Abstract 072: Risk Factors for Developing Ischemic Strokes While Taking Non‐Vitamin K Antagonist Oral Anticoagulants

Introduction Over the past decade, non‐vitamin K antagonist oral anticoagulants (NOACs) have replaced warfarin as the initial strategy for stroke prevention in non‐valvular atrial fibrillation. However, ischemic strokes occurring in patients taking NOACs are becoming increasingly more frequent. In this study, we aimed to determine the clinical, echocardiographic, and neuroimaging markers associated with developing ischemic strokes in patients taking NOACs for atrial fibrillation. Methods Among atrial fibrillation patients on NOACs with and without ischemic stroke, brain MRIs were reviewed for the presence of neuroimaging markers of small vessel disease. Transthoracic echocardiography reports were reviewed to obtain left atrial volumes, presence of left ventricular hypertrophy, ejection fractions, and presence of wall motion abnormalities. Clinical and radiologic variables with significant differences between groups were entered into a multivariable regression model to determine predictors of ischemic stroke. In the ischemic stroke group, a Cox regression analysis was constructed to determine predictors of ischemic stroke recurrence during follow‐up. Results 112 patients with ischemic stroke and 94 controls (free of ischemic stroke) were included in the study. The mean age (77±9 years vs. 74±11 years, p = 0.13) and proportion of patients with CHA2DS2‐VASc>2 were similar between controls and ischemic stroke patients (82% vs. 83%, p = 0.86). The variables that were significantly different between groups included apixaban use, dabigatran use, prior cerebrovascular events, hemoglobin A1c (HbA1c), left ventricular hypertrophy, left atrial volume index, and severe white matter hyperintensities. After multivariable adjustment, prior cerebrovascular events (aOR 23.86, 95% CI [6.02–94.48]), HbA1c levels (aOR 2.36, 95% CI [1.39–3.99]), left ventricular hypertrophy (aOR 2.73, 95% CI [1.11–6.71]) and left atrial volume index (aOR 1.05, 95% CI [1.01–1.08]) increased the risk of stroke, whereas apixaban use appeared to decrease the risk (aOR 0.38, 95% CI [0.16–0.92]). Among patients with ischemic stroke, there were 103 survivors, 82 (80%) of which had follow‐up data. The median follow‐up duration for these patients was 2.6 (0.5–4.7) years. Of these 82 patients, 11 (13%) experienced an ischemic stroke during the follow‐up period. Malignancy was associated with ischemic stroke recurrence (aHR 4.90, 95% CI [1.35–18.42]) after adjustment for age and chronic renal failure. Conclusion Prior cerebrovascular events, diabetes, left ventricular hypertrophy, and left atrial size are associated with the development of ischemic stroke among NOAC users. Future studies should aim to investigate nonpharmacological stroke prevention strategies in patients with the vascular risk factors and echocardiographic risk markers identified in the present study.