Continued warfarin better than heparin bridging during device implantation


A randomised trial has shown that performing device implantation surgery without interruption of warfarin in patients with an annual risk of stroke equal to or over 5% was associated with a significant reduction in device pocket haematoma. The investigation also showed that the strategy was not associated with major perioperative bleeding and was favoured by the majority of patients.

The findings of the BRUISE CONTROL trial were presented by co-principal investigator David H Birnie, director of the Arrhythmia Service at the University of Ottawa Heart Institute, Canada, at Heart Rhythm 2013 (8–11 May, Denver, USA) and published online in The New England Journal of Medicine. The study evaluated pacemaker and implantable defibrillator surgery without interruption of oral anticoagulation-either with continued use of warfarin to or temporary cessation with heparin bridging in patients before and after device surgery.

“Between 14% and 35% of patients receiving arrhythmia devices require chronic oral anticoagulation and their periprocedural management presents a dilemma to physicians. This is particularly true for the subset of patients at moderate to high risk (≥5% per year) or thrombotic events,” Birnie told delegates at a Late Breaking Trial session at Heart Rhythm (HRS). “Current guidelines recommend interruption of warfarin and ‘bridging’ with intravenous or low-molecular-weight heparin around the time of surgery.”

However, Birnie noted that there are a number of “potential drawbacks” to bridging patients with heparin. He commented: “Bridging with heparin consumes considerable healthcare resources. It also involves a short period of normal coagulability with associated risk of thromboembolism. Case series have found a substantial risk of device pocket haematoma when heparin bridging is used.”

Although some centres have started performing device surgery without interruption of warfarin, there have been little clinical trial data to support the safety and efficacy of this approach and two small randomised trials have been inconclusive, Birnie stated. In the first study (Tolosana et al), four of 51 patients (7.8%) in the heparin-bridging arm and four of 50 (8%) in the warfarin arm developed pocket haematomas following implantation. In the second study (Cheng et al) only seven patients received heparin bridging. “We sought to resolve this dilemma with an adequately powered randomised controlled trial,” Birnie said.

The BRUISE CONTROL trial enrolled patients at 17 centres in Canada and one in Brazil. In the continued warfarin arm, the international normalised ratio (INR) on the day of surgery was targeted to be three or lower, except for patients with mechanical valves, for whom an INR of 3.5 or lower was permitted. In the heparin-bridging group, patients had warfarin discontinued five days before the procedure and started low-molecular-weight heparin or intravenous heparin three days before the procedure. Heparin was restarted 24 hours after the surgery.

Data on 668 patients were reviewed by the Data and Safety Monitoring Board (DSMB) at the second pre-specified interim analysis on 27 February 2013, at which time they recommended study termination. All randomised patients who had device surgery before that date were followed until study completion. At HRS, Birnie presented data from 681 patients-338 randomised to heparin bridging and 343 to continued warfarin-enrolled between October 2009 and February 2013. Complete follow-up was available for 325 patients in the heparin bridging group and 334 in the warfarin group.

The study found that clinically significant haematomas were 80% less frequent in patients that received uninterrupted warfarin (16% with heparin bridging vs. 3.5% with warfarin, p<0.001). Compared to the heparin arm of the trial, patients receiving warfarin had fewer haematoma prolonged hospitalisation (1.2% vs. 4.7%, p=0.006), fewer haematomas interruption of anticoagulation (3.2% vs. 14.2%, p<0.001) and fewer haematomas requiring reoperations (0.6% vs. 2.7%, p=0.03). Patients in the continued warfarin arm also had greater satisfaction with their peri-operative anticoagulation management.

In terms of secondary outcomes, Birnie noted that there were no embolic events in the heparin-bridging arm and two in the continued warfarin arm. Both patients with embolic events had non-valvular atrial fibrillation and high risk of stroke as determined by CHADS2 score. In both cases the INR had become subtherapeutic. There was one episode of cardiac temponade requiring pericardiocentesis in the heparin bridging group. There were six (1.8%) device system infections in the heparin bridging arm and two (0.6%) in the continued warfarin group (p=0.17). All eight infections required complete system extraction.

Birnie commented: “These results suggest that continuation of warfarin during pacemaker or defibrillator surgery may be preferable to heparin bridging, at least for patients like those enrolled in our trial.”

Co-principal investigator Vidal Essebag, director of Cardiac Electrophysiology, McGill University Health Center, Montreal, Canada, stated, “To many, the substantial reduction in pocket haematoma that we observed with continued warfarin may be counterintuitive. One explanation that has been proposed is the concept of an ‘anticoagulant stress test.’ That is, if patients undergo surgery while fully anticoagulated, any excessive bleeding will be detectable and appropriately managed while the wound is still open. In contrast, when surgery is performed with heparin bridging, such bleeding may remain latent, and appear only when full anticoagulation is resumed postoperatively.”

The BRUISE CONTROL trial was funded by the Canadian Institutes of Health Research (CIHR).