ABSTRACT
We studied 21 patients with lumbosacral radiculopathy with segmental somatosensory evoked potentials (SEPs) recorded over both spine and scalp following saphenous, superficial peroneal, and sural nerve stimulation. Spinal SEPs were abnormal in 10 patients. In 3 patients, SEPs detected abnormalities not seen on EMG examination. With 1 exception, all anatomic levels of SEP abnormalities matched that of radiographic, EMG, or clinical abnormalities. SEPs were abnormal in 41% of nerve roots shown to be involved by other techniques. SEPs added to the clinical evaluation in 4 patients, but were less accurate than a combination of EMG and radiography in indicating the extent of nerve root involvement. We conclude that spinal SEPs following segmental sensory stimulation are useful in the evaluation of lumbosacral radiculopathies and complement information provided by the EMG. In contrast, scalp-recorded segmental SEPs rarely provide additional useful clinical information.
Subject(s)
Evoked Potentials, Somatosensory , Spinal Cord/physiopathology , Spinal Nerve Roots , Adult , Electromyography , Humans , Lumbosacral Region , Middle Aged , Nervous System Diseases/diagnosis , Nervous System Diseases/physiopathology , Scalp/physiopathology , Sensitivity and Specificity , Spinal Nerve Roots/physiopathologyABSTRACT
Dorsal root function cannot presently be measured directly. The H-reflex is an indirect measure of dorsal root function but only for the S1 root. Spinal somatosensory evoked potentials (SEPs) following dermatomal stimulation of the legs have the potential of providing direct data reflecting dorsal root function but have not been reliably recorded in normal subjects. We have developed a reliable technique for recording SEPs at the lumbar root entry zone following segmental sensory stimulation of the legs. The saphenous, superficial peroneal, and sural nerves were stimulated representing the L3/L4, L5 and S1 roots respectively. Reproducible responses (N-wave) were recorded over the lumbar spine in all 60 normal limbs examined. The N-wave peak latency was significantly correlated with lower limb length. The conduction velocities from the stimulation sites to the lumbar spine were similar to published values for peripheral conduction velocities in these nerves. The mean inter-limb latency differences for the N-wave peak were: L3/L4 0.61 msec; L5 0.35 msec; and S1 0.57 msec. The mean N-wave amplitudes were: L3/L4 0.11 microV; L5 0.28 microV; and S1 0.23 microV. This technique is a direct measure of dorsal root integrity. Unlike scalp recorded SEPs, the lumbar N-wave is not state-dependent and is unaffected by lesions within the brain and rostral cord.
