Abstract
Introduction
The direct oral anticoagulants (DOACS) such as dabigatran (D), rivaroxaban (R), apixaban (A), and edoxaban (E) have become the leading anticoagulation strategy in atrial fibrillation (AF). Many studies have shown that the safety of DOACS is similar or better than warfarin (W) periprocedural pre- and post AF ablation. It is unclear if they fare the same compared to vitamin K antagonists (VKA) other than W, if individual DOACS have the same benefits, or if interrupted or uninterrupted regimens are better. Since complication rates of AF ablation are low, individual studies preclude an ability to answer this issue with certainty, and a meta-analysis would provide a better estimation.
Methods
We have evaluated all published manuscripts (n=53 enrolling 31,589 patients) that directly compare peri-AF ablation complication rates in patients receiving DOACS vs. VKA. Bleeding (vascular complications, pericardial effusion, or other bleeding) and embolic complications were the primary endpoints. Additionally, individual types of DOACS and VKA, interrupted and uninterrupted regimens, were compared. We used the Mantel-Haenszel random effect model to pool the study results, with a random effects model for heterogeneous samples/results.
Results
The 14270 patients on a DOACS (6943 D, 4269 R, 2457 A, 495 E) were similar to the 17319 patients on VKA. DOACS were either uninterrupted or minimally interrupted, and an uninterrupted VKA (UVKA) regimen was used in 13794 (80%) patients. Composite bleeding rates were significantly lower in patients treated with DOACS (6.6%) compared to VKA (8%) (OR 0.80, 95% CI 0.69-0.93; I2 = 45%, p = 0.003), whereas the composite embolic rates were similar in both groups (0.38% in DOACS vs 0.38% in VKA, OR 0.76, 95% CI 0.53-1.1; I2 = 0%, p = 0.15). The interrupted DOACS regimen had lower bleeding events than uninterrupted VKA (OR 0.71, CI 0.57-0.89, I2 = 34%, p = 0.02), with no difference in thromboembolic complications. D was associated with lower bleeding complications than UVKA (OR 0.81, 95%, CI 0.66-0.98, I2 = 32%, p = 0.03), while R and A were similar to UVKA.
Conclusions
This meta-analysis demonstrates that DOACS should be preferred to VKA for peri-AF ablation anticoagulation. Minimally interrupted DOACS demonstrate lower bleeding complications with similar risk for thromboembolism. Future studies should determine the optimal time to stop and restart of the DOACS relative to the ablation.