exposure and to decrease the need for brain retraction,
Occasionally, part of the orbital roof is removed inadvertently, but this is of little concern . If the patient
has a thick frontal bone, it may be helpful to remove the inner one-half table of bone between the
sphenoid win g and orbital bur hole to increase exposure and to minimize brain retraction.
9
Neurosurgery Books
Opening of the Dura Mater

If the dura mater is going to be closed with a graft, it is best to open the dura mater and to tack it to
the bone and muscle margins. If this is done, there is no risk that the underlying cerebral cortex will be
injured inadvertently by a tack suture. However, if the dura mater is going to be closed primarily, it is
necessary to tack the dura before it is opened by placing the tack sutures in the outer layer of the dura
mater. The dura mater is opened in a way that allows it to be reflected over the sphenoid wing and
tacked to the temporalis muscle. This will decrease the amount of blood running into the wound. A small
“t” is extended over the temporal lobe and tacked to the edge of the bone to facilitate exposure of the
Sylvian fissure. The remaining dura mater is tacked to the bone margins but left intact to protect the
brain.
Division of the Sylvian Fissure

The Sylvian fissure can be divided with either a medial to lateral or a lateral to medial approach. In
most circumstances, the latter method is better. However, in young patients or in some patients with a
severe subarachnoid hemorrhage, the Sylvian fissure is obliterated or not present. In this case, the
surgeon must temporarily retract the frontal lobe to initially identify the carotid artery adjacent to the
clinoid process and then work from medial to lateral to divide the fissure. It is for this reason—that is, to
minimize the degree of brain retraction in case an initial subfrontal approach is required—that it is
important to ensure that the frontal cut made during the craniotomy is low and just above the orbital
ridge.
In the more standard dissection of the Sylvian fissure, the frontal lobe first is protected with
hemostatic fabric (Surgicel) covered by a large cottonoid (Americot). By placing the hemostatic fabric
10
Neurosurgery Books
first, it will be easier to remove the cottonoid from the brain at the end of the operation and there will be
less risk of subpial hemorrhage. Gently, the frontal lobe is retracted medially and upward, which places
tension on the arachnoid between the frontal and temporal lobes. Under the operating microscope, a
small incision is made in the arachnoid with a #11 blade knife. Any large Sylvian veins should be kept
lateral with the temporal lobe. After the initial incision is made in the arachnoid, it is enlarged with
either a straight microscissors or bipolar forceps. Occasionally, small veins that bridge the Sylvian
fissure must be cauterized and divided. There is wide variation in vascular anatomy among patients, and
in some, the middle cerebral artery is quite superficial and prone to injury. After the outer, or superficial,
Sylvian fissure has been divided, the frontal retractor is repositioned to provide new tension on the deep
arachnoid. The dissection of the arachnoid along the Sylvian fissure is extended down to the carotid
bifurcation.
At this point, it is best to incise the arachnoid on both sides of the carotid artery to achieve vascular
control, especially in cases of cerebral aneurysm. After the carotid bifurcation has been dissected free of
arachnoid, the lesion dictates the approach to dissecting the arachnoid off the anterior cerebral artery.
Specifically, if the operation is for a craniopharyngioma, the arachnoid between the anterior cerebral
artery and the optic chiasm should be incised. This will allow the anterior cerebral artery to be retracted,
with the frontal lobe still protected by arachnoid. Alternatively, if the lesion is an anterior
communicating artery aneurysm, the arachnoid on both sides of the proximal anterior cerebral artery
needs to be incised for control of proximal blood vessels.

11
Neurosurgery Books
NEUROSURGICAL ANATOMY

Because the optic nerve is the anatomic landmark and serves as a reference to all other structures in
this region, it should be identified early in the exposure. There is significant variation among patients
regarding the diameter and length of the internal carotid artery, the location of the carotid bifurcation,
the size and extent of the anterior clinoid process, and whether the optic chiasm is prefixed or postfixed.
The anatomic publications of Yasargil and Rhoton and their colleagues are required reading.
Ophthalmic Artery

The ophthalmic artery originates medially from the internal carotid artery as it emerges from the
cavernous sinus underneath the anterior clinoid process. The ophthalmic artery arises from the subdural
portion of the internal carotid artery in 90 percent of patients and at the carotid-dural ring in 2 percent. In
the other 8 percent of patients, its origin is extradural, from the cavernous carotid artery. As the
ophthalmic artery leaves the internal carotid artery, it runs along the inferior surface of the optic nerve,
delicately attached by loose connective tissue. The artery enters the optic canal inferiorly to the optic
nerve by piercing the dural sheath of the optic nerve. As the artery runs through the optic canal, it is
inferior to the optic nerve. The artery penetrates the orbit and curves medially, either above or below the
optic nerve.
Superior Hypophysial Artery

Several small arteries, named the superior hypophysial arteries, usually exit from the inferior medial
portion of the internal carotid artery underneath the optic nerve and supply the pituitary stalk, anterior

12
Neurosurgery Books
lobe of the pituitary, and the inferior surface of the optic nerve and chiasm. These small arteries
anastomose with their counterparts from the opposite internal carotid artery and with the inferior
hypophysial arteries form a vascular plexus around the pituitary stalk.
Posterior Communicating Artery

The next major branch of the internal carotid artery is the posterior communicating artery. It
originates from the inferior lateral wall of the supraclinoid internal carotid artery, 2 to 7 mm distal to the
anterior clinoid process, and exits from the carotid cistern by penetrating the arachnoid posteriorly and
inferiorly to enter the interpeduncular cistern. This portion of the posterior communicating artery is
covered by arachnoid. As the artery runs posteriorly, it is adjacent to the posterior clinoid process, to
which it is occasionally adherent. In approximately 60 percent of patients, the posterior communicating
artery is 2 mm or less in diameter. Its diameter is larger in children than in adults. In approximately 10
percent of patients, the posterior communicating artery is hypoplastic or absent. Rarely, it can have a
duplication or fenestration. After it leaves the internal carotid artery, the posterior communicating artery
runs medially and, thus, is hidden from view if the surgeon uses a subfrontal approach to this region. If a
more lateral or temporal approach along the axis of the sphenoid wing is used, the artery can be
followed further into the interpeduncular cistern. Approximately 3 mm from its origin, the posterior
communicating artery gives rise to 2 to 10 branches that run posteriorly, inferiorly, and medially into the
interpeduncular cistern to supply the optic chiasm and tract, tuber cinereum, mammillary bodies,
hypothalamus, and inferior thalamus. These branches are long and variable in caliber.
Anterior Choroidal Artery

13
Neurosurgery Books
 The anterior choroidal artery arises 2 to 5 mm distal to the origin of the posterior communicating
artery and runs laterally to this artery, following the optic tract posteriorly. The anterior choroidal artery
varies greatly in diameter (0.5 to 1.5 mm) and may arise as more than one vessel (it is duplicated in 30
percent of patient s). If it originates as a single trunk that divides into two vessels, one of them, the
“uncal artery,” ramifies almost immediately to supply the uncus, part of the amygdala, and the anterior
hippocampus. The main trunk of the anterior choroidal artery continues posteriorly inferior to the optic
tract to the choroidal fissure, where its supplies the choroid plexus of the temporal horn and anastomoses
distally with branches of the posterior lateral choroidal artery. In its course through the carotid cistern,
the anterior choroidal artery gives off branches that supply the inferior surface of the optic chiasm, the
posterior two-thirds of the optic tract, the medial globus pallidus, the genu of the internal capsule, the
middle third of the cerebral peduncle, part of the red nucleus, the subthalamus, and thalamic nuclei. As
the anterior choroidal artery enters the choroidal fissure to supply the choroid plexus of the temporal
horn, it gives off branches that supply the anterior lateral half of the lateral geniculate body, the anterior
half of the posterior internal capsule, the retrolenticular part of the internal capsule, and the optic
radiations. Considering the territory that is irrigated by the anterior choroidal artery, it is clear that loss
of this vessel has devastating neurologic consequences for the patient.
Middle Cerebral Artery

The middle cerebral artery is 2.5 to 4.6 mm in diameter at its origin. The part of this artery that
extends from its origin to its bifurcation—approximately 15 mm in length—is referred to as the “M1”
“sphenoidal,” or “pterional segment.” The M1segment gives off two sets of branches: the superior
lateral (or temporal) and the inferior medial (or deep) perforating branches. Typically, 1 to 3 branches
form the superior lateral group and are labeled the “uncal,” “polar temporal,” and “anterior temporal
arteries.” Occasionally, the uncal branch originates from the posterior internal carotid artery just distal to14
Neurosurgery Books
the anterior choroidal artery. Often, the polar temporal artery is small or absent. In this case, the
remaining anterior temporal artery is large. Occasionally, the polar temporal and anterior temporal
arteries are both absent and are replaced by a large branch from the anterior trunk of the M2 segment of
the middle cerebral artery.
The inferior medial branches of the M1 segment are the lenticulostriate arteries. There usually are 3
to 5 of these arteries, although up to 15 perforating vessels have been reported. They enter the lateral
two- thirds of the anterior perforated substance and supply the anterior commissure, the putamen, the
lateral globus pallidus, the superior half of the internal capsule, and the head and body of the caudate
nucleus. Sometimes, there is only one large lenticulostriate artery that divides into many smaller
branches that pe rfuse the above territory. This larger lenticulostriate artery usually originates just
proximal to the bifurcation of the middle cerebral artery and runs medially. This vessel can be injured if
the clip is placed too deep during repair of a middle cerebral artery aneurysm.
The part of the middle cerebral artery that is distal to its bifurcation is called the “M2 segment.” It
usually consists of a superior branch and inferior branch, but in approximately 20 percent of patients,
there are three branches-a trifurcation instead of a bifurcation. The proximal branches of the M2
segment include the lateral orbitofrontal, prefrontal, frontal auricular, precentral, central sulcus, angular,
and posterior temporal arteries. The branches originating from the superior trunk typically supply
inferior frontal cortex, frontal auricular cortex, and parietal and central sulcus regions. Branches from
the inferior trunk usually supply the temporal, temporooccipital, angular, and posterior parietal gyri.
Rarely, an accessory middle cerebral artery has been reported to originate from the anterior cerebral
artery complex.
15
Neurosurgery Books
Anterior Cerebral Artery

The size and configuration of the anterior cerebral artery complex and its branches vary
significantly. The A1 segment of the anterior cerebral artery extends from the bifurcation of the internal
carotid artery to the anterior communicating artery. Typically, one AI segment is dominant, with the
opposite A1 segment being either quite small in caliber or absent. The average diameter of the A1
segment is 2.5 mm. Perforating arteries arise from the
inferior posterior portion of the proximal anterior
cerebral artery and supply the optic chiasm and
penetrate the anterior perforated substance to supply
part of the fornix, the anterior limb of the internal
capsule, the anterior inferior part of the striatum, and

Figure 1-1.
The optic nerve and the supraclinoid part of the internal
carotid artery serve as the landmarks or refere nce points for
all the structures exposed during a standard pterional
craniotomy. The approach to these structures is along the
long axis of the sphenoid wing. The two routes of access are
the trans-Sylvian and the subfrontal. Which of these two
routes is used is determined primarily by whether the patient
has a Sylvian fissure that can be divided easily. In young
patients or those who are obese or have a subarachnoid
hemorrhage, the Sylvian fissure is often obscured. In this
case, a subfrontal approach can be used, and the medial
aspect of Sylvian fissure can be identified and incised from
16
medial to lateral.
Neurosurgery Books
the anterior hypothalamus. These branches of the A1 segment may originate from one larger vessel, the
medial proximal striate artery, which may have a reciprocal relationship with the lenticulostriate arteries
that originate from the M1 segment of the middle cerebral artery.
The anterior communicating artery is 0.1 to 3.0 mm long. Typically, a large A1 branch divides into
two distal A2 segments, and the anterior communicating artery is the apparent attachment or entry point
of the hypoplastic contralateral A1 segment. In 75 percent of patients, the anterior communicating artery
is a single vessel, and in the other 25 percent, it has various anomalies, including fenestrations or
duplications. Perforating arteries, which vary in number and diameter, arise from the posterior inferior
aspect of the anterior communicating artery and the distal A1 segment, at its junction with the anterior
communicating artery. They supply the infundibulum, opticc chiasm, and preoptic area of the
hypothalamus. Yasargil reported that if the A1 segments are equal in size, the perforating vessels arise
from the midportion of the anterior communicating artery. However, if the A1 segments are unequal, the
perforating branches arise from the anterior communicating artery on the side of the larger A1 segment.
Because this relationship is similar to that observed for aneurysms in this location, these vessels
typically are found on the anterior inferior side of the neck of the aneurysm.
Recurrent Artery of Heubner

A recurrent artery of Heubner is almost always present and usually originates from the A1 or A2
segment, adjacent to the origin of the anterior communicating artery. It originates from the A2 segment
in approximately 80 percent of patients and is bilateral in 95 percent. The reported diameters for the
recurrent artery of Heubner range from 0.2 to 2.9 mm. It runs parallel to the anterior cerebral artery
before entering the anterior perforated substance to supply the anterior part of the caudate nucleus, the
17
Neurosurgery Books
anterior putamen, part of the globus pallidus, and the anterior limb of the internal capsule. During
exposure of the anterior comlllunicating artery complex, any vessel that runs laterally along the inferior
frontal lobe must be presumed to be the recurrent artery of Heubner and, therefore, protected. This is
particularly true if partial resection of the gyrus rectus is necessary. Of note is that the recurrent artery of
Heubner occasionally originates from the frontal polar branch of the anterior cerebral artery.
Figure 1-2.
Step 1. A, The patient’s head is fixed in a pinion,
extended slightly, and rotated approximately 25 to 30
degrees opposite the side of the craniotomy. With
extension of the head, the frontal lobe falls away
from the floor of the frontal cranial fossa, thereby
lessening the degree of brain retraction.

Step 1. B, The keyhole bur hole allows access to both

the frontal and middle cranial fossae. It is important to
extend the bone Hap just medial to the mid-pupillary
line to increase exposure in case a subfrontal approach
is required. 18
Neurosurgery Books
 Figure 1-3.
Step 2. Several techniques can be used to cui the bone flap, including
standard bur holes and Gigli saws or newer craniotomes. It is helpful to
mark the extent of the bone flap when using a craniotome to prevent
minimizing the size of the craniotomy. It is important to extend the bone
flap down toward the base of the middle cranial fossa. Sometimes, the
optimal method is to perform two-thirds of the craniotomy with a
craniotome and then use a high-speed air drill with a diamond bur to
make the last one-third of the cut along the floor of the middle cranial
fossa over the outer sphenoid wing up to the keyhole bur hole. With this
method, the bone flap is not minimized.

Figure 1-4.
Step 3. A, After removing the bone flap. the outer sphenoid wing is
removed down to the superior orbital fissure. This increases the
exposure by approximately 1.5 cm. The dura mater along the sphenoid
wing is sharply dissected free with a periosteal elevator.
B, The outer sphenoid wing is removed with a small orbital
rongeur. A
C, A Yasargil retractor is used to displace the dura mater, and,
under the microscope, a high-speed air drill with a diamond bur is used
to remove the deeper portion of the sphenoid wing. In fact, removal of
the sphenoid wing can be extended deep to include the clinoid process,
which allows the anterior clinoid process to be removed without
incising the dura mater. This is described later in the chapter. In some
patients, the frontal bone is thick and the diamond bur can be used to
remove the inner one-half of the frontal bone, which also increases
exposure, especially if a subfrontal approach to the internal carotid B
artery is being used.
Neurosurgery Books
C
 Figure 1-5.
Step 4. The dura mater is tacked to the bone margins. If a dural graft is
going to be used, the dura mater can be opened and then tacked. In this way,
there is no risk of inadvertently injuring the underlying cerebral cortex.
However, if the dura mater is going to be closed primarily, it is best to tack it
to the bone before opening it. In patients with a large frontal sinus, the frontal
tack stitch should be sewn to the muscle to avoid a tack hole going into the
frontal sinus, which can lead to rhinorrhea postoperatively.

Figure 1-6.
Step 5. The dura mater is opened and reflected over the sphenoid wing
and temporalis muscle to prevent blood from running into the operative
field. The frontal lobe is lined with hemostatic fabric and cottonoids and
then gently retracted to place the Sylvian fissure under tension. Correct
orientation of the Yasargil bar, head, and arm will increase the effectiveness
of this retractor. As described by Sundt, the Yasargil arm should hang
straight down and touch the drapes approximately 5 cm from the blade,
forming a gentle “S” shape. With this arrangement, gravity assists the
retraction and the distal arm has a fulcrum. To achieve this straight up-and-
down orientation, the Yasargil head off the bar needs to be in the correct
position. The surgeon can determine this by attaching the arm to the
unsecured Yasargil head on the bar. The arm is held in the correct
orientation, and the head is rotated around the bar until there is no torque.
Next, the head is secured to the bar. A common mistake is not to insert the
Yasargil bar far enough into the clamp along the side of the operating room
table to prevent excessive outward looping of the arm
Neurosurgery Books
Figure 1-7.
Step 6. With a #5 or #7 straight suction tip and small cottonoid, the
frontal lobe adjacent to the lateral Sylvian fissure is retracted gently to
place the arachnoid under tension. In the process of dividing the Sylvian
fissure, both the suction tip and coltonoid are “walked down” the
fissure, placing under tension the next part of the arachnoid to be cut.

Figure 1-8.
This incision in the arachnoid is extended downward or medially
along the Sylvian fissure by separating the arachnoid with
bipolar forceps. Alternatively, the arachnoid can be lifted and cut
with a straight microscissors. Large Sylvian veins should be left
intact and displaced laterally with the temporal lobe. Small
bridging veins across the Sylvian fissure can be safely cauterized
and divided.
Neurosurgery Books
 Figure 1-9.
As the Sylvian fissure becomes
more broadly split, the frontal lobe is
increasingly retracted upward to
place the deep arachnoid under
tension. If a lumbar drain has been
placed, it is best not to remove
cerebrospinal fluid until the Sylvian
fissure has been divided. Removal of
cerebrospinal fluid through the drain
will decompress the Sylvian fissure
and, thus, make it more difficult to
dissect the arachnoid. The arachnoid
incision is carried down on top of the
internal carotid artery.
Neurosurgery Books
 Figure 1-10.
With a straight or angled dissector, the
arachnoid is teased off the internal carotid
artery both medially and laterally,
especially in cases of aneurysm.
Accordingly, immediate proximal control
has been obtained. Next, the arachnoid is
incised medially between the anterior
cerebral artery and the optic chiasm. This
incision is carried over to the opposite
optic nerve. By cutting the arachnoid
between the anterior cerebral artery and
the optic chiasm, the A1 segment and its
perforating vessels are lifted with the
frontal lobe and, therefore, moved out of
harm’s way. If an anterior cerebral artery–
anterior communicating artery aneurysm is
being treated, both sides of the proximal
A1 segment need to be dissected to provide
proximal arterial control. After the Sylvian
fissure is divided, the temporal lobe can be
retracted laterally if necessary. There often
are bridging veins off the tip of the
temporal lobe that are under tension.
These can be supported and protected with
a piece of absorbable gelatin sponge
(Gelfoam). If required, these bridging veins
can safely be cauterized and divided.
Neurosurgery Books
POSTERIOR COMMUNICATING ARTERY ANEURYSM

Aneurysms in this location are best exposed by first identifying the medial surface of the internal
carotid artery adjacent to the optic nerve. This is called the “optic-carotid triangle.” In most instances,
these aneurysms project laterally toward the tentorium cerebelli and the oculomotor nerve. Not
unexpectedly, they often are associated with hemorrhage into the temporal lobe. Occasionally, these
aneurysms are located proximally, necessitating resection of the anterior clinoid process.
After the medial surface of the internal carotid artery is identified, a dissector is used to incise the
arachnoid over the lateral aspect of the artery. This allows proximal arterial control to be achieved and, if
necessary, a temporary clip to be placed. Typically, the aneurysms arise from the axilla of the posterior
communicating artery, at its origin from the internal carotid artery. The base of the aneurysm usually
projects at a 45-degree angle away from the origin of the posterior communicating artery. It is important
to dissect the origin of the posterior communicating artery from the anterior wall of the aneurysm so that
the artery can be preserved when a clip is placed across the base. After the posterior communicating
artery is identified and the arachnoid that tethers the aneurysm to the artery is incised, a piece of
absorbable gelatin sponge (Gelfoam) can be placed to displace this artery. A dissector is then used to
identify the origins of the anterior choroidal artery. A piece of absorbable gelatin sponge is also placed
between this artery and the neck of the aneurysm. Placement of small bits of absorbable gelatin sponge
keep the important arteries out of harm’s way when an aneurysm clip is placed. In rare instances, the
posterior communicating artery originates from the neck of the aneurysm in such a way that the
posterior communicating artery has to be occluded at its origin from the internal carotid artery before the
aneurysm can be obliterated. Before concluding that this is indeed the case, it is mandatory that the

24
Neurosurgery Books
surgeon ensure that there is no plane between the artery and neck of the aneurysm and that the posterior
communicating artery has a small diameter. Furthermore, the preoperative angiogram should show
retrograde filling of the posterior communicating artery on the vertebral study. In this rare case, it is
acceptable to place the aneurysm clip in a way that both the neck of the aneurysm and the origin of the
posterior communicating artery are occluded. This will be tolerated by the patient because the posterior
communicating artery fills retrogradely.
For all aneurysms, it is necessary to ensure that after the clip has been placed, the aneurysm has
been obliterated completely and no vessels are trapped by the jaws of the clip. It is useful to aspirate the
aneurysm, especially large lesions, to collapse the structure. After this is·done, the dome of the aneurysm
can be manipulated with a #5 or #7 straight suction tip, and the underside of the neck of the aneurysm
and the clip can be inspected. After a posterior communicating artery aneurysm is clipped, the anterior
choroidal artery is viewed from a lateral perspective to make sure that it has not been compromised. The
same is true for the posterior communicating artery. Depending on the location of the aneurysm clip, it
may be difficult to visualize the posterior communicating artery. If this is so, the artery and its branches
can be viewed from a medial perspective between the optic nerve and the internal carotid artery by
gently displacing the medial wall of the internal carotid artery laterally. After these vessels have been
inspected, papaverine can be applied topically with a cottonoid to relieve mechanically induced
vasospasms.

25
Neurosurgery Books
Figure 1-11.
Typical operative exposure used for
aneurysms of the posterior communicating
artery, anterior choroidal artery, or
bifurcation of the internal carotid artery.
Depending on the location and projection of
the aneurysm, a retractor placed on the
temporal lobe can be useful.

Figure 1-12.
Step 1. Under the operating microscope, it is
mandatory to dissect the neck of the aneurysm of the
posterior communicating artery off both the anterior
choroidal artery and the posterior communicating artery.

Neurosurgery Books
Figure 1-13.
Step 2. A useful step in all aneurysm surgery is to
displace dissected vessels from the neck of the aneurysm by
using small pieces of absorbable gelatin sponge (Gelfoam).
Neurosurgery Books
 Figure 1-14.
Step 3. A, Usually a straight, bayonet,
A or 45-degree-angled aneurysm clip is best
for aneurysms of the posterior
communicating artery or anterior
choroidal artery. When placing the clip, it
is important to ensure that neither the
perforating vessels nor the internal
carotid artery is compromised or
constricted.
B, After the aneurysm is clipped, its
dome is aspirated if the surgeon believes
that the neck has been occluded
completely with the clip. Aspirating the
dome of the aneurysm enhances the
B exposure, thus allowing better inspection
of the underlying perforating vessels.
Neurosurgery Books
 Figure 1-15.
Step 4. In all aneurysm
repairs, it is mandatory to ensure
that both the parent vessel and
adjacent perforators are
structurally intact. The takeoff of
the anterior choroidal artery can
be visualized best from a lateral
view. Often, this is true also for
the posterior communicating
artery. Occasionally, the clip
prevents the posterior
communicating artery from being
visualized. In that case, it can be
viewed from a medial to lateral
perspective, looking between the
optic nerve and internal carotid
artery through the optic-carotid
triangle.
Neurosurgery Books
ANTERIOR COMMUNICATING ARTERY ANEURYSM

It is easiest to expose an aneurysm of the anterior communicating artery from a more subfrontal than
sphenoidal approach through a pterional craniotomy. Therefore, the head should not be rotated more
than 30 degrees opposite the side of the craniotomy. It usually is best to approach the aneurysm from the
side of the dominant feeding artery, unless there has been severe hemorrhage into the opposite gyrus
rectus. Because it occasionally is necessary to remove part of the gyrus rectus to expose the aneurysm, it
is preferable to remove tissue that has already been damaged by a hemorrhage. After the Sylvian
fissure is divided, the carotid bifurcation is identified. A hemostatic fabric and cottonoids are placed over
the frontal lobe, which is then retracted upward and posteriorly. This increases tension on the arachnoid
over the anterior cerebral artery. The arachnoid between the anterior cerebral artery and the chiasmatic
cistern is opened and the proximal A1 segment is identified in case a temporary clip has to be placed
should the aneurysm rupture. The arachnoid is dissected free between the anterior cerebral artery and the
optic chiasm. It is not necessary to separate the arachnoid completely between the frontal lobe and the
anterior cerebral artery, because this might increase the risk of injury to the perforating arteries that
originate from the A1 segment. To prevent inadvertent rupture during exposure, the surgeon must bear in
mind the direction of the dome of the aneurysm.
The incision in the arachnoid is extended medially over the optic chiasm so that the opposite A1
segment can be visualized. With a #11 blade knife, the cistern of the lamina terminalis in the
interhemispheric fissure is incised between the two gyri recti. If the lamina terminalis can be identified
clearly, it may be useful to incise it to drain cerebrospinal fluid from the third ventricle to increase brain
relaxation. If a lumbar spinal drain is being used, this maneuver is not required. If necessary, a small

30
Neurosurgery Books
portion of the gyrus rectus can be removed with a bipolar cautery and suction. It is important not to
coagulate or to divide any perforating vessels in this area. Furthermore, the frontal polar artery must not
be mistaken for the recurrent artery of Huebner. After the gyrus rectus is removed. the ipsilateral A2
segment is identified. Therefore, at this point in the operation, three of the four major trunks have been
visualized except for the opposite A2 segment. With a typical posteriorly pointing aneurysm, it is best to
start dissecting the aneurysm at the junction between the ipsilateral A2 segment and the anterior
communicating artery. This dissection usually starts posteriorly behind the aneurysm, with the aneurysm
being elevated with a #5 or #7 straight suction tip and small cottonoid. Typically, perforating vessels are
ensheathed in their own arachnoid and can be dissected off the back wall of the aneurysm as a single
sheet or layer. As the aneurysm is gently rotated anteriorly, the opposite A2 segment usually can be
identified and dissected free. Thereafter, a small piece of absorbable gelatin sponge is placed between
the opposite A2 segment, arachnoid and perforating vessels, and the ipsilateral A2 segment and the neck
of the aneurysm. After clipping, it is important to decompress the aneurysm to allow manipulation of the
dome to verify that the anterior communicating artery complex and its perforating vessels are intact and
not entrapped by the jaws of the clip.
In large aneurysms, it may be necessary to perform a thrombectomy in order to place a clip and to
preserve the surrounding vasculature. In this case, the dominant feeding A1 segment is temporarily
occluded. If perforating vessels originate from the A1 segment, it is preferable to place the temporary
clip distal to their takeoff, without compromising exposure of the neck of the aneurysm. An incision is
made in the dome far distal to the neck. If an incision is made too close to the neck, then subsequent
placement of an aneurysm clip may be compromised. Thrombectomy is performed with curets. If back
bleeding is significant, it may be necessary to occlude the contralateral A1 segment.
31
Neurosurgery Books
 Figure 1-16.
Step 1. The approach to an anterior
communicating artery aneurysm is more
subfrontal than sphenoidal. After the
Sylvian fissure is divided, the frontal lobe
retractor is repositioned to lift the frontal
lobe up and posteriorly to expose the
carotid bifurcation. The arachnoid over
the A1 segment is incised to allow access
in case a temporary clip has to be placed
because of intraoperative rupture of the
aneurysm. The arachnoid between the
anterior cerebral artery and the optic
chiasm is incised to identify the opposite
A1 segment.

Neurosurgery Books
 Figure 1-17.
Step 2. Despite the Sylvian
fissure being divided and despite
cerebrospinal fluid being
aspirated from the
suprachiasmatic cistern and
drained through a lumbar drain
or an incision in the lamina
terminalis, exposure of the
aneurysm can still be difficult. If
exposure is a problem, part of
the ipsilateral gyrus rectus may
be removed with bipolar cautery
and delicate suction. Any artery
that runs lateral and horizontal
to the A1 segment should be
protected and presumed to be
the recurrent artery of Heubner.
Neurosurgery Books
 Figure 1-18.
Step 3. The frontal
retractor is placed deeper to
allow additional retraction of
the frontal lobe and gyrus
rectus. It is important to be sure
that the interhemispheric
arachnoid has been incised to
decrease tension on the
opposite frontal lobe. Before
the aneurysm is dissected, it is
mandatory to ensure good
access to the dominant A1
segment in case temporary
clipping is necessary.
Neurosurgery Books
 Figure 1-19.
Step 4. A, The neck of the aneurysm is
carefully dissected free from the underlying
perforating vessels and the adjacent A1 and A2
segments. It is best first to isolate the dominant
A1 segment and then the opposite A1 for good
proximal control. A spatula or dissector is used to
dissect free the ipsilateral A2 segment. The
contralateral A2 segment is the last structure
identified. The perforating vessels originating
from the anterior communicating artery usually
are sheathed in arachnoid and can be separated
in a plane en bloc. This is similar to dissecting
perforating vessels off the basilar caput. The
perforating vessels adjacent to the ipsilateral A2
segment are identified first. The plane between
these perforators and the neck of the aneurysm is
developed and extended around the backside of
the lesion.
B, A piece of absorbable gelatin sponge is
used to displace these vessels before the
aneurysm is clipped.
C, After placement of the clip, the dome is
aspirated and manipulated to allow inspection of
the clip and its relationship to all adjacent
vascular structures.

Neurosurgery Books
 Figure 1-20.
Illustrated here is repair of a giant partially
thrombosed aneurysm of the anterior
communicating artery complex. With these lesions,
thrombectomy must be performed. A temporary
clip is placed on the dominant A1 segment distal to
any perforating vessels. The dome of the aneurysm
is incised, and ring curets are used to remove the
thrombus. It is important to make sure that the
incision in the dome of the aneurysm is as far from
the neck as possible. If there is significant back
bleeding, it may be necessary to occlude
temporarily the opposite anterior cerebral artery
adjacent to the opposite optic nerve.
Neurosurgery Books
SUPRACLINOID INTERNAL CAROTID ARTERY ANEURYSM

As emphasized by Sundt, giant aneurysms of the supraclinoid internal carotid artery can be divided
by the direction of their projection, including superior, posterior, medial, and lateral. In nearly all
instances, it is necessary to resect the anterior clinoid process. This resection can be performed through
either an intradural or an extradural approach. If the intradural approach is chosen, the dura mater
overlying the anterior clinoid process is cauterized for a distance of 5 to to mm. The bone is then
removed with a high-speed diamond air bit. When using the highspeed air drill, it is critical to remove
all cottonoids from the wound. because the suction vacuum created by the drill can quickly wrap these
cottonoids around the drill and create an “egg beater” in the wound. The extradural approach to
removing the anterior clinoid process is described later in this chapter.
Posteriorly projecting aneurysms, as illustrated in this case, are typically the most treacherous
because it often is difficult to reconstruct the parent internal carotid artery. Therefore, the surgeon must
be prepared to perform a bypass graft either from the cervical carotid or petro us carotid artery, using a
saphenous vein harvested from the leg. In this regard, it is useful to know preoperatively the potential
for collateral blood flow by performing a trial balloon occlusion during the initial angiography, perhaps
in combination with a xenon blood flow or a single photon emission computed tomographic (SPECT)
blood flow study. Alternatively, the surgeon can expose the cervical carotid artery and perform a
temporary occlusion in conjunction with intraoperative electroencephalographic or cerebral blood flow
monitoring. If on inspection of the aneurysm it appears that it is not technically feasible to reconstruct
the internal carotid artery and the patient does not tolerate the temporary occlusion, a reasonable option
is to perform a saphenous vein bypass graft.

37
Neurosurgery Books
 After the Sylvian fissure is divided, the internal carotid artery and optic nerve are first identified.
The relationship between the internal carotid artery between the anterior clinoid process and the carotid
bifurcation is inspected to determine whether primary clipping will be feasible technically. If the anterior
clinoid process has not been removed through an extradural approach, it is removed through an
intradural approach. This allows identification of the proximal neck of the aneurysm, which typically
originates immediately under the anterior clinoid process.
After the proximal neck of the aneurysm is identified, attention is turned to the distal neck to
establish the relationship between it and the anterior choroidal artery. The posterior communicating
artery is often absent in posteriorly projecting aneurysms of the internal carotid artery. However, the
anterior choroidal artery is often stretched just underneath the carotid bifurcation and must be dissected
free. After the anterior choroidal artery is identified, a piece of absorbable gelatin sponge is placed
between it and the neck of the aneurysm.
For most giant aneurysms in this location, temporary vessel occlusion and thrombectomy are
necessary to facilitate clip placement. The internal carotid artery is occluded either in the neck just distal
to the carotid bifurcation or intracranially as the carotid artery exits the cavernous sinus. In that regard, it
is best to incise the distal carotid dural ring, which allows a temporary clip to be placed more
proximally, thus increasing the amount of space along the proximal neck of the aneurysm. The second
temporary clip is placed on the distal internal carotid artery if there is sufficient room. Otherwise, it may
be necessary to occlude the proximal A1 segment if back bleeding is sufficient after the aneurysm has
been opened and decompressed. The patient’s blood pressure is increased to a systolic pressure of 150 to
160 mm Hg to increase collateral blood flow. Intraoperative electroencephalography can be helpful in
38
Neurosurgery Books
determining whether collateral cerebral blood flow is sufficient. If, during preoperative planning, the
surgeon anticipates that occlusion of the internal carotid artery will be longer than 20 minutes because of
the complexity of the aneurysm and trial balloon occlusion was not tolerated, serious consideration
should be given to performing the operation under profound hypothermia or constructing a preocclusion
bypass graft to the distal middle cerebral artery complex.
After the aneurysm is temporarily trapped, an incision is made in the dome as distal to the parent
artery as possible. Thrombectomy is performed with ring curets. After the aneurysm is collapsed, a clip
is placed across its neck in one of two directions. Placement of an angled clip underneath the internal
carotid artery from lateral to medial is technically easier than using a cutout clip placed parallel to the
internal carotid artery. Although it often is written that a cutout clip is less likely to stenose or to occlude
the internal carotid artery, it is difficult to place the clip in a position in which the neck of the aneurysm
is obliterated but the origin of the anterior choroidal artery is not compromised. For this reason, it is best
to start out using an angled clip placed lateral to medial to determine whether the aneurysm can be
obliterated without stenosing the carotid artery. After the clip is placed, blood flow through the carotid
artery is confirmed with intraoperative Doppler, ultrasonography, or angiography. The anterior choroidal
artery is inspected visually to make sure that it is patent.

39
Neurosurgery Books
Figure 1-21.
Step 1. It is necessary to remove the anterior clinoid process and to expose the proximal internal
carotid artery to obtain proximal control of the vessel and to identify the neck of the aneurysm. Some of
these aneurysms can be transitional in that the neck originates from the clinoid segment and the dome of
the aneurysm extends intradurally. Typically, the posterior communicating artery is not present. However,
invariably, the anterior choroidal artery is intimately associated with the distal neck of the aneurysm and
must be dissected free. It is necessary to separate widely the Sylvian fissure to identify the carotid
bifurcation. It may be usef l to place a retractor on the temporal lobe. However, depending on the angle of
the clip to be placed, the temporal lobe retractor may be in the way and may need to be removed. With
giant aneurysms in this location, exposure of the cervical carotid artery provides a safety net in case the
aneurysm ruptures before proximal control of the carotid artery
Neurosurgery is achieved.
Books
Figure 1-22.
Step 2. There are two ways to remove the anterior
clinoid process. As illustrated here, it can be removed
through an intradural approach. A small spatula is
used to cauterize and to peel the dura mater off the
anterior clinoid process. A small diamond bur is used
to gently drill the clinoid process and expose the
supraclinoid carotid artery. When the anterior clinoid
process is thin, a small bone rongeur is used to bite off
the rest of the bone. During removal of the anterior
clinoid process, it is important to remove the bar of
bone between the anterior clinoid process and the
superior orbital fissure. Also, it is best to remove the
optic strut bone just lateral to the optic nerve. This
removal of bone medial and lateral to the anterior
clinoid process creates more room, thereby increasing
the options for placing and manipulating the clip. The
other way to remove the anterior clinoid process is
through an extradural approach, as pioneered by Dol-
enc. The advantage of the extradural approach is that
there is dura mater between the bony removal and the
aneurysm. Also, the surgeon can be more aggressive
about the degree of bony removal of the medial
sphenoid wing, which increases exposure of the intracavernous carotid artery. With the extradural approach, the
outer sphenoid wing is removed with a small orbital rongeur. Under the operating microscope, the medial sphenoid
wing is drilled off with a diamond bur, first identifying the superior orbital fissure. Next, the diamond bur is used to
thin the bony bridge between the superior orbital fissure and the anterior clinoid process. The drill is also used to
thin and to remove the orbital strut of bone between the optic canal and the anterior clinoid process. After these
medial and lateral bony bridges have been thinned, small bone cups or forceps are used to bite bone and to wiggle
the anterior clinoid process free. These same bony cups can be used to remove additional thin bone over the optic
nerve medially and the superior orbital fissure laterally.
Neurosurgery Books
Figure 1-23.
Step 3. A, After the anterior clinoid process
is removed, the internal carotid artery is
mobilized by incising the distal dural ring. This
distal dural ring is continuous with the dura
mater that surrounds the optic nerve medially
and the fal iform ligament laterally.
B, A ball-tip dissector can be used to dissect
the plane between the internal carotid artery
and the dural ring.
A
C, When this plane is established, a
microscissors such as a micro Potts can be used
to incise the dural ring. It is best to stay on top
of the internal carotid artery during this
incision. After the outer dural ring is cut, a small
spatula can be used to sweep the dural B
reflections medially and laterally. In most
transitional carotid aneurysms, the origin of the
ophthalmic artery defines the medial extent of
the neck of the aneurysm. Often, the ophthalmic
artery originates from the extradural carotid
artery. Therefore, it is necessary to carry the
dissection proximal to the anterior genu of the
internal carotid artery. Often, the very lateral
and medial attachments of the carotid dural C
rings are adherent to the neck of the aneurysm
and must be sharply dissected free with a
spatula or, occasionally, a #11 blade knife.
Neurosurgery Books
 Figure 1-24.
Step 4. For most giant aneurysms, it is
necessary to temporarily occlude some of
the major arteries and to perform
thrombectomy. In this example, removal of
the anterior clinoid process is not sufficient
to expose the proximal internal carotid
artery to temporarily place a clip.
Therefore, the cervical internal carotid
artery is occluded. A temporary clip is also
placed on the A1 segment. Occasionally, it
is possible to place a temporary clip
proximal to the anterior choroidal artery, in
which case there is continued perfusion of
the anterior choroidal artery and the
middle cerebral artery complex through the
anterior communicating artery. More
typically, as illustrated here, there is not
enough room to place temporary clips
proximal to the bifurcation. When incising
the dome of the aneurysm, it is important to
make the incision as distal to the neck as
possible so that placement of the permanent
clip will not be compromised.

Neurosurgery Books
 Figure 1-25.
Step 5. There are two ways to place the aneurysm clip—the one
illustrated here and the one shown in Figure 1-26. Placing the clip
perpendicular to the internal carotid artery is technically easier in terms
of preserving the anterior choroidal artery. However, it may also cause
kinking of the carotid artery and lead to stenosis or occlusion.

Figure 1-26.
An alternative clip placement is a cutout that runs parallel with the
internal carotid artery. There is less risk of stenosis of the artery if the
cutout has sufficient caliber. Regardless of the way the clip is placed, the
more aggressive the thrombectomy, the easier the placement of the clip and
the less risk of stenosis of the parent internal carotid artery. After the clip is
placed in these complex cases, it is mandatory to determine that the
internal carotid artery is patent. Although intraoperative micro Doppler or
ultrasonography can be used to demonstrate patency, intraoperative
angiography is superior because it reveals whether there is significant
stenosis adjacent to the clip. If the ipsilateral cervical carotid artery has
already been exposed, intraoperative angiography can be performed
through a direct common carotid artery puncture.
Neurosurgery Books
CRANIOPHARYNGIOMA

There are several good ex posures for removal of suprasellar craniopharyngiomas, including a
frontotemporal orbital craniotomy and a standard pterional craniotomy. The true subfrontal approach,
with removal of one orbital roof, can be advantageous when there is a large extension of the
craniopharyngioma into the third ventricle (described in Chapter 3). However, a large extension into the
third ventricle can also be removed through a standard pterional craniotomy by vigorous internal
decompression of the tumor, often in combination with an incision in the lamina terminalis posterior to
the optic chiasm. With a pterional craniotomy, the expanded optic-carotid triangle can be used to the
surgeon’s advantage. Sectioning the lamina terminalis during a pterional approach is often more difficult
than expected, because the posterior margin of the optic chiasm may not be visible. In this case, it is best
to perform an aggressive debulking of the tumor through the optic-carotid triangle and to collapse the
third ventricular extension of the tumor into the subchiasmatic cistern. Occasionally, the infundibulum
can be preserved, especially if the craniopharyngioma is largely cystic. However, if the tumor is more
calcified and grumous, the infundibulum typically enters the tumor capsule and cannot be preserved.
After the Sylvian fissure is divided and the retractors are placed, the optic-carotid triangle is
identified. Typically, the ipsilateral carotid artery is displaced laterally, the anterior cerebral artery and its
perforating branches are pushed upward, and the optic chiasm is bowed medially and upward. The
arachnoid is first incised between the tumor capsule and medial surface of the internal carotid artery.
This incision is then carried distally between the anterior cerebral artery and the tumor capsule. Finally,
the arachnoid between the optic nerve and chiasm and the tumor is incised. The arachnoid between the

45
Neurosurgery Books
anterior cerebral artery and the frontal lobe is left intact. A Yasargil retractor is placed deeper along the
inferior aspect of the frontal lobe, which helps to move the anterior cerebral artery out of harm’s way.
The next step is to make a linear incision, 5 to 10 mm, in the tumor capsule. An internal
decompression is then performed with ring curets. After the tumor has been partially decompressed, the
lateral aspect of the tumor capsule is dissected off the medial aspect of the internal carotid artery.
Typically, the posterior communicating artery and its branches are ensheathed in arachnoid, which
facilitates dissection and separation from the lateral aspect of the tumor capsule. After this has been
achieved, cottonoids are placed between the outer tumor capsule and the inner aspect of the internal
carotid artery, the posterior communicating artery, and the anterior choroidal artery.
With a #5 or #7 straight suction tip and a cottonoid, the tumor is gently retracted medially and
dissection is performed behind the tumor capsule, with preservation of the arachnoid overlying the
basilar artery and its perforating vessels. The tumor is further decompressed internally to facilitate
manipulation of the capsule wall. After the basilar artery, the basilar caput, and, possibly, the posterior
cerebral arteries have been identified, cottonoids are placed deep to protect them.
Next, attention is directed to dissecting the tumor off the opposite anterior clinoid process and
internal carotid artery. Typically, the arachnoid covers the internal carotid artery and posterior
communicating artery, and if it is left intact, it will protect these structures. The tumor capsule is
retracted toward the surgeon with a #5 or #7 straight suction tip, which stretches the adhesions between
the capsule and the internal carotid artery, making it easier to divide the adhesions.

46
Neurosurgery Books
 After the tumor has been dissected off the opposite internal carotid artery, the angle of the
microscope is repositioned to allow a more direct view underneath the optic chiasm. It usually is
necessary, once again, to debulk the tumor to facilitate retraction of the capsule. The tumor capsule is
again retracted toward the ipsilateral internal carotid artery with the use of a #5 or #7 straight suction tip
and cottonoid, which stretches the adhesions between the tumor capsule and the optic chiasm and the
hypothalamus. They can be dissected free more easily, after which cottonoids are placed to protect the
underside of the chiasm and the hypothalamic region.
It is important to emphasize that the residual tumor capsule should be removed through the optic-
carotid triangle with gentle force. While the capsule is being removed, all aspects are inspected to make
sure that no residual adhesions remain between the tumor capsule and the surrounding vascular and
parenchymal structures. After the tumor is removed, the surgical bed is inspected. A malleable fiberoptic
probe or mirrors are useful in inspecting the underside of the optic chiasm and the third ventricle for
possible residual tumor.

47
Neurosurgery Books
Figure 1-27.
Step 1. The Sylvian fissure is divided
to identify the bifurcation of the internal
carotid artery. The frontal lobe is lined
with hemostatic fabric and cottonoids and
retracted. The arachnoid between the
tumor and medial aspect of the internal
carotid artery and anterior cerebral
artery is incised. The arachnoid between
the anterior cerebral artery and the
frontal lobe is left intact, so that when the
frontal lobe is retracted the anterior
cerebral artery is gently displaced.
During the operation, most of the work is
performed through the expanded optic-
carotid triangle. Tumor debulking can
also be performed between the two optic
nerves if the chiasm is postfixed.
Neurosurgery Books
 Figure 1-28.
Step 2. After the lateral
aspect of the tumor capsule has
been defined, a 5- to 10-mm
incision is made in the tumor.
Various ring curettes are used to
begin debulking the tumor.
During this process, it is
important to make sure that
traction is kept to a minimum on
the optic apparatus.
Neurosurgery Books
Figure 1-29.
Step 3. After the tumor is partially debulked,
the surgeon dissects the lateral margin of the tumor
off the ipsilateral internal carotid artery. There
typically is a layer of arachnoid between the tumor
capsule and the posterior communicating and
anterior choroidal arteries and their perforating
arteries. Respecting this arachnoid barrier will
protect these vascular structures. The tumor capsule
is gently displaced medially with a #5 or #7 straight
suction tip and cottonoid. After the perforating
arteries that come off the internal carotid artery are
identified, they are protected with cottonoids.

Neurosurgery Books
Figure 1-30.
Step 4. The dissection is carried deeper to identify the
basilar artery. The arachnoid that invests the posterior
communicating artery extends deep between the tumor capsule
and basilar artery. This arachnoid should not be violated. The
tumor is retracted medially with a #5 or #7 straight suction tip
and a cottonoid. After the posterior component of the tumor
capsule has been identified, cottonoids are placed over the
basilar artery.

Figure 1-31.
Step 5. Next, the tumor capsule is dissected off the
contralateral internal carotid artery and the anterior
clinoid process. In this case, there is a postfixed chiasm;
thus, the dissection is performed between the two optic
nerves. The arachnoid plane between the carotid artery
and its branches and the tumor capsule should be
respected
Neurosurgery Books
Figure 1-32.
Step 6. The tumor is further decompressed
internally with ring curets. To work across the
middle cranial fossa, the operating microscope
is repositioned to provide a more lateral or
horizontal view of the ipsilateral optic nerve.
This permits dissection of the interface between
the tumor and the optic nerve and chiasm. There
is no arachnoid plane between these structures.
The tumor capsule should be retracted
downward; oftentimes, this retraction peels the
tumor capsule off the underside of the chiasm. It
is important to recognize that the primary blood
supply to the chiasm is from the underside, and,
therefore, all minute blood vessels must be
preserved. In patients with a large extension of
craniopharyngioma into the third ventricle or a
prefixed optic chiasm, it is useful to remove this
ventricular component through the lamina
terminalis. This is described in Chapter 3.

Neurosurgery Books
Figure 1-33.
Step 7. After the tumor has
been dissected off the chiasm
and extracted from the third
ventricle, it is removed through
the optic·carotid triangle with a
forceps. After the tumor has
been removed, the underside of
the chiasm and the third
ventricle are inspected with
either mirrors or a malleable
fiberoptic probe to make sure
that the tumor has been resected
completely.

