Intracanalicular acoustic neuroma: early surgery for preservation of hearing STEPHEN J. HAINES, M.D., AND SAMUEL C. LEVINE, M.D.

Transcription

1 J Neurosnrg 79:55-50, 99 Intracanalicular acoustic neuroma: early surgery for preservation of hearing STEPHEN J. HAINES, M.D., AND SAMUEL C. LEVINE, M.D. Center for Craniofacial and Skull Base Surgery, Departments of Neurosurgery and Otolaryngology/ Head and Neck Surgery, University of Minnesota, Minneapolis, Minnesota ~" Diagnosis of very small acoustic neuromas has become much more common since the advent of magnetic resonance imaging. Many of the patients so diagnosed have minimal unilateral hearing loss as their only symptom. Because limited information is available on the natural history and prognosis of these lesions, the choice of treatment is controversial. The authors review their recent experience with the surgical treatment of inlracanalicular acoustic neuroma. The records of 4 consecutive patients with inlracanalicular acoustic neuroma were reviewed with respect to type of presentation, pre- and postoperative facial and auditory nerve function, surgical approach, and complications. Detailed results for patients operated on to preserve hearing are presented. Presenting symptoms were nearly equally divided among diminished hearing, vertigo, and tinnitus. Eleven of the 4 patients had serviceable hearing preoperatively and nine (8%) remained in this condition postoperatively. Facial nerve function was unchanged by operation in patients. Seven operations were performed through the middle fossa, five through the posterior fossa, and two by the translabyrinthine approach. The probability of preserving hearing during surgical excision of intracanalicular acoustic neuroma in patients with serviceable hearing exceeds 80%. Given the relative infrequency of serious complications and the likelihood of progressive hearing loss in the untreated patient, excision of such small tumors shortly after diagnosis may offer the best chance of long-term hearing preservation. KEY WORDS acoustic neuroma 9 hearing preservation 9 timing of surgery G ADOLINIUM (Gd)-enhanced magnetic resonance (MR) imaging has revolutionized the diagnosis of acoustic neuroma. We have previously reported our finding that this technique is the most sensitive test for acoustic neuroma) ~ While intracanalicular tumors could previously be identified by air-contrast computerized tomography (CT) cisternography, the ability of Gd-enhanced MR imaging to identify heretofore invisible tumors restricted to the internal auditory canal has increased the frequency with which such tumors are diagnosed. We report our experience with intracanalicular acoustic neuromas since Gd-enhanced MR imaging has been available. Clinical Material and Methods The diagnostic studies were reviewed of all 9 patients with acoustic neuroma identified after Gd-enhanced MR imaging became available at our institution. The patients in whom the tumors were confined to the internal auditory canal with no measurable extension into the posterior fossa and who were operated on with the intention of preserving heating are included in this report. Salient characteristics of the patients including age at diagnosis, gender, presenting symptoms, side of tumor, size of tumor in the long axis of the internal auditory canal, pre- and postoperative speech reception threshold (SRT) (average of values obtained at 500, 000, and 000 Hz), and pre- and postoperative word recognition scores were extracted from the medical records. All patients with serviceable hearing (Gardner-Robertson Grades I or II ) preoperatively underwent operations with the intention of preserving their hearing. Intraoperative monitoring of heating function was accomplished by the technique of transtympanic electrocochleography. Facial nerve function was monitored by intraoperative evoked electromyography and classified according to the House and Brackmann grading system. 6 Results During the period of the study (986 to 99) 9 patients were operated on for acoustic neuroma; in 4 (%) of these the tumors were intracanalicular. The J. Neurosurg. / Volume 79/October, 99 55

2 S. J. Haines and S. C, Levine patients ranged in age from 6 to 70 years (mean 48.4 years): eight were women and six were men. The tumors ranged from 4 to 0 mm in length and were equally divided between the right and left side. Seven were operated on through the middle fossa, five through the posterior fossa, and two via the translabyrinthine route. The translabyrinthine approach was used only for patients with recurrent tumor. These two patients are, therefore, excluded from the remainder of this report. The presenting symptoms in the remaining patients are shown in Table. All patients were symptomatic, with either hearing loss, vertigo, or tinnitus as the first symptom. Eleven of the patients had serviceable hearing preoperatively. One patient had a 5-dB SRT, but only 0% word recognition score. The results with respect to hearing preservation are shown in Tables and. Average results for patients undergoing the middle fossa approach include data for the six patients with serviceable preoperative hearing