Postoperative hearing preservation in patients undergoing retrosigmoid craniotomy for resection of vestibular schwannomas: A meta-analysis of 1,249 patients
- Author(s): Ong, Vera
- Kwan, Isabelle
- Advisor(s): Yang, Isaac
- et al.
Objective To investigate the estimated hearing preservation rates in patients with sporadic VS (vestibular schwannomas) after RS (retrosigmoid) surgery through a meta-analysis based on the published hearing outcomes within the current literature.
Data Sources Both the PubMed and Cochrane databases were used to identify retrospective and prospective studies which involved hearing outcomes of VS patients who underwent the RS approach. The final search was performed on November 11, 2017. The first and third authors independently reviewed the aggregated articles, and final decisions of study inclusion were dependent on consensus.
Study Selection Of 313 articles, 16 studies (5.1%) met eligibility criteria. Together, these 16 studies comprised a total of 1,249 patients with serviceable preoperative hearing in the affected ear, from whom postoperative hearing preservation could be evaluated.
Data Extraction Both preoperative and postoperative hearing grades were extracted to determine the hearing preservation rate presented in the group of patients in each article. Additionally, demographics regarding gender, age, mean tumor size, and mean FU (follow-up) time pertaining to the patients in each article were extracted as well.
Data Synthesis Study heterogeneity was quantified through computation of t2,Q, and I2 statistics. A Wald-type Q statistic was used to assess statistical significance of study heterogeneity. Assessment of study bias was performed using standard funnel plot analysis and an Egger’s test for funnel plot asymmetry. Significant cross-study heterogeneity was found, with rates of hearing preservation ranging from 12-79% across studies. Aggregate hearing preservation was 23% under a fixed effects model and 37% allowing for random study effects. Clear systematic bias was also apparent, with disproportionate numbers of (mostly smaller) studies reporting hearing preservation rates markedly higher than the aggregate estimates (P< 0.0001). As expected, rates of hearing preservation were also strongly dependent on preoperative tumor size, with rates of 59%, 37%, and 11% observed for intracanalicular, small, and large (>20 mm) tumors, respectively.
Conclusions Hearing preservation rates are likely dependent on multiple factors and tumor size has a strong effect. It is critical to discuss the patient’s expectations for hearing preservation when deciding on treatment plans for VS.