ANALYSES OF NEUROOTOLOGICAL COMPLICATIONS AFTER RADIOSURGERY FOR ACOUSTIC NEURINOMAS

Purpose: To find out the optimum treatment parameters and the proper i ndications for treatment of acoustic neurinomas, univariate and multiv ariate actuarial analyses of neuro-otological complications after ster eotactic radiosurgery for acoustic neurinomas were performed. Methods and Materials: The subjects were 46 patients with acoustic neurinomas who underwent unilateral radiosurgery between June 1990 and June 1994 and were followed up at the University of Tokyo. Age ranged from 13 to 77 years (median, 54 years). Tumor diameter ranged from 0 to 25 mm (m ean, 12 mm) at the cerebellopontine angle and from 2 to 15 mm (mean, 8 .3 mm) in the internal auditory meatus. Maximum tumor doses ranged fro m 20 to 40 Gy (mean, 31.4 Gy), and peripheral doses from 12 to 25 Gy ( mean, 16.8 GS) One to eight isocenters were used (mean, 3.2). Median f ollow-up was 39 months. Eight events concerning neurootological compli cations were chosen, and the potential risk factors for them were anal yzed by the actuarial analyses (univariate and multivariate). The even ts examined include hearing loss, vestibular function loss, facial pal sy, and trigeminal nerve dysfunction. In order to point out potential risk factors for neuro-otological complications, univariate analyses w ere performed using both the Wilcoxon test and the log rank test, and multivariate analyses were performed with the Cox proportional hazards model. Variables nominated as potential risk factors were 1) demograp hic variables such as patient age and sex, 2) tumor dimensions, 3) tre atment variables such as tumor doses and number of isocenters, and 4) pretreatment hearing levels. A variable with significant p-values (p < 0.05) in two or more of the three actuarial analyses (two univariate and one multivariate) was considered a possible risk factor. Results: The possible variables that increase the risk for each event analyzed were: neurofibromatosis type II (NF2) and the number of isocenters for total hearing loss; experience of prior operation, the tumor diameter in the internal auditory meatus, and NF2 for hearing threshold elevat ion; peripheral tumor dose for vestibular function loss; patient age o r midporus transverse tumor diameter (the two variables were correlate d), and the number of isocenters for facial palsy; and the number of i socenters for trigeminal neuropathy. Conclusion: NF2 and the tumor dia meter were the common risk factors for hearing loss in previous studie s and ours. For the 5th/7th nerve dysfunction, the tumor diameter was the common risk factor. The risk of using more isocenters remains cont roversial. The difference in risk factors for hearing impairment and v estibular function loss suggests different mechanisms for the two. Fur ther studies with larger populations and longer follow-up periods are required in order to draw conclusions on the risk factors in radiosurg ery. (C) 1997 Elsevier Science Inc.