Abstract

BACKGROUND: Short and rare episodes of atrial fibrillation (AF) are commonly detected using implanted devices (device-detected AF) in patients with prior stroke or transient ischemic attack (TIA). The effectiveness and safety of oral anticoagulation in patients with prior stroke or TIA and device-detected AF but with no ECG-documented AF is unclear. METHODS AND RESULTS: This prespecified analysis of the NOAH-AFNET 6 (Non-Vitamin K Antagonist Oral Anticoagulants in Patients With Atrial High Rate Episodes) trial with post hoc elements assessed the effect of oral anticoagulation in patients with device-detected AF with and without a prior stroke or TIA in the randomized, double-blind, double-dummy NOAH-AFNET 6 trial. Outcomes were stroke, systemic embolism, and cardiovascular death (primary outcome) and major bleeding and death (safety outcome). A prior stroke or TIA was found in 253 patients with device-detected AF randomized in the NOAH-AFNET 6 (mean age, 78 years; 36.4% women). There was no treatment interaction with prior stroke or TIA for any of the primary and secondary time-to-event outcomes. In patients with a prior stroke or TIA, 14 out of 122 patients experienced a primary outcome event with anticoagulation (5.7% per patient-year). Without anticoagulation, there were 16 out of 131 patients with an event (6.3% per patient-year). The rate of stroke was lower than expected (anticoagulation: 4 out of 122 [1.6% per patient-year]; no anticoagulation: 6 out of 131 [2.3% per patient-year]). Numerically, there were more major bleeding events with anticoagulation in patients with prior stroke or TIA (8 out of 122 patients) than without anticoagulation (2 out of 131 patients). CONCLUSIONS: Anticoagulation appears to have ambiguous effects in patients with device-detected AF and a prior stroke or TIA in this hypothesis-generating analysis of the NOAH-AFNET 6 in the absence of ECG-documented AF, partially due to a low rate of stroke without anticoagulation.