as well as for all seven in this group; the six showed no significant change in hearing, and three actually showed minor improvement. In the group with a posterior fossa approach, one patient had unmeasurable hearing postoperatively. As this represents infinite hearing loss, it is not possible to include this patient in the averages. Therefore the average results in this group include only the four patients with TABLE Presenting symptoms in patients with preoperative hearing Presenting Middle Posterior Total Symptom Fossa Route Fossa Route Cases hearing loss 5 vertigo tinnitus 4 total cases 7 5 TABLE Hearing preservation in seven patients with middle fossa operations Speech Word Recog- Case Reception nition (%) No. Threshold (db) Preop Postop Preop Postop NT* case average? , 6-case median? case average~ case median~ * NT = not tested. "? Includes data for six patients with serviceable hearing. {: Includes data for all seven patients. measurable postoperative hearing, but the median values include all five patients. Facial nerve function is shown in Table 4. All had normal facial function (Grade I) preoperatively. One patient in the middle fossa approach group had House- Brackman Grade IV facial nerve function (moderately severe dysfunction) postoperatively and one in the posterior fossa approach group had Grade III function (moderate dysfunction), All others had normal function after surgery. Complications other than hearing loss are shown in Table 5. Discussion Acoustic neuroma confined to the internal auditory canal is not a life-threatening disease. The rationale for treatment is the likelihood of hearing loss and the eventual growth of the lesion to a size that makes treatment more difficult and dangerous, and may eventually become a threat to neurological function or to the patient's life. Therefore, to be justified, any treatment for this disease must have better results than the natural history of the tumor with respect to both life and hearing preservation. Mortality There were no deaths in this series. Samii, et al., 5 also reported no deaths in their series, and we have found no report of death related to surgical treatment of purely intracanalicular acoustic neuroma. While TABLE Hearing preservation in five patients with posterior fossa operations* Case No. Speech Reception Word Recog- Threshold (db) nition (%) Preop Postop Preop Postop NT 88 NT a-case average case median * Average results include only patients with measurable postoperative hearing. NT = not tested. TABLE 4 Facial nerve junction* Operative Preop Data Postop Data Approach Function No. of Cases Function No. of Cases middle fossa I 7 I 6 IV posterior fossa 5 I 4 IIl * Facial nerve function graded according to House and Brackmann. ~ 56 J. Neurosurg. / Volume 79 / October, 99

3 Hearing preservation in acoustic neuroma surgery death remains a theoretical complication of any operation, it must be an extraordinary event in the surgical treatment of such small tumors and therefore should not be a major consideration in decisions regarding treatment of such lesions. Hearing Preservation Since Walsh t9 first suggested the possibility in a report on McKissock's patients, the preservation of hearing during surgery for the removal of acoustic neuromas has been an ideal achieved relatively rarely. Through the 980's, the average success rate in operations carried out with the intention of preserving hearing was about %, but this is highly dependent on the definition of serviceable hearing used and the size of tumor being removed. ~ For purposes of comparing and predicting results, the classification of hearing proposed by Gardner and Robertson has been most helpful. Both SRT and word TABLE 5 Complications other than hearing loss resulting from operation Complication Middle Fossa Posterior Fossa Route Route CSF leak/meningitis* I f 0 seizure persistent vertigo 0 0 pulsatile tinnitus 0 * CSF = cerebrospinal fluid. f These complications affected the same patient. recognition are considered, and the classification used is that indicated by the worst of these two measurements. We define a patient with preserved hearing as one who had "serviceable" (that is Gardner-Robertson Grade I or II) hearing both pre- and postoperatively. In this series of tumors restricted to the internal auditory canal, we have achieved a hearing-preservation rate of 8% (nine of cases). This compares favorably to the % to 50% hearing-preservation rates reported for series incorporating tumors of all sizes mentioned above, and suggests that one important predictor of successful hearing preservation is the size of tumor. Based on the more stringent requirement of a word recognition score of 70%, in our series six of the seven patients with a middle fossa approach and four of the five with a posterior fossa approach met this criterion preoperatively. All six patients undergoing the middle fossa approach (the seventh patient in this group had no useful hearing preoperatively) and three of the four posterior fossa approach patients met the criterion postoperatively (90% preservation of useful hearing). In 99, Samii, et al., ~5 reported a series of surgically