J Neurosurg 96:453–463, 2002

Trigeminal schwannomas: removal of dumbbell-shaped

tumors through the expanded Meckel cave and outcomes
of cranial nerve function

OSSAMA AL-MEFTY, M.D., SAMER AYOUBI, F.R.C.S.(I),

AND ESAM GABER, M.B.B.CH., M.CH., PH.D.

Department of Neurosurgery, University of Arkansas for Medical Sciences, Little Rock, Arkansas

Object. As in patients with vestibular schwannomas, advances in surgical procedures have markedly improved
outcomes in patients with trigeminal schwannomas. In this article the authors address the function of cranial nerves
in a series of patients with trigeminal schwannomas that were treated with gross-total surgical removal. The authors
emphasize a technique they use to remove a dumbbell-shaped tumor through the expanded Meckel cave, and dis-
cuss the advantage of the extradural zygomatic middle fossa approach for total removal of tumor and preservation
or improvement of cranial nerve function.
Methods. Within an 11-year period (1989–2000), 25 patients (14 female and 11 male patients with a mean age
of 44.4 years) with benign trigeminal schwannomas were surgically treated by the senior author (O.A.) with the aim
of total removal of the tumor. Three patients had undergone previous surgery elsewhere. Trigeminal nerve dysfunc-
tion was present in all but two patients. Abducent nerve paresis was present in 40%. The approach in each patient
was selected according to the location and size of the lesion. Nineteen tumors were dumbbell shaped and extended
into both middle and posterior fossae. All 25 tumors involved the cavernous sinus. The zygomatic middle fossa ap-
proach was particularly useful and was used in 14 patients. The mean follow-up period was 33.12 months. In pa-
tients who had not undergone previous surgery, the preoperative trigeminal sensory deficit improved in 44%, facial
pain decreased in 73%, and trigeminal motor deficit improved in 80%. Among patients with preoperative abducent
nerve paresis, recovery was attained in 63%. Three patients (12%) experienced a persistent new or worse cranial
nerve function postoperatively. Fifth nerve sensory deficit persisted in one of these patients, sensory and motor dys-
function in another, and motor trigeminal weakness in the third patient. In all patients a good surgical outcome was
achieved. One patient died 2 years after treatment from an unrelated cause. In three patients the tumors recurred af-
ter an average of 22.3 months.
Conclusions. Preservation or improvement of cranial nerve function can be achieved through total removal of a
trigeminal schwannoma, and skull base approaches are better suited to achieving this goal. The zygomatic middle
fossa approach is particularly helpful and safe. It allows extradural tumor removal from the cavernous sinus, the in-
fratemporal fossa, and the posterior fossa through the expanded Meckel cave.

KEY WORDS • cranial nerve • cavernous sinus • Meckel cave •

trigeminal schwannoma • cystic tumor • brain neoplasm • skull base surgery

is possible of course that a method may some day be
I
T
evolved whereby a Gasserian neurinoma or meningio-
ma, even after it has crossed the ridge, may be safely
approached and removed. Should this come to pass, it will
be another conquest for neurosurgery.—Harvey Cushing11
THE treatment of trigeminal schwannomas is fast approach-
ing the fulfillment of Dr. Cushing’s prediction. With the ad-
vent of microsurgical techniques and skull base approach-
es, a remarkable improvement has been seen in outcomes
in patients undergoing surgical treatment for trigeminal
schwannomas,12,13,31,35,38,42,48,49,52 a trend that parallels that of
Abbreviations used in this paper: CT = computerized tomog-
raphy; MR = magnetic resonance; SPECT = single-photon emis-
sion CT.
patients with vestibular schwannomas. Consequently, the
emphasis of treatment has shifted to the patient’s quality of
life, with the aim to recover or preserve cranial nerve func-
tion. In this report we focus on cranial nerve function and
examine the eventual outcome of cranial nerve deficits,
whether they are present preoperatively or develop post-
operatively. Various surgical approaches have been used
for the treatment of these tumors; the proper selection of
the surgical approach minimizes postoperative deficits. We
emphasize the advantage of using the zygomatic middle
fossa approach for the extradural removal of even large,
dumbbell-shaped tumors, which are known to be the most
challenging. By expanding the Meckel cave, a large portion
of the posterior fossa can be removed carefully through the
middle fossa without drilling the petrous apex. We present
the anatomical basis of and the surgical techniques for this
maneuver.

J. Neurosurg. / Volume 96 / March, 2002 453

 O. Al-Mefty, S. Ayoubi, and E. Gaber

TABLE 1
Clinical findings in 25 patients with trigeminal schwannoma*
Signs
Symptoms
Corneal
Case DOS Numb- Sensory TM Reflex Cere- Brain-
No. (mos) Facial Pain ness Other Deficit Deficit Deficit Other CN bellar stem

1 12 tic douloureux
 V2 N N   
2 12 V1, V2
headache V1 N N  ataxia 
3 3 tic douloureux V1–3 nausea, headache V1, V2

  
4 12 periorbital  headache, diplopia N N N VI ataxia 
5 ? tic douloureux V2  V2 N N   
6† 12

diplopia V1–3 N
 ataxia 
7 ?  V2 blind spot: uveitis V2 N N   
8† 4 retrobulbar  hemicrania N wasting N   
9 4

vertigo N N
  
10† 36

diplopia V1–3

VI, IX, X, XII ataxia 
11 24  V1, V2 diplopia V1–3 N N VI  
12 96 tic douloureux V1–3 diplopia V1–3 atrophy N VI ataxia 
13 60

diplopia V1–3

VI  
14† 12  V1–3 diplopia V1–3 N N IV  
15 24 
diplopia V2, V3 N
VI, VIII, IX, X  
16† 4 
diplopia V1, V2 N
VI  
17 2  V2  V2 N
  
18 5  tingling diplopia V2, V3

IV, VI  
19† 19

diplopia, visual V1–3 N N   
obscuration
20† ?

diplopia, headache, V2, V3 N N IV, VI nystag-

blurred vision mus
21† 2   ear fullness, ringing N N N VIII  
22† 9 retroorbital  diplopia N N N IV, VI  
23 48 
imbalance V2, V3 N N II, VIII ataxia 
24 3   dizziness N N N VIII  
25† 5 
 N N N   
total 18.5‡ 14 (56%) 19 (76%) 21 (84%) 18 (72%) 6 (24%) 9 (36%) 14 (56%) 7 (28%) 1 (4%)
* CN = cranial nerve; DOS = duration of symptoms; N = normal; TM = trigeminal motor;
= present; – = absent; ? = unknown.
† Cystic lesion.
‡ Value represents the mean duration of symptoms.

