Clinical experience with motor and cerebellar evoked potential monitoring.

We are reporting 98 cases of motor evoked potential (MEP) monitoring performed between 1982 and 1986. These were divided into supratentorial, posterior fossa, and spinal cord categories. We observed in this sample that the peripheral nerve or electromyographic response was substantially more sensitive than the spinal cord response to injury and hypotension during operation and was often present bilaterally. Reversible weakening or loss of the peripheral nerve response was not associated with a deficit. However, weakening of the peripheral nerve response without recovery could warn of a motor defect after operation. The spinal cord responses can change so little in amplitude and latency with injury conditions that their reliable use during operation without accompanying peripheral nerve and muscle response monitoring may be compromised, especially in view of the often difficult recording environment in the operating room. With spinal cord monitoring, the MEP seemed closely correlated with the stresses that we imposed on the cord, as well as with subsequent clinical outcome. In posterior fossa cases, we observed sensitivity of the MEP to manipulation of the nervous system and reliable indication of the outcome in the cases monitored. Supratentorial cases present a more complex environment for monitoring; a potential pitfall is to stimulate an area that produces responses in the peripheral nerves and spinal cord, but that is not being compromised by the injury process. Alternatively, stimulating too large an area of cortex or stimulating with too high a current, which penetrates to white matter below the gray matter, may not show an injury. Although we did not encounter cases where permanent deficits were missed by the motor evoked potential, we were concerned by the appearance of temporary deficits. These may be related to technical limitations of our methods and indicate that monitoring of the supratentorial process requires substantial methodological advance for high reliability. Modulation of the MEP by prior somatosensory evoked potential stimulation seems useful and promising. Cerebellar evoked potential responses were present in humans and were sensitive to injury. Overall, this test is promising.