Relative Efficacy of Transcranial Motor Evoked Potentials, Mechanically-Elicited Electromyography, and Evoked EMG to Assess Nerve Root Function During Sustained Retraction in a Porcine Model

Study Design. This is an animal experiment using transcranial motor evoked potentials (TcMEP), mechanically elicited electromyography (EMG), and evoked EMG during spinal nerve root retraction in a pig model. Objective. To compare the sensitivity of these 3 electro-physiological measures for a constant retraction force applied to an isolated lumbar nerve root for a specific duration of time. Summary of Background Data. The incidence of nerve root injury during lumbar spine surgery ranges from 0.2% to 31%. Direct retraction of spinal nerve roots may cause these injuries, but the amount and duration of force that may safely be applied is not clear. Using an established porcine model, we examined the changes occurring to multimyotomal TcMEPs, mechanically elicited EMGs, and evoked EMGs during continuous retraction of a nerve root at a constant force applied over 10 minutes. Methods. TcMEP, mechanically elicited EMG, and evoked EMG responses were recorded from the tibialis anterior (TA) muscle in 10 experiments. The dominant root innervating the TA was determined with evoked EMG; preretraction TcMEP and nerve root stimulation threshold (NRT) was obtained. The dominant root was retracted at 2 Newton (N) for 10 minutes. TcMEP trials were elicited every minute during retraction. NRT was measured immediately after retraction. TcMEP and NRT were measured after 10 minutes of recovery. Results. During the 10 minutes of retraction at 2 N, the amplitude of the TA muscle progressively decreased in all trials in a highly significant curvilinear fashion. The mean TcMEP amplitude decreased 59% +/- 14% from baseline values. The mean NRT after 10 minutes of retraction at 2 N rose to 1.8 +/- 0.7 mA (P < 0.01 vs. baseline). The NRT increase after retraction strongly correlated with the decrease in motor evoked potentials amplitude in the TA (R-2 = 0.90, P < 0.001). EMG activity was variable; tonic EMG was observed in only 2 nerve roots (20%). Conclusion. Three electrophysiologic methods were used intraoperatively to assess neural function during retraction of a single nerve root. Retraction produced consistent changes in TcMEPs and evoked EMG. These 2 methods show promise for assessing the limits on the force and duration of nerve root retraction during spine surgery. Mechanically elicited EMG was not sensitive to the amount and duration of nerve root retraction.