Clinical Material and Methods
Within an 11-year period (1989–2000), a series of 25 pa-
tients with histologically verified benign trigeminal nerve
schwannomas underwent surgery performed by the senior
author (O.A.). Data from some of these patients were in-
cluded in a previous publication on the overall treatment
of tumors of the cavernous sinus.14 Three patients were re-
ferred to our department because they harbored a growing
residual tumor after they had undergone a previous subtotal
resection elsewhere. All patients underwent detailed gener-
al, ophthalmological, and neurological examinations. Au-
diograms were obtained in 12 patients.
Neuroimaging evaluation of these patients included pre-
operative MR imaging, CT scanning, and, early in the se-
ries, cerebral angiography with cross-compression studies
or a balloon-occlusion test with SPECT scanning to evalu-
ate cerebral collateral circulation. Magnetic resonance an-
giography replaced angiography later in the series.
All patients underwent surgical excision aimed at total
removal of the tumor. Intraoperative monitoring included
recording of brainstem auditory evoked responses and so-
matosensory evoked potentials with median nerve stimula-
tion. Electromyographic recordings from the appropriate
muscle were obtained for monitoring the third, fifth, sixth,
and seventh cranial nerves. When necessary, the lower cra-
nial nerves were monitored using electromyography.
The postoperative evaluation and follow-up review of
patients involved general and neurological examinations. A
CT scan was obtained early after surgery. Within 1 week af-
ter surgery MR imaging was usually performed to evaluate
the extent of tumor excision. Additional MR images were
obtained approximately 6 months after surgery and, later,
annually to detect any recurrence. The pathological exami-
nation included intraoperative frozen section and formalin-
fixed specimens to which routine and special stains were
applied.
Results
This series consists of 14 female and 11 male patients.
The mean age of the patients was 44.4 years (range 11–80
years), and the peak incidence of tumor discovery occurred
during the patient’s fourth decade. Table 1 lists the clinical
findings in the 25 patients. The mean duration of symptoms
before treatment (known for 22 patients) was 18.5 months
(range 2–96 months).
Signs and Symptoms
At the time of surgery, all but two patients manifested tri-
geminal nerve dysfunction (sensory, motor deficit, or pain).
Fourteen patients experienced facial pain, four of whom
displayed the typical tic douloureux. The other 10 patients

454 J. Neurosurg. / Volume 96 / March, 2002

Trigeminal schwannomas
described their pain as periorbital or retroorbital (three pa-
tients), or as facial pain with undetermined characteristics
(seven patients). Nineteen patients reported having numb-
ness along one or more divisions of the trigeminal nerve.
During the physical examination, 21 (84%) of the 25 pa-
tients displayed trigeminal dysfunction. Eighteen experi-
enced sensory disturbances, namely, hypesthesia, dyses-
thesia, or hypalgesia, along one or more divisions of the
trigeminal nerve. One patient exhibited a reduced corneal
reflex only, without reduced sensation in any of the three
dermatomes, and another patient only had only fifth nerve
motor weakness. Nine patients (36%) displayed a de-
creased or absent corneal reflex. Decreased motor function,
appearing as weakness or atrophy of the temporal or mas-
seter muscles, was seen in six patients (24%).
Thirteen patients (52%) complained of diplopia, which
was a presenting symptom in three patients. Four patients
(16%) exhibited a deficit of the fourth cranial nerve, and 10
patients (40%) a deficit of the sixth nerve. Two patients
each (8%) displayed deficits of the ninth and 10th cranial
nerves (one due to tumor and the other as a result of previ-
ous surgery), and one patient (4%) displayed a deficit of the
12th nerve (due to a previous surgery). One patient experi-
enced episodes of obscured vision associated with reduced
visual acuity and restricted visual fields, and two patients
experienced visual symptoms due to an unrelated patho-
logical condition. Six patients displayed cerebellar signs.
Of the 12 patients who underwent audiography, five had
normal hearing, one had a positive decay reflex on the af-
fected side, and five had high-frequency hearing loss.
Neuroimaging Evaluation
Preoperative CT scans were obtained in 18 patients and
revealed bone erosion in six (33%). Contrast enhancement
demonstrated a homogeneously enhanced lesion in 10 pa-
tients. In one patient, the tumor was not apparent on CT
scans and was only revealed by MR imaging studies.
Magnetic resonance images were obtained routinely in
all patients. The lesion was isointense or slightly hypoin-
tense on T1-weighted images, hyperintense on T2-weight-
ed images, and homogeneously enhanced after administra-
tion of contrast agent. In 10 tumors a cystic component was
identified. Magnetic resonance arteriography, which was
performed in 15 patients, revealed narrowing of the inter-
nal carotid artery in only one patient. Magnetic resonance
venography, which was performed in 10 patients, visual-
ized the vein of Labbé and the exact venous sinus anatomy
in all 10 patients.
The results of the CT and MR imaging studies and the
surgical findings helped us accurately diagnose the trigem-
inal schwannomas and clearly identify the tumors’ loca-
tions, extensions, and origins. The tumor was located in the
middle fossa in six patients (24%) and in both middle and
posterior fossae in 19 patients (76%). The lesion extended
to the orbit in one patient, and the maxillary sinus and in-
fratemporal fossa were each involved in one patient.
Early in the series, cerebral angiography was performed
in seven patients and a balloon occlusion test in six. One
patient could not tolerate the occlusion, and tumor blush
was detected in another. In two patients angiography dem-
onstrated displacement of the nearby vessels. In three of the
patients who had undergone angiography and tolerated the
J. Neurosurg. / Volume 96 / March, 2002
balloon occlusion, SPECT was performed. Normal perfu-
sion was seen in all three cases. In one patient CT angiogra-
phy was performed to distinguish the tumor from a throm-
bosed aneurysm.
Surgical Procedure
The surgical approach was selected after a thorough
study of each patient’s preoperative radiological findings
and was tailored to each case.
Because of the large size of the tumor in most patients,
its extension into both middle and posterior fossae, and the
involvement of the cavernous sinus, a skull-base approach
was deemed advantageous for most patients. The cranioor-
bital zygomatic4 approach was used in three patients, the
petrosal3 or extended petrosal2 in six patients, and the pter-
ional in one patient. One extensive tumor required a com-
bination of the cranioorbital zygomatic and the petrosal
approaches (Fig. 1). We used the zygomatic middle fossa
approach in 14 patients and found it particularly helpful
and safe. It allowed extradural removal of tumor from the
cavernous sinus and removal of tumor extension in the in-
fratemporal fossa. We removed extensions into the poste-
rior fossa through the expanded Meckel cave without sec-
tioning the tentorium or drilling the petrous apex. Only
large caudal extensions could not be reached in this way.
Thus, the anatomical basis of the procedure and the zygo-
matic middle fossa approach deserve a description here.
Microsurgical Anatomy of the Meckel Cave. Johann Fried-
rich Meckel the Younger (1781–1833) described the dural
cleft that was named after him in the following manner:
. . . at the upper edge of the petrous portion of the temporal
bone it (the trigeminal nerve) enters a rounded and oblong
sheath of the dura matter, which generally is entirely separated
from the cavernous sinus. This sheath is at first loose, but after-
wards is placed strongly on its surface. It thus goes from before
downward and from behind forward on the anterior face of the
petrous portion of the temporal base.32
The Meckel cave is a dural recess extending from the
posterior fossa into the posteromedial portion of the middle
cranial fossa, providing a natural pathway that allows ac-
cess from the middle fossa to the posterior fossa. Trigemi-
nal schwannomas expand the Meckel cave, allowing entry
to the posterior fossa (Fig. 2).
The oval mouth (porus trigeminus) of the Meckel cave is
situated just below the superior petrosal sinus. The cave
is elongated forward, like a three-fingered glove, covering
the roots, ganglion, and divisions of the fifth nerve until
they reach their respective foramina. The wrist of this glove
is located where the roots are transmitted. The cave extends
anteriorly to a level abreast of the anterior wall of the pitu-
itary fossa, laterally to the foramen ovale, and upward to
half the height of the cavernous sinus. The trigeminal nerve
carries its arachnoid sheath into the middle cranial fossa. A
probe inserted into the cave along the lateral wall may
reach these levels without much resistance.29 Due to this
invagination of the dura of the posterior fossa under that
of the middle cranial fossa, a plane of cleavage is formed
laterally between the two fused layers. This plane serves
as the anatomical basis for the extradural approach to the
Meckel cave.22,23
Surgical Procedure via the Zygomatic Middle Fossa Ap-
proach. The initial steps for the zygomatic middle fossa
455
 O. Al-Mefty, S. Ayoubi, and E. Gaber

