Detection of atrial fibrillation using an implantable loop recorder in patients with previous cryptogenic stroke: the SPIDER-AF registry (Stroke Prevention by Increasing DEtection Rates of Atrial Fibrillation).

Abstract

The use of implantable loop recorder (ILR) to detect atrial fibrillation (AF) in patients with a history of cryptogenic stroke (CS) has seldom been investigated in "real-world" settings. This study aimed to present the results of the Stroke Prevention by Increasing DEtection Rates of Atrial Fibrillation (SPIDER-AF) registry. SPIDER is a multicentric, observational registry, including 35 facilities all over Germany. It enrolled outpatients without AF history and with CS before ILR implantation. We included 500 patients (mean age 63.1 ± 12.7years), 304 (60.8%) were males, with previous stroke (72.6%) and TIA (27.4%). After inclusion, all were followed for at least 12months. AF was detected in 133(26.8%), 25.3% of patients with previous stroke and 29.9% with TIA (p = 0.30). AF was detected after a median of 121.5days (IQR 40.5-223.0). AF was primarily paroxysmal (95.7% after stroke, 100% after TIA). In 19.5% (n = 26) anticoagulation was not administered after AF detection. Mean CHA2DS2-VASc score (OR 1.22; CI 1.01-1.49; p = 0.037) and age > 70years (OR 0.41; CI 0.19-0.88; p = 0.0020) correlated with AF. Repeat cerebrovascular events at follow-up (28; 5.6%) correlated with AF (OR = 2.17; CI 1.00-4.72; p = 0.049). AF (p = 0.91) and recurrent events (p = 0.43) occurred similarly in females and males. In the context of cryptogenic strokes ILRs are a valuable tool for early detection of AF. In a real-world scenario, the anticoagulation therapy remains inadequate despite diagnosed AF in a relevant subset of patients requiring optimization of patient management.