Hemispheric atrophy as a predictor for naming recovery following left hemisphere ischemic stroke

Numerous large-scale epidemiological studies investigating the trajectory of cognitive recovery after ischemic stroke have presented data suggesting an immediate drop in cognition acutely post-stroke followed by persistent, accelerated decline over time when averaged as a group. We sought to further examine this trend, speculating that the average persistent decline may be a reflection of two subgroups with vastly different prognoses: 1) a minority experiencing decline secondary to neurodegenerative processes like vascular dementia and Alzheimer's disease, and 2) a majority without marked progressive brain atrophy who typically see improvement. Our team thus investigated atrophy's association with language recovery, hypothesizing that declining naming performance in the year after left hemisphere ischemic stroke would be correlated to atrophy of the contralesional hemisphere. We postulated that volume loss within the lesioned hemisphere would be less informative due to separate confounding processes related to the stroke itself like Wallerian degeneration and encephalomalacia. Participants (n=72; MSD age=6011) in a longitudinal cohort study of language following left hemisphere ischemic stroke were included if they completed an MRI both acutely and chronically (either 6- or 12-months post-stroke). Naming performance was assessed using the Boston Naming Test, stroke volumes were extracted from acute imaging, and atrophy was measured as the monthly percent change in hemispheric volume from baseline to chronic scan for each individual. Pearson's correlations were calculated to determine the relationship between lesion volume and atrophy along with atrophy and change in Boston Naming Test score. Lesion volume was negatively correlated to the monthly percent change of volume in the left (ipsilesional) hemisphere (r=-0.48; p<0.0001) but was not correlated to rate of right (contralesional) hemisphere volume loss. While there was no clear relationship between atrophy of the left hemisphere and language recovery, we found that volume changes of greater negative magnitude within the right hemisphere (increased atrophy) were associated with worse functional recovery of language (r=0.38; p=0.0025). By showing that atrophy of the right hemisphere was not significantly impacted by left hemisphere lesion size, we suggest that accelerated volume loss in the non-lesioned hemisphere after stroke may be indicative of a separate pathology. We then go on to support this claim with behavioral data showing that greater rates of volume loss within the non-lesioned hemisphere were associated with poorer naming recovery. Together, these findings imply that contralesional atrophy after stroke may have negative implications for recovery and could serve as a useful imaging signature for separate neurodegenerative processes.