RESUMO
BACKGROUND AND PURPOSE: Infarct topology is a key determinant in classification of a stroke as potentially embolic, with cortical and multifocal lesions being presumed embolic. Whether isolated subcortical multifocal infarcts are likely embolic has not been well studied. METHODS: A prospective, single-center cohort study of consecutive patients with acute multifocal strokes confirmed on diffusion-weighting imaging (DWI) was queried, and patients compared according to the presence of isolated subcortical infarct topology versus cortical ± subcortical topology. Descriptive statistics and multivariable logistic regression were used to determine independent predictors of cryptogenic, subcortical infarcts. RESULTS: Of 1739 patients screened, 743 had complete diagnostic testing with DWI evidence of acute infarction, 183 (24.6%) of whom had a multifocal stroke pattern. Isolated subcortical involvement was disproportionate among patients with ESUS (64.9%) when compared to patients with cardioembolic (24.3%) or large vessel disease (10.8%, p<0.01). Following multivariable adjustment, independent predictors of isolated subcortical multifocal infarction were milder strokes (OR 0.94, 95%CI 0.89-0.98) and higher grade Fazekas score (OR 2.32, 95%CI 1.02-5.29), while cardioembolism (OR 0.30, 95%CI 0.08-1.13) and large vessel disease (OR 0.27, 95%CI 0.08-0.91) remained inversely associated (as compared to ESUS). CONCLUSIONS: These data suggest that multifocal subcortical infarctions are less likely to have an associated proximal embolic source than multifocal infarctions with cortical involvement. The strong association with chronic microvascular disease suggests this topology is more consistent with acute-on-chronic microvascular injury rather than an occult embolic source.