Reduced Upper Limb Recovery in Subcortical Stroke Patients With Small Prior Radiographic Stroke.

Research imaging costs limit lesion-based analyses in already expensive large stroke rehabilitation trials. Despite the belief that lesion characteristics influence recovery and treatment response, prior studies have not sufficiently addressed whether lesion features are an important consideration in motor rehabilitation trial design. Using clinically-obtained neuroimaging, evaluate how lesion characteristics relate to upper extremity (UE) recovery and response to therapy in a large UE rehabilitation trial. We reviewed lesions from 297 participants with mild-moderate motor impairment in the Interdisciplinary Comprehensive Arm Rehabilitation Evaluation (ICARE) study and their association with motor recovery, measured by the UE Fugl-Meyer (UE-FM). Significant lesion features identified on correlational and bivariate analysis were further analyzed for associations with recovery and therapy response using longitudinal mixed models. Prior radiographic stroke was associated with less recovery on UE-FM in participants with motor impairment from subsequent subcortical stroke (-5.8 points) and in the overall sample (-3.6 points), but not in participants with cortical or mixed lesions. Lesion volume was also associated with less recovery, particularly after subcortical stroke. Every decade increase in age was associated with 1 less point of recovery on UE-FM. Response to specific treatment regimens varied based on lesion characteristics. Subcortical stroke patients experienced slightly less recovery with higher doses of upper extremity task-oriented training. Participants with cortical or mixed lesions experienced more recovery with higher doses of usual and customary therapy. Other imaging features (leukoaraiosis, ischemic vs. hemorrhagic stroke) were not significant. ICARE clinical imaging revealed information useful for UE motor trial design: stratification of persons with and without prior radiographic stroke may be required in participants with subcortical stroke, the majority of motor rehabilitation trial participants. Most of the prior radiographic strokes were small and cortically-based, suggesting even minor prior brain injury remote to the acute stroke lesion may limit spontaneous and therapy-related recovery. Lesion location may be associated with response to different therapy regimens, but the effects are variable and of unclear significance.