Abstract Body (Do not enter title and authors here): Background. The implantable loop recorder (ILR) is the most sensitive method for detecting atrial fibrillation (AF) following a cryptogenic stroke. However, to date, no study has demonstrated that it significantly reduces the rate of recurrent ischemic stroke.

Objective. To assess the cumulative incidence of recurrent ischemic stroke in patients with a prior cryptogenic stroke who received an ILR, and to evaluate whether stroke recurrence is associated with device-detected AF.

Methods. We conducted an observational study including consecutive patients with cryptogenic stroke who received an ILR in four academic centres in France. Prior to implantation, all patients underwent a standardized diagnostic workup, including cerebral CT angiography, brain MRI, carotid ultrasound, echocardiography, and at least 24 hours of ECG monitoring. Patients were followed from the date of ILR implantation until the first recurrence of stroke (primary endpoint), death from any cause (competing event), or end of follow-up (defined as ILR explant, last remote transmission, or final device interrogation). Device-detected AF and initiation of oral anticoagulation were analysed as time-dependent covariates in the univariate and multivariate analyses. Cause-specific Cox proportional Hazards model was used as the primary analysis.

Results. We included 1,001 patients (median age 68 years [IQR 59–75]; 41% female; median CHA2DS2-VASc score 4 [IQR 3–5]). Over a median follow-up of 1.9 years, 67 recurrent ischemic strokes were recorded, yielding a cumulative incidence of 3.4% per year. ILR-detected AF occurred in 275 patients (27%). In univariate analysis, AF not treated with anticoagulation was associated with a significantly higher risk of recurrent stroke (HR 3.47, 95% CI 1.08–11.11, p=0.037), whereas AF with anticoagulation was not (HR 1.26, 95% CI 0.67–2.34, p=0.473). These associations remained consistent after multivariate adjustment for age, sex, peripheral artery disease, and renal function (HR 3.27, 95% CI 1.01–10.61, p=0.048 for untreated AF; HR 1.01, 95% CI 0.52–1.98, p=0.971 for anticoagulated AF).

Conclusion. Among patients with cryptogenic stroke monitored with an ILR, recurrent ischemic stroke remains a significant concern. Device-detected AF is associated with a higher risk of recurrent stroke only when not treated with oral anticoagulation. These findings support early initiation of anticoagulation in patients with device-detected AF to reduce risk of stroke recurrence.