Principal Findings:

A total of 15,526 patients were enrolled at 766 centers in 44 countries (5.5% in North America), 5,174 to rivaroxaban 2.5 mg, 5,176 to rivaroxaban 5 mg, and 5,176 to placebo. The mean duration of treatment with a study drug was 13.1 months. Baseline characteristics were similar between the two arms. About 37% of the patients were ≥65 years, 73% were White, 32% had diabetes mellitus, and 27% had a prior history of myocardial infarction (MI). The median body mass was 78 kg, with a median creatinine clearance of 85 ml/min. The index diagnosis was ST-segment elevation MI (STEMI) in 50%, non-STEMI in 26%, and unstable angina in 24%. About 61% underwent revascularization with either percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG).

Over a median follow-up of about 2 years, the primary endpoint of cardiovascular (CV) death/MI/stroke was significantly lower in the rivaroxaban arm, as compared with placebo (8.9% vs. 10.7%, hazard ratio [HR] 0.84, 95% confidence interval [CI] 0.74-0.96, p = 0.008), and among STEMI patients, the primary outcome was 8.4% vs. 10.6% (p = 0.019). The benefit on the primary endpoint was seen in both the 5 mg dose and the 2.5 mg dose individually (see below). Individual components including CV mortality (3.3% vs. 4.1%, p = 0.04) and MI (5.5% vs. 6.6%, p = 0.047) were also reduced; rates of stroke were similar (1.6% vs. 1.2%, p = 0.25). Stent thrombosis rates (2.3% vs. 2.9%, p = 0.02) were lower. The results were generally consistent across the subgroups tested, with the possible exception of a small subgroup of patients with previous stroke or transient ischemic attack. In the subgroup that underwent PCI (n = 9,631), ARC definite and probable stent thrombosis was significantly reduced with rivaroxaban (1.5% vs. 1.9%, p = 0.017).

The primary safety endpoint of non–CABG-related TIMI (Thrombolysis in Myocardial Infarction) major bleeding was significantly elevated in the rivaroxaban arm, as compared with placebo (2.1% vs. 0.6%, HR 3.96, 95% CI 2.46-6.38, p < 0.001). This major bleeding finding was true among all patients and also those with STEMI. Rates of TIMI minor bleeding (1.3% vs. 0.5%, p = 0.003) and intracranial hemorrhage (ICH) (0.6% vs. 0.2%, p = 0.0009) were also higher in the rivaroxaban arm. Bleeding appeared to be higher after 6 months, with numerically higher rates in patients with lower body weight and older age.

When analyzing the benefit by the dose of rivaroxaban, the greatest efficacy for reduction in ischemic endpoints was noted with the 2.5 mg daily dose. The primary endpoint was significantly lower in this arm as compared with placebo (9.1% vs. 10.7%, HR 0.84, 95% CI 0.72-0.97, p = 0.02). CV death was reduced by 34% (2.7% vs. 4.1%, p = 0.002) and all-cause mortality by 32% (2.9% vs. 4.5%, p = 0.002). In comparison, while the 5 mg daily dose was associated with a reduction in the primary endpoint (8.8% vs. 10.7%, p = 0.03), no reductions were noted in CV mortality (p = 0.63) or all-cause mortality (p = 0.66) as compared with placebo. Both doses had similar increases in non-CABG TIMI major bleeding as compared with placebo (1.8% vs. 0.6%, and 2.4% vs. 0.6%, respectively). However, the 2.5 mg dose had lower rates of TIMI minor bleeding as compared with the 5 mg dose (0.9% vs. 1.6%, p = 0.046), but not ICH (0.4% vs. 0.7%).

Premature drug discontinuation was noted in 26.9% of patients in the 2.5 mg daily arm, 29.4% in the 5 mg daily arm, and 26.4% in the placebo arm. About 94% of patients were compliant with the study drug at least 85% of the time. Alanine aminotransferase >3x the upper limit of normal occurred in 0.2% of patients in each arm.