Moisés A. Arriaga, Douglas A. Chen. Facial Function in Hearing Preservation Acoustic Neuroma Surgery. Arch Otolaryngol Head Neck Surg. 2001;127(5):543–546. doi:10.1001/archotol.127.5.543
To determine if facial function is worse after hearing preservation acoustic neuroma surgery (retrosigmoid and middle fossa) than in translabyrinthine surgery.
Retrospective medical record review.
Private neuro-otology subspecialty practice of patients operated on in a tertiary care hospital.
This study evaluated 315 consecutive acoustic neuroma surgical procedures between April 1989 and July 1998. A total of 209 translabyrinthine procedures and 106 hearing preservation surgical procedures were performed. The hearing preservation procedures were equally divided between retrosigmoid (n = 48) and middle fossa (n = 58) procedures.
Medical records were reviewed and tabulated for tumor size, surgical approach, and House-Brackmann facial function grade at short-, intermediate-, and long-term intervals.
Postoperative facial function in hearing preservation surgical procedures at short- and long-term follow-up was not worse than facial function after translabyrinthine surgical procedures in comparably sized tumors.
Concern about postoperative facial function should not be the deciding factor in selecting hearing preservation vs nonhearing preservation acoustic neuroma surgery.
THE OVERALL objectives of acoustic tumor (vestibular schwannoma) surgery are straightforward. The first objective is patient safety, that is, removing the life-threatening risks of an expanding mass in the posterior fossa. The second objective is functional preservation, that is, preserving facial function and, if possible, hearing. Before the microsurgical era, the risk of mortality from acoustic tumors was high. A number of factors have combined to improve the success of acoustic tumor surgery, including the routine use of the microscope, improved anesthetic techniques, intraoperative monitoring of facial function, better understanding of the perioperative physiologic stresses of posterior fossa surgery, and improved diagnostic techniques. Modern series of surgically managed acoustic neuromas routinely report mortality rates of approximately 1% and anatomic facial nerve preservation rates greater than 90%.1
Three principal surgical approaches are available for managing acoustic neuroma: middle fossa craniotomy (subtemporal), retrosigmoid craniotomy (suboccipital), and translabyrinthine craniotomy. The first 2 surgical approaches offer the possibility of hearing preservation since the otic capsule is not violated. Each of these approaches has particular strengths. With regard to hearing preservation, the middle cranial fossa offers complete exposure of the internal auditory canal with limited posterior fossa exposure. The retrosigmoid approach offers excellent exposure of the posterior fossa; however, exposure of the internal auditory canal, especially the lateral one third, is hampered by the posterior semicircular canal and vestibule. The translabyrinthine approach offers excellent posterior fossa and internal auditory canal exposure, but hearing is sacrificed with removal of the semicircular canals. The technical demands of hearing preservation acoustic neuroma surgery are greater than in translabyrinthine surgery, since the surgeon must focus on preservation of the cochlear nerve and its blood supply in addition to the usual considerations of safe tumor removal and facial nerve preservation.
Our center is convinced that each of these approaches has merit, and the appropriate approach for a particular patient is individualized based on the patient's medical status, the anatomic size and location of the tumor, and the patient's preferences. For most patients, facial function preservation is a higher priority than hearing preservation, since most of these patients have already become accustomed to a certain degree of hearing loss in the affected ear. Nonetheless, hearing preservation and facial function preservation are not mutually exclusive goals. As more patients are presenting with tumors at sizes amenable to hearing preservation surgery, the question of whether hearing preservation surgical approaches pose a greater risk than translabyrinthine surgery to facial function outcome after acoustic neuroma surgery has become a significant issue. The purpose of this study is to address the question, "Is facial function worse after hearing preservation surgery (retrosigmoid and middle fossa) than after translabyrinthine surgery?" The answer to this question provides guidance if concern over postoperative facial function should be a deciding factor in selecting hearing preservation approaches vs nonhearing preservation approaches in acoustic neuroma surgery.
This study retrospectively evaluated 315 consecutive acoustic neuroma surgical procedures performed by Pittsburgh Ear Associates surgeons (M.A.A. and D.A.C.) between April 1989 and July 1998. Facial function was graded according to the House-Brackmann scale.2 All facial nerve grading was performed by the authors. Facial nerve function was graded at 3 intervals after surgery: immediately after surgery (within 48 hours), at an intermediate time after surgery (at the 2-week postoperative visit), and in the long term (at least 3 months after surgery). Data were obtained by reviewing office records, hospitalization records, operative reports, and discharge summaries. Tumor measurements are the largest dimension in centimeters (including internal auditory canal and posterior fossa component).