Neurosurgery Books
SPHENOID WING MENINGIOMA

Meningiomas located along the medial two-thirds of the sphenoid wing are best approached through
a standard pterional craniotomy, whereas those located on the lateral one-third should be approached
through a larger frontotemporal craniotomy. It is important to recognize that often one or two perforating
arteries off the middle cerebral artery enter the deep aspect of the tumor capsule. Thus, to prevent
evulsing these branches, the tumor must not be pulled out before the underside or deep component of the
tumor is dissected.
After the bone flap is removed, the outer sphenoid wing is removed with a high-speed air drill,
under the operating microscope. This is useful in increasing exposure and devascularizing part of the
tumor. The dura mater is opened widely, and any branches off the middle meningeal artery that were
exposed when the outer sphenoid wing was removed are cauterized and divided because they typically
feed the tumor. The brain is lined with hemostatic fabric and cottonoids, and the tumor capsule is
separated from the overlying frontal and temporal lobes.
Separating the tumor from the overlying parenchyma is made easier by placing the brain under
slight tension with a Yasargil retractor and providing countertraction on the tumor capsule with a #5 or
#7 straight suction tip. This places tension on the adhesions between the tumor capsule and the
arachnoid, easing separation. First, the lateral margin of the tumor is dissected free from the temporal
lobe. Second, the distal Sylvian fissure is divided, identifying the interface between the tumor and the
middle cerebral artery. After the middle cerebral artery is identified, cottonoids are placed over it for
protection. The interface between the frontal lobe and the tumor is then identified and dissected free, and
cottonoids are placed to protect the frontal and temporal lobes.
54
Neurosurgery Books
 After an initial dissection has been performed between the tumor and the frontal and temporal lobes,
the tumor capsule is cauterized and the tumor is decompressed internally with curets and an ultrasonic
aspirator. The greater the internal decompression of the tumor, the easier it is to retract the tumor capsule
from the neurovascular structures.
Typically, the blood supply of these tumors is from the dura mater of the sphenoid wing. Therefore,
after debulking some of the tumor and clearly distinguishing the borders of the tumor from the frontal
and temporal lobes, the tumor attachments to the sphenoid wing are cauterized. The tumor capsule is
retracted with a #5 or #7 straight suction tip and cottonoids, thereby allowing better definition of the
interface between the tumor and sphenoid wing. Bipolar cautery with irrigation is used to obtain
hemostasis of the dura mater, which often is quite vascular.
The critical component in tumor resection is the deep aspect of the tumor and its relationship to the
supraclinoid carotid artery, the middle cerebral artery, and the optic nerve. Occasionally, one can work
from lateral to medial along the Sylvian fissure, identifying the middle cerebral artery and following it
proximal to the carotid bifurcation. When this is achieved, much of the tumor can be removed. This
leaves a small piece of tumor along the anterior clinoid process.
Underneath the anterior clinoid process, the internal carotid artery and optic nerve usually are
separated from the tumor by overlying arachnoid. It is important to preserve this arachnoid (if possible)
because it protects these structures. The residual bit of tumor is rolled outward along the sphenoid wing
to identify the internal carotid artery and its bifurcation. Although the tumor may wrap itself around the
internal carotid artery, the arachnoid plane between this artery and the tumor often allows removal of the
tumor between the internal carotid artery and its branches laterally and the tentorial edge medially.
55
Neurosurgery Books
Figure 1-34.
The meningioma illustrated involves much of the
sphenoid wing, both medially and laterally.
56
Neurosurgery Books
Figure 1-35.
Step 1. The interface between the tumor and the temporal
lobe and, then, the frontal lobe is identified and separated.
Typically, a Yasargil retractor is used to retract the brain and a
#5 or #7 straight suction tip and cottonoid are used to place
countertraction on the tumor capsule. This places the arachnoid
adhesions under tension, allowing easier separation of the
tumor from the brain.

Figure 1-36.
Step 2. After the tumor is separated from the frontal and
temporal lobes, it is decompressed with curets, laser, or an
ultrasonic aspirator.
Neurosurgery Books
 Figure 1-37.
Step 3. The outer or lateral Sylvian fissure is divided, and the
relationship between the tumor and the middle cerebral artery is
dissected free. Often, one or two branches off the middle cerebral
artery are incorporated into the tumor capsule. Because of this, the
underside of the tumor must be separated completely from the middle
cerebral artery before the tumor is removed. Otherwise, some of the
perforating branches of the middle cerebral artery may be evulsed.
Also, the tumor may parasitize several of these vessels. Branches of
the middle cerebral artery that clearly supply the tumor may be
safely cauterized and divided.

Figure 1-38.
Step 4. With increasing exposure, further internal
decompression of large tumors is performed. This internal
decompression facilitates the manipulation of the tumor capsule
off the deeper neurovascular structures.
Neurosurgery Books
Figure 1-39.
Step 5. After extensive debulking of the tumor and dissection of
its lateral and medial margins, the attachments of the tumor to the
dura mater of the sphenoid wing are further cauterized and divided
down to the anterior clinoid process.

Figure 1-40.
Step 6. Typically, there is a layer of arachnoid between the deepest
part of the tumor capsule and the underlying internal carotid artery and
optic nerve. This arachnoid prevents injury to these structures during
dissection of the tumor. The attachment of the tumor to the anterior
clinoid process and the planum sphenoidal is cauterized and divided.
Countertraction is provided by a #5 or #7 straight suction tip placed on
the tumor capsule.
Neurosurgery Books
Figure 1-41.
Step 7. The residual piece of
tumor is rotated medially and
laterally as the tumor is
dissected off the internal carotid
artery and its branches. Often
the tumor encircles part of the
internal carotid artery, but it
usually can be removed from the
artery with gentle suction, fine
spatulas, and bipolar cautery.
Occasionally, one can visualize
the tumor invading the adventitia
of the internal carotid artery.
Fine spatulas should be used to
scrape the tumor off the artery.
Any bleeding should be
controlled with an absorbable
gelatin sponge instead of bipolar
cautery. Cauterizing small
bleeders on the internal carotid
artery can lead to arterial
dissection and thrombosis.

Neurosurgery Books
TRANSITIONAL CAVERNOUS CAROTID ANEURYSM

Anatomy

It is mandatory for surgeons to understand exactly the three-dimensional anatomic relationships of

the cavernous sinus before operating on a transitional cavernous carotid aneurysm. The important
features are the course of the internal carotid artery through the cavernous sinus, the relationship of the
cranial nerves to the internal carotid artery, and the enveloping and tethering reflections of the dura
mater. When considering the anatomy and surgery of the cavernous sinus, neurosurgeons are indebted to
the pioneering work of Parkinson and Dolenc. The following discussion of the anatomy of the cavernous
sinus should be considered a bare minimum, or cursory, and relevant primarily to transitional cavernous
carotid aneurysms.
The internal carotid artery enters the cavernous sinus through the foramen lacerum and runs
superiorly and curves anteriorly toward the superior orbital fissure (Figure 1-42). The m ningeal
hypophysial trunk usually arises from the superior medial aspect of the internal carotid artery just distal
to this curve. The three branches of the meningeal hypophysial trunk include the tentorial artery of
Bernasconi and Cassinari, the dural meningeal artery, and the inferior hypophysial artery. There also
may be an artery of the inferior cavernous sinus, which originates from the lateral side of the internal
carotid artery and runs anteriorly to cross the abducens nerve and eventually anastomose with branches
of the maxillary artery. Some of the anastomoses include the artery of the foramen rotundum, the
recurrent meningeal artery along the superior orbital fissure, the accessory meningeal artery at the
foramen ovale, and the middle meningeal artery at the foramen spinosum. Approximately 10 percent of
61
Neurosurgery Books
patients have McConnell’s capsular arteries, which originate from the medial side of the internal carotid
artery and supply part of the pituitary gland.
Within the cavernous sinus, a dense venous plexus connects the ophthalmic vein, the superior and
inferior petrosal sinuses, the pterygoid plexus, and the basilar venous plexus. This venous plexus
surrounds the internal carotid artery. The oculomotor, trochlear, and trigeminal nerves run in the lateral
wall of the cavernous sinus. The dura mater of the cavernous sinus is continuous with the connective
tissue that tethers these cranial nerves to the wall of the sinus. The abducens nerve runs within the
cavernous sinus and enters the superior orbital fissure under the ophthalmic division of the trigeminal
nerve.
Several nomenclatures have been proposed for divisions of the internal carotid artery. The first was
that of Fisher in 1938, subsequently modified by Fukushima in 1988. In this classic nomenclature, the
C1 segment begins at the carotid bifurcation and extends to the origin of the posterior communicating
artery. The C2 segment extends from the posterior communicating artery to the distal dural ring. This
has also been described by Day as the “ophthalmic segment.” The C3 segment is the extradural
extracavernous part of the artery and lies underneath the anterior clinoid process. The C4 segment is the
intracavernous segment, which extends to the origin of the meningeal hypophysial trunk. The C5
segment extends from the meningeal hypophysial trunk to underneath the trigeminal nerve. The C6
segment, the last segment, is the intrapetrous portion of the internal carotid artery, from where it crosses
underneath the mandibular division of the trigeminal nerve to its entrance at the foramen lacerum. More
recently, van Loveren and colleagues introduced a nomenclature that essentially reverses and modifies

62
Neurosurgery Books
the numbering sequence to reflect the direction of blood flow. To avoid confusion, terms such as
“horizontal segment” and
“anterior genu” are used in this
book. Figure 1-42.

There are two so-called

carotid dural rings, and they are
important anatomic landmarks
(Figure 1-43). The proximal ring
denotes where the internal
carotid artery exits from the
cavernous sinus to become the
clinoid segment of the internal
carotid artery. The distal ring is
the point where the internal
carotid artery becomes
intradural. Accordingly, these
two rings delineate the clinoid
segment of the internal carotid
artery. The distal dural ring
surrounds the internal carotid
artery and is continuous with the
dura mater of the falciform
63
Neurosurgery Books
ligament, the anterior Figure 1-43.
clinoid process, and the
cavernous sinus. It is
attached to the adventitia of
the internal carotid artery.
The proximal dural ring,
which incompletely
surrounds the internal
carotid artery, is made up of
periosteum arising from
the anterior clinoid
process. The membrane
between the internal
carotid artery and the
oculomotor nerve is
sometimes called the
“caroticooculomotor
membrane.” The nine
triangles of the cavernous
sinus region need to be
considered (Figure 1-44).
1. Anterior Medial
64
Neurosurgery Books
Figure 1-44. Triangle
The anterior medial triangle is defined by the medial
border of the optic nerve medially and the oculomotor
nerve laterally. The base of the triangle is the dural edge of
the tentorium cerebelli. The anterior loop of the internal
carotid artery lies in the floor of the triangle, along with
trabeculated venous channels. This is called the “clinoid
segment” of the internal carotid artery, and it is neither
intradural nor intracavernous. Most transitional cavernous
carotid aneurysms arise from this clinoid segment and
penetrate through the distal dural ring to enter the
subarachnoid space. Fibrous connective tissue in the apex
of the triangle forms the proximal ring that ensheaths the
internal carotid artery and delineates the anteromedial
border of the cavernous sinus.
2. Paramedial Triangle
The two sides of the paramediai triangle are defined medially by the medial border of the
oculomotor nerve and laterally by the lateral border of the trochlear nerve. The base of the triangle is the
dural edge of the tentorium cerebelli. The anterior loop and part of the horizontal segment of the
cavernous carotid artery can be visualized through this aperture. Some surgeons have extended the para
medial triangle to a “medial triangle” that has the above-mentioned margins but a base that extends
65
Neurosurgery Books
 Figure 1-44. along the dura mater from the anterior to the
posterior clinoid process; “the oculomotor trigone.”
Incising the dura mater along this oculomotor
trigone and then extending the incision into the
paramedial triangle increases the exposure of the
horizontal segment of the cavernous carotid artery.
3. Parkinson’s Triangle
Parkinson’s triangle is defined medially by the
trochlear nerve and laterally by the first division of
the trigeminal nerve. The base of this triangle is
also the dural edge of the tentorium cerebelli.
Access through this triangle provides visualization
of the horizontal segment of the cavernous carotid
artery.
4. Anlerior Lateral Triangle
The anterior lateral triangle is defined medially by the ophthalmic division of the trigeminal nerve
and laterally by the maxillary division of the trigeminal nerve. The base of the triangle is formed by a
line running anteriorly between the superior orbital fissure and the foramen ovale. This triangle can be
used to expose the superior orbital vein and to gain access to the anterior portion of the cavernous sinus.
5. Lateral Triangle
66
Neurosurgery Books
 Figure 1-44. The lateral triangle is defined medially by the maxillary
division of the trigeminal nerve and laterally by the
mandibular division of the trigeminal nerve. The base is
formed by a line extending from the foramen rotundum to
the foramen ovale. The lateral triangle can be used when
exposure of the anterior lateral part of the cavernous
sinus is necessary for tumor excision.
6. Posterior Lateral Triangle (Glasscock’s Triangle)
The posterior lateral triangle is defined by the posterior
rim of the foramen ovale, the foramen spinosum, the
posterior border of the mandibular division of the
trigeminal nerve, and the cochlear apex (greater petrosal
nerve). This triangle is important for obtaining proximal
control of the horizontal segment of the intrapetrous
internal carotid artery for temporary occlusion or bypass procedures. This space is approached
extradurally. Often, the internal carotid artery can be seen or palpated because it is covered only by a
fibrous membrane instead of bone. Exposure of this triangle is achieved by using a diamond bur. After
control of the middle meningeal artery is achieved, drilling commences just medial to the foramen
spinosum and progresses along the posterior border of the mandibular branch of the trigeminal nerve.
The greater petrosal nerve must be sectioned to prevent traction on the geniculate ganglion.

67
Neurosurgery Books
 Figure 1-44. 7. Posterior Medial Triangle (Kawase’s Triangle)
The posterior medial triangle is defined laterally by the
greater superior petrosal nerve and medially by the
petrosal sinus. The base is the trigeminal nerve. No
pertinent neurovascular structures occur in this triangle.
This ridge of petrous bone within Kawase’s triangle may
safely be removed to provide greater access to the
posterior cranial fossa when performing a transtentorial
approach. The posterior medial triangle also provides
greater exposure of the trigeminal nerve.
8. Inferior Medial Triangle
The inferior medial triangle is defined medially by a
line between the posterior clinoid process and the
abducens nerve at Dorello’s canal and laterally by a line running between Dorello’s canal and the
trochlear nerve at the edge of the tentorium. The base of the triangle is the petrous apex.
9. Inferior Lateral Triangle
The inferior lateral triangle is defined medially by a line running between Dorello’s canal and the
trochlear nerve at the edge of the tentorium cerebelli and laterally by a line between Dorello’s canal and
the petrosal vein at the petrosal sinus. The base is the tentorium cerebelli.

68
Neurosurgery Books
Operation

Most transitional carotid cavernous aneurysms can be approached through a modified pterional
craniotomy that requires more removal of temporal bone. Some surgeons prefer to approach these
aneurysms through a frontotemporal zygomatic craniotomy. If the aneurysm extends proximally to the
proximal carotid ring, removal of the zygoma will allow better exposure of the lateral cavernous sinus.
In most larger cavernous carotid aneurysms, proximal control of the internal carotid artery is
desirable. This can achieved through an extradural exposure of the petrous carotid artery. Alternatively,
the cervical carotid artery can be exposed through a linear incision anterior to the sternocleidomastoid
muscle. Exposure of the cervical carotid artery has the advantages of being easier technically and
providing a route to perform intraoperative angiography if so needed. Exposure of either the cervical or
petrous carotid offers the surgeon some degree of safety should the aneurysm inadvertently rupture
intraoperatively.

69
Neurosurgery Books
Figure 1-45.
Step 1. Most transitional cavernous carotid aneurysms can be explored through an enlarged pterional craniotomy
that has some extension into the middle cranial fossa. As illustrated here, the cervical carotid artery is exposed to
provide proximal blood vessel control should an inadvertent rupture occur intraoperatively. An alternative to this is a
frontotemporal zygomatic craniotomy, with extradural exposure of the petrous carotid artery. Exposure of the cervical
carotid artery is technically easier and quicker, and it provides an easier route for performing intraoperative
angiography if needed. Note that the position of the patient’s head is slightly more lateral than for a true pterional
craniotomy to facilitate dissection of the cavernous sinus.
Neurosurgery Books
 Figure 1-46.
Step 2. After the bone flap is elevated, the frontal and temporal dura
mater are dissected off the sphenoid wing and protected with malleable
retractors. The outer sphenoid wing is removed with a high-speed diamond
air drill. Approximately 2 to 3 cm deep to the outer sphenoid wing, the drill
is used to make a hole into the orbital roof.

Figure 1-47.
Step 3. The hole created in the orbital roof is extended by using a small bone
forceps or punch. Both small spatulas and ball-tip dissectors can be used to
dissect the inner sphenoid wing and anterior clinoid process off both the orbital
fascia and the frontal and temporal dura mater. As the dissection is extended
deeper, two struts of bone tether the anterior clinoid process. Both of these struts,
the optic strut and the lateral rim over the superior orbital fissure, can be
removed with small bone forceps. After these two bony struts are removed, a fine
spatula is used to dissect the anterior clinoid process off the dura mater. The
dura mater is then grabbed with the bone forceps, gently rocked, and rotated
medially and laterally until it is loose, and then it is pulled out of the junction
between the frontal dura mater, temporal dura mater, and orbital fascia.
Neurosurgery Books
Figure 1-48.
Step 4. This figure illustrates the location of the aneurysm in relation to the removed part of the medial
sphenoid wing and the anterior clinoid process. This extradural removal is advantageous because dura mater
overlies the aneurysm, the internal carotid artery, and the optic nerve. The relationship of the optic nerve to the
medial bony optic strut indicates why removal of this bone with bone forceps decreases the risk of thermal
injury to the optic nerve, which might occur with a high-speed air drill.
The dura mater is opened in a curved fashion and tacked to the overlying temporalis muscle. This dural
flap prevents blood from running into the wound. At the end of the operation, the dura mater can be closed
either primarily or with a graft. In this way, the only dural defect that remains is the one in the cavernous sinus
region, and the risk of postoperative leakage of cerebrospinal fluid and wound complications is minimized.
Neurosurgery Books
 Figure 1-49.
Step 5. As in a standard pterional
craniotomy, the next step is to separate the
Sylvian fissure. In this illustration, the fissure is
being divided from lateral to medial. However,
as mentioned above, in some situations it may
be necessary to divide the fissure from medial to
lateral, especially in young patients or after a
significant subarachnoid hemorrhage in which
blood is packed in the Sylvian fissure. Although
small bridging veins overlying the Sylvian
fissure may safely be cauterized and divided, it
is best to preserve the temporal veins by
displacing them laterally with the temporal
lobe. This will help decrease the risk of
postoperative seizures and contusions of the
temporal lobe.
Neurosurgery Books
 Figure 1-50.
Step 6. The Sylvian fissure is divided, and after
the brain is protected with hemostatic fabric and
cottonoids, the frontal lobe retractor is again
repositioned deeper. The arachnoid over the
supraclinoid carotid artery is divided to identify the
bifurcation of the internal carotid artery. Typically,
the rostral portion of the aneurysm is immediately
apparent. Most often, the optic nerve is displaced
medially by the medial projection of the dome of the
aneurysm.
Exposure of the cavernous portion of the
aneurysm is first accomplished by incising the dura
mater linearly along the presumed axis of the
internal carotid artery over the site where the
extradural bone was removed. Before this incision is
made, a small spatula is used to gently depress the
dura mater intended to be cut to make sure that no
part of the aneurysm dome lies immediately beneath
the site.
Neurosurgery Books
Figure 1-51.
Step 7. A small ball-tip dissector is used to separate the distal
dural ring from the underlying aneurysmal dome. The ball-tip
dissector is swept medially to laterally. The freed dural ring is then
incised with either a scalpel, micro Potts scissors, or standard
straight microscissors. The advantage of either the micro Potts or
microscissors is that the dural ring can be lifted off the dome of the
aneurysm and cut more easily than with a scalpel.
After the distal dural ring is incised, it is necessary to dissect
the medial and lateral portions of the aneurysm. Most often, the
dome of the aneurysm is adherent to the lateral portion of the wall
of the cavernous sinus. A small spatula can be used to sharply
dissect the dome off this wall. It is important to recognize that
cranial nerves run in this lateral cavernous wall. At this level, the
primary nerve at risk for injury is the oculomotor nerve.
Medially, it is necessary to identify the origin of the ophthalmic
artery. Most often, the origin of this artery marks the junction
between the aneurysm and the parent carotid artery along the
medial border. This is an important landmark because it may be
difficult to identify this transition zone. Sometimes, it is not enough
to remove the bony optic strut. If so, a small punch can be used to
remove more bone to allow a better line of site to identify the
junction between the aneurysm and the parent artery.
It also is important to identify the proximal margin of the
aneurysm and its point of takeoff from the parent carotid artery.
Usually, this proximal neck occurs at the anterior genu of the
internal carotid artery, as it starts to curve laterally and deep.
Neurosurgery Books
 Figure 1-52.
Step 8. The dome of the aneurysm is displaced with a #5
or #7 straight suction tip and cottonoid to allow exact
identification of the junction of the ophthalmic artery, parent
carotid artery, and medial neck of the aneurysm. The best clip
to use for most of the aneurysms is a forward-angled straight
clip passed along the long axis of the artery. Helpful visual
and tactile guides to use for the initial placement of the clip
are the ophthalmic artery medially and the anterior wall of
the cavernous sinus. If the clip is placed just above the
ophthalmic artery and inserted until there is resistance when
the tips of the clip hit the anterior bony wall of the cavernous
sinus, it is likely that this initial clip placement will
successfully obliterate the neck without compromising the
internal carotid artery.
After the location of the jaws of the aneurysm clip have
been inspected visually, the dome is aspirated for
decompression of the aneurysm. Thereafter, correct placement
of the aneurysm clip is reconfirmed. At this point, it is useful
to have objective documentation that the aneurysm has been
obliterated and that the parent internal carotid artery is not
stenosed or occluded. Intraoperative Doppler or
ultrasonography of the internal carotid–middle cerebral
artery complex is useful. However, currently, intraoperative
angiography is the best method of documentation. Therefore,
if available, a radiolucent headholder and table should be
used. Intraoperative angiography can be performed through
either a direct cervical carotid puncture or by femoral artery
catheterization.
Neurosurgery Books
Chapter 2

Frontotemporal
Approach

Neurosurgery Books
Chapter 2: Frontotemporal Approach
ANATOMIC CONSIDERATIONS

Gross Anatomy

The temporal lobe occupies the anterior portion of the cerebral hemisphere below the lateral, or
Sylvian, fissure and anterior to the parietooccipital sulcus (Figure 2-1). The supramarginal gyrus of the
parietal lobe caps the Sylvian fissure, and the angular gyrus curves around the end of the superior
temporal sulcus. Wernicke’s area (the posterior speech area) is located in the region of the supramarginal
and angular gyri and the posterior portion of the superior temporal gyrus of the dominant hemisphere.
An imaginary line connecting the parietooccipital sulcus and the temporooccipital notch marks the
posterior limit of the temporal lobe. The temporal lobe has three surfaces: lateral, inferomedial, and
superior. The lateral surface has three longitudinal gyri—the superior, middle, and inferior temporal gyri
—separated by the superior and inferior temporal sulci. If the floor of the middle cranial fossa is not
exposed during a craniotomy, often only the first two gyri will be visualized. The superior temporal
sulcus is parallel with the Sylvian fissure, and its ascending ramus ends in the angular gyrus.
The inferior temporal gyrus curves onto the inferomedial surface of the temporal lobe. From lateral
to medial on the inferomedial surface are the inferior temporal gyrus, the fusiform (or lateral
occipitotemporal) gyrus, and the parahippocampal gyrus. The fusiform gyrus occurs only in the posterior

78
Neurosurgery Books
half of the temporal lobe and is set off by the occipitotemporal sulcus laterally and the collateral sulcus
medially. The posterior end of the parahippocampal gyrus curves around the splenium and becomes the
cingulate gyrus. The portion of this gyrus that is immediately posterior and inferior to the splenium is
called the “isthmus” of the cingulate gyrus. The anterior end of the parahippocampal gyrus curves
medially and turns posteriorly to form the uncus. Deep to the parahippocampal gyrus lies the convoluted
hippocampal formation.
The superior surface of the temporal lobe (the opercular surface) contains the transverse gyrus of
Heschl, or auditory cortex. Posterior to this gyrus is the planum temporale.
Auditory cortex (Brodmann areas 41 and 42) is tonotopically organized, with low frequencies
represented anterolaterally and high frequencies posteromedially. Damage to auditory cortex is
associated with subtle auditory impairment (mainly, impaired ability to localize sound in space), because
the auditory cortex in each cerebral hemisphere receives input from both ears. In addition, areas 41 and
42 are connected with the corresponding areas in the opposite hemisphere through the corpus callosum.
Area 22, primarily responsible for the comprehension of spoken language, is located on the lateral
temporal surface adjacent to areas 41 and 42. It receives projections from auditory cortex.
Because of the bilaterality of the projections in the auditory system, including projections through
the corpus callosum, deafness in one ear does not produce aphasia. However, destruction of area 22 in
the dominant cerebral hemisphere can result in auditory aphasia, as can bilateral destruction of areas 41
and 42. Injury to an area inferior and adjacent to area 22 can produce anomia (amnesic aphasia), the
inability to recall names.

79
Neurosurgery Books
 It is important to recognize that the location of eloquent language cortex in both the temporal and
frontal lobes can vary from patient to patient. For surgical procedures in which potential injury of these
cortical areas is a concern, preoperative mapping or awake surgery with intraoperative mapping is
necessary.
Part of the visual system is also contained in
the temporal lobe, specifically, Meyer’s loop of
the geniculocalcarine tract (visual radiations).
Meyer’s loop is the portion of the
geniculocalcarine tract that extends forward into
the temporal lobe and curves around the tip of
the temporal horn of the lateral ventricle. These
visual fibers represent the superior quadrant of
the visual field. Temporal lobectomy may
involve Meyer’s loop. However, the degree of
visual impairment depends on how far
posteriorly the resection extends. Resections that
extend more than 9 cm from the tip of the
temporal lobe produce a complete homonymous
hemianopia. The temporal lobe also contains
visual association cortex important in the
recognition of shapes, including faces.

80
Neurosurgery Books
THE HIPPOCAMPUS

Structure

The hippocampal formation is a special area of cerebral cortex and consists of the dentate gyrus, the
hippocampus proper (also called “Ammon’s horn” and usually referred to as “hippocampus”), and the
subiculum and presubiculum. The subiculum and presubiculum are part of the parahippocampal gyrus
on the inferomedial surface of the temporal lobe. The dentate gyrus and hippocampus form a bulge in
the medial wall of the temporal horn of the lateral ventricle. immediately inferior to the choroidal
fissure. This bulge extends the length of the temporal horn. Anteriorly, the bulge is prominent and
indented and resembles a paw, hence the name “pes hippocampi.”
The hippocampus is considered the oldest part of the cerebral cortex and is called “archicortex.”
Unlike most of the cerebral cortex (“neocortex”), which has six recognized layers, the hippocampus
consists of only three layers. The middle layer is composed of the cell bodies of hippocampal pyramidal
cells, and the inner and outer layers are composed mainly of dendrites and axons. The major afferents to
the hippocampal formation synapse on granule cells in the dentate gyrus, the doorway to the
hippocampus. The axons of these granule cells synapse on hippocampal pyramidal cells. Many of the
axons of the hippocampal pyramidal cells, in turn, project to the subiculum and presubiculum. Other
hippocampal axons enter the fornix. The myelinated axons that join the fornix go to the ventricular
surface of the hippocampus and form a thin layer of white matter called the “alveus.” The axons of the
alveus converge and form the fimbria, a flange of white matter on the superior surface of the
hippocampus. Posteriorly, the fimbria becomes the fornix.
81
Neurosurgery Books
 On entering the temporal horn of the lateral ventricle, the surgeon sees the prominent pes
hippocampi covered by the thin white veil of the alveus and fimbria and the choroid plexus. In the
choroid plexus lie the anterior choroidal artery, the inferior choroidal vein (a branch of the basal vein of
Rosenthal), and the lateral posterior choroidal artery. If the choroid plexus is retracted superiorly and
medially, the ambient cistern is visible. This cistern contains the posterior cerebral artery, the basal vein
of Rosenthal, the cerebral peduncles, and the oculomotor and trochlear nerves.
Connections

The hippocampus is connected with a large portion of the rest of the cerebral cortex and with many
subcortical areas. The major source of afferents to the hippocampal formation is the entorhinal cortex of
the parahippocampal gyrus. Entorhinal cortex receives afferents from areas of sensory association cortex
—visual, auditory, ancl somatosensory. Axons of pyramidal cells in the entorhinal sensory. Axons of
pyramidal cells in the entorhinal cortex form the perforant path, which ends among the granule cells of
the dentate gyrus. The hippocampus and, to a larger extent, the subiculum and presubiculum project
back to the entorhinal cortex and, through it, to sensory association cortex. Other axons from the
hippocampus and subiculum and presubiculum project to subcortical areas through the fornix.
The Fornix

Near the posterior end of the hippocampus, the fimbria forms a thick bundle of axons called the
“fornix.” On each side, the fornix arches upward and forward under the splenium of the corpus
callosum. This part of the fornix is called the “crux” (plural, crura). The left and right crura are joined by
the commissure of the fornix, called the “psalterium.” The fornix continues anteriorly under the body of
the corpus callosum. This is the “body of the fornix.” Anteriorly, at the level of the foramen of Monro,
82
Neurosurgery Books
the fornix curves posteriorly and enters the substance of the diencephalon. Some of the axons pass
anterior to the anterior commissure; this is the precommissural component of the fornix. These axons are
distributed to the septal nuclei, the preoptic area, and the anterior portion of the hypothalamus. The rest
of the fornix curves posterior to the anterior commissure, forming the postcommissural component.
These axons are distributed to the anterior and intralaminar nuclei of the thalamus, the medial nucleus of
the mammillary body, and the midbrain reticular formation.
Functional Importance

The dentate gyrus, hippocampus, subiculum, presubiculum, and entorhinal cortex are of central
importance in declarative memory function, that is, the formation and the recall of memories of facts and
places. Lesions of subcortical areas associated with the fornix, especially the dorsomedial nucleus of the
thalamus and the mammillary bodies of the hypothalamus, have been implicated in the memory loss that
is part of Korsakoff syndrome, or amnestic confabulatory syndrome, a sequela of Wernicke’s
encephalopathy.

THE AMYGDALA

Structure

The amygdala is a collection of nuclei located in the anteromedial part of the parahippocampal
gyrus, deep to the region of cerebral cortex called the “uncus.” The amygdala is incompletely separated
from the pes hippocampi by the tip of the temporal horn of the lateral ventricle. The amygdala is divided
into a basolateral group and a corticomedial group of nuclei separated by the central nucleus.
83
Neurosurgery Books
Connections

The olfactory system is a major source of afferents to the corticomedial group of nuclei. The
basolateral group receives most of its afferents from the cerebral cortex, especially sensory association
cortex.
The basolateral group of nuclei project back to the cerebral cortex through the entorhinal cortex,
thus influencing function in sensory association cortex. In addition, the basolateral group sends axons to
the central nucleus, which is the origin of many subcortical projections from the amygdala, especially
through the stria terminalis. The amygdala also projects to subcortical areas through the ventral
amygdalofugal pathway. The subcortical areas that receive axons from the amygdala include the
thalamus, hypothalamus, and brain stem.
Functional Importance

It has been demonstrated experimentally in animals and, to a lesser extent, in humans that the
amygdala affects a wide range of behavioral, visceral, and endocrine functions roughly grouped into
“flight” and “fight” behaviors. For example, electrical stimulation of unanesthetized animals can lead to
intense behavioral arousal that begins with the arrest of all ongoing activity, the so-called freezing
reaction. This reaction may represent the initial phase of flight or fight behavior. Also, stimulation of
relatively discrete areas of the amygdala can produce specific reactions, including pupillary dilatation,
piloerection, and disturbances of awareness. The basolateral group of nuclei appear to be especially
important in learning associations between unconditioned and conditioned stimuli, for example, learning
to associate a buzzing sound with a damaging blast.

84
Neurosurgery Books
 Bilateral destruction of the amygdala produces profound emotional disturbances. Animals become
placid and do not evince reactions of rage, aggression, or fear. Similar changes have been observed in
humans after bilateral damage to the amygdala.
Vascular Supply

The temporal lobe is supplied by three arteries: the middle cerebral artery, the anterior choroidal
artery, and the posterior cerebral artery. The predominant blood supply is from the middle cerebral
artery, which supplies the superior and lateral surfaces of the temporal lobe. The middle cerebral artery
is divided into four segments: M1, from the origin of the middle cerebral artery to the bottom of the
Sylvian fissure; M2, from the bifurcation-trifurcation of the artery in the Sylvian fissure to the distal
edge of the insula; M3, the opercular segment; and M4, the cortical segment distal to the operculum.
Typically, three to five small perforating vessels originate from the proximal M1 segment; these are the
lenticulostriate arteries. The size of the medial group of lenticulostriate arteries is inversely related to the
size of the ipsilateral recurrent artery of Heubner and the medial lenticulostriate arteries from the
anterior cerebral artery. The middle cerebral artery may have either a bifurcation or trifurcation. Two or
three large perforating branches usually originate just proximal to the bifurcation of the middle cerebral
artery and go medially into the insular region to supply lateral and posterior lenticular structures (lateral
lenticulostriate arteries). Often, no perforating branches arise from the trunk of the middle cerebral
artery between the origin of the medial and lateral lenticulostriate arteries. Therefore, this short length of
segment M1 is fit for temporary occlusion of the middle cerebral artery if necessary. Approximately 50
percent of patients have a branch off the proximal portion of the M1 segment that runs lateral to the
anterior temporal lobe.

85
Neurosurgery Books
 The anterior choroidal artery arises from the internal carotid artery approximately 2 mm distal to the
origin of the posterior communicating artery. It runs lateral to the posterior communicating artery,
parallel with the optic tract. This vessel varies considerably in diameter and may arise as more than one
vessel; it is duplicated in 4 to 30 percent of patients. Occasionally, it can originate from either the middle
cerebral artery or the posterior communicating artery. The anterior choroidal artery is divided into a
cisternal segment and a plexal segment. The cisternal segment is the source of perforating branches that
go medially to supply the uncus, optic tract, lateral geniculate body, cerebral peduncle, anterior
perforated substance (putamen and internal capsule), hippocampus, and pulvinar. The plexal segment
enters the choroidal fissure to supply the choroid plexus. Several branches usually originate from the
proximal trunk of the plexal segment to supply the hippocampus.
The posterior cerebral artery is the source of two arterial supplies of the temporal lobe. First,
approximately three to four branches of the P2 segment of the posterior cerebral artery supply the
hippocampus; these vessels enter the hippocampus through the choroidal fissure. Second, the posterior
temporal artery arises at the junction of segments P2 and P3. This artery goes along the floor of the
middle cranial fossa underneath the temporal lobe and perfuses the inferior surface and adjacent lateral
surface of the temporal lobe.
Venous drainage of the temporal lobe includes the deep venous and superficial systems. The
superficial system involves the Sylvian vein, which has connections with the inferior and superior great
anastomotic veins posteriorly and the cavernous sinus and sphenoparietal sinus anteriorly. The deep
venous system involves the deep Sylyian vein, which drains into the basal vein of Rosenthal. The
anterior hippocampal vein, the uncal veins, the anterior and posterior longitudinal hippocampal veins,
86
Neurosurgery Books
the inferior ventricular vein, and the inferior choroidal vein all drain into the basal vein of Rosenthal.
The basal vein of Rosenthal drains into the internal cerebral vein. Cortical polar veins often come off the
tip of the temporal lobe and drain directly into the sphenoparietal sinus.

TEMPORAL LOBECTOMY

Conceptually, there are two types of temporal lobectomy, depending on the goals of the operation.
One, temporal lobectomy often is performed on patients with a temporal lobe neoplasm to provide
evidence for diagnosis, to produce decompression, and to decrease the risk of uncal herniation. As much
neocortical tissue as needed is removed to afford maximal decompression. Two, temporal lobectomy
performed in combination with resection of the amygdala and hippocampus is the primary means of
treating intractable partial seizures originating in the temporal lobe. As little neocortex as possible is
removed to gain access to the hippocampus. Subsequently, an aggressive resection of medial temporal
structures is performed.
Surgeons debate about the maximal length of temporal neocortex that can safely be resected, as
measured from the temporal tip. In the left temporal lobe, it should be less than 6.0 cm; in the right
temporal lobe, a 7- to 8-cm length of neocortex can be removed with a minimal risk. For both temporal
lobes, the more posterior the resection, the greater the risk of producing a contralateral homonymous
hemianopia. Resection of a length longer than 9.0 cm results in complete homonymous hemianopia.
With resection of a 5-cm length of neocortex, the risk of superior quadrantanopia is about 10 to 15
percent. When operating on the left temporal lobe, the primary concern is the risk of language
dysfunction. When performing a standard left temporal lobectomy to resect a glioma, it is prudent to
87
Neurosurgery Books
limit the extent of the superior temporal resection to 3.5 cm. The posterior margin can be angled back to
remove 6 cm of the inferior temporal gyrus. With these cortical margins, the risk of significant
permanent language dysfunction is less than 5 percent.
It is important to consider the vascular structures at risk during temporal lobectomy. Along the
Sylvian fissure, the middle cerebral artery complex can be injured, and along the tentorial notch, the
anterior choroidal and posterior cerebral arteries can be injured. More posteriorly, adjacent to the edge of
the tentorium cerebelli, the posterior cerebral artery can be harmed. The best means of preventing
vascular injury is to respect the overlying arachnoid along the Sylvian fissure and the basal cistern.
Therefore, it is best to perform a subpial resection of the superior temporal and parahippocampal gyri. A
Penfield #1 dissector or fine spatula can be used to peel the superior temporal gyrus off the Sylvian
fissure. One or two small arteries probably will need to be cauterized and divided. Otherwise, any
bleeding or oozing originating from the arachnoid of the Sylvian fissure should be controlled with
hemostatic fabric (Surgicel). It is important never to cauterize along the Sylvian fissure because it can
cause dissection or thrombosis of the middle cerebral artery. Also, it is important to peel the uncus out of
the tentorial notch without violating the arachnoid. This not only protects the underlying anterior
choroidal and posterior cerebral arteries but also prevents blood from seeping into the basal cistern.
The cranial nerves at risk during temporal lobectomy include the oculomotor, the trochlear, and the
trigeminal nerves. The oculomotor nerve may be injured if the arachnoid over the tentorial notch is
violated. This nerve is exquisitely vulnerable to manipulation. In fact, irrigation of the operative bed
with cold saline may result in the patient temporarily experiencing blurred vision. The trochlear nerve
may be injured if the edge of the tentorium cerebelli is manipulated. The trigeminal nerve may be
88
Neurosurgery Books
injured if the dura mater along the middle cranial fossa is excessively cauterized. Often during temporal
lobectomy, bleeding occurs from veins and sinuses along the dura mater. Bleeding from these small
vessels within the dura mater of the middle cranial fossa is best controlled with bits of absorbable gelatin
sponge (Gelfoam) instead of cauterization. In fact, cauterization may lead to retraction of the dura mater
along these venous channels and make hemostasis more difficult.
The operation described below is performed for intractable right temporal lobe seizures. Therefore,
the anterior neocortex of the temporal lobe is removed to provide access to the hippocampus. It often is
easier to perform this resection in two steps instead of performing an en bloc resection.
Figure 2-2.
Step 1. Diagram of the position of the patient’s head,
the incision, and the craniotomy. The position of the head
that is advocated is not that of a standard temporal
lobectomy, in which the head is parallel to the floor.
Instead the head is angled, as illustrated here, to allow the
surgeon to look down the long axis of the hippocampus. In
this way, the hippocampus can be resected back to the level
of the tectal plate or atrium. It is important to have the
craniotomy extend low down along the middle cranial
fossa to facilitate resection of the inferior temporal and
fusiform gyri.

89
Neurosurgery Books
Figure 2-3.
Step 2. After the craniotomy, the dura mater is tacked to the margins of the bone and muscle and
then opened so that only the cerebral cortex destined for removal is exposed. In this example,
intraoperative electrocorticography will be performed. Accordingly, part of the inferior frontal lobe
is exposed to facilitate placement of the strip electrode
90
Neurosurgery Books
 Figure 2-4.
Step 3. Before the neocortex is resected, it is
best to examine the surface anatomy. In this
illustration, a large anterior vein runs from the
lateral fissure to the inferior surface of the
temporal lobe. A major vein such as this may be
an anterior vein of Labbé, which must be
preserved. Also, before resection, it is best to take
a Penfield #1 dissector and determine the exact
length from the tip of the temporal lobe to the cut
margin of the dural opening. In this way, the
exact length of the neocortical resection can be
determined. Surface inspection often reveals one
or two branches of the middle cerebral artery
that emanate from the Sylvian fissure or the
superior temporal sulcus and run posteriorly.
These arteries should be preserved during
neocortical resection, especially in the left
temporal lobe, to decrease the risk of language
dysfunction caused by vascular injury of the
posterior portion of the superior temporal gyrus.
Before commencing the lobectomy, it is useful to
take into account these surface observations and
to outline the margins of the proposed resection
with a bipolar cautery.

91
Neurosurgery Books
 Figure 2-5.
Step 4. Resection of the temporal lobe begins along the posterior margin, extending from the superior
temporal gyrus to the inferior temporal gyrus. A #5 or #7 straight suction tip should be the primary
instrument for removing the neocortex along the junction of the posterior resection line with the incision
of the superior temporal gyrus. In patients with intractable epilepsy, there typically is atrophy of the
temporal lobe, and the middle cerebral artery may be more superficial than expected. Use of the straight
suction tip instead of the bipolar cautery to divide the white matter at this junction limits the risk of
inadvertently injuring a more superficially located middle cerebral artery. After the Sylvian fissure or the
middle cerebral artery has been identified, a cottonoid (Americot) is placed for protection.

MIDDLE CEREBRAL ARTERY ANEURYSM

The most common approach to a middle cerebral artery aneurysm is through a frontotemporal
craniotomy with a trans-Sylvian dissection. Some surgeons advocate exposure of these aneurysms
92
Neurosurgery Books
 Figure 2-6.
Step 5. A, The posterior margin incision is deepened
along the middle temporal gyrus until the lateral ventricle is
B, The resection line is extended inferiorly through the The ventricle is perhaps the most important
identified.
inferior temporal gyrus into the fusiform gyrus. Occasionally,
landmark during temporal lobectomy. After the ventricle has
it can be difficult to distinguish among the marginsbeen
of the
entered, a cottonoid is placed in it to prevent blood from
inferior temporal gyrus, the fusiform (lateral entering the occipital horn of the lateral ventricle.
occipitotemporal) gyrus, and the parahippocampal gyrus.
There are always partial sulci between these gyri that contain
invaginating arachnoid and several small blood vessels. It is
useful to identify the sulci by visualizing the invaginating
arachnoid. The resection of the inferior temporal lobe is
carried through the fusiform gyrus to the collateral sulcus.

93
Neurosurgery Books
Figure 2-7.
Step 6. A and B, After the resection of the posterior margin down to the collateral sulcus has been
completed, the superior temporal gyrus is resected. The arachnoid just lateral to the Sylvian vein is cauterized
and incised. A small suction tip is used to remove the underlying superior temporal gyrus off the Sylvian fissure.
A small jeweler’s forceps can be used to hold the arachnoid of the Sylvian fissure medially to provide
countertraction as the suction tip or Penfield dissector is used to pull the superior temporal gyrus off the Sylvian
fissure. During this process, it is critical to make sure that the arachnoid over the Sylvian fissure is not violated.
After the Sylvian fissure has been identified, cottonoids are placed over it to protect the vascular structures
within it. The incision in the superior temporal gyrus is carried progressively deeper from back to front with
bipolar cautery. A retractor is placed on the temporal lobe to lift it partially out of the wound to allow better
visualization. Resection of this deeper white matter is angled to run over the top of the lateral ventricle toward
the collateral sulcus.
94
Neurosurgery Books
Figure 2-8.
Step 7. A and B, At the depth of the wound along the floor of
the middle cranial fossa, the arachnoid along the inferior
temporal lobe is cauterized and incised. In fact, a large margin of
arachnoid is left intact over the edge of the tentorium cerebelli to
protect the neurovascular structures and to prevent blood from
entering the basal cistern. At this point, the white matter
underlying the fusiform gyrus, which is lateral to the lateral
ventricle, is being incised.
95
Neurosurgery Books
Figure 2-9.
Step 8. A and B, The neocortex has been removed down to the roof of the temporal horn of
the lateral ventricle. This uncapping allows visualization of the hippocampus. The operating
microscope is rotated to allow the surgeon to look down the long axis of the hippocampus. A new
cottonoid is placed in the lateral ventricle to prevent blood from entering the temporal horn. The
posterior margin of the neocortical resection and that along the Sylvian fissure are lined with
hemostatic fabric and cottonoids. A Yasargil or self-retaining retractor is placed along the edge of
the resection. This retractor is placed under tension to retract the remnant of the temporal lobe in
an anterior-to-posterior direction. The temporal lobe tolerates this type of anterior-to-posterior
retraction.

96
Neurosurgery Books
 Figure 2-10.
Step 9. A and B, The self-retaining retractor is
repositioned to allow a better view of the posterior
hippocampus. The roof of the lateral ventricle is
opened anteriorly with a bipolar cautery. The
choroid plexus is held medially with a cottonoid and
a #5 or #7 straight suction tip.

97
Neurosurgery Books
Figure 2-11.
Step 10. A and B, After the roof of the lateral ventricle has been opened anteriorly, the choroidal
fissure can be partially identified. It is almost translucent. Usually, four or five small blood vessels
penetrate the arachnoid and enter the choroidal fissure to supply the medial side of the hippocampus. These
Ammon’s horn vessels must be cauterized and divided. They usually are contained in a thin layer of
arachnoid as they pass through the choroidal fissure. Therefore, there is almost a double layer of triangular
arachnoid. This triangular arachnoid and vessels can be displaced medially. A #5 or #7 straight suction tip
is used to gently rotate or extract the medial edge of the hippocampus from the curved gutter of arachnoid
overlying the posterior cerebral artery. In this way, the arachnoid over the posterior cerebral artery is not
violated.
98
Neurosurgery Books
 Figure 2-12.
Step 11. The body of the hippocampus is
mobilized off the posterior cerebral artery, and
the triangular arachnoid containing Ammon’s
horn vessels is cauterized and displaced
medially. The uncus, which contains the
amygdala, is pulled out of the tentorial notch
with the Penfield #1 dissector and suction tip.
Again, the arachnoid over the tentorial notch is
not violated. This protects the anterior
choroidal and the posterior cerebral arteries.
There typically is a small bit of white matter
between the uncus and the forward-most cut
along the roof of the lateral ventricle that must
be removed with suction and bipolar cautery.

99
Neurosurgery Books
 Figure 2-13.
Step 12. At this point, most of the amygdala
and hippocampus has been mobilized. The
posterior edge of the hippocampus is cauterized
and divided. In this way, the hippocampus is
removed en bloc. After the hippocampus has been
removed, the self-retaining retractor is
repositioned along the roof of the lateral ventricle
and again retracted in an anterior-to-posterior
direction. Typically, a #7 straight suction tip can
be used to remove the tail of the hippocampus
back to the level of the tectal plate. When
removing the tail of the hippocampus, it is
important to continue to identify the choroid
plexus, which marks the lateral ventricle. The
tissue between the lateral ventricle and the
collateral sulcus is removed with suction in a
subpial fashion. Preserving the tissue lateral to
the collateral sulcus decreases the risk of visual
field loss.

100
Neurosurgery Books
 Figure 2-14.
This postresection illustration
highlights some of the pertinent
arachnoid structures. The deepest
arachnoid visualized is that overlying
the posterior cerebral artery and the
oculomotor nerve. As this arachnoid
passes medially, it runs underneath the
choroid plexus and turns laterally in a
triangular pattern. This contains
branches of the posterior cerebral
artery, called “Ammon’s horn vessels,”
that feed the hippocampus. The outer
layer of this triangular arachnoid
continues more superficially and joins
the arachnoid over the Sylvian fissure.

101
Neurosurgery Books
through a subpial superior temporal gyrus approach. In the standard trans-Sylvian approach, it is
important to remember that the aneurysm usually is located more superficially than expected. This is
why a frontotemporal craniotomy is superior to a pterional craniotomy for exposure.
One of the first major decisions the surgeon must make during exposure is whether to start splitting
the Sylvian fissure proximally or distally. The advantage of a proximal exposure is that it provides
access to the supraclinoid carotid artery and the proximal middle cerebral artery before the aneurysm is
dissected. Thus, if inadvertent rupture occurs, a temporary vascular clip can be applied. However, if the
patient has had a significant subarachnoid hemorrhage, perhaps associated with a temporal lobe
hematoma, the brain may be edematous and obtaining exposure of the proximal or deep Sylvian fissure
early may be more difficult than anticipated. The disadvantage of dissecting the lateral Sylvian fissure
early in the procedure is that the surgeon comes down on the dome of the aneurysm before achieving
proximal control of the blood vessel. A useful rule to follow is that if a hemorrhage has occurred, early
proximal splitting of the Sylvian fissure is best (if possible). For aneurysms that have not hemorrhaged,
a lateral-to-medial division of the Sylvian fissure can be performed with minimal retraction of the
frontal or temporal lobe.
Although middle cerebral artery aneurysms appear to be low risk because of easy access, the
reported rate of ischemic stroke complications is high, probably for several reasons. One, many middle
cerebral artery aneurysms have a fusiform origin and the takeoff of the M2, segments may be from the
neck of the aneurysm. Accordingly, placement of the clip may compromise the origin of one of these
vessels, especially if the base of the aneurysm is partially calcified. Two, often, a lenticulostriate artery
originates from the deep side of the middle cerebral artery distally along the M1, segment close to the
102
Neurosurgery Books
bifurcation. This vessel may be occluded inadvertently by the tips of the aneurysm clip. Thus, after the
aneurysm clip is placed, it is important to visually inspect the tips of the aneurysm clip and to auscultate
the M2 branches with a micro Doppler probe or ultrasonography, if available.

