Surgery for sporadic vestibular schwannoma. Part IV. Predictive factors influencing facial nerve function after surgery

Abstract

Objective

To analyze the impact of various clinical, radiological and perioperative factors that could influence the facial nerve intraoperative disruption risk (CNVII-IDR) and its long-term function (CNVII-LTF) after vestibular schwannoma (VS) surgery.

Material and methods

The study included 212 patients operated on for sporadic VS with no history of previous treatment for VS or CNVII palsy. The mean size of the tumor was 30mm. Gross (210) or near-total (2) resections were carried out using the retrosigmoid (210) or translabyrinthine (2) approach. Correlation studies and multivariate regression analysis (RA) were performed.

Results

In correlation studies, the CNVII-IDR was increased by: headaches and cerebellar ataxia if one of them was the first symptom of the tumor (33% and 29%, respectively, p=0.008); preoperative hydrocephalus (40% vs. 9%, p=0.01), tumor size >3cm (18% vs. 5%, p<0.01), tumor volume >10cm3 (19% vs. 4%, p<0.01), right-sided location 15% vs. 6%, p=0.047), lateral “park-bench” position (19% vs. 5% for supine position, p<0.01) and the procedure order (16% for the first 106 procedures vs. 6% for the last 106 procedures, p<0.05). In RA the tumor volume (p=0.012), side of the tumor (p=0.028) and patient's position during surgery (p=0.016) independently affected the CNVII-IDR. The following factors correlated significantly with satisfactory CNVII-LTF (HB grades I–III): tumor stage <T4 (p=0.000), tumor size ≤3cm (p=0.000), tumor volume ≤10cm3 (p=0.001), and left-sided location (p=0.048). Additional factors correlated significantly with very good CNVII-LTF (HB grades I–II): anterior CNVII displacement (p=0.044), nimodipine use (p=0.016), the absence of postoperative complications (p=0.019), CNVII responsive on final intraoperative EMG stimulation (p=0.000) and supine position during surgery (p=0.018). However, an independent impact on very good CNVII-LTF proved to be the tumor size (p=0.0000), side (p=0.0175), and nimodipine use (p=0.0349).

Conclusions

In our series, the factors related to size and side of the tumor confirmed an independent impact on CNVII-IDR and CNVIILTF. The significance of patient positioning may reflect the impact of learning curve as only the first 91 patients were operated on using the lateral “park-bench” position. An independent impact on CNVII-LTF was exerted by the perioperative use of nimodipine.