FIG. 1. Preoperative (A and B) and postoperative (C and D) MR images demonstrating a giant trigeminal schwanno-
ma, the treatment of which necessitated a combination of the cranioorbital zygomatic and petrosal approaches. Note the
extensive cystic degeneration. E and F: Photographs showing that the patient has recovered from a preexisting sixth
cranial nerve dysfunction and a postoperative third cranial nerve dysfunction, but still has facial hypesthesia.

approach are similar to those of previously described pro-
cedures.1 The dura mater of the middle fossa is elevated
from the floor of the temporal fossa to expose the middle
meningeal artery, which is coagulated and sectioned. The
foramen ovale and rotundum are identified. Posterior to
the foramen ovale, the lesser and greater petrosal nerves
are identified, and traction on the greater petrosal nerve is
avoided. Proximal control of the cavernous carotid artery is
obtained in the petrous carotid artery canal, as previously
described.46
The sheaths of the second and third divisions of the tri-
geminal nerve are elevated to expose the lower cavernous
sinus. The outer layer of the lateral wall of the cavernous si-
nus is peeled medially and posteriorly to expose the tumor,
which usually displaces the gasserian ganglion, expands
the Meckel cave, and splays the divisions of the trigeminal
nerves (Fig. 3). The tumor capsule is entered between the
divisions of the trigeminal nerve or behind the ganglion.
The lesion is debulked using suction and followed poste-
riorly into the Meckel cave. It is then carefully dissected
from the roots, ganglion, and divisions of the nerve to spare
fascicles that are not the origin of the tumor in a manner
similar to that used for other schwannomas. After the tumor
is extirpated from the middle fossa and cavernous sinus
extradurally, it is followed into the cerebellopontine angle
through the mouth of the Meckel cave without drilling the
petrous apex or sectioning the tentorium. The mouth can be
further enlarged by an upward small incision toward the su-
perior petrosal sinus, which might require coagulation and
sectioning of the most anterior portion of the superior pe-
trosal sinus. The tumor is carefully dissected from the fifth
nerve rootlets and from the brainstem and the basilar artery
and its branches by maintaining intraarachnoidal dissection
(Figs. 4 and 5).
Tumors that extend into the infratemporal fossa or the
pterygopalatine fossa are extirpated after removal of the
floor of the middle fossa between the foramen ovale and ro-
tundum. Tumors that extend along the superior orbital fis-
sure are excised after removal of the floor of the middle
fossa between the rotundum and the opening of the superi-
or orbital fissure. For patients harboring large tumors, judi-
cious placement of abdominal fat in the cavernous sinus
can prevent venous bleeding and a cerebrospinal fluid leak,
and will facilitate follow-up MR imaging examinations.
The temporal muscle is split and the partially vascularized
flap is laid on the floor of the middle fossa to support the
closure with vascularized tissue.
Surgical Complications and Outcome
Twenty-seven surgical procedures were performed in 25

