Intraoperative Nerve Monitoring Parameters Predict Facial Nerve Outcome in Parotid Surgery

Presentation: AHNS-042
Topic: Salivary
Type: Oral
Date: Thursday, May 2, 2019
Session: 6:00 AM - 7:00 AM Scientific Session 5 - Value I
Authors: Catherine T Haring, MD, Andrew J Rosko, MD, Susan E Ellsperman, MD, Paul Kileny, PhD, Deborah Kovatch, MA, MBA, CCCA, Bruce Edwards, AuD, Matthew E Spector, MD
Institution(s): University of Michigan, Department of Otolaryngology- Head & Neck Surgery

Background: Facial nerve injury is the most dreaded complication of parotidectomy. Intraoperative nerve monitoring has gained widespread use to facilitate facial nerve identification and prevent injury.  Prior studies have demonstrated that the use of facial nerve monitoring decreases the rate of immediate post-operative facial nerve weakness, however there are no published data on normative values for these parameters or cutoff values to prognosticate facial nerve outcomes.

Objective: To identify intraoperative nerve monitoring parameters that predict postoperative weakness and to establish cutoff values for these parameters under which nerve function can be assured.

Methods: A retrospective case series of parotid surgery performed with intraoperative nerve monitoring by audiology was conducted. Patients who underwent primary parotid surgery for benign disease were included. Free running and evoked electromyography were utilized. Nerve monitoring parameters evaluated included nerve stimulation threshold, mechanical events and spasm events. Receiver operating curves were used to determine the accuracy of these parameters in predicting postoperative nerve outcomes (early and late) and to define optimal cutoff points to maximize the sensitivity and specificity of the chosen parameters in predicting outcomes.

Results: 223 patients were included. The rate of temporary facial nerve paresis was 45% and the rate of permanent paralysis was 1.3%. The mean pre-dissection threshold was 0.22 (range:0.1-0.6) and the mean post-dissection threshold was 0.24 (range: 0.08-1.0). The average number of mechanical events was 9 events (ranging from 0-66 events), and spasm events was 1 (ranging from 0-12). Both post-dissection threshold and the number of mechanical events predicted early post-operative facial nerve outcome (AUC 0.68, p<0001 and AUC 0.58, p=0.02, respectively), while the number of spasm events did not (Figure 1). The optimal cutoff value for threshold was 0.25 mA and for mechanical events was 8 events. If the threshold was greater than 0.25 mA, there was a 69% chance of post-operative weakness versus 35% if the threshold was less than or equal to 0.25 mA. If there were greater than 8 mechanical events, there was a 53% chance of post-operative weakness compared to 40% if there were 8 or fewer mechanical events. There were no parameters that predicted permanent facial nerve injury.

Conclusions: Facial nerve monitoring is a valuable adjunct to the knowledge of surgical anatomy and technique. Normative values of nerve monitoring parameters and optimal cutoff values provide surgeons with a baseline for interpretation of data and predict post-operative nerve function. Post dissection threshold and the number of mechanical events predict immediate post-operative facial nerve outcomes. Prediction of facial nerve function after parotid surgery is not only useful to provide anticipatory guidance to patients but may also provide surgeons with intraoperative feedback allowing them to adjust their operative techniques to improve outcomes.

Figure 1. Receiver Operating Curves and Optimal Cuffoff Values for Intraoperative Nerve Monitoring Parameters