103
Neurosurgery Books
 Figure 2-15.
Step 1. In this example, the Sylvian fissure
is widely divided. It often is helpful to use a
temporal lobe retractor. The retractor should
be placed along the medial temporal lobe so
that manipulation of the clip and the clip
holder is not obstructed if the trajectory of clip
placement is horizontal from lateral to medial.
This illustration emphasizes the presence of a
trifurcation of the middle cerebral artery, with
a small branch exiting deep to the dome of the
aneurysm. Also, a single lenticulostriate trunk,
with its arcade, comes off the medial wall of
the M1 segment.

Neurosurgery Books
 Figure 2-16.
Step 2. A, After the Sylvian fissure has been
divided, the neck of the aneurysm must be separated
from the M2 branches.
B, Often, part of the neck or dome is adherent to the
takeoff of the M2 branch. A fine spatula is useful in the
dissection. After the neck has been identified, placement
A of a small piece of absorbable gelatin sponge helps
protect the takeoff of the M2 branches. In this
illustration, a curved clip has been applied. This is one
of the more common trajectories for clip placement for
these aneurysms.
B
C, After the clip has been placed, it is important to
aspirate the dome to collapse the aneurysm. Thereafter,
the jaws of the clip are inspected to make sure that no
vessels are trapped along the back wall. Finally. the M2
branches should be assessed immediately after clip
placement with a micro Doppler probe or
ultrasonography.

Neurosurgery Books
 A B C

Figure 2-17.
A-C, This sequence of drawings illustrates a giant, partially thrombosed aneurysm of the middle cerebral artery.
In this situation, partial thrombectomy of the aneurysm is almost always required. Therefore, temporary clips are
placed on the distal M1 segment just proximal to the bifurcation. If possible, the temporary clip should be placed in a
way to preserve blood flow to the lenticulostriate arteries that originate from the M1 segment. Note that a temporary
clip has been placed on one branch of the middle cerebral artery just proximal to a bifurcation to allow retrograde
collateral blood flow (curved arrow). Generally, before the vessels are occluded in this type of aneurysm repair, a
metabolic suppressant is administered intravenously, such as thiopental (2 to 3 mg/kg). Also, the patient’s systolic
blood pressure is increased to 130 to 150 mm Hg to increase leptomeningeal collateral blood flow. Temporary
occlusion without reperfusian for less than 10 minutes is safe, with low risk of residual ischemic injury. If major
vessels are occluded for more than 20 minutes, the risk of ischemic injury increases, unless there is good collateral
blood flow. Currently, the effects of intermittent reperfusion in this type of surgical scenario have not been determined.

Neurosurgery Books
A

Figure 2-18.
A, A technically difficult middle
cerebral artery aneurysm that has dense
adhesions with several branches of the
B middle cerebral artery.
B, After the vessels have been dissected
free from the neck of the aneurysm, it is
useful to place a small piece of absorbable
gelatin sponge when placing the clip.
C, The sponge protects the vessel at risk
when the clip is inserted. With complex
aneurysms in which the neck is the origin of
distal vessels, it is mandatory to have
C objective evidence that the vessels are
patent after clip placement. This evidence
can be obtained with Doppler,
microultrasonography, or angiography.

Neurosurgery Books
TEMPORAL LOBE ARTERIOVENOUS MALFORMATION

When considering surgical resection of an arteriovenous malformation, the surgeon must have a
detailed understanding of the three-dimensional anatomy of the lesion, the feeding arteries, and the
draining veins. After these relationships are understood, an operation can be devised to minimize the risk
of resection. In temporal lobe arteriovenous malformations, the arterial blood supply is derived
superficially from the middle cerebral artery along the Sylvian fissure and inferiorly from the posterior
cerebral artery and associated branches. The deep blood supply is complicated and derived from
branches of the posterior cerebral artery, the anterior choroidal artery, and the middle cerebral artery. The
venous drainage is through the Sylvian veins, the parietal veins, and the posterior temporal veins,
including the vein of Labbé.
The general approach to a temporal lobe arteriovenous malformation is first to isolate the superficial
arterial supply from the Sylvian fissure. Next, the blood supply along the inferior and posterior margins
is cauterized and divided. The arteriovenous malformation is then retracted to expose the many arterial
feeding vessels along thetemporal horn of the lateral ventricle. During the resection, every attempt is
made to preserve all the draining veins until after all the arterial blood supply has been isolated and
divided. One of the best intraoperative indications that the resection of the arteriovenous malformation is
complete is a loss of arterialized venous blood.
General principles of surgery on arteriovenous malformations include the following: one,
preoperative embolization is beneficial and should be pursued aggressively if an experienced
endovascular interventionalist is available. Embolization significantly decreases the severity of
108
Neurosurgery Books
intraoperative bleeding. Also, it allows a graded accommodation of the surrounding brain tissue and
blood vessels to changes in blood flow, thereby decreasing the risk of a postoperative hyperperfusion
syndrome and hemorrhage. If embolization is performed, it is important that surgery not be delayed too
long; otherwise, the deeper and more inaccessible blood supply to the arteriovenous malformation
increases and is more difficult to control. Two, it is beneficial to tightly regulate the patient’s blood
pressure for approximately 48 hours postoperatively, depending on the size of the lesion. The larger the
lesion, the greater the risk of a postoperative hyperperfusion syndrome. Tight regulation of blood
pressure decreases this risk. With large arteriovenous malformations, the author prefers to keep systolic
blood pressure less than 100 mm Hg for 48 hours postoperatively. Three, during resection of an
arteriovenous malformation, it is important to preserve all surrounding veins. Veins that have arterialized
blood may drain both the malformation and normal brain, and the inadvertent sacrifice of these veins
can result in hemorrhagic infarction postoperatively. Four, irrigating bipolar cautery and small wire
microclips are most helpful in obliterating small arterial feeding vessels along the ventricular surface.
Five, postoperative imaging is needed to confirm complete resection of the arteriovenous malformation
unless the surgeon is absolutely certain that this has been accomplished. If a small residual of
arteriovenous malformation is left untreated, the risk of future hemorrhage is significant.

109
Neurosurgery Books
 Figure 2-19.
Step 1. Both the craniotomy and
the opening of the dura mater must be
performed with caution to avoid any
inadvertent tearing of the arteriovenous
malformation, especially venous varices
that may have eroded partially through
the dura mater. After the dura mater has
been opened, the first step is to examine
the arteriovenous malformation to
understand its surface anatomy. It is
important to identify the draining veins,
including those that drain both the
malformation and normal brain. It often
can be difficult to distinguish between
an arterialized vein and a feeding
artery. A useful tip is to follow the
vessel in question distally along the
cortical surface. Often, the smaller
branches or ramifications assume the
characteristics associated with a
thicker walled arteriole or a thin-walled
venule.

Neurosurgery Books
 Figure 2-20.
Step 2. In this example, a large
partially arterialized Sylvian vein
makes it difficult for the surgeon to
enter the Sylvian fissure. Therefore, the
arachnoid over the Sylvian fissure is
divided and the vein is retracted
laterally. Many small arteries usually
run out of the Sylvian fissure into the
substance of the arteriovenous
malformation. Each of these should be
cauterized and divided.

Neurosurgery Books
 Figure 2-21.
Step 3. As the division of the
Sylvian fissure is extended, some
larger branches of the M1 and
M2 segments become apparent.
Some of these clearly may enter
the arteriovenous malformation;
they can be cauterized. However,
for other vessels, it may not be
clear that they contribute to the
malformation; in fact, they may
be vessels “en passage.” Vessels
that may not feed the
arteriovenous malformation can
be occluded temporarily with
small microclips until their
anatomy has been defined more
completely. A useful visual clue
for identifying feeding vessels of
an arteriovenous malformation is
that their origin from the middle
or posterior cerebral artery often
has a corkscrew shape.

Vessel “en passage”

Middle cerebral artery
“Corkscrew” feeding vessel
Neurosurgery Books
 Figure 2-22.
Step 4. After potential feeding
vessels from the Sylvian fissure have
been cauterized or temporarily
clipped, an incision is made in the
cerebral cortex just posterior to the
arteriovenous malformation. In the
example shown here, the
arteriovenous malformation has two
obvious draining veins. The
posterior draining vein has been
cauterized and divided to facilitate
rotation of the arteriovenous
malformation.

Neurosurgery Books
 Figure 2-23.
Step 5. A retractor is placed gently on
the arteriovenous malformation as the
dissection is carried deeper. There are
always several branches from the posterior
temporal artery, a branch of the posterior
cerebral artery. These branches run
underneath the temporal lobe and enter the
arteriovenous malformation laterally, along
the floor of the middle cranial fossa. Also,
enlarged Ammon’s horn vessels usually feed
the arteriovenous malformation. They
should be cauterized and divided. To protect
the posterior cerebral artery, every attempt
should be made to preserve the arachnoid
overlying the artery.

Neurosurgery Books
 Figure 2-24.
Step 6. The apex of most
arteriovenous malformations, regardless
of location, is periventricular. Small
arteries along the ependymal surface of
the ventricle always feed the arteriovenous
malformation. These arteries are
extremely thin walled and, thus,
exceedingly difficult to cauterize.
Patience, an irrigating bipolar, and
microclips are the best tools.

Neurosurgery Books
 Figure 2-25.
Step 7. After ensuring that all the
arterial blood supply to the arteriovenous
malformation has been obliterated, the
last draining veins can be safely
cauterized and divided. After the
arteriovenous malformation has been
resected, it is important to inspect
visually, with the operating microscope,
all the surface veins to make sure that
there is no residual arterialized blood.
The presence of arterialized venous blood
indicates that some of the malformation
was not resected.

Neurosurgery Books
Chapter 3
Frontal Approach

Neurosurgery Books
Chapter 3: Frontal Approach

ANATOMIC CONSIDERATIONS

The lateral surface of the frontal lobe has three serpentine gyri: the superior, middle, and inferior
frontal gyri (Figure 3-1). The superior frontal gyrus (Brodmann areas 6, 8, and 9) extends from the
frontal pole to the precentral gyrus, a distance of 10 to 11 cm. Anteriorly, it merges with the middle
frontal, orbital, and rectus gyri. Approximately 8 to 10 transverse gyri interdigitate with the middle
frontal gyrus. The medial surface of the superior frontal gyrus is also called the “medial frontal
gyrus” (Brodmann areas 6, 8, 9, 10, 11, 12, and 32). It is adjacent to the parolfactory sulcus. It extends
ventrally toward the frontal pole and curves posteriorly around the cingulate gyrus. The cingulate sulcus
separates the medial frontal gyrus from the cingulate gyrus. There usually are several intertwining
transverse gyri between the medial frontal and cingulate gyri.
The middle frontal gyrus (Brodmann areas 6, 8, 9, and 10) is inferior to and parallel with the
superior frontal gyrus and extends from the frontal pole to the precentral gyrus. Several transverse sulci
divide this gyrus.
The inferior frontal gyrus (Brodmann areas 44, 45, and 47) is just above the Sylvian fissure,
overlying the antenor half of the insula. It forms the “W-shaped” frontal operculum, which is divided
into three segments (pars orbltahs, pars tnangularis. and pars opercularis) by ascending rami of the
118
Neurosurgery Books
Sylvian fissure and the inferior precentral sulcus. The pars orbitalis (Brodmann area 47) fuses superiorly
with the anterior segment of the middle frontal gyrus and anteriorly with the lateral orbital gyrus. The
pars triangularis (Brodmann area 45) is demarcated by the ascending rami of the Sylvian fissure and
merges anteriorly with the orbital gyri and posteriorly with the pars opercularis. The pars opercularis
(Brodmann area 44) lies between the pars triangularis anteriorly and the inferior precentral and inferior
frontal sulci posteriorly.
Anatomically, the pars opercularis corresponds to Broca’s area (Brodmann area 44). However,
intraoperative cortical stimulation indicates that the pars triangularis (Brodmann area 45) should be
included within Broca’s area. This cortical region has many convolutions with tremendous variability.
Also, deep within the sulci are transverse gyri that are not visible with inspection of the surface.
The precentral gyrus is perpendicular to the superior, middle, and inferior frontal gyri and
incompletely separated from them by the precentral sulcus. The posterior boundary of the gyrus is the
central sulcus. The precentral gyrus extends from the cingulate sulcus on the medial surface of the
hemisphere (where it comprises the anterior portion of the paracentral lobule) to the Sylvian fissure on
the lateral surface. A useful tip for identifying the precentral gyrus on magnetic resonance images is to
follow the superior frontal sulcus, which terminates posteriorly in two ramifications that form a “T.”
This “T” (part of the precentral sulcus) abuts the precentral gyrus.
The precentral gyrus is the primary motor cortex (Brodmann area 4), which traditionally is
considered the cortical area for voluntary movement. As shown with electrical stimulation of the cortical
surface, the muscles of the body are represented systematically (but disproportionately) along the gyrus.
This topographic organization is epitomized in the so-called motor homunculus of Penfield. Muscles of
119
Neurosurgery Books
the throat (including the pharyngeal muscles) are represented in the most inferior part of the gyrus: the
part abutting the Sylvian fissure. Proceeding superiorly are the representations of the muscles of the
face, thumb, fingers, hand, distal arm, proximal arm, and shoulder. The muscles of the trunk and hip are
represented near the superior edge of the hemisphere. where the gyrus continues onto the medial surface
of the hemisphere. The representations of the muscles of the leg, feet, and toes are on this medial surface
of the gyrus, as are the representations for the muscles of the bowel and bladder.
Immediately anterior to the primary motor cortex is the premotor cortical area (Brodmann area 6).
The portion on the lateral surface of the hemisphere is called the “premotor cortex” and that on the
medial surface is the “supplementary motor cortex.” Electrical stimulation of the primary motor cortex
usually elicits contraction of only one or a few muscles on the opposite side of the body. In contrast,
electrical stimulation of the premotor areas produces a more complicated reaction that may consist of the
contraction of several muscles on both sides of the body, causing movement of several joints.
The blood supply to the medial frontal lobe is from branches of the anterior cerebral artery (Figure
3-2). These branches include the orbitofrontal, frontopolar, anterior internal frontal, middle internal
frontal, and posterior internal frontal arteries. The blood supply to the lateral frontal lobe is from
branches of the middle cerebral artery, including the orbitofrontal, operculofrontal, and central sulcus
arteries.

120
Neurosurgery Books
 Figure 3-2 Precentral Gyrus

Superior frontal sulcus Central sulcus

Superior frontal gyrus Postcentral sulcus

middle frontal gyrus Supramarginal gyrus

Angular gyrus

Orbital gyri

Pars orbitalis
Inferior
frontal Pars triangularis
gyrus
Pars opercularis

Superior frontal Sulcus

Superior frontal Sulcus

Superior frontal Sulcus

Neurosurgery Books
 Paricallosal artery
Figure 3-2
Middle internal frontal Cingulate sulcus
Posterior inferior frontal

Paracental artery
Anterior internal frontal artery Callosal Sulcus

Callosomarginal artery Precuneal artery

Parietoocipital artery

Frontopolar artery

Orbitofrontal artery

Recurrent artery of Heubner

Internal carotid artery Calcarine artery

Posterior communicating artery Posterior temporal artery

Posterior cerebral Anterior Temporal artery

Basilar Artery

Neurosurgery Books
 FRONTAL LOBECTOMY

Figure 3-3
Step 1. The patient’s head is held in a donut or fixed in a pinion, as illustrated here. The head should be slightly
flexed and higher than the heart to promote venous drainage. To prevent partial kinking of the internal jugular
veins, it is important not to overflex the head. The incision is a standard bicoronal incision. The medial edge of the
bone nap should be along the superior sagittal sinus. This provides easier access to the interhemispheric fissure,
with better visualization of the anterior cerebral artery and its branches. The posterior margin of the bone flap is at
the coronal suture. The lateral margin extends from the junction of the coronal suture with the superior temporal
line to the “keyhole.” Neurosurgery Books
Figure 3-4.
Step 2. After the bone flap has been removed, the dura mater is tacked to the margins of the bone. Ensure that
tack holes along the anterior or ventral edge of the bone do not enter the frontal sinus, because if they do, it may
be a source of rhinorrhea. Occasionally, the frontal sinus is entered along the medial aspect of the craniotomy. If
this occurs, the mucosa should be stripped with a small punch. A piece of temporalis muscle is harvested and used
to plug the ostium of the sinus. Thereafter, a small galeal-periosteal flap is rotated with a #11 blade knife and
tacked to the adjacent dura mater. This effectively obliterates the sinus.
Neurosurgery Books
Figures 3-5 and 3-6.
Step 3. The margin of the frontal lobectomy is outlined using bipolar cautery. Some of the brain
regions at risk during frontal lobectomy include the precentral gyrus, the frontal operculum, and the
deeper basal ganglia. If the operation is performed without preoperative or intraoperative mapping of
eloquent brain regions, the following guidelines are useful for preventing neurologic dysfunction
postoperatively. The posterior resection line should be approximately 1.0 cm anterior to the coronal
suture. This line should project lateral to the sphenoid wing and medial to the olfactory bulb. In a left
frontal lobectomy, this will spare the frontal operculum by a safe margin.

Neurosurgery Books
Figure 3-7.
Step 4. The incision is deepened with the lise of bipolar cautery and suction. If the operation is performed
for treatment of frontal seizures only, there is no need to extend the resection into the deep white matter. After
cutting through the cerebral cortex, the surgeon’s trajectory can then angle anteriorly to the olfactory bulb. If the
operation is performed for tumor debulking for relief of mass effect, it is necessary to extend the resection deeper
toward the tip of the frontal horn of the lateral ventricle. Laterally, along the sphenoid wing, it is important to
respect the pial barrier, which in turn will prevent inadvertent injury to the middle cerebral artery in the Sylvian
fissure.

Neurosurgery Books
 Figure 3-8.
Step 5. At risk medially are the pericallosal and
callosomarginal arteries. As in most lobectomies, the
best way to prevent vascular injury is to use subpial
resection. Specifically, a #7 suction tip is used to remove
the cerebral cortex beneath the pia mater. After the pia
mater is devoid of blood vessels, it can be cauterized
and cut.

Neurosurgery Books
Figure 3-9.
Step 6. In this example. the frontal horn of
the lateral ventricle has been entered. A
cottonball is used to plug the ventricle while the
resection is completed. This helps decrease the
amount of blood that seeps into the lateral
ventricle. To decrease the risk of postoperative
aseptic meningitis, it is important to irrigate the
ventricle to flush out blood products.

Neurosurgery Books
PERICALLOSAL ARTERY ANEURYSM

Collosomarginal artery

Pericollosal artery

Figure 3-10. Most pericallosal aneurysms

originate at the bifurcation of the
callosomarginal and pericallosal arteries. The
more frontal the craniotomy, the more likely the
surgeon will be to come down upon the neck
instead of the dome of the aneurysm.

Neurosurgery Books
 Figure 3-11.
Step 1. As illustrated here, two
different types of bone flaps can be used to
gain access to the aneurysm. Both of these
flaps can also be used to enter the lateral
ventricle through an interhemispheric
approach and to section the anterior
corpus callosum for treatment of certain
types of epilepsy. The patient’s head is
supine and straight up to aid in surgical
orientation. The medial edge of the bone
flap should be on the superior sagittal
sinus; this provides greater access with
less retraction of the medial frontal lobe.
In patients with pericallosal aneurysms,
the venous phase of the angiogram should
be studied to determine whether there are
major draining veins. If there are, the
bone flap can be planned accordingly.

Neurosurgery Books
Figure 3-12. Step 2. B, Every attempt needs be made to
Step 2. A, The dura mater has been opened and preserve the vein by separating the hemisphere
tacked to the margins of the bone. In this example, a anterior to it. Leaving the enveloping arachnoid
rather large vein crosses the surgical field. intact helps to preserve these bridging veins. In
addition, a piece of absorbable gelatin sponge
(Gelfoam) can be placed on the vein to act as a
buttress. If the hemisphere is swollen because of
significant hemorrhage, brain relaxation can be
achieved with lumbar or ventricular drainage of
cerebrospinal fluid or with mannitol. The square
bone flap provides easier access should a
ventricular tap be required.
Neurosurgery Books
 Figure 3-13.
Step 3. A, Most pericallosal aneurysms
can be approached from the right, thereby
eliminating retraction on the medial left
(dominant) frontal lobe. The right
hemisphere is lined with hemostatic fabric
(Surgicel) and cottonoids (Americot) and
gently retracted. The thin bridging
arachnoid between the two hemispheres
can easily be divided with a dissector or a
bipolar forceps. As true for most aneurysm
surgery, it is best to first identify the
proximal parent artery.
B, Placement of a small piece of
absorbable gelatin sponge around the
neck of the aneurysm will facilitate
placement of the clip. C, After the
aneurysm has been clipped, the dome is
aspirated and the jaws of the clip
reinspected.

Neurosurgery Books
THIRD VENTRICULARCOLLOID CYST - INTERHEMISPHERIC APPOROACH

Figure 3-14.
A standard coronal section through
the brain at the level of the foramen
magnum highlighting the relationship of
the fornix and choroid plexus with the
foramen of Monro.

Neurosurgery Books
Figure 3-15.
This axial section depicts the
relationship of the coronal suture to the
lateral ventricle and the foramen of Monro.

Neurosurgery Books
 Anterior septal vein

Anterior caudate

Posterior caudate vein

Thalamostriate vein

Superior choroidal vein

Figure 3-16.
This figure emphasizes the pertinent
neurovascular anatomy of the lateral ventricle. The
Interior cerebral vein first landmark the surgeon visualizes on entering the
lateral ventricle is the choroid plexus, which
Posterior septal vein overlies the choroidal fissure. Typically, a small vein
called the “superior choroidal vein” runs on top of
Basal vein of
the choroid plexus. This choroid plexus sits between
the body of the fornix and the thalamus. Lateral to
Vein of Galen
the choroid plexus is the convex surface (“bulge”)
of the thalamus, which forms the floor of the lateral
ventricle. The next structure observed is the
thalamostriate vein, which has two major
tributaries: the anterior caudate and the posterior
caudate veins. The thalamostriate vein is the largest
contributor to the internal cerebral vein.
Neurosurgery Books
Figure 3-16. (cont.)
The thalamostriate vein runs in the groove between the caudate nucleus and the thalamus, above
the stria terminalis. At the foramen of Monro, the thalamostriate vein curves around the anterior
tubercle of the thalamus and joins the internal cerebral vein. In the frontal horn of the lateral ventricle,
the anterior septal vein runs along the septum pellucidum and crosses the fornix to join the internal
cerebral vein. A small posterior septal vein may be seen posteriorly in the body of the lateral ventricle.
The posterior septal vein also runs along the septum pellucidum and goes around the fornix to enter the
midportion of the internal cerebral vein. This figure does not show the small veins in the atrium of the
ventricle that pass through the choroidal fissure to enter the basal vein of Rosenthal, the internal
cerebral vein, and the vein of Galen.
The paired internal cerebral veins originate at the foramen of Monro and go posteriorly in the
velum interpositum. As they go posteriorly, they curve lip over the superolateral aspect of the pineal
gland and follow the curve of the inferior surface of the splenium of the corpus callosum. The left and
right internal cerebral veins join at the posteroinferior surface of the splenium to form the vein of
Galen.
The basal vein of Rosenthal begins at the level of the anterior perforated substance and runs
posteriorly between the midbrain and the temporal lobe. It drains blood from the medial temporal lobe
and the midbrain. In the posterior incisural space, it joins the internal cerebral veins and vein of Galen.
In the midline is the septum pellucidum, which may contain a space called the “cavum septi
pellucidi.” Deeper (along the inferior edge of the septum pellucidum) is the fornix, which consists of
axons that originate in the hippocampal formation, extend around the thalamus, and end in the
mammillary bodies in the floor of the third ventricle. The fornix is divided into four parts. Near its
origin in the hippocampal formation, the axons form the fimbria. The fimbria passes posteriorly to form
the crus (plural, crura) of the fornix. Each crus goes anteriorly along the posterior surface of the
pulvinar. At the level of the atrium and body of the lateral ventricle, the two crura join and form the
body of the fornix, which continues anteriorly along the medial wall of the lateral ventricle. At the level

Neurosurgery Books
Figure 3-16. (cont.)
of the foramen of Monro, the body of the fornix splits into two columns, each of which forms the
anterior margin of the foramen of Monro. Each column enters the substance of the brain and goes
posteriorly through the hypothalamus to end in the ipsilateral mammillary body.
By using a midline approach and dividing the septum pellucidum, the surgeon will encounter the
fornix, a thin layer of tela choroidea, the internal cerebral veins, and then another layer of tela
choroidea.
When the surgeon enters the lateral ventricle, the junction of the anterior septal vein, the
thalamostriate vein, and the choroid plexus point to the foramen of Monro. Occasionally, the surgeon
may be confused about which ventricle has been entered. The pulsating septum pellucidum can be used
as the medial landmark for orientation.

Neurosurgery Books
 Figure 3-18
Step 2. Two basic craniotomies, square (A) and
triangular (B), can be used. The square craniotomy
provides slightly more room for retraction. However,
either craniotomy is excellent for an
interhemispheric approach. If angiography was
performed preoperatively, the bridging veins
should be identified on the angiogram and
the bone flap positioned accordingly.

Figure 3-17.
Step 1. The patient’s head is
fixed in a pinion. Note that the
head is at 0 degrees for
orientation and slightly flexed and
elevated above the level of the
heart to decrease intracranial
pressure.

Neurosurgery Books
Figure 3-19.
Step 3. The dura mater is opened,
reflected over the superior sagittal sinus, and
tacked to the margins of the bone. After the
right frontal lobe has been protected with
hemostatic fabric and cottonoids, it is gently
retracted. The arachnoid over any bridging
veins should be preserved to reinforce these
vascular structures.

Neurosurgery Books
Figure 3-20.
Step 4. The author prefers to approach third ventricular
colloid cysts from the right lateral ventricle instead of using an
interforniceal approach. In this way. there is less risk of injury to
the fornix. The corpus callosum is identified by its whitish
appearance. The incision in the corpus callosum is approximately
1.5 cm long.

Neurosurgery Books
Figure 3-21.
Step 5. The primary means of removing a colloid cyst is decompression with suction.

Neurosurgery Books
 Figure 3-22.
Step 6. After the lateral ventricle has
been entered, it is mandatory that the
surgeon find the pertinent structures, which
include the choroid plexus, the
thalamostriate vein, the anterior septal vein,
and the anterior caudate vein. Although the
bulge of the colloid cyst will be apparent, it
is helpful to identify these structures to
become oriented and to identify the fornix.
Afterward, an incision is made in the colloid
cyst bulging through the foramen of Monro

Neurosurgery Books
Figure 3-23.
Step 7. After the colloid cyst has
been incised, a #5 or #7 suction tip is
inserted into the cyst and the contents
are evacuated. This usually provides
immediate decompression of the cyst
wall. The cyst frequently is attached to
the posterior aspect of the foramen of
Monro, anterior to the internal
cerebral vein. It usually is necessary
to gently cauterize and then to cut this
attachment before extirpating the wall.

Neurosurgery Books
Figure 3-24.
An alternative to an interhemispheric approach is the transcortical route through the middle frontal
gyrus. This can be achieved through a small trephine type craniotomy that is 3 cm from the midline and 2 cm
in front of the coronal suture. This is the same location for placing an external ventricular drain into the
right frontal horn of the lateral ventricle.

Neurosurgery Books
Figure 3-25.
A transcortical approach provides excellent access to the colloid cyst because of the angle.
However, the disadvantage of this approach is the cortical incision that must be made and the
small risk of postoperative seizures.

Neurosurgery Books
CRANIOPHARYNGIOMA—SUBFRONTAL APPROACH

Figure 3-26.
A subfrontal approach for resection of a
craniopharyngioma is useful, especially when
there is a large extension of the tumor into the
third ventricle. In comparison with a standard
pterional craniotomy (described in Chapter
1), the subfrontal approach, with removal of
the orbital rim, provides a better angle for
access to the rostral portion of the tumor.
Although removing the orbital ridge increases
the surgical time for performing the
craniotomy, it is worthwhile.

Neurosurgery Books
 Figure 3-27.
Step 1. The patient’s head is extended to
decrease the need for frontal lobe retraction.
Despite the head being extended, the chest is still
angled to promote venous drainage from the

Neurosurgery Books
 Figure 3-28.
Step 2. There are three bone flaps that can be
used in a subfrontal approach to the third ventricle
A, The standard frontal craniotomy is easiest
to perform.
B, Removing the orbital ridge unilaterally
greatly decreases the need for frontal brain
retraction.
C, In massive tumors that have lateral
extensions, removal of both orbital ridges may be
advantageous. In this case, both frontal lobes are
retracted to provide excellent exposure for not only
the rostral but also the lateral extensions of the
neoplasm. However, this approach has several
disadvantages, including the time required to
remove both orbital ridges, the need to deal
completely with the frontal sinus, and the likely loss
of both olfactory tracts even if they are carefully
dissected off the frontal lobes. All three
craniotomies can be achieved through a bicoronal
incision.

Neurosurgery Books
 Figure 3-29.
Step 3. The craniotomy is achieved
through two bone flaps. First, a standard
frontal bone flap is performed, with its
medial margin along the superior sagittal
sinus. The outlines of the orbital removal
are indicated by the dashed line

Neurosurgery Books
 Figure 3-30.
Step 4. A, It is important to preserve the
supraorbital nerve. Often, the supraorbital
foramen is a notch, and a small spatula can be
used to dissect the nerve out of its foramen and
preserve it with the skin flap. Occasionally, the
inferior rim of the foramen must be removed with
a small rongeur. The dura mater, the temporalis
muscle, and the orbital fascia are dissected off
the bone with a curved periosteal elevator and a
Penfield #1 dissector.
B, Both the medial and lateral bone cuts are
made with an oscillating saw. While the bone
cuts are being made, it is important to protect the
brain and the orbit with spatulas. A sterilized
tablespoon is ideal for protecting the orbit. After
the two cuts have been completed, an orbital
rongeur is used to grab the orbital ridge and to
rock it gently. The posterior bone will crack
along lines extending to the superior orbital
fissure. Any shelves of bone along the inferior
surface of the frontal lobe can then be removed
with the rongeur

Neurosurgery Books
 Figure 3-31.
Step 5. Next, the dura mater is opened and
folded over the orbit and retracted under tension
with fishhooks

Neurosurgery Books
 Figure 3-32.
Step 6. The right frontal lobe is lined with
hemostatic fabric and cottonoids and gently
retracted. The amount of retraction needed is
usually quite small, because of removal of the
orbital ridge. If more brain retraction is
necessary for large tumors, the cisterns that
surround the optic nerves and internal carotid
arteries can be incised to promote drainage of
the cerebrospinal fluid. With this approach, the
lamina terminalis is incised. The lamina
terminalis usually is extremely thin and merges
almost imperceptibly with the capsule of the
tumor. After the lamina terminal is has been
incised and the tumor capsule has been
entered, the tumor is decompressed internally
and debulked with ringed curets. When
incising the lamina terminalis, it is best to
leave 3 to 4 mm of tissue behind the visually
identified posterior margin of the optic chiasm.

Neurosurgery Books
 Figure 3-33.
Step 7. After the tumor has been
debulked, the capsule is manipulated
with a cottonoid and a #5 or #7 suction
tip. Fine spatulas or bipolar cautery
forceps are used to peel the capsule
gently off the optic tracts and optic
chiasm. It is important to recognize that
deep to the tumor is a layer of arachnoid
that overlies the basilar artery and its
performing branches. Laterally, this
arachnoid overlies the posterior
communicating artery. Preserving this
arachnoid will protect these structures.

Neurosurgery Books
 Figure 3-34.
Step 8. After the tumor has been
removed, the margins are inspected with
small angled Applebaum mirrors or a
fiberoptic scope. Do not cauterize any
bleeding from the walls of the third
ventricle, which at this point are the walls
of the hypothalamus. Instead, use gentle
irrigation and small pieces of hemostatic
fabric to achieve hemostasis. After the
resection has been completed, the
retractors are withdrawn and the frontal
lobe is inspected. The dura mater is closed
primarily or with periosteum. The bone flap
is repositioned with small metal plates.

Neurosurgery Books
OLFACTORY GROOVE–PLANUM SPHENOIDALE MENINGIOMA

Meningiomas that lie in the olfactory groove and extend posteriorly along the planum sphenoidale
are resected through a subfrontal approach. A more standard pterional or frontotemporal craniotomy can
be used for small tumors and those with a unilateral extension along one orbital roof. If this type of
craniotomy is used, it is important to make sure that the bone flap extends to the inner canthus of the eye
to facilitate the subfrontal trajectory. The advantage of a unilateral subfrontal approach is that only one
cerebral hemisphere is retracted. Also, if the meningioma is posterior along the planum sphenoidale, the
opposite olfactory nerve may be preserved.
For large bilateral tumors, a bifrontal craniotomy is advantageous because it facilitates dissection of
the lateral and posterior borders of the tumor from the adjacent neurovascular structures. Along those
margins, the tumor is intimately related to the optic chiasm, the optic nerves, and branches of the
anterior cerebral artery. Bifrontal craniotomy has several disadvantages. One, it is technically more
difficult and time consuming. Two, the frontal sinus is often violated and must be exonerated. Three, the
superior sagittal sinus must be ligated. Four, the likelihood of the bilateral loss of the olfactory tracts is
high. However, with a large tumor along the planum sphenoidale, it is difficult to preserve these tracts
regardless of the approach.
For lesions at the base of the frontal skull, such as an esthesioneuroblastoma, a combined
otorhinolaryngologic and neurosurgical approach is optimal. With esthesioneuroblastomas, the
cribriform plate is eroded. A low frontal trephine or rhomboid-shaped craniotomy is made through a
bicoronal incision. The craniotomy extends anteriorly to the nasion and laterally to the orbital roofs. The
155
Neurosurgery Books
approach is epidural. Therefore, a periosteal elevator or Penfield #1 dissector is used to dissect the dura
mater off the orbital roofs. Eventually, the dura mater is dissected off the crista galli, which is then
removed with a bone rongeur. The dura mater is sectioned at the level of the olfactory bulbs. These two
rents in the dura mater are oversewn with 4-0 monofilament (Prolene) sutures. The dura mater is
continually stripped back along the planum sphenoidale, almost to the tuberculum sella. Subsequently,
the frontal lobes protected by dura mater are retracted bilaterally with self-retaining retractors. As the
facial surgeon works from below, the neurosurgeon removes the cribriform plate with an orbital rongeur
and small angled Kerrison rongeur. Occasionally, a small chisel is needed. This unroofs the sphenoid
sinus. Additional bone can be removed posteriorly back to the tuberculum sellae. The neurosurgeon can
assist in the dissection of the rostral and caudal extents of the esthesioneuroblastoma sitting in the
posterior sphenoid sinus and laterally in the ethmoid sinuses.

156
Neurosurgery Books
 Figure 3-35.
The relationship of the tumor with the
olfactory tracts, planum sphenoidale, and
neurovascular structures.

Neurosurgery Books
 Figure 3-36.
The vascular structures at risk are the anterior
cerebral artery and its branches, including the
orbitofrontal artery. If a frontopolar artery takes off
relatively low, it may be encountered laterally along the
rostral part of the tumor. The drawing indicates that the
lower the frontal craniotomy, the less need for frontal
brain retraction.

Neurosurgery Books
Figure 3-37.
Step 1. A, The patient’s head is fixed in a pinion, Step 1B, The dura mater is tacked laterally to the
and a bifrontal craniotomy is made through a bone margins and opened low over both frontal lobes.
bicoronal incision. Note that in this example both Penfield #1 or small retractor blades are used to
frontal sinuses have been entered. After the mucosa separate the mesial frontal lobes off the superior
has been stripped from the walls of the sinuses, pieces sagittal sinus and the adjacent falx cerebri. The
of temporalis muscle are used to plug the frontal superior sagittal sinus is then ligated with braided
sinuses. A small vascularized periosteal flap can be 3-0 sutures and severed.
harvested from the skin flap and sewn over the muscle
packing and tacked to the dura mater.

Neurosurgery Books
 Figure 3-38.
Step 2. The frontal lobes are
lined with hemostatic fabric and
cottonoids and gently retracted.
Next, the falx cerebri is cut posterior
to the crista galli.

Neurosurgery Books
Figure 3-39.
Step 3. As the frontal lobes are retracted,
the tumor is encountered. The blood supply to
the tumor runs along the underside of the
tumor, along the olfactory groove and dura
mater of the planum sphenoidale. Therefore, it
is advantageous first to dissect the underside
of the tumor with bipolar cautery and suction.
The tumor then is debulked internally.

Neurosurgery Books
 Figure 3-40.
Step 4. The tumor is progressively debulked
using bipolar cautery, suction, and an ultrasonic
aspirator. Sometimes, dividing the
interhemispheric arachnoid may facilitate
retraction of one frontal lobe without causing
undue traction on the opposite frontal lobe. As
true for most operations on meningiomas, it is
important to respect the arachnoid plane between
the tumor capsule and the cerebral cortex.

Neurosurgery Books
Figure 3-41.
Step 5. The difficult portion of the tumor is its
posterior, or deep, extension along the optic
chiasm. With very large tumors, the bifurcation of
the internal carotid artery is displaced laterally
and the optic chiasm is pushed posteriorly. A layer
of arachnoid along the anterior clinoid process
overlies the optic nerves, the optic chiasm, and the
internal carotid arteries. Because this arachnoid
plane separates the tumor from these
neurovascular structures, it is important to identify
this arachnoid. The tumor usually can be retracted
with a #5 or #7 suction tip and cottonoid, while the
tumor capsule is separated off the arachnoid with a
small dissecting spatula.

Neurosurgery Books
Figure 3-42.
This postoperative view demonstrates
that both olfactory tracts were lost. It is
better to control bleeding from the olfactory
bulb with absorbable gelatin sponge than
with cautery. It is important to cauterize the
dura mater along the planum sphenoidale.

Neurosurgery Books
Chapter 4
Parietal Approach

Neurosurgery Books
Chapter 4: Parietal Approach
ANATOMIC CONSIDERATIONS

Central Lobe

The demarcation between the frontal and parietal lobes is the central sulcus. However. many
anatomists and surgeons prefer to describe a “central lobe” that consists of the precentral, postcentral,
and paracentral gyri. Thus, the central lobe sits between the frontal and parietal lobes. The rationale for a
central lobe is based partly on the integrated sensorimotor function of these gyri.
The precentral gyrus (Brodmann area 4) begins at the middle third of the Sylvian fissure and runs
vertically toward the interhemispheric fissure. Medially, it is continuous with the anterior part of the
paracentral lobule. Near the midportion of the precentral sulcus, the precentral gyrus abuts the middle
frontal gyrus. Posteriorly, the precentral gyrus is separated from the postcentral gyrus by the central
sulcus. Within the central sulcus are several transverse gyri that interdigitate with the postcentral gyrus.
The precentral gyrus is 10 to 12 cm long and 10 to 15 mm wide, and its cerebral cortex is 3.5 to 4.5 mm
thick. This gyrus is considered to be responsible primarily for skilled voluntary movements. Cortical
mapping studies have demonstrated that despite moderate variability there generally is a topographic
representation of movements (“motor map”) along the precentral gyrus. Commencing with the most
inferior part of the precentral gyrus adjacent to the Sylvian fissure and going toward the

166
Neurosurgery Books
interhemispheric fissure, cortical stimulation causes movements of the jaw, lips, upper face, thumb,
fingers, wrists, elbows, and shoulders.
The postcentral gyrus (Brodmann areas 1, 2, and 3) is separated from the precentral gyrus by the
central sulcus. Medially, it extends into the paracentral lobule. Several incomplete sulci partially divide
the postcentral gyrus.
The paracentral gyrus (Brodmann areas 1, 2, 3, 4, and 5 on the medial surface of the hemisphere), or
lobule, is a rectangular region that contains the medial interhemispheric aspect of both the precentral and
postcentral gyri. It is separated anteriorly from the medial frontal gyrus by the paracentral sulcus and
posteriorly from the precuneus by the marginal limb of the cingulate sulcus. The paracentral gyrus often
has several incomplete sulci. Stimulation of the anterior portion of the paracentral gyrus elicits
movements of the lower extremity. This region is also involved with bowel and bladder functions.
Parietal Lobe

The parietal lobe consists of four lobules (Figures 4-1 and 4-2): the precuneus and the superior,
inferior, and middle parietal lobules. The precuneus (Brodmann areas 7 and 31) is located on the medial
surface of the parietal lobe between the cingulate gyrus and the superior parietal lobule. Anteriorly, it is
separated from the paracentral lobule by the marginal limb of the cingulate sulcus, and posteriorly, it is
separated from the cuneus (of the occipital lobe) by the parietooccipital sulcus. Grossly, the precuneus
appears to consist of obliquely running gyri that are continuations of the cingulate gyrus, especially the
isthmus of the cingulate gyrus. On the lateral cortical surface, the precuneus is continuous with the
postcentral gyrus and the superior parietal lobule.

167
Neurosurgery Books
 The superior parietal lobule (Brodmann areas 5 and 7) sits between the interhemispheric fissure and
the interparietal sulcus. Anteriorly, it is associated with the postcentral gyrus and medially and
superiorly, with the precuneus. Injury to the right parietal lobule can lead to disturbances of space
perception, as manifested by dressing apraxia, central extinction, and constructional apraxia.
The inferior parietal lobule (Brodmann areas 39 and 40) is divided into the inferior lobule
(supermarginal or circumflex gyrus) and the middle lobule (angular gyrus). They are demarcated by the
distal Sylvian fissure, the superior temporal sulcus, and the interparietal sulcus. The middle parietal
lobule is posterior to the inferior lobule and merges with the occipital lobe posteriorly. The arterial
supply to the convexity of the central and parietal lobes is from the middle cerebral artery, and the
vessels supplying the medial surface and the apical portion of the convexity are primarily from the
anterior cerebral artery.
Vascular Supply

The middle cerebral artery gives off five to eight branches along its insular course. The Sylvian
triangle is defined angiographically by the following landmarks. The Sylvian point or apex of the
Sylvian triangle is established by the most posterior branch of the middle cerebral artery as it exits from
the Sylvian fissure. The superior margin is demarcated by the branches of the superior ramifications of
the middle cerebral artery. The inferior margin of the Sylvian triangle is outlined by the inferior loops of
the middle cerebral artery.
The middle cerebral artery usually has either a bifurcation or trifurcation depending on whether the
anterior temporal branch originates at or just proximal to the main bifurcation. The trunk of the middle

168
Neurosurgery Books
cerebral artery typically
bifurcates to yield an
anterior and a posterior Superior frontal gyrus

group of arteries. The first

Middle frontal gyrus
branch of the anterior
division is the Inferior frontal gyrus
orbitofrontal artery, which
Superior frontal sulcus
supplies the inferolateral
aspect of the frontal lobe.
Precentral gyrus
The second division of the
ascending frontal group is Central sulcus
the operculofrontal artery,
which forms a candelabra Postcentral gyrus

of branches. By definition,
this group of Postcentral sulcus
operculofrontal arteries
consists of the arteries Superior parietal lobule
anterior to the central
Inferior parietal lobule
sulcus arteries. They
supply most of the middle
and inferior frontal gyri,
including Broca’s area and
169
Neurosurgery Books
part of the central lobe through the anterior parietal artery and its ramifications.
There usually are one to two central sulcus arteries (Rolandic arteries), which are the last branches
of the anterior bifurcation of the middle cerebral artery. The Rolandic arteries supply much of the central
lobe.
The posterior bifurcation of the middle cerebral artery typically has three major branches, which
arise in the Sylvian fissure, loop over the insula, and go posterolaterally across the cortical surface. The
first major posterior branch is the posterior parietal artery. It supplies most of the parietal lobe posterior
to the central lobe. The second major branch is the angular artery, which often has a common trunk with
the posterior parietal artery. The angular artery usually is the largest cortical branch of the middle
cerebral artery. As it emerges from the apex of the Sylvian fissure, it runs posteriorly and superiorly to
supply the posterolateral parietal lobe, the lateral occipital lobe, and the superior temporal gyrus. The
third major posterior branch is the posterior temporal artery. which supplies the temporal lobe.
Venous Drainage

The major venous drainage of the central and parietal lobes consists of the vein of Trolard, vein of
Labbé, the superficial middle cerebral vein. and several unnamed veins anterior to the vein of Trolard,
which run laterally to medially and terminate in the superior sagittal sinus.

170
Neurosurgery Books
 Cingulate sulcus
Cingulate gyrus Paracentral lobule
Collosal sulcus
Central sulcus
Medial frontal gyrus

Marginal limb cingulate sulcus

Precuneus

Parietooccipital sulcus

Cuneus

Clacarine sulcus

Gyrus rectus

Anterior commissure

Uncus

Parahippocampal gyrus Lingual gyrus

Collateral sulcus Fusiform (occipitotemporal) gyrus

Neurosurgery Books
FALCINE MENINGIOMA

Falcine meningiomas can occur anywhere along the falx cerebri in the interhemispheric fissure.
They may originate along the superior sagittal sinus and invade it, even occlude it. They also may extend
inferiorly to involve the inferior sagittal sinus. Some surgeons divide these tumors into outer and inner
falcine meningiomas on the basis of whether the epicenter of the tumor arises from the more superficial
or deeper aspect of the falx cerebri. The primary blood supply to these tumors is from the dura mater.
Before operating on a falcine meningioma, it is necessary to know the status of the adjacent sagittal
sinus. Whether the sinus is patent or thrombosed has serious implications for the surgical management
of the tumor. Specifically, if the sagittal sinus is thrombosed because of tumor invasion, complete tumor
removal (which requires excision of the involved sagittal sinus) is likely to be tolerated. However, if the
tumor has invaded the outer margin of the sinus but the sinus is still intact, the sinus must be
preserved. In this circumstance, the outer leaf of the dura mater is cauterized extensively
with a bipolar cautery.
In most large falcine meningiomas, a large cortical draining vein inevitably overlies the tumor
exposure. It is important to preserve these cortical veins because injury to them can lead to venous
hemorrhagic infarction. If the location of the cortical veins can be determined with preoperative imaging
studies, a bone flap can sometimes be planned to avoid unnecessary exposure.
It is important to make sure that the bone flap exposes the superior sagittal sinus. This will decrease
the amount of retraction needed along the medial surface of the hemisphere. After the superficial aspect

172
Neurosurgery Books
of the tumor is exposed, the arterial blood supply originating from the falcine dura mater is cauterized.
Thereafter, the tumor is debulked and collapsed into the operative field. An arachnoid plane is usually
present between the tumor and the callosomarginal and pericallosal arteries.

Figure 4-3.
In this example, the
epicenter of the
meningioma is located
along the inferior aspect
of the falx cerebri. The
precuneus overlies the
tumor. The primary
blood vessel at risk is
the pericallosal artery
and its branches.

173
Neurosurgery Books
 Figure 4-4.
Step 1. Several head positions are good for exposing a
posterior falcine meningioma. The slouch position is ideal
because it provides good orientation, promotes venous drainage,
and provides a good line of sight for visualizing the relationship
between the superior sagittal sinus, dura mater, and tumor,
thereby minimizing the need for lateral retraction of the brain.
An alternative position is the prone position, with a slight-to-
moderate reverse Trendelenburg position to promote venous
drainage.

Figure 4-5.
Step 2. It is important to make sure that the bone flap (dashed line)
is far enough medial to expose most of the width of the superior sagittal
sinus. There are several options for making this medial bone cut. As
illustrated here, two bur holes are made. A Penfield #3 dissector is used
to dissect the dura mater off the underside of the bone flap. The cut can
be made with a craniotome, a Gigli saw, or a high-speed air drill with a
diamond bur. This medial cut should be made last, so that if an
inadvertent tear is made in the superior sagittal sinus the bone flap can
be removed quickly to achieve hemostasis.

Neurosurgery Books
Figure 4-6.
Step 3. It is important to tack the dura mater to the margins of the surrounding bone. Any oozing of the blood
from the superior sagittal sinus is best controlled by placing a piece of absorbable gelatin sponge (Gelfoam) or
muscle on top of the sinus and holding it in place by tacking a reflected flap of dura mater.

Neurosurgery Books
Figure 4-7.
Step 4. In this example, the line of sight to the tumor lies between two cortical surface draining veins, which
must be protected. The ensheathing arachnoid should be left intact. A piece of absorbable gelatin sponge can be
placed around the junction of the vein with the sagittal sinus to buttress it. The medial hemisphere is lined with
hemostatic fabric (Surgicel) and cottonoids (Americot) and then gently retracted laterally.

Neurosurgery Books
 Figure 4-8.
Step 5. The blood supply to
the meningioma from the falx
cerebri is cauterized with a
bipolar cautery.

Neurosurgery Books
 Figure 4-9.
A-C, The tumor is debulked with a bipolar
cautery, curets, and Ultrasonic aspirator. Next,
the tumor is collapsed into the operative field.
This will lessen the degree of hemispheric
brain retraction. Arachnoid usually overlies the
pericallosal artery and its branches.
A Respecting this arachnoid will help preserve
these vessels.