treated intracanalicular tumors with a hearing preservation rate of 57%. However, they used a different classification of hearing function, based solely on SRT. Table 6 compares our results to theirs, using their hearing classification. Table 7 makes the same comparison using the Gardner-Robertson classification. Because they did not provide word recognition data for all patients in their report, some of their patients have been classified using only SRT. This will overestimate their classification because adding the word recognition score can only lower the final grade. These comparative TABLE 6 Comparison of hearing-preservation results using criteria of Samii, et al. * Preop Haines & Levine Series: Postop Results Samii, el al., Series: Postop Results Grade Good Fair Bad Deaf Total Good Fair Bad Deaf Total good fair bad total * Comparison of postoperative hearing based on the hearing classification used by Samii, etal. ~5 Preop Grade TABLE 7 Comparison of hearing-preservation results using present criteria* Haines & Levine Series: Postop Grade Samii, et al., Series: Postop Grade I Ill IV V Total I II IlI IV V Total I ! 4 III IV V total I * Comparison of postoperative hearing between the present series and that of Samii, et at.] 5 based on the Gardner-Robertson classification system. J. Neurosurg. / Volume 79/October, 99 57

4 S. J. Haines and S. C. Levine TABLE 8 Follow-up evaluation of nonsurgical cases of intracanalicular acoustic neuroma with intact hearing* Author & Year Thomson & Tos, 988 Zrllner & Bockenheimer, 985 Luetje, el al., 988 Gardner & Robertson, 988 Case Growth Initial Hearing Final Hearing Rate Follow-Up No. (mm/yr) SRT (db) WR (%) SRT (db) WR (%) Period (yrs) Surgery no no "serviceable" no change ,5 0,8 0.8 no no AL NA 0.5 no RP 0,5 95 NR no change. no MP no change.8 no RG NA 0.8 no VH no PJ no * SRT = speech reception threshold: WR = word recognition; NA = not available; NR = not recorded. results strongly suggest that the better the preoperative hearing, the better is the chance that it will be preserved postoperatively. With regard to the surgical approach (Tables and ), we lost serviceable hearing in one patient in each of the operative approach groups. Hearing also improved in one patient in each group (Table 6). However, there was a better chance of maintaining the preoperative hearing level (or actually improving it slightly) when the middle fossa approach was used. We attribute this to the fact that the cochlear portion of the nerve is more remote and therefore protected from inadvertent manipulation when the internal auditory canal is approached from above rather than from behind. This series is too small to draw a definitive conclusion in this regard, but we continue to prefer the middle fossa approach for superior vestibular nerve tumors because of this tendency. Facial Nerve Function Only two of our patients experienced permanent postoperative deterioration in facial nerve function, one with House-Brackmann Grade III and one with Grade IV function. One facial nerve complication occurred early in our experience and represents a technical error. The second was in a patient showing bulbous enlargement of the internal auditory canal without enlargement of the medial end (porus acustieus). This radiographic finding was believed to be associated with unusual adherence of the tumor to the facial nerve (a "sticky" tumor) and to confer an increased risk of facial paresis. In fact, of the patients with normal preoperative facial function (House-Brackmann Grade I), eight (6%) had normal function immediately postoperatively and three (%) had Grade II function (mild dysfunction). We consider that this indicates a very low risk of facial dysfunction for such small tumors in the hands of an experienced surgeon. Other Complications The first patient in whom we performed a middle fossa operation experienced an early postoperative generalized seizure. Since that time we have administered phenytoin prophylactically in a loading dose just prior to surgery and continued the drug for 4 to 6 weeks. No additional seizures have been seen. One case of cerebrospinal fluid (CSF) leak was treated with a lumbar drain, but a nursing error in the management of the drain led to meningitis. No other patients have experienced CSF leakage. Pulsatile tinnitus is present in two patients with early treatment in the middle fossa group. This resulted from exposure of the ossicles to the temporal dura, and has been prevented in recent cases by placement of Gelfilm* over the bone defect in the middle fossa floor prior to closure. Natural History of Acoustic Neuroma The little information available on the natural history of acoustic neuroma was reviewed in 986 by Gardner, et al' The fact that small acoustic neuromas are found in nearly % of autopsies of apparently asymptomatic patients suggests that many tumors never reach a size sufficient to produce symptoms prompting evaluation and disclosure of their presence. 5'9''6 This statement must be interpreted with care, however. Intracanalicular tumors of the size most often found at autopsy may produce only relatively mild symptoms of decreased hearing, tinnitus, and vertigo. Such symptoms can easily be attributable to the normal aging process. Before the advent of Gd-enhanced MR imaging, such tumors could be demonstrated only by invasive tests and even the symptomatic patient might not be evaluated until ness or other neurological symptoms occurred. Fur- * Gelfilm manufactured by Upjohn Pharmaceuticals, Kalamazoo, Michigan. 58 J. Neurosurg. / Volume 79 / October, 99

