Aaron Mohanty, M.Ch., Sastry Kolluri Venkatrama, M.S., Basrur Ravimohan Rao, M.Ch., Bangalore A. Chandramouli, M.Ch., Peruvumba N. Jayakumar, M.D., Bhabani Shankar Das, M.Ch.
Departments of Neurosurgery (AM, SKV, BRR, BAC, BSD) and Neuroradiology (PNJ), National Institute of Mental Health and Neurosciences, Bangalore, India
OBJECTIVE: Cerebellopontine angle (CPA) epidermoids, although of benign nature, are of considerable neurosurgical interest because of their close proximity and adherence to the cranial nerves and the brain stem. We describe our experience and attempt to correlate the final outcomes with the extent of surgical removal.
METHODS: Twenty-five consecutive patients with CPA epidermoids that were surgically treated were reviewed, and the final outcomes were assessed.
RESULTS: Thirteen patients had trigeminal neuralgia. In 7 of the 13 patients, trigeminal neuralgia was the only presenting feature. The epidermoid was confined to the CPA in each of 6 patients, and in each of 18, it had varying degrees of supratentorial extension. One patient had a predominant supratentorial epidermoid with extension to the CPA. The lesions were totally excised in 12 patients. Near-total removal was accomplished in eight patients, and in the remaining five, partial removal was accomplished. Transient worsening of the cranial nerve functions occurred in 11 patients, probably as a result of aggressive dissection of the capsule from the cranial nerves. In 9 of the 11 patients, the cranial nerve functions improved by the time of discharge. All of the patients who had trigeminal neuralgia were relieved of their symptoms. Eighteen patients were followed up for a mean period of 42 months, and none had symptoms of recurrence.
CONCLUSION: Aggressive surgical removal results in transient but significant cranial nerve dysfunction in the postoperative period. A conservative approach is indicated for patients in whom the capsule is adherent to the brain stem and the cranial nerves.
(Neurosurgery 40:24-30, 1997)
Key words: Cerebellopontine angle, Cysts, Epidermoid cyst, Outcome
Epidermoids represent 0.2 to 1% of all primary intracranial tumors (28). These slow-growing benign tumors commonly occur in the parapontine or parapituitary regions. Uncommonly, their location can also be purely intracerebral, at the brain stem, or intraventricular. Cerebellopontine angle (CPA) epidermoids constitute 40% of all intracranial epidermoids (24) and are of considerable surgical interest. Their close proximity to the exquisitively sensitive cranial nerves and the brain stem and their tendency to adhere to them pose technical difficulties in their complete removal. Although total removal is considered to be ideal (35), it is often associated with significant morbidity in the postoperative period. On the other hand, partial removal risks recurrence and subsequent surgery (2). We summarize our experience with 25 patients with CPA epidermoids during a 9-year period and attempt to correlate the final outcomes with the extent of surgical removal.
Twenty-five patients were operated on for CPA epidermoids at the National Institute of Mental Health and Neurosciences, Bangalore, India, during a 9-year period (1986-1994). The age group varied from 18 to 42 years, with the majority of the patients being between 20 and 40 years of age (Table 1). No definite sex predilection was noted.
Thirteen patients presented with ipsilateral trigeminal neuralgia (Table 2). It was the only presenting feature in seven patients. In another, it was associated with hemifacial spasm (painful tic convulsif). In three patients, it was atypical in nature. Two patients presented with seizures. The majority of the patients had varying degrees of cranial nerve dysfunction. Six patients had papilledema.