Neurosurgery Books
B

Neurosurgery Books
 Figure 4-10.
In this example, the origin of the tumor
from the falx cerebri is excised to decrease the
risk of recurrence. The falx cerebri is first
cauterized and then incised with a #11 blade
knife. A tooth forceps is used to hold the falx
cerebri laterally, and a small dissecting scissors
is used to excise the falx. It is important to make
sure that the inferior sagittal sinus is well
cauterized. If the tumor is attached to the outer
dura mater of a patent sagittal sinus, the only
option is aggressive bipolar cauterization.

Neurosurgery Books
CONTRALATERAL APPROACH TO FALCINE MENINGIOMA

Rarely, a large cortical vein may

overlie the middle of the operative
field, putting the vessel at high risk
during tumor removal. In this
circumstance, the surgeon may
consider a contralateral approach to the
tumor. The advantage of the
contralateral approach is that the major
cortical draining vein will be protected.
However, this approach has at least two
disadvantages. First, both cerebral
hemispheres are placed at risk, because
to achieve exposure the opposite
hemisphere must be retracted. Second,
the relationship of the tumor to the
pericallosal and callosal marginal
arteries is less well visualized than with Figure 4-11.
an ipsilateral approach. The bone flap has been planned to consider all three cortical
draining veins.

181
Neurosurgery Books
 A

Figure 4-12. A and B,

The contralateral hemisphere is gently
retracted. The falx cerebri is incised below
the superior sagittal sinus. Typically, the
falx cerebri is red and vascular. Palpating
the falx cerebri with a small spatula will
give the surgeon a tactile sense about
where the tumor is located. The falx cerebri
is cauterized and incised with a #11 blade
knife. A bipolar cautery is used to sever the
attachments and the blood supply between
the tumor and the falx cerebri.

Neurosurgery Books
Figure 4-13.
As with all meningiomas, it is important to debulk the tumor and to collapse the tumor capsule
into the operative field. Although the final “surgical specimen” is less impressive in size than the
original tumor, aggressive debulking minimizes the need for brain retraction.
Neurosurgery Books
TRIGONE VENTRICULAR TUMOR

The most common neoplasms of the trigone of the lateral ventricle are meningiomas and choroid
plexus papillomas. Other tumors, including gliomas of all types, subependymomas, and giant cell
astrocytomas, are more frequent more anteriorly in the body of the ventricle. Still more anteriorly,
adjacent to the septum pellucidum and the foramen of Monro, colloid cysts and central neurocytomas
are more frequent.
Meningiomas originate from the arachnoid cap cells of the choroid plexus and tela choroidea. The
most common site for an intraventricular meningioma is the trigone, although this tumor may rarely
arise in the third ventricle or, least commonly, in the fourth ventricle. Choroid plexus papillomas, in
comparison, occur most commonly in the fourth ventricle, with extensions into the cerebellopontine
angle, or in the trigone.
The primary blood supply of both meningiomas and choroid plexus papillomas is derived from the
anterior and posterior choroidal arteries. Three surgical approaches can be considered for gaining access
to tumors in the trigone. The first and oldest is through the middle posterior temporal gyrus. The
advantage of this approach is that it allows the surgeon good access to the anterior blood supply of the
tumor along the choroid plexus. A disadvantage is that an incision in the posterior left temporal lobe has
the risk of causing language dysfunction. Also, significant brain retraction must be applied to visualize
the posterior aspect of the trigone.

184
Neurosurgery Books
 The second approach is through the superior parietal lobule. This is described in the sequence of
illustrations shown here. The disadvantage of this approach is that a cortical incision must be made in
the parietal lobe. A cortical incision in and retraction of the left superior parietal lobule is well tolerated,
with little risk of injury; however, a significant incision in or retraction of the right superior parietal
lobule can cause the patient marked spatial disorientation postoperatively. Another disadvantage is that it
is more difficult to access the anteriorly located blood supply of the tumor.
The third approach is an interhemispheric-splenial approach through a parietooccipital craniotomy,
with an incision in the junction between the precuneus and the splenium of the corpus callosum. The
advantage of this approach is that it does not require an incision in the temporal or parietal cortex. The
disadvantage is that moderate lateral retraction of the cerebral hemisphere may be required to access
both the lateral margin of the tumor and the anterior and inferior blood supply from the choroid plexus.

185
Neurosurgery Books
 Figure 4-14.
This drawing shows a tumor in the
trigone and its relationship with its arterial
blood supply from the anterior and
posterior choroidal arteries. The surgical
approach chosen for this tumor is a
stereotactic craniotomy through the left
superior parietal lobule.

Neurosurgery Books
A B

Figure 4-15.
Step 1. The use of stereotaxis, either frame (A) or frameless (B) allows the surgeon
to plan many surgical approaches that allow the best line of sight for the surgeon with
the least amount of brain retraction and cortical disruption

Neurosurgery Books
Figure 4-16.
Step 2. When planning a stereotactic approach through the superior parietal lobule, the
surgeon can often see on magnetic resonance imaging a deep or enlarged sulcus around the
planned trajectory. If such a sulcus is present, the surgeon should use it during the approach to the
lateral ventricle. This intrasulcal approach lessens the degree of cortical injury. As depicted here,
the disadvantage of the superior parietal approach is that the arterial blood supply of the tumor is
deep and more difficult to access than it is with the traditional middle temporal gyrus approach.

Neurosurgery Books
 A

Figure 4-17.
Step 3. A, A trephine craniotomy
is shown: with the use of an operating
microscope, the sulci are inspected for
an intrasulcal approach. In this
example, there is an appropriate
sulcus just posterior to a cortical
vessel. All the cortical arteries and
veins in this region should be
protected. Opening the sulcus is
similar to dividing the Sylvian fissure.
B
B, A sharp #11 blade knife is used
to stab the arachnoid and to lift it up
as it is severed. A microscissors can be
used to continue the cut in the
arachnoid that overlies the sulcus.
Usually, about a 1.5-cm length of
sulcus has to be exposed to obtain
adequate exposure of the lateral
ventricle.

Neurosurgery Books
 Figure 4-18.
Step 4. After the arachnoid has been incised, bipolar
cautery is used to spread the sulcus until the deeper arcuate
fibers (“U fibers”) of the white matter are encountered. Two
self-retaining retractors are used to retract the adjacent gyri
after they are lined with hemostatic fabric and cottonoids.
The white matter is divided using the bipolar cautery and
suction. Whenever using an intrasulcal approach to a deep-
seated tumor, it is best to consider the direction of the
radiations within the corona radiata, which is being entered.
Dividing the white matter parallel with, or in the plane of,
these radiations will decrease the risk of neurologic injury.

Neurosurgery Books
Figure 4-19.
Step 5. As the tumor is approached, the
ependyma of the lateral ventricle is seen bulging
out into the surgical field; it forms a thin cap over
the tumor. After the ependyma has been incised,
the tumor capsule becomes visible. The tumor is
centrally debulked, using primarily the bipolar
cautery and suction. By using the bipolar cautery
instead of curets or an ultrasonic aspirator,
hemostasis will be earlier and better, which has
the advantage of decreasing the amount of blood
entering the lateral ventricle. Spilling of blood
into the lateral ventricle may cause aseptic
meningitis postoperatively. The administration of
corticosteroids perioperatively lessens the severity
of such aseptic meningitis. After a space has been
created between the ependyma and the tumor
capsule, cottonballs or cottonoids can be placed
deep to help prevent blood from spilling into the
lateral ventricle. After the tumor has been
resected, it is important to irrigate the lateral
ventricle copiously with warm saline solution.
After the tumor has been debulked, its capsule is
collapsed into the surgical view. The
anteroinferior feeding arteries arising from the
choroid plexus can be cauterized and divided. The
tumor is debulked further and subsequently
extracted from the lateral ventricle.

Neurosurgery Books
LESIONECTOMY—METASTATIC TUMORS AND CAVERNOUS HEMANGIOMAS

The most common cerebral metastases occur from neoplasms of the lung, breast, skin, kidney, and
gastrointestinal tract. Their distribution within the central nervous system mirrors the volume of brain
compartments and lobes. Therefore, the most common locations are the frontal lobe, the parietal lobe,
and the cerebellum. In the cerebral cortex, these tumors tend to occur at the junction of the gray and
white matter and they typically are well circumscribed. Accordingly, a localized stereotactic craniotomy
is ideal for resecting these subcortical tumors. In the central or parietal lobe, the major issue is the
relationship of the tumor to eloquent cortex.
Cavernous hemangiomas are vascular malformations that at surgery are lobulated and well
demarcated; they have a reddish-brownish mulberry appearance. They tend to grow like slow neoplasms
and have repetitive microhemorrhages that often cause seizures. Cavernous hemangiomas are found in
all lobes of the cerebral cortex, the brain stem (with a propensity to localize to the pons), and the basal
ganglia. Surgical resection of cortical lesions usually produces good results; for example, lesionectomy
alone produces seizure control in 80 to 90 percent of patients. Surgical resection of cavernous
hemangiomas is technically easy because they are well-circumscribed, low-flow lesions that do not have
a significant arterial blood supply. Furthermore, there is a good plane of separation between the
cavernous hemangioma and the surrounding gliotic plane that results from the repetitive hemorrhages. In
large cavernous hemangiomas, the various cysts can be decompressed to collapse the wall of the tumor
into a small operative field. The major difficulty with resection of small cortical cavernous
hemangiomas is their location and relationship to eloquent cortex. For this reason, use of stereotactic

192
Neurosurgery Books
guidance and the intrasulcal approach described below is ideal for resecting both metastatic tumors and
cavernous hemangiomas.

Superior frontal gyrus

Middle frontal gyrus

Inferior frontal gyrus

Superior frontal sulcus

Precentral gyrus

Central sulcus

Postcentral gyrus

Figure 4-20.
Postcentral sulcus Step 1. A stereotactic
trephine craniotomy is made
Superior parietal lobule
using a frameless system;
therefore, the skull is held
Inferior parietal lobule rigid.

193
Neurosurgery Books
Figure 4-21.
Two good head positions are the lateral decubitus position, with the head parallel to the floor. and the
supine position, with the head straight up. Both of these positions are helpful in surgical orientation. If the
lateral decubitus position is chosen, it is important to make sure that the neck is not overrotated. Placing
towels or blankets under the shoulder relieves tension on the neck.

Neurosurgery Books
 Figure 4-22.
Step 2. The dura mater is tacked to the bone
margins and opened. In most metastases, the
subcortical location may be difficult to visualize
by inspection of the surface with the operating
microscope. Therefore, stereotaxis helps to
minimize the size of the craniotomy and brain
exposure. Because the exact location of the
precentral gyrus usually is not clear, as in this
example, somatosensory evoked potential
monitoring is used to identify the central sulcus.
The wave reversal points to the location of the
central sulcus. In this example, electrode “A” is
on top of the precentral gyrus, “B” represents the
central sulcus, and “C” identifies the postcentral
gyrus. An alternative to this type of mapping is
direct electrical stimulation of the cerebral cortex
with assessment of motor function, with the patient
either awake or given general anesthesia without
inhalation anesthetics. As preoperative functional
imaging improves, this type of intraoperative
stimulation or mapping may not be necessary.

Neurosurgery Books
 Figure 4-23.
Step 3. An intrasulcal approach to the
tumor has been chosen. The arachnoid is
incised with a #11 blade knife and spread with
the bipolar cautery forceps. The adjacent
cortical vein is preserved. The intrasulcal
approach is extended down to the “U” fibers,
where a metastatic tumor is typically
encountered. If necessary, small self-retaining
retractors can be used to spread the sulcal
opening. A bipolar cautery and suction are used
to separate the tumor from the adjacent brain
parenchyma.

Neurosurgery Books
Figure 4-24
. Several small feeding arteries
usually enter a cavernous hemangioma.
They should be cauterized and divided
individually. Remember that irrigating
the tips of the bipolar cautery helps while
these vessels are being cauterized, as in
resection of an arteriovenous
malformation. In the example shown here
is an adjacent venous angioma that
should be preserved

Neurosurgery Books
TRIGONE ARTERIOVENOUS MALFORMATION

Arteriovenous malformations of the trigone and body of the lateral ventricle derive their primary
arterial blood supply from the anterior and posterior choroidal arteries and branches off segments P1,
and P2, and P3 of the posterior cerebral artery. The primary venous drainage is through subependymal
veins, which drain into the basal vein of Rosenthal. If the arteriovenous malformation extends into the
lateral ventricle, it usually involves the posterior limb of the internal capsule and the thalamus, and the
venous drainage also includes the posterior septal and thalamostriate veins, which drain into the internal
cerebral vein and the basal vein of Rosenthal. The location of the intraventricular arteriovenous
malformation dictates the surgical approach. An arteriovenous malformation or tumor located in the
posterior medial quadrant of the trigone can be resected through a posterior transcallosal approach, using
a midline parietooccipital craniotomy. A lesion located more laterally or posteriorly in the lateral
ventricle can be accessed better through an intrasulcal transcortical approach through the posterior
parietal lobe. An inferolaterally located lesion that involves the temporal horn is best approached
transcortically through the middle temporal gyrus. In many ventricular arteriovenous malformations, the
initial presentation is that of a cerebral hemorrhage, with extension of the blood clot into brain
parenchyma. The blood clot can be used to the surgeon’s advantage. First, it often dissects some aspects
of the margin of the arteriovenous malformation from the surrounding parenchyma. Second, if the
hemorrhage approaches the cortical surface, it often provides a route for reaching the arteriovenous
malformation with little risk of causing more neurologic injury.

198
Neurosurgery Books
 Some general principles need to be followed for all resections of arteriovenous malformations
regardless of their location. Yasargil originally emphasized the importance of the location of an
arteriovenous malformation, whether it was superficial or deep, supratentorial or infratentorial.
Subsequently, Spetzler and Martin introduced a grade I to V classification that considers size,
relationship to eloquent cortex, depth, location, and venous drainage.
First, as in any operation, it is mandatory that the surgeon have a realistic assessment of the expected
risk of resecting the arteriovenous malformation. This is especially true with the advent of alternative
options for treatment, including stereotactic irradiation.
Second, preoperative embolization is beneficial for several reasons. The process of embolization
forces the surgeon and interventionalist to thoroughly study and understand the three-dimensional
anatomy of the arteriovenous malformation. Selective Amytal injections can help determine vessels “en
passage” that perfuse eloquent cortex. There is no doubt that multiple-stage preoperative embolization
facilitates intraoperative hemostasis. Also, staged preoperative embolization of a large arteriovenous
malformation allows time for the brain vasculature to reequilibrate to alterations in cerebral blood flow.
This decreases the risk of postoperative hemorrhage and hyperperfusion complications.
Third, during the surgical resection, it is important to have clean irrigating bipolar cautery. Because
blood vessels of an arteriovenous malformation do not have a normal media, cauterization of the smaller
vessels is difficult. Small microclips are useful for occluding small blood vessels along the ependymal
surface that are difficult to coagulate. Small wire clips are also useful for temporarily occluding
suspected “en passage” blood vessels.

199
Neurosurgery Books
 Fourth, it is important to inspect the surface anatomy of the arteriovenous malformation before
resection and to identify the draining veins. These draining veins contain arterialized blood. The best
intraoperative confirmation that an arteriovenous malformation has been completely resected is
resolution of this arterialized venous blood. At least one major draining vein should be preserved
throughout the resection of the malformation and cauterized last. After the arteriovenous malformation
has been resected, the entire operative bed should be thoroughly inspected under the operating
microscope. Although some surgeons prefer to resect arteriovenous malformations with the patient
mildly hypotensive, it is mandatory that systolic blood pressure be normal when the postoperative bed is
inspected.
Fifth, in patients with large arteriovenous malformations, the systolic blood pressure is decreased to
approximately 90 mm Hg for 48 to 72 hours postoperatively. Cerebrovascular autoregulation may not be
normal in the cerebrovasculature surrounding the resected arteriovenous malformation; therefore,
controlling the patient’s postoperative blood pressure will decrease the risk of postoperative hemorrhage.
Sixth, in very large complex arteriovenous malformations, it is important to consider performing
angiography immediately postoperatively while the patient is still intubated. In the author’s experience,
intraoperative angiograms do not have the same degree of resolution of angiograms obtained in a formal
angiographic suite. Therefore, if angiography is considered immediately postoperatively, the craniotomy
should be closed quickly and the patient transported to the angiographic suite while under general
anesthesia. Thus, if the postoperative angiogram demonstrates residual arteriovenous malformation, the
patient can be taken back immediately to the operating room, with little waste of time. One must be
aware that interpreting an immediate postoperative angiogram is difficult. Often, there is stagnant blood
200
Neurosurgery Books
flow or contrast material in draining veins
that have been occluded. These should not
be mistaken for early draining or shunting
veins, which are the hallmark of residual
arteriovenous malformation. These early
draining veins are best seen in the middle
portion of the arterial phase of an
angiogram. Any residual arteriovenous
malformation must be treated because the
risk of hemorrhage is high. If residual
arteriovenous malformation is detected in
eloquent or difficult to access cerebral
cortex, an argument can be made for early
postoperative stereotactic irradiation.

Figure 4-25.
The relationship of the ventricular system with the
superficial and deep venous drainage. The primary venous
drainage from an arteriovenous malformation in the trigone
is the basal vein of Rosenthal and the internal cerebral vein
through small subependymal veins, posterior septal veins,
and the thalamostriate vein.

201
Neurosurgery Books
Figure 4-26.
An incision through the
posterior middle temporal gyrus
provides good access to an
arteriovenous malformation that is
located in the temporal horn and
extends into the trigone. The arterial
blood supply is from the anterior
choroidal artery, the posterior
choroidal artery, and the perforating
vessels that come off segments P2
and P3 of the posterior cerebral
artery. Most of the arterial blood
supply is deep to the surgical
approach.

Anterior choroidal artery Calcarine artery

Medial posterior choroidal artery Parietooccipital artery

Basal vein
Lateral posterior choroidal artery
of Rosenthal

Neurosurgery Books
 Figure 4-27.
Step 1. A horseshoe-shaped incision
centered posterior to the ear provides good
access to the posterior temporal lobe. It is
important to make the inferior bone cut with
care to prevent inadvertent tearing of the
transverse sinus. The lower margin of the
craniotomy should also be along the floor of
the middle cranial fossa. In very large
arteriovenous malformations of the trigone
with massive branches of the posterior
cerebral artery, a subtemporal approach to
these feeding vessels can be considered if
preoperative embolization was not successful
in obliterating them.

Neurosurgery Books
Figure 4-28.
Step 2. A cortical incision is made
in the middle temporal gyrus over a
length of approximately 2.0 cm. If there
is a significant sulcus between the
superior and middle temporal gyri, an
intrasulcal approach to the lateral
ventricle can be attempted. However, the
sulci in the posterior temporal lobe
usually are not well formed; therefore,
an intrasulcal approach offers little
advantage. After the lateral ventricle has Basal vein of
Rosenthal
been entered, two self-retaining
retractors are used to retract the
ependymal wall. The choroid plexus is
identified. Invariably, many arterial
feeding vessels come out of the choroid
plexus and go to the arteriovenous
malformation. They are branches of both
the anterior and the posterior choroidal
arteries. They should be meticulously
cauterized and divided. Microclips may
be helpful.

Neurosurgery Books
Figure 4-29.
Step 3. The medial, deeper aspect of the
arteriovenous malformation is dissected free. This will
allow access to the branches that come off the P2 and
P3 segments. The arteriovenous malformation is gently
retracted with a #5 or #7 suction tip and cottonoid. In
the example shown here is a large draining vein that
presumably goes to the basal vein of Rosenthal. These
medial draining veins and arteries enter the lateral
ventricle through the choroidal fissure; therefore, there
is a layer of arachnoid overlying the posterior cerebral
artery. Respecting this arachnoid will help prevent
inadvertent injury of this artery.
Step 4. The arteriovenous malformation is rotated
anteriorly to allow exposure and dissection of the
posterior feeding arteries and subependymal draining
veins. Each is individually cauterized and divided.
Step 5. After the arteriovenous malformation has
been completely isolated from its arterial blood supply,
the preserved draining veins should be inspected. If
there is arterialized blood, residual arteriovenous
malformation is still present and needs to be identified
and resected.
Neurosurgery Books
Chapter 5
Occipital
Approach and
Combined
Suboccipital
Approach

Neurosurgery Books
Chapter 5: Occipital Approach and
Combined Occipital-Suboccipital
Approach
ANATOMIC CONSIDERATIONS

The occipital lobe forms the posterior part of the cerebral hemisphere. Its lateral surface includes the
superior and inferior occipital gyri, which are separated by the lateral occipital sulcus. The major gyrus
on the inferior, or tentorial, surface of the lobe is the fusiform, or occipitotemporal, gyrus. The medial
surface of the occipital lobe is divided by the calcarine sulcus into two parts, called the “cuneus” and the
“lingual gyrus” (or “lingula”).
The cuneus (Brodmann areas 17, 18, and 19) is the triangularly shaped area between the
parietooccipital sulcus superiorly and the calcarine sulcus inferiorly. The apex of the triangle verges on
the isthmus of the cingulate gyrus. The lingual gyrus (Brodmann areas 17, 18, and 19) lies between the
calcarine sulcus superiorly and the collateral sulcus inferiorly. It is continuous anteriorly with the

207
Neurosurgery Books
parahippocampal gyrus of the temporal lobe. The lingual gyrus is separated from the fusiform
(occipitotemporal) gyrus by the collateral sulcus.
The optic radiations, also called the “geniculocalcarine tract,” project from the lateral geniculate
body of the thalamus to the calcarine cortex (Brodmann area 17), which is the primary visual cortex.
Calcarine cortex is the cerebral cortex that borders the calcarine fissure. The optic radiations are
organized retinotopically. The portion of the radiations that represents the upper retinal quadrants (lower
visual field) terminates along the superior bank of the calcarine sulcus, part of the cuneus. The portion
that represents the lower retinal quadrants (upper visual field) terminates along the inferior bank of the
calcarine sulcus, part of the lingual gyrus. Moreover, the peripheral retina is represented anteriorly in the
calcarine cortex and the central retina (macula) is represented posteriorly, including the occipital pole.

ARTERIAL BLOOD SUPPLY

The blood supply to the occipital lobe is from the posterior cerebral artery, which originates at the
bifurcation of the basilar artery on the ventral surface of the midbrain. From its origin, the posterior
cerebral artery turns laterally and posteriorly and enters the interpeduncular cistern. The P1 segment of
the artery extends from the bifurcation of the basilar artery to the origin of the posterior communicating
artery. This segment gives rise to the posterior thalamoperforating arteries, which vary in number and
size. They run superiorly in the interpeduncular cistern to supply the posterior limb of the internal
capsule, the thalamus, the hypothalamus, and the subthalamic nuclei.

208
Neurosurgery Books
 From the interpeduncular cistern, the posterior cerebral artery turns posteriorly into the ambient
cistern. This is segment P2, which extends from the junction between the posterior communicating and
the posterior cerebral arteries to the dorsal aspect of the midbrain. The thalamogeniculate arteries
originate from segment P2 and supply part of the posterior limb of the internal capsule, the lateral and
medial geniculate bodies, part of the optic tract, and part of the thalamus. Other perforating branches
arising from segment P2 supply the cerebral peduncles. This segment also gives rise to the medial
posterior choroidal artery, which occasionally originates from segment P1. The medial posterior
choroidal artery runs anteriorly around the brain stem and turns superomedially to enter the roof of the
third ventricle. As it runs anteriorly toward the foramen of Monro, it supplies the pineal body, the
choroid plexus of the third and lateral ventricles, and the thalamus. Segment P2 also gives rise to one or
more lateral posterior choroidal arteries. The posterior choroidal arteries may also originate from the
proximal segments of cortical branches of the posterior cerebral artery. The lateral posterior choroidal
arteries enter the choroidal fissure laterally and curve anteriorly around the pulvinar to supply the
choroid plexus, part of the cerebral peduncle, the lateral geniculate body, the posterior limb of the
internal capsule, the thalamus, the fornix, and the caudate nucleus.
Four primary cortical branches originate from the posterior cerebral artery. One of these, the inferior
temporal artery, often originates from segment P2. Branches of the inferior temporal artery supply the
inferior part of the temporal lobe. These branches are named the “hippocampal,” “anterior,” “middle,”
and “posterior temporal arteries.” Occasionally, branches of the posterior temporal artery extend
posteriorly to perfuse part of the visual cortex.

209
Neurosurgery Books
 Segment P3 of the posterior cerebral artery extends from the quadrigeminal cistern to the calcarine
sulcus. It often divides into branches before entering the calcarine fissure. Its cortical branches are as
follows: the parietooccipital artery usually originates in the ambient cistern and extends posteriorly
along the parietooccipital sulcus. It supplies part of the precuneus, the cuneus, and the superior occipital
gyrus. It also may perfuse part of the parietal and central lobes. The calcarine artery runs in the calcarine
sulcus and supplies much of the cuneus and lingual gyrus, that is, primary visual cortex (Brodmann area
17). Some small splenial arteries that arise from segment P3 or the parietooccipital artery supply the
splenium of the corpus callosum.

TORCULAR MENINGIOMA

Often, meningiomas that arise along the posterior falx cerebri and tentorium cerebelli are best
removed through an interhemispheric approach, using a parietooccipital craniotomy. The blood supply
to the tumor originates from the tentorium cerebelli. The relationship of the tumor to the surrounding
venous sinuses and torcular Herophili must be considered. If the tumor invades a patent sinus, the best
plan is to create a plane of dissection parallel to the dura mater and sinus using the bipolar cautery.
Residual tumor along the wall of the sinus is cauterized aggressively.
Large tumors that have a deep extension into the tentorial notch are more difficult to excise because
they are in proximity to the vein of Galen along the midline and to the posterior cerebral arteries
laterally. The arachnoid layer between the tumor capsule and these vascular structures must be
respected. In tumors that extend caudally, sectioning the tentorium cerebelli is beneficial. An incision is
made parallel with and approximately 5 to 10 mm lateral to the straight sinus. For orientation, the vein
210
Neurosurgery Books
of Galen and the posterior cerebral arteries are
above the tentorium cerebelli, and the superior
cerebellar arteries and oculomotor nerves are
below it.

Figure 5-1.
A midline occipitoparietal craniotomy works
well for resecting tumors along the falx cerebri and
those on or above the tentorium cerebelli. The
medial bone cut must be on the superior sagittal
sinus, and the inferior cut should extend to the top
of the transverse sinus. In this way, there will be less
retraction along the occipital pole. The cuts along
the superior sagittal and transverse sinuses must be
made with care. Placing bur holes and then
stripping the dura mater with Penfield #1 and #3
dissectors before making the bone cuts decreases
the risk of inadvertently tearing a sinus. If there is
significant hyperostosis, it is best to use a high-
speed air drill with a diamond bur.

211
Neurosurgery Books
 Figure 5-2.
Two patient positions that work well for a midline parietooccipital
craniotomy are the semisitting position (A) and the prone position
(B). The semisitting, or slouch, position affords excellent brain
relaxation, with good orientation and surgical access. However, with
the head higher than the level of the heart, the risk of air embolism
must be considered because the surgeon is working in proximity to
the major venous sinuses. Therefore, the anesthesiologist should
place both a transesophageal echo and a right atrial catheter. The
risk of air embolism may be decreased with the prone position
A because the head is not elevated above the heart. However, a
disadvantage of the prone position is the risk of increased
intracranial pressure because of decreased venous drainage from the
brain. If the prone position is used. it is important to make sure that
the chest rolls are placed correctly to decrease intrathoracic pressure.
A reverse Trendelenburg body
position combined with head
flexion promotes venous
drainage. In B, note that
intermittent compression boots
have been placed to help
decrease the risk of postoperative
deep thrombosis, which is a
recognized complication after
operations on intracranial
meningiomas.
B

Neurosurgery Books
 Figure 5-3.
Step 1. A, A U-shaped flap is used for
exposure. Because the broad flap is
supplied by branches of the occipital artery,
there is no risk of ischemic scalp necrosis.
As illustrated here, three bur holes are
made over the transverse and sigmoid
sinuses. A Penfield #3 dissector is used to
dissect the dura mater off the underside of
the bone flap. A craniotome is used to make
the outer cut. The author’s preference is to
use a high-speed air drill with a diamond
bur to make the medial and inferior cuts
over the sinus.
A Step 2. B. The dura mater is opened
and reflected over the transverse and
sigmoid sinuses and tacked to the margins
of the bone. By exposing the bone over the
superior sagittal sinus and then tightly
tacking the dura mater, the sinuses are
retracted slightly to provide greater
exposure.

Neurosurgery Books
 Figure 5-4.
Step 3. A, A #5 or #7 suction tip and a
A bipolar cautery are used to investigate the plane
between the capsule of the tumor and the
occipital pole. As true for most meningiomas, an
arachnoid-pial plane separates the tumor from
the enveloping cerebral cortex.
Step 4. B, The occipital lobe is retracted
gently. The base of the tumor along the
tentorium cerebelli is sharply divided using a
bipolar cautery. Similarly, the attachment of the
tumor to the falx cerebri is also cauterized.
There usually is an intimate relationship
between the base of the tumor and the straight
sinus. Bleeding from the straight sinus usually
B can be controlled with absorbable gelatin
sponge (Gelfoam).

Neurosurgery Books
 Figure 5-5.
Step 5. A, The tumor is debulked using the
bipolar cautery, suction, ultrasonic aspirator,
or cutting loops. Aggressive debulking of the
tumor allows large lesions to be removed with
little retraction of the brain.
Step 6. B, The most medial and deep
aspect of the tumor is cauterized. Occasionally,
along the edge of the tentorium cerebelli, there
are unexpectedly large arterial branches from
A the internal carotid artery.

Neurosurgery Books
A

Figure 5-6.
A, There is a tear in the straight sinus. For all sinus
B tears, the best way to control hemostasis is first to place a
piece of absorbable gelatin sponge and a cottonoid
(Americot) with tamponade. Next, the surgical assistant
harvests a piece of muscle from the skin flap, which is used
to replace the absorbable gelatin sponge.
B, The muscle is sewn in place with several permanent
3-0 tacking sutures. A braided suture such as silk is better
than monofilament. When tying the surgical knot, the first
two ties should be thrown in the same direction. This allows
the surgeon to cinch down the knot on top of the muscle
plug. The third tie is thrown in the opposite direction to lock
the knot.

Neurosurgery Books
PINEAL TUMOR

The parietooccipital transtentorial interhemispheric approach to pineal tumors may be advantageous

if the neoplasm has a significant caudal extension toward the fourth ventricle. If the majority of the
tumor lies above the tentorium cerebelli, a suboccipital approach is superior. In particularly large tumors,
both a parietooccipital and a suboccipital craniotomy can be performed to allow multiple angles for
removing the tumor.

217
Neurosurgery Books
Figure 5-7.
The intracranial venous system. A and
B illustrate the pertinent anatomy of the
intracranial venous system that needs to be
considered when approaching a tumor in
the region of the pineal body. The basal vein Vein of Trolard
of Rosenthal is formed by the anterior and
deep middle cerebral veins and small Inferior sagittal sinus
branches from the insula and cerebral
Anterior septal vein
peduncles. As the basal vein of Rosenthal
goes posteriorly around the cerebral Caudate vein
peduncle, it receives tributaries from the
temporal horn and the medial temporal lobe Vein of Labbe
and is joined by the inferior striate veins,
which drain part of the basal ganglia. The Thalamostriate vein
basal vein of Rosenthal may also receive
Choroidal vein
blood from the lateral mesencephalic vein.
The basal vein of Rosenthal curves Internal cerebral veins
posteriorly around the cerebral peduncle
and collicular (tectal) plate to join the vein Basal vein of Rosenthal
of Galen. Vein of Galan

The septal and thalamostriate veins

join along the inferior aspect of the foramen
of Monro to form the paired internal
cerebral veins. Each internal cerebral vein
penetrates the subarachnoid space to run
just lateral to the midline in the velum
interpositum. As the vein goes posteriorly,

Neurosurgery Books
Figure 5-7 (cont.).
small subependymal veins and the
posterior septal veins join it. Just inferior to
the splenium of the corpus callosum, the
pair of internal cerebral veins join the pair
of basal veins of Rosenthal to form the vein Septum pellucidum
of Galen.
The vein of Galen is short and V- Fornix

shaped as it curves posteriorly and

Pineal recess third ventricle
superiorly around the splenium. The vein of
Galen ends at the apex of the tentorial Pulvinar
notch by joining the inferior sagittal sinus
to form the straight sinus. The vein of Galen Pineal gland
also receives small venous tributaries from
the occipital lobe, the corpus callosum, the Superior colliculi
mesencephalon, and the cerebellum. An
Inferior colliculi
important landmark is the midline
precentral vein that extends from the Trochlear nerve
cerebellum to the underside of the vein of
Galen.

Neurosurgery Books
 Figure 5-8.
Step 1. A U-shaped incision is made for this
exposure. The midline and inferior cuts
should be along the superior sagittal and
transverse sinuses, respectively. Note that the
head is rotated slightly to the right to allow
the surgeon a better line of sight. For pineal
tumors with large lateral, eccentric
extensions, head rotation can be used
accordingly. For example, if the patient has a
large tumor extending far right, the head
should be rotated slightly to the left.
Alternatively, a left-sided craniotomy can be
performed to allow a “crosscourt” approach.
The degree of head flexion is not as
pronounced as it is for a suboccipital
approach.

Neurosurgery Books
 Figure 5-9.
Step 2. The dura mater is tacked to the margins of
the bone, and the occipital lobe is lined with
hemostatic fabric (Surgicel) and cottonoids and gently
retracted. Next, an incision is made in the tentorium
cerebelli approximately 5 to 10 mm lateral to the
straight sinus. This incision is carried forward to the
quadrigeminal cistern. It is important to recognize
that the quadrigeminal cistern is dark. Sitting
underneath the arachnoid that covers the
quadrigeminal cistern is the right basal vein of
Rosenthal and the vein of Galen. Accordingly, when
initially opening this cistern, the surgeon needs to be
careful to avoid inadvertently tearing these veins,
which are displaced laterally and posteriorly by the
tumor. The best technique is to make a small nick with

Neurosurgery Books
 Figure 5-10.
Step 3. After the quadrigeminal
cistern has been opened, the splenium
of the corpus callosum is retracted
gently. This reveals the right basal
vein of Rosenthal and, likely, the right
internal cerebral vein.

Neurosurgery Books
 Figure 5-11.
Step 4. The medial occipital lobe and the
splenium of the corpus callosum are retracted
more vigorously to expose the rostral portion of
the tumor. Note that the pertinent veins lie within
a layer of arachnoid.

Neurosurgery Books
Figure 5-12.
Step 5. The tumor is inspected visually to determine the largest aperture for tumor debulking.
This usually occurs between the ipsilateral internal cerebral vein and the basal vein of Rosenthal.
The arachnoid and tumor capsule are incised. Small angled curets are used to perform internal
debulking of the tumor. A small spatula is used to gently dissect the veins within the ensheathing
arachnoid off the tumor capsule. Specifically, the right internal cerebral vein is dissected medially
and the basal vein of Rosenthal is dissected laterally and caudally. It is important to perform as
vigorous a debulking of the tumor as possible before manipulating the tumor capsule.
Neurosurgery Books
Figure 5-13.
Step 6. This
illustration of a midsagittal
section of the brain
emphasizes why a
parietooccipital
craniotomy with a
transtentorial approach
can be useful in removing
pineal tumors that have a
marked caudal extension.
It is difficult to remove the
inferior extension of a
pineal tumor through a
suboccipital approach
because of the need for
significant retraction of the
cerebellum. Alternatively, a
transtentorial approach
allows a downward line of
sight because the occipital
lobe is retracted laterally
and the surgeon works
through the
interhemispheric fissure. The cerebellum is protected with a cottonoid and displaced downward with a #5 or
#7 straight suction tip. Long angled curettes are used to debulk the tumor. The tumor capsule is worked
gently upward into the operative field. The collicular plate is usually caudal and ventral to the tumor.

Neurosurgery Books
ARTERIOVENOUS FISTULA OF THE LATERAL AND SIGMOID DURAL SINUSES

Dural arteriovenous fistulas of the lateral and sigmoid sinuses typically present with intractable
pulsatile tinnitus and headaches. On neurologic examination, a bruit and, often, papilledema are
detected, indicative of increased intracranial pressure. The nidus of the arteriovenous fistula is found at
the junction of the transverse and sigmoid sinuses. It has been proposed that these fistulas result from
thrombosis or stenosis of the transverse sinus. The blood supply to these lesions is complicated but
includes feeding vessels from the external carotid artery, including the occipital artery, the posterior and
the middle meningeal arteries, and vascularized petrous bone. There almost always is a blood supply
from the internal carotid artery through arterial branches that originate from the meningohypophysial
trunk. Occasionally, these fistulas may parasitize cortical arterial branches. The goal of the operation is
to disconnect the fistula. The fistula not only causes pulsatile tinnitus, but it can lead to cortical venous
hypertension. The presence of cortical venous hypertension with this type of fistula indicates an
increased risk of seizures and hemorrhage.
It is important to determine whether the sigmoid sinus on the side of the fistula is patent or
thrombosed. If it is patent, it must be preserved. The vein of Labbé likely drains antegradely into it. In
this circumstance, the arterial feeding vessels from the dura mater over the occipital lobe and the
cerebellum, the tentorium cerebelli, and the petrous bone are disconnected. If the sigmoid sinus is
thrombosed. it is easier technically to disconnect the fistula. Specifically, the transverse sinus is ligated
approximately 2 cm from the torcular Herophili and rotated outward and used as a handle to expose the
inner blood supply coming from the tentorium cerebelli. Next, the transverse sinus is devascularized

226
Neurosurgery Books
down to its junction with the sigmoid sinus, and the sigmoid sinus is packed with hemostatic fabric. This
leads to additional thrombosis of the sigmoid sinus and obliteration of the fistula at the level of the
petrous bone.
In the following example, the sigmoid sinus is patent. Therefore, the goal of the operation is to
disconnect the fistula without inducing thrombosis of the sigmoid sinus. Technically, this is more
difficult to achieve because it is difficult to visualize the tentorium cerebelli along the inner side of the
transverse sinus. The occipital lobe and the cerebellum both are retracted. This allows the tentorium
cerebelli to be incised, cauterized. and divided along the inner side of the transverse sinus down to its
junction with the sigmoid sinus. Thereafter, the bipolar cautery is used to extensively cauterize the dura
mater along the petrous bone, both in the posterior cranial fossa and along the occipital lobe. In effect,
this will also cauterize and induce thrombosis of the petrosal sinus, which usually participates in the
fistula.
Preoperative embolization is useful for obliterating much of the blood supply from the external
carotid artery. This facilitates intraoperative hemostasis. However, embolization or ligation of the
external carotid artery in itself does not lead to obliteration of the lesion. In fact, it has an adverse effect
by driving the blood supply deeper, making subsequent surgical resection more difficult and dangerous.

227
Neurosurgery Books
 Figure 5-14.
Step 1. The position of the
patient’s head is similar to that
used in a standard suboccipital
approach for acoustic neuroma or
microvascular decompression
surgery. Specifically, the body is
supine, with the head rotated
parallel to the floor and held in a
pinion. It is important not to
overrotate the neck, because this
causes significant postoperative
neck discomfort and might lead to
compression of the jugular vein
and, subsequently, increased
intracranial pressure.
Alternatively, the operating table
can be rotated laterally. In this
operation, it is important to have a
lumbar drain in place to facilitate
drainage of the cerebrospinal fluid,
because the cisterna magna will
not be immediately available for
drainage after bone removal.

Neurosurgery Books
Figure 5-15.
Step 2. The craniotomy is made with a high-speed air drill and
a diamond bur. A diamond bur is best because it helps with
hemostasis by cauterizing the venous and arterial channels in the
bone diploë. The air drill should be held at approximately a 30-
degree angle to prevent drilling vertically through the bone. Drilling
should proceed uniformly around the craniotomy.

Neurosurgery Books
 Figure 5-16.
Step 3. After the inner cortical layer has
been identified around the circumference of
the craniotomy, the bone flap can be removed
with a small curved periosteal elevator. As the
bone flap is removed, bleeding may occur from
the dural arterial and venous vessels. This can
be controlled with a large piece of absorbable
gelatin sponge. Preoperative embolization
decreases the risk of significant hemorrhage
during removal of the bone flap.

Neurosurgery Books
 Figure 5-17.
Step 4. A high·speed air drill and a diamond bur
are used to unroof the sigmoid sinus by removing the
lateral aspect of the mastoid bone. It is unnecessary to
remove the mastoid to the point at which the
semicircular canals are identified. It is necessary to
visualize the dura mater medial to the sigmoid sinus so
it can be cauterized. The dashed lines on the figure
indicate the two parallel incisions that are made above
and below the transverse sinus. It is important to
ensure that these incisions are approximately 7 to 10
mm lateral to the sinus to allow a good edge for
subsequent sewing of the dural grafts. These incisions
are made by first cauterizing the dura mater and then
incising it with a #11 blade knife. A cottonoid is placed
through the incision to protect the underlying
parenchyma of the brain. The incisions are extended
laterally with a small dissecting scissors or knife. As
the junction between the transverse and sigmoid
sinuses is approached, the dura mater becomes more

Neurosurgery Books
 Figure 5-18.
Step 5. The dura mater has been cauterized
and divided as far as the junction of the
transverse and sigmoid sinuses. Cottonoids have
been placed to protect both the cerebellum and
the occipital lobe. The next step is to section the
medial tentorium cerebelli. This is difficult
because the sigmoid sinus is to be preserved and,
therefore, cannot be rotated outward as a handle.
It is best to use the microscope for good
illumination. An incision is made in the tentorium
with a #11 blade knife. The underlying brain
parenchyma is protected with cottonoids. The
incision is carried to the petrous bone using
cautery and the #11 blade knife.

Neurosurgery Books
 Figure 5-19.
Step 6. At this point, in addition to the occipital and
suboccipital dura mater having been sectioned, the
tentorium cerebelli has been divided to the petrous apex.
The bipolar cautery is used to extensively cauterize the
dura mater lying along the inner side of the petrous bone.
The dura mater is cauterized from both an occipital and a
suboccipital approach. It is important to irrigate the
surgical field while cauterizing to prevent burning a hole
through this lateral dura mater. If the petrosal sinus is
opened inadvertently, it is packed with a small piece of
hemostatic fabric. The dura mater lateral to the sigmoid
sinus that was revealed by the earlier unroofing of the
mastoid sinus is also cauterized. Afterward, the two dural
openings are closed with dural grafts.

Neurosurgery Books
Chapter 6
Modified Pterional
Approach

Neurosurgery Books
Chapter 6: Modified Pterional Approach

The modified pterional craniotomy, or anterior temporal approach, developed by Sundt provides
good access to the upper basilar artery and the basilar caput. One of the major principles of the modified
pterional approach is that the temporal lobe is retracted in an anterior-to-posterior direction to allow
access to an aneurysm along the floor of the middle cranial fossa. This type of anterior-to-posterior
retraction is well tolerated as long as the Sylvian fissure is widely divided. Another advantage of this
approach is that it usually is possible to repair the aneurysm without a cutout clip, which sometimes can
be difficult to place precisely. Furthermore, it is easier to identify and to dissect the opposite P1 segment
of the posterior cerebral artery through a modified pterional approach than with a subtemporal approach.
When contemplating the approach to a basilar caput aneurysm, two important considerations are the
relationship of the neck of the aneurysm to the posterior clinoid process and the projection of the dome
of the aneurysm. The modified pterional approach is best suited for aneurysms that are within 1.5 cm in
either direction of the posterior clinoid process. For high basilar caput aneurysms, a subtemporal
approach through a frontotemporal zygomatic craniotomy may offer a better line of site, with less
retraction of the brain. For low basilar caput or basilar trunk aneurysms, a subtemporal transtentorial
approach works well.
Approximately one-half of the aneurysms of the basilar caput have a dome that projects posteriorly.
Therefore, the back wall of the aneurysm is related intimately to the thalamoperforating arteries that
235
Neurosurgery Books
arise from the P1 segment and the caput Visualization and dissection of these perforating arteries may be
extremely difficult, especially with a large posteriorly projecting dome. In that case, additional surgical
maneuvers, including temporary occlusion of the basilar artery or deep hypothermic circulatory arrest,
may be necessary to collapse the dome to facilitate the dissection of these perforating arteries off the
back wall of the aneurysm.
The second most common projection is straight superior. Larger superiorly projecting aneurysms
may present with obstructive hydrocephalus. In smaller superiorly projecting aneurysms, dissecting the
thalamoperforating arteries is technically easier than it is for posteriorly projecting domes. The least
common projection is anteriorly toward the dorsum sellae. When performing the dissection, a small
cottonball can sometimes be placed between the basilar artery and the clivus to facilitate the dissection
of the underside of the neck of the aneurysm.

ANATOMY

Perforating arteries originate from the posterior aspect of the basilar artery as it ascends along the
pons. Small paramedian perforating arteries can be found within 2 to 3 mm caudal to the basilar
bifurcation. One perforating artery may arise from the basilar caput. These basilar perforating vessels
enter the interpeduncular cistern along with the posterior thalamoperforating arteries that originate from
the P1 segment of the posterior cerebral artery.
The posterior thalamoperforating arteries, branches of the P1 segment, are asymmetric and may
originate as one or more thalamoperforating trunks that branch within the interpeduncular cistern. The

236
Neurosurgery Books
anterior thalamoperforating arteries originate from the posterior communicating artery. Also, branches of
the P2 segment of the posterior cerebral artery are referred to as the “thalamogeniculate” and
“peduncular perforating arteries.” These perforating arteries perfuse the mammillary bodies, the
posterior thalamus, the hypothalamus, the subthalamus, the substantia nigra, the oculomotor and
trochlear nuclei, the red nucleus, the mesencephalic reticular formation, the pretectum, the rostral medial
floor of the fourth ventricle, and part of the posterior limb of the internal capsule. Injury to these
perforating vessels leads to devastating neurologic outcomes.
The short and the long lateral circumflex arteries originate from the medial aspect of the P1 and P2
segments. They wrap around the midbrain in parallel with the posterior cerebral artery to perfuse the
geniculate bodies, the collicular plate, and the cerebral peduncles.
As emphasized by Sundt, in superiorly projecting aneurysms, the P1 segments are frequently
adherent to the base of the aneurysm, especially when the bifurcation is low. If the P1 segments are
densely adherent to the neck and cannot be sharply dissected free, Drake cutout clips are necessary to
obliterate the aneurysm, with the thalamoperforating arteries passing through the aperture of the cutout
clip.

237
Neurosurgery Books
 Figure 6-1.
With the modified pterional
approach, the patient’s head is rotated
15 to 20 degrees and slightly flexed,
approximately 15 degrees. Therefore,
there is no head extension compared to
a standard pterional craniotomy.
Flexion of the head rotates the
posterior clinoid process forward and
brings the basilar caput into the line of
sight through a middle cranial fossa
trajectory. With this approach, it is
necessary to retract the temporal lobe
in an anterior-to-posterior direction.
The posterior communicating artery
may clutter the microoperative field. If
so, it often can be ligated safely at its
junction with the posterior cerebral
artery as long as its origin from the
internal carotid artery appears to be
sufficiently large and there is no fetal-
type circulation. The posterior
communicating artery and its
perforating branches are then swept
forward.

238
Neurosurgery Books
Figure 6-2.
This approach uses a craniotomy that exposes
more of the floor of the middle cranial fossa than a
standard pterional approach does. Therefore, the
skin incision must curve posteriorly above the ear.
The temporalis muscle is reflected anteriorly with
the skin flap. The lower edge of the craniotomy is
made with a high-speed air drill to remove the
inferior portion of the temporal squamous bone,
thereby allowing a middle cranial fossa approach.
As the operation progresses, the angle of view
progresses laterally. One, initially, the view is
along the Sylvian fissure, to aid in its dissection.
Two, after the temporal lobe has been retracted
posteriorly, the arachnoid along the edge of the
tentorium cerebelli and the oculomotor nerve is
dissected free. Three, after this is accomplished,
the head of the operating microscope is shifted
more laterally as the surgeon works across the
middle cranial fossa to approach the basilar caput.
Because of this shifting line of sight, it is important
that the scrub nurse and the overlying brain table
be sufficiently caudal so that the space along the
patient’s shoulder is open and free. In this way, the
surgeon can sit comfortably with his or her arm resting on the patient’s shoulder, with adequate room for the
microscope head. A standard frontotemporal craniotomy is used for the exposure.

239
Neurosurgery Books
 Figure 6-3.
Step 1. After the frontotemporal
craniotomy has been performed, the dura
mater is opened and tacked to the margins
of the bone. Under the operating
microscope, the Sylvian fissure is widely
divided. The more completely the bridging
arachnoid is divided, the less traction there
will be on the frontal lobe as the temporal
lobe is displaced posteriorly.