5 Hearing preservation in acoustic neuroma surgery thermore, there was little reason to consider treatment of such small tumors before techniques allowing a high chance of hearing preservation were developed. As the condition is increasingly easy to diagnose, and the value of early treatment has been shown, the natural history of these tumors should be reconsidered with particular regard to the prognosis for hearing. Attempts have been made to estimate the growth rate of acoustic neuromas, with estimates ranging from 0 to 6 cm per year?.7-8~ There are many technical problems with such estimates. Some are based on posterior fossa myelography or CT scanning, techniques unable to detect small changes in size with precision. The choice of which dimension, dimensions, or volume estimates of the tumor to use for measurement, the problem of obtaining comparable images over many years of follow-up evaluation, and the possibility that a tumor may remain dormant for some time and then grow rapidly are all factors that need to be taken into account. Lesser, et at., ~~ measured the Ki-67 fraction of eight acoustic neuromas and concluded that this fraction may correlate with growth rate (although radiographically determined growth rates were available for only four patients). In their review of published acoustic neuroma growth rates, they concluded that it was possible to group acoustic tumors into three categories by growth rate: no growth, progression approximately 0. cm/yr, or progression approximately cm/yr. Unfortunately, no factor measurable before operation can predict the subsequent growth rate of the tumors. Indeed, there is no evidence that the growth rate of any single tumor is constant over time. Valvassori and Shannon ~8 found that 50% of tumors not treated surgically grew in a variable follow-up period. The prognosis for hearing once an intracanalicular acoustic neuroma is discovered in a patient with serviceable hearing is not well known. Early diagnosis of these tumors was very difficult before Gd-enhanced MR imaging became available and since then many patients with tumors discovered by that means have been operated on without long follow-up evaluation; therefore, no large series of unoperated intracanalicular acoustic tumors in patients with serviceable hearing has been reported. From the literature, we have identified 5 patients with intracanalicular tumor for whom there are sufficient radiological and audiometric follow-up data to make an assessment of both growth rate and heating prognosis (Table 8). Of these 5 patients, eight had serviceable hearing at the time of diagnosis, seven of whom underwent audiometric follow-up evaluation. Two experienced no change in their hearing in a -year follow-up period, One patient, with Mtni~re's disease in the opposite ear, actually showed no change in the SRT and a 0% increase in word recognition in a - year follow-up period. Another patient had a stable SRT but showed a 0% decrease in word recognition over 8 months. Three patients progressed to nonserviceable hearing over a follow-up period ranging from to t5 years. Four of the seven evaluahle patients, therefore, experienced progressive hearing loss. It is also noted that the follow-up period of those whose hearing remained unchanged is relatively short ( years). In addition, Nedzelski, et al.," identified eight elderly patients with serviceable hearing whose tumors were followed without surgery and whose SRT deteriorated at a mean rate of 6 db/yr. Unfortunately, insufficient detail was supplied regarding the actual audiometry and tumor size to include these patients in Table 8. The growth rates of the 5 tumors in Table 8 are also of interest. Seven tumors showed no evidence of growth in follow-up periods ranging from months to 5 years, seven showed growth rates of 0. to 0 mm/yr during observation periods of up to years. One tumor actually regressed at a rate of approximately.5 mm/yr over a 4-year follow-up period. It is interesting to note that growth rate (measured by relatively crude radiological techniques) did not correlate with hearing prognosis. Two patients with serviceable hearing that progressed to ness showed no measurable growth in their tumors; one patient whose hearing remained stable over a -year period showed t cm of growth in that same interval. In summary, the very limited available information suggests that about 50% of patients with serviceable hearing at the time of diagnosis of an intracanalicular acoustic tumor will show progressive heating loss over the next years. Coupled with the fact that the majority of patients in whom such a diagnosis is made are evaluated because asymmetrical hearing loss has been noted, this strongly suggests that intracanalicular acoustic