Characteristics of multi-channel Br(E)-MsEP waveforms for the lower extremity muscles in thoracic spine surgery: comparison based on preoperative motor status.

Purpose: To evaluate the characteristics of brain-evoked muscle action potential [Br(E)-MsEP] waveforms of lower limb muscles in thoracic spine surgery.
Methods: The subjects were 159 patients who underwent thoracic spine surgery with intraoperative Br(E)-MsEP monitoring from January 2009 to December 2015, using a total of 2226 muscles in the extremities. The waveform derivation rate for each lower extremity muscle was examined at baseline and intraoperatively. Data were interpreted based on the preoperative motor status.
Results: The preoperative ambulatory and non-ambulatory rates were 38% (60/159, McCormick grades I and II) and 62% (99/159, grades III-V), respectively. Eleven cases (all non-ambulatory) had undetectable baseline waveforms in all muscles, and in 19 cases (12%) a baseline waveform could only be derived from the abductor hallucis (AH). The waveform derivation rate in all lower limb muscles was significantly higher in ambulatory cases (p < 0.05), and the rates for the AH were the highest in both groups (p < 0.05). Postoperative paralysis occurred in 31 cases (19%). A decrease in intraoperative amplitude of ≥ 70% from baseline occurred in 54 cases and had sensitivity of 100% and specificity of 82% for prediction of postoperative motor deficit.
Conclusions: This is the first study of Br(E)-MsEP waveforms for each lower limb muscle based on preoperative ambulatory status. Detection of waveforms from distal muscles was still possible in a case with preoperative motor deficit, and the AH had an especially high derivation rate, even in cases with preoperative muscle weakness. Collectively, the results support use of Br(E)-MsEP monitoring using the AH in the lower extremities.