OUP user menu

Safety and efficacy of dabigatran etexilate during catheter ablation of atrial fibrillation: a meta-analysis of the literature

Stefan H. Hohnloser, A. John Camm
DOI: http://dx.doi.org/10.1093/europace/eut241 1407-1411 First published online: 16 August 2013


Aims To summarize and analyse the published literature on the efficacy and safety of dabigatran for anticoagulation during atrial fibrillation (AF) ablation.

Methods and results A comprehensive search of the literature in the English language was conducted and 10 studies were identified which compared dabigatran therapy with warfarin. Data were extracted and meta-analysed using appropriate methodology. There was significant clinical heterogeneity among studies concerning study designs and methodologies. Most studies consisted of retrospective single-centre observational studies. A total of 3648 patients were included of whom 2241 were receiving warfarin and 1407 dabigatran. Only 12 thrombo-embolic events were reported, 3 during warfarin and 9 during dabigatran therapy [odds ratio (OR) 2.38, 95% confidence interval (CI) 0.82–6.85, P = n.s.). The pooled rate of major bleeding events was similar for warfarin and for dabigatran (OR 1.05, 95% CI 0.62–1.80; P = n.s.). Rates of minor bleeding events were also similar between the two treatment groups.

Conclusion Dabigatran has similar efficacy and safety compared with warfarin when used for periprocedural anticoagulation during AF ablation.

  • Atrial fibrillation
  • Ablation
  • Anticoagulation
  • Dabigatran
  • Novel oral anticoagulants

What's new?

  • This meta-analysis of 10 mainly observational studies assessed the efficacy and safety of peri-procedural anticoagulation using dabigatran vs warfarin in 3648 patients who underwent ablation for AF.

  • There was no significant difference in thromboembolic events, major and minor bleeding between dabigatran and warfarin.

  • The meta-analysis emphasises the need for a well-designed large-scale clinical trial to firmly establish the safety and efficacy of dabigatran and other NOACs in peri-procedural management of anticoagulation.


Catheter ablation is an established treatment modality in many patients with atrial fibrillation (AF). Atrial fibrillation ablation is technically challenging and has been shown to be associated with a definite risk of periprocedural thrombo-embolic and bleeding complications.13 The increased thrombo-embolic risk is related to the underlying prothrombotic state associated with AF, and to specific ablation-related issues such as thrombus generation at the ablation catheter or within the transseptal sheaths. Accordingly, adequate periprocedural anticoagulation is of paramount importance to protect patients.1 Optimal anticoagulation protocols are still a matter of much debate, and currently, observational and randomized controlled clinical trials favour the uninterrupted administration of warfarin at an international normalization ratio between 2 and 3 over bridging with heparin.4 At the same time, large-scale randomized trials in the general AF population have demonstrated superior efficacy and safety of novel oral anticoagulants (NOACs) such as dabigatran, rivaroxaban, and apixaban over warfarin. In addition, these NOACs are much easier to use than vitamin K antagonists. Hence, electrophysiologists are keen to examine whether these NOACs are suitable for periprocedural anticoagulation during AF ablation. Dabigatran has been available for therapy of AF for almost 3 years, and in the meantime, several studies on the efficacy and safety of this compound during AF ablation have been published.514 To overcome some of the limitations of these studies, particularly the relative small sample sizes, the present meta-analysis of all respective studies aims to provide comprehensive data on dabigatran in this clinical scenario.


Literature search strategy

A Medline search of the English literature was performed using the search terms AF, catheter ablation, periprocedural anticoagulation, and dabigatran. The authors reviewed all manuscripts, and publications were selected for this meta-analysis if they met the following criteria: (i) evaluated periprocedural anticoagulation therapy in patients undergoing AF ablation, (ii) compared warfarin therapy to treatment with dabigatran etexilate, and (iii) had been published as peer-reviewed full-size articles. Data analysis was performed using a pooling across all trials for common reported event rates and a meta-analysis using fixed-effect Mantel–Haenszel odds ratio (OR) for the overall comparative result was performed. Effect sizes from each trial were weighted by the inverse variance to calculate a weighted mean effect size, and the χ2 test was used to assess for statistical heterogeneity between studies. The effect of dabigatran vs. vitamin K antagonists treatment is presented using the OR.


