Monitoring of Motor and Somatosensory Evoked Potentials during Brain Surgery

Abstract

BACKGROUND: During brain surgery, intraoperative monitoring of somatosensory evoked potentials (SEPs) and motor evoked potentials (MEPs) is important for prediction of post-operative motor deterioration. The aim of this study is to identify which cut-off point of SEPs and MEPs changes are the most reliable for the prediction of postoperative motor deterioration after brain surgery. METHODS: By medical chart review among patients who underwent brain surgery between December 2015 and December 2016, 100 patients with intraoperative monitoring records of SEPs and MEPs were screened. Muscle strength was assessed by Medical Research Council (MRC) scale in all patients a day before surgery, within 48 hours postoperatively, and 4 weeks later. We analyzed sensitivity and specificity of prediction of motor deterioration using each patient’s changes in intraoperative SEPs or MEPs. To find the best cut-off point of SEPs and MEPs, receiver operating characteristic (ROC) curve analysis was conducted. RESULTS: In this study, non-tumor disease was 61(61%) and tumor disease was 39(39%). The sensitivity of pre-existing alarm criteria; latency delay more than 10% from baseline SEPs was 8.3% and amplitude reduction more than 50% from baseline MEPs was 58.3%, and the specificity for predicting motor deterioration was 97.7% for SEPs and 87.5% for MEPs, respectively. By ROC curve analysis, the maximally discriminating point for intraoperative SEPs latency change and intraoperative MEPs amplitude change were 7.1% and 59.5%, respectively. For clinical utility, 7.0% SEPs change and 60.0% MEPs change values were used for best-cut off point. With these cut-offs, the sensitivity of SEPs and MEPs was 66.7% and 58.3%, and the specificity was 80.7% and 90.9%, respectively. Significant intraoperative MEPs changes over the best cut-off value showed higher sensitivity than that of SEPs changes for postoperative motor deterioration. Sensitivity of patients with either SEPs or MEPs changes was 83.3% and specificity was 73.9%. A correlation test proved significant association between SEPs and MEPs changes (P=0.04). CONCLUSION: For brain tumor surgery, we suggest new alarm criteria which are cut-off value of 7% delay of SEPs latency or 60% reduction of MEPs amplitude. Further accumulation of cases would provide a more precise cut-off value to use as the alarm criteria.