456 J. Neurosurg. / Volume 96 / March, 2002

Trigeminal schwannomas

patients. Three patients who had permanent cranial nerve

deficits after undergoing previous surgery elsewhere were
excluded, and their corresponding nerve functions were not
included in the analysis of recovery. Table 2 lists postoper-
ative improvements as they relate to preoperative findings.
Preoperative deficits of all other involved cranial nerves
resolved by 4 to 6 months after surgery, except for two of
the eight cases of abducent nerve deficit. All cerebellar and
brainstem manifestations resolved postoperatively.
The surgically induced cranial nerve dysfunction de-
scribed here includes new deficits or worsening of present-
ing deficits in these nerves. It occurred in a total of 11 pa-
tients (44%) in different combinations and was alleviated in
all but three patients (12%). No intraoperative complica-
tions occurred. In seven patients (28%; Cases 1, 2, 7, 8, 9,
14, and 21) trigeminal nerve function (sensory, motor, or
corneal reflex) worsened after surgery. Five patients (Cases
2, 7, 8, 9, and 21) experienced numbness of the face, al-
though two of them had not noted any numbness before
surgery. In one patient (Case 7) facial pain developed after
surgery. Two patients (Cases 1 and 8) displayed fifth crani-
al nerve motor weakness and four patients (Cases 1, 2, 14,
and 21) exhibited a decreased corneal reflex.
Sensory function returned to nearly normal in three of
five patients, and the corneal reflex returned to normal in
all but one patient. In two patients motor function of the
fifth cranial nerve, which was affected by surgery, recov-
ered in one and persisted mildly in the other.
Other cranial nerves were affected in four patients (Cas-
es 7, 9, 11, and 20). These included trochlear nerve deficits
in two patients (8%) that fully recovered 4 months postop-
eratively. Transient third nerve palsy in one patient and
facial palsy in another patient observed 4 days after surgery
recovered completely after 1 year. Function of the abducent FIG. 2. Upper: Photograph showing an anatomical model of a
nerve was not worsened by surgery. Cerebrospinal fluid dumbbell-shaped tumor that expands the orifice of the Meckel cave
rhinorrhea occurred in one patient (4%) and resolved after and compresses the trigeminal rootlet and ganglion. Lower: Con-
lumbar drainage. One patient experienced Dilantin intox- trast-enhanced axial MR images demonstrating a comparable ex-
ication, but recovered completely. Frontalis muscle weak- panded Meckel cave with a dumbbell-shaped schwannoma preop-
ness from the peripheral frontal branches of the facial nerve eratively (lower left) and 6 months postoperatively (lower right).
over the temporal muscle occurred in four patients, but re-
solved during the follow-up period. Table 3 lists postoper-
ative complications and follow-up clinical findings in all 25 changes, but we continue to follow them closely by per-
patients. forming MR imaging examinations.
Three patients (13%) experienced tumor recurrence after
a mean period of 22.3 months (range 13–36 months). In the
first patient regrowth was identified 18 months after total
tumor removal. This recurrence was detected on MR im- Discussion
ages and the patient underwent a repeated operation 4 years First reported by Dixon in 1846, schwannomas that arise
and 10 months after the first surgery. She was free from re- from the intracranial portion of the trigeminal nerve are
currence for 5 years after the second operation, but the tu- rare, accounting for 0.07 to 0.33% of intracranial tumors
mor again recurred. The second patient who experienced and 0.8 to 8% of intracranial schwannomas.5,10,26,38,52 In
tumor recurrence exhibited ataxia and diplopia 3 years after 1952, Cuneo and Rand10 published an excellent detailed
total resection of tumor. The recurrent tumor was extirpat- history on the treatment of these lesions and stated that, in
ed radically, and he was still free from recurrence 1 year 1927, Peet had found 63 cases in the literature. Subsequent
after surgery. In the third patient tumor recurrence was re- reviews have revealed approximately 402 cases to date.5,26,52
vealed by MR imaging 13 months after surgery. There were Because of the different origins and extensions of these
no clinical manifestations of the disease and, therefore, the tumors along the trigeminal nerve, several classification
patient was followed up for 2 years until he died of an un- systems have been suggested that have implications on the
related condition at the age of 82 years. In four patients, clinical findings, surgical approach, and outcomes of pa-
postoperative MR images demonstrated a small area of tients with these tumors. Jefferson20 classified these tumors
increased enhancement that did not change on subsequent into three types: Type A, tumors located mainly in the mid-
MR images. These areas are believed to be postoperative dle fossa that arise from the gasserian ganglion; Type B, tu-

J. Neurosurg. / Volume 96 / March, 2002 457

 O. Al-Mefty, S. Ayoubi, and E. Gaber

FIG. 3. Photograph of an anatomical dissection (left) and artist’s illustrations (right and overlap) demon-
strating the surgeon’s extradural middle fossa view of a dumbbell-shaped trigeminal schwannoma with ex-
posure of the petrous carotid artery, three divisions and ganglion of the fifth cranial nerve, and entry into the
expanded Meckel cave.

mors located predominately in the posterior fossa that arise
from the root of the trigeminal nerve; and Type C, tumors
with significant components in both middle and posteri-
or fossae, which he described as dumbbell-shaped or hour-
glass tumors. Several modifications of this classification
have been described.26,38,52
Goal of Treatment
Trigeminal schwannomas are overwhelmingly benign
lesions; only a few malignant cases have been reported.6,10,
12,17,28
Even when large, they displace surrounding neuro-
vascular structures rather than engulf them. Therefore, total
removal is believed to offer the best chance of cure,13,42 and
subtotal removal is reported to be associated with higher
rates of recurrence.35 Subtotal removal is also reported to be
a source of postoperative bleeding from the tumor bed with
associated instances of morbidity and mortality.5,28 In a re-
view of the literature, Sindou and Pelissou40 found that total
removal of the tumor had been achieved in only 50% of
cases because of the close relationship of the lesion to the
cerebellopontine angle, petrous apex, cavernous sinus, and
important cranial nerves. Despite the emphasis on total re-
moval in later series,31,45 this has not always been possible
because of risks to vital structures.8,31,35 In their review of
the literature, Samii and colleagues38 found that total or
near-total removal was achieved in approximately 70% of
cases during the past 15 years. Our aim in treating patients
with these tumors has been to achieve total removal, even
if the tumor is large and involves the cerebellopontine an-
gle, petrous apex, or cavernous sinus (Table 4).
Early attempts to remove these tumors were associated
with high rates of mortality.5 Hence, the emphasis on early
detection of tumor is essential in preventing instances of
morbidity and mortality.27,44 With the advent of microsur-
gical procedures, there has been a remarkable improve-

FIG. 4. Photograph of an anatomical dissection and an artist’s illustration demonstrating the surgeon’s view after the
posterior fossa extension of the tumor has been reached and removed through the Meckel cave, exposing the basilar artery
and the brainstem.

