Distribution of lumbar spinal evoked potentials and their correlation with stimulation-induced paresthesiae.

In 7 awake patients with neuropathic lower extremity pain, spinal somatosensory evoked potentials (SEP) were elicited from the non-painful leg by electrical stimulation of the peroneal nerve and mechanical stimulation of the hallux ball. Recording was made epidurally in the thoraco-lumbar region by means of an electrode temporarily inserted for trial of pain-suppressing stimulation. In response to peroneal nerve stimulation, two major SEP complexes were found. The first complex consisted, as has been described earlier, of an initial positivity (P12), a spike-like negativity (N14), a slow negativity (N16) and a slow positivity (P23). The second complex consisted of a slow biphasic wave, conceivably mediated by a supraspinal loop. Both complexes had a similar longitudinal distribution with amplitude maxima at the T12 vertebral body. The SEP evoked by mechanical hallux ball stimulation had a relatively small amplitude, and there was no significant second complex. The relationship between stimulus intensity and SEP amplitude was negatively accelerating. The longitudinal distribution of spinal SEP was compared with the somatotopic distribution of paresthesiae induced by stimulation through the epidural electrode. It was found that stimulation applied at the level of maximal SEP generally induced paresthesiae in the corresponding peripheral region. Therefore, spinal SEP may be used as a guide for optimal positioning of a spinal electrode for therapeutic stimulation when implanted under general anesthesia. An attempt was made to record the antidromic potential in the peroneal nerve elicited from the dorsal columns by epidural stimulation. The antidromic response was, however, very sensitive to minimal changes of stimulus strength and body position of the patient, and was also contaminated by simultaneously evoked muscular reflex potentials. Thus, peripheral responses evoked by epidural stimulation appeared too unreliable to be useful for the permanent implantation of a spinal electrode for therapeutic stimulation.

Epidurally recorded cervical spinal activity evoked by electrical and mechanical stimulation in pain patients.

Spinal SEPs to electrical and mechanical stimulation of the upper limb of the non-painful side in 7 pain patients were recorded from the cervical epidural space. In response to electrical stimulation of the median nerve, the longitudinal distribution of the spinal postsynaptic negativity (N13) along the cord had a distinct level of maximal amplitude at the C5 vertebral body. When recorded at increasing distances cranial or caudal to this level, the latency of N13 was successively prolonged, in agreement with a spread-out near-field generator in the dorsal horn. Similar patterns of distribution and levels of maximal amplitude were demonstrated for the N13 wave evoked by electrical stimulation of the ulnar and thumb nerves as well as by mechanical stimulation of the thumb ball. The amplitude ratios of the N13 waves evoked by electrical stimulation of the median nerve and the thumb nerves, and by mechanical stimulation of the thumb ball were 3.9 to 1.4 to 1. The slow positive wave (P18), which has been assumed to represent recurrent presynaptic activity, had a somewhat different distribution, with a lower maximal amplitude and a less marked falling off in amplitude along the cord, as compared to the N13 component. The initial presynaptic positivity (P10) appeared with an almost constant amplitude along the cord. Tactile stimuli produced responses with considerably longer latency and duration than those obtained with electrical stimulation. There seemed to be a non-linear relationship between the amplitude of the response and the depth of skin indentation. The presented data contribute a more detailed picture of epidurally recorded spinal SEPs than previous studies. They will serve as a reference for further analysis of SEPs evoked by stimulation of the affected side in pain patients, to explore whether the painful state is associated with altered SEPs before or after therapeutic intervention.