240
Neurosurgery Books
 Figure 6-4.
Step 2. The tip of the temporal lobe is
lined with hemostatic fabric (Surgicel) and
cottonoids (Americot) and gently retracted
in a posterior direction. Typically, bridging
veins come off the tip of the temporal lobe
and must be cauterized and divided. After
the temporal lobe has been retracted
posteriorly, any bridging arachnoid deep
along the edge of the tentorium cerebelli
between the frontal and temporal lobes must
be separated. As the temporal lobe is
retracted, brain relaxation is required.
Draining cerebrospinal fluid through a
lumbar needle and administering mannitol
intravenously, help facilitate brain
relaxation. In rare circumstances, the brain
may still be tight and the surgeon will not be
able to obtain good access along the edge of
the tentorium cerebelli. In this case, the tip
of the temporal lobe and uncus, which
contains the amygdala, can be removed. The
risk of memory impairment with removal of
the nondominant medial temporal lobe is
minimal.

241
Neurosurgery Books
 Figure 6-5.
Step 3. Under high magnification, the edge of
the tentorium cerebelli is identified. The arachnoid
overlying it is incised and separated with a fine
dissecting spatula. It may be necessary to place a
small retractor along the underside of the frontal
lobe, which will displace the internal carotid artery
medially. Along the edge of the tentorium cerebelli,
the most important landmark is the oculomotor
nerve. Overlying this nerve is the posterior cerebral
artery, which the surgeon follows to the basilar
caput. After the oculomotor nerve has been
identified, the edge of the tentorium cerebelli is
retracted laterally by placing a tacking suture into
the temporal dura mater. When placing the suture, it
is important to avoid the trochlear nerve and to make
sure that the suture is as tight as possible to provide
maximal retraction of the edge of the tentorium
cerebelli. The best way to tie the suture is to throw
the first two knots in the same direction. This allows
cinching the knot downward. The third throw is done
in reverse to lock the knot.

242
Neurosurgery Books
 Figure 6-6.
Step 4. The medial temporal lobe,
including the uncus, is retracted to better
identify the edge of the tentorium cerebelli.
The oculomotor nerve is identified, and the
posterior cerebral artery is followed medially
to the basilar caput. It is necessary to dissect
the arachnoid (Liliequist’s membrane)
between the oculomotor nerve and the
posterior cerebral artery to provide access to
the basilar trunk. With low-lying bifurcations,
access to the basilar trunk is provided
between the oculomotor nerve and the
tentorium cerebelli. Before dissecting out the
aneurysm, it is best to ensure that the basilar
artery is visualized and free in case a
temporary clip must be placed. In fact, it is
best to have the scrub nurse load the clip and
the surgeon practice placing it to make sure
that an aneurysm clip of correct length has
been chosen. It also is important to have two
suction tips available: a smaller #5 or #7
suction tip is used for the dissection and a
large #15-French suction tip is hooked to the
wall suction and is immediately available
should the aneurysm rupture prematurely.

243
Neurosurgery Books
 Figure 6-7.
Step 5. The posterior communicating artery
is often in the line of sight. It usually can be
ligated close to the posterior cerebral artery
and, along with its perforating arteries, be
swept forward in its own arachnoid sheath
out of harm’s way. Although small wire clips
can be used to ligate the posterior
communicating artery, these clips
occasionally can be in the way. Therefore,
ligation with an 8-0 or 9-0 monofilament
suture is best. Dissection of the aneurysm
starts along the proximal posterior cerebral
artery to identify the thalamostriate arteries
that originate from the P1 segment and their
relationship to the neck of the aneurysm.
These perforating arteries can be swept off as
a sheet in continuity with the perforating
arteries along the basilar caput on the back
side of the neck of the aneurysm, because
they are ensheathed in a thin layer of
arachnoid. Therefore, first identifying the
correct plane along the proximal P1 segment
will aid in dissection of the small perforating
arteries along the back side of the aneurysm.

244
Neurosurgery Books
Figure 6-8.
Step 6. After the thalamogeniculate perforating arteries have been dissected off the ipsilateral
posterior cerebral artery and basilar caput, a piece of absorbable gelatin sponge (Gelfoam) is placed
between them and the neck of the aneurysm. This keeps these small vessels out of harm’s way when the
aneurysm clip is placed.

245
Neurosurgery Books
 Figure 6-9.
Step 7. The contralateral P1 segment is dissected
off the neck of the aneurysm to identify the
thalamogeniculate trunk that originates from that P1
segment. A small ball-tip dissector or angled spatula
works best. The dome of the aneurysm is manipulated
gently with a #5 or #7 suction tip and a cottonoid. If the
surgeon is concerned about rupturing the aneurysm,
there are several options to consider. One, the patient’s
systolic blood pressure can be decreased to 80 to 90 mm
Hg to soften the dome. Two, the basilar artery can be
occluded temporarily just above or below the
oculomotor nerve. Depending on the position of the
basilar caput, placement of a temporary aneurysm clip
can be difficult and may clutter the operative field. After
the contralateral P1 segment has been dissected off the
neck of the aneurysm, a small piece of absorbable
gelatin sponge is placed if a major thalamogeniculate
trunk has been identified. An aneurysm clip is then
placed across the neck of the aneurysm, parallel to the
posterior cerebral arteries. The absorbable gelatin
sponge is removed, and the dome of the aneurysm is
aspirated. It is mandatory to make sure—absolutely sure
—that no perforating vessels are caught within the
aneurysm clip.

246
Neurosurgery Books
 Figure 6-10.
A, In the sequence illustrated here, the
aneurysm ruptured prematurely during dissection.
The best strategy is to change immediately to a
larger suction tip and to capture the dome. The
neck can quickly be dissected free and the
aneurysm clip placed.
B, Alternatively, a temporary aneurysm clip
can be placed on the basilar artery between the
posterior cerebral artery and the superior
cerebellar artery. Depending on the degree of
collateral blood flow through the posterior
communicating arteries, there may be a surprising
amount of back bleeding through the aneurysm
despite temporary occlusion of the basilar artery.
After the aneurysm has been clipped, the temporary
clip is removed to restore blood flow. Thereafter,
the dome of the aneurysm is manipulated to ensure
that all the perforating vessels are free. If
prolonged temporary occlusion is anticipated, it is
best to have the anesthesiologist normalize blood
pressure and to administer a cerebral protective
agent such as thiopental (2 to 3 mg/kg). Increasing
the patient’s blood pressure increases potential
collateral blood flow, and the cerebral protective
agent decreases metabolic demand.
247
Neurosurgery Books
 248
Neurosurgery Books
Chapter 7
Temporal
Approach

Neurosurgery Books
Chapter 7: Temporal Approach
A temporal craniotomy necessary for a subtemporal approach is useful to access aneurysms of the
basilar caput at or below the level of the posterior clinoid process and those involving the upper one-
third of the basilar trunk. This approach also provides good exposure for neoplasms, cysts, and vascular
malformations that involve the tentorial notch, the medial temporal lobe, or the lateral aspect of the
mesencephalon.
Regardless of the lesion, several principles should be adhered to in using the subtemporal approach.
One, it is important to optimize the position of the patient’s head so that it is nearly horizontal and
extended approximately 10 degrees. Maintaining the head in a horizontal position aids in surgical
orientation, and extending the head lessens the need for retraction of the temporal lobe. Two, it is
important to remove residual temporal bone with rongeurs until the floor of the middle cranial fossa is
reached. Three, it is necessary to protect the vein of Labbé, which nearly always is visualized along the
posterior margins of the craniotomy. The vein of Labbé can be buttressed with absorbable gelatin sponge
(Gelfoam) and cottonoids (Americot). Four, it usually is necessary to enhance brain relaxation with the
use of lumbar spinal drainage and mannitol. Fifth, because medial temporal lobe structures are retracted,
the perioperative use of anticonvulsants is prudent. Sixth, some mastoid air cells often are exposed when
the temporal squamous bone is removed with rongeurs. A watertight dural closure decreases the risk of
postoperative otorrhea.

250
Neurosurgery Books
BASILAR CAPUT ANEURYSM
Figure 7-1.
In most cases, the upper basilar artery and
caput should be approached from the right side
in patients who have a dominant left
hemisphere. This decreases the risk of injury to
the dominant temporal lobe, medial temporal
lobe structures, and vein of Labbé. However, in
a patient with a P1-P2 aneurysm on the left
side, a left subtemporal approach is required.
Illustrated here is the vascular anatomy
visualized through a subtemporal approach.
The basilar caput is at the level of the dorsum
sellae in approximately 50 percent of patients,
superior to the dorsum sellae in 30 percent, and
inferior in 20 percent. Approximately 80
percent of all aneurysms of the basilar caput
are within 1.0 cm of the dorsum sellae. In tall
persons, the bifurcation of the basilar artery
tends to be at a lower level, below the level of
the posterior clinoid process. Often, the P1
segments are attached to the neck of the
aneurysm, requiring the use of a fenestrated or
cutout clip. It has been suggested that with
increasing age, the basilar artery becomes
progressively more tortuous and the level of its
bifurcation becomes progressively more inferior.
251
Neurosurgery Books
Figure 7-1 (Cont.)
The perforating arteries include ones that originate from the posterior communicating artery, the posterior
aspect of the basilar bifurcation, and the P1 segment of the posterior cerebral artery. The perforating arteries from
the posterior communicating artery are termed the “anterior thalamoperforating arteries,” and those from the P1
segment are called the “posterior thalamoperforating arteries.” Perforating branches from the P2 segment of the
posterior cerebral artery are termed the “thalamogeniculate” and “peduncular perforating arteries.” Some of the
thalamoperforating arteries supply the mammillary bodies, and the rest go through the interpeduncular fossa to
supply the posterior thalamus, hypothalamus, subthalamus, substantia nigra, red nucleus, oculomotor and
trochlear nuclei, mesencephalic reticular formation, pretectum, rostral medial floor of the fourth ventricle, and part
of the posterior limb of the internal capsule.
Circumflex arteries also originate from the distal P1 and P2 segments and encircle the midbrain, running
parallel and medial to the posterior cerebral artery. The short circumflex arteries primarily perfuse the geniculate
bodies, and the long circumflex arteries supply the superior and inferior colliculi, the geniculate bodies, part of the
cerebral peduncle, and part of the mesencephalic tegmentum. With the subtemporal approach, the primary vein at
risk is the vein of Labbé, which runs along the posterior aspect of the craniotomy.

Neurosurgery Books
 Figure 7-2.
Step 1. The patient’s head is positioned
nearly horizontal at 10 to 15 degrees and
extended approximately 15 degrees. The U-
shaped incision is centered over the ear. To
avoid injury to the facial nerve, it is
important to make sure that the anterior limb
of the incision ends approximately 1.0 cm
above the midportion of the zygoma.
253
Neurosurgery Books
 Figure 7-3.
Step 2. The temporalis muscle is turned with the
skin flap and retracted with fishhooks. A bur hole is
made along the posteroinferior aspect of the exposed
calvarium. By placing this bur hole in the region of
the transverse sinus, a #1 or #3 Penfield dissector
can be used to peel the dura mater off the underside
of the bone flap before the craniotomy is performed,
thereby minimizing the risk of inadvertent injury to
the transverse sinus. After the bone flap has been
removed, it is important to use Adson rongeurs to
remove the inferior temporal squamous bone down to
the floor of the middle cranial fossa. The more bone
that is removed, the better the angle of view along the
temporal floor, with less retraction on the brain.
Thereafter, the dura mater is tacked to the margins of
the bone and then opened in a U-shaped fashion.
Next, the dura mater is retracted by repositioning the
fishhooks previously used to hold the temporalis

254
Neurosurgery Books
Figure 7-4.
Step 3. The vein of Labbé is protected with absorbable gelatin sponge and cottonoids. The mid-
temporal lobe is retracted with a Yasargil self-retaining retractor. During this retraction, the temporal lobe
is elevated approximately 1.5 to 2.0 cm. Brain retraction is facilitated with the use of lumbar spinal
drainage and mannitol. The edge of the tentorium cerebelli is retracted with a toothed pickup, and the
arachnoid is incised and opened with a #11 blade knife. During this incision and dissection, it is important
to identify both the oculomotor and the trochlear nerves to ensure that they are not inadvertently injured.
255
Neurosurgery Books
 Superior cerebellar artery

Posterior cerebral artery

Posterior communicating artery

Ocularmotor nerve

Figure 7-5.
Step 4. The edge of the tentorium cerebelli is tacked lateral to the floor of the middle
cranial fossa with an interrupted suture. The stitch is placed anterior to the trochlear nerve,
which runs along the underside of the tentorium cerebelli. The oculomotor nerve is the
landmark for orientation.

256
Neurosurgery Books
 Figure 7-6.
Step 5. The Yasargil retractor is placed
deeper to elevate the uncus. The oculomotor
nerve is followed medially to identify the
basilar trunk and the posterior cerebral
artery. If a subarachnoid hemorrhage has
occurred, the blood clot is removed gently
with controlled suction and irrigation. The
arachnoid lateral to the oculomotor nerve is
incised to allow identification of the
superior cerebellar artery. Next, Liliequist’s
membrane is divided to remove the blood
clot from the prepontine cistern. If the
bifurcation of the basilar artery is low, it
usually is not necessary to divide the
arachnoid along the underside of the
oculomotor nerve. After the basilar trunk
has been identified, a small spatula is used
to dissect the basilar trunk free should a
temporary clip have to be placed. The
proximal ipsilateral posterior cerebral
artery is then identified. The neck of the
aneurysm is retracted posteriorly with a #7
suction tip and a cottonoid to allow the
contralateral P1 segment to be identified.

257
Neurosurgery Books
 Figure 7-7.
Step 6. A plane of dissection is
created starting with the perforating
arteries that originate from the
ipsilateral P1 segment and working
along the posterior wall of the
aneurysm toward the contralateral P1
segment. This typically requires
retraction of the neck of the aneurysm
anteriorly, using a small suction tip
and cottonoid. Deliberately
maintaining systolic blood pressure at
70 to 80 mm Hg softens the neck.
Alternatively, the basilar trunk can be
occluded temporarily. If temporary
occlusion is chosen, hypotension
should be avoided. Maintaining
reasonable perfusion through the
perforating vessels makes their
identification and subsequent
dissection easier. Note that the
combination of occlusion of the
proximal basilar trunk and
hypotension may increase the risk of
ischemic injury.

258
Neurosurgery Books
 Figure 7-8.
Step 7. The plane of dissection that starts along the ipsilateral P1
segment is carried along the back wall of the aneurysm. Generally,
the perforating arteries that come off both the PI segments and the
basilar caput can be dissected off the neck of the aneurysm as a sheet
because they usually are attached to arachnoid. The vessels at
greatest risk from this exposure are the perforating arteries that
originate from the contralateral PI segment. Accordingly, it is
mandatory to dissect the neck of the aneurysm free from the left PI
segment and to make sure there are no thalamoperforating vessels.
After this dissection has been completed, a piece of absorbable
gelatin sponge can be used to displace these vessels. Before the
aneurysm clip is placed, the back wall of the neck of the aneurysm is
retracted anteriorly and the sheath of perforating arteries is
reinspected, including identifying of the piece of absorbable gelatin
sponge sitting between the neck of the aneurysm and the
contralateral P1 segment. After this has been identified, the
aneurysm clip is placed. If the proximal P1 segment has been
dissected off the neck of the aneurysm, a straight aneurysm clip can
be used. However, if this proximal ipsilateral P1 segment cannot be
dissected off the neck of the aneurysm, a fenestrated clip must be
placed to preserve the posterior cerebral artery and its perforating
arteries. In this case, some of the perforating arteries may run
through the fenestration of the clip. Thus, it is technically more difficult to place a fenestrated clip. After the
aneurysm has been secured, the dome is aspirated with a small needle to ensure that complete obliteration has
been achieved. After the dome has been collapsed, the back wall of the neck of the aneurysm and aneurysm clip
are reinspected to make sure that no perforating arteries have been entrapped.
259
Neurosurgery Books
 Figure 7-9.
Illustrated here is the use of a
temporary basilar clip during
intraoperative rupture of an aneurysm.
Under ideal circumstances, the best clip
placement is between the posterior
cerebral artery and the superior
cerebellar artery. Placement of a
temporary clip can in itself be difficult
depending on the relationship of the
basilar artery to the tentorium cerebelli.
Specifically, placement of a small clip is
desirable because it is less likely to be
in the way during subsequent dissection.
However, depending on the degree of
exposure, it is possible that only a very
long straight clip can be used to occlude
temporarily the basilar artery. Thus,
before beginning dissection on the
aneurysm, the surgeon should choose
several potential aneurysm clips to be
used to occlude temporarily the basilar
artery and to practice placing them on
the basilar trunk to determine which
clip will work and will not be in the way
of the dissection. After the appropriate
temporary clip has been chosen, it
should be loaded on the aneurysm clip
applier for quick use.
Neurosurgery Books
 Figure 7-10.
If the aneurysm ruptures, the suction
tip is used to aspirate the dome. The
surgeon can then quickly maneuver the
dome with the suction tip while dissecting
off the perforating vessels.

261
Neurosurgery Books
 Figure 7-11.
llustrated here is the typical use
of a Drake-type or fenestrated
aneurysm clip thal encircles the
proximal P1 segment. When placing
this clip, it is important to make sure
not only that the opposite P1 segment
is patent but that the perforating
arteries that come off the ipsilateral
P1 segment have not been entrapped.
In attempting to preserve the opposite
P1 segment, it often is necessary to
leave a small remnant of the neck of
the aneurysm to prevent
encroachment.

262
Neurosurgery Books
POSTERIOR CEREBRAL ARTERY ANEURYSM

Figure 7-12.
Most aneurysms of the proximal
posterior cerebral artery tend to occur along
the P2 segment, as the artery runs adjacent
to the tentorium cerebelli. Perhaps trauma to
the posterior cerebral artery by compression
against the tentorium cerebelli promotes the
development of aneurysms in this location. In
addition to the posterior cerebral artery, the
vessels at risk include the thalamogeniculate
arteries that may originate from the P2
segment. It is important to emphasize that the
temporal lobe must be retracted carefully
when using a subtemporal approach to
aneurysms in this location. Excessive
retraction along the temporal lobe may tear
the dome of the aneurysm, which usually is
attached to or embedded in the uncus.

263
Neurosurgery Books
 Figure 7-13.
The exposure is identical to that used for an
aneurysm of the basilar caput. It is reasonable to
place a small temporary clip along the posterior
cerebral artery just distal to the posterior
communicating artery if space permits. The neck of
the aneurysm is dissected free from the posterior
cerebral artery. It is not necessary to work the
dome of the aneurysm out from the temporal lobe.
Either a right angled clip placed parallel to the
posterior cerebral artery or a straight clip placed
perpendicularly can be used. The angled clip is
superior in preserving patency of the posterior
cerebral artery. After the aneurysm clip has been
placed, it is important to make sure that the distal
posterior cerebral artery is patent. A small micro
Doppler probe is useful in determining whether the
artery is patent.

264
Neurosurgery Books
SUPERIOR CEREBELLAR ARTERY ANEURYSM

Figure 7-14.
Step 1. The subtemporal approach to
an aneurysm of the superior cerebellar
artery is identical to that used for an
aneurysm of either the basilar caput or the
proximal posterior cerebral artery. After
the appropriate craniotomy has been
completed, the vein of Labbé is protected
and the temporal lobe is gently retracted.

265
Neurosurgery Books
 Figure 7-15.
Step 2. The arachnoid between the posterior
cerebral artery and the superior cerebellar artery is
incised after the location of the trochlear nerve has
been identified. A #11 blade knife is used to incise
the arachnoid. A small ball-tip dissector can be used
to peel the arachnoid anteriorly toward the basilar
trunk. If space permits, it is best first to identify the
basilar trunk proximal to the superior cerebellar
artery in case a temporary clip has to be applied.

266
Neurosurgery Books
 Figure 7-16.
Step 3. A small angled ball-tip dissector works
well for dissecting the neck of the aneurysm off both
the posterior cerebral artery and the superior
cerebellar artery.

267
Neurosurgery Books
 Figure 7-17.
Step 4. It is important to dissect the oculomotor
nerve off the neck of the aneurysm. A small
piece of absorbable gelatin sponge can be
placed between the neck of the aneurysm and
the oculomotor nerve and the superior
cerebellar artery to keep these structures safe.
A straight aneurysm clip is used to obliterate
the neck of the aneurysm. As in all aneurysm
repairs, it is important to aspirate the dome of
the aneurysm to make sure that complete
occlusion has occurred.

268
Neurosurgery Books
BASILAR TRUNK ANEURYSM

Figure 7-18.
The aneurysm of the basilar trunk
illustrated here occurs close to the
origin of the anterior inferior
cerebellar artery. Through a
subtemporal approach, it will be
necessary to incise the tentorium
cerebelli to obtain exposure of the
posterior cranial fossa. The variation
in the basilar artery and its perforating
arteries is great. In addition to the
major branches, many paramedian and
short and long circumflex pontine
arteries arise from the basilar trunk.
The paramedian arteries originate from
the dorsal aspect of the basilar artery
and supply the pons. Occlusion of these
paramedian arteries can produce
significant dysfunction, including
hemiplegia, quadriplegia, and
disconjugate eye movements. Occlusion
of one of the circumferential vessels
will also likely produce cerebellar

269
Neurosurgery Books
 Hypoglossal nerve Vertebral artery

Vagal nerve Posterior inferior cerebellar artery

Glossopharyngeal nerve

Abducens nerve

Facial and vestibulocochlear nerve

Trigeminal nerve
Troclear nerve
Hypoglossal nerve Anterior inferior cerebellar artery

Superior cerebellar artery

Hypoglossal nerve
Posterior cerebellar artery

Posterior communicating artery

Anterior choroidal artery

Middle cerebral artery

Figure 7-19.
Anterior cerebral artery Illustrated here are the neurovascular relationships
along the basilar trunk that need to be considered when
repairing aneurysms of the basilar trunk.

270
Neurosurgery Books
 Figure 7-20.
Step 1. A standard
subtemporal craniotomy is
performed. The vein of Labbé is
protected. The mid-temporal lobe
is retracted gently to identify the
edge of the tentorium cerebelli
adjacent to the cerebral peduncles.
The trochlear nerve is identified
beneath the arachnoid. Next, the
tentorium cerebelli is incised
posterior to the trochlear nerve,
approximately 1.5 cm behind the
petrous ridge, for a distance of
approximately 2.0 cm. The anterior
leaf of the incision is cauterized,
folded forward, and tacked to the
floor of the middle cranial fossa.
The incision in the tentorium
cerebelli is carried up to, but not
through, the petrosal vein.

271
Neurosurgery Books
 Aneurysm

Trigeminal nerve

Basilar artery

Facial and vestibulocochlear nerves

Figure 7-21.
Step 2. The arachnoid underneath the tentorium cerebelli is
opened with a #11 blade knife. The trigeminal nerve is medial, and the
facial nerve and the vestibulocochlear nerve are lateral in this
exposure. After the petrosal vein has been visualized, it should be
cauterized and divided. It may be necessary to extend the incision in the
tentorium cerebelli laterally after the petrosal vein has been divided.
272
Neurosurgery Books
 Figure 7-22.
Step 3. A, A small retractor is placed
between the trigeminal nerve medially and
the facial and vestibulocochlear nerves
laterally. If the aneurysm has hemorrhaged,
blood can be removed gently with suction
and irrigation.
Step 4. The anterior inferior cerebellar
A artery is identified laterally and followed to
the basilar artery.
Step 5. B, Aneurysms in this location
typically arise at the junction of the anterior
inferior cerebellar artery and the basilar
trunk. Although they usually project laterally,
the configuration of the aneurysm with the
anterior inferior cerebellar artery can be
complicated. The dome of the aneurysm is
retracted gently with a #5 or #7 suction tip
and cottonoid, and a spatula is used to
dissect the neck of the aneurysm free.

C C, A small straight or curved clip works

well in this location. However, depending on
B the degree of exposure, a long straight clip
may be needed to obliterate the neck of the
aneurysm.
273
Neurosurgery Books
TENTORIUM MENINGIOMA

Figure 7-23.
Tumors such as meningiomas along
the edge of the tentorium cerebelli
typically have extensions into both the
middle and the posterior cranial fossae.
Smaller tumors can be approached
directly through a subtemporal approach,
with sectioning of the tentorium cerebelli.
However, for large tumors that have a
significant extension into the posterior
cranial fossa, a subtemporal approach
may need to be combined with a
suboccipital approach. Accordingly, it is
necessary to plan an incision that allows
the surgeon access to both the middle and
posterior cranial fossae. The position of
the patient’s head is similar to that
described for an aneurysm of the basilar
trunk, in which the head is 10 to 15
degrees to the horizontal and extended
approximately 15 degrees. A question-
mark incision behind the ear can be
extended down if a suboccipital
craniotomy is necessary.
274
Neurosurgery Books
 Figure 7-24.
A Step 1. A, A subtemporal approach is used in
the series of illustrations shown here. After the
vein of Labbé has been protected, a Yasargil
retractor is used to elevate the temporal lobe
gently to identify the tumor along the tentorium
cerebelli. These tumors are vascular, deriving their
blood supply from the dura mater of the middle
and posterior cranial fossa. Accordingly, before
the tumor is debulked, it is best to attack this blood
supply by vigorously cauterizing its dural
attachment using a bipolar cautery.
Step 2. B. After the blood supply of the tumor
B has been cauterized, the tumor is debulked with
curets, an ultrasonic aspirator, or a laser.
275
Neurosurgery Books
 Figure 7-25.
Step 3. After some of the tumor
has been debulked, the tentorium
cerebelli is incised posterior to the
petrous ridge. A cottonoid can be
placed along the underside of the
edge of the tentorium cerebelli to
prevent inadvertent injury to the
trigeminal nerve and petrosal vein.
276
Neurosurgery Books
 Figure 7-26.
Step 4. After the tentorium
cerebelli has been incised, the
medial edge of the tentorium and
the tumor are retracted laterally
to dissect the underside of the
tumor off the arachnoid that
overlies the trigeminal nerve, the
pons, the facial nerve, and the
vestibulocochlear nerve. The
petrosal vein is often seen
running into the underside of the
tumor, close to the trigeminal
nerve, and must be cauterized
and divided.
277
Neurosurgery Books
 Anterior inferior cerebellar artery Figure 7-27.
Step 5. After most of the tumor
has been debulked and removed,
the extension of the tumor along
the underside of the tentorium
cerebelli in the posterior cranial
Oculomotor nerve
fossa is visualized. Ring curets
usually can be used to scrape the
tumor off the dura mater. If the
tumor has a large extension into
the cerebellopontine angle, a
suboccipital craniotomy should
be performed to allow better
access to this extension into the
posterior cranial fossa. Because
of the relationship of the tumor
to the cranial nerves,
Trochlear nerve
intraoperative
electromyographic monitoring of
the trigeminal, facial, and
vestibulocochlear nerves is
necessary. Respecting the
arachnoid layer will protect
these cranial nerves and the
adjacent vascular structures.

278
Neurosurgery Books
Chapter 8
Lateral Suboccipital
Approach

Neurosurgery Books
Chapter 8: Lateral Suboccipital Approach
ARTERIAL SUPPLY

The arterial supply of the posterior fossa consists of the vertebral and basilar arteries and their
branches. The pertinent anatomic features of the arterial and venous systems are illustrated in Figures
8-1 to 8-3.

280
Neurosurgery Books
 Posterior cerebral artery

Superior cerebral artery

Optic tract

Posterior communicating artery

Oculomotor nerve
Trochlear nerve
Basilar Artery

Trigeminal nerve

Anterior inferior cerebellar artery

Abducens nerve
Nervus intermedius
Vestibulocochlear nerve
Glossopharyngeal nerve
Vagus nerve
Hypoglossal nerve

posterior inferior cerebral artery

Vertebral artery Spinal accessory nerve

Anterior spinal artery

Neurosurgery Books
 Inferior sagittal sinus
Internal cerebral veins
Vein of Galen
Basal vein of Rosenthal Precentral vein
Superior vermain vein

Posterior mesencephalic vein Straight sinus

Lateral mesencephalic vein

Anterior pontomesencephalic
vein
Petrosal (Dandy’s) vein
Superior sagittal sinus

Anterior medulary vein

Hemispheric vein

Inferior vermian vein

Anterior spinal vein

282
Neurosurgery Books
Quadrangular lobule
Precentral vein

Primary fissure Culmen

Lobulus simplex
Superior vermian vein

Superior posterior
cerebellar vein

Superior semilunar lobule

Declive
Inferior semilunar lobule Tuber vermis
Inferior posterior cerebellar vein
Horizontal fissure

283
Neurosurgery Books
Vertebral Artery

The vertebral artery originates as the first branch of the subclavian artery and enters the transverse
foramen of C-6. At C-3, the artery turns laterally and enters the foramen of C-2. After exiting through
the foramen of C-1, it curves behind the atlantooccipital joint to lie horizontally along the posterior arch
of C-1. This horizontal segment is called the “pars atlantica.” As the vertebral artery courses medially, it
curves rostrally and passes through the foramen magnum. Along the transverse segment of the vertebral
artery, there are usually muscular branches that anastomose with branches of the external carotid artery.
The posterior meningeal branch often originates from the middle portion of this segment. It also has
been reported that the posterior inferior cerebellar artery originates from the pars atlantica in
approximately 5 percent of patients.
After the vertebral artery penetrates the dura mater, it goes laterally and then ventrally or anteriorly
to join the contralateral vertebral artery to form the basilar artery. In approximately two-thirds of
patients, this union occurs at or just below the inferior pontine sulcus. In approximately 15 percent of
patients, one vertebral artery is dominant, and in 5 to 10 percent a nondominant vertebral artery ends as
the posterior inferior cerebellar artery and does not join the contralateral vertebral artery.
Branches of the vertebral artery include the posterior inferior cerebellar artery, direct perforating
branches to the medulla, the anterior spinal artery, small meningeal branches, and, occasionally, a
posterior spinal artery.

284
Neurosurgery Books
Posterior Inferior Cerebellar Artery

The posterior inferior cerebellar artery is quite variable. In approximately 50 percent of patients, it
originates from the proximal third of the intracisternal portion of the vertebral artery. In 5 to 10 percent
of patients, one posterior inferior cerebellar artery is absent. In these patients, the ipsilateral anterior
inferior cerebellar artery is hypertrophic or dominant and perfuses the area normally supplied by the
posterior inferior cerebellar artery. Occasionally, both posterior inferior cerebellar arteries may be
absent; alternatively, there can be duplications of the vessels.
The first segment of the posterior inferior cerebellar artery goes laterally around the medulla and is
called the “anterior medullary segment.” The artery then curves caudally to form a loop along the laleral
aspect of the medulla; this is called the “lateral medullary segment.” This segment typically courses
between the rootlets of the spinal accessory nerve. At the posterior margin of the medulla, the posterior
inferior cerebellar artery turns superiorly to complete the caudal loop. The portion of the artery that
passes superiorly to form a cranial loop is called the “posterior medullary segment.” Branches that
originate from the apex of the cranial loop go to the choroid plexus of the fourth ventricle and the
cerebellar tonsil. Distal to the origin of these branches, the posterior inferior cerebellar artery crosses the
cerebellar tonsil to form the “supratonsillar segment.” Originating from the supratonsillar segment are
medial branches that supply the vermis and lateral branches that supply the cerebellar hemisphere.

285
Neurosurgery Books
Basilar Artery

In approximately 80 percent of patients, the left and right vertebral arteries join to form the basilar
artery. This occurs along the anterior surface of the medulla, close to the pontomedullary junction. With
increasing age, the basilar artery tends to become more curved and, sometimes, ectatic. If the vertebral
artery is large on one side, the basilar artery is usually deviated to the opposite side. It also may have
fenestrations. The basilar artery extends rostrally along the ventral surface of the pons and terminates in
the interpeduncular cistern by dividing into the left and right posterior cerebral arteries. In
approximately one-half of patients, the basilar caput is at the level of the posterior clinoid process.
Branches of the basilar artery include the anterior spinal, pontine, labyrinthine, anterior inferior
cerebellar, superior cerebellar, mesencephalic, and posterior cerebral arteries. The internal auditory or
labyrinthine artery originates from the anterior inferior cerebellar artery in approximately 85 percent of
patients. In the other 15 percent, it originates from the trunk of the basilar artery. The labyrinthine artery
joins the vestibulocochlear nerve. Medial and lateral pontine branches originate from the trunk of the
basilar artery. Other pontine branches originate from both the anterior inferior cerebellar and superior
cerebellar arteries. These pontine branches perfuse the pons and medulla.

Anterior Inferior Cerebellar Artery

The anterior inferior cerebellar artery originates, in most patients, from the proximal two-thirds of
the basilar artery. Typically, the left and right anterior inferior cerebellar arteries originate at the same
level. Duplication of one of these arteries occurs in approximately 10 to 15 percent of patients. The

286
Neurosurgery Books
artery courses laterally and inferiorly over the anterior surface of the pons. The artery often has a second
loop adjacent to the internal acoustic meatus. At its origin in about 80 percent of patients, the anterior
inferior cerebellar artery is anterior to the origin of the abducens nerve. As mentioned above, the
labyrinthine artery is a branch of the anterior inferior cerebellar artery. Branches of the anterior inferior
cerebellar artery perfuse the pons, middle cerebellar peduncle, flocculus, tegmentum, and cerebellar
hemisphere.

Superior Cerebellar Artery

In 85 percent of patients, the left and right superior cerebellar arteries arise from the basilar artery as
a single trunk. Variations include two smaller branches of the superior cerebellar artery, one of which
runs medially and the other laterally; origin from the posterior cerebral artery; and dual trunks that unite
to form a single superior cerebellar artery. At its origin, the superior cerebellar artery is separated from
the posterior cerebral artery by the oculomotor nerve. The segments of the superior cerebellar artery
include the mesencephalic segment below the oculomotor nerve, the lateral pontine mesencephalic
segment below the trochlear nerve and separated from the posterior cerebral artery by the edge of the
tentorium cerebelli, the cerebellar mesencephalic segment located between the cerebellum and midbrain,
and the cortical segment. The lateral branch of the superior cerebellar artery supplies the superior aspect
of the cerebellar hemispheres, the superior cerebellar peduncle, the dentate nucleus, and a portion of the
middle cerebellar peduncle. The medial branch perfuses the rostral surface of the cerebellar hemisphere
and the vermis.

287
Neurosurgery Books
VENOUS DRAINAGE

The venous drainage of the posterior fossa can be divided into three systems: the Galenic system,
anterior petrosal system, and posterior (or tentorial) system.
Galenic System

The Galenic system includes the precentral cerebellar vein, the superior vermian vein, the anterior
pontomesencephalic vein, and the posterior and lateral mesencephalic veins.
The precentral cerebellar vein is located in the midline in the fissure between the lingula and the
central lobule of the vermis. It is formed by the union of veins from the brachium pontis and vermis. It
sits posterior to the collicular plate and precentral lobule of the vermis to enter the vein of Galen. It is an
important landmark during supracerebellar approaches to the pineal and tectal plate region.
The superior vermian vein curves along the anterior surface of the culmen and receives tributaries
from the vermis. It also enters the vein of Galen, either with or just adjacent to the precentral cerebellar
vein. The posterior mesencephalic vein originates from the outer aspect of the cerebral peduncle and
curves around the brain stem in the ambient cistern. Often, it is adjacent to the basal vein of Rosenthal.
One of the main tributaries to the posterior mesencephalic vein is the lateral mesencephalic vein, which
is an important landmark for identifying the junction of the tegmentum and the cerebral peduncle. The
anterior pontomesencephalic vein is a complex of small veins that runs along the ventral and rostral
surfaces of the pons and mesencephalon. This vein curves along the interpeduncular fossa below or

288
Neurosurgery Books
caudal to the cerebral peduncle to drain into either the basal vein of Rosenthal or the posterior
mesencephalic vein.
Petrosal System

The petrosal system of veins receives multiple tributaries from the cerebellum, pons, and medulla.
They unite to form the petrosal vein. The petrosal vein, or “Dandy’s vein,” is the most important
cisternal bridging vein encountered during cerebellopontine angle operations. It drains laterally, adjacent
to the trigeminal nerve, into the superior petrosal sinus. It collects venous blood from the middle and
superior cerebellar peduncles, ventral cerebellum, pons, flocculus, medulla, and lateral region of the
fourth ventricle. These tributary veins collect close to the dorsal root entry zone of the trigeminal nerve
and form the bridging petrosal vein. Despite the relatively large drainage territory of the petrosal vein, it
can be cauterized and divided, with low risk of venous infarction.
Posterior System

The posterior system of veins drains directly posterolaterally into the torcular Herophili. The inferior
vermian vein curves posterosuperiorly along the inferior surface of the vermis and receives blood from
the hemispheric veins of the cerebellar tonsils and caudal cerebellum.
The inferior petrosal sinus originates in the posterior region of the cavernous sinus. It travels along
the apex of the petrous ridge, where it is crossed by the abducens nerve in Dorello’s canal, and then runs
down along the clivus to enter the jugular foramen. The left and inferior petrosal sinuses communicate
with each other through the basilar plexus, which is located along the ventral aspect of the clivus. Also,
some small veins from the hypoglossal canal, condyle region, and foramen magnum enter the inferior
289
Neurosurgery Books
petrosal sinus. When a far lateral suboccipital approach is used, with removal of the occipital condyle,
these veins can be the source of considerable nuisance bleeding.
Acoustic neurilemomas are divided into four grades, depending on size. The relevance of the
grading system is related to the prognosis regarding preservation of the facial nerve. Grade 1 tumors are
intracanalicular in location and have a longitudinal diameter of 1 to 10 mm. Grade 2 tumors are both
intracanalicular and intracisternal and have a longitudinal diameter up to 20 mm. Grade 3 tumors are
intracisternal, have a longitudinal diameter of 30 mm, and abut the pons. Grade 4 tumors have a
longitudinal diameter greater than 30 mm and displace the pons. In grade 3 or 4 acoustic neurilemomas,
the facial nerve is displaced anteriorly or ventrally in approximately 70 percent of patients, rostrally in
10 percent, inferiorly in 10 percent, and posteriorly or dorsally in approximately 5 percent. Most
commonly, the facial nerve is ventral and slightly rostral along the tumor capsule, but near the internal
acoustic meatus, the nerve tends to be more posterior or dorsal. Therefore, when initially debulking the
tumor, it is important to stay away from the porus acoustica. In approximately two·thirds of patients
with acoustic tumors not related to Recklinghausen’s disease, the facial nerve is preserved as a thin
bundle, and in the other one·third, it is thin and splayed against the tumor capsule. In patients with
Recklinghausen’s disease, the facial nerve most commonly is splayed and thin, and the risk of facial
nerve paralysis is greater. In most patients, the tumor originates from either the inferior or superior
vestibular nerve, but subsequently it may involve the cochlear division.

290
Neurosurgery Books
ACOUSTIC NEURILEMOMA

Figure 8-4.
A gentle S-shaped incision provides
access to the cerebellopontine angle. The
extent of bony removal, through either a
craniotomy or craniectomy, depends on
the type of lesion and surgical approach.
For most lesions, it is best to identify the
junction of the transverse and sigmoid
sinuses. This visual identification
facilitates the surgeon’s orientation. The
asterion, which is formed by the junction
of the lambdoid suture and the temporal
squamous suture, lies over this junction
and is a useful landmark when planning
bony removal. Also, there may be a
mastoid emissary vein that originates
from the proximal part of the sigmoid
sinus. For resection of an acoustic
neurilemoma, bony removal should
extend down to the foramen magnum.
Electromyographic monitoring of the
cranial nerves is essential when
resecting an acoustic neurilemoma.

291
Neurosurgery Books
 Figure 8-5.
Two patient positions can be used for
resecting acoustic neurilemomas. A, If the
sitting position is chosen, the head should be
slightly flexed and rotated toward the
surgeon by 15 to 20 degrees. This rotation
helps decrease the need for cerebellar
retraction and shortens the distance between
the tumor and the surgeon. In this way, the
surgeon’s arms are less extended, and they
can be placed more comfortably on an
armrest. B, If the supine position is chosen,
it is important not to overrotate or
overextend the patient’s head, because this
may cause compression of the contralateral
jugular vein and possibly an increase in
intracranial pressure. Propping up the
patient’s ipsilateral shoulder decreases the
degree of neck rotation. However, it is
important that this “propped up” shoulder
not protrude too high, otherwise it will
interfere with the surgeon comfortably
resting his or her arm on the patient’s
shoulder while resecting the tumor.

Neurosurgery Books
 Figure 8-6.
Step 1. The asterion is identified,
and a bur hole is placed. The
craniectomy is performed with Adson
rongeurs. A small angled Kerrison
rongeur can be used to undermine and
to remove bone to identify the junction
of the transverse sinus and the sigmoid
sinus. However, when placing the foot
plate of the Kerrison rongeur
underneath the bone, the surgeon must
be careful, especially in young children,
not to tear the sigmoid sinus, which
may interdigitate with the overlying
bone. An alternative is to use a high-
speed air drill and diamond bur.
Removing sufficient bone laterally to
identify the first 4 to 6 mm of the edge
of the sigmoid sinus decreases the
degree of retraction on the cerebellum.
The mastoid air cells are waxed. After
the dura mater is closed, a piece of
muscle is harvested and placed against
the mastoid air cells for reinforcement.

Neurosurgery Books
Figure 8-7.
Step 2. The dura mater is opened and tacked to the margins of the muscle. When approaching
large tumors of the cerebellopontine angle, a useful maneuver is to incise the arachnoid over the
cisterna magna. This provides immediate release of cerebrospinal fluid and considerably enhances the
ease of exposure, It is best to incise the arachnoid closest to the cerebellum. In this way, the arachnoid
will drape over and protect the lower cranial nerves during the operation. Rarely with massive tumors,
it may be necessary to partially resect the lateral 1.0 cm of the cerebellar hemisphere to gain exposure.

294
Neurosurgery Books
 Figure 8-8.
Step 3. The exposed cerebellum is lined with
hemostatic fabric (Surgicel), after which a
cottonoid (Americot) is placed. The hemostatic
fabric prevents the cottonoid from being stuck to
the cerebellum at the end of the operation. Next,
the cerebellum is gently retracted with a #7
suction tip. When first gaining access to the
cerebellopontine angle, rotation of the operating
table improves the angle of microscopic view.
Arachnoid adhesions occur between the
cerebellum and tumor capsule. These adhesions
should be peeled back toward the cerebellum
with a bipolar forceps or a small flat dissector.
Several tributaries of the petrosal vein usually
lie on the cerebellum and pons. These veins are
extremely useful in identifying the pial surface
and, thus, the junction between the pons and the
tumor capsule. It is important to keep these veins
and arachnoid adhesions with the cerebellum
and pons. This pial layer merges with the thin
layer of arachnoid that separates the facial and
vestibulocochlear nerve complex from the tumor
capsule. If the pial plane becomes obscure
during the dissection, the surgeon should dissect
another aspect of the tumor and work back
toward the area of difficulty.
295
Neurosurgery Books
 Figure 8-9.
Step 4. Typically, the trigeminal
nerve is displaced rostrally and
sometimes ventrally. An arachnoid
layer separates the trigeminal nerve
from the tumor capsule. This
arachnoid is continuous with the
arachnoid overlying the facial and
vestibulocochlear nerve complex. It
also is continuous with the pial layer
of the pons. Therefore, the trigeminal
nerve is dissected off the tumor
capsule with its arachnoid intact.
Similarly, the arachnoid overlying
the lower cranial nerves is preserved.
The arachnoid over the
glossopharyngeal and vagus nerves
is continuous with the arachnoid
overlying the facial and
vestibulocochlear nerve complex. It
also is continuous with the pial layer
of the lower pons.

296
Neurosurgery Books
 Figure 8-10.
Step 5. Tumor debulking is done
with a bipolar cautery or an
ultrasonic aspirator. It is
important not to be overly
aggressive along the ventral
aspect of the tumor when
debulking the center. Otherwise,
the surgeon may inadvertently
punch through the capsule or
the deep side of the tumor and
injure the facial and
vestibulocochlear nerve
complex. It is important to
recognize that an ultrasonic
aspirator can cause vibratory
injury to the facial nerve.
Moreover, aggressive bipolar
cautery without irrigation can
cause heat injury.

297
Neurosurgery Books
 Figure 8-11.
Step 6. After the tumor has been
debulked, the suction tip with a small
cottonoid is used to gently rotate and
lift the tumor away from the pons.
However, before this maneuver is
performed, it is important to identify
the arachnoid over the trigeminal nerve
and lower cranial nerves and the place
where it merges with the pial surface of
the pons is identitied. Often, the thin
facial nerve is translucent. However, a
very thin layer of arachnoid usually
separates the facial and cochlear
nerves from the tumor capsule. As the
tumor is retracted laterally, a small
spatula is used to push the thin
arachnoid away from the tumor
capsule, and an attempt is made to
maintain the integrity of this layer, with
the more clearly defined arachnoid
along the underside of the trigeminal
nerve and the arachnoid lying along
the upper aspect of the lower cranial
nerves.

298
Neurosurgery Books
 Figure 8-12.
Step 7. At the junction of the
pons with the tumor capsule,
bipolar stimulation in
combination with
electromyographic monitoring
is used to identify the facial
nerve. With a spatula or
forceps, the arachnoid
adhesions are dissected or
peeled with the facial nerve.
These adhesions are never
sharply dissected off the tumor
capsule.

299
Neurosurgery Books
 Figure 8-13.
Step 8. By step 8. the majority
of tumor has been removed. The
tumor within the internal acoustic
meatus is exposed with careful
drilling of the porus acoustica. If it
is difficult to identify the facial
nerve along the tumor capsule
during the initial dissection, the
porus acoustica should be removed
earlier. The facial and cochlear
nerves then can be identified and
followed medially into the
cerebellopontine angle.
300
Neurosurgery Books
 A

Superior vestibular nerve

Inferior vestibular nerve

Nervus intermedius
Cochlear nerve

Facial nerve

Figure 8-14.
A, The normal anatomy within the internal acoustic meatus. The internal acoustic meatus is covered
by a thin layer of dura mater. Within it are the facial and vestibulocochlear nerves, including the nervus
intermedius. Typically, there is at least one small labyrinthine artery, and in approximately one-half of
patients, a lateral loop of the anterior inferior cerebellar artery enters the proximal portion of the
internal acoustic meatus. The facial nerve is against the anterosuperior aspect of the internal acoustic
meatus, and the cochlear nerve is against the anteroinferior aspect. The superior and inferior vestibular
nerves are posterior to the facial and cochlear nerves. The superior and inferior vestibular nerves are
separated by the transverse crest. The small nervus intermedius is intimately related to the facial nerve
and usually posterior to it, adjacent to the cochlear nerve.

301
Neurosurgery Books
B B, After the porus acoustica has been removed with a high-speed air drill, the
thin dura mater is incised with a sickle knife.
C, Next, normal-appearing superior and inferior vestibular nerves are
identified and sectioned at their junction with the tumor, and then the facial and
cochlear nerves are identified in the internal acoustic canal. The tumor is gently
dissected off these nerves with a small spatula. As the facial nerve exits from the
internal acoustic meatus to enter the subarachnoid space, it usually is quite thin
and compressed against the opening or rim. The surgeon works back and forth
from both a brain stem and internal acoustic meatus perspective until the tumor
is dissected off the arachnoid that overlies the thin facial nerve. After the tumor
has been resected, the facial nerve is stimulated
at its origin from the brain stem to assess C
residual conduction. Some mild bleeding usually
occurs along the facial nerve. This is treated
best with small bits of hemostatic agent, such as
absorbable gelatin sponge (Gelfoam) or
hemostatic fabric, followed by irrigation,
instead of bipolar cautery. After immaculate
hemostasis has been achieved, the surface of the
cerebellum is inspected to ensure that there are
no retraction hemorrhages. Next, the dura mater is closed with a graft in a
watertight fashion, and the mastoid air cells are rewaxed. If a craniotomy was
performed, the bone flap is reapproximated with small titanium plates, wires, or
monofilament sutures. If a craniectomy was performed, a new bone plate can be
fashioned with either acrylate or newer bone matrix materials. According to
anecdotal experience with removal of acoustic neurilemoma, replacement of a
bone flap decreases postoperative headache. Before suturing, the muscle is
injected with 0.25 percent bupivacaine to decrease postoperative neck pain.

