Retain Warfarin During AF Ablation for Fewer Strokes: Randomized Trial

April 23, 2014

DALLAS, TX — Maintaining warfarin anticoagulation is safer than interrupting it for catheter-ablation procedures for atrial fibrillation, concludes a randomized study with >1500 patients at elevated risk for thromboembolic stroke[1].

In the study, patients without periprocedural warfarin, who were bridged with low-molecular weight heparin in standard fashion, showed a >10-fold increased odds of ischemic stroke or transient ischemic attack (TIA) in the 48 hours after ablation, compared with those on uninterrupted warfarin. Those who stayed on the oral anticoagulant did not have more major or minor bleeding; in fact, their risk of minor bleeding was significantly less.

Bridging patients with parenteral anticoagulation, typically enoxaparin and heparin, was long the standard of care in warfarin-managed patients going to AF ablation, although based more on expert consensus than solid data, according to Dr Andrea Natale (Texas Cardiac Arrhythmia Institute at St David's Medical Center, Austin), who is principal investigator of the current trial, called COMPARE . But most of the high-volume centers in the US have recently been moving toward the uninterrupted-warfarin approach, as observational data have emerged more and more in its favor, he told heartwire . It's becoming more common in Europe as well, he said, though more slowly.

Natale said his group initially tried cutting the enoxaparin dosage while weaning patients off warfarin to address the long-recognized risk of access-site bleeding; the current study took that step further to test a no-bridging strategy with a focus on maintaining a therapeutic INR throughout the procedure. The results, which Natale said provide for the first time a randomized-trial evidence base for the continued-warfarin approach, were published April 17, 2014 in Circulation with lead author Dr Luigi Di Biase (Texas Cardiac Arrhythmia Institute).

One of its key messages, according to Natale: the greater protection provided by uninterrupted warfarin in the setting of AF ablation may be most important for patients with chronic AF and others at especially increased risk of thromboembolic events.

COMPARE prospectively randomized 1584 patients slated for radiofrequency catheter ablation of paroxysmal, persistent, or "longstanding persistent" AF who had a CHADS2 score of >1; patients also must have had an INR of 2.0 to 3.0 on warfarin in the "three to four weeks" prior to ablation. One group received standard periprocedural anticoagulation consisting of warfarin suspension two to three days before ablation with bridging using subcutaneous enoxaparin followed by IV unfractionated heparin, with the process reversed after ablation. Warfarin was continued uninterrupted in the other group. Both groups received an added heparin bolus before the transseptal puncture required of the cath procedure.

Events Within 48 Hours After Catheter Ablation for Atrial Fibrillation, by Periprocedural Anticoagulation Strategy

End point Warfarin withdrawn, n=790 (%) Warfarin maintained, n=794 (%) p
Stroke or TIA* 4.9 0.25 <0.001
Stroke 3.7 0.25 <0.001
TIA 1.3 0.0 <0.001
Major bleeding 0.76 0.38 0.31
Minor bleeding 22 4.1 <0.001
*Primary end point
TIA=transient ischemic attack

Of the trial's 41 strokes or TIAs, 85% were in patients undergoing ablation for longstanding persistent AF, although such patients made up only about half the population. The patients with longstanding persistent AF included the only two thromboembolic events, both strokes, in the uninterrupted-warfarin group.

Significant multivariate predictors of thromboembolic events included being female (p=0.03), CHADS2 >2 (p<0.001), and longstanding persistent AF (p<0.001), as well as warfarin discontinuation, for which the odds ratio was 13 (95% CI 3.1–55.6, p<0.001).

It makes sense that the benefit of an uninterrupted-warfarin strategy would be greater in patients with chronic AF, Natale observed, because their AF requires longer, more intense ablation. The more ablation, the greater the periprocedural thromboembolic risk; and the more risk, the more the apparent benefit. Similarly, there was more apparent benefit from maintaining warfarin in patients with CHADS2 >2.

Natale cautions against extrapolating the COMPARE findings to the "new oral anticoagulants." Dabigatran (Pradaxa, Boehringer Ingelheim), in particular, has been explored in the setting of AF ablation without parenteral anticoagulation, with favorable results. But many of those studies include predominantly patients with paroxysmal AF, he said. "That really doesn't give you a test of [the oral anticoagulant's] real protection. Because the risk of thromboembolic events in the paroxysmal patient is so small, you'd need thousands and thousands of patients to be able to say it is safe."

Therefore, "If you test the new anticoagulants in low-risk patients, you might wrongly believe that the drug is safe. It's important to realize that the newer anticoagulants should be tested in the high-risk patient before we establish that they are safe enough as an alternative to warfarin."

Di Biase discloses serving as consultant for Hansen Medical, Biosense Webster, and St Jude Medical. Natale discloses receiving honoraria for speakers' bureau service for St Jude Medical, Boston Scientific, Medtronic, and Biosense Webster and consulting for Biosense Webster and St Jude Medical. Disclosures for the coauthors are listed in the article.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.