Integrity of arcuate fasciculus is a good predictor of language impairment after subcortical stroke

Subcortical aphasia results from damage to subcortical regions, but the exact underlying mechanism is unclear. We investigated the influence of arcuate fasciculus (AF) damage on the severity of subcortical aphasia after stroke using brain diffusion tensor imaging (DTI). We enrolled subjects from the “Stroke Outcome Prediction (STOP)” database: 41 patients with first-ever, left hemispheric subcortical stroke (33 with hemorrhagic, eight with ischemic stroke) who underwent the Western Aphasia Battery (WAB) evaluation and DTI. We performed a correlation analysis between lesion volume (LV) and the DTI parameters of the AF, including fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), axial diffusivity (AD), and AF type, and the WAB scores, and a classification and regression tree (CART) analysis to identify significant neuroimaging predictors for subcortical aphasia. FA significantly correlated with the aphasia quotient (AQ) and the naming score, and LV with the AQ and several WAB subtests (fluency, repetition, naming). At the subgroup level, FA and AF type showed significant correlations with the AQ, comprehension, and naming in the ischemic stroke group, while LV was associated with naming in the hemorrhagic stroke group. LV and FA were significant predictors of the AQ in all subjects, while FA was the only predictor of the language score in the hemorrhagic stroke group. Damage to the AF may be an important underlying mechanism of subcortical aphasia, and the FA value representing the integrity of the AF is an important predictor of the severity of subcortical aphasia.