Preoperative Assessment of Cervical Vestibular Evoked Myogenic Potentials (cVEMPs) Help in Predicting Hearing Preservation After Removal of Vestibular Schwannomas Through a Middle Fossa Craniotomy

Prognostic Factors for Hearing Preservation Surgery in Small Vestibular Schwannoma

OBJECTIVE

to evaluate recent contributions to the literature on prognostic factors of hearing preservation in small vestibular schwannoma microsurgery.

METHODS

review of the most recent studies.

RESULTS

factors such as tumor size, preoperative hearing status, tumor growth rate, tumor origin, surgical approach, radiological characteristics, results of preoperative neurophysiological tests, preoperative symptoms and demographic features have been investigated and some of them reported to be significant in the prediction of hearing preservation.

CONCLUSIONS

tumor size and preoperative hearing status are the most impactful factors and play a key role in patient selection for hearing preservation surgery. Other features such as fundal extension, tumor origin and impaired ABR could have prognostic value on hearing preservation. Tumor growth rate, preoperative impedance, cVEMPs and age have also recently been found to be significant, but more studies are needed. The role of preoperative tinnitus, vertigo and gender is lacking and controversial, whereas the differences between available surgical approaches have been smoothed out in recent years.

Auditory Brain Stem Response Predictors of Hearing Outcomes after Middle Fossa Resection of Vestibular Schwannomas

Objectives To analyze the relationship between preoperative and intraoperative auditory brain stem response (ABR) characteristics and hearing outcomes in patients with vestibular schwannomas (VS) undergoing hearing preservation (HP) surgery via a middle cranial fossa (MCF) approach.

Design Prospective study.

Setting Academic tertiary skull base referral center.

Methods Pre- and postoperative pure-tone average (PTA) and word recognition score (WRS) were examined. Intraoperative ABR wave III latency, wave V latency, and amplitude were recorded. HP was defined as postoperative WRS ≥50%.

Participants Adult patients with VS and WRS ≥50% who underwent MCF tumor resection between November 2017 and September 2019.

Main Outcome Measures Postoperative hearing outcomes.

Results Sixty patients were included. Mean tumor size was 9.2 mm (range, 3–17). HP rates were 56.7% for the cohort and 69.7% for tumors <10 mm. A complete loss of wave V was associated with an 82.9% increase in postoperative PTA (p < 0.001) and 97.2% decrease in WRS (p < 0.001), whereas a diminished wave V was correlated with 62.7% increase in PTA (p < 0.001) and 55.7% decrease in WRS (p = 0.006). A diminished or absent wave V, but not increased wave III/V latency or decreased wave V amplitude, was correlated with a decline in postoperative hearing class (r = 0.735, p < 0.001). Receiver-operating characteristic analysis demonstrated that a stable wave V has the highest accuracy in predicting HP (sensitivity of 82.6%, specificity of 84.8%).

Conclusion Of the examined preoperative and intraoperative ABR characteristics, a stable wave V intraoperatively was the strongest predictor of HP after MCF resection of VS.

Level of Evidence Level III.

Machine learning approach for prediction of hearing preservation in vestibular schwannoma surgery

In vestibular schwannoma patients with functional hearing status, surgical resection while preserving the hearing is feasible. Hearing levels, tumor size, and location of the tumor have been known to be candidates of predictors. We used a machine learning approach to predict hearing outcomes in vestibular schwannoma patients who underwent hearing preservation surgery: middle cranial fossa, or retrosigmoid approach. After reviewing the medical records of 52 patients with a pathologically confirmed vestibular schwannoma, we included 50 patient’s records in the study. Hearing preservation was regarded as positive if the postoperative hearing was within serviceable hearing (50/50 rule). The categorical variable included the surgical approach, and the continuous variable covered audiometric and vestibular function tests, and the largest diameter of the tumor. Four different algorithms were lined up for comparison of accuracy: support vector machine(SVM), gradient boosting machine(GBM), deep neural network(DNN), and diffuse random forest(DRF). The average accuracy of predicting hearing preservation ranged from 62% (SVM) to 90% (DNN). The current study is the first to incorporate machine learning methodology into a prediction of successful hearing preservation surgery. Although a larger population may be needed for better generalization, this study could aid the surgeon’s decision to perform a hearing preservation approach for vestibular schwannoma surgery.