Study characteristics

A total of 10 studies fulfilled the selection criteria and were selected for this meta-analysis.514 Table 1 provides details of the selected studies. Eight studies were retrospective analyses of patients who underwent AF ablation with periprocedural anticoagulation utilizing warfarin or dabigatran. One small study randomized 90 patients to warfarin or dabigatran.12 Finally, the remaining study randomized only a subset of the study population to either of the two treatments whereas the majority of patients were analysed retrospectively.13

View this table:
Table 1

Comparative studies of warfarin and dabigatran

StudyDesignPatientsWarfarin dosingDabigatran dosingTarget ACT during ablation (s)Thrombo- embolic eventsP valueMajor bleedingP valueMinor bleedingP value
Kaseno et al.6Retrospective observational211Uninterrupted110 mg BID; held 12 h prior to CA, resumed 12–15 h3000/1010/110n.s.2/1010/110n.s.11/1015/110n.s.
Snipelisky et al.5Retrospective observational156Stopped 24 h prior to CA150 mg BID held 12 h prior to CA3500/1250/31n.s.0/1250/31n.s.21/1256/31n.s.
Lakkireddy et al.7Multicentre, matched case–control290Uninterrupted150 BID; held 12 h prior to CA, resumed 3 h post-haemostasis300–4000/1453/145n.s.1/1459/1450.0198/14512/145n.s.
Kim et al.8Retrospective observational763Uninterrupted150 mg BID; held 24 h prior to CA, resumed 4 h post-haemostasis300–3500/5720/191n.s.12/5724/191n.s.19/5725/191n.s.
Imamura et al.9Prospective, observational227Heparin bridging110/150 mg BID; held 12–24 h prior to CA, resumed 3 h post-haemostasis300–3500/1261/101n.s.4/1263/101n.s.5/1265/101n.s.
Kaiser et al.10Retrospective observational257Uninterrupted or heparin bridging150 mg BID; held 24–30 h prior to CA, resumed 4–6 h post-haemostasis300–3501/1353/122n.s.1/1352/122n.s.10/1353/1220.09
Maddox et al.11Retrospective observational463Uninterrupted150 mg BID; uninterrupted350–4000/2511/212n.s.6/251a2/212an.s.n.a.n.a.n.a.
Nin et al.12Prospective, randomized90Warfarin stopped 24 h prior to CA, no bridging110 mg BID; held 25 h prior to CA, resumed after haemostasis300–4001/450/45n.s.0/450/45n.s.9/4519/450.013
Yamaji et al.13Retrospective observational, partially randomized5031 group bridging, 1 group uninterrupted110/150 mg BID; held 12 h prior to CA, resumed 3 h post-haemostasis300–3500/397b0/106n.s.4/397b0/106n.s.11/397b2/106n.s.
Bassiouny et al.14Matched case–control688Uninterrupted150 m BID; held 12–24 h prior PVI, resumed 3 h post-haemostasis350–4501/3441/344n.s.5/3444/344n.s.9/3446/344n.s.
Total36483/2241 (0.13%)9/1407 (0.64%)35/2241 (1.56%)24/1407 (1.71%)103/2241 (4.60%)63/1407 (4.48%)
  • ACT, activated clotting time; CA, catheter ablation; BID, twice daily.

  • aOnly total bleeding complications reported.

  • bBoth warfarin groups summarized.

Uninterrupted warfarin therapy was used during ablation in 7 of the 10 studies (Table). Dabigatran therapy was held 12–30 h prior to ablation and was generally resumed 3–4 h after achieving haemostasis. All studies used concomitant heparin therapy during the procedure aiming at a target activated clotting time of 300–450 s.

Outcome events

In total, outcomes in 3648 patients were reported, of whom 2241 had been treated with warfarin and 1407 with dabigatran. The primary efficacy outcomes, stroke/transient ischaemic attack (TIA) or non-central nervous system embolism, were consistent across studies. Only 12 thrombo-embolic events were observed in all studies yielding a pooled rate of 0.13% in warfarin patients and 0.63% in dabigatran patients. In a meta-analysis the Mantel–Haenszel OR resulted in 2.38 for dabigatran, but due to very wide 95% confidence intervals (CIs; 0.82–6.85) this was not statistically significant (Figure 1).

