Introduction: Cortically-based subclinical infarcts are considered risks for embolic stroke, but few studies have stratified subcortical SBI by penetrating versus medullary artery location. We hypothesized that subclinical medullary infarcts are due to small emboli and are predictors of embolic stroke.
Methods: Stroke-free participants in the Northern Manhattan Study underwent a brain MRI to assess for subclinical biomarkers of vascular disease. Subcortical brain infarcts were defined voids >3mm on axial T1 and FLAIR images, with perilesional FLAIR hyperintensities referred to as pathology-informed subclinical brain infarcts (PI-SBI). Each subcortical PI-SBI was rated as penetrating or medullary by two vascular neurologists blinded to stroke subtypes. Participants were followed prospectively for incident stroke. Two vascular neurologists ascertained ischemic stroke subtypes independent of brain MRI imaging at baseline. Embolic stroke required a cardiac source or based on a superficial location. Cox proportional risk models generated hazard ratios and 95% confidence intervals (HR, 95% CI) adjusting for age, sex, ethnicity and traditional vascular risk.
Results: The sample included 1290 NOMAS participants (mean age 71±9 years, 60% women, 66% Hispanic, 75% with hypertension) who were followed on average 9±3 years. At baseline, 19% of participants had PI-SBI (11% medullary artery, 7% penetrating artery and 3% cortical). During follow up, 80 participants (6.2%) had stroke (3.6% embolic, 2.4% due to intracranial artery disease (i.e. small and large), and 0.2% other subtypes). In a fully adjusted model, medullary artery (2.04, 1.00-4.14) and not penetrating artery PI-SBI (1.64, 0.99-2.70) predicted risk of embolic stroke. Distal field PI-SBI (i.e. cortical + medullary artery) were even more robust predictors of embolic stroke (2.17, 1.11-4.25). Penetrating artery (1.98, 1.09-3.61) and not medullary artery PI-SBI (HR 1.03, 0.34-3.06) predicted risk of intracranial artery stroke.
Conclusions: Subtyping PI-SBI by location and plausible mechanisms may help with risk stratification for clinical trials testing stroke prevention strategies. Our data suggest not all subcortical infarcts are due to small vessel disease.