Should ximelagatran replace warfarin for stroke prevention in patients with atrial fibrillation?

BACKGROUND Warfarin can successfully lower stroke risk in patients with nonvalvular atrial fibrillation (AF); however, adverse effects, drug-food interactions, and the constant dose monitoring that warfarin requires, mean that an alternative, fixed-dose, therapy is needed. OBJECTIVE The objective of this trial was to establish whether the oral, fixed-dose, direct thrombin inhibitor ximelagatran is as efficacious as dose-adjusted warfarin in preventing systemic embolism and stroke in patients with nonvalvular AF. DESIGN This North American, multicenter, double-blinded trial involved patients with persistent or paroxysmal nonvalvular AF, who had at least one of the following risk factors for stroke: previous systemic embolism; stroke or transient ischemic attack; hypertension; left-ventricular dysfunction; age over 65 years with diabetes mellitus or coronary disease; or age over 75 years. INTERVENTION Patients were randomized to receive 36 mg ximelagatran twice daily or adjusted-dose warfarin, with a target international normalized ratio of 2.0-3.0. Apart from limited aspirin and nonsteroidal anti-inflammatory drugs, all antithrombotic therapy was withheld. Patients were followed up after 1, 4 and 6 weeks, after 2, 3, 4, 5, 6, 8, 10 and 12 months and at 3-month intervals thereafter. Primary endpoints were identified as soon as possible and stroke severity was evaluated 3 months after the event. OUTCOME MEASURES The primary endpoint was stroke (either ischemic or hemorrhagic), or systemic embolism. RESULTS In total, 3,922 patients were randomized: 1,960 received ximelagatran and 1,962 received warfarin (69% male, average age 72 years). A total of 6,405 patient years were assessed and the mean follow-up time was 20 months for both groups. During the study, 88 patients had primary events: 51 patients treated with ximelagatran and 37 patients treated with warfarin. There was no significant difference between the primary endpoint incidence rate of 1.16% per year for warfarin and 1.61% for ximelagatran. When analyzed by intention to treat, ximelagatran was noninferior to warfarin within the prespecified absolute 2.0% per year margin with regard to difference in primary endpoint occurrence (P < 0.001). Major bleeding rates were similar in both groups, with total bleeding slightly lower in the ximelagatran group compared with the warfarin group (63 vs 84). Rates of adverse events were similar in the ximelagatran and warfarin patients; however, alanine aminotransferase levels three times the upper normal limit were observed in 117 (6.0%) ximelagatran patients-typically between 2-6 months of treatment initiation-compared with 15 (0.8%) warfarin patients. CONCLUSION Ximelagatran at a fixed daily dose is as efficacious as warfarin in preventing stroke in patients with nonvalvular AF The authors highlight that increased levels of serum alanine aminotransferase in patients treated with ximelagatran stresses the need for frequent monitoring and further studies to investigate the cause of hepatotoxicity.