Neurosurgery Books
CEREBELLOPONTINE ANGLE MENINGIOMA Figure 8-15.
Step 1. The patient position is similar to
that for resection of an acoustic neurilemoma.
Specifically, a supine, park bench, or sitting
position can be used. The craniectomy or
craniotomy is identical to that used for an
acoustic neurilemoma: the junction between
the transverse sinus and the sigmoid sinus is
identified, and the bone is removed as far as
the foramen magnum. Approximately 2 cm
medially from the transverse sinus and
sigmoid sinus, the dura mater is opened and
tacked to the margins of the muscle. The rest
of the dura mater is left intact to protect the
cerebellum, keeping it out of harm’s way. The
cerebellum is lined with hemostatic fabric and
cottonoids and gently retracted. Lateral
rotation of the operating room table facilitates
entry into the cerebellopontine angle, with less
retraction on the cerebellum. Typically, the
trigeminal nerve is displaced rostrally at the
margins of the exposure adjacent to the
petrosal vein. The facial and
vestibulocochlear nerves are usually displaced
dorsally and encountered during the initial
retraction of the cerebellum. The lower
cranial nerves are displaced caudally.
303
Neurosurgery Books
Figure 8-16.
Step 2. A, The trajectory or approach for resecting the tumor depends on
the way the cranial nerves have been displaced. In this illustration, the
facial and vestibulocochlear nerves have been displaced rostrally.
Therefore, the tumor will initially be approached caudally, below the
facial and vestibulocochlear nerves. If during the initial exposure the
facial nerve has been displaced caudally, the tumor is approached
between the trigeminal and the facial and vestibulocochlear nerves.
Regardless of the approach, it is valuable to preserve the arachnoid that
overlies the facial and vestibulocochlear nerves. This arachnoid layer
helps to protect these nerves and their vascular supply from injury.
During the operation, brain stem auditory evoked potentials are a
valuable tool in helping to preserve audition. For example, an increasing
latency in wave 5 during the resection suggests that cerebellar retraction
is excessive. In this case, the retractor blade should be loosened. Constant
monitoring of the facial nerve during the dissection prevents excessive
manipulation. During any exposure of the cerebellopontine angle, it is
important to watch the petrosal vein. Often, it is best to cauterize
“Dandy’s vein” if it appears to be under tension or it obscures exposure.
When cauterizing this vein, it is important to cauterize it completely and to cut it only partially with a
microscissors, after which the vein is again cauterized and then severed. Bleeding or oozing may occur from the
petrosal vein at its junction with the transverse sinus. It is better to treat this with absorbable gelatin sponge
packing instead of attempting cauterization. Occasionally, bleeding may occur from the top of the cerebellum
because of tearing of small bridging veins. These veins usually cannot be visualized directly. Packing with
absorbable gelatin sponge is also useful in this situation. Following the resection, it is important to remove the
absorbable gelatin sponge from the top of the cerebellum and to irrigate to ensure that no hematoma has been
trapped. B, After the arachnoid has been dissected off the tumor capsule, an incision is made in the midportion of
the meningioma, and a bipolar cautery, laser, or ultrasonic aspirator is used to core out the tumor.
304
Neurosurgery Books
 Figure 8-17.
Step 3. After the tumor has been
debulked, a small ball tip or similar
dissector is used to dissect the cranial
nerves off the tumor capsule. The lower
cranial nerves usually can be swept off
in a sheath of arachnoid. Dissecting the
facial and vestibulocochlear nerves off
the tumor capsule must be done with
care. During this dissection, it is
important to press against the tumor
capsule instead of against the nerves.

305
Neurosurgery Books
A

Figure 8-18.
Slep 4. A, After the tumor has been dissected
off the cranial nerves, additional debulking is
performed. Next, a small spatula is used to dissect
the tumor off the trigeminal nerve. Also, a spatula
or ball-tip dissector is used to dissect the tumor off
the distal facial and vestibulocochlear nerves.
Remember that preserving the arachnoid over the
B facial and vestibulocochlear nerves helps to protect
these structures. The tumor usually flattens these
nerves against the internal acoustic meatus.
Step 5. B, The residual nubbin of tumor along
the petrous ridge is cauterized with a bipolar
cautery and then removed. Small curets can be
useful in removing less well visualized tumor from
around corners. After the tumor has been resected,
it is important to meticulously cauterize the dural
attachment of the tumor in a wide fashion to
decrease recurrence.
306
Neurosurgery Books
MICROVASCULAR DECOMPRESSION FOR TRIGEMINAL NEURALGIA

Figure 8-19.
Step 1. A straight or gentle curve incision is used in
this exposure. The craniectomy must identify the junction
between the transverse sinus and the sigmoid sinus. It does
not need to extend down to the foramen magnum. The first
bur hole is placed at the asterion. An Adson rongeur is
used to bite bone laterally until the junction of the
transverse and sigmoid sinuses is identified. Although
electromyographic monitoring is not mandatory, it
provides a degree of security in protecting the cranial
nerves. Placement of a lumbar needle can be helpful in
younger patients. Specifically, withdrawal of
approximately 10 to 20 mL of cerebrospinal fluid through
the lumbar needle facilitates cerebellar displacement.
Often, a retractor is not necessary. One disadvantage of
the lumbar needle is the risk of postoperative low-pressure
headaches. After the craniectomy, the dura mater is opened
in a curved fashion. The cerebellum is lined with
hemostatic fabric and a cottonoid and gently retracted with
a #7 suction tip. The arachnoid bulging over the trigeminal
nerve is incised. The immediate drainage of cerebrospinal
fluid provides adequate relaxation of the cerebellum. The
surgeon’s view and approach to the trigeminal nerve can
be improved by rotating the operating room table from side
to side.
307
Neurosurgery Books
Figure 8-20.
Step 2. A, With gentle retraction of the A
cerebellum, the petrosal vein is encountered. If the
petrosal vein is under tension or is at risk of being
injured during the dissection, it should be cauterized,
partially divided, recauterized, and then completely
severed. From this operative approach, the facial and
vestibulocochlear nerves are visualized at the caudal
end of the exposure. The arachnoid over these
structures should be left intact.
Step 3. B, It is mandatory to clearly identify and
expose the dorsal root entry zone of the trigeminal
nerve. For patients with pain in the ophthalmic or
maxillary distribution, the site of vascular
compression is usually along the rostral side of the
dorsal entry zone; however, if the pain is in the
mandibular distribution, the site of compression is B
located caudally. A focal area of discoloration or
demyelination is often present at the site of vascular
compression. A small dissector or spatula is used to
gently dissect the tethering arachnoid and to lift off
the vascular loop.
Step 4. C, Before permanently displacing the
C
vascular loop, it is important to examine the
circumference of the dorsal root entry zone to ensure

308
Neurosurgery Books
 Step 4. C, (cont.) there are no other sites of vascular compression. Next, a small piece of polytef (Teflon) is
used to displace the vascular compression off the dorsal root entry zone. Some surgeons use cotton or another type
of packing material, including muscle, and others place a suture in the underside of the tentorium cerebelli and
create a sling around the artery.
If a site of vascular compression cannot be identified, a difficult decision must be made about what the next
step should be. There are several alternatives. First, the nerve can be compressed three to five times with a bipolar
forceps. Such traumatic compression often is successful in alleviating the pain. Second, the trigeminal nerve can be
partially sectioned. If the patient has pain in the maxillary or mandibular distribution, the lower one-half to two-
thirds of the nerve can be sectioned. However, if the pain is in the ophthalmic distribution, it is better to compress
the nerve instead of performing a rhizotomy, which may result in corneal anesthesia dolorosa. Before either nerve
compression or rhizotomy is considered, the surgeon must be absolutely certain that there is no arterial or venous
compression. The possibility of not finding any arterial or venous compression and the alternatives to this should
be discussed thoroughly with the patient before the operation

309
Neurosurgery Books
MICROVASCULAR DECOMPRESSION FOR HEMIFACIAL SPASM
Figure 8-21.
Step 1. Bone removal for treatment of hemifacial
spasm should extend lower than that typically used for
surgical treatment of trigeminal neuralgia. It is
valuable to outline the junction of the transverse sinus
and sigmoid sinus at the upper end of the craniectomy.
The rationale for having more inferior exposure of
bone is that the facial nerve should be approached from
a caudal direction, between it and the
glossopharyngeal nerve. This is contrary to
approaching the facial nerve directly laterally or
rostrally below the trigeminal nerve. The latter two
approaches have an increased risk of causing hearing
loss. Electromyographic monitoring for hemifacial
spasm is required for several reasons. First, it helps
preserve cochlear nerve function. Second, patients with
hemifacial spasm have an abnormal electromyographic
lateral spread reflex. Lifting off the arterial
compression leads immediately to the loss of the lateral
spread reflex. Thus, the surgeon immediately has
intraoperative confirmation that the correct vascular
compression has been identified. This is in contrast to
microvascular decompression for trigeminal neuralgia,
in which there is no direct confirmation that successful
decompression has been achieved.
310
Neurosurgery Books
Figure 8-22.
Step 2. A, The dura mater is opened in a curved fashion and A
tacked to the margin of the muscle. The cerebellum is lined with
hemostatic fabric and a cottonoid and gently retracted. As mentioned
above, the facial nerve is approached from a more caudal or inferior
approach, above the glossopharyngeal nerve. A lumbar needle can
be placed to drain cerebrospinal fluid to enhance the ease of
exposure. Alternatively, the cisterna magna can be opened during the
initial retraction of the cerebellum for cerebrospinal fluid drainage.
After gentle retraction of the cerebellum, the arachnoid below the
facial and vestibulocochlear nerves is incised.
Step 3. B, In most circumstances, the vascular compression is B
due to a loop of the posterior inferior cerebellar artery. The vascular
loop is gently lifted off the dorsal root entry zone of the facial nerve
with a small dissector. There typically are a few small branches of the
posterior inferior cerebellar artery that supply the facial nerve; these
must be preserved. After the vascular loop has been dissected and
lifted off the facial nerve, electromyographic monitoring should show
an immediate loss of the lateral spread reflex. The surgeon can let the
vascular loop fall back onto the facial nerve to determine whether the
C
lateral spread reflex returns. This is additional confirmation that the
correct vascular compression has been identified.
Step 4. C. The vascular loop is displaced from the dorsal root entry zone by packing the space with polytef.
During this packing, it is important for the technician to continue to monitor electromyographically the lateral
spread reflex. The return of this reflex during the packing is a strong indication that the vascular loop had slipped
back onto the facial nerve.
311
Neurosurgery Books
PONTINE CAVERNOUS HEMANGIOMA

Figure 8-23.
Step 1. A. The craniotomy or craniectomy for
exposure of a pontine cavernous hemangioma is
similar to that for microvascular decompression of the
facial nerve. Specifically, the bony exposure must
identify the junction between the transverse sinus and
the sigmoid sinus for orientation. It also should extend
sufficiently inferiorly or caudally that the surgeon can
identify the glossopharyngeal and vagus nerves. After
the dura mater has been opened, the cerebellum is
lined with hemostatic fabric and cottonoids and gently
A displaced with a small tapered retractor. With most
pontine cavernous hemangiomas, some hemosiderin
or bluish discoloration of the pial surface is apparent
under high magnification.
B, The best place to enter the cavity of the pontine
hematoma is between the trigeminal and facial nerves
in an avascular portion of the pons. Although a large
area of the pial surface may be stained with
hemosiderin, it is best to enter at the point of bluish
B discoloration that indicates the location of the
hematoma cavity.
312
Neurosurgery Books
 Figure 8-24.
A Step 2. A, The pial surface of the pons is
cauterized and incised vertically over a length of
approximately 6 to 7 mm. Two small spatulas are
moved in a back and forth manner to deepen the
incision until the hematoma capsule is identified.
Step 3. B, The hematoma cavity is entered and
the clot removed. Within the clot, the cavernous
hemangioma is identified as a purplish-reddish
mulberry lesion that usually is small, approximately
4 to 5 mm. Generally, three to five small arterial
feeding vessels supply the cavernous hemangiomas,
which should be cauterized individually and divided
with a microscissors.
Step 4. C, After the cavernous hemangioma has
B been isolated, it is removed with a small cup
forceps.

313
Neurosurgery Books
ANEURYSMS OF THE POSTERIOR INFERIOR CEREBELLAR ARTERY
Figure 8-25.
Step 1. A, The position of the patient is the same as that used for removing
an acoustic neurilemoma. Either a sitting or supine position can be used. A
A supine or park bench position is used most often; however, for lower
lying aneurysms of the posterior inferior cerebellar artery, a sitting
position actually works better. One of the risks of this surgical procedure is
injury to the vagus nerve. With low-lying aneurysms approached through a
sitting position, the cerebellar tonsil is lifted up and the aneurysm is
approached behind the vagus nerve instead of going through the nerve,
which often is necessary when the patient is supine or in the park bench
position. Regardless of the position of the patient, the craniectomy or
craniotomy must be similar to that for an acoustic neurilemoma. It is best
to identify the junction between the transverse sinus and the sigmoid sinus
for surgeon orientation. Also, it is mandatory to remove bone down to the
foramen magnum. After the dura mater has been opened, the cerebellum is
gently retracted and the lower cranial nerves are identified. It is best to B
open up the arachnoid caudally, below the spinal accessory nerve. This
allows early identification and isolation of the vertebral artery.
Step 2. B, The dissection is carried distally along the vertebral artery until
the aneurysm is identified. In almost all circumstances, the
glossopharyngeal, vagus, and spinal accessory nerves lie over the aneurysm to some extent. Ideally, it is best to
work either below or above the vagus nerve instead of working through its rootlets. When working from a more
rostral approach, the glossopharyngeal nerve will serve as a buffer and protect the vagus nerve. After the
proximal neck of the aneurysm has been identified, a small piece of absorbable gelatin sponge can be placed to
preserve the separation of the neck from the vertebral artery.
314
Neurosurgery Books
A
Figure 8-26.
Step 3. A, The distal neck of the aneurysm must
be identified. Nearly always, it is intimately associated
with the origin of the posterior inferior cerebellar
artery. As illustrated here, the vagus nerve lies over
the neck; therefore, the surgeon must work through the
rootlets of the nerve.
Step 4. B, After the neck has been identified, the
aneurysm is repaired, usually with a small curved or
B bayonet clip.
Step 5 C, The dome of the aneurysm is aspirated
and the origin of the posterior inferior cerebellar
artery is inspected. A small Doppler probe is valuable
in confirming patency and flow through both the
posterior inferior cerebellar and vertebral arteries.
During the immediate postoperative convalescence,
the patient should be observed for potential airway
C obstruction or aspiration from possible injury to the
vagus nerve.
315
Neurosurgery Books
ANEURYSMS OF THE VERTEBROBASILAR JUNCTION

Figure 8-27.
Illustrated are the two patient positions used for exposure of an aneurysm of either the posterior
inferior cerebellar artery or the vertebrobasilar trunk.

316
Neurosurgery Books
 Figure 8-28.
Step 1. Two different incisions can be
used for exposure of an aneurysm of the
vertebrobasilar junction. Bone removal
must extend from the junction of the
transverse and sigmoid sinuses down to
the foramen magnum. Furthermore, it
usually is best to remove the medial third
of the occipital condyle. A
hemilaminectomy of C-l must also be
performed. It is easier to expose the
foramen magnum and the occipital
condyle through the inverted S-shaped
incision. Note that in the sitting position,
the head is rotated approximately 20 to 25
degrees. For a large or giant aneurysm of
the vertebrobasilar junction or basilar
trunk, deep hypothermic circulatory
bypass is usually required to provide the
degree of brain relaxation and protection
necessary for repair of these difficult
aneurysms. If this is the case, the supine
or park bench position is required.

317
Neurosurgery Books
 Figure 8-29.
Step 2. Following the
hemilaminectomy of the arch of C-1, the
vertebral artery is identified, and the
medial one-third of the occipital condyle
is removed with a high-speed air drill and
diamond bur. Next, the dura mater is
opened, as shown here.

318
Neurosurgery Books
 Figure 8-30.
Step 3. A, Note that lateral bone removal allows
direct access to the cerebellopontine angle. If the
lateral exposure is not sufficient to identify the
vertebrobasilar junction after the cerebellum has
been retracted, the surgeon must consider a
presigmoid sinus approach. This requires additional
removal of bone over the sigmoid sinus, with a partial
mastoidectomy.
Step 4. B, The arachnoid is incised over the
cisterna magna to drain cerebrospinal fluid to aid
cerebellar relaxation.

319
Neurosurgery Books
 Figure 8-31.
Step 5. The cerebellum is retracted
medially and upward. The lower
cranial nerves are identified. The
posterior inferior cerebellar artery is
followed proximally to the vertebral
artery.

320
Neurosurgery Books
 A

Figure 8-32.
Step 6. A, The cerebellum is retracted
more aggressively. The vertebral artery is
B followed distally to identify the
vertebrobasilar junction.
Step 7. B. The cerebellum is further
retracted. It is usually necessary to gently
retract the lower pons to better identify the
vertebrobasilar junction. Aneurysms in this
location generally point caudally. The
aneurysm is manipulated gently with a #5 or
#7 suction tip and cottonoid. A small spatula
is used to dissect the neck of the aneurysm and
to identify the anterior spinal artery, which
likely is present.

321
Neurosurgery Books
 Figure 8-33.
Step 8. After the aneurysm has been
dissected, a small straight or slightly
curved aneurysm clip is used to obliterate
the neck, and the dome of the aneurysm is
aspirated to ensure that successful
clipping has been achieved. The anterior
spinal artery is inspected to make sure
that its origin has not been compromised.

322
Neurosurgery Books
JUVENILE CEREBELLAR ASTROCYTOMA

Figure 8-34.
Step 1. Juvenile cerebellar astrocytomas can be
approached through several positions, such as those
illustrated here. Bony removal should expose most of the
involved cerebellar hemisphere and, thus, extend from
the sigmoid sinus to the midline and down to the
foramen magnum.

323
Neurosurgery Books
 Figure 8-35.
Step 2. A, After the dura mater has
been opened, the folia of the cerebellum
are inspected. With cystic cerebellar
astrocytomas, the folia overlying the cyst
cavity are usually expanded and enlarged.
An incision is made parallel to the folia.
Step 3. B, The cyst wall is identified
and entered. Drainage of the
xanthochromic fluid leads to immediate
relaxation of the cerebellum. A small
retractor can be used to keep the cyst
cavity open for inspection.

324
Neurosurgery Books
 Figure 8-36.
Step 4. With juvenile astrocytomas,
the neoplastic portion is a fleshy
nodular·appearing mass along one
wall of the cyst. It is distinct from
the compressed gliotic cerebellar
tissue. Surgical cure is achieved by
resecting the neoplastic mass. It is
not necessary to resect the entire
cyst wall. However, a healthy
margin of approximately 0.5 cm
from the edge of the fleshy
neoplastic mass should be used
during resection to ensure
complete removal of the tumor.

325
Neurosurgery Books
CEREBELLAR HEMANGIOBLASTOMA
Figure 8-37.
Step 1. The extent of bony removal for a cerebellar
hemangioblastoma is similar to that for resection of other
hemispheric lesions, such as juvenile astrocytoma and metastatic
tumor. Specifically, much of the bone over the involved cerebellar
hemisphere should be removed, extending laterally from near the
A
midline to the sigmoid sinus.
Step 2. A, After the dura mater has been opened, the
cerebellum is inspected. Most hemangioblastomas are associated
with a large cyst. Therefore, the folia are often enlarged and
expanded over the epicenter of the cyst. An incision is made in an
B expanded folium and in parallel with the folium.
Step 3. B, The cyst is encountered and entered. Drainage of the
xanthochromic fluid provides immediate relaxation of the
cerebellum. Occasionally, it is necessary to use a small brain
spatula to keep the incision open within the cyst after the
cerebellum collapses. The interior of the cyst is inspected until the
reddish mulberry hemangioblastoma is identified.
Step 4. C, After the hemangioblastoma has been identified, the
C small feeding arteries are cauterized and divided. The
hemangioblastoma is gently retracted with a #7 suction tip and
cottonoid, and the bipolar cautery is used to separate the tumor
from the surrounding cerebellar cyst cavity.

326
Neurosurgery Books
CEREBELLAR METASTASIS Figure 8-38.
Step 1. A metastasis to the cerebellar hemisphere
can be approached with the patient in either a
sitting or a supine position. Either a midline or
lateral incision over the involved hemisphere can
be used. Bone removal should be generous,
extending laterally from close to the midline to
almost the transverse and sigmoid sinuses. It also
should extend down close to the foramen
magnum. This large bony removal will afford
some decompression of the edematous
cerebellum. For a metastasis in the vermis, a
midline incision with a bilateral suboccipital
craniectomy should be performed.
Step 2. The dura mater is opened in a curved
fashion and tacked to the margins of the muscle.
Often, inspection of the cerebellar surface is not
revealing about the location of the metastatic
tumor. If available, ultrasonography or
stereotaxis can be useful in pinpointing the
location of the tumor and minimizing the incision
in the cerebellum. The incision in the cerebellum
should be parallel with the folia and
approximately 1 to 1.5 cm long. Most metastatic
lesions are encountered within 1.0 cm of the pial
surface.
327
Neurosurgery Books
 Figure 8-39.
Step 3. After the tumor has been
located, a bipolar cautery and #5 or #7
suction tip are used to separate the
tumor from the surrounding cerebellar
tissue. If necessary, small brain
spatulas can be used to gently retract
the cerebellum. Some metastatic
tumors, such as renal cell carcinoma,
can be extremely vascular. The best
way to control bleeding is to not enter
the tumor but to stay on its periphery,
quickly separating it from the
surrounding cerebellum.

328
Neurosurgery Books
 Figure 8-40.
Step 4. After the tumor has
been extracted, the resection
bed is examined for hemostasis.
Secondary closure of the dura
mater with pericranium,
muscle, or graft is effective
treatment if the cerebellum
looks edematous.

329
Neurosurgery Books
Chapter 9
Midline Suboccipital
Approach

Neurosurgery Books
Chapter 9:
Midline Suboccipital Approach

ANATOMY

The pertinent neurosurgical anatomy for midline suboccipital approaches is illustrated in Figures 9-1
to 9-3.

331
Neurosurgery Books
 Third ventricle

Posterior commissure
Thalamus
Habenula

Pineal body

Intercollicular sulcus

Superior colliculus

Inferior colliculus Superior cerebellar peduncle

Trochlear nerve Facial colliculus

Taenis pontis

Lingula Middle cerebellar peduncle

Rhomboid fossa Flocculus

Hypoglossal trigone

Vagal trigone

Obex

Fasciculus gracilis

Fasciculus cuneatus

332
Neurosurgery Books
 Posterior commissure

Internal cerebral vein Inferior sagittal sinus

Septum of corpus callosum
Fornix Pineal body
Basal vein of Rosenthal
Thalamus
Great cerebral vein
Superior colliculs
Inferior colliculus
Anterior commissure
Culmen
Primary fissure
Lamina terminalis
Optic chiasm Declive

Mammillary body
Superior sagittal sinus
Pituitary

Oculomotor nerve Straight sinus

Aqueduct
Folium
Central lobe
Tuber
Superior medullary velum
Lingula
Fourth ventricle Transverse sinus

Nodulus

Uvula Pyramis

333
Neurosurgery Books
 Denticulate ligament

334
Neurosurgery Books
CHIARI MALFORMATION

A
Figure 9-4.
Illustrated are two of
the most common patient
positions used for midline
suboccipital approaches.
A, The sitting position
works well for several
reasons. First, compared
with the prone position,
there is less intrathoracic
pressure and, thus, better
venous return.
Accordingly, brain
relaxation is excellent.
Second, for less
experienced surgeons, the
sitting position facilitates
surgeon orientation. Third,
in the sitting position,
blood drains out from the lower end of the incision, but in the prone position blood coming from the cut edge of
muscle tends to run into the operative field. Disadvantages of the sitting position include the increased risk of
air embolus, pneumocephalus, and brain collapse, resulting in subdural hygroma or hematoma.

335
Neurosurgery Books
B

Figure 9-4 (cont.)

B, If the prone position is used, it is important to make sure that the chest rolls are optimally placed to
prevent an increase in intrathoracic pressure. Before the patient’s head is flexed, it is necessary to confirm that
the endotracheal tube is adequately secured. An advantage of the prone position is that the surgeon’s assistant
can be involved more actively in the operation instead of standing to the side, which occurs if the patient is in
the sitting position.
With the sitting position, a central venous catheter, transthoracic Doppler, and transesophageal
echocardiography (TEE) are used. If there is Doppler or TEE evidence of a venous air embolus, the
anesthesiologist should attempt to aspirate the air through the central venous catheter. At the same time, the
surgeon should immediately inspect the operative field to determine the source of the air embolus. The most
common sites for air absorption into the venous circulation are through bone diploë, veins located laterally

Neurosurgery Books
Figure 9-4 (cont.)
along the craniovertebral junction and between vertebrae C-1 and C-2, and the cut edge of muscle.
Accordingly, the surgeon should rewax the bone edges, place absorbable gelatin sponge (Gelfoam) in the
epidural space between vertebrae C-1 and C-2 laterally and at the craniovertebral junction, and press wet
gauze against the edge of the muscle. If nitrous oxide is used as an anesthetic, it is discontinued. If the above
surgical maneuvers do not stop the source of air absorption, jugular vein compression should be performed to
facilitate identification of the air leak by causing bleeding from the site that is the source of entry of air into
the systemic circulation. Simultaneously, the surgical bed is rotated into a more Trendelenburg position, which
lowers the head in relationship to the heart but still allows the surgeon to visualize the wound. If TEE
continues to show air absorption, if end-expired carbon dioxide decreases, if end-expired nitrogen increases, if
arterial blood pressure begins to decrease, or if TEE shows paradoxical air embolus in the left ventricle, the
surgeon must consider rapid closure of the surgical wound. In practice, the need to close a wound because of
air embolus with the patient in the sitting position is rare.

Neurosurgery Books
 Figure 9-5.
Step 1. In the case of a Chiari
malformation, the main goal of the
operation is decompression of the
posterior fossa. A midline incision is
made from the inion down to the cervical
lamina one level below the lowest extent
of the cerebellar tonsils, as identified on
preoperative imaging. It is uncommon
for the incision to extend below the level
of vertebra C-3. The craniectomy should
extend rostrally approximately 1 to 2 cm
below the torcular and transverse sinus
and laterally approximately 1.0 to 1.5
cm from the sigmoid sinus. At the
foramen magnum, bone removal should
be carried lateral to the occipital
condyles. A laminectomy of C-1 is also
performed. Depending on the extent of
the tonsillar herniation, it may be
necessary to perform additional cervical
laminectomies.

338
Neurosurgery Books
 Figure 9-6.
Step 2. Note that in the process of exposing
the calvarium, a triangular pedicle of muscle
and fascia is left attached to the external
occipital protuberance. This will facilitate
muscle closure at the end of the operation. In the
example shown here, a laminectomy of both C-1
and C-2 has been performed. The dura mater is
opened in the standard Y-shaped fashion, which
is used in almost all midline approaches to the
posterior fossa. Some patients have a midline
occipital sinus that can be the source of bleeding
when the dura mater is opened. In this case, it is
best to open the dura mater over each cerebellar
hemisphere first before crossing the midline.
After the dura mater has been opened, the dural
leaves are tacked to the margins of the muscle, If
there is oozing from the torcular, a piece of
absorbable gelatin sponge or muscle can be
packed against it before the upper dural fold is
tacked.

339
Neurosurgery Books
 A
Figure 9-7.
Step 3. A, In Chiari malformations, the
arachnoid is often thickened, and it acts like
tethering bands against the cerebellar tonsils at the
foramen magnum. Therefore, it is necessary to
incise the arachnoid. A small jeweler’s forceps can
be used to hold the arachnoid while a sharp #11-
blade knife or microscissors is used to cut the
arachnoid. After a hole has been made in the
arachnoid, cerebrospinal fluid immediately drains
through it. B, The rest of the arachnoid can easily be
opened with a small ball-tip dissector or knife.
B Sometimes, arachnoid adhesions tether the posterior
inferior cerebellar artery to the dura mater,
requiring sharp dissection.
340
Neurosurgery Books
 Figure 9-8.
Step 4. After the arachnoid has been
opened, some surgeons tack it to the dura mater.
An alternative is to excise the arachnoid over
the cerebellar tonsils. Thereafter, the fourth
ventricle should be inspected to make sure there
are no arachnoid adhesions that prevent
communication of the fourth ventricle with the
cisterna magna. In patients with an associated
syrinx, a small ball-tip dissector can be used to
identify the obex and to make sure that
adhesions do not prevent the flow of
cerebrospinal fluid out of the syrinx into the
subarachnoid space.

341
Neurosurgery Books
 Figure 9-9.
Step 5. After the cerebellar tonsils have
been decompressed and the floor of the fourth
ventricle and the obex have been inspected, a
loose dural graft is sewn in place in a watertight
fashion. Options for this graft include fascia lata
harvested from the leg, a bovine or cadaver
dural graft, or possibly pericranium. It usually is
difficult to obtain a piece of pericranium large
enough to cover the dural opening for this
particular operation, using a midline incision.

342
Neurosurgery Books
SUBOCCIPITAL SUPRACEREBELLAR APPROACH TO PINEAL REGION TUMORS

Figure 9-10.
Step 1. This operation can be performed with the
patient in either a sitting or prone position. However,
the sitting position offers an easier approach to
tumors in this region. Regardless of body position, it
is essential to make sure that the head is maximally
flexed so that the tentorium cerebelli is parallel with
the surgeon’s line of sight. This will lessen the need
for downward retraction of the cerebellum. When
positioning the patient’s head, the anesthesiologist
typically places two fingers underneath the chin to
prevent overflexion. Excessive flexion of the head
leads to kinking and compression of the jugular
veins, which can increase intracranial venous
distention and intracranial pressure. A midline
incision is used, similar to the one described above
for treating a Chiari malformation. However, bone
removal is extended more rostrally to expose
approximately 5 mm of the torcular and transverse
sinus. It also is advantageous to perform a
laminectomy of C-1 to decompress the foramen
magnum. The arachnoid over the cisterna magna
should be opened for drainage of cerebrospinal fluid
to provide relaxation of the cerebellum.
343
Neurosurgery Books
 Figure 9-11.
Step 2. After the arachnoid over the
cisterna magna has been opened, the
cerebellum will sag caudally. The top of
the cerebellum is lined with hemostatic
fabric (Surgicel) and cottonoids and
gently retracted with a #5 or #7 suction
tip and bipolar cautery. It is important
to limit the number of cottonoids used,
otherwise they will narrow the exposure.
Several small veins usually come off the
top of the cerebellum and run to the
underside of the tentorium cerebelli;
they should be cauterized and divided.
Note how the torcular and transverse
sinus are retracted upward by tacking
the dura mater, thereby providing an
additional 3 to 4 mm of exposure.
344
Neurosurgery Books
 Figure 9-12.
Step 3. Under higher magnification,
the surgeon extends the dissection deeper
toward the precentral vein. If necessary, a
single retractor blade can be placed over
the roof of the cerebellum. The precentral
vein is ensheathed by arachnoid and
extends to the vein of Galen. The precentral
vein may have several tributaries. This vein
should be completely cauterized and
divided.

345
Neurosurgery Books
 Figure 9-13.
Step 4. After the precentral vein has been
divided, the caudal aspect of the tumor is
usually encountered, covered in arachnoid.
The arachnoid should be divided and peeled
laterally to better define the posterior margin
of the tumor capsule.

346
Neurosurgery Books
 Figure 9-14.
Step 5. There are usually adhesions between the
tumor capsule and the basal veins of Rosenthal and the
internal cerebral veins. A, A dissecting spatula can be
used to separate the tumor capsule from these venous
structures. It is best to incise the tumor and perform
internal debulking. Because of the complexity of tumors
in this region, as much specimen as possible should be
preserved for pathologic examination. B, After internal
debulking has been performed, the anterior margin of the
tumor capsule is dissected off the internal cerebral veins,
which lie rostrally and ventrally to the tumor. Depending
on the type of tumor, complete removal may not be
possible. If necessary, remnants of the tumor should be
left intact instead of injuring the major deep draining
veins.
347
Neurosurgery Books
 Figure 9-15.
Step 6. With the supracerebellar
approach, access to tumors with a
significant caudal extension may be
difficult. It may be helpful to retract
the cerebellum and to rotate the
operating room table to alter the
surgeon’s line of sight. Usually, a
spatula can be used to gently work
the tumor capsule upward into the
operative field. A malleable
fiberoptic scope or mirror may be
useful to inspect this deeper portion
to better understand the dissecting
plane if direct visualization is not
optimum. For tumors with a very
large caudal extension, a combined
occipital transtentorial and
supracerebellar approach can be
used, as described in Chapter 5.

348
Neurosurgery Books
 Figure 9-16.
Step 7. This is the typical view
following removal of a pineal
tumor. The third ventricle is framed
laterally by the splenium and basal
vein of Rosenthal, anteriorly by the
internal cerebral vein, and
inferiorly by the thalamus. It is
important to irrigate the third
ventricle thoroughly to flush out
any debris. The veins should be
inspected to make sure there is no
thinning of the vascular wall. If the
structural integrity of one of the
veins is questioned, the vein should
be buttressed with absorbable
gelatin sponge. The cerebellum is
inspected to make sure there are no
hematomas or subpial
hemorrhage. The author prefers to
close the dura mater with a graft
and not to replace the bone flap.
This provides good decompression
of the posterior fossa.

349
Neurosurgery Books
MEDULLOBLASTOMA
Figure 9-17.
Step 1. The approach to a
medulloblastoma is similar to that for a
Chiari malformation. The patient can be
placed in either a sitting or prone
position. The patient’s head does not need
to be flexed as much as for a
supracerebellar approach to a tumor in
the pineal region. Bony removal is similar
to that for Chiari decompression. In
addition to removing the posterior rim of
the foramen magnum, a laminectomy of
vertebra C-1 should also be performed.
The dura mater is opened in a Y-shaped
fashion and tacked to the margins of the
muscle. Often, the tumor can be seen
bulging out underneath the arachnoid,
splaying the two cerebellar tonsils
laterally. The arachnoid is incised first
over the cisterna magna for drainage of
cerebrospinal fluid. The arachnoid
incision is then extended upward. The
arachnoid can be held laterally by
tacking it to the dura mater with either a
suture or small hemostatic clips.

350
Neurosurgery Books
A

Figure 9-18.
Step 2. A, The caudal, or inferior, portion
of the tumor is separated first from the medulla.
A Penfield #1 dissector works well for this
maneuver.
B, After the tumor has been lifted off the
floor of the fourth ventricle, several cottonoids
are placed between the tumor and the floor of
the ventricle and the cisterna magna. In this
way, tumor debris is more likely to be flushed
B out of the surgical wound, decreasing the risk of
drop metastases entering the spinal canal. After
the caudal portion of the tumor has been
dissected, the plane between the lateral capsule
of the tumor and cerebellar tissue is identified
and defined by placement of cottonoids.

351
Neurosurgery Books
 Figure 9-19.
Step 3. The attachment of the tumor is at the
superior medullary velum. A #5 or #7 suction tip and
bipolar cautery are used to develop the plane over
the rostral and ventral portions of the tumor. It is
important to ensure that the edge of the tumor is
identified and that the dissection is approximately 5
to 10 mm above this edge to allow for a more
aggressive resection of this infiltrative zone. At the
depths of the rostral dissection, the surgeon will
punch through into the fourth ventricle. It is
important not to injure the floor of the ventricle. For
small tumors, during the initial exposure of the
tumor at the foramen magnum and before beginning
the dissection, the surgeon can insert a cottonoid all
the way up along the floor of the fourth ventricle
underneath the tumor. This maneuver is more
difficult to accomplish with large tumors.
With large tumors, it can be advantageous to
retract both cerebellar tonsils with self-retaining
retractors. Alternatively, the tonsil can be retracted
with a suction tip and cottonoid. Using the suction
tip as a retractor is faster because the surgeon can
move it rapidly as the dissection is carried laterally
and anteriorly.

352
Neurosurgery Books
 A

Figure 9-20.
Step 4. A, With large tumors, internal
debulking is advantageous to collapse the
tumor and to decrease the need for
cerebellar retraction. Before the tumor is
debulked, cottonoids should be placed over
B the cisterna magna to prevent tumor debris
from entering the spinal canal.
B, After the tumor has been debulked,
its lateral capsule is peeled off the overlying
cerebellar tissue. In most circumstances,
this plane is easy to establish and preserve.
As tumor resection is carried deeper, the
surgeon must follow this plane carefully to
prevent inadvertent injury to the middle
cerebellar peduncle.
353
Neurosurgery Books
 Figure 9-21.
Step 5. After the tumor has been resected, the
underside of the cerebellar vermis and its extension
into the superior medullary velum should be
inspected thoroughly for any remnants of the tumor.
Thereafter, it is meticulously cauterized. With most
medulloblastomas, the floor of the fourth ventricle
is not invaded. Any thickened or abnormal-
appearing arachnoid along the cisterna magna
should be harvested and examined pathologically
for evidence of tumor dissemination.

354
Neurosurgery Books
EPENDYMOMA Figure 9-22.
Step 1. The position of the patient, the skin
incision, and the bony exposure for resection of an
ependymoma are identical to those used for
medulloblastoma or Chiari malformation. After
the dura mater has been opened, the arachnoid
over the cisterna magna is opened to drain
cerebrospinal fluid to decrease pressure within the
posterior fossa. After the arachnoid has been
opened, the caudal portion of the tumor closest to
the surgeon is dissected or lifted off the
cervicomedullary junction. For large tumors, there
may be adhesions between the tumor capsule and
the arachnoid and vascular structures in this
region that require sharp dissection After this
posterior tongue of tumor has been dissected,
cottonoids are placed between it and the medulla
and the cisterna magna. In sequence, each
cerebellar tonsil is retracted laterally so that the
tumor capsule can be dissected off the overlying
cerebellum. The tumor rarely invades the cerebral
hemispheres; therefore, this plane usually is easy
to identify and to maintain by placing cottonoids
on the lateral side of the tumor capsule. The tumor
is incised and internally debulked using suction,
bipolar cautery, or an ultrasonic aspirator.

355
Neurosurgery Books
 Figure 9-23.
Step 2. It is important to
recognize that ependymomas
originate from the floor of the
fourth ventricle. Therefore, as
debulking of the tumor proceeds
into the fourth ventricle, the
surgeon must exercise every
caution to prevent inadvertent
injury to the medulla.
Furthermore, it may be difficult to
separate the tumor from the
lateral aspect of the ventricular
floor, at its junction with the
middle cerebellar peduncle. As the
surgeon enters the fourth
ventricle, a small cottonball can
be placed rostral to the tumor at
the opening of the aqueduct to
prevent excessive drainage of
cerebrospinal fluid from the
lateral ventricles. This maneuver
is especially helpful when the
patient is in the sitting position
and a mass in the fourth ventricle
is being resected.
356
Neurosurgery Books
 Figuralre 9-24.
Step 3. Under high
magnification, the tumor is
separated from the floor of the
fourth ventricle, but small deposits
of infiltrative tumor will remain.
Creation of this plane is based on
visual cues. Electromyographic
monitoring of the lower cranial
nerves can be useful when
dissecting a tumor off the floor of
the fourth ventricle, because it
alerts the surgeon to the location
of functional tissue along the
facial and vagal trigones.

Neurosurgery Books
FORAMEN MAGNUM TUMOR

Figure 9-25.
Step 1. The position of the patient, the skin
incision, and the bony removal for resection
of a foramen magnum tumor are similar to
those used for treating Chiari malformation.
In a foramen magnum tumor, the extent of
cervical laminectomy is determined by the
caudal extension of tumor into the cervical
spinal canal. It is not necessary to expose
bone up to the torcular or lateral to the
sigmoid sinus. However, at the foramen
magnum, it is necessary that the bony
removal extend to the occipital condyle on
the side of the tumor. The medial one-third of
the occipital condyle can be removed with a
high-speed air drill and small diamond bur,
which limits bleeding from the small bony
venous channels.

358
Neurosurgery Books
 Figure 9-26.
Step 2. After the dura mater has
been opened, the arachnoid over the
cisterna magna is opened and tacked to
the dura. Most foramen magnum
tumors originate from the anterolateral
foramen magnum. Thus, the
cervicomedullary junction is typically
displaced posteriorly and laterally by
the tumor mass. This natural
displacement assists the surgeon, and
retraction of the spinal cord is not
necessary. A single stitch can be placed
in the dentate ligament to preserve this
displacement until tumor resection has
been completed. The rootlets of C-1
and C-2 usually overlie the tumor. If
necessary, these posterior rootlets can
be sectioned to provide additional
exposure. The spinal accessory nerve is
usually displaced laterally against the
dura mater of the posterior fossa. The
position of the vertebral artery varies,
and the artery may be encased by
tumor. This artery usually is anterior,
especially when the tumor originates

359
Neurosurgery Books
 A

Figure 9-27.
Step 3. A, The arachnoid,
which may partially overlie the
tumor capsule, is peeled laterally.
The plane between the medial
aspect of the tumor and the lateral
cervicomedullary junction is easily
identified. It is advantageous to
aggressively debulk the tumor
internally, working above and
below the rootlets of C-1 and C-2.

B Step 4. B, After the tumor has

been debulked, its attachment to the
dura mater of the foramen magnum
should be cauterized to decrease
the vascular supply of the tumor.

360
Neurosurgery Books
 Figure 9-28.
Step 5. As the resection of the
tumor extends deeper into the
foramen magnum, it is important to
identify the vertebral artery, which
may be encased by tumor. A suction
tip and small dissecting spatula
work well for peeling tumor off the
artery. If tumor is adhesive to
either the vertebral or basilar
artery or their perforating vessels,
it may be necessary to leave small
remnants of tumor. This may occur
if a tumor of the foramen magnum
has extended up to involve the
lower two-thirds of the clivus.
361
Neurosurgery Books
Chapter 10
Transsphenoidal
Approach

Neurosurgery Books
Chapter 10 Transsphenoidal Approach

The transsphenoidal route is an excellent approach for resecting various sellar lesions, including
pituitary adenomas, Rathke’s cleft cysts, and intrasellar craniopharyngiomas. On preoperative
radiographic imaging, the best clue that a transsphenoidal approach will work well for resecting a sellar/
suprasellar lesion is enlargement of the sella turcica. Even pituitary macroadenomas with massive
suprasellar extensions can be removed grossly through a transsphenoidal approach. However, an
apparent extension or bulging of a tumor, with displacement of the diaphragma sellae inferiorly, without
enlargement of the pituitary fossa (as often seen in a craniopharyngioma) indicates that the tumor needs
to be removed through a transcranial approach. Cushing initially developed the translabial transseptal
approach to the sella turcica but subsequently abandoned it for a transcranial route because of a high
incidence of leakage of cerebrospinal fluid and meningitis. Transsphenoidal surgery was subsequently
rejuvenated by Guiot, Thibaut, Hardy, and Wigser, with excellent surgical results.
Most transsphenoidal surgery is performed for resection of pituitary adenomas, These tumors may
be either functional or nonfunctional. Approximately 75 percent of all pituitary adenomas are functional,
and 40 to 50 percent are prolactinomas, 15 to 25 percent are growth hormone-secreting tumors, and 5
percent are adrenocorticotropic hormone (ACTH)–producing tumors. Of prolactinomas, approximately
70 to 75 percent of microadenomas and 30 percent of macroadenomas have long-term cure rates after
transsphenoidal surgery, with normalization of serum levels of prolactin. One of the most significant
363
Neurosurgery Books
prognostic indicators for successful prolactinoma surgery is the preoperative level of prolactin.
Specifically, if the serum level of prolactin is less than 500 ng/mL preoperatively, the cure rate after
transsphenoidal surgery is approximately 50 to 70 percent. However, if the preoperative serum level is
greater than 1,000 ng/mL, the cure rate is less
than 25 percent.
The surgical results are similar for growth
hormone–producing tumors. Overall, the cure
rate for somatotroph tumors is approximately 60
percent, with postoperative growth hormone
levels less than 5 ng/mL. However, similar to
prolactinomas, there is an inverse relationship
between preoperative levels of growth hormone and
long-term surgical cure. The success rate for
surgical tumor control appears to decrease
significantly when the preoperative level of
growth hormone is greater than 50 ng/mL.
The surgical treatment of Cushing’s disease is
complicated. Reportedly, approximately 85
percent of patients with biochemical
This is the test I want
documentation of Cushing’s disease have a pituitary adenoma even if the results of preoperative
magnetic resonance imaging are unremarkable. Some experienced surgeons have reported a long-term
364
Neurosurgery Books
cure rate of 80 to 90 percent with noninvasive microadenomas. With large ACTH-producing tumors,
there often is an extension into the cavernous sinus and long-term surgical cure rates accordingly fall
below 50 percent. At surgery, the presence of a microadenoma may be difficult to find because the tumor
occasionally is located within the midportion of the pituitary gland and not lateralized. Also, in some
patients with Cushing’s disease, the source of excessive ACTH production is pituitary hyperplasia
instead of a discrete pituitary adenoma. Cushing’s disease is a life-threatening condition; therefore, the
surgeon should strongly consider performing a complete hypophysectomy if a discrete adenoma is not
found at surgery, because if transsphenoidal surgery for Cushing’ disease fails, additional procedures and
treatments will be required, for example, bilateral adrenalectomy. If there is residual pituitary adenoma,
Nelson syndrome is a risk after adrenalectomy. Consequently, irradiation of the pituitary fossa is often
recommended after adrenalectomy if the original biochemical data clearly support the presence of
Cushing’s disease instead of Cushing’s syndrome. Because of the potential for the sequence of events, a
strong argument can be made for performing a near-complete hypophysectomy if a discrete
microadenoma is not found intraoperatively. A complete hypophysectomy can be performed with
minimal traction on the infundibulum of the pituitary. Thus, the need for vasopressin postoperatively is
often limited. The long-term endocrine implications for management of patients with Cushing’s disease
who undergo hypophysectomy instead of resection of a microadenoma are not dramatic in adult males.
However, in children and in women who might desire to have children, the implications of a complete
hypophysectomy are significant. It is debatable, but in children and in women desiring children, bilateral
adrenalectomy instead of complete hypophysectomy might be best after a failed transsphenoidal
exploration.

365
Neurosurgery Books
ANATOMY

The sella turcica is bounded by the sellar floor anteriorly and inferiorly, the dorsum sellae
posteriorly, the diaphragma sellae superiorly, and the dura mater of the cavernous sinus laterally. The
diaphragma sellae may be incompetent; therefore, a small pouch of arachnoid may be present within the
sella turcica. Within the diaphragma sellae is the circular venous sinus, which has a variable anatomy
and consists of the following venous structures. Typically, small venous channels connect the cavernous
sinuses and run posteriorly to the dorsum sellae. Also, a venous sinus runs anteriorly at the junction of
the tuberculum sellae and the anterior wall of the pituitary fossa. Collectively, these small venous
sinuses are the “circular sinus” and usually are of no consequence during transsphenoidal surgery.
However, in patients with Cushing’s disease, the circular sinus often appears larger than normal and may
be a source of nuisance bleeding during transsphenoidal surgery.
The pituitary gland is divided into an anterior lobe (the pars distalis and the pars intermedia) and a
posterior lobe (the pars nervosa). The pars intermedia is a remnant of Rathke’s pouch, which separates
the anterior and posterior lobes and occasionally may be seen at surgery as a small cleft. The location of
Rathke’s cleft explains why, during transsphenoidal resection of a cyst of Rathke’s cleft, the surgeon first
encounters a thin but normal pituitary gland after removing the sellar floor. The posterior pituitary lobe
is the end of the pituitary stalk, which consists of axons whose cell bodies are in the supraoptic and
paraventricular nuclei of the hypothalamus. This explains why a minimally traumatic hypophysectomy
often produces only transient diabetes insipidus. A thin layer of cells extends from the anterior lobe and
encircles the pituitary stalk to form the pars tuberalis. Immunohistochemical data suggest that

366
Neurosurgery Books
corticotropin cells (ACTH) are located in the midline in the anterior lobe, lactotrophs (prolactinomas)
posterolaterally, somatotrophs (growth hormone) laterally, and thyrotrophs (rare thyroid secreting
tumors) anteriorly.
The blood supply to the pituitary gland is derived primarily from the superior and inferior
hypophysial arteries. The superior hypophysial artery originates from the paraclinoid internal carotid
artery and perfuses the pituitary stalk and hypothalamus. The inferior hypophysial artery originates from
the meningohypophysial trunk and supplies most of the pituitary gland. The venous drainage of the
pituitary gland is a portal system that originates in the hypothalamus and drains into the cavernous sinus
through the lateral hypophysial veins. The parasellar structures, including the cavernous sinus, optic
chiasm, and carotid artery, are discussed in Chapter 1.

367
Neurosurgery Books
Figure 10-1.
For all transsphenoidal operations, the patient is positioned supine, with the operating table flexed and the
patient’s head supported in a horseshoe head rest. The horseshoe is advantageous over a rigid pinion system
because it allows the head to be manipulated during the operation. The patient’s nose should be parallel with the
floor and angled approximately 30 to 45 degrees from the long axis of the body to allow the surgeon to stand
comfortably and to rest his or her hands on the patient’s chest. For large tumors with a suprasellar extension, a
lumbar spinal drain can be placed for intraoperative injection of approximately 10 to 15 mL of air into the thecal
sac. Because the patient’s head is slightly elevated, the air ascends into the suprasellar cistern and helps displace
the tumor downward into the operative field. The abdomen or right thigh is prepared for a possible fat or muscle
graft.
368
Neurosurgery Books
 Figure 10-2.
It is important to use an image intensifier during the
operation. It allows the neurosurgeon to confirm that the
sella turcica has been reached through the
transsphenoidal approach. With large tumors, the sella
turcica is enlarged and thin and its identification is
apparent; however, with microadenomas, it is less
obvious. Radiographic confirmation of the pituitary floor
is important before bone is removed. Inadvertent removal
of the frontal fossa bone at the junction of the anterior
sella turcica has the risk of cerebrospinal fluid leakage
because the dura mater in that area is thin. Unplanned
removal of the upper clivus carries a risk of
neurovascular injury, bleeding from small venous
sinuses, and cerebrospinal fluid leakage. The use of an
intraoperative image intensifier also allows the surgeon
to visualize the location of the ring curets during
resection of the suprasellar component of the tumor.

