Abstract
Chronic language impairments, aphasia, are a common consequence of stroke, yet the factors influencing their severity remain poorly understood. Although lesion location is widely recognized as a key determinant of aphasia severity and recovery, the impact of gene expression within the affected brain regions is largely unexplored. In this proof-of-concept study, we introduce a novel computational approach termed gene expression-based lesion-symptom mapping (GLSM). We sought to investigate whether mapping stroke lesions to brain regions with heightened expression of certain genes could offer insights into aphasia severity. For demonstration purposes, we chose the gene FOXP2 due to its controversial yet intriguing relationship with language processing. We applied GLSM to a well-characterized cohort of 91 individuals with chronic aphasia resulting from a left-hemisphere stroke (36 females, 55 males). Our findings revealed that individuals with lesions to left hemispheric language-related areas with high FOXP2 expression presented worse aphasia severity than individuals with left hemispheric, dorsal stream lesions in regions with low FOXP2 expression. Adding FOXP2 gene expression lesion load explained more variance in aphasia severity than lesion load alone. Through GLSM, we introduce a novel perspective on the relationship between brain lesions and functional deficits, shedding light on previously unexplored nuances. Our study underscores the potential of GLSM as a valuable tool in unraveling the intricate mechanisms underlying language impairments poststroke.