Cerebrospinal Fluid Leak following Vestibular Schwannoma Surgery—Factors that Impact Outcomes

Cerebrospinal fluid leaks are associated with significant morbidity following acoustic neuroma surgery. The incidence of cerebrospinal fluid leak varies greatly in the literature and the best method for prevention is still debated. Our objective in this study was to evaluate the incidence of cerebrospinal fluid leak following vestibular schwannoma surgery for retrosigmoid approaches, translabyrinthine approach with mesh cranioplasty, or translabyrinthine approach with water-tight periosteum closure. Our secondary objective was to determine the association of nonsurgical variables with higher rates of cerebrospinal fluid leak following acoustic neuroma surgery. This is a retrospective chart review of 219 patients who underwent sporadic vestibular schwannoma resection via a translabyrinthine or retrosigmoid approach between 2000 and 2019. These surgeries were performed at a tertiary academic referral center. Patient variables that were analyzed included patient’s age at time of surgery, sex, body mass index, and the largest dimension of the tumor based on preoperative magnetic resonance imaging. Surgical variables that were analyzed included length of surgery, from incision to the end of closure, and surgical approach as well as method of closure in the translabyrinthine approach group (watertight periosteal closure vs. titanium mesh plate closure). Postsurgical variables that were analyzed included total number of hospital days (including total hospital days during the initial stay as well as hospital days during any re-admission within 30 days), number of days spent in the intensive care unit, cerebrospinal fluid leak, the site of the cerebrospinal fluid leak, the need for a second operation and pseudomeningocele occurrence. Our overall cerebrospinal fluid leak rate was 12.3% with a leak rate of 17.2% in our retrosigmoid group, 12.8% in our translabyrinthine titanium mesh closure group, and 0% in our translabyrinthine periosteal closure. A significant difference was found in the cerebrospinal fluid leak rate between the two translabyrinthine groups (p = 0.019) and between the watertight periosteal closure translabyrinthine and retrosigmoid groups (p = 0.005). There was no difference in cerebrospinal fluid leak rate between the titanium mesh translabyrinthine and retrosigmoid groups (p = 0.413). There was no statistically significant effect of age, body mass index, and size of tumor on the incidence of cerebrospinal fluid leak in these postsurgical patients. There was also no statistically significantly difference between the three groups on length of operative time or number of days spent in the ICU. Total hospital days demonstrated a trend parallel to cerebrospinal fluid leak rates, with the shortest amount of hospital days in the watertight periosteal translabyrinthine group (average of 5.2), 6.4 days in the mesh translabyrinthine group, and 7.4 days in the mesh retrosigmoid group; however, this did not reach statistical significance (p = 0.119). Watertight periosteal closure of a translabyrinthine approach is an effective means of closing the temporal bone defect that allows for lower cerebrospinal fluid leak rates in comparison to closure with titanium mesh.