Figure 1

Thrombo-embolic events.

Major or minor bleeding events constituted safety outcomes in all studies but there was no uniform definition of these events. Major bleeding events were reported at a pooled rate of 1.56% in warfarin patients compared with 1.71% in dabigatran patients. The meta-analysis obtained a result near unity (OR 1.05; 95% CI 0.62–1.80) (Figure 2). Minor bleeding episodes were observed in 4.60% of warfarin and 4.48% of dabigatran patients (OR 0.95; 95% CI 0.67–1.35) (Figure 3 and Table 1). One study11 did not report incidences of minor bleeding events. The two treatment strategies were not statistically different from each other with respect to these safety measures.


A recent world-wide survey on safety and efficacy of AF ablation comprising 20 850 procedures reported a periprocedural stroke/TIA incidence of 0.94%.2 There is no consensus on the best anticoagulation approach to protect patients from thrombo-embolic events during ablation.1,15 However, a recent randomized international, multicentre trial has demonstrated that performing catheter ablation of AF without warfarin discontinuation significantly decreases the rate of periprocedural stroke and bleeding complications compared with a strategy of discontinuing warfarin with heparin bridging.4 Given the superior efficacy and safety of dabigatran over warfarin in the general AF population as established in the RE-LY trial,16 many AF patients are or will be receiving this NOAC. Given its ease in use, dabigatran constitutes therefore an attractive alternative to vitamin K antagonists also in the setting of AF ablation. There are already 10 clinical studies published which examine this approach. With the exception of one small prospective randomized study,13 the others are retrospective and in some instances matched case–control studies. This meta-analysis incorporating all of these studies shows that dabigatran has similar efficacy in terms of stroke prevention and similar safety with respect to major or minor bleeding events. No particular pattern of dabigatran interruption or continuation was associated with thrombo-embolic events or bleeding.

There were only 12 thrombo-embolic events among 3648 patients, 3 among patients receiving warfarin, and 9 in patients on dabigatran, for a total incidence of <1%. This is consistent with the observations made in the world-wide survey on AF ablation,2 although higher incidences have been recently reported from a randomized trial comparing warfarin discontinuation vs. uninterrupted warfarin therapy.4 Importantly, studies reviewed here followed different designs and methodologies, particularly with respect to dabigatran withdrawal prior to ablation and dabigatran dosing. Hence, the non-significant higher stroke/TIA incidence in the dabigatran vs. the warfarin group needs to be considered with caution. The overall safety profile of dabigatran is in line with a recent report from the RE-LY trial on periprocedural bleeding and thrombo-embolic events.17 This study demonstrated similar rates of periprocedural bleeding of dabigatran and warfarin, including patients having undergone urgent surgery. Finally, other recently published work on dabigatran in the setting of AF ablation confirmed its safety and effectiveness when used in a series of 259 ‘real-world’ patients.18

The current evidence seems to indicate that dabigatran can usually be stopped 12–30 h prior to AF ablation and that dabigatran therapy can be safely resumed 3–4 h after achieving haemostasis. Shorter stopping intervals before the ablation or even uninterrupted dabigatran therapy seem worthwhile to be systematically evaluated. Since dabigatran is ∼80% excreted via the kidneys, some modification of these recommendations seems appropriate in patients with impaired renal function resulting in longer interruption periods before the intervention. Dabigatran is available at two different dosages in many countries whereas only the 150 mg twice daily (BID) dose is approved in the USA, where many of the current studies were conducted. Hence, more information is needed concerning the use of the lower dabigatran dose (110 mg BID). Of note, the issue of periprocedural use of the NOACs in patients undergoing AF ablation has not been addressed in international guidelines.15


There is limited experience with dabigatran and other NOACs for peri-procedural management of anticoagulation in patients undergoing ablation for AF. Although meta-analysis of 10, mainly observational, studies found no statistically significant difference in the rates of thromboembolic events and major and minor bleeding between patients managed on dabigatran compared with warfarin, this meta-analysis has not enough power to firmly establish the efficacy and safety of dabigatran in the setting of catheter ablation of AF. This implies the need for a well-designed large-scale clinical trial to firmly establish the safety (and possibly the efficacy) of dabigatran (and other NOACs) in the setting of AF ablation.


View Abstract