Abstract 131: Association Between Left Ventricular Injury Detected by Cardiac MRI and Incident Ischemic Stroke: The Multi-Ethnic Study of Atherosclerosis

Abstract

Background: The association between left ventricular injury (LVI) detected by cardiac magnetic resonance imaging (CMR) and the risk of a subsequent ischemic stroke is unknown. Methods: We studied stroke-free participants in the Multi-Ethnic Study of Atherosclerosis study who underwent a steady-state free precession CMR at Exam 5 (2010-2012). The exposure of interest was LVI defined by either LV ejection fraction (EF) < 50% or circumferential strain > -10 in any LV apical wall. The primary outcome was incident ischemic stroke. We performed Cox proportional hazard analysis to model the association between LVI and incident ischemic stroke, adjusting for demographic and clinical covariates. Results: The analytic sample consisted of 2,584 participants (n=302 or 11.7% with LVI, n=2,282 or 88.3% without LVI). There were a total of 18 ischemic stroke events among those with LVI (6%) and 65 among those without LVI (3%). Participants with LVI were significantly older, male, Black, had higher BMI, and lower HDL. Participants with LVI were more likely to have diabetes, tobacco use, and atrial fibrillation (24% versus 14%, p<0.001). There was no significant difference in history of cardiovascular disease events (6% versus 4%, p=0.143). LVI was significantly and independently associated with incident ischemic stroke (adjusted hazard ratio 1.80, 95% C.I. 1.04-3.12). When participants were stratified into 3 mutually exclusive groups of no LVI, LVEF > 50% and LV apical circumferential strain < -10, and LVEF < 50% and LV apical circumferential strain > -10, the hazard of incident ischemic stroke was significantly higher for each of the two latter groups compared to the no LVI group (log-rank p-value=0.003; Figure 1 ). Conclusion: LVI identified by CMR is a significant, independent predictor of incident ischemic stroke. LV apical circumferential strain > -10% may be a novel and quantitative stroke risk factor. Further studies are needed to validate these CMR biomarkers of stroke risk.