Magnetic Resonance Appearance of the Inner Ear After Hearing-Preservation Surgery

Abstract

To determine whether the appearance of the inner ear on T2-weighted follow-up magnetic resonance imaging correlates with hearing status after hearing-preservation surgery for vestibular schwannoma. Retrospective chart review. Tertiary referral medical center. The study includes patients undergoing hearing-preservation surgery for vestibular schwannoma from 1998 to 2003. Diagnostic evaluation with magnetic resonance imaging and audiometric evaluation. Hearing results as reported in charts was correlated with appearance of membranous labyrinth on T2-weighted magnetic resonance images obtained at least 1 year after surgery. Twenty-nine patients were identified, 16 of whom satisfied the inclusion criteria. All 16 of the patients underwent middle fossa removal of vestibular schwannoma. Serviceable hearing according to American Academy of Otolaryngology-Head and Neck Surgery criteria was preserved in eight patients (50%). Of the eight patients without serviceable hearing, six had the cochlear nerve sacrificed at the time of surgery. All patients with serviceable hearing had normal appearing cochleovestibular signal on T2-weighted images, whereas six of eight patients (75%) with no hearing or poor hearing had abnormal low signal in the inner ear, suggesting inner ear ossification. The positive predictive value of a normal labyrinth for preserved hearing was 90%, whereas the negative predictive value of an abnormal labyrinth for no hearing was 100%. All but one patient who had the cochlear nerve sacrificed showed abnormal morphology of the labyrinth on T2-weighted magnetic resonance imaging. We describe the T2-weighted magnetic resonance findings after hearing-preservation surgery for acoustic tumor removal. Loss of inner ear signal on T2-weighted images correlates with loss of hearing postoperatively, whereas preserved inner ear signal correlates with hearing preservation after middle fossa surgery for vestibular schwannoma removal.