458 J. Neurosurg. / Volume 96 / March, 2002

Trigeminal schwannomas

FIG. 5. Operative photographs demonstrating extradural expo-

sure via the middle fossa and the removal of the posterior fossa
extension through an expanded Meckel cave. Upper Left: The
tumor (T) is exposed by peeling the outer dural wall (D) of the
cavernous sinus, and the stretched rootlets (arrow) can be seen.
Upper Right: View after the tumor has been dissected, with pres-
ervation of the trigeminal nerve including its divisions (V1,V2,
and V3), ganglion (G), and roots (arrowhead). The posterior fossa
arachnoid is seen through the mouth of the Meckel cave (arrow).
Lower: Posterior fossa neurovascular structures as seen through
the Meckel cave after tumor removal. B = basilar artery; P = pons.

ment in the ability to remove trigeminal schwannomas to- the sixth and seventh cranial nerves can be a complication
tally with low risks of mortality and morbidity.8,12,13,31,35,42,45,52 of surgery. Good health with only trigeminal nerve symp-
Fifth cranial nerve deficits that developed postoperatively toms is considered to be an excellent outcome.49
have been almost universal.26 The absence of mortality or Currently, the emphasis of treatment has shifted to the
major morbidity in our series corresponds to other excellent patient’s quality of life through the preservation or im-
results reported in recent series.12,13,42,52 provement of cranial nerve function, including that of the
Stereotactic radiosurgery has been used recently to treat
patients with trigeminal schwannomas.19 Tumor shrinkage
was achieved in 56% of treated patients and tumor growth TABLE 2
arrest in 44%. No neurological deficit was observed after a
mean follow-up period of 44 months. Although the authors Relationship between preoperative findings and
recommended that stereotactic radiosurgery be used as the postoperative improvement*
primary treatment in patients with small and moderate- Preop Clinical Findings No. of Cases W/ Postop Improvement (%)
sized trigeminal schwannomas, we reserve the use of radio-
surgery as adjuvant therapy for residual or recurrent tumors facial pain 8 of 11 (73)
that cannot be removed surgically, for patients unable to facial numbness 6 of 16 (38)
trigeminal motor deficit 4 of 5 (80)
undergo surgery and for patients who opt not to have sur- cerebellar signs 4 of 4 (100)
gery. We agree with others that the best treatment for tri- brainstem signs 1 of 1 (100)
geminal neurinomas is complete microsurgical removal of CN deficits
the lesion.13 II 1 of 1 (100)
IV 3 of 3 (100)
Cranial Nerve Function VI 6 of 8 (75)
VIII 3 of 4 (75)
Injury or permanent damage to the trigeminal branches IX 1 of 1 (100)
has been inevitable in many cases.49 Good health with tri- X 1 of 1 (100)
geminal nerve symptoms and other neurological deficits is * Three patients with prior surgically induced cranial nerve deficits were
considered to be a good outcome, even though injuries to excluded.

J. Neurosurg. / Volume 96 / March, 2002 459

 O. Al-Mefty, S. Ayoubi, and E. Gaber

TABLE 3
Surgical intervention and follow-up results*
Follow Up

Case Surgical Duration Tumor

No. Approach Postop Complication Clinical Findings (mos) Recurrence (mos)

1 ZMF weak masseter muscle, mild fron- preop V2 & V3 numbness remained, trigeminal pain improved; 60 none
talis muscle weakness,† & de- postop corneal reflex & frontalis & masseter muscles re-
creased corneal reflex covered
2 pet‡ weak frontalis muscle,† decreased preop facial pain persisted; postop numbness, corneal deficit, & 29 none
V2, & decreased corneal reflex frontalis muscle recovered
3 ZMF none preop V2 & V3 numbness & corneal deficit persisted, pain & 75 none
masseter weakness improved
4 COZ phenytoin toxicity preop CN VI palsy & pain improved 62 stable on MR image
5 ZMF none preop facial pain persisted 134 18, reop at 5 yrs,
(2) 2nd recurrence 5
yrs later
6§ pet none preop facial pain recovered 48 36, reop; 1 yr
(2) free from recurrence
7 pet V1 & V2 numbness, facial preop V1 & V2 numbness improved; postop mild to moderate 42 none
pain, weak CN IV pain persisted, CN IV recovered completely
8† ZMF V2 & V3 numbness preop pain & masseter muscle weakness persisted; postop V2 & 10 none
V3 numbness persisted
9 ZMF V2 & V3 numbness, facial preop pain recovered; postop headache & facial palsy recovered, 4 none
palsy, headache numbness largely recovered
10§ COZ‡ none preop V2 & V3 numbness, decreased corneal reflex, & masseter 60 none
muscle weakness persisted; CN VI improved; lower CNs improved
11 pter CN IV dysfunction preop V1–3 numbness & CN VI weakness persisted; postop CN IV 24 13, died 2 yrs later
dysfunction resolved of other disease
12 COZ none preop V2 & V3 numbness persisted, pain & masseter muscle 60 none
improved, CN VI palsy improved after 8 yrs of paralysis preop
13 pet none preop numbness & pain improved, but retroorbital pain persisted 15 none
14§ pet dysphasia, decreased postop dysphasia & corneal reflex fully recovered 20 none
corneal reflex
15 pet CSF rhinorrhea preop V2 & V3 numbness improved, pain & CN VI dysfunction 58 none
resolved, CSF leak stopped
16† ZMF none preop CN V & VI improved 24 none
17 ZMF mild frontalis muscle preop numbness & corneal reflex improved; postop frontalis 37 none
weakness† branches improved
18 ZMF none preop CN V & VI weakness & corneal reflex recovered 12 none
19§ ZMF none preop pain improved 5 none
20§ COZ & transient CN III paresis preop ophthalmoplegia recovered, but CN V deficit persisted 24 none
pet‡
21§ ZMF weak frontalis muscle,† V1 preop CN V numbness increased; postop frontalis branches 5 none
numbness, decreased corneal improved
reflex
22§ ZMF none preop CN IV improved, persistent CN VI pain improved 7 none
23 ZMF none preop CN V & cerebellar function improved 6 none
24 ZMF none normal 3 none
25§ ZMF none preop V1 & V2 numbness; postop V2 numbness persisted 4 none
* Mean duration of follow up was 33.12 months during which three (15%) of 25 patients experienced tumor recurrence. The outcome in each case was
good. Abbreviations: COZ = cranioorbital zygomatic; CSF = cerebrospinal fluid; pet = petrosal; pter = pterional; ZMF = zygomatic middle fossa.
† From skin flap.
‡ Underwent previous surgery elsewhere.
§ Cystic lesion.

