[Origin of short latency somatosensory evoked potential following median nerve stimulation]

Short latency somatosensory evoked potentials (S-SEP) to median nerve stimulation recorded from scalp-shoulder derivation consisted of four positivities (P1, P2, P3 and P4) and one negativity (N). There is still, however, significant controversy regarding the generators of these peaks. The purpose of the present study is to attempt to solve some of the controversies by comparing the potentials recorded on surface montages with intrasurgical tracing obtained while monitering evoked potentials. This is complemented by analysis of S-SEP in patients with disorders of the cervical cord, brain stem or thalamus. Average latency of each component in normal subjects (N = 17) was as follows; P1: 7.6 +/- 0.6 ms P2: 9.6 +/- 0.7 ms, P3: 11.8 +/- 0.8 ms, P4: 12.9 +/- 0.5 ms, N: 17.7 +/- 0.8 ms. The peak latency of P1 was always shorter than that of negativity recorded over the Erb's point following the median nerve stimulation. The peak latency of P2 corresponded to that of the earliest negativity of spinal evoked potential, recorded from the epidural dorsal surface of the lower cervical cord during the operation, which was considered as the potential of afferent volley at the posterior root or spinal entry portion. In a case with severe spondylotic radiculopathy, P2 and following components were abolished. The peak latency of P3 corresponded to that of negative potential recorded over the high cervical dorsal column or cuneate nucleus.(ABSTRACT TRUNCATED AT 250 WORDS)