Open Access

Stefan H. Hohnloser, Alan John Camm, Riccardo Cappato, Hans-Christoph Diener, Hein Heidbuchel, Lluís Mont, Carlos A Morillo, Hans-Joachim Lanz, Heiko Rauer, Paul-Egbert Reimitz, Rüdiger Smolnik, Josef Kautzner

• Aims: This post hoc analysis of ELIMINATE-AF evaluated requirements of unfractionated heparin (UFH) and procedure-related bleeding in atrial fibrillation (AF) patients undergoing ablation with uninterrupted edoxaban or vitamin K antagonist (VKA) therapy. Methods and results: Patients were randomized 2:1 to once-daily edoxaban 60 mg (or dose-reduced 30 mg) or dose-adjusted VKA (target international normalized ratio: 2.0–3.0). Uninterrupted anticoagulation was mandated for 21–28 days’ pre-ablation and 90 days’ post-ablation. During ablation, UFH administration targeted an activated clotting time (ACT) of 300–400 s. Periprocedural bleeding was differentiated between procedure-related (bleeding at puncture side, cardiac tamponade) and unrelated events. Of 614 randomized patients, 553 received study drug and underwent catheter ablation (edoxaban n = 375; VKA n = 178). The median (Q1–Q3) time from last dose to ablation procedure was 14.8 (13.3–16.5) vs. 16.5 (14.8–19.5) h (edoxaban vs. VKA group, respectively). Mean ACT (SD) ≥300 s was observed in 52% edoxaban- vs. 76% VKA-treated patients, despite a higher mean (SD) UFH dose in the edoxaban vs. VKA group [14 261 (6397) IU vs. 11 473 (4300) IU; exploratory P-value < 0.0001]. In the edoxaban group, 13 patients (3.5%) had procedure-related bleeds of whom 9 had received an UFH dose above the median (13 000 IU). In the VKA arm, 7 patients (3.9%) had procedure-related bleeds of whom 3 had received an UFH dose above the median (10 225 IU). Conclusion: The rate of procedure-related major/clinically relevant non-major bleeding did not differ between the treatment arms despite higher doses of UFH used with edoxaban vs. VKA to achieve a target ACT during AF ablation.