Background: Atrial fibrillation (AF) can be a cause of previously diagnosed cryptogenic stroke. However, AF can be paroxysmal and asymptomatic, thereby making detection with routine ECG methods difficult. Oral anticoagulation is highly effective in reducing recurrent stroke in patients with AF, but its initiation is dependent on the detection of AF. Cryptogenic Stroke and Underlying Atrial Fibrillation (CRYSTAL AF) is the first randomized study to report the detection of AF in cryptogenic stroke patients using continuous long-term monitoring via insertable cardiac monitors (ICM).
Methods and results: Patients with prior cryptogenic stroke were randomized to control (n=220) or ICM (n=221) and followed for ≤36 months. Cumulative AF detection rates in the ICM arm increased progressively during this period (3.7%, 8.9%, 12.4%, and 30.0% at 1, 6, 12, and 36 months, respectively), but remained low in the control arm (3.0% at 36 months). This resulted in oral anticoagulation prescription in 94.7% of ICM patients with AF detected at 6 months, 96.6% at 12 months, and 90.5% at 36 months. Among ICM patients with AF detected, the median time to AF detection was 8.4 months, 81.0% of first AF episodes were asymptomatic, and 94.9% had at least 1 day with >6 minutes of AF.
Conclusions: Three-year monitoring by ICM in cryptogenic stroke patients demonstrated a significantly higher AF detection rate compared with routine care. Given the frequency of asymptomatic first episodes and the long median time to detection, these findings highlight the limitations of using traditional AF detection methods. The majority of patients with AF were prescribed oral anticoagulation therapy.
Clinical trial registration: Clinicaltrials.gov; Unique identifier: NCT00924638.
Keywords: anticoagulation; atrial fibrillation arrhythmia; diagnosis; electrophysiology; insertable cardiac monitor; stroke.
© 2016 American Heart Association, Inc.