Warfarin continued in the periprocedural period yielded lower rates of stroke, TIA, and minor bleeding than use of heparin to bridge catheter ablation surgery for AF.
Results of multiple randomized studies, most recently the BRUISE CONTROL study1, have shown that uninterrupted anticoagulation with warfarin during pacemaker or defibrillator surgery decreases the risk of bleeding associated with bridging therapy with heparin. Observational data suggest that uninterrupted anticoagulation may also be beneficial to avoid ischemic endpoints in persons at high risk for thromboembolic events. However, a recent randomized controlled trial (RCT) of more than 1500 patients from Natale and colleagues2 demonstrated for the first time that there is a decreased risk of thrombotic complications (notably ischemic stroke) with uninterrupted anticoagulation in the periprocedural setting during atrial fibrillation (AF) catheter ablation.
In the COMPARE trial, published in Circulation in early 2014, 1584 patients with AF and a CHADS2 score >1 were randomized 1:1 prospectively and in an open-label fashion to either bridging with heparin therapy (low molecular weight heparin [LMWH] followed by unfractionated heparin [UFH] before the ablation procedure and UFH followed by LMWH after the procedure) or to uninterrupted warfarin therapy. All patients had documented INRs between 2 and 3 for at least 3 to 4 weeks prior to ablation. All patients also received intraprocedural heparin before the transseptal puncture needed for the ablation procedure.
The results were unequivocal-there was a much higher rate of stroke, TIA, or systemic thromboembolic event (the primary endpoint) in the 48 hours following the procedure in the bridging group as compared to the warfarin group (4.9% vs. 0.25%, p<.001), with the stroke and TIA endpoints each being statistically significant. There was also a significantly lower rate of minor bleeding in the warfarin group (4% vs. 22%, p<.001) but no difference was observed in major bleeding, which was quite low in both arms (0.38% vs. 0.76%, p=0.31). Notably, the neurologists and physicians who were ascertaining the primary endpoint were blinded to the treatment assignment.
The COMPARE results provide another example of the safety and efficacy of uninterrupted periprocedural warfarin in patients who have a high thromboembolic risk. The results are encouraging because switching from one anticoagulant to another can create periods of over- or underanticoagulation, which can predispose to bleeding or thromboembolism.
Although it is an open-label study, this trial is likely to be another “game changer” given its fairly large sample size and randomized design and will likely influence practice. An important caveat is important here: Despite these impressive results, the findings cannot be extrapolated to the novel oral anticoagulants during such procedures.
As the level of evidence accumulates that procedures such as AF ablation and pacemaker surgeries can be performed safely on therapeutic anticoagulation, perhaps this will open the door for other subspecialists to feel more comfortable about performing minor procedures (eg, screening colonoscopies) using uninterrupted anticoagulation.