Effect of lesion location on upper limb motor recovery after stroke.

BACKGROUND AND PURPOSE The purpose of this study was to assess the effects of stroke involvement of primary and secondary hemispheric motor systems and corticofugal tracts on arm and hand recovery. METHODS Forty-one patients participating in an inpatient stroke rehabilitation database, admitted 17+/-2 (SEM) days after initial unilateral hemispheric ischemic stroke, with neuroimaging studies performed >48 hours after stroke and with minimal upper limb (UL) movement (admission Fugl-Meyer UL motor scores 18, synergies+isolated movements. Lesions affecting the following structures were recorded: primary motor cortex, premotor area, supplementary motor area, anterior half of the middle third of corona radiata (secondary motor efferents), posterior half of the middle third of corona radiata (primary motor efferents), genu, anterior and posterior limbs of the internal capsule (PLIC), basal ganglia, and thalamus. chi(2) Analysis and ANOVA were used to study the significance of stroke location on UL motor recovery. RESULTS The effect of involvement of primary, premotor, or supplementary motor areas on motor recovery did not reach statistical significance. Patients with purely cortical stroke were likely to recover UL isolated movement (3 of 4) compared with purely subcortical (1 of 17) or mixed cortical-subcortical stroke location (2 of 20) (P:=0.009). Of those with cortical, subcortical, or mixed cortical plus subcortical lesions sparing the PLIC, 5 of 13 recovered isolated UL movement (P:=0.01). Only 1 of 28 patients with involvement of the PLIC plus adjacent corona radiata, basal ganglia, or thalamus recovered isolated UL movement (P:=0.01). Patients with small lacunar strokes affecting only the PLIC did not have sufficient motor deficits 2 weeks after stroke to meet inclusion criteria. CONCLUSIONS The probability of recovery of isolated UL movement decreases progressively with lesion location as follows: cortex, corona radiata, and PLIC. This is consistent with our current understanding of redundant cortical motor representation and convergence of corticofugal motor efferents as they pass through the corona radiata to the PLIC.