Electromyography in the detection of mechanically induced spinal motor tract injury: observations in diverse porcine models.

ABSTRACT

OBJECT Porcine spinal cords were mechanically injured at the thoracic level while recording muscle-derived electrically stimulated transcranial motor evoked potentials (TcMEPs) and electromyography (EMG) readings from the same electrode derivations. The authors postulated that midthoracic spinal cord injury caused by diverse methods can trigger hindlimb EMG activity. Early detection of hindlimb EMG activity may permit avoidance of motor conduction block (TcMEP loss). METHODS: Twelve pigs underwent midthoracic spinal cord exposure. Spinal cord sectioning was performed to define dorsal column versus lateral spinal cord contribution to muscle-derived electrically stimulated spinal cord motor evoked potentials (SC MEPs) and TcMEPs (in 2 pigs). A bipolar needle stimulator was placed within intramedullary sites to 1) acquire electrically stimulated motor evoked potentials in the hindlimbs, and 2) induce mechanically stimulated hindlimb EMG activity at sites responsive to electrical stimulation (in 2 pigs). Transcranial MEPs and EMG recordings were observed during spinal cord distraction (in 3 pigs), slow and rapid extradural spinal cord compression with a metal caliper (in 3 pigs), and rapid extradural spinal cord compression with a spring-loaded clip (in 2 pigs). RESULTS: Lateral cord (but not dorsal column) sectioning abolished both SC MEPs and TcMEPs. Intramedullary electrical and mechanical stimulation within the lateral (but not dorsal) cord elicited ipsilateral hindlimb MEPs and EMG activity ("EMG injury discharge"), respectively. Distraction inconsistently produced EMG injury discharges concomitant with TcMEP loss. Rapid extradural spinal cord compression with a metal caliper or spring-loaded clip consistently induced EMG injury discharges (in 4 of 4 pigs); slow compression did not elicit EMG activity. Brief extradural spring-loaded clip compressions (1-2 seconds) elicited EMG injury discharges without TcMEP loss; 14-second clip compression effected EMG injury discharges and TcMEP loss, which recovered after clip removal. CONCLUSIONS: Electromyographic activity (referred to as "EMG injury discharges" in the present study) can be elicited both by intramedullary manipulation and rapidly applied transaxial spinal cord compression. Preliminary observations suggest that these EMG injury discharges precede and may anticipate TcMEP loss. Presumably, rapid deformation of spinal motor tracts (which appear to lie within the lateral porcine spinal cord) generates descending volleys which can bring to firing threshold lumbar motor neurons (and recording of EMG injury discharges). Intraoperative neuromonitoring of high-risk spinal surgeries at the spinal cord level may benefit from the addition of EMG recording to tests of spinal cord motor conduction such as TcMEP. Further clinical trials are required to examine EMG efficacy in this context.