TABLE 1. Cerebellopontine Angle Epidermoids: Age and Sex Distribution | |
Distribution | No. of Patients |
Age (yr) | |
<20 | 1 |
21-30 | 12 |
31-40 | 10 |
>41 | 2 |
Sex | |
Male | 15 |
Female | 10 |
Total | 25 |
TABLE 2. Cerebellopontine Angle Epidermoids: Clinical Features | |
Feature | No. of Patients |
| |
Trigeminal neuralgia | 13 |
Painful tic convulsif | 1 |
Vth sensory impairment | 17 |
Vth motor impairment | 13 |
Facial paresis | 13 |
Hearing impairment | 12 |
Lower cranial nerve dysfunction | 4 |
Cerebellar signs | 16 |
Pyramidal signs | 5 |
Papilledema | 6 |
Seizures | 2 |
Total | 25 |
TABLE 3. Cerebellopontine Angle Epidermoids: Radiology | ||
Procedure | Finding | No. of Patients |
Cranial x-ray | Raised intracranial tension | 1 |
(n = 25) | Destruction of petrous apex | 1 |
Computed | Nature | |
tomography | Hypodense nonenhancing | 22 |
(n = 25) | Hypodense ring-enhancing | 1 |
Hypodense with calcification | 2 | |
Location | ||
Confined to cerebellopontine angle | 6 | |
Extension to supratentorial compartment | 18 | |
Predominantly supratentorial with extension to cerebellopontine angle | 1 | |
Extension anterior to brain stem | 9 | |
Hydrocephalus | 13 | |
Mild | 6 | |
Moderate | 5 | |
Severe | 2 | |
Magnetic resonance imaging (n = 2) | Hypointense in T1-weighted and hyperintense in T2-weighted images | 2 |
The duration of symptoms ranged from 3 months to 14 years, with a mean of 35 months. Patients who presented with only trigeminal neuralgia had a mean duration of symptoms of 29.4 months, as compared to 38.4 months for those who did not have trigeminal neuralgia.
The radiological examination (Table 3) included cranial radiography, computed tomography (CT), and magnetic resonance imaging (MRI). As revealed by a routine cranial radiograph, one patient had features of raised intracranial tension. Another had destruction of the petrous apex, suggesting chronic local pressure effect. CT of the brain with intravenous contrast was performed for all patients. In each of 22 patients, the lesion was hypodense and nonenhancing, and in another, there was evidence of peripheral enhancement. Two others had hypodense nonenhancing lesions with peripheral specks of calcification. The epidermoid was confined to the CPA in each if six patients. In each of another 18 patients, it had varying degrees of extension to the supratentorial compartment through the tentorial hiatus (Fig. 1). One of the 18 patients had a significant supratentorial extension. In another, the lesion was predominantly supratentorial but had significant CPA extension. Extension anterior to the brain stem and deformation of the brain stem were present in nine patients. Thirteen patients showed evidence of obstructive hydrocephalus, which was mild in six, moderate in five, and severe in the remaining two. MRI was performed for two patients, who were treated during the later part of the study period. The evidence was hypointense on T1-weighted images and hyperintense on proton density and T2-weighted sequences (Fig. 2). In the proton density sequence, the epidermoid had a different signal intensity from that of cerebrospinal fluid (CSF), thus suggesting the diagnosis. As expected, MRI demonstrated the extent of the tumor and deformation of the brain stem better than did CT for these two patients.
Four patients underwent ventriculoperitoneal shunting preoperatively. Primarily, 24 patients underwent retromastoid craniectomy. For another patient, who had a predominant supratentorial mass, a subtemporal approach was used (Table 4). At the initial surgery, microscopic total excision was performed in 12 patients, near-total removal in 8, and partial removal in 5. The excision was considered to be total when the lesion was completely removed, near-total when parts of the capsule were left behind, and partial when more than 20% of the tumor was left behind. One patient underwent a reexploration for the residual epidermoid, which could then be near-totally removed. Of the 18 patients who had supratentorial extensions, complete removal of the parts extending through the hiatus, using the posterior fossa approach, was achieved in 15 (Fig. 1A). In another two, partial removal was achieved. Another patient who had a significant supratentorial extension required a second subtemporal approach at a later date (Fig. 1B). Because of dense adhesions of the capsule to the adjacent brain stem and the cranial nerves, a conservative approach was used and the capsule was partially removed in 11 patients. Inadvertent injury to the VIIth cranial nerve occurred in one patient and inadvertent injury to the VIIIth cranial nerve occurred in one patient when attempts were made to dissect the capsules from the nerves, aiming at total excision.
