THE ABSENCE OF BRACHIAL-PLEXUS INJURY IN STROKE

Brachial plexus injury and proximal mononeuropathy have been reported as a potential complications in the hemiplegic shoulder after a stroke . The diagnosis of brachial plexus injury and proximal mononeuropathy in the hemiplegic extremity is complicated by the upper motor neuron f indings on physical examination and by the diffusely abnormal electrod iagnostic test results frequently seen in hemiplegic limbs. This study investigated the incidence of brachial plexus injury and proximal mon oneuropathy after a thromboembolic stroke. Hemiplegic patients (n = 50 ) underwent physical examination, needle electromyography of the hemip legic extremities and nerve conduction studies across the brachial ple xus within 4 months after a stroke. Combining the physical examination and electromyographic findings we were unable to make a diagnosis of brachial plexus injury or proximal mononeuropathy in any hemiplegic pa tient. Spontaneous electromyographic activity was observed in 68% of t he arms and 70% of the legs examined on the hemiplegic side. The sever ity and incidence of spontaneous activity was evenly distributed in up per and lower trunk muscles. Mean central latencies across the lower b rachial plexus were slightly delayed (12.5 +/- 2 v 11.6 +/- 2.2 ms, P < 0.01) compared with the contralateral normal limb, but in no case wa s the F wave unilaterally unelicitable. The mean hypothenar compound m uscle action potential amplitude was diminished (7 +/- 2.7 v 9.2 +/- 4 .1 mV, P < 0.01) in the hemiplegic hand compared with the normal side and the degree of amplitude loss inversely corresponded (r = -0.6, P < 0.01) to the amount of spontaneous electromyographic activity observe d in the first dorsal interosseus muscle. There was no correlation bet ween compound muscle action potential amplitude and muscle strength, a nd no individual patient had an absent or severely reduced compound mu scle action potential amplitude, even when there was no volitional mov ement in the corresponding muscle (n = 16). There was a negative corre lation (r = -0.8, P < 0.001) between spontaneous activity and muscle s trength in the hemiplegic arm. A positive correlation (r = 0.5, P < 0. 01) was observed between the number of days following stroke that need le examination was performed and the degree of spontaneous electromyog raphic activity observed. In the context of the literature we conclude d that these findings probably reflect minimal motor axon degeneration secondary to the cortical lesion rather than any brachial plexus inju ry or proximal mononeuropathy with severe axonal loss or complete cond uction block.