369
Neurosurgery Books
 Figure 10-3.
This drawing illustrates the transnasal
transsphenoidal approach to the sella turcica.
The introduction of the endoscope has allowed
resection through a transnasal approach
instead of a sublabial approach. The transnasal
approach has several advantages. There is no
gingival incision or resection of the anterior
nasal spine; therefore, there is less numbness of
the upper teeth. Also, because the approach is
less traumatic, the patient has less
postoperative discomfort. A disadvantage of the
transnasal route is that the approach is more
angled and, thus, slightly crosscourt when
working in the sella turcica. After the rostrum
of the sphenoid sinus has been removed, a
bivalved Hubbard or Hardy speculum is
inserted. At this point, the operating
microscope is brought into view. The rest of the
lateral sphenoid sinus is removed to enhance
exposure. Small bony septa and sinus mucosa
are removed with a small bone punch. After
this, the floor of the sella turcica is identified
visually and confirmed on the image intensifier.

370
Neurosurgery Books
 Figuralre 10-4.
There are several techniques for opening the floor of
the sella turcica. If the floor is enlarged and thin, a
microrongeur can be used to bite off the bone. Small
angled Kerrison rongeurs can be used to enlarge the
opening. If the sellar floor is not thin, a small chisel or
high-speed air drill can be used to make the opening. In
this illustration, a small chisel is used to make the initial
cracks in the sellar floor. A bone forceps is used to crack
off the initial body fragments, and an angled Kerrison is
used to enlarge the bony removal. Regardless of the
technique, it is mandatory to make sure that removal of
the sellar floor is maximized. Specifically, it is important
to remove the bone laterally as far as the junction of the
sellar dura mater and the cavernous sinus and up to the
junction with the floor of the frontal fossa. Similarly, it is
important to remove bone down toward the clivus.
Because most microadenomas are lateral, the lateral
bony removal to the cavernous sinus is most important.
Removal of the inferior floor down toward the clivus is
necessary only for large macroadenomas or if
hypophysectomy is planned.

371
Neurosurgery Books
 Figure 10-5.
After the floor of the sella turcica has
been removed, the dura mater is cauterized
with either a bipolar cautery or suction
cautery. Occasionally, the tumor infiltrates or
extends through the dura mater. It is
important to cauterize up to the junction of
the normal dura mater with the cavernous
sinuses. Usually, the sinuses can be identified
by a slight bluish discoloration of the dura
mater. After the dura mater has been
cauterized, an incision is made with a #11
blade knife and the dura mater is further
cauterized. When incising the dura mater, it is
important to stay 2 to 3 mm lateral to the
apparent location of the cavernous sinuses.
This safety margin will help decrease
intraoperative venous bleeding. Sometimes
bleeding occurs from venous sinuses at the
junction of the dura mater and the bony
removal. This bleeding usually can be
controlled with small pieces of absorbable
gelatin sponge (Gelfoam) placed under
tamponade.
372
Neurosurgery Books
 Figure 10-6.
After the dura mater has been
opened, a subdural dissection is
performed with a small ball-tip or flat
dissector. This helps to “milk” the tumor.
In large macroadenomas, the tumor
typically exudes or fungates out of the
dural opening. It is removed with small
cups and various angled ring curets.
Both rounded and sharp-edged ring
curets can be used. Sharp-edged ring
curets are good for cutting the tumor and
scraping it off normal pituitary tissue.

373
Neurosurgery Books
Figure 10-7.
As the tumor is removed, progressively larger ringed
curets are used. It is important to explore methodically all
recesses of the sella turcica. With large adenomas, large
Vanderbilt ring curets can be used. The image intensifier is
useful in confirming the location of the ringed curets,
especially when the surgeon is above the clinoid process.
Injection of 10 mL of air through the spinal needle helps to
push the tumor down into the operative field. Most often,
the tumor descends into the sella turcica without the
injection of thecal air, because these tumors are usually
under pressure and the dural opening creates an automatic
route for decompression. When scraping tumor out of the
lateral gutters, it is important to hold the ring curets gently
between the thumb and index finger to avoid inadvertent
entry into or injury to the cavernous sinus. Thickened,
exposed dura mater should be cauterized. Because of the
location of these macroadenomas, the pituitary gland is
usually displaced rostrally. Therefore, as more tumor is
removed, the pituitary gland descends into the pituitary
fossa. Most macroadenomas are easily suctioned, but the
normal gland is not. This can be an important technique
for differentiating tumor from normal gland. Also, the
normal gland has a more pinkish or reddish hue, whereas
tumors are typically yellow-white. Furthermore, a thin layer of reactive tissue or fibrosis often occurs between
the normal gland and the adenoma.
374
Neurosurgery Books
 Figure 10-8.
After resection of large macroadenomas,
it is prudent to reconstruct the floor of the
sella turcica. If cerebrospinal fluid is leaking
or oozing from the sella turcica, it is best to
harvest an abdominal fat or thigh muscle
graft and pack the sella. Importantly, the
sella turcica should not be overly packed.
Some surgeons advocate packing a large
sella turcica even if there is no cerebrospinal
fluid leak to prevent downward herniation of
the optic chiasm. After the sella turcica has
been packed, a piece of nasal cartilage is
placed subdurally to reconstruct the sellar
floor. With the transnasal approach, little
cartilage is usually available for
reconstruction. If there is a definite
cerebrospinal fluid leak, it is best to pack the
entire sphenoid sinus with abdominal fat.

375
Neurosurgery Books
 Figure 10-9.
Sequence for removal of a pituitary
microadenoma. In contrast to macroadenomas,
microadenomas usually do not cause
enlargement of the sella turcica. Therefore, the
surgeon has to work in a smaller space.
Thorough study of the preoperative magnetic
resonance images and biochemical data in
Cushing’s disease (if petrosal sinus sampling was
performed) are helpful in developing an idea
about where the tumor is located. After the bony
floor has been removed, the dura mater is
cauterized with a bipolar cautery and opened
with a #11 blade knife and further cauterized. It
is important to prevent heat injury to the normal
pituitary gland when cauterizing the dura mater.
Heat injury to the normal pituitary gland may
make it more difficult to identify the adenoma.
Some surgeons have reported the use of
microultrasonography in locating
microadenomas within the pituitary gland.
376
Neurosurgery Books
 Figure 10-10.
A subdural dissection is performed with a
small ball-tip or flat dissector. This helps to “milk
out” the tumor. An aggressive subdural dissection
can be useful in identifying the presence of a
lateral microadenoma. The subdural dissection
occasionally causes oozing of tumor exudate or
necrotic material. This early visual identification is
important in locating the site of the tumor. In the
operation illustrated here, the preoperative
magnetic resonance imaging study indicated that
the tumor likely was located within the right side of
the pituitary gland. Therefore, a #11 blade knife
was used to incise the right lateral gland. This
usually leads to oozing of necrotic tumor.

377
Neurosurgery Books
Figure 10-11.
After the tumor cavity has been identified, small curets are
used to work within the tumor to debulk it. A sharp ring curet can
be useful in peeling the tumor off its junction with normal pituitary
tissue. For microadenomas that are located laterally, it is
important to use ring curets to dissect the lateral gutter adjacent to
the cavernous sinus. In patients with Cushing’s disease,
identification of the microadenoma may be a problem. However, an
idea about where the tumor is located can be developed on the
basis of preoperative imaging studies and petrosal sinus sampling.
Therefore, the suspicious part of the pituitary gland should be
incised first. If no obvious adenoma is identified, the lateral one-
third of the pituitary should be removed en bloc and sent for
pathologic examination of frozen sections. If no tumor is identified,
an identical procedure should be performed with removal of the
opposite lateral one-third of the pituitary. If neither side (one-third)
of the pituitary contains tumor, an incision should be made in the
midline to determine whether the adenoma is sitting in the
midportion of the pituitary. If necessary, this remaining one-third of
the pituitary can be removed to perform a hypophysectomy. When
performing a hypophysectomy, it is important not to exert any
downward traction on the middle third of the gland to prevent long-
term injury to the infundibulum.

378
Neurosurgery Books
 Figure 10-12.
Sequence for removal of a cyst of
Rathke’s cleft cyst or intrasellar
craniopharyngioma. In both cases, the sella
turcica is usually enlarged and the bony
floor is thin. After the standard removal, the
dura mater is cauterized and opened in a
cruciate fashion. Whether a cyst or
craniopharyngioma, the first tissue
encountered is thin pituitary gland.
Therefore, pituitary tissue sits between the
surgeon and the lesion. The pituitary gland
is incised vertically to encounter the cyst.
The pituitary gland usually is no more than
2 to 3 mm thick. When the cyst is
encountered, there is expression of

379
Neurosurgery Books
 Figure 10-13.
After the cyst has been entered, the
pituitary gland collapses into the operative
field because of the decompression. Therefore,
a suction tip and small cottonoid (Americot)
are used to hold one wall open while the
contents of the cyst are explored. The wall of a
true cyst of Rathke’s cleft is thicker than normal
arachnoid. Parts of the cyst wall generally can
be dissected off the normal pituilary gland with
a long thin spatula or dissecting probe.
Occasionally, the involved cyst is so thin that it
cannot be removed without injuring the
underlying pituitary gland. In this case, it is
best to leave those small filament attachments
alone. When exploring the interior of a cyst of
Rathke’s cleft, it is important to look for any
grumous, yellowish firm material. The presence
of this material suggests that the tumor is in
fact a craniopharyngioma. Any solid tissue or
tumor should be removed with small ring curets
and punches.

380
Neurosurgery Books
 Figure 10-14.
The entire contents of the cyst should be
examined to make sure there is no solid
tissue. Small malleable endoscopes may be
useful in this regard. With intrasellar
craniopharyngiomas, the capsule is often
adherent to the diaphragma sellae and
pituitary stalk, requiring sharp dissection.
With intrasellar craniopharyngiomas or a
cyst of Rathke’s cleft, there is often a
cerebrospinal fluid leak. Therefore, it is
necessary to pack the sella turcica with
abdominal fat and to reconstruct the sellar
floor. It also is necessary to pack the
sphenoid sinus with abdominal fat,
especially if a future transcranial resection
of a craniopharyngioma is anticipated. It
can be difficult to control postoperative
rhinorrhea if a patient undergoes a cranial
approach after a transsphenoidal procedure
unless the pituitary fossa and the sphenoid
sinus have been adequately packed.

381
Neurosurgery Books
Chapter 11
Extracranial Vascular
Approaches

Neurosurgery Books
Chapter 11:
above the clavicle, usually along a fold of skin.
The skin incision is made with a scalpel and
extended through the platysma muscle. This skin

Extracranial incision usually severs the transverse cervical

nerve, which can be relatively large in some
patients. Sectioning of this nerve leads to

Vascular numbness anterior to the skin incision underneath

the mandible. Several small arteries tend to run
through the platysma muscle. Hemostasis
Approaches throughout this operation is achieved best with a
bipolar cautery. Use of bipolar instead of
monopolar current decreases the risk of injury to
EXPOSURE OF THE CAROTID cranial nerves and blood vessels. Because heparin
will be administered to the patient, it is best to
ARTERY spend several minutes using the bipolar cautery to
cauterize small arteries along the platysma and
A curvilinear incision along the anterior the underside of the skin margin. It usually is not
border of the sternocleidomastoid muscle is used necessary to use skin clips or hemostats.
to expose the common carotid artery. The top of
The surgeon and assistant retract the two
the incision curves posteriorly approximately 1.5
edges of the skin incision with sponges. A scalpel
cm underneath the mastoid process. The lower
is used to carry the incision deeper throughout its
end of the incision curves anteriorly 2 to 3 cm
length to identify the anterior border of the
383
Neurosurgery Books
sternocleidomastoid muscle. The upper end of the omohyoid muscle is also identified. The angle formed
by the sternocleidomastoid muscle and the omohyoid muscle points to the common carotid artery. With
dissecting scissors, the deeper relatively avascular fascia between the sternocleidomastoid muscle and
the trachea is incised and spread to identify the artery. At this point, fishhook retractors are placed
throughout the length of the incision, both medially and laterally. The fishhooks should be placed to grab
the fascia that envelops both the trachea and the sternocleidomastoid muscle, thus retracting these
structures medially and laterally, respectively. Here, the dissecting scissors are used to incise the deep
cervical fascia on top of the common carotid artery up toward its bifurcation.
It is critical when exposing the common carotid artery to dissect on top of it. The surgeon should not
dissect medially because of possible injury to the recurrent laryngeal nerve. At the level of the
bifurcation of the common carotid artery, the surgeon should never work medial to the superior thyroid
artery. When dissecting the proximal common carotid artery, it is important to clear the underside of the
artery to facilitate placement of vascular loops. However, one should not dissect underneath the
bifurcation of the common carotid artery because of the possibility of dislodging emboli or of injuring
the superior laryngeal nerve. Leaving the underside of the bifurcation attached to the deep cervical fascia
decreases the risk of these two complications.
At or just above the level of the bifurcation of the common carotid artery, the facial vein is
identified. It should be dissected sharply free to make sure that the hypoglossal nerve is not hidden along
its backside. Thereafter, the facial vein is doubly ligated and divided.
It can be useful to identify the descending hypoglossal nerve and to follow it proximally to the
hypoglossal nerve. The descending hypoglossal nerve can be severed close to its junction with the
384
Neurosurgery Books
hypoglossal nerve to allow better mobilization of the hypoglossal nerve distally. The superior thyroid
artery is identified at the level of the bifurcation of the common carotid artery. A temporary hemostatic
clip can be placed on this structure.
With sharp scissors, the external carotid artery is dissected free from the medial carotid sinus.
Approximately 1 to 2 cm above the bifurcation of the common carotid artery, the external carotid artery
gives off several branches, including the facial and occipital arteries. A vascular loop is placed around
the proximal trunk of the external carotid artery. The surgeon then identifies the medial aspect of the
internal carotid artery and frees it from the carotid sinus. Sharp scissors are used to dissect the distal
internal carotid artery above the bifurcation of the common carotid artery by at least 3.0 cm. For a
carotid endarterectomy, the surgeon can gently palpate the internal carotid artery and determine where
the atherosclerotic plaque ends. Occasionally, the plaque can be visualized under magnification. In either
case, it is best to have at least 1.0 cm of the internal carotid artery free above the presumed end of the
plaque to facilitate placement of a shunt if one is needed. As the dissection is carried distally under the
angle of the jaw, the fishhook retractors are placed under more tension. This tends to lift the carotid
artery complex up out of the wound. The dissection usually is carried to the belly of the digastric
muscle.
The tissue above the bifurcation of the common carotid artery between the external and internal
carotid arteries contains the nerves that innervate the carotid body and the carotid sinus. Injecting
approximately 1.0 mL of 1 percent lidocaine into this tissue helps to decrease the side effects of
excessive stimulation of the carotid sinus, including fluctuations in blood pressure and cardiac rhythm. It
is best to look between this neural sheath and the medial side of the external carotid artery to determine
385
Neurosurgery Books
whether an ascending pharyngeal artery originates from the backside of the bifurcation of the common
carotid artery. If this artery is present, a temporary hemostatic clip is used to occlude it.
At this point in the operation, two vascular loops are placed around the common carotid artery and
one around the external carotid artery. The author prefers not to place any vascular loops around the
distal internal carotid artery. After the loops have been placed, the surgeon must again determine
whether the dissection has been extended far enough distally. Typically, increased exposure can be
obtained by using blunt dissection with the finger and developing the plane between the anterior border
of the sternocleidomastoid muscle and the parotid gland underneath the angle of the jaw. The fishhooks
can be placed deeper into the belly of the sternocleidomastoid muscle close to its junction with the
parotid gland. Occasionally, additional exposure is required.

EXPOSURE OF THE DISTAL INTERNAL CAROTID ARTERY

Exposure of the distal internal carotid artery is an important ancillary technique that can be useful in
performing a complex carotid endarterectomy, in which the bifurcation of the common carotid artery is
high or the atherosclerotic plaque has a distal extension. It also is important to expose the distal internal
carotid artery close to the base of the skull when operating on extracranial aneurysms of the internal
carotid artery or large tumors of the carotid body. During high exposures, the cranial nerves at risk
include the facial, glossopharyngeal, vagus, and hypoglossal nerves. Careful anatomic dissection of the
distal internal carotid artery minimizes injury to these functionally important cranial nerves.

386
Neurosurgery Books
 In exposing the distal internal carotid artery, the following three points are important. One, it is
useful to have the patient intubated nasally. This allows the mouth to be closed, which increases
exposure under the angle of the jaw by approximately 1.0 to 1.5 cm. Although some surgeons report
planned dislocation of the mandible during high exposures, this has not been necessary in the author’s
experience. Two, it is important to have the anesthesiologist prepared to use induced hypertension
during cross-clamping of the internal carotid artery. With very high exposures, as for large aneurysms,
placing the shunt beyond the lesion at the base of the skull may not be technically feasible. Induced
hypertension can be a valuable technique in providing cerebral protection during cross-clamping by
increasing collateral blood flow. Three, use of the operating microscope is important during high
dissections. Both illumination and magnification are essential in minimizing injury to the cranial nerves.
Because the operative field can become cluttered, self-retaining retractors will compromise the
exposure. Accordingly, fishhooks are useful not only for optimizing exposure but for preventing injury
to the lower cranial nerves from excessive retraction.
The lower part of the skin incision parallels the anterior border of the sternocleidomastoid muscle
and curves posteriorly just underneath the earlobe. The superior portion of the incision sits in the
postauricular sulcus around the earlobe and ascends in a pretragal skin crease in front of the ear. In rare
cases, the incision extends up to the zygoma, which allows complete mobilization of the parotid gland
and the facial nerve. This incision results in a triangular bend underneath the ear and allows for better
alignment of the incision at closure.
The initial exposure is described in the preceding section. The descending hypoglossal nerve is
followed rostrally to identify the hypoglossal nerve. Because the hypoglossal nerve will be retracted
387
Neurosurgery Books
upward and medially under the mandible, the descending hypoglossal nerve must be severed. It is best to
leave a good stump of this nerve branch because a suture can be placed through it into the
submandibular fascia to retract the hypoglossal nerve. Occasionally, a muscular branch of the occipital
artery that supplies the sternocleidomastoid muscle can be identified. Because this muscular branch can
tether the hypoglossal nerve, it may be beneficial to doubly ligate and divide this artery as it sweeps over
the superior margin of the hypoglossal nerve, thus facilitating displacement of the nerve.
Next, the sternocleidomastoid muscle is dissected rostrally, often bluntly, with the surgeon’s finger.
This better defines the border between the anterior edge of the sternocleidomastoid muscle and the
caudal edge of the parotid gland. There is some loose fascia that is relatively avascular; it can easily be
divided with dissecting scissors. At this point, the digastric muscle usually is seen along the upper
margins of the operative exposure. Depending on the degree of exposure required, the digastric muscle
can be retracted upward with either an Army-Navy type retractor or a self-retaining Henley retractor. If a
higher exposure is required, the digastric muscle must be divided. The digastric muscle should be
followed to its insertion in the mastoid groove and divided there. However, before this muscle is ligated,
it is important to make sure that the main trunk of the facial nerve is not compromised. The stylohyoid
muscle, which lies superior and parallel to the digastric muscle, can also be divided, which exposes the
deeper stylomandibular ligament. This deeper stylomandibular ligament also can be resected to increase
exposure of the internal carotid artery at the base of the skull.
Before the digastric and stylohyoid muscles are ligated, it is useful to identify the facial nerve.
Accordingly, the incision at the lower end in front of the ear is made and extended caudally to connect
with the incision in front of the sternocleidomastoid muscle. Deep to this incision, the parotid gland is
388
Neurosurgery Books
identified in its enveloping fascia. Several hemostats or an Army-Navy retractor can be placed on the
parotid gland and used as retraction by the surgeon’s assistant. The posterior border of the parotid gland
is exposed further and elevated. The anteroinferior surface of the tragus is followed deep to the medial
triangular projection of cartilage. This deep cartilaginous projection “points” to the facial nerve. The
parotid fascia is incised further between the mastoid process and the posterior margin of the parotid
gland. Placement of the surgeon’s finger on the mastoid tip directed anteriorly and forward points to and
overlies the main trunk of the facial nerve. After the main trunk of the facial nerve has been identified,
the lower division and the marginal mandibular nerve, which form the upper limit of the deep dissection,
can be traced forward by sharp dissection and safely elevated by using the mobilized parotid tissue. At
this point, the digastric and stylohyoid muscles can be ligated safely. Sectioning of the stylomandibular
ligaments facilitates displacement of the mandible medially. Additional distal exposure of the internal
carotid artery can be achieved by removing the styloid process with a small rongeur.
The Facial Nerve

The facial nerve exits the base of the skull through the stylomastoid foramen and travels forward
and laterally into the parotid gland. It lies superior to the digastric muscle and, thus, can easily be
damaged. Before the digastric muscle is ligated, it is necessary to identify the main trunk of the facial
nerve. This is why it is essential to dissect the fascia between the parotid gland and the
sternocleidomastoid muscle adjacent to the mastoid process. Placing fishhooks in the submandibular
fascia and the parotid gland during prolonged operations helps limit injury to the marginal mandibular
branch of the facial nerve.

389
Neurosurgery Books
The Glossopharyngeal Nerve

The glossopharyngeal nerve contains important sensory fibers of the soft palate and the oropharynx.
Damage to the branches of this nerve may result in the patient having difficulty initiating swallowing.
Small filaments from the glossopharyngeal and the vagus nerves to the pharynx are sometimes anterior
and lateral to the internal carotid artery.
The Vagus Nerve

The pharyngeal branch of the vagus nerve exits from the trunk of the nerve at the level of the first
cervical vertebra. It enters the superior constrictor muscle of the pharynx and supplies all the muscles of
the pharynx and the soft palate except the stylopharyngeus and the tensor palati muscles. These motor
branches leave the vagus nerve and course laterally and deep to the external carotid artery, well above
the bifurcation of the common carotid artery. Therefore, during a typical carotid endarterectomy, they
are never visualized. However, they are at risk during an operation for a large aneurysm of the cervical
internal carotid artery or a carotid body tumor. The branches of these nerves are quite small, and they
may be either anterior or posterior to the internal carotid artery.
The superior laryngeal nerve leaves the vagus nerve at the lower margin of the first cervical vertebra
and descends. It is medial to both the internal and the external carotid arteries. The superior laryngeal
nerve divides into an internal and an external branch. The internal branch is sensory and innervates the
mucus membrane covering the epiglottis and the portion of the larynx above the vocal cords. The
external branch is motor and innervates the cricothyroid muscle and the inferior constrictor muscle of
the pharynx. Damage to the superior laryngeal nerve results in marked difficulty with swallowing
because the patient is unaware of food particles lodged in and around the area of the epiglottis and
390
Neurosurgery Books
because there is partial weakness of the inferior constrictor muscle of the pharynx and unilateral
paralysis of the cricopharyngeus muscle (which allows the opposite cricopharyngeus muscle to distort
the normal position of the cricoid cartilage). The superior laryngeal nerve can be injured during a routine
carotid endarterectomy when the surgeon dissects underneath the bifurcation of the common carotid
artery. Therefore, it is best not to mobilize underneath this bifurcation, not only to help prevent injury to
the superior laryngeal nerve but also to decrease the risk of emboli from excessive mobilization. The
superior laryngeal nerve also can be injured if the surgeon dissects medial to the superior thyroid artery.
Superiorly, the primary trunk of the vagus nerve descends between the internal carotid artery and the
internal jugular vein, and inferiorly, between the common carotid artery and the internal jugular vein. In
most patients, the vagus nerve is posterior to the artery and vein and, thus, is rarely injured.
Occasionally, the vagus nerve can have a wandering course and be anterior to the carotid artery. It also
can originate from the hypoglossal nerve and be confused with the descending hypoglossal nerve.
Accordingly, any large nerve on top of the common carotid artery should be preserved and protected.
The recurrent branch of the vagus nerve is never visualized during a routine exposure of the internal
carotid artery. It ascends between the trachea and the esophagus. On the right side, the recurrent branch
arises from the vagus at the base of the neck in front of the right subclavian artery. It then loops around
this artery (passing below and behind the artery) and runs superiorly and medially, crossing obliquely
behind the common carotid artery to gain access to the groove between the trachea and the esophagus,
where it ascends to the lower border of the cricoid cartilage. At this point, the recurrent branch of the
vagus nerve becomes the inferior laryngeal nerve and innervates many of the intrinsic muscles of the

391
Neurosurgery Books
larynx. Damage to the recurrent laryngeal nerve leads to unilateral paralysis of the vocal cords and
hoarseness.
The Spinal Accessory Nerve

The spinal accessory nerve is identified adjacent to the jugular vein as it descends over the
transverse process of the first cervical vertebra and disappears into the deep surface of the
sternocleidomastoid muscle. Thus, this nerve has a medial to lateral course. While the
sternocleidomastoid muscle is dissected off the mastoid process during a high internal carotid artery
exposure, the spinal accessory nerve can be injured. Damage to this nerve leads to paralysis of the
sternocleidomastoid and the trapezius muscles. Injury to the trapezius muscle leads to a partial shoulder
drop and painful bursitis.
The Hypoglossal Nerve

As indicated above, the hypoglossal nerve can be mobilized rostrally after the descending
hypoglossal nerve has been severed. This stump can be used to manipulate the hypoglossal nerve; it
even can be tacked to the fascia underneath the mandible for improved exposure. Occasionally, the
artery to the sternocleidomastoid muscle must be severed to allow the hypoglossal nerve to be displaced
distally.
The Great Auricular Nerve

The cervical plexus has three major sensory nerves: the occipital nerve, the great auricular nerve,
and the transverse cervical nerve. The great auricular nerve and the transverse cervical nerve enter the
subcutaneous tissue by piercing the external cervical fascia along the posterior border of the
392
Neurosurgery Books
sternocleidomastoid muscle. The transverse cervical nerve usually is cut during the skin incision, and
this may result in patchy numbness along the incision. The great auricular nerve is encountered during
the distal exposure when the parotid gland is mobilized. Severing the great auricular nerve leads to
numbness of the earlobe. Depending on the location of this nerve in relation to the incision, it often can
be dissected and mobilized. However, necessary exposure of the distal internal carotid artery should not
be compromised by the anatomic preservation of the great auricular nerve. The cervical plexus
contributes to the ansa cervicalis. The ansa hypoglossi supplies the motor innervation to the deep
muscles of the neck. These branches usually are severed during exposure of the carotid artery without
any resulting sequelae.
Carotid Endarterectomy

Approximately 3 minutes before the common carotid artery is cross-clamped, 5,000 units of heparin
are administered intravenously. Also, the anesthesiologist is asked to use induced hypertension to
increase the patient’s systolic blood pressure to approximately 170 to 180 mm Hg. This induced
hypertension facilitates collateral blood flow and significantly decreases the risk of cross-clamp
ischemia and the need for a shunt. After time has been allowed for the heparin to circulate, the common
carotid artery is occluded as low as the dissection permits with one “click” of a soft-shoe Fogarty clamp.
Before the common carotid artery is occluded, it is important for the surgeon to palpate this vessel with
the thumb and index finger to try to determine whether the clamp can be placed below any obvious
atherosclerotic plaque in the common carotid artery. The external carotid artery is occ1uded by placing
tension on the vascular loop or by using a temporary aneurysm clip. A small temporary aneurysm clip is
placed across the distal internal carotid artery as far as the exposure will allow. These temporary

393
Neurosurgery Books
aneurysm clips have a closing pressure of approximately 75 to 100 g. Use of a higher pressure closing
clip has the risk of injuring the intima of the distal internal carotid artery.
The common carotid artery is opened with a stab wound made with a #11 blade knife. A Potts
scissors is used to carry the incision distally through the bifurcation of the common carotid artery into
the internal carotid artery above the level of the atherosclerotic plaque. It is important to extend the
arteriotomy in the middle of the vessel instead of veering off to one side, which makes closure of the
arteriotomy more difficult. After the initial arteriotomy, a small suction tip is used to remove blood in the
lumen. The atherosclerotic plaque is inspected under magnification. It usually is necessary to extend the
arteriotomy further, both distally and proximally, to expose completely the upper and lower ends of the
lesion. In many patients, the plaque extends well down toward the aortic arch; therefore, clinical
judgment must be used to determine how low the plaque is removed from the common carotid artery.
The atherosclerotic plaque is removed with a small spatula. Typically, the plane between the
atherosclerotic plaque and the underlying intima can be identified along the medial or lateral side
adjacent to the bifurcation of the common carotid artery. By using the spatula in a gentle back and forth
motion along the long axis of the artery, the plaque is freed from the underlying intima. Usually, the
plaque feathers free or breaks cleanly from the intima of the distal internal carotid artery. In some cases,
the plaque blends imperceptibly with the intima distally. In these situations, a small incision with a
microscissors is used to start the break. The break then can be continued by a peeling action of the
plaque, with countertraction applied against the wall of the vessel. In most cases, the best separation of
the plaque from the intima involves a tearing instead of a cutting motion. After a good separation of the

394
Neurosurgery Books
plaque from the distal internal carotid artery has been achieved, the dissection is carried down along the
common carotid artery.
After the atherosclerotic plaque has been removed from the internal carotid artery, the dissection is
carried proximally along the lateral wall of the common carotid artery to the point that the surgeon
thinks it is best to end plaque resection, at which place the plaque is incised with a Potts scissors. This is
carried circumferentially around to the medial side of the common carotid artery. The spatula is used to
dissect the plaque off the medial wall of the common carotid artery up to its bifurcation.
It is important to obtain good removal of the atherosclerotic plaque up the external carotid artery to
prevent retrograde stump emboli. This is achieved best by using the spatula to work up the external
carotid artery circumferentially around the plaque. The vascular loop is loosened gently and a hemostat
is placed up in the external carotid artery as far as possible to grasp the plaque. In a simultaneous
motion, the external carotid artery is everted and the hemostat is used to pull the plaque down, ideally
obtaining a clean break in the distal external carotid artery. After the plaque has been removed, the entire
endarterectomy bed is inspected meticulously for small flaps. Irrigation with a heparin solution usually
will cause these small filaments to float upward, allowing better visualization and removal. It is
important to use magnification to ensure that the endarterectomy bed is as smooth as possible. Again, the
distal endarterectomy bed and its junction with the internal carotid artery are inspected to make sure
there are no flaps or loose filamentous attachments, which can be the source of thrombus formation and
emboli.
It should be emphasized that the complete atherosclerotic plaque up the distal internal carotid artery
must be removed during the endarterectomy. It is common to find a subintimal extension of the plaque
395
Neurosurgery Books
along the posterior wall of the artery. Strict attention to the plane between the plaque and the intima
facilitates removal of the plaque from its subendothelial layer and the distal internal carotid artery. If
after the plaque has been removed there is a ledge or concern about possible future flap formation, 6-0
double-arm monofilament sutures are used as tack sutures. They are placed from the interior of the
vessel to the exterior. Use of double-arm instead of single sutures allows better placement of these
tacking sutures, without stenosing the distal internal carotid artery.
Shunt Placement

Currently, the role of monitoring and the use of shunts are a matter of controversy. The author
believes that continuous monitoring with intraoperative electroencephalography and selective shunting
significantly decrease the risk of intraoperative ischemic complications. The criterion used for placement
of a shunt is a significant electroencephalographic change, which occurs in approximately 10 percent of
patients who undergo carotid endarterectomy. It is safe to place the shunt, if needed, after the plaque has
been removed. This allows for better plaque removal and easier shunt placement. The shunt is placed by
inserting it into the common carotid artery and securing it with two vascular loops. The Fogarty clamp is
removed temporarily, allowing the shunt to be flushed. The common carotid artery is reoccluded, and
the shunt is inserted into the internal carotid artery. After the shunt has been inserted into the internal
carotid artery, it is secured with a Sundt-Kees clip graft. Thereafter, flow is restored up the shunt.
Typically, within 1 to 2 minutes after the shunt has been placed, the electroencephalogram should
normalize. If a significant electroencephalographic change persists, strong consideration should be given
to performing intraoperative angiography through the shunt to determine whether there is an embolus in
the middle cerebral artery.

396
Neurosurgery Books
Arteriotomy Closure

If the surgeon prefers, the vessel can be closed primarily with a running 6-0 monofilament suture.
Clearly, the primary closure should be performed with magnification to avoid stenosis of the vessel
because of incorporation of too much vessel wall in the suturing.
If the surgeon prefers to close with either a fabric or saphenous vein patch graft, it is sewn in place
with a double-arm running 5-0 monofilament suture. The first stitch is placed through the distal end of
the graft and arteriotomy up the internal carotid artery. This stitch is tied together with three overhand
knots. One end of the monofilament suture is tagged to provide some countertraction. The other stitch is
used to carefully anastomose the graft to the internal carotid artery. The first five or so stitches should be
placed accurately to allow good taper of the graft to the internal carotid artery. These sutures should be
approximately 1 mm deep and 1 mm apart. Each suture should purchase a full thickness of the intima
and the vessel wall. Also, the suture should also go from inside the vessel to the outside to allow
eversion of the suture line to the luminal surface. Accordingly, one suture line is performed largely with
a backhand stitch. After the anastomosis is down into the common carotid artery, the backhand stitch can
be rotated to allow a forehand suture from the abluminal to the luminal surface if desired. All sutures
should be parallel, with an equal purchase of artery and graft, to prevent any irregularities along the
suture line, which possibly could serve as a source of thrombus formation. As the graft approaches the
proximal arteriotomy along the common carotid artery, it is tailored to provide a better fit. When sewing
the graft to the artery, it is important to use these graft sutures as tacking sutures to secure the proximal
plaque along the common carotid artery to the wall of the vessel. If a saphenous vein is used as a patch,
it is best to reinforce it with a thin sheet of polytef (Teflon) to prevent formation of a pseudoaneurysm.

397
Neurosurgery Books
Alternatively, if a synthetic expanded polytetrafluoroethylene (Gore-Tex) graft is used, a reinforcing
polytef graft is not required.
Restoring Flow

Just before the closure of the arteriotomy is completed, the common carotid artery must be occluded
if a shunt has been placed. The distal securing clip is loosened, and the shunt is removed. A temporary
clip is placed again on the internal carotid artery, and then the final sutures in the arteriotomy are placed.
The last suture is kept loose to allow for back bleeding. The sequence of back bleeding and restoring
flow is critical in preventing emboli. First, the clip is removed temporarily from the internal carotid
artery to allow back bleeding from this vessel. After several seconds of back bleeding, the clip is
reapplied. Second, the Fogarty clamp is loosened gently to allow bleeding up through the site of the last
few stitches of the arteriotomy closure. After several seconds of bleeding up the common carotid artery,
the Fogarty clamp is reapplied. Third, the clip on the internal carotid artery is again removed temporarily
to allow for a final back bleeding. After this clip has been reapplied, the monofilament suture is tied. In
tying the suture, a gentle force is used to tighten the entire suture line. Before flow is restored, the
anesthesiologist is directed to reduce systolic blood pressure to 150 mm Hg to decrease the risk of
hyperperfusion. Flow is restored, first, by removing the Fogarty clamp and, second, by loosening the
vascular loop on the external carotid artery. During this time, the internal carotid artery is occluded.
Accordingly, any loose fragments or emboli in the common carotid artery are flushed up the external
carotid artery. After blood flow has been restored up the external carotid artery for 1 or 2 minutes, the
clip is removed from the internal carotid artery. During this reperfusion period, the anesthesiologist is
requested to palpate the pulse of the temporal artery by the ear. A good pulse means that the external
398
Neurosurgery Books
carotid artery is patent. However, the absence of a pulse suggests there is an intimal flap in the distal
external carotid artery. In this case, it is best to explore the external carotid artery by temporarily
occluding it at its takeoff from the common carotid bifurcation and as far distally as the exposure will
allow. An arteriotomy is made with a #11 blade knife and extended with Potts scissors up the external
carotid artery. The surgeon usually will find a poor breakage of the atherosclerotic plaque. This
generally can be dissected off the intima of the external carotid artery. Thereafter, the arteriotomy is
closed primarily with a running 6-0 monofilament suture.

399
Neurosurgery Books
Figure 11-1.
This drawing shows the pertinent
neurovascular anatomy and typical incision
for carotid endarterectomy. If a high
exposure is required for partial mobilization
of the parotid gland, the incision is carried
anterior to the ear.

External carotid artery

Mandibular branch of facial nerve
Superior thyroid artery

Common carotid artery

Facial nerve

Digastric muscle

External jugular vein

Jugular vein

Ligated facial vein

Hypoglossal nerve

Descending hypoglossal nerve

Superior laryngeal nerve

Vagus nerve
Figure 11-2.
Step 1. After 5,000 units of
heparin have been given
intravenously, the common carotid
artery is occluded with a soft-shoe
Fogarty clamp, the internal carotid
artery with a temporary aneurysm
clip, and the external carotid artery
with a vascular loop. An arteriotomy
is made in the common carotid
artery with a #11 blade knife.
401
Figure 11-3.
Step 2. The arteriotomy is extended with Potts
scissors up through and beyond the
atherosclerotic plaque in the internal carotid
artery. A plane of cleavage between the
atherosclerotic plaque and intima at the
bifurcation of the common carotid artery is
identified with a spatula. The plaque first is
removed up the internal carotid artery. The
plaque generally breaks free and feathers out of
the internal carotid artery. Occasionally, it is
necessary to use microscissors to make a small
cut in the plaque. After the starting cut has been
made, the plaque is torn with a combination of
countertraction and use of the spatula. Only
atraumatic vascular forceps are used to grasp
the vessel wall. Furthermore, to avoid traumatic
injury, the intima of the internal carotid artery
beyond the endarterectomy is never held
directly with the forceps.

402
Figure 11-4.
Step 3. After the distal
plaque in the internal carotid
artery has been removed, the
dissection is carried proximally
down the common carotid artery
with the use of the spatula and
countertraction between the
plaque and the vessel wall. The
Potts scissors is used to cut the
plaque in the common carotid
artery in a circumferential
pattern.

403
Figuralre 11-5.
Step 4. The spatula is used to dissect the plaque in a blind fashion circumferentially in the
external carotid artery. The plaque is grasped with a hemostat, and an eversion technique is
used to pull it out of the external carotid artery. After the plaque has been removed, the entire
endarterectomy site is irrigated with heparin solution. Loose filaments are removed with
microforceps. The internal carotid artery is inspected to make sure there are no flaps or ledges.
404
Figure 11-6.
Step 5. In large caliber vessels, the arteriotomy can be closed, under magnification, with a running 6-0
monofilament suture. Each suture should be full thickness, including the adventitia, to ensure that the closure
has normal tensile strength. Just before the suture is tied, the aneurysm clip on the internal carotid artery is
removed temporarily to allow back bleeding down this vessel. The artery is then reoccluded. The Fogarty clamp
is removed temporarily from the common carotid artery to flush air and any grumous material out through the
arteriotomy opening. The final suture is then tied. Blood flow is restored by removing the Fogarty clamp and
loosening the loop around the external carotid artery. After 10 to 20 seconds, the clip on the internal carotid
artery is removed. Accordingly, any thrombus or grumous material is flushed up the external carotid artery.
This decreases the risk of an embolic event occurring during the restoration of blood flow after the carotid
endarterectomy.
405
Figure 11-7.
In some cases, intraoperative monitoring,
such as electroencephalography, indicates a
significant change with crossclamping of the
common carotid artery. In such case, it is best to
place a shunt. Before the shunt is placed, it is best
to remove at least the plaque up the internal
carotid artery. Often, the entire endarterectomy
can be performed before a shunt is placed, thereby
ensuring good plaque removal. First, the shunt is
placed down the common carotid artery. The
Fogarty clamp is released temporarily, and the
shunt is flushed to dislodge any air. Second, the
shunt is inserted into the internal carotid artery.
The temporary clip is removed from the internal
carotid artery to flush out this vessel. After the
shunt has been inserted up the internal carotid
artery, it is secured with a clip graft. The Fogarty
clamp is then removed and blood flow is restored
up the shunt. If the electroencephalogram does not
normalize after several minutes, intraoperative
angiography through the shunt should be
considered to look for a possible embolus in the
middle cerebral artery.
406
Figure 11-8.
The internal carotid artery often
has a small caliber. In this case, it is
best to close the arteriotomy
secondarily with a patch graft. A
saphenous vein or synthetic graft can
be used. Illustrated here is repair of
the arteriotomy with a graft and an
indwelling shunt. A double-arm 24-
inch 5-0 monofilament suture works
best. It is optimum to sew from the
graft to the vessel, thereby leading to
eversion of the suture line. The first
five or so stitches in the internal
carotid artery should be
approximately 1 mm wide and 1 mm
deep. Thereafter, the stitches should
be 2 mm wide and 2 mm deep. Exact
placement of the sutures will lead to
a smooth suture line, which will
decrease turbulence and the risk of
thromboembolic complications.

407
Figure 11-9.
Just before the suture line is
closed, the common carotid artery is
occluded again with a Fogarty
clamp. The securing clip on the
internal carotid artery is removed,
and the shunt is extracted. This will
also allow back bleeding down the
internal carotid artery.

408
Figure 11-10.
After several seconds of back bleeding, the
internal carotid artery is occluded again with a
temporary aneurysm clip. The arteriotomy is
then closed. The Fogarty clamp is removed,
after which the vascular loop on the external
carotid artery is loosened. This allows any
potential embolic material to be flushed up the
external carotid artery. After approximately 10
to 20 seconds, the clip on the internal carotid
artery is removed.

409
CAROTID BODY TUMOR

Surgical Technique

It is useful to consider several general principles when resecting large tumors of the carotid body.
First, it is important to perform the operation with intraoperative monitoring such as
electroencephalography. If the common carotid artery must be occluded temporarily for hemostasis, it is
important to know the status of the collateral blood flow to the ipsilateral cerebral hemisphere. Second,
distal exposure of large tumors is obtained by mobilization of the parotid gland, as discussed above. This
approach facilitates identification of the lower cranial nerves cephalad to the tumor and aids in their
preservation. Third, these tumors are dissected in the capsular-adventitial plane instead of the
subadventitial plane advocated by some surgeons. The capsular-adventitial plane is developed with
bipolar coagulation techniques and magnification with either surgical loops or an operating microscope.
Use of this plane minimizes the risk of injury to the arterial wall and, ultimately, hemorrhage. Fourth,
great effort is taken to maintain the integrity of the external carotid artery. Fifth, although some authors
recommend routine use of a shunt in patients with large tumors, shunts are used only when the
electroencephalogram demonstrates insufficient perfusion if carotid occlusion is required. This
procedure minimizes the risks associated with the use of a shunt. In most cases, meticulous dissection
eliminates the need for either temporary occlusion of the carotid artery or placement of a shunt.
In all patients, the ipsilateral lower leg is prepared and draped in case a saphenous vein graft is
required. The proximal common carotid artery is exposed, as described above, by dissection of the deep

410
fascia anterior to the sternocleidomastoid muscle. This dissection is extended upward to the bifurcation
of the common carotid artery, at which point the caudal limit of the tumor is encountered.
The common facial vein is often incorporated in the tumor capsule and must be ligated together with
several draining veins surrounding the tissue. The tumor is isolated along its medial and lateral borders.
The proper plane of dissection is identified, using the bipolar forceps under visual magnification
between the lower pole of the tumor and the common carotid artery. Because the main blood supply of
the tumor is from the bifurcation of the common carotid artery and the external carotid artery, the
dissection delineates these attachments first. There usually is an areolar plane between the tumor and the
artery, except for the subadventitial attachment of the tumor at the posterior wall of the bifurcation of the
common carotid artery. With the use of a bipolar cautery, the many perforating arteries arising from the
vasa vasorum are coagulated and divided. The tumor is often fed by large proximal branches of the
external carotid artery, which should be ligated individually, as should the large feeding arteries from the
vertebral and thyrocervical trunk that develop in very large tumors. The tumor is grasped with forceps
and rotated superolaterally to expose the interface between the tumor and the carotid artery. By
dissecting in the periadventitial layer close to the arteries, the risk of injuring the superior and recurrent
laryngeal nerves is minimized.
Rarely, the vagus nerve is incorporated in the tumor bed; if it is, it must be carefully dissected free.
The identification of the vagus nerve in relation to the tumor is facilitated by mobilization of the parotid
gland. The other cranial nerve that can be injured at this point in the dissection is the hypoglossal nerve.
The tumor usually displaces the hypoglossal nerve posteriorly and superiorly. It is important to identify
this nerve in the submandibular region to preserve it. The mandibular branch of the facial nerve can be
411
injured by excessive retraction under the angle of the jaw; however, mobilization of the parotid gland
usually provides sufficient room for the dissection. In tumors that have a large lateral extension, the
surgeon must be alert for the accessory spinal nerve.
After the feeding branches from the common and external carotid arteries have been dissected and
ligated, the lateral and superior poles of the tumor are further mobilized. Laterally and somewhat
posteriorly, it is common for the tumor to derive a major share of its blood supply from the carotid
sheath. These vessels often are quite large, but they can be well controlled with bipolar coagulation
techniques. As the exposure is extended higher, the parotid gland is mobilized so that the most cephalic
portion of the tumor can be identified. The digastric muscle generally is severed at its connection with
the mastoid process to aid in dissecting the upper limits of large tumors. The last portion of the tumor
that is dissected free is the medioposterior attachment between the internal and external carotid arteries.
In this manner, most of the tumor can be rotated to visualize better the vagus nerve and its extremely
important superior laryngeal branch. As the tumor is elevated from its bed, the superior laryngeal nerve
is worked away from the tumor capsule. After the tumor has been excised, the arteries are inspected
carefully for any injury to the arterial wall. If a segment of artery appears to be injured, the artery is
occluded and a local arteriotomy is made. The appropriate arterial repair and, if necessary,
endarterectomy are performed.
Rarely, a massive tumor of the carotid body may extend up to and erode the foramen lacerum and
the petrous bone. Tumors that extend into the posterior fossa require suboccipital craniectomy. Tumors
that invade the foramen lacerum can be approached through a petrosectomy, but often these tumors are
not completely resectable.
412
Figure 11-11.
Step 1. The typical exposure of a
carotid body tumor requires some
mobilization under the parotid gland.
The digastric muscle must be either
retracted or sectioned at its
attachment to the mastoid process.
The facial vein usually is
incorporated in the capsule of the
tumor.

413
Figure 11-12.
Step 2. The dissection
starts at the origin of the
carotid body tumor at the
bifurcation of the common
carotid artery. The proximal
internal and external
carotid arteries are
identified and dissected
sharply free. There typically
are numerous vascular
feeding vessels from the
bifurcation of the common
carotid artery and the
external carotid artery that
414
Figure 11-13.
Step 3. The dissection proceeds up the
external carotid artery, where major arterial
feeding vessels can be identified. Some of
these vessels need to be ligated individually.
The tumor capsule is somewhat fibrous;
therefore, with greater mobilization, the
tumor can be rotated to facilitate the
dissection. After the tumor has been freed
from the external carotid artery, the surgeon
identifies the proximal internal carotid artery
and dissects it off the tumor capsule.

415
Figure 11-14.
Step 4. At the upper limit of
the tumor, the hypoglossal nerve
is usually identified under the
belly of the digastric muscle. It is
dissected sharply free from the
tumor capsule. The tumor is
displaced or retracted downward
to facilitate identification of this
plane.

416
Figure 11-15.
Step 5. The attachment of the tumor to the
deep fascia between the internal and external
carotid arteries is dissected sharply free. It is
important to recognize that branches of the
superior laryngeal nerve that run in this fascia
need to be protected, otherwise the patient will
have difficulty swallowing.

