Abstract TP216: Atrial Fibrillation Detection in Patients With an Implantable Loop Recorder After Acute Embolic Stroke of Unknown Source (ESUS)

Abstract

Introduction: The embolic stroke of unknown source concept was introduced as a more rigid analysis of patients with cryptogenic stroke representing a superselection of patients with cardioembolic stroke. These patients are particularly candidates for intermittent AF. As long as AF has not been documented, current concepts do not recommend oral anticoagulation.Implantable loop recorders (ILR) in patients with ESUS may detect AF and establish the indication for oral anticoagulation. The aim of this study was to prospectively assess and predict AF occurrence in patients with ILR after ESUS. Methods: In patients with ESUS (MR imaging based cardioembolic stroke, exclusion of structural cardiac stroke source by TEE, no AF detectable by 72h Stroke Unit monitoring and 24h holter ECG, exclusion of other stroke causes such as symptomatic carotid stenosis) an ILR was implanted and AF detection assessed by daily remote monitoring. The ILR was implanted on average 20 days after stroke. We analyzed the predictive value of different clinical and imaging characteristics for AF detection. Results: By daily remote monitoring of 124 Patients over a period of 12.7±5.5 months, AF was documented and manually confirmed in 29 of 124 patients (23.4%). First AF detection occurred on average after 3.6±3.4 months of monitoring. Characteristics of patients with and without AF detection are shown in the table. Conclusions: Patients with ESUS and asymptomatic AF detected only by long-term continuous monitoring with an ILR were on average older, had a higher CHA2DS2-VASc score and had more often microangiopathy. Other clinical parameters and features of cerebral imaging in ESUS did not increase the probability of AF detection in these preselected patients with ESUS. Importantly, ESUS selection almost doubled AF detection rate compared to recent studies such as the ILR group of the Crystal-AF trial (23.4% in 12.7±5.5 months compared to 12.4% in 12 months).