Mr. Wilson, age 48, has severe aortic stenosis, and needs an aortic valve replacement. He finds no reason to quarrel with his doctor’s recommendation for a mechanical valve, understanding that it is more durable than a bioprosthetic valve. But when the doctor explains that he will need lifelong anticoagulation with warfarin, and then details the specifics of warfarin treatment, Mr. Wilson balks. “I don’t want to take warfarin” he tells his doctor. “Aren’t there any other choices?”
Mr. Wilson’s reaction is not uncommon. Warfarin, a vitamin K antagonist, has been used successfully in patients with mechanical heart valves to protect against thromboembolic complications, yet it has well known limitations, including the risk of serious bleeding, multiple food and drug interactions, and a requirement for lifelong international normalized ratio (INR) monitoring. The recent development of novel oral anticoagulants has thus been welcomed.
One such anticoagulant is dabigatran etexilate, a direct thrombin inhibitor. Dabigatran received FDA approval in October 2010 for use in patients with atrial fibrillation, after the RE-LY trial showed that it is a clinically acceptable alternative to warfarin in this setting. When early animal studies showed promising results for dabigatran in preventing valve thrombosis, it was hoped that dabigatran might replace warfarin for patients with mechanical heart valves. But that hope has dimmed.
In this week’s NEJM, Eikelboom et al report the results of RE-ALIGN, a phase 2, prospective, open-label study designed to test a dagibatran dosing regimen for prevention of thromboembolic complications in patients with mechanical valves. The trial included patients in the immediate post-operative period (nearly 80% were in this group), as well as those who had undergone valve replacement at least three months earlier. Eligible participants were randomized to receive either dabigatran or warfarin for 12 weeks.
The dabigatran dosing algorithm was based primarily on pharmacokinetic data from the RE-LY trial, and assumed that the risk of thromboembolism in patients with mechanical heart valves was similar to that of patients with atrial fibrillation. Starting doses were determined by creatinine clearance, and trough plasma levels of dabigatran were measured at pre-specified intervals. The target minimum trough plasma level was 50 ng/ml, as the RE-LY trial had shown increased thromboembolic risk below that threshold. Appropriate INR targets were established for patients randomized to receive warfarin. The primary outcome of the study was trough plasma concentration of dabigatran.
However, after 252 patients had been enrolled, an independent data safety monitoring board noted an excess of thromboembolic and bleeding complications in the dabigatran group, and the trial was terminated prematurely. The composite of stroke, transient ischemic attack, systemic embolism, myocardial infarction or death occurred in 15 patients (9%) randomized to dabigatran, compared to 4 patients (5%) randomized to warfarin. Major bleeding occurred in 7 patients (4%) in the dabigatran group and 2 patients (2%) in the warfarin group. There appeared to be no relationship between plasma levels of dabigatran and thromboembolic or bleeding events.
The study authors raise the possibility that a different dosing regimen, one that produced a higher trough but lower peak, might have produced a better outcome. They also note that multiple coagulation pathways contribute to thrombus formation in patients with mechanical valves. Differences in mechanism of action, they claim – warfarin inhibits the synthesis of coagulation factors VII, IX, X and thrombin, while dabigatran inhibits only thrombin – likely made warfarin more effective than dabigatran in this clinical setting.
In an accompanying editorial, Dr. Hylek of Boston University School of Medicine raises several issues potentially contributing to the RE-ALIGN results, including the drawbacks of testing a fixed-dose regimen of dabigatran in the early postoperative period, when frequent individualized dose adjustments of warfarin are often necessary, as well as the limitations of transferring pharmacokinetic and dose information from one clinical setting to another.
NEJM Deputy Editor John Jarcho emphasizes that the message from the RE-ALIGN trial is clear, namely, that “given the current evidence, dabigatran and other novel anticoagulants should not be used in patients with mechanical valves.”