The authors of this study note that periprocedural thromboembolic events (TE) are one of the most worrisome complications of radiofrequency catheter ablation (CA) for atrial fibrillation, because of the potential long-term effect on functionality. Although ablation procedures performed without warfarin discontinuation seem to be associated with lower TE risk compared to strategies of warfarin discontinuation, no randomised controlled trial currently exists. Their research therefore sought to examine this issue further.
In the COMPARE study, 1584 adults with INR 2.0-3.0 in the four weeks prior to CA and a CHADS score ≥1, with no known bleeding disorders were randomised to a strategy of discontinued warfarin or continuous warfarin. For those randomised to the warfarin discontinuation group, warfarin was stopped 2-3 days prior to ablation and they were bridged with enoxaparin (1mg/kg twice daily prior to ablation, then enoxaparin 0.5mg/kg twice daily after the procedure). Warfarin was restarted the night of the procedure and the enoxaparin stopped when the INR >2.
The incidence of TE events (stroke, TIA or systemic thromboembolism) during the 48 hours post-ablation (primary endpoint) was 4.9% in the warfarin discontinuation group (3.7% stroke; 1.3% TIA) and 0.25% in the continuous warfarin group. The unadjusted relative risk is presented [no adjusted RR is stated] – this was 0.051 (95% CI 0.012 to 0.211). The majority of the TE events occurred in patients with long-standing persistent AF.
There were no statistically significant differences in the incidence of major bleeding complications (0.76% of the warfarin discontinuation group and 0.38% of the continuous warfarin group); minor bleeding complications were however more frequent in the discontinuation group (22% v 4.1%; p<0.001).
The authors note that the heparin bolus given prior to the procedure was larger in the ‘off’ warfarin group; activated coagulation time was however kept at a high level in both arms and it is therefore unlikely that this was responsible for the higher bleeding rate in this group – they suggest instead that this should be attributed to the use of LMWH.
They go on to discuss the limitations of their study, including the lack of blinding of operators to anticoagulation management. They suggest that their results not be extrapolated to the newer oral anticoagulants, and that further research should address their role in high-risk patients and compare them to ‘on’ warfarin treatment.