TABLE 4. Cerebellopontine Angle Epidermoids: Surgical Approach and Extent of Removal | ||
No. of Patients | ||
Surgical approach | ||
Retromastoid | 23 | |
Subtemporal | 1 | |
Combination | 1 | |
Extent of removal | ||
Total | 12 | |
Near total | 8a | |
Partial | 5 | |
a One patient underwent subsequent
surgery for the residual epidermoid, which could be near-totally removed. |
Eleven patients experienced deterioration of their cranial nerve functions in the postoperative period. For 2 of the 11 patients, in whom the nerves were injured at surgery, the deterioration was permanent. For the remaining nine, the deterioration was transient and improved by the time of discharge (Table 5). Three of the nine patients experienced dysfunction of two cranial nerves each, and another experienced multiple cranial nerve dysfunction. Two patients worsened in cerebellar functions. Two others had chemical meningitis, which was successfully managed with steroids. One patient required a ventriculoperitoneal shunt in the postoperative period.
Two patients died during the hospital stay. One of the two patients, who had significant supratentorial extension, underwent multiple procedures for the residual epidermoid and developed pneumonic consolidation and succumbed to it. The other patient died of an unrelated cause in the postoperative period.
All of the patients who had trigeminal neuralgia and hemifacial spasm were relieved of their symptoms. The two patients with postoperative worsening of cerebellar signs experienced improvement by the time of discharge. Cranial nerve paresis improved by the time of discharge in all except the two patients in whom it occurred preoperatively. Of the 23 patients who were discharged from the hospital, 19 experienced improved neurological statuses, 2 remained unchanged, and 2 experienced persistent cranial nerve paresis (Table 6).
Eighteen patients were followed up for periods ranging from 1 to 8 years, with a mean duration of 42 months. None experienced recurrence of their symptoms. All the patients resumed their previous occupations.
TABLE 5. Cerebellopontine Angle Epidermoids: Postoperative Deficits (n = 25)a | |||
Total Removal (n = 12) | Near-total Removal (n = 8) | Partial Removal (n = 5) | |
Cranial nerve dysfunction | |||
Vth (n = 1) | 1 | - | - |
VIIth (n = 4) | 2 (1) | 1 | 1 |
VIIIth (n = 2) | 2 (1) | - | - |
VIIth and VIIIth (n = 3) | 2 | 1 | - |
VIIth, VIIIth, IXth, and Xth (n = 1) | 1 | - | - |
Total (n = 11) | 8 (2) | 2 | 1 |
Cerebellar dysfunction (n = 2) | 1 | 1 | - |
Chemical meningitis (n = 2) | - | 1 | 1 |
a Numbers in parentheses indicate permanent paresis. -, not applicable. |
TABLE 6. Cerebellopontine Angle Epidermoids: Outcome and Follow-up | |
No. of Patients | |
Outcome at discharge (n = 23) | |
Better than preoperative state | 19 |
Same as preoperative state | 2 |
Persistent cranial nerve deficits | 2 |
Follow-up (n = 18; mean duration, 42 mo; range, 1-8 yr) | |
Persistent cranial nerve paresis | 2 |
Able to resume to previous occupation | 18 |
Postoperative elective CT was performed for 13 patients, either at the time of discharge or during the follow-up examination. Eight of the 13 patients had undergone total excision of the tumor, 4 had undergone near-total removal, and 1 had undergone subtotal removal. The hypodense areas and the deformity of the brain stem, although less than in the preoperative state, persisted at the site of the surgery, as revealed by CT, even when the tumor was completely or near-totally removed. For one patient, who had undergone a total excision, the hypodense area persisted even 5 years after surgery (Fig. 3, A and B); MRI did not reveal any tumor reccurence but indicated CSF filling (Fig. 3, C and D).
Epidermoids are uncommon and represent 0.2 to 1% of all primary intracranial tumors (28). They are benign extracerebral intradural lesions and are commonly located in the CPA and sellar, parasellar, and intraventricular regions. Uncommonly, they may be located purely intracerebrally or in the region of the pineal gland (8, 18, 31). They can extend in both supratentorial and infratentorial compartments, either along the basilar artery or assuming an hourglass configuration in the paramedian position (1). Of all the epidermoids, 40% occur in the CPA, making it the most common site. The incidence of epidermoids among all the CPA tumors is approximately 5% (24).
Epidermoids have an extremely slow linear growth rate (3). Hence, the duration of symptoms is often prolonged and the patient presents late in the course of the illness. However, as depicted in this study, patients often present relatively early when the predominant symptom is trigeminal neuralgia, probably because of the associated pain.