neuroma causes progressive hearing loss in the majority of affected patients. As the best predictor of the quality of postoperative hearing is the quality of the patient's preoperative hearing function, hearing will be best preserved when intracanalicular acoustic neuromas are removed shortly after diagnosis. As more and more of these tumors are discovered at evaluation for other reasons, it becomes increasingly important to clarify these issues in the natural history of acoustic neuroma. Problems of Diagnosis An aggressive search for intracanalicular tumors will occasionally uncover other lesions that may mimic acoustic neuroma, These include certain aneurysms of the anterior inferior cerebellar artery, Ramsey-Hunt syndrome, 4 presumed viral neuritis, 4 arteriovenous malformation of the eighth cranial nerve, 4 and small facial neuromas. 4 While we have not experienced a falsepositive Gd-enhanced MR study, we understand that surgical exploration of patients with false-positive imaging has occurred. Obviously, this should be avoided, and we have adopted a policy of repeating a study on the internal auditory canal 6 to weeks after the initial diagnosis. Persistent enhancing lesions that are otherwise consistent with acoustic neuroma are removed. Conclusions Although the hearing prognosis associated with an intracanalicular acoustic neuroma is not well defined, existing information suggests that more than one-half of such patients will experience inexorably progressive heating loss. Current techniques allow preservation of hearing at or near preoperative levels in a high percent- J, Neurosurg / Volume 79/October, ]99 59

6 S. J. Haines and S. C. Levine age of such patients. While careful studies of the natural history of untreated intracanalicular acoustic neuroma and the recurrence rate of treated tumors are needed, the best opportunity to preserve hearing in these patients appears to be with surgery soon after the diagnosis is made. References. Gardner G, Moretz WH Jr, Robertson JH, et al: Nonsurgical management of small and intracanalicular acoustic tumors. Otolargyngol Head Neck Surg 94:8-, 986. Gardner G, Robertson JH: Hearing preservation in unilateral acoustic neuroma surgery. Ann Otol Rhinol Laryngo97:55-66, 988. Gleeson M J, Cox TC, Strong A J: Aneurysm of the anterior inferior cerebellar artery mimicking an intracanalieular acoustic neuroma. J Laryngol Otol 0:07-0, Han MH, Jabour BA, Andrews JC, el al: Nonneoplastic enhancing lesions mimicking intracanalicular acoustic neuroma on gadolinium-enhanced MR images. Radiology 79: , Hardy M, Crowe S J: Early asymptomatic acoustic tumor. Report of six cases. Arch Surg :9-0, House JW, Brackmann DE: Facial nerve grading system. Otolaryngol Head Neck Surg 9:46-47, Kasantikul V, Netsky MG, Glasscock ME Ill, et al: Acoustic neurilemmoma: clinicoanatomical study of 0 patients. J Neurosurg 5:8-5, Kasantikul V, Netsky MG, Glasscock ME Ill, el al: lntracanalicular neurilemmomas clinicopathological study. Ann Otol 89:9-, Leonard JR, Talbot ML: Asymptomatic acoustic neurilemoma. Arch Otolaryngol 9:7-4, Lesser THJ, Janzer RC, Fisch U: Clinical growth rate of acoustic schwannomas: correlation with the growth fraction as defined by the monoclonal antibody Ki-67. Skull Base Surg : - 5, 99. Levine S, Antonelli PJ, Le CT, et al: Relative value of diagnostic tests for small acoustic neuromas. Am J Otol :4-46, 99. Luetje CM, Whittaker CK, Davidson KC, et al: Spontaneous acoustic tumor involution: a case report. Otolaryngol Head Neck Surg 98:95-97, 988 [. Nedzelski JM, Canter RJ, Kassel EE, et al: Is no treatment good treatment in the management of acoustic neuromas m the elderly? Laryngoscope 96:85-89, Parnes LS, Lee DH, Peerless SJ: Magnetic resonance imaging of facial nerve neuromas. Laryngoscope 0: -5, I99 5. Samii M, Matthies C, Tatagiba M: Intracanalicular acoustic neurinomas. Neurosurgery 9:89-99, Schuknecht HF: Pathology of vestibular schwannoma (acoustic neurinoma), in Silverstein H, Norrell H (eds): Neurological Surgery of the Ear. Birmingham, Ala: Aesculapius, 977, pp Thomsen J, Tos M: Acoustic neuromas. Diagnostic delay, growth rate and possible non-surgical treatment. Aeta Otolaryngol Suppl 45:6-, Valvassori GE, Shannon M: Natural history of acoustic neuromas. Skull Base Surg :65-67, Walsh L: The surgery of acoustic neuromas. Proc R Soc Med 58:07-076, Wazen J, Silverstein H, Norrell H, et al: Preoperative and postoperative growth rates in acoustic neuromas documented with CT scanning. Otolaryngol Head Neck Surg 9:5-55, 985. Whittaker CK, Luetje CM: Vestibular schwannomas. J Neurosurg 76: , 99 (Editorial). Winzenburg SM, Margolis RH, Levine SC, et al: Tympanic and transtympanic electrocochleography in acoustic neuroma and vestibular nerve section surgery. Am J Otol 4:6-69, 99. Z6ner C, Bockenheimer S: The growth rote of acoustic neuromas: a report of three cases. Arch Otorhinolaryngol 4:59-64, 985 Manuscript received November 8, 99. Accepted in final form April 8, 99. Address reprint requests to. Stephen J. Haines, M.D., Department of Neurosurgery, 40 Delaware Street S.E., Box 96 UMHC, Minneapolis, Minnesota J. Neurosurg. / Volume 79 / October. 99