Statistical comparisons of tumor sizes were made with the t test, and facial function results were evaluated with the χ2 test. The statement regarding no statistical difference between groups means that a statistical test is performed and failed to reject the null hypothesis (P>.05).
Facial function results are listed separately according to the 6 House-Brackmann grades and in a grouped fashion. The groups were excellent (grades I and II), acceptable (grades III and IV), and poor (grades V and VI).
In the total group at long-term follow-up, 80% demonstrated excellent function, 13% demonstrated acceptable function, and 7% had poor function. In this group of 315 patients, 58 were operated on through the middle fossa approach, 48 through the retrosigmoid approach, and 209 through the translabyrinthine approach. The mean ± SD size was 0.75 ± 0.31 cm for middle fossa tumors, 1.9 ± 1.09 cm for retrosigmoid tumors, and 2.06 ± 1.12 cm for translabyrinthine tumors. Middle fossa tumors were significantly smaller than retrosigmoid and translabyrinthine tumors.
Table 1 lists the facial outcome for these patients according to surgical approach and time after surgery. Table 1 lists all patients in the series regardless of tumor size, previous operation, anatomic status of the facial nerve at the conclusion of surgery, primary and revision surgery, and whether or not data were available at all 3 intervals. This last factor explains the different totals at the different intervals. These raw data are provided to facilitate comparison with other reports.
Overall, at long-term follow-up, 88% of middle fossa patients had excellent facial function compared with 91% of retrosigmoid patients and 77% of the translabyrinthine patients. Four patients experienced facial nerve transection: 1 middle fossa, 1 retrosigmoid, and 2 translabyrinthine. In contrast, 2% of middle fossa patients had poor long-term facial function outcome compared with 7% of retrosigmoid patients and 7% of translabyrinthine patients.
To adequately compare the effects of the operative approach, we chose to compare the facial function outcomes in patients with similarly sized tumors. We selected tumors 1.5 cm and smaller because this was the largest tumor excised through the middle fossa approach in our series. To fully assess the effects of the operative procedures, we excluded patients who had previously received treatment for their acoustic tumors (either surgery or radiation). Also, we excluded patients who did not present with preoperative grade I facial function. The facial function of cases that met the criteria for comparison are presented in Table 2. This group of 167 acoustic tumors measuring 1.5 cm or less were initially treated through the middle fossa (n = 57), retrosigmoid (n = 27), or translabyrinthine (n = 83) surgical approach as their initial management. In this series of 167 tumors measuring 1.5 cm or less, facial function was not significantly different between either hearing preservation approach and the translabyrinthine approach at the immediate, intermediate, or long-term time frame. Long-term facial function was excellent in 90% of translabyrinthine, 89% of middle fossa, and 100% of retrosigmoid cases. Viewed from the perspective of poor functional outcomes, only 1 patient (2% of middle fossa cases) had poor long-term facial function.
The intervals of these data permitted us to assess delayed facial palsy. We define delayed palsy as a deterioration of function from excellent or acceptable to poor from the immediate to intermediate interval. No patients in the retrosigmoid group experienced delayed palsy, whereas 5 patients (9%) in the middle fossa group had delayed palsy and 7 patients (8%) in the translabyrinthine group had delayed palsy. With the exception of 1 patient whose facial function went from grade II (immediate) to grade IV (intermediate) and remained at grade IV (long term), all patients with delayed palsy recovered to within 1 grade of their immediate postoperative function.
The surgical complication rate overall was low. Among the 167 patients in the small tumor group, 1 patient had meningitis and 7 had cerebrospinal fluid leakage, with none requiring surgical repair (3 in the translabyrinthine group [3.6%], 2 in the middle fossa group [3.5%], and 2 in the retrosigmoid group [7.4%]).
