Value of somatosensory evoked potentials in thoraco- brachial outlet syndrome

INTRODUCTION: Brachial plexus involvement in symptoms of thoracic outlet syndrome (TOS) is often difficult to assess from clinical data. Conventional EMG and nerve conduction studies (NCS) do not seem reliable to all authors. For this reason, our investigations of this syndrome were complemented by study of somatosensory evoked potentials (SEP) in order to compare the results of these different techniques. PATIENTS AND METHODS: Ten patients were studied, all of whom had prominent vascular symptoms which led to their consulting a vascular surgeon. Only one had hand wasting without hypoesthesia. None had cervical rib or cervical spine anomaly. In all cases, diagnosis was confirmed by arteriography or phlebography. Operations were decided on clinical data and results of vascular investigations. Patients were tested with conventional motor and sensitive NCS F-wave studies. Needle EMG was performed in abductor pollicis brevis, first dorsal interosseus or abductor digitiminimi. Their SEP were performed as for controls. Ten controls were studied whose SEP were obtained at Erb's point (N9) and C2 cervical spine level (N13) after percutaneous stimulation of median and ulnar nerves at the wrist on both sides. The criterion of abnormality was the mean of controls + 2.5 SD for latencies. Amplitude was considered as low when it was less than 50% of the contralateral one. RESULTS: For 2 patients EMG, NCS and SEP were abnormal. One had hand wasting and denervation in hand muscles as well as slowed median and ulnar sensory conduction with low amplitude responses. SEP at Erb's point were slightly delayed after ulnar stimulation. No cervical response was obtained after ulnar stimulation. The second one had normal responses at Erb's point but delayed responses at the cervical level. In addition, N13 amplitude after ulnar stimulation was low. Four patients had normal EMG, NCS and SEP. Two patients had normal EMG and NCS, but their SEP was questionable since latencies were normal, even though amplitude was low after median and ulnar stimulation. This was not considered this to be abnormal since it was bilateral. For the remaining 2 patients (F.1), EMG and NCS and Erb's point SEP were normal, but C2 median and ulnar responses were delayed in one case and C2 ulnar response amplitude was very low on one side only in the other. In conclusion, SEP were abnormal for 4 patients out of 10 but gave more information than conventional EMG and NCS for only 2 patients. SEP abnormalities prevailed after ulnar stimulation.