Transcranial Motor Evoked Potential Alarm Criteria to Predict Foot Drop Injury During Lumbosacral Surgery.

STUDY DESIGN

A retrospective cohort analysis.

OBJECTIVE

This study aims to investigate whether waveform alterations in transcranial motor evoked potentials (TCMEPs) can reliably predict postoperative foot drop.

SUMMARY OF BACKGROUND DATA

Nerve injury leading to foot drop is a potential complication of lumbosacral surgery. Very limited data exist on the use of intraoperative TCMEPs to identify iatrogenic foot drop.

METHODS

We retrospectively reviewed neuromonitoring data from 130 consecutive spine surgeries with instrumentation involving L4-S1. TCMEP waveform analysis included amplitude (A), area under the curve (AUC), latency (L), and duration (D). Patient outcomes were correlated with neuromonitoring results. Intraoperative alert criteria were established on the basis of observed intraoperative changes.

RESULTS

Three patients developed severe foot drop with a muscle weakness functional grade ranging from 0/5 to 3/5. Two patients developed a mild foot drop with functional grade 4/5. Twenty-three patients had preoperative weakness in an L5 distribution. One-hundred two patients who had neither preoperative nor postoperative neurological complications served as a control group. Amplitude significantly decreased in patients with a severe postoperative deficit (P = 0.005) as did AUC and duration (P < 0.05). Intraoperative alert criteria defined as a >65% decrease in AUC resulted in a sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 100%, 91.4%, 12%, and 100%, respectively. When defining an alert criteria as a >50% decrease in amplitude, sensitivity, specificity, PPV, and NPV were 100%, 87.9%, 8.8%, and 100%, respectively.

CONCLUSION

Reduction of TCMEP waveform associated with postoperative severe foot drop can be detected during lumbar surgery. Other waveform parameters such as AUC may predict foot drop better than the amplitude. Additional examinations in larger samples of foot drops are needed to validate these alert threshold findings.

LEVEL OF EVIDENCE