trigeminal nerve. Most patients have only a slight deficit
in trigeminal function, which should remain the same13 or
even improve. Our findings indicate that nontraumatic mi-
crosurgical dissection of the remaining fibers of the trigem-
inal roots, ganglion, and divisions effectively preserves or
even improves trigeminal nerve function.
The clinical sequence of symptoms and cranial nerve de-
ficits is well described in the literature.5,12,13,15,18,24,31,35,38,39,42,45,
47,49,51,52
The tumors in our patients fit the usual pattern of
presentation. After surgery, preoperative facial numbness
improved in 44% of patients, facial pain was alleviated in
73%, and trigeminal motor deficit improved in 80% of pa-
tients. The preoperative deficits of all other involved cra-
nial nerves resolved by 4 to 6 months after surgery, except
for two of the eight cases of abducent nerve deficit. All cer-
ebellar and brainstem manifestations resolved postopera-
tively. Table 5 shows the effects of surgery on initial cranial
nerve symptoms in major reported series. Cranial nerve de-
ficits as a result of surgery and their clinical progress are
listed in Table 6.
Our experience in this small series confirmed the report-
ed value of neurophysiological monitoring in skull-base

460 J. Neurosurg. / Volume 96 / March, 2002

Trigeminal schwannomas

TABLE 4 nerve dysfunction was seen in 70% of patients after con-

Total removal of tumor reported in the literature ventional approaches. Yoshida and Kawase52 also found
that rates of total tumor removal and clinical outcomes
Temporal Fossa Posterior Fossa Dumbbell Shape were significantly better in patients treated via a skull base
No. of Total No. of Total No. of Total
approach than in those treated via conventional approaches.
Authors & Year Cases Removal Cases Removal Cases Removal Skull base approaches were used in all but one of our pa-
tients, and gross-total resection was achieved in all patients.
McCormick, et al., 4 3 3 2 7 4 There was no incidence of mortality or major morbidity.
1988 Dumbbell-shaped tumors that involved both middle and
Pollack, et al., 1989 2 2 4 4 6 2
Yasui, et al., 1989 2 2 2 2 4 4 posterior fossae (Type C in Jefferson’s classification20 and
Samii & Matthies, 5 5 1 0 5 4 Type 4 in the classification of Lesoin and colleagues26) are
1997 distinct in their clinical and surgical considerations. Some
Day & Fukushima, 15 14 9 6 9 7 investigators have suggested that total removal of a dumb-
1998* bell-shaped tumor is more difficult when the retrosigmoid
Yoshida & Kawase, 9 6 5 3 12 10 and subtemporal approach is used.38 In most reported cases,
1999
present study 6 6 0 0 19 19 however, the ability to remove these tumors totally was not
total 43 38 (88%) 24 17 (71%) 62 50 (81%) significantly reduced (Table 4).
Dumbbell-shaped tumors were dominant in our series
* Excluding malignant cases in series.
(19 cases), constituting 76% of all tumors. Tumors in all
patients involved the cavernous sinus. Intraorbital, intra-
maxillary, intrasphenoidal, and infratemporal fossa exten-
surgery.34 It assists in locating cranial nerves displaced by sions appeared in our series and did not preclude total re-
tumor before they are inadvertently damaged and continu- moval of the tumor.
ally confirms their function during surgery.16,33,41,50 Electric As the literature shows, patients who underwent extra-
stimulation is also a useful predictor of postoperative func- dural surgery did well, even as early as the 1950s.5,10,27 In
tion.33,50 the 1980s, surgeons mainly used an intradural approach to
these tumors.6,8,31,35,49 Lesoin and colleagues26 advocated ex-
tradural approaches only for small tumors. More recently,
Role of Skull Base Approaches
the extradural approach has been used for extracranial tu-
Recently, authors have used skull base approaches to mors with intracranial extension,38 peripheral and gasserian
treat patients with trigeminal schwannomas.12,13,42,52 In a lesions,12,13 and small posterior fossa tumors when an ante-
comparison of conventional and skull base approaches, Ta- rior transpetrosal approach is added.52
ha and colleagues42 found that skull base approaches al- This extradural tumor removal is readily achieved
lowed better exposure of these tumors, multiple working through the zygomatic middle fossa approach from the cav-
angles with minimal brain retraction, and more complete ernous sinus and the infratemporal fossa. We add a zygo-
removal without increased morbidity. They found residual matic osteotomy to minimize retraction of the temporal
or recurrent tumors in 65% of patients who had undergone lobe and allow better access to the anterior temporal fossa.
conventional approaches compared with 10% of those who A posterior fossa extension is removed through the expand-
had undergone skull base approaches. In addition, cranial ed Meckel cave. Only large caudal extensions cannot be

TABLE 5
Literature review of effects of surgery on preoperative symptoms of trigeminal schwannoma*
CN V
CN CN CN Overall CN CN Lower
No. Pain Sensory Motor Overall III IV VI Diplopia VII VIII CNs
of
Authors & Year Cases N I N I N I N I N I N I N I N I N I N I N I

McCormick, et al., 14 8 7 11 0 4 0 — — 1 — 0— — — 2 3 2 3 2 4 3 — —
1988
Bordi, et al., 1989 6 — — — — — — 5 0 — — — — — — — — — — — — — —
Pollack, et al., 1989 16 7 ? 11 ? 3 ? — — 1 ? 1 ? 5 ? 5 ? 5 ? 7 ? 2 ?
Yasui, et al., 1989 8 3 ? 7 0 1 ? — — — — — — — — — — — — — — — —
Dolenc, 1994 44 3 3 — — — — 23 11 — — — — 5 5 5 5 — — — — — —
Samii, et al., 1995 12 2 ? 6 0 — — — — 3 ? 4 ? 4 ? 6 ? 2 ? — — — —
Taha, et al., 1995 15 3 100% 14 19% 5 — — — 2 — 2 — 2 — — 67% 1 — 3 — — —
Day & Fukushima, 38 13 11 32 2 — — — — 2 — — — 5 — 7 4 — — — — — —
1998
Yoshida & Kawase, 27 3 ? 14 ? 3 ? 21 ? 3 ? 2 ? 4 ? — — 2 ? 3 ? — —
1999
present study 22† 11 8 16 7 5 4 — — 0 0 3 3 8 6 9 6 0 0 4 3 1 1
* I = number of patients with improvement; N = total number of patients; — = data not available.
† Three patients with prior surgically induced cranial nerve deficits were excluded.

