Abstract WMP65: Prospective Validation of Predictive Features of Paroxysmal Atrial Fibrillation (PROPhecy): An Interim Analysis

Background: Atrial fibrillation (AF) has a distinct antithrombotic regimen for secondary stroke prevention. While 30-day cardiac monitoring has a greater detection rate for AF than 24-hour Holter, it is not widely accessible. Risk stratification may identify patients who could benefit most from prolonged monitoring. PROPhecy aims to prospectively validate predictive features for detection of AF found in EMBRACE, a trial using 30-day monitoring in individuals with an embolic stroke of undetermined source (ESUS). Methods: Participants were > 55 years and within six months of ESUS, without evidence of AF/flutter on 24-hour Holter. All were given an event-triggered external loop recorder for 30 days. Primary outcome was detection of sustained ( > 30 sec) or non-sustained AF/flutter on 30-day monitoring. Results: 150 of a planned 250 participants have completed long-term monitoring to date. Baseline characteristics are compared to EMBRACE (Table 1). Any AF/flutter was detected in 19.3% (EMBRACE, 16.1%). Burden of atrial premature beats in PROPhecy was low in comparison to EMBRACE and did not predict presence of AF on monitoring (Table 2). Left atrial volume index was a significant predictor of AF in both univariable and multivariable regression adjusted for age and sex (OR1.04 per mL/m 2 , 95% 1.01-1.08, p=0.02). Conclusion: Recruitment is ongoing. AF was detected in ~1/5 participants. The burden of atrial ectopy in our cohort is much lower than in EMBRACE despite similar patient characteristics and AF burden. Further work is required to assess the nature of these differences. Left atrial volume index may be helpful for risk stratification.