Mesencephalic corticospinal atrophy predicts baseline deficit but not response to unilateral or bilateral arm training in chronic stroke.

OBJECTIVE Stroke survivors with motor deficits often have pyramidal tract atrophy caused by degeneration of corticospinal fibers. The authors hypothesized that the degree of atrophy correlates with severity of motor impairment in patients with chronic stroke and predicts the response to rehabilitation training. METHODS They performed a post hoc analysis of 42 hemiparetic patients (>6 months) who had been randomized to either 6 weeks of bilateral arm training with rhythmic auditory cueing (BATRAC) or dose-matched therapeutic exercise (DMTE). Arm function was measured using the Fugl-Meyer (FM) and modified Wolf Motor Function Test (WMFT). Structural MRI and diffusion tensor imaging (DTI) on the pontine level measured corticospinal tract (CST) atrophy by planimetric measurement of the mesencephalon (mesencephalic atrophy ratio) and fractional anisotropy (FA), respectively. Voxel-based lesion symptom mapping (VLSM) was used to determine the lesions associated with highest degrees of atrophy. The predictive value of CST atrophy for impairment and training response was analyzed. RESULTS CST atrophy predicted baseline motor arm function measured by the FM and WMFT. The authors found only a trend for the correlation with FA. No measure of atrophy predicted response to either BATRAC or DMTE. CST atrophy was higher with larger lesions and those that affected the CST. VLSM identified internal capsule lesions as being associated with highest CST atrophy. CONCLUSION Larger lesions, internal capsule lesions, and those overlapping the pyramidal tract are associated with greater CST atrophy. CST atrophy explains in part the variability of baseline deficits but does not seem to predict the response to BATRAC or unilateral arm training on upper-extremity function.