The symptoms and signs are caused by displacement of the adjacent neural and vascular structures. Trigeminal neuralgia was seen in 80% (10 of 13 patients) and hemifacial spasms in 8% (1 of 12 patients) of the patients with posterior fossa epidermoids reported by Revilla (27). Painful tic convulsion (hemifacial spasm with ipsilateral trigeminal neuralgia) occurs rarely (33) and was present in one of our patients. Direct compression of the nerve at the root entry zone (9, 16, 22, 23), displacement of the trigeminal nerve and compression against a blood vessel (6, 17, 26) at the root entry zone, or a combination of the two (4) have been postulated to be the cause.
Destruction of the petrous apex can occur because of slow compression of the hourglass epidermoid and was seen in one of our patients. High resolution CT and MRI are currently the investigation modalities of choice and are considered by some to be complimentary to each other (29). As revealed by CT, epidermoids are characteristically hypodense nonenhancing lesions (11); however, they occasionally are hyperdense on plain computed tomographic scans (15). Enhancement with contrast is uncommon and often indicates a malignant transformation (13, 19). Enhancement can also occur because of perilesional inflammatory granulation as a result of leakage of the irritative cyst fluid. In the present study, contrast ring enhancement was present in one patient. However, a histological examination did not reveal any evidence of malignant change. Direct coronal sections at the level of the tentorial hiatus can demonstrate the extension into the tentorial incisura and helps in the presurgical planning and approach.
MRI reveals epidermoid tumors to have long T1 and T2 relaxation times, possibly because of the presence of keratin in a solid crystalline state (12). The CSF present in the interstices of the tumor can also cause long T1 and T2 relaxation times (12). The proton density sequences can differentiate between the CSF and the tumor and, hence, can differentiate epidermoid from arachnoid cyst (25, 32).
Hydrocephalus is said to be uncommon because of the long-standing nature of the lesion and also because CSF can permeate through the crevices of the lesion. However, in the present study, 50% of the patients had ventriculomegaly and four of them required preoperative CSF diversion procedures. Another patient required such a procedure in the postoperative period because of a CSF leak. In a previously reported series, 12 of 30 patients with CPA epidermoids required shunt surgery for hydrocephalus (10).
The surgical approach is generally determined by the location and the extent of the lesion. The lesion, when confined to the CPA, is approached by a retromastoid craniectomy, whereas significant supratentorial extension needs a combined retromastoid and subtemporal approach or a staged procedure. However, the tumor commonly extends into the hiatus and this incisural part can be completely removed with the posterior fossa approach (4, 35). In 15 of our 18 patients who had supratentorial extension, the lesion could be removed completely through the retromastoid approach, whereas in another two, it could be partially removed. In the patient with significant supratentorial extension, a second surgery by a subtemporal route was required.
There is a controversy regarding the extent of removal. Although the aim of surgery is for complete removal, few authors advocate total removal of the tumor (2, 20, 35). It has been suggested that with microscopic meticulous sharp dissection, every bit of the capsule should be removed to prevent a recurrence (20). However, adherence of the capsule to the important neurovascular structures in and around the brain stem often leads to its incomplete removal (5). A conservative approach with decompression and the removal of the nonadherent portion of the capsule has been suggested by others (6, 10, 21, 30, 32). Coagulation of the residual capsule is also advocated to minimize recurrence (10) but is not widely practiced. It also may be dangerous to coagulate the remnant adherent capsule near the exquisitely sensitive cranial nerves, brain stem, or important vessels in the CPA. The occurrence of the high incidence of postoperative cranial nerve dysfunction (40%) in the present series is probably because of our initial aggressive approach in removing every bit of the capsule from the cranial nerves. A relatively high incidence of cranial nerve dysfunctions has been mentioned in the previously reported series, the majority of which, however, improve by the time of discharge or during the follow-up period (6, 29, 32, 35). In all of our patients, except those in whom nerve function was damaged at surgery, the cranial nerve functions improved. We think the risks and benefits of total removal should be considered on an individual basis, and when the capsule is densely adherent, it may be left behind because the lesion is benign and has an extremely slow growth rate.
