Clinical findings and electrodiagnostic testing in 108 consecutive cases of lumbosacral radiculopathy due to herniated disc

STUDY AIM: This prospective study aim to examine whether clinical findings and electrodiagnostic testing (EDX) in patients with lumbosacral monoradiculopathy due to herniated disc (HD) differ as a function of root involvement level (L5 vs. S1) and HD zone (paramedian vs. intraforaminal). PATIENTS AND METHODS: All patients with L4, L5 or S1 monoradiculopathy were prospectively enrolled at four electromyography (EMG) labs over a 2-year period. The diagnosis was based on a congruence between patient history and MRI evidence of HD. We compared the sensitivities of clinical findings and EDX with respect to both root involvement level and HD zone. Multivariate logistic regression was performed in order to verify the association between abnormal EMG, clinical, and neuroradiological findings. RESULTS: One hundred and eight patients (mean age 47.7 years, 55% men) were consecutively enrolled. Sensory loss in the painful dermatome was the most frequent finding at physical examination (56% of cases). EMG was abnormal in at least one muscle supplied by femoral and sciatic nerves in 45 cases (42%). Inclusion of paraspinal muscles increased sensitivity to only 49% and that of proximal muscles was useless. Motor and sensory neurography was seldom abnormal. The most frequent motor neurographic abnormalities were a delay of F-wave minimum latency and decrease in the compound muscle action potential amplitude from extensor digitorum brevis and abductor hallucis in L5 and S1 radiculopathies, respectively. Sensory neurography was usually normal, the amplitude of sensory nerve action potential was seldom reduced when HD injured dorsal root ganglion or postganglionic root fibres. Multivariate logistic regression analysis showed that EMG abnormalities could be predicted by myotomal muscular weakness, abnormal deep reflexes, and paraesthesiae. The only clinical and electrophysiological differences with respect to root involvement level concerned deep reflexes and motor neurography of deep peroneal and tibial nerves. CONCLUSIONS: Only some EDX parameters are helpful for the diagnosis of lumbosacral radiculopathy. EMG was abnormal in less than 50% of cases and its abnormalities could be predicted by some clinical findings. However, neurography is useful as a tool for differential diagnosis between radiculopathy and more diffuse disorders of the peripheral nervous system (polyneuropathy, plexopathy).