Principal Findings:

The primary outcome, incidence of cardiovascular death, myocardial infarction, or stroke, occurred in 4.1% of the rivaroxaban plus aspirin group vs. 4.9% of the rivaroxaban alone group vs. 5.4% of the aspirin alone group (p < 0.001 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.12 for rivaroxaban alone vs. aspirin alone). The primary efficacy outcome was the same in all tested subgroups.

Secondary outcomes:

• All-cause mortality: 3.4% for rivaroxaban plus aspirin, 4.0% for rivaroxaban alone, vs. 4.1% for aspirin alone (p = 0.01 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.67 for rivaroxaban alone vs. aspirin alone)
• All stroke: 0.9% for rivaroxaban plus aspirin, 1.3% for rivaroxaban alone, vs. 1.6% for aspirin alone (p < 0.001 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.12 for rivaroxaban alone vs. aspirin alone)
• Ischemic stroke: 0.7% for rivaroxaban plus aspirin, 0.9% for rivaroxaban alone, vs. 1.4% for aspirin alone (p < 0.001 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.004 for rivaroxaban alone vs. aspirin alone)
• Hemorrhagic stroke: 0.2% for rivaroxaban plus aspirin, 0.3% for rivaroxaban alone, vs. 0.1% for aspirin alone (p = 0.33 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.005 for rivaroxaban alone vs. aspirin alone)
• Major bleeding: 3.1% for rivaroxaban plus aspirin, 2.8% for rivaroxaban alone, vs. 1.9% for aspirin alone (p < 0.001 for rivaroxaban plus aspirin vs. aspirin alone, p < 0.001 for rivaroxaban alone vs. aspirin alone)
• Cancer diagnosis: 4.0% for rivaroxaban plus aspirin, 4.0% for rivaroxaban alone vs. 3.8% for aspirin alone. In 77.1%, the timing of gastrointestinal (GI) cancer was within 6 months of a GI bleeding event, while in 88.7%, the timing of genitourinary (GU) cancer was within 6 months of a GU bleeding event.

COMPASS-PAD: Among 7,470 participants with peripheral arterial disease (PAD), 4,129 had symptomatic PAD, 1,919 had carotid disease, and 1,422 had coronary artery disease plus ankle-brachial index <0.9.

• Major adverse cardiac events (MACE): 5.0% for rivaroxaban plus aspirin, 6.0% for rivaroxaban alone, vs. 7.0% for aspirin alone (p = 0.005 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.19 for rivaroxaban alone vs. aspirin alone)
• Major adverse limb events (MALE): 1.5% for rivaroxaban plus aspirin, 1.9% for rivaroxaban alone, vs. 2.6% for aspirin alone (p = 0.01 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.07 for rivaroxaban alone vs. aspirin alone)
• Major bleeding: 3.0% for rivaroxaban plus aspirin, 3.0% for rivaroxaban alone, vs. 2.0% for aspirin alone (p = 0.009 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.004 for rivaroxaban alone vs. aspirin alone)

COMPASS-Lower Extremity PAD (n = 4,129 participants):

• MACE: 5.2% for rivaroxaban plus aspirin vs. 7.2% for aspirin alone (p < 0.05 for rivaroxaban plus aspirin vs. aspirin alone)
• MALE: 1.8% for rivaroxaban plus aspirin vs. 3.4% for aspirin alone (p < 0.05 for rivaroxaban plus aspirin vs. aspirin alone)
• Major bleeding: 3.3% for rivaroxaban plus aspirin vs. 1.9% for aspirin alone (p < 0.05 for rivaroxaban plus aspirin vs. aspirin alone)

COMPASS-CAD: There were 24,824 subjects with coronary artery disease (CAD).

• MACE: 4.0% for rivaroxaban plus aspirin, 5.0% for rivaroxaban alone, vs. 6.0% for aspirin alone (p < 0.0001 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.094 for rivaroxaban alone vs. aspirin alone)
• Major bleeding: 3.0% for rivaroxaban plus aspirin, 3.0% for rivaroxaban alone, vs. 2.0% for aspirin alone (p < 0.0001 for rivaroxaban plus aspirin vs. aspirin alone; p < 0.0001 for rivaroxaban alone vs. aspirin alone)

COMPASS-CABG: This substudy examined 1,448 COMPASS trial patients 4-14 days after CABG.

• Graft failure: 9.1% for rivaroxaban plus aspirin, 7.8% for rivaroxaban alone, vs. 8.0% for aspirin alone (p = 0.45 for rivaroxaban plus aspirin vs. aspirin alone; p = 0.75 for rivaroxaban alone vs. aspirin alone)

COMPASS-PCI: This substudy examined 9,862 COMPASS trial CAD patients with a history of percutaneous coronary intervention (PCI).

• Among those with a history of PCI, the incidence of MACE was 4.0% for rivaroxaban plus aspirin vs. 5.5% for aspirin alone.
• Among those without a history of PCI, the incidence of MACE was 4.4% for rivaroxaban plus aspirin vs. 5.7% for aspirin alone (p for interaction = 0.85).

COMPASS-Heart Failure: There were 5,902 subjects with a history of heart failure and CAD or PAD.

• MACE: 5.5% for rivaroxaban plus aspirin, 6.3% for rivaroxaban alone, vs. 7.9% for aspirin alone; rivaroxaban plus aspirin vs. aspirin alone (hazard ratio [HR] 0.68, 95% confidence interval [CI] 0.53-0.86), rivaroxaban alone vs. aspirin alone (HR 0.80, 95% CI 0.63-1.01)
• Hospitalization for heart failure: Rivaroxaban plus aspirin vs