Abstract P603: Atrial Cardiopathy and Brain Infarction in Multiple Vascular Territories in the ARCADIA Trial

Abstract

Introduction: Some embolic strokes of undetermined source (ESUS) are likely caused by occult cardiac embolism. One potential cardioembolic source is atrial cardiopathy without atrial fibrillation (AF). Patients with cardiac embolism more often have brain infarcts in multiple vascular territories than those with stroke mechanisms not involving a central embolic source. Hypothesis: In patients with ESUS, atrial cardiopathy is associated with brain infarction in multiple vascular territories. Methods: The ARCADIA trial is enrolling ESUS patients, screening them for atrial cardiopathy, and randomly assigning those with atrial cardiopathy to aspirin or apixaban. In the trial, atrial cardiopathy is defined as ≥1 of the following: P-wave terminal force >5,000 μV*ms in ECG lead V 1 , serum NT-proBNP >250 pg/mL, and left atrial diameter index ≥3 cm/m 2 on echocardiogram. Site investigators report whether the index CT or MRI showed brain infarction in the left carotid, right carotid, or posterior circulation, or some combination. In this analysis, our primary outcome was brain infarction in more than one of these three vascular territories. Our secondary outcome was infarction in both the left and right carotid territories. Results: Among 1,707 ESUS patients enrolled in ARCADIA, 679 (39.8%) met the trial’s randomization criteria for atrial cardiopathy and 213 (12.5%) had multi-territorial brain infarcts. The prevalence of brain infarction in more than one vascular territory was 14.0% in those with atrial cardiopathy versus 11.5% in those without (prevalence ratio, 1.22; 95% CI, 0.95-1.57). The prevalence of brain infarction in both the left and right carotid territories was 9.1% in those with atrial cardiopathy versus 7.4% in those without (prevalence ratio, 1.24; 95% CI, 0.90-1.70). Conclusions: These preliminary analyses from ARCADIA suggest a possible association between atrial cardiopathy and brain infarction in multiple vascular territories, but further analysis of a larger sample is needed to conclusively test whether our atrial cardiopathy definition is associated with the classic neuroimaging profile of cardiac embolism.