Branch atheromatous disease diagnosed as embolic stroke of undetermined source: A sub-analysis of NAVIGATE ESUS

Background Branch atheromatous disease (BAD) is distinctive from large and small arterial diseases, which is single subcortical infarction larger than lacunar stroke in the territories of deep perforators without relevant arterial stenosis. BAD meets the current criteria of embolic stroke of undetermined source. We performed an exploratory analysis of BAD in patients recruited to NAVIGATE embolic stroke of undetermined source, a randomized controlled trial to compare rivaroxaban and aspirin in embolic stroke of undetermined source patients. Methods and results Among 3972 stroke patients in cerebral hemispheres with intracranial arterial imaging, 502 (12.6%) patients met the criteria for BAD. BAD was associated with younger age (years; OR: 0.97, 95% CI: 0.96–0.98), race (Asian; OR: 1.78, 95% CI: 1.44–2.21), region (Eastern Europe; OR: 2.49, 95% CI: 1.87–3.32), and higher National Institute of Health Stroke Scale (OR: 1.17, 95% CI: 1.12–1.22) at randomization. During follow-up, stroke or systemic embolism (2.5%/year vs. 6.2%/year, p = 0.0022), stroke (2.1%/year vs. 6.2%/year, p = 0.0008), and ischemic stroke (2.1%/year vs. 5.9%/year, p = 0.0013) occurred less frequently in BAD than non-BAD patients. There were no differences in annual rates of stroke or systemic embolism (2.5%/year vs. 2.5%/year, HR: 1.01, 95% CI: 0.33–3.14) or major bleeding (1.3%/year vs. 0.8%/year, HR: 1.51, 95% CI: 0.25–9.05) between rivaroxaban and aspirin groups among BAD patients. Conclusions BAD was relatively common, especially in Asian and from Eastern Europe among embolic stroke of undetermined source patients. Stroke severity was higher at randomization but recurrence of stroke was fewer in BAD than non-BAD patients. The efficacy and safety of rivaroxaban and aspirin did not differ among BAD patients.