J. Neurosurg. / Volume 96 / March, 2002 461

 O. Al-Mefty, S. Ayoubi, and E. Gaber

TABLE 6
Cranial nerve deficits after surgery for trigeminal schwannoma and their progress*
No. of Patients
CN V
W/ CN CN CN CN CN
Postop Sensory Motor Total III IV VI Diplopia VII VIII
Compli-
Authors & Year Total cations N P N P N P N P N P N P N P N P N P

Lesoin, et al., 1986 6 3 — — — — 1 1 — — — — — — — — 2 1 — —

McCormick, et al., 1988 14 12 — — — — 9 9 — — — — — — 3 1 3 1 4 1
Pollack, et al., 1989 16 14 7 — — 14 7 — — — — 4 1 4 1 3 1 2 0
Yasui, et al., 1989 8 8 7 7 — — 7 7 1 1 1 1 4 4 4 4 4 4 — —
Dolenc, 1994 44 — — — — 10 10 — — — — — — — — — — — —
Samii, et al., 1995 12 9 — — — — 8 8 — — — — — — — — 2 0 2 2
Taha, et al., 1995 15 70% 70% 56% 56% — — — — — — — — 63% 7% — — — —
Day & Fukushima, 1998 38 12 — — — — 11 11 — — — — — — 1 1 — — — —
Yoshida & Kawase, 1999 27 17 — — — — 10 10 3 2 — — 5 2 6 2 2 0 — —
present study 25 11 5 4 2 1 7 3 1 0 2 0 0 0 3 0 1 0 0 0
* N = total number of patients with postoperative cranial nerve dysfunction; P = number of patients with permanent deficits.

reached in this way. In these cases, the presigmoid petrosal
approach allows access to large, dumbbell-shaped tumors
and their posterior fossa components, which extend below
the level of the seventh and eighth cranial nerves.
Cystic Schwannomas
Magnetic resonance imaging is indispensable in the diag-
nosis of trigeminal schwannomas and in surgical planning
for patients harboring these tumors.7,25,30,36,43 It is particular-
ly accurate in demonstrating cystic schwannomas, which
occurred in 10 (40%) of our patients. In contrast to what has
been reported for acoustic neurinomas,9 there was no cor-
relation between the appearance of cystic lesions on MR
images and the totality of tumor removal or neurological
outcomes in our series. In patients with acoustic schwan-
nommas, a cystic component has been associated with a
higher risk of facial nerve injury9,37 and postoperative hem-
orrhage.37 Cystic acoustic schwannomas have been found
to be associated with larger size, a short clinical history,
atypical initial symptoms, and histologically higher nuclear
atypia.21 This was not the case in our trigeminal schwanno-
ma series.
Acknowledgments
The authors are grateful to Mr. Ron Tribell for the artwork and to
Mrs. Julie Yamamoto and Mrs. Amy Keeland for their editorial as-
sistance.
References
1. Al-Mefty O: Operative Atlas of Meningiomas. Philadelphia:
Lippincott-Raven, 1998, p 33
2. Al-Mefty O: Operative Atlas of Meningiomas. Philadelphia:
Lippincott-Raven, 1998, pp 318–321
3. Al-Mefty O, Ayoubi S, Smith RR: The petrosal approach: in-
dications, technique, and results. Acta Neurochir Suppl 53:
166–170, 1991
4. al-Mefty O, Smith RR: Tailoring the cranio-orbital approach.
Keio J Med 39:217–224, 1990
5. Arseni C, Dumitrescu L, Constantinescu A: Neurinomas of the
trigeminal nerve. Surg Neurol 4:497–503, 1975
6. Beck DW, Menezes AH: Lesions in Meckel’s cave: variable pre-
sentation and pathology. J Neurosurg 67:684–689, 1987
7. Beges C, Revel MP, Gaston A, et al: Trigeminal neuromas: as-
sessment of MRI and CT. Neuroradiology 34:179–183, 1992
8. Bordi L, Compton J, Symon L: Trigeminal neuroma. A report of
eleven cases. Surg Neurol 31:272–276, 1989
9. Charabi S, Tos M, Borgesen SE, et al: Cystic acoustic neuromas.
Results of translabyrinthine surgery. Arch Otolaryngol Head
Neck Surg 120:1333–1338, 1994
10. Cuneo HM, Rand CW: Tumors of the gasserian ganglion. Tu-
mor of the left gasserian ganglion associated with enlargement
of the mandibular nerve. A review of the literature and case re-
port. J Neurosurg 9:423–431, 1952
11. Cushing H, Eisenhardt L: Meningiomas: Their Classification,
Regional Behaviour, Life History, and Surgical End Results.
Springfield, IL: Charles C Thomas, 1938, p 223
12. Day JD, Fukushima T: The surgical management of trigeminal
neuromas. Neurosurgery 47:233–241, 1998
13. Dolenc VV: Frontotemporal epidural approach to trigeminal neu-
rinomas. Acta Neurochir 130:55–65, 1994
14. Eisenberg MB, Al-Mefty O, DeMonte F, et al: Benign non-
meningeal tumors of the cavernous sinus. Neurosurgery 44:
949–955, 1999
15. Findler G, Fiensod M, Sahar A: Trigeminal neurinoma with un-
usual presentation. Report of a case with trigeminal somatosen-
sory-evoked response. Surg Neurol 19:351–353, 1983
16. Harper CM, Daube JR: Facial nerve electromyography and oth-
er cranial nerve monitoring. J Clin Neurophysiol 15:206–216,
1998
17. Hedeman LS, Lewinsky BS, Lochridge GK, et al: Primary ma-
lignant schwannoma of the Gasserian ganglion. Report of two
cases. J Neurosurg 48:279–283, 1978
18. Hitchon PW, Graf CJ, McCabe BF, et al: Trigeminal neurilem-
moma presenting as a sphenoid sinus mass. Surg Neurol 11:
305–307, 1979
19. Huang CF, Kondziolka D, Flickinger JC, et al: Stereotactic
radiosurgery for trigeminal schwannomas. Neurosurgery 45:
11–16, 1999
20. Jefferson G: The trigeminal neurinomas with some remarks on
malignant invasion of the gasserian ganglion. Clin Neurosurg
1:11–54, 1955
21. Kameyama S, Tanaka R, Kawaguchi T, et al: Cystic acoustic
neurinomas: studies of 14 cases. Acta Neurochir 138:695–699,
1996
22. Kapila A, Chakeres DW, Blanco E: The Meckel cave: comput-
ed tomographic study. Part I: Normal anatomy; Part II: Pathol-
ogy. Radiology 152:425–433, 1984