Chemical meningitis caused by spillage of the cyst contents occurs in the postoperative period, which usually is transient and self-limiting (35). It occurred in two patients and could be managed successfully with steroids. Excision of the capsule by sharp dissection (14), irrigation of the CPA cisterns with hydrocortisone solution during the surgery (6), and delayed withdrawl of steroids in the postoperative period (7, 20, 30) have been advocated as possible measures for preventing chemical meningitis. The same procedure was practiced in all patients in the present series. Communicating hydrocephalus can develop as a sequelae to meningitis and might require CSF diversion procedures (6, 29, 34).
Of the two patients who died during the postoperative period, one death was unrelated to the primary disease or the surgical procedure. The other was caused by aspiration pneumonia secondary to the lower cranial nerve paresis while an attempt was made to excise the adherent capsule. Hence, partial removal is suggested for patients with adherent capsules to minimize the morbidity and mortality.
Partial removal of a lesion leads to recurrence, which often occurs after a prolonged period as a result of the slow growth rate of the tumor (2, 21). As the hypodense areas revealed by CT persist for a prolonged period, even after complete removal of the tumor, possibly because of a long-standing deformation of the neural structures, a diagnosis of recurrence at an early stage is often not possible. MRI is useful to diagnose an early recurrence (20, 25, 29). However, subsequent surgery is often indicated only when the recurrent lesion is symptomatic (29).
Received, February 14, 1996.
Accepted, July 31, 1996.
Reprint requests: Aaron Mohanty, M.Ch., Department of Neurosurgery, National Institute of Mental Health and Neurosciences, Hosur Road, Bangalore 560 029, India.
The authors present a large series of cerebellopontine angle (CPA) epidermoid cysts that were surgically treated within a 9-year period in Bangalore, India. The authors retrospectively review the clinical presentation, surgical management, and results for 25 patients. Although this report does not add any new data, this series represents an important contribution to the world literature. The authors' conclusions are in many aspects similar to those we have recently published (1).
CPA epidermoid cysts may achieve large extensions into the middle fossa and to the foramen magnum. In our series of 40 CPA epidermoids, the most frequent symptoms were hearing loss and vertigo and then trigeminal problems. In the present series, no information is provided regarding VIIIth cranial nerve affection but trigeminal problems were present in half of the patients. We agree with the authors that most CPA epidermoids with supratentorial extension can be totally resected by a simple retrosigmoid approach, although some cysts with extreme extension into the middle fossa may need a combined subtemporal-retrosigmoid approach.
In the majority of cases, the cranial nerve deficits resolve during the postoperative period. Although most epidermoid cysts can be totally resected with good postoperative outcomes, there still are cases in which total resection is impossible without resulting in severe neurological deficits. Severe injury of the lower cranial nerves, particularly in cases of bilateral cysts, may result in life-threatening aspiration pneumonia. Long follow-up radiological studies are necessary to depict cyst recurrence or regrowth.
Marcos Tatagiba
Hannover, Germany
The authors provide a broad experience in the treatment and management of epidermoids of the CPA. Because these are slow-growing tumors, the follow-up period of 42 months is short to ensure permanent cure. However, my personal experience indicates that as the authors suggest, an absolute total removal of the capsule and interstices of an epidermoid tumor is not necessary to ensure long-term cure. I have had patients who have gone 25 years after radical but subtotal removal of the capsule material without any evidence of recurrence. Therefore, the best attempt at removal of the contents as well as the capsule should be performed but not to the point of sacrificing vital structures, such as cranial nerves and blood vessels of the posterior fossa. Vigorous attempts to remove every last scrap have resulted in unwarranted cranial nerve deficit as well as injury to arteries. The authors present, overall, a large experience with excellent results in these benign lesions.
New York, New York
Mohanty et al. present their clinical experience with CPA epidermoids in 25 patients. This is a relatively large series of CPA epidermoids, and the authors further review the pertinent literature. The authors conclude that they prefer a somewhat conservative approach with subtotal removal for these CPA epidermoid tumors, which is in accordance with the majority of the previous literature. With the advent of the keyhole microneurosurgical technique and precise sharp dissection, it is possible to completely resect this tumor with negligible morbidity.
Through my experience with more than 120 cases of this type of tumor, I have obtained extremely satisfactory results with radical total resection, and I think that this benign process has to be resected in total fashion. The follow-up period of 3 to 4 years is not enough for this benign process. The epidermoid tumor will recur for 10 to 15 years.
Pittsburgh, Pennsylvania