We have previously presented a detailed assessment of hearing preservation results using our criteria for patient selection.3 Our strategy individualizes the surgical approach to the patient's medical characteristics, tumor anatomy, and patient preferences. In general, we encourage hearing preservation in patients with a speech reception threshold of 30 dB or better and speech discrimination of 70% or better. However, we believe that a patient with a speech reception threshold of 50 dB and speech discrimination of 50% is an acceptable candidate. In certain circumstances, depending on patient preferences, we have offered hearing preservation to patients with even poorer hearing. We used the middle cranial fossa approach for tumors involving the internal auditory canal with extension of 0.5 cm or less into the cerebellopontine angle. In contrast, we use the retrosigmoid approach for patients with tumors with greater than a 0.5-cm cerebellopontine angle extension but not reaching the lateral third of the internal auditory canal. Using these restrictive criteria, we have published an overall useful hearing preservation rate of 67%, with 74% useful hearing preservation in middle fossa and 58% useful hearing preservation in retrosigmoid cases.2 Since that publication, the success rate for hearing preservation has improved with useful hearing preservation in 78% of middle fossa tumors and 64% of retrosigmoid tumors (M.A.A. and D.A.C., unpublished data). The increasing emphasis by referring physicians and patients on hearing preservation prompted us to review our facial function outcomes in acoustic neuroma surgery in a formal fashion.
In this series, the rate of anatomic preservation of the facial nerve was 99% overall, with 98% in middle fossa, 98% in retrosigmoid, and 99% in translabyrinthine cases. Modern series are consistently reporting high rates of anatomic preservation,4 but function is the principal concern. This study confirmed our previously identified findings regarding the chronology of facial function after acoustic neuroma surgery. Specifically, this study confirms that postoperative facial function after acoustic neuroma undergoes a temporary decrease in the weeks after surgery but then recovers during the ensuing months.5 In addition, the relationship of better postoperative facial function in smaller tumors was also confirmed. Since these facial function outcomes were consistent with previous statistical trends, the validity of the functional outcomes observed in the remainder of the study was enhanced.
Although the difference was not significant, translabyrinthine tumors at the immediate and intermediate intervals showed a trend toward better facial results than middle fossa tumors in our series of acoustic neuromas 1.5 cm or smaller. This is consistent with the generally accepted surgical impression that the facial nerve undergoes greater manipulation in middle fossa than translabyrinthine surgery for similarly sized tumors, since the facial nerve must be mobilized to permit removal of the adjacent acoustic neuroma. Nonetheless, the rate of delayed facial paralysis was similar in the middle fossa and translabyrinthine series (8% and 9%, respectively). The rate of excellent facial function in the retrosigmoid cases was 100% at each of the 3 intervals for the retrosigmoid tumors. This trend in the present study can be explained by our selection method for offering retrosigmoid surgery. This group of tumors represents lesions in which the tumor did not reach the fundus of the internal auditory canal. Thus, most of the dissection of tumor from the facial nerve was not performed within the confines of the bony limits of the internal auditory canal. It can be argued that this potentially lessens the trauma to the facial nerve. Despite these trends, these findings did not reach statistical significance.
Our results confirm the previous findings of no significant difference between facial function and surgical approaches in a series comparing only translabyrinthine and middle fossa surgery performed at another center.6 Overall, facial function outcome was not significantly different from either middle fossa or retrosigmoid vs translabyrinthine resection of acoustic neuromas 1.5 cm or smaller in this series.
Our series may be biased toward poorer results at the long-term interval, since we accepted 3-month data for this interval. We know from chronology studies that the facial function improvement continues for at least a year after the decrement in function seen at the postoperative visit. Despite this factor, the long-term facial function rates were excellent in 89% of middle fossa cases, 90% of translabyrinthine cases, and 100% of retrosigmoid cases with tumors 1.5 cm or smaller. Despite a slight trend toward better function in the translabyrinthine and retrosigmoid groups at the intermediate interval, the difference is not significant. Also, the delayed paralysis rate was the same for middle fossa and translabyrinthine cases.
Obviously, the decision of how to manage acoustic tumors is complex and involves many variables, including tumor size, patient age, patient health, and patient preferences. A patient with useful hearing and a small acoustic tumor should understand that hearing preservation surgery may expose them to a slight (not statistically significant) risk of temporary facial weakness in comparison with nonhearing preservation surgery. For many patients, the probability of successful hearing preservation outweighs this risk. We conclude that concerns about postoperative facial function should not be the only deciding issue for hearing preservation vs translabyrinthine surgery.
Accepted for publication September 22, 2000.
We thank Robert Rubin, MD, PhD, and Karen Berliner, PhD, for statistical consultation and Shirley Simonic for manuscript preparation.
Corresponding author and reprints: Moisés A. Arriaga, MD, Pittsburgh Ear Associates, 420 E North Ave, Suite 402, Pittsburgh, PA 15212.