462 J. Neurosurg. / Volume 96 / March, 2002

Trigeminal schwannomas
23. Kehrli, P, Maillot C, Wolff MJ: Anatomy and embryology of the
trigeminal nerve and its branches in the parasellar area. Neurol
Res 19:57–65, 1997
24. Kubota T, Hayashi M, Yamamoto S: Subarachnoid hemorrhage
due to trigeminal neurinoma. Surg Neurol 16:157–160, 1981
25. Lee BCP, Kneeland JB, Deck MDF, et al: Posterior fossa le-
sions: magnetic resonance imaging. Radiology 153:137–143,
1984
26. Lesoin R, Rousseaux M, Villette L, et al: Neurinomas of the tri-
geminal nerve. Acta Neurochir 82:118–122, 1986
27. Levinthal R, Benston JR: Detection of small trigeminal neurino-
mas. J Neurosurg 45:568–575, 1976
28. Levy WJ, Ansbacher L, Byer J, et al: Primary malignant nerve
sheath tumor of the gasserian ganglion: a report of two cases.
Neurosurgery 13:572–576, 1983
29. Lockhart RD, Hamilton GF, Fyfe FW: Anatomy of the Human
Body. Philadelphia: JB Lippincott, 1959, pp 453–454
30. Martini D, Har-El G, Johnson C: Trigeminal neurinoma. Ann
Otol Rhinol Laryngol 103:652–654, 1994
31. McCormick PC, Bello JA, Post KD: Trigeminal schwannoma.
Surgical series of 14 cases with review of the literature. J Neu-
rosurg 69:850–860, 1988
32. Meckel JF, Jourdan AJL, Doane AS: Manual of General, De-
scriptive, and Pathological Anatomy. Philadelphia: Carey &
Lea, 1832, Vol 3, pp 88–92
33. Moller AR: Neuromonitoring in operations in the skull base.
Keio J Med 40:151–159, 1991
34. Nuwer MR: Intraoperative monitoring, in Al-Mefty O (ed): Me-
ningiomas. New York: Raven Press, 1991, pp 309–319
35. Pollack IF, Sekhar LN, Jannetta PJ, et al: Neurilemomas of the
trigeminal nerve. J Neurosurg 70:737–745, 1989
36. Rigamonti D, Spetzler RF, Shetter A, et al: Magnetic resonance
imaging and trigeminal schwannoma. Surg Neurol 28:67–70,
1987
37. Samii M, Matthies C: Management of 1000 vestibular schwan-
nomas (acoustic neuromas): surgical management and results
with an emphasis on complications and how to avoid them.
Neurosurgery 40:11–23, 1997
38. Samii M, Migliori MM, Tatagiba M, et al: Surgical treatment of
trigeminal schwannomas. J Neurosurg 82:711–718, 1995
39. Shimabukuro H, Masuzawa T, Miyagi K, et al: Trigeminal neu-
rinoma revealed by intratumoral hemorrhage following a minor
head injury. Surg Neurol 19:346–350, 1983
40. Sindou M, Pelissou I: Trigeminal neurinomas. A special type of
cavernous sinus tumors, in Dolenc VV (ed): The Cavernous Si-
J. Neurosurg. / Volume 96 / March, 2002
nus. A Multidisciplinary Approach to Vascular and Tumor-
ous Lesions. Wien: Springer-Verlag, 1987, pp 355–376
41. Stechison MT: Neurophysiologic monitoring during cranial base
surgery. J Neurooncol 20:313–325, 1994
42. Taha JM, Tew JM Jr, van Loveren HR, et al: Comparison of con-
ventional and skull base surgical approaches for the excision of
trigeminal neurinomas. J Neurosurg 82:719–725, 1995
43. Tanaka A, Takaki T, Maruta Y: Neurinoma of the trigeminal
root presenting as atypical trigeminal neuralgia: diagnostic val-
ues of orbicularis oculi reflex and magnetic resonance imaging.
A case report. Neurosurgery 21:733–736, 1987
44. Tancioni F, Gaetani P, Villani L, et al: Neurinoma of the trigem-
inal root and atypical trigeminal neuralgia: case report and re-
view of the literature. Surg Neurol 44:36–42, 1995
45. Vaquero J, Cabezudo JM, Leunda G, et al: Simultaneous poste-
rior and middle cranial fossa neurinomas. Acta Neurochir 55:
321–327, 1981
46. Wascher TM, Spetzler RF, Zabramski JM: Improved transdural
exposure and temporary occlusion of the petrous internal carotid
artery for cavernous sinus surgery. Technical note. J Neurosurg
78:834–837, 1993
47. Yamashita J, Asato R, Handa H, et al: Abducens nerve palsy as
initial symptom of trigeminal schwannoma. J Neurol Neuro-
surg Psychiatry 40:1190–1197, 1977
48. Yasargil MG: Microneurosurgery. New York: Thieme, 1996,
Vol 1, pp 124–131
49. Yasui T, Hakuba A, Kim SH, et al: Trigeminal neurinomas: op-
erative approach in eight cases. J Neurosurg 71:506–511, 1989
50. Yingling CD: Intraoperative monitoring of cranial nerves in
skull base surgery, in Jackler RK, Brackmann DE (eds): Neu-
rotology. St Louis: Mosby, 1994, pp 967–1002
51. Yonas H, Jannetta PJ: Neurinoma of the trigeminal root and
atypical trigeminal neuralgia: their commonality. Neurosurgery
6:273–277, 1980
52. Yoshida K, Kawase T: Trigeminal neurinomas extending into
multiple fossae: surgical methods and review of the literature. J
Neurosurg 91:202–211, 1999
Manuscript received March 8, 2001.
Accepted in final form October 29, 2001.
Address reprint requests to: Ossama Al-Mefty, M.D., Depart-
ment of Neurosurgery, University of Arkansas for Medical Sci-
ences, 4301 West Markham Street, Slot 507, Little Rock, Arkansas
72205.
463