Monitoring of evoked potentials during spinal cord ischaemia: experimental evaluation in a rabbit model.

OBJECTIVES

Somatosensory evoked potentials (SEPs), spinal evoked potentials (Spinal-EPs), and motor-evoked potentials (MEPs) were monitored in a rabbit model of spinal cord ischaemia to evaluate their accuracy and relationship to clinical status.

METHODS

A modified rabbit spinal cord ischaemia model of infrarenal aortic occlusion for 21 min was employed (30 rabbits). After baseline SEPs, Spinal-EPs, and MEPs were obtained, evoked potentials were recorded continuously during and after clamping of the aorta (30 min). Neurological outcome at 24 h was correlated with evoked potentials, and histopathological findings.

RESULTS

Fifteen animals became paraplegic. MEPs were always abolished after clamping of the aorta while Spinal-EPs and SEPs remained. The sensory evoked potentials (SEPs and Spinal-EPs) were the least sensitive to spinal cord ischaemia, and their presence had no correlation with the final clinical status (50% of false negatives). This was consistent with histopathological examination that showed damage almost entirely confined to the anterior horn, while the dorsal columns were generally well preserved. High spine MEPs evoked by twitch stimulation was the best predictor of clinical outcome (0% of false negatives, 0% of false positives).

CONCLUSIONS

SEPs and Spinal-EPs cannot be used as safe monitors of ischaemia of the spinal cord. High spine MEPs evoked by twitch stimulation was the most useful for real-time evaluation of spinal cord ischaemia, and the best predictor of neurologic outcome during reperfusion.