RESUMO
OBJECTIVE: This study evaluated the risk of subclinical atrial fibrillation (AF) in patients with central retinal artery occlusion (CRAO) compared to those with cryptogenic stroke using implantable loop recorders (ILR). METHODS: We conducted a retrospective analysis of 273 consecutive patients who had ILRs inserted at our institution for either cryptogenic stroke (n = 227) or CRAO (n = 46). Our primary endpoint was a time to event analysis for the new diagnosis of AF by ILR. Univariable and multivariable Cox proportional hazard models were used to determine the predictors of time-to-AF. RESULTS: A total of 64 patients were found to have newly diagnosed AF by remote monitoring of the ILR. AF was detected in 57 of 227 (25%) cryptogenic stroke patients by the end of a maximum 5.1 years follow-up and in seven of 46 (15%) CRAO patients by the end of a maximum 3.6 years follow-up (P = .215, log-rank test). The Kaplan-Meier estimates for freedom from AF was 59.4% for CRAO and 66.6% for cryptogenic stroke (P = NS, log-rank test). Baseline variables predicting AF included older patients, higher CHADS2 VASC score, longer PR interval on initial EKG evaluation, and mitral annular calcification on transthoracic echocardiogram. CONCLUSIONS: Patients with CRAO are at risk for subclinical AF, similar to those with cryptogenic stroke. Long-term monitoring to detect AF may lead to changes in pharmacotherapy to reduce the risk for subsequent stroke.