Atrial fibrillation (AF) catheter ablation has been given a Class I recommendation (2012 AHA/ACC guidelines), for patients with symptomatic paroxysmal atrial fibrillation who are intolerant or refractory to at least one antiarrhythmic agent.¹ Traditionally, patients undergoing catheter ablation for AF had anticoagulation (Vitamin K antagonist [VKA] e.g. warfarin) discontinued 3 to 5 days prior to the procedure and were bridged with heparin or a low molecular weight heparin after the procedure until the patients were therapeutically anticoagulated with warfarin. However, there has been an increasing trend of the uninterrupted anticoagulation with VKA in patients undergoing AF ablation. In a recent meta-analysis, Nairooz et al² evaluated the role of uninterrupted VKA versus interrupted VKA with heparin bridging in AF ablation. They included 13 studies (1 randomized and 12 observational) with 17, 434 patients undergoing AF ablation, of which 7877 patients managed with uninterrupted approach and 9557 with interrupted approach. There was a significantly lower incidence of major bleeding (OR 0.72, 95% CI 0.54-0.95; p=0.02), minor bleeding (OR 0.33, 95% CI 0.21-0.52; p<0.0001), combined stroke and transient ischemic attack (OR 0.25, 95% CI 0.10-0.62; p=0.003) with uninterrupted VKA as opposed to an interrupted VKA and bridging heparin/enoxaparin strategy. Similarly, with advent of newer oral anticoagulant (NOAC’s)(especially dabigatran, rivaroxaban and apixaban, numerous studies have demonstrated thromboembolic safety profile in patients not only with non-valvular AF but also undergoing AF ablation with lower bleeding rates. In light of these findings, 3 studies have been conducted so far (1 retrospective and 2 prospective) assessing the safety and efficacy of uninterrupted apixaban in patients undergoing AF ablation as compared to VKA.^3-5 In view of these studies, we aim to perform a meta-analysis to assess for safety uninterrupted periprocedural apixaban in patients undergoing AF ablation.

We searched PubMed, The Cochrane Library, EMBASE, EBSCO, Web of Science and CINAHL databases from the inception through October 15, 2015 comparing uninterrupted apixaban with uninterrupted VKA peri AF ablation. All endpoints were reported through the longest follow up available. The primary safety outcome in our study was a composite of thromboembolic event and overall bleeding. Individual outcomes assessed in our study were thromboembolic event, overall bleeding, major and minor bleeding, groin hematomas and pericardial effusion. Random effects model was followed to estimate the odds ratios (OR) and respective 95% confidence intervals (CI) using Cochrane Collaborative software, RevMan 5.3. Measure of heterogeneity between the studies was assessed using the chi square test and was considered significant if p values < 0.10 or I2 > 50%.

Three trials (2 prospective, 1 retrospective) with a total of 1,057 patients were included in the analysis, of those 410 on uninterrupted apixaban versus 647 uninterrupted VKA. Characteristics of included studies and periprocedural anticoagulation strategy are described in table 1. Zero thromboembolic events were reported in the apixaban group versus 1 in the VKA group with no statistical difference (OR 0.75; 95% CI 0.03-18.49). No statistical significant difference in the primary outcome (OR 0.92; 95% CI 0.54-1.55) was observed in our study. Consistently, no difference was observed in risk of overall bleeding (OR 0.94, 95% CI 0.55- 1.58), major bleeding (OR1.37; 95% CI 0.33-5.67), minor bleeding (OR 0.89; 95% CI 0.50-1.55), pericardial effusion (OR 0.50; 95% CI 0.18-1.38) and groin hematoma (OR 1.36; 95% CI 0.70-2.65) in the apixaban group as compared to VKA group (figure 1).

Figure 1 Forest plot of primary outcome, thromboembolic events, overall bleeding events, groin hematoma, pericardial effusion in patients undergoing ablation of atrial fibrillation with uninterrupted apixaban compared to uninterrupted vitamin K antagonist

---

---

This is the first meta-analysis of the currently available literature comparing uninterrupted apixaban to uninterrupted VKA in patients undergoing AF ablation. Among 1,047 patients who underwent AF ablation, we demonstrated that there was no difference in the primary outcome (i.e. composite of thromboembolic events and bleeding) in either group. Although non significant there were no thromboembolic events in the apixaban group as compared one event in the VKA group. Interestingly there was also no significant difference observed in the secondary endpoints of major bleeding, minor bleeding, pericardial effusion or groin hematoma in either strategy. Interestingly, the periprocedural activated clotting time in VKA group was significantly higher in all 3 studies: 258±26 versus 288±34, p<0.001;⁵ 342.1±23.1 versus 363.1±26.5, p<0.001;⁴ 275±54 versus 313±47, p<0.001,³ apixaban and VKA groups respectively. However no difference in the bleeding events were observed in the both groups.

Based on our study, apixaban was the non-inferior to uninterrupted VKA in patients undergoing AF ablation. In a recent meta-analyses by Lu et al assessing the safety and efficacy of apixaban in patients undergoing atrial fibrillation ablation, apixaban was as effective as VKA (that corroborates with our study), but our study is essentially different from this meta-analyses as we only included trials in which apixaban was used in uninterrupted fashion, and excluded all trials where apixaban was either stopped a night before or started on postprocedure day 1 (as included in Lu et al meta-analysis).⁶

Also other NOAC agents that have been compared to VKA include dabigatran and rivaroxaban. There has been conflicting data with Steinberg et al⁷ and Sardar et al⁸ demonstrating an increase in neurologic complications with dabigatran compared to VKA. This is contrary to the meta-analysis conducted by Honhloser et al⁹ and Providencia et al¹⁰ who demonstrated no significant differences in the composite of neurologic and bleeding complications between the dabigatran and uninterrupted VKA groups. Rivaroxaban has even lesser data and no meta-analysis to compare the uninterrupted Rivaroxaban against uninterrupted VKA.

Given the lack of literature, a randomized controlled trial AFAXA (Apixaban During Atrial Fibrillation Catheter Ablation: Comparison to Vitamin K Antagonist Therapy; NCT02227550) will provide valuable information as the study compares fixed dose apixaban (5 mg two times a day) with uninterrupted VKA (goal INR 2.0-3.0) in AF ablation.¹¹ The primary outcome of the study is composite of allcause mortality, stroke (ischemic stroke, subarachnoid hemorrhage and hemorrhagic stroke), and major bleeding events with follow-up period of up to 4 months.

One of the major limitations of the current meta-analysis includes reliance on retrospective and prospective and lack on any randomized controlled trials. In the trial by Kaess et al,⁵ irrespective of the patient’s home apixaban dose or renal function, a fixed dose of 2.5 mg was administered to apixaban group, which could potentially result in decreased bleeding events. Despite, heterogeneity in trials design, study protocols, and baseline characteristics of study population, the test of heterogeneity was non significant.

In this meta-analysis of patients undergoing AF ablation, apixaban was safe and non-inferior to uninterrupted VKA without any increase the risk of thromboembolic event, major and minor bleeding. Although the results from randomized controlled trials are pending, our study demonstrates that safety and efficacy of apixaban in comparison to uninterrupted VKA, and hence supports its use in AF ablation.

None.