417
CAROTID ARTERY ANEURYSM

Surgical resection of an aneurysm of the cervical internal carotid artery uses many of the techniques
described above for carotid endarterectomy and high exposures. It is important to emphasize that it is
mandatory to obtain adequate exposure distal to the aneurysm. Often, this requires partially mobilizing
the parotid gland and sectioning the digastric and stylohyoid muscles. It is beneficial to have a
neuroradiologist perform a trial balloon occlusion of the internal carotid artery proximal to the aneurysm
during preoperative transfemoral cerebral angiography. This trial balloon occlusion in combination with
blood flow studies can provide valuable information about the degree of collateral blood flow. Thus,
before the operation, the surgeon knows whether a shunt will be necessary when the carotid artery is
occluded and the aneurysm is resected.
If a trial balloon occlusion cannot be performed, the following intraoperative assessment with
electroencephalography should be performed. After the common carotid artery and its bifurcation have
been exposed, it is best to do a test clamping of the carotid artery to assess the potential for collateral
blood flow when the arteries are occluded for resection of the aneurysm. During this test occlusion, the
common carotid artery and the external carotid artery are occluded. Alternatively, the internal carotid
artery is occluded just beyond the bifurcation of the common carotid artery if this exposure has already
been achieved. During this cross-clamping trial, the intraoperative electroencephalogram is observed for
3 to 5 minutes to determine whether there is sufficient collateral blood flow. If this trial occlusion
produces a change in the electroencephalogram, blood flow is restored immediately and the patient is
made hypertensive. Next, a second trial occlusion is performed to determine whether this induced

418
hypertension has significantly increased collateral blood flow. These measures are necessary because it
is considerably easier to perform an interposition graft without having to use a shunt.
After the degree of collateral flow and the potential need for a shunt have been determined, the
aneurysm is dissected free to identify the distal internal carotid artery. During this distal dissection, care
is taken to identify and to protect the hypoglossal nerve and possibly filaments coming from the vagus
nerve. After the distal and the proximal ends of the aneurysm have been identified, the arteries are
occluded with temporary aneurysm clips and the aneurysm is resected.
The interposition graft between the stump of the internal carotid artery and the distal internal carotid
artery can be performed with either a dilated saphenous vein or a synthetic vascular graft. Several
anastomoses are illustrated to provide a smooth transition between vessels and grafts of different caliber.
The distal anastomosis is performed first, using an operating microscope and interrupted 6-0
monofilament sutures. The anastomosis commences with the back wall. It is best to use interrupted
instead of running sutures to avoid stenosis of the anastomosis site. Typically, the angle-spatulated
anastomosis works best. After the distal anastomosis has been performed, the graft is irrigated with a
solution of heparin and then sewn to the proximal internal carotid artery using several different types of
anastomoses, depending on the difference in caliber between the graft and the proximal internal carotid
artery. Although interrupted sutures are more time-consuming, they tend to work best. Just before the
last several sutures are tied, it is best to back bleed both down the distal and the proximal internal carotid
arteries to prevent emboli from going up through the graft.

419
Figure 11-16.
The primary treatment of an aneurysm
of the cervical internal carotid artery is
resection of the aneurysm and
reconstruction of the internal carotid
artery. Reconstruction of the internal
carotid artery can be performed by direct
reanastomosis of the proximal and distal
internal carotid artery if there is sufficient
redundancy in the vessel. Otherwise, an
interposition graft will be necessary. It
often is necessary to work underneath the
angle of the jaw and to partially mobilize
the parotid gland to expose the distal
internal carotid artery beyond the
aneurysm. As illustrated here, the
digastric muscle and the stylohyoid
muscle have been severed and reflected
medially.

420
Figure 11-17.
After 5,000 units of heparin have been administered, the arteries are occluded. A soft-shoe Fogarty
clamp is used to occlude the common carotid artery, and a temporary aneurysm clip is used to occlude the
distal internal carotid artery beyond the aneurysm. The external carotid artery can be secured with either
an aneurysm clip or a vascular loop. Next, the aneurysm is dissected from the surrounding tissue. The
nerves at risk for injury include the hypoglossal nerve, the vagus nerve, and the superior laryngeal nerve.

421
Figure 11-18. Various anastomoses are useful for
sewing interposition grafts. The goal of these
anastomoses is to prevent stenosis at the site of
the anastomosis. A, The first anastomosis is a fish
mouth maneuver. It is important to use double-
arm monofilament sutures and to sew from inside
to outside to allow eversion of the vessel and,
therefore, proximation of the intima with either the
intima of the saphenous vein graft or the inner
lining of the fabric graft. B, The second
anastomosis is a spatulated anastomosis. Again, it
is important to make sure that the anastomosis
line is everted. Although a running suture can be
used, it has a higher risk of stenosis. Therefore,
interrupted sutures are often preferable for
reconstructing smaller vessels, although these
sutures take more time. C, The third anastomosis
is a back wall–roofpatch reconstruction. This
anastomosis is useful in providing the best taper
between two vessels of significantly different
diameters. The back wall is sewn with interrupted
6-0 monofilament sutures. If a saphenous vein is
used as the interposition graft, a segment of the
vein can be used for the patch graft. If a synthetic
graft is used, either the graft itself or a small vein
can be used as the patch.

422
Figure 11-19. In this illustration, an interposition graft is being constructed with a
polyethylene terephthalate (Dacron) fabric ribbed graft. The distal hookup is a
spatulated anastomosis.

423
Figure 11-20. A proximal anastomosis using a back wall–roofpatch graft reconstruction. As in any
interposition graft, it is important to make sure that the interposition vessel is of sufficient length
that neither anastomosis is under excessive tension. After the reconstruction has been performed,
the Fogarty clamp is removed from the common carotid artery, and the external carotid artery is
reopened. This allows any thrombus or grumous material to be flushed up the externalcarotid
artery. Thereafter, blood flow is restored through the internal carotid artery.

424
RECURRENT CAROTID STENOSIS

Reoperation for recurrent carotid stenosis is becoming a more frequent surgical procedure. The most
common causes of recurrent stenosis are recurrent atherosclerosis and myointimal hyperplasia.
Operating on a recurrent carotid stenosis is more difficult than performing the original operation because
of the dense scar tissue that forms in the neck after the initial procedure. Dissecting the carotid artery
can be hazardous not only because of injury to the arteries but also to the cranial nerves. Also, in
approximately one-half of the patients with recurrent atherosclerosis, endarterectomy cannot be
performed because the plaque is tightly adherent to the underlying intima by scar tissue. It quickly
becomes apparent to the surgeon whether the recurrent plaque can be removed. If a plane cannot be
identified easily, in all likelihood an interposition graft from the common carotid artery to the internal
carotid artery above the recurrent stenosis will be required.
The interposition graft can be fashioned from either a dilated saphenous vein or a synthetic
expanded polytetrafluoroethylene graft. The techniques and anastomoses for this graft are identical to
those described above for surgical resection of aneurysms of the cervical internal carotid artery. It is
important to emphasize that induced hypertension before cross-clamping the carotid artery can
significantly reduce the need for placement of a shunt. Performing an interposition graft without an
indwelling shunt is technically easier. It also is important to administer 5,000 units of heparin before
cross-clamping the carotid artery. Although this leads to temporary leaking at the two sites of
anastomosis, it decreases the incidence of thromboembolic complications.

425
 The treatment of recurrent stenosis due to myointimal hyperplasia is enlargement of the artery with a
roofpatch graft. In myointimal hyperplasia, the lesion consists of a thickening of the entire arterial wall
in the region of the previous endarterectomy. Because there is no plane in which an endarterectomy can
be performed, angioplasty with a patch graft works best.

426
Figure 11-21. An arteriotomy is made in
a standard fashion extending from the
common carotid artery up through the
internal carotid artery. In approximately
50 percent of patients, no plane
separates the recurrent atherosclerotic
plaque from the underlying intima; thus,
an interposition graft must be performed.
The external carotid artery is doubly
ligated and divided. The superior thyroid
artery is also ligated. The common
carotid artery below the arteriotomy and
the internal carotid artery distal to the
arteriotomy are cut, leaving stumps of
each vessel. It is unnecessary—and in
fact dangerous—to remove the
bifurcation of the common carotid artery
because it will be scarred to its
underlying fascia, and the superior
laryngeal nerve will be exposed to injury
during attempted removal. Thus, it is best
just to lay the interposition graft within
the opening vessel and to leave the
bifurcation of the common carotid artery
in situ.

427
Figure 11-22. In the case illustrated here, a saphenous vein has been harvested from the leg. It is being
dilated with cold papaverine-heparin solution. Use of this balloon vein dilatation kit prevents
overdilatation and injury to the intima of the vein. The saphenous vein is sewn in place, anastomosing it
from the common carotid artery to the internal carotid artery. The distal anastomosis is performed first,
usually with a fish mouth anastomosis. Next, the proximal anastomosis is completed. Just before the last
few sutures of the proximal anastomosis are tied, both antegrade and retrograde back bleeding are
performed from the internal and the common carotid arteries to prevent an embolus from traveling up the
internal carotid artery.

428
MOYAMOYA DISEASE

Several surgical techniques have been used in the treatment of moyamoya disease. The operation
advocated here always uses an encephalomyosynangiosis, with grafting of the temporalis muscle to the
cortical surface. This is combined with either an encephaloarteriosynangiosis or a direct superficial
temporal artery-to-middle cerebral artery anastomosis, depending on the caliber of the temporal artery
and the cortical vessels. Specifically, if the temporal artery or cortical branches of the middle cerebral
artery are small, an arteriosynangiosis is performed by dissecting the temporal artery free but leaving it
in continuity with the scalp. This temporal artery graft is placed onto the cortical surface and held in
place by two bands of dura mater. The surrounding bony defect is covered with the temporalis muscle.
Alternatively, if the temporal artery and recipient branches of the middle cerebral artery are of sufficient
caliber, a direct superficial temporal artery-to-middle cerebral artery anastomosis is performed. The
remaining craniotomy defect is repaired with a temporalis muscle graft.
It is important to emphasize that cerebral blood flow in patients with moyamoya disease is tenuous.
Several rules need to be followed. First, the patient’s PaCO2 should be kept normal and hyperventilation
should be avoided. Hyperventilation can lead to an intraoperative steal phenomenon, resulting in strokes
both in and out of the circulation that is being treated surgically. Second, the patient should be kept
euvolemic, with close attention paid to blood loss throughout the surgical procedure. Third, the bone
flap should be reattached to the skull with sutures to avoid rigid fixation. Because the temporalis muscle
sits underneath the bone, close approximation of the bone flap to the cranium can result in a significant
mass effect and a stroke. Fourth, because the temporal artery is harvested from the skin flap, it is

429
important to design a flap that has a broad pedicle. Furthermore, the topical application of nitroglycerin
paste to the skin flap can be beneficial in decreasing ischemic necrosis of the scalp.

Figure 11-23. Step 1. The skin flap is

designed so that the temporal artery can be
harvested from the scalp and dissected free
of the temporalis muscle but preserved in
continuity in case an
encephalomyoarteriosynangiosis will be
performed. This type of skin flap works
better than a horseshoe type because there is
less risk of ischemic necrosis of the scalp
flap.

430
Figure 11-24. Step 2. A and B, The
temporal artery is dissected free from
the galea aponeurotica but preserved
in continuity. The temporalis muscle is
dissected off the calvarium and split to
allow passage of the temporal artery in
preparation for an
encephalomyoarteriosynangiosis.

431
432
Figure 11-25. Step 3.
The temporalis muscle is
retracted with fishhooks.
A large
frontotemporoparietal
craniotomy is performed
with a high-speed air
drill and diamond bur
or craniotome. During
this process, the
temporal artery pedicle
is protected.

433
Figure 11-26. Step 4. The dura mater is
opened widely and tacked to the margins
of the bone. If an
encephalomyoarteriosynangiosis is to be
performed, two bands of dura mater are
left intact. The temporal artery pedicle is
sewn to these dural bands to approximate
the graft to the cortical surface. After the
dura mater has been opened and tacked to
the margins of the bone, numerous
incisions are made in the arachnoid with
use of the operating microscope. Usually,
a sharp #11 blade knife is used to make
the initial incision in the arachnoid over a
cortical vessel, and then a straight
microscissors is used to lift up the
arachnoid and to cut it over the length of
the vessel as far as the subarachnoid
space permits.

434
Figure 11-27. Step 5. After the
arachnoid has been incised,
the temporal artery pedicle is
approximated to the two dural
bands with either interrupted
or running sutures. With each
successive incision in the
arachnoid, there will be some
drainage of cerebrospinal
fluid, which makes the next
incision more difficult.
Accordingly, the surgeon
should start by incising the
arachnoid over the largest
vessels closer to the ear and
work toward the midline. The
collection of cerebrospinal
fluid tends to be gravity-
dependent; therefore, despite
the incisions in the arachnoid
over the temporal lobe and
Sylvian fissure, there usually
are some large subarachnoid
spaces along the parietal lobe
close to the inferior edge of the
bone that can be incised.
435
Figure 11-28. Step 6. The
myosynangiosis is
performed by sewing the
temporalis muscle and its
fascia to the cut edge of the
dura mater covering all the
exposed cerebral cortex.
After the
encephalomyoarteriosynan
giosis has been performed,
the bone flap is applied
loosely with monofilament
sutures. The skin flap is
closed in two layers.
Nitroglycerin paste is
applied to the scalp, and a
loose head dressing is
applied.

436
Figure 11-29. If the temporal artery and cortical
vessels are of sufficient size, a direct superficial
temporal artery-to-middle cerebral artery
anastomosis should be performed. The recipient
cortical vessel is dissected free from its
arachnoid over a distance of approximately 1.5
cm. A rubber dam is placed under the vessel to
prop it up. Before the artery is occluded
temporarily, thiopental (2 to 3 mg/kg) is
administered to provide some cerebral
protection. The artery is occluded temporarily
with two microwire clips. An arteriotomy is
made with a broken razor blade or a sharp #11
blade knife and extended with straight
microscissors. Depending on the surgeon’s
preference, the anastomosis can be performed
either with running or interrupted 10-0 or 9-0
monofilament sutures. During preparation of the
cortical vessel, the temporal artery is irrigated
with a papaverine-heparin solution and then
occluded at its base to prevent thrombus
formation within the vessel. After the
anastomosis has been performed, some oozing
may occur from the suture line. This usually can
be stopped by applying an absorbable gelatin
sponge (Gelfoam).

437
Figure 11-30. After the
superficial temporal artery-
to-middle cerebral artery
bypass has been completed,
the myosynangiosis is
performed by anastomosing
the temporalis muscle and its
fascia to the cut edge of the
dura mater. The wound is
closed as described above.

438
VERTEBRAL ARTERY TRANSPOSITION

The incision is placed approximately one fingerbreadth above the clavicle. It extends for 7 to 8 cm
laterally from the suprasternal notch. The platysma, sternocleidomastoid, sternohyoid, and sternothyroid
muscles are divided. This exposes the common carotid artery, the internal jugular vein and the vagus
nerve. The artery and vein are separated free from their loose fascial tissue. The vertebral artery call be
exposed by two routes. First, the jugular vein and the vagus nerve can be retracted laterally and the
vertebral artery identified in the deep cervical fascia. Exposure in this way does not require sectioning
the anterior scalene muscle or displacing the phrenic nerve. On the right side, the surgeon should be alert
for the recurrent laryngeal nerve, which needs to be protected if visualized. On the left side, the thoracic
duct may be visualized; if it is, it should be ligated. The second type of exposure requires displacement
of the jugular vein medially. The phrenic nerve is identified on top of the scalene muscle, running from
lateral to medial. This nerve is retracted laterally by using a small vascular loop. The antenor scalene
muscle is incised horizontally to allow the proximal vertebral artery to be identified. The vertebral artery
is dissected free from its surrounding tissue and its submergence behind the longest colli muscle.
Occasionally, sympathetic fibers cross the vertebral artery; they are divided.
After the vertebral artery has been mobilized, it is ligated proximally and divided. The common
carotid artery is occluded temporarily with two soft-shoe Fogarty clamps. A Goosen punch is used to
create an opening in the common carotid artery. The anastomosis of the vertebral artery to the common
carotid artery is performed under the operating microscope using interrupted 8-0 monofilament sutures.
It is best to use interrupted sutures because a running suture may lead to stenosis of the anastomosis. The

439
back wall is sewn in place first. Before the opening is made in the common carotid artery, it is important
to determine the best angle and course of the vertebral artery that will allow a gentle curve without
causing kinking or tension along the suture line. Occasionally, if there is significant atherosclerotic
plaque in the common carotid artery, it may be necessary to perform an endarterectomy before the
vertebral artery transposition. To increase collateral blood flow, induced hypertension should be used
during the operation, as should intraoperative encephalography.

Figure 11-31. The incision is made approximately one

fingerbreadth above the clavicle. It extends for 7 to 8 cm
laterally from the suprasternal notch. As illustrated here, the
jugular vein and the vagus nerve are retracted medially and
the phrenic nerve is retracted laterally. This allows exposure
of the anterior scalene muscle, which is divided to expose the
vertebral artery. The thoracic duct may be visualized on the
left side; it should be ligated. A sufficient length of the
vertebral artery needs to be mobilized.

440
Figure 11-32. A small vascular
loop can be placed around the
phrenic nerve and gently
retracted laterally to keep this
nerve out of harm’s way. After the
anterior scalene muscle has been
sectioned, the vertebral artery is
dissected free from the
surrounding soft tissue and
ligated as low as possible in the
operative field. The more the
vertebral artery is mobilized, the
easier the anastomosis.

441
Figure 11-33. The common
carotid artery is occluded with
two softshoe Fogarty clamps. A
Goosen punch is used to create
the opening in the common
carotid artery. It is important to
place this opening in such a
location that the vertebral artery
makes a gentle curve without
kinking. Under the operating
microscope, the vertebral artery
is sewn in place to the common
carotid artery with interrupted
8-0 monofilament sutures. During
the carotid occlusion, induced
hypertension is used to increase
collateral blood flow.

442
SAPHENOUS VEIN BYPASS

Saphenous vein bypass procedures can be beneficial in treating both ischemic disease and large
broad-based fusiform type giant intracranial aneurysms. For example, a saphenous vein bypass graft
from the cervical carotid artery to the posterior cerebral artery may be useful in treating medically
refractory high-grade stenosis of the basilar artery. Giant aneurysms of the internal carotid or middle
cerebral artery can sometimes be treated by the combination of proximal ligation or trapping and a distal
bypass.
An extremely important component of this operation is correct harvesting of the saphenous vein.
First, it should not be overly manipulated during the dissection, and the adventitia surrounding the vessel
should be preserved. The small branches and tributaries are ligated with 5-0 or 6-0 monofilament sutures
instead of bipolar cautery. Before the saphenous vein is removed, a 6-0 monofilament suture is sewn
loosely into the adventitia from proximal to distal to allow for the correct orientation of the vein (Garrett
line). This prevents the vein from twisting during its placement through the subcutaneous tunnel from
the craniotomy to the neck. After the vein has been harvested, it is irrigated with a cold solution of
papaverineheparin. A vein dilatation kit is used to prevent overdistention of the vein.
For procedures on the anterior circulation, either the supraclinoid internal carotid artery or the
middle cerebral artery is exposed through a standard pterional craniotomy. For anastomosing into the
middle cerebral artery, an M2 segment is best. Sewing into the middle cerebral artery is technically
easier than performing a direct end-to-end or end-to-side anastomosis to the supraclinoid internal carotid
artery. The anastomosis is performed with interrupted 9-0 or 8-0 monofilament sutures. Although a
443
running suture requires less time, there is a small risk of stenosis at the site of the anastomosis. The
length of the anastomosis is approximately 1.0 cm. Before the middle cerebral artery or the internal
carotid artery is occluded, thiopental (3 to 5 mg/kg) or etomidate is administered intravenously. Induced
hypertension is used to increase collateral blood flow.
For procedures on the posterior circulation, the subtemporal approach is used to expose the posterior
cerebral artery. The craniotomy is carried low to the floor of the middle cranial fossa. Also,
cerebrospinal fluid is drained through a lumbar needle and mannitol is injected to achieve brain
relaxation. In the subtemporal approach, it is important to protect the vein of Labbé with absorbable
gelatin sponges and cottonoids (Americot). The entire temporal lobe is protected with hemostatic fabric
(Surgicel) and cottonoids before it is retracted. This is an important point because before the posterior
cerebral artery is occluded, 5,000 units of heparin are administered. As the posterior cerebral artery
courses around the peduncle, it is dissected free from the arachnoid. Typically, the anterior half of the P2
segment is chosen for the anastomosis posterior to the oculomotor nerve. This segment of the posterior
cerebral artery usually has no perforating vessels. Specifically, the thalamoperforating arteries arise
along the P1 segment and the posterior choroidal and posterior temporal arteries arise from the distal P2
segment. The distal anastomosis in a posterior circulation bypass is an extremely exacting and
technically difficult procedure. The anastomosis is approximately 1.0 cm long. It is performed with
running 9·0 monofilament sutures. A running suture is preferable to interrupted sutures because limited
space at the depth of the wound makes tying the suture difficult; therefore, the use of running sutures
saves time.

444
 Before the distal anastomosis is performed, a #20-French Argyle trocar catheter is inserted into the
temporalis muscle and passed over the zygoma, through the deep layers of the subcutaneous tissue in
front of the ear, superficial to the parotid gland, and down into the neck wound. A #2 silk suture is then
brought through this catheter with a trocar. The saphenous vein is tied to the silk suture and brought
through the Argyle catheter. The previously placed Garrett line is used to make sure that the vein is not
rotated in the Argyle catheter.
The importance of the proximal anastomosis needs to be emphasized. A poorly constructed proximal
anastomosis will lead to occlusion of the vein graft. Various proximal anastomoses can be used. For a
posterior circulation bypass, the saphenous vein should be sewn to the stump of the external carotid
artery with either a fish mouth or spatulated anastomosis. For giant aneurysms of the anterior circulation
in which trapping is performed, it still is useful to sew to the external carotid artery. In this way, a
saphenous vein can be constructed completely before the external carotid artery is occluded, thereby
decreasing the fisk of ischemic injury. If the external carotid artery is small, an end-to-side anastomosis
to the common carotid artery can be performed using a Goosen punch. If a preoperative trial balloon
occlusion suggests that temporary occlusion of the internal carotid artery can be tolerated, the saphenous
vein can be sewn directly end-to·end to the internal carotid artery.

445
Figure 11-34. This is the general scheme for
intracranial-to-extracranial bypass using a
long-vein saphenous vein bypass. In this
illustration, a giant fusiform aneurysm of the
middle cerebral artery is being trapped.
Blood flow to the distal middle cerebral
artery is preserved through a saphenous vein
graft from the external carotid artery to the
middle cerebral artery beyond the aneurysm.

446
Figure 11-35. A critical step in the preparation for this
type of bypass is harvesting the saphenous vein from the
leg. The saphenous vein is identified adjacent to the
medial malleolus and dissected upward just beyond the
popliteal fossa. During this dissection, the various
branches of the saphenous vein are ligated with a 6-0
monofilament suture instead of bipolar cautery. The
adventitia of the vein should be preserved. Before the vein
is removed from the leg, a Garrett line is sewn in place
with a 6-0 monofilament suture. This Garrett line is
important in maintaining the correct orientation of the
vein as it is passed through the subcutaneous tunnel from
the craniotomy to the neck. After the vein is removed, a
mechanical spasm usually occurs. This spasm can be
broken with gentle massage. Dilatation of the vein is
achieved with a standard vein dilatation kit, using a cold
papaverine solution. After the vein has been expanded, it
is irrigated with a heparin solution.

447
Figure 11-36. The
subcutaneous tunnel from
the craniotomy to the
cervical carotid artery is
achieved with a #20-French
Argyle trocar catheter. The
trajectory of this catheter is
in front of the ear and
superficial to the parotid
gland. After the Argyle tube
has been placed, a #2 silk
suture is brought through it.
This is attached to the vein,
which then is also brought
through the catheter using
the Garrett line to preserve
the correct orientation.

448
Figure 11-37. In this illustration, a branch of the
M2 segment has been chosen for the site of the
anastomosis so that blood will still flow through
most of the middle cerebral artery complex during
the temporary occlusion required for the
anastomosis. A rubber dam is useful in propping
up the middle cerebral artery. The length of the
anastomosis is approximately 1.0 cm. Although
running sutures work well, there is a small risk of
stenosis at the site of the anastomosis; thus,
interrupted sutures are preferable. After the distal
anastomosis has been achieved, the temporary
clips on the middle cerebral artery are removed.
The saphenous vein is irrigated again with
heparin, and a temporary clip is applied close to
the distal anastomosis to prevent blood from
running down the saphenous vein and possibly
causing thrombus formation because of
stagnation. When performing an extracranial-to-
intracranial saphenous vein bypass for ischemic
disease, 5,000 units of heparin should be
administered systemically before the initial
anastomosis is performed. However, whether
heparin should be used during bypass surgery for
intracranial aneurysms is a matter of controversy.
449
Figure 11-38. As illustrated here, the saphenous
vein is being sewn to the external carotid artery
just beyond the bifurcation of the common
carotid artery. With this procedure, blood now is
maintained through the internal carotid artery
throughout the operation. Furthermore, after the
aneurysm of the middle cerebral artery has been
trapped, the internal carotid artery will continue
to perfuse the anterior cerebral, the posterior
communicating, and the anterior choroidal
arteries. Often, the best proximal anastomosis is
a spatulated type, using interrupted 6-0
monofilament sutures. Just before the last few
sutures are tied, it is best to use back bleeding
down the saphenous vein and up the external
carotid artery to flush out any residual air in the
graft. In the example shown here, the aneurysm
is trapped with aneurysm clips after flow is
restored. For ischemic disease procedures, the
heparin is not reversed. It is important to give
aspirin to all these patients after the operation.

450
Chapter 12
Cerebrospinal
Fluid Drainage
Chapter 12: Cerebrospinal Fluid
Drainage Procedures

EXTERNAL VENTRICULAR DRAIN

452
 Figure 12-1. Placement of an external
ventricular drain can be performed at the
bedside, in the emergency room, or in the
operating room. The ease of placement depends
on the size of the ventricles and the presence or
absence of a midline shift. If the ventricle is
very small, the use of stereotaxis or external
scalp support devices should be considered.
The patient’s head is held in a donut or round
wash basin in a straight up, slightly flexed
position, with 0 degrees of rotation. It is useful
to have the nurse or assistant hold the patient’s
chin underneath the drapes to prevent
movement. The twist drill or bur hole should be
placed 3 cm lateral to the midline and 1 to 2 cm
anterior to the coronal suture. The ventricular
catheter is passed perpendicular to the brain.
Therefore, the trajectory will point to an
imaginary line extending back from the inner
canthus of the ipsilateral eye. In most cases, the
surgeon will enter the lateral ventricle at 5 to 6
cm from the calvarium. Typically, there is a
feeling of resistance and then popping as the
lateral ependymal surface is punctured. After the
ventricle has been cannulated, the catheter should
be inserted an additional 2 cm. The metal stylus
within the catheter is removed to ensure that there is
good flow of cerebrospinal fluid. The catheter is
reoccluded to prevent excessive drainage of

453
Figure 12-2. It is optimum to have the catheter run
underneath the galea for approximately 5 to 6 cm
before it is brought out through the skin. Some
evidence suggests that placing the catheter in a
funnel decreases the risk of contamination and
infection. The wounds are closed with a
monofilament suture. Collodion can also be
applied to the incision overlying the twist drill. A
full head dressing is applied. Various draining
systems can be used to regulate the amount of
cerebrospinal fluid drainage and intracranial
pressure.

454
VENTRICULOPERITONEAL SHUNT

Figure 12-3. Step 1. Illustrated here is the general body

position for placement of a ventriculoperitoneal shunt. Under
ideal circumstances, the head should be rotated laterally,
parallel to the floor. Placement of a towel underneath the
shoulder helps with head rotation. In older patients with
cervical spondylosis, head rotation may be difficult. In fact,
placement of a ventricular shunt through a frontal bur hole
may be technically easier. For parietooccipital shunt
placement, the bur hole should be made 6 cm up from the
external occipital protuberance and 3 cm lateral from the
midline. This will approximate the bur hole over the
lambdoid suture. Placing the bur hole on top of the lambdoid
suture is advantageous because frequently the dura mater is
stuck to the underlying calvarium, which means there often is
less epidural bleeding. Either a midline or a right subcostal
incision can be used to expose the peritoneum.

455
Figure 12-4. Step 2. After the bur hole has been placed and the peritoneum exposed, a trocar is used to
create a tunnel between the two incisions. A 2-0 suture is tied to the trocar and passed through the
subcutaneous tunnel. Often, it is necessary to make a third incision near the clavicle. The subcutaneous
tunnel adjacent to the occipital incision is spread with hemostats to create a subcutaneous pocket to
contain the shunt valve.

456
Figure 12-5. Step 3. The shunt tube is passed between the two incisions. It is important not to let the
shunt tube touch the skin. Because of this, the patient is draped with iodine-impregnated sterile (Steri)
drapes to cover the skin overlying the surgical sites. After the shunt tube has been brought through the
tunnel, the longer abdominal end is wrapped in a towel.

457
Figure 12-6. Step 4. Many different shunt systems are available, and many of them
have a peritoneal catheter already attached to the valve. If this is not the case, the
chosen shunt valve is attached to the peritoneal catheter.

458
Figure 12-7. Step 5. The dura mater is cauterized and opened in a cross fashion. It is important to not
cut the dura mater back to the bone edge, otherwise nuisance bleeding may occur along the dural edge,
which cannot be cauterized because of the overlying bone. The pia mater is cauterized and incised with
a #11 blade knife, avoiding any major vessels in the area. Next, the ventricular catheter is passed into
the ventricle using a trajectory that points the catheter toward a line extending posteriorly from the
inner canthus of the ipsilateral eye. Because of this trajectory, the ventricular catheter will be
perpendicular to the brain at its entry. It is important to recognize that a more lateral trajectory has the
risk of injury to the posterior limb of the internal capsule. Alternatively, if the trajectory is too medial,
the catheter will miss the ventricle and hit the falx cerebri. At approximately 5 to 6 cm, the catheter
should have punctured the ependyma and be sitting within the atrium of the lateral ventricle. If the
ventricle has not been cannulated by 7 cm, it is best for the surgeon to withdraw the catheter and try a
slightly more medial instead of lateral trajectory.

459
Figure 12-8. Step 6. After the ventricle has
been cannulated, the stylet is removed; the
surgeon’s assistant should immediately
pinch the catheter with atraumatic forceps
to prevent cerebrospinal fluid drainage.
One cause of subdural hygroma in
patients with normal pressure
hydrocephalus is ventricular collapse due
to excessive removal of cerebrospinal fluid
at the time of shunt placement. There are
several preferences for the length of the
catheter that should be inserted into the
ventricle. Some surgeons advocate
placement of 10 cm of catheter, which
places the tip in the frontal horn of the
ipsilateral ventricle away from the choroid
plexus. Other surgeons advocate a 7- to 8-
cm placement, which places the catheter
adjacent to the foramen of Monro.
Whatever the choice, it is important to
advance the catheter without a stylet to
make sure that the catheter remains within
the ventricle.

460
Figure 12-9. Step 7. Often, the
valve is connected to the shunt
with an L-shaped connector. It is
best to tack the valve to the
adjacent pericranium with a
small monofilament suture. The
valve is pumped while the distal
end is observed to make sure
that there is expression of
cerebrospinal fluid. Thereafter,
approximately 25 cm of catheter
is placed into the peritoneal
cavity. All wounds should be
closed in multiple layers. The
use of absorbable monofilament
instead of braided sutures may
decrease the risk of shunt
infection.

461
 VENTRICULOATRIAL SHUNT

Figure 12-10. Step 1. Illustrated here is the position for placement of a ventriculoatrial shunt. After induction, the anesthesiologist should
insert a transesophageal echocardiography probe. This instrument can be quite useful in confirming the correct location of the atrial catheter
within the right atrium. If transesophageal echocardiography is not available, other techniques are available to confirm correct catheter
placement, including intraoperative radiography and electrocardiography. For the latter technique, an electrocardiographic clip is attached to
the ventricular catheter, which is filled with saline. As the catheter is inserted from the superior vena cava into the right atrium, the P waves
become progressively larger. As the catheter progresses into the right ventricle, the P waves become biphasic. At this point, the catheter is
withdrawn back into the right atrium. It is important to fill the catheter with saline, because otherwise a current will not be conducted and the
electrocardiogram will be useless. Overall, intraoperative transesophageal echocardiography is more accurate and easier to use. An
additional advantage is that after the shunt has been placed, pumping of the valve causes an efflux of cerebrospinal fluid and air bubbles into
the right atrium, which can be visualized on the echocardiogram, confirming shunt patency. The parietooccipital bur hole is placed 6 cm above
the external occipital protuberance and 3 cm lateral from the midline. One of the more difficult aspects of a ventriculoatrial shunt procedure is
identification of the facial vein. A horizontal incision two fingerbreadths below the angle of the jaw is used. The surgeon can usually palpate
the notch created by the facial artery in the mandible, just anterior to the angle of the jaw. Running along with this artery is the facial vein. A
horizontal incision should be placed below this notch.

462
Figure 12-11. Step 2. After the occipital bur
hole has been placed, the facial vein is
identified through a horizontal incision two
fingerbreadths below the facial artery notch in
the mandible. It is useful but not mandatory to
isolate the facial vein down to its junction with
the internal jugular vein. One suture is placed
distally around the facial vein and held with
some tension using a hemostat. A second loose
suture is placed around the proximal facial
vein under tension to prevent back bleeding.
Thereafter, a linear incision is made, and the
atrial catheter is inserted down into the jugular
vein toward the superior vena cava. Before
insertion, it is important to make sure that the
atrial catheter has been flushed with saline and
attached to a syringe. If significant resistance
is met when passing the catheter, it is likely
that the catheter is either in the external
jugular vein or has turned laterally into the
subclavian vein. The catheter should be
withdrawn and reinserted. If resistance is met
again, the cannulated vein should be dissected
further to make sure that it is the facial vein
that is entering the internal jugular vein.

463
Figure 12-12. Step 3. The atrial catheter is passed from the neck incision to the cranial incision through a
subcutaneous tunnel previously made with a trocar. The valve is then attached. It is important during this time to
make sure that the atrial catheter is filled with saline and occluded with a small atraumatic clamp to prevent
blood from sitting in the atrial catheter and possibly forming thrombus. After the valve has been connected, it is
useful to attach a manometer to ensure that the opening and closing pressures are correct. Next, the dura mater
is cauterized and opened in a cross fashion. The pia mater is cauterized, and a ventricular catheter is inserted,
similar to that described for placement of a ventriculoperitoneal shunt. This is hooked to the valve. The valve is
then pumped to make sure there is normal compression and refilling. As mentioned above, echocardiography can
be used to confirm shunt patency by visualizing air bubbles exiting from the atrial catheter.

464
 Figure 12-13. Placement of a third
ventriculostomy is an extremely useful
THIRD VENTRICULOSTOMY procedure if the patient has obstructive
hydrocephalus. In this illustration, the child has
aqueductal stenosis due to a neoplastic process
involving the tectal plate. Step 1. The patient is
positioned supine with the head straight up,
with 0 degrees of rotation, and slightly flexed.
The use of stereotaxis increases the ease and
success of a third ventriculostomy by permitting
the surgeon to determine the best trajectory that
allows the tuber cinereum to be punctured just
anterior to the mammillary bodies by passing
an instrument or scope through the foramen of
Monro. Stereotaxis is especially useful when
there is significant ventricular distortion caused
by an inflammatory or neoplastic process. If a
free hand approach is chosen, the bur hole is
made 1 cm anterior to the coronal suture and 3
cm off the midline. This is slightly posterior to
the bur hole used for placement of an external
ventricular drain. A bur hole is used instead of
a twist drill, because it is important to have
some room to maneuver the scope in passing
through the foramen of Monro.

465
Figure 12-14. Step 2. A, Illustrated here is the typical
appearance of the right foramen of Monro as visualized
through an endoscope. Usually, the foramen of Monro is
enlarged because of the fornix being stretched. The typical
intraventricular landmarks visualized include the medial
septal vein, choroidal plexus, thalamostriate vein, and lateral
anterior caudate vein. It is important not to traumatize the
fornix as the endoscope passes through the foramen of
Monro. Most endoscopes have a side port that can be used for
irrigation. Irrigating through the scope as it is inserted
through the foramen will gently push the fornix out of harm’s
way. Step 3. B, As the scope is inserted through the foramen
of Monro, the mammillary bodies become more visible. The
point of puncture is anterior to the mammillary bodies in the
tuber cinereum. Looking through the endoscope, anterior is
the dorsum sellae and clivus. If the tuber cinereum is thin, the
outline of either the basilar or the posterior cerebral artery
may be seen. Step 4. C, There are several techniques for
puncturing the tuber cinereum. For example, the endoscope
itself can be used. More sophisticated scopes have a side port
that allows passage of a dilating balloon to enlarge the third
ventriculostomy. Often, there is slight bleeding, which will
stop with irrigation. D, The endoscope can be inserted deeper
to better visualize the basilar artery. Usually, there are
strands of arachnoid between the clivus and the basilar
artery. The surgeon should not attempt to cut or to lyse these
arachnoid trabeculations. If the patient has a functioning
shunt, it is important to ligate it to make sure that the flow of
cerebrospinal fluid is directed through the third
ventriculostomy to increase the likelihood of patency.
466
Chapter 13
Percutaneous Procedures for
Trigeminal Neuralgia
Chapter 13: Percutaneous Procedures
for Trigeminal Neuralgia
Facial pain can be caused by many different disorders. It is essential that an accurate assessment of
the pain be obtained before any treatment is initiated. The first step is a detailed history that includes a
description of the onset, nature, location, and character of the pain and factors that either elicit or relieve
the pain. On the basis of this information, it may be necessary to perform a computed tomographic or
magnetic resonance imaging examination or to select laboratory studies to exclude neoplastic,
infectious, or vascular causes of the pain. After this evaluation has been completed, neuralgic or non-
neuralgic facial pain can be diagnosed in most patients.
Neuralgic facial pain is commonly classified as classic trigeminal neuralgia, trigeminal neuropathy,
or atypical trigeminal neuralgia. Classic trigeminal neuralgia is characterized by intermittent, shock-like
facial pains that can be exacerbated by various triggers. Patients with classic trigeminal neuralgia are
pain-free between the episodes of facial pain. Patients with trigeminal neuropathy have a constant,
unremitting pain that cannot be elicited with cutaneous triggering events. Atypical trigeminal neuralgia
is defined as a combination of both episodic and constant facial pain.
Typical trigeminal neuralgia can be either idiopathic or secondary. Most authors believe that in
idiopathic cases vascular compression of the dorsal root entry zone causes the trigeminal nerve to

468
become hyperactive and dysfunctional. Secondary causes of trigeminal neuralgia include multiple
sclerosis and tumors of the skull base that impinge on the trigeminal nerve or ganglion. For most
patients with classic trigeminal neuralgia, medical management is successful in relieving pain for several
years. However. if the patient becomes unresponsive to medical therapy or is unable to tolerate the side
effects of the medications, surgical intervention may be required.
The goals of surgery for trigeminal neuralgia are twofold: one, to have the patient pain-free and not
require medications after the procedure, and two, to minimize the chance of the tic pain converting into
a differentiated pain syndrome such as anesthesia dolorosa. Microvascular decompression, as described
in Chapter 8, is often recommended because it does not produce a destructive lesion in the trigeminal
nerve or ganglion; thus, few patients have significant postoperative facial numbness. The incidence of
anesthesia dolorosa is also quite low. However, many patients and some physicians are wary of the risks
associated with a major intracranial procedure. Furthermore, many patients with trigeminal neuralgia are
elderly and may be considered at increased risk for complications from general anesthesia. In these
circumstances, less invasive techniques to treat the pain are desirable.
Several percutaneous techniques have been developed to treat pain by injuring the trigeminal
ganglion. The modalities used to damage the ganglion include thermocoagulation by radio-frequency
currents, chemical destruction with neurotoxic agents such as glycerol, and compression with a
microballoon. Each procedure has its relative advantages and disadvantages. Of the percutaneous
procedures, radio-frequency rhizotomy is associated with the lowest rate of pain recurrence. Studies
have documented that pain-free outcomes are strongly associated with the degree of postoperative facial
numbness after radio-frequency procedures. Thermal rhizotomy of the trigeminal nerve is associated
with the highest incidence of anesthesia dolorosa. Glycerol rhizotomy is less likely to produce changes
in facial sensation, and the incidence of postoperative differentiation pain is very low; however, the
469
recurrence rate of pain has been estimated to range from 30 to 50 percent. Similar to thermal rhizotomy,
the goal of balloon compression is facial numbness. Postoperative weakness of the masseter muscle
occurs in approximately 25 percent of patients after ganglion compression.
Some physicians recommend injection into the peripheral facial nerves as a simple, effective means
to treat facial pain. Although such procedures are generally considered palliative in nature, they offer
many advantages, especially for an elderly patient with pain in a discrete region. The procedure is
simple to perform, can be performed under local anesthesia, and has immediate results. The duration of
pain relief is short, with a median duration of 6 to 12 months.
In contrast to the results of surgery for classic trigeminal neuralgia, in which approximately 90
percent of patients become pain-free, the results of peripheral procedures for patients with atypical facial
neuralgias is poor. The mechanism of atypical facial neuralgia is thought to be located centrally in either
the nucleus caudalis, the spinothalamic tract, or the ventral posteromedial nucleus of the thalamus. A few
authors have reported good results after trigeminal tractotomy, mesencephalotomy, or thalamotomy for
atypical facial neuralgias. In general, procedures aimed at either the trigeminal ganglion or nerve should
not be performed on these patients.

ANATOMY

The successful completion of any peripheral trigeminal procedure depends on a thorough knowledge
of the pertinent anatomy. The trigeminal nerve is the primary cutaneous sensory nerve of the head and
face. The tactile and nociceptive fibers that compose the trigeminal nerve are axons of first-order
sensory neurons with cell bodies in the trigeminal ganglion. The central processes of these sensory
470
neurons synapse in either the main sensory nucleus or the spinal nucleus of the trigeminal nerve. The
trigeminal (Gasserian) ganglion rests in a depression in the medial petrous temporal bone. A dural
envelope surrounds the ganglion, producing a small cerebrospinal fluid cistern known as Meckel’s cave.
From the trigeminal ganglion emerge the three divisions of the trigeminal nerve. Each of these exits the
cranium through foramina at the base of the skull: the ophthalmic division (V1) goes through the
superior orbital fissure, the maxillary division (V2) goes through the foramen rotundum, and the
mandibular division (V3) goes through the foramen ovale. After each division exits the skull, it divides
further into the peripheral nerves that supply the skin of the head and face.
Percutaneous procedures aimed at the Gasserian ganglion require that the injurious agent (electrode,
glycerol, or microballoon) be placed intracranially adjacent to the ganglion. Entrance into the
intracranial space is achieved by passing a needle through the foramen ovale. If the tip of the needle is
directed too medially, it will meet the lateral pterygoid plate and need to be repositioned. Fluoroscopic
guidance facilitates the accurate and safe placement of the needle. Nearby foramina that can be
penetrated inadvertently by the needle include the superior orbital fissure, carotid canal, and jugular
fissure. Intraoperative fluoroscopy is also helpful in preventing the needle from penetrating too deep into
the middle cranial fossa. Passage of the needle through the cheek and infratemporal fossa may injure
branches of the external carotid artery. The usual consequence of such an injury is a hematoma of the
cheek, but this resolves.

471
PERCUTANEOUS RETROGASSERIAN GLYCEROL RHIZOTOMY

Patients who are to have percutaneous retrogasserian glycerol rhizotomy are admitted to the hospital
on the day of the operation. They are instructed to take their normal morning medications with a sip of
water before coming to the hospital. However, aspirin or other nonsteroidal anti-inflammatory
medications that may affect platelet function are not to be taken in the days preceding surgery.
Preoperative studies are few. If the patient has no history of clotting difficulties, it is not necessary to
perform coagulation studies or other blood tests.
Although general anesthesia is not required or desirable, intraoperative anesthetic management is
crucial. An intravenous catheter is inserted while the patient is in the outpatient surgery area, and the
patient is brought to the operating room or radiology suite. The patient’s condition is monitored with
intermittent cuff measurements of blood pressure, electrocardiography, and continuous pulse oximetry.
Nasal oxygen is routinely given. To reduce anxiety, patients are often given a short-acting
benzodiazepine (midazolam, 1.5-mg increments) upon arrival. Ideally, the patient will remain relaxed
but awake and able to assist with positioning of the needle throughout the operation. Sometimes, it is
also necessary to control pain with narcotics (fentanyl citrate, 25-50-g increments). To reduce the chance
of cardiac complications and hematoma formation, the patient’s systolic blood pressure should be less
than 150 mm Hg at the beginning of the procedure. Blood pressure greater than 150 mm Hg is generally
associated with an anxious patient who will have difficulty with the procedure. In such cases. the
operation is delayed until blood pressure can be brought under better control. Bradycardia and asystole
can occur in response to placing the needle through the foramen ovale or to injecting the glycerol.
Consequently, the anesthesiologist should be prepared to administer cholinergic blocking agents
(atropine sulfate, 0.4 mg) if the patient has a significant vasovagal reaction. To minimize patient

472
Figure 13-1. The safe and effective completion of the procedure requires high-
quality fluoroscopic guidance. Depending on the surgeon’s preference, the
procedure can be performed in either a radiology suite where multiplane
fluoroscopy is available or in the operating room with C-arm fluoroscopy. All
personnel in the operating room should be adequately protected with lead
gowns from radiation exposure. The procedure begins with the patient supine on
an operating room table. The patient’s head is supported throughout the
procedure by a standard cerebellar headholder.
discomfort, either a short-acting barbiturate (methohexital sodium) or propofol (10-20 mg intravenous
bolus) is given before the needle penetrates the foramen ovale. These measures are important to blunt
cardiovascular responses during the procedure. Although rare, postoperative cardiac ischemia has been
reported after percutaneous needle procedures for the treatment of trigeminal neuralgia.

Figure 13-2. Two primary methods are used to ensure accurate

placement of the needle through the foramen ovale. The first,
developed by Håkanson, uses a combination of anterior-posterior
(AP) and lateral projections to guide needle placement. A true AP
projection is obtained so that the petrous ridge appears at the level
of the inferior orbital rim. The lateral projection should have the
internal auditory canals superimposed on one another. The second
method, advocated by Apfelbaum, depends on an oblique,
submental projection to visualize the foramen ovale.

474
PERCUTANEOUS GASSERIAN MICROCOMPRESSION

Percutaneous microcompression of the trigeminal ganglion developed by Mullan uses an approach

to the foramen ovale similar to that for glycerol or radio-frequency rhizotomy. Under fluoroscopic
control, an 18-gauge needle is initially used to locate the rim of the foramen ovale, then a blunt-tip 14-
gauge needle is threaded over a wire stylet and inserted into the foramen ovale. Following radiographic
confirmation, a #3 or #4 Fogarty catheter is passed through the 14-gauge needle into the foramen and
inflated to 0.5 to 0.75 mL for 1 or 2 minutes. The pain relief following balloon compression is reported
to be equal to that after glycerol rhizotomy. One disadvantage of this approach is the indiscriminate
compression of the Gasserian ganglion, which may lead to hypesthesia (more apparent in V2 and V3
divisions of the trigeminal nerve).

475
Figure 13-3. After the patient has been properly positioned
and the fluoroscopic equipment has been tested, the
procedure can begin. Towels are placed on the chest and
neck, and the face is cleansed with alcohol. The needle used
for the procedure is a 20- or 22-gauge spinal needle, 90 mm
long. The usual site of needle entry is 2.5 cm lateral to the
angle of the mouth.

476
Figure 13-4. From this point, two lines are drawn with
a marking pen. They serve as the initial guide for
needle placement. The first line directs needle
placement in the AP and superior-inferior planes and
extends along the zygoma to a point 25 mm in front of
the ear. The second line directs needle placement in
the left-right direction and extends to the ipsilateral
medial canthus. The skin at the site of needle entry is
infiltrated with lidocaine. As the needle is passed
through the cheek, the surgeon places his or her finger
in the patient’s mouth to provide proprioceptive
guidance and to ensure that the needle does not
penetrate the oral cavity. If the needle enters the
mouth, it is withdrawn and the procedure is started
again with a new sterile needle. Fluoroscopy should be
checked frequently to ensure the proper trajectory of
the needle. Ideally, the needle can be placed on the
first attempt, to minimize the chance of injury to a
branch of the external carotid artery. If the trajectory
needs to be corrected, it is best to withdraw the needle
several centimeters before redirecting it in a new

477
Figure 13-5. Injection of absolute alcohol into one of the peripheral
branches of the trigeminal nerve is often useful in elderly patients with
pain localized to a discrete region. The procedure generally can be
performed under local anesthesia in a physician’s office. However, it
often is more comfortable for the patient to have some anesthetic
support to relieve anxiety and to minimize discomfort. Patients
generally complain of paresthesias in the appropriate distribution
whenever the needle is placed adjacent to the desired nerve. A test
injection with a local anesthetic solution (lidocaine) is performed
before the injection of alcohol. If the patient states that the pain relief is
satisfactory, the injection is repeated with absolute alcohol (0.5 mL). No
preoperative studies are needed.

The patient is placed in the supine position. The appropriate

region is then cleansed with alcohol. For injection into the supraorbital
and supratrochlear nerves, the medial one-third of the superior orbital
rim is palpated for the supraorbital notch. After the overlying skin has
been anesthetized, the needle is advanced into the supraorbital foramen
in an inferior-superior direction and absolute alcohol is injected. The
infraorbital nerve can be injected with alcohol at the infraorbital
foramen. The foramen can be palpated at a point 1.5 cm lateral to the
nasolabial fold and 1 cm below the inferior rim of the orbit The needle
should be advanced in an inferior-superior and slightly medial to
lateral direction. For lower jaw pain, the mental nerve can be injected
with alcohol where it exits the mental foramen. The mental foramen is
often difficult to palpate. To estimate its position, a line is drawn
through the supraorbital and infraorbital foramina onto the mandible.
The needle is advanced in a superior-inferior and slightly lateral to
medial direction. Again, to ensure a satisfactory clinical result, it is
important that the patient report paresthesias in the expected
distribution during needle placement before alcohol is injected.
Injections into peripheral nerve can be repeated as required, although
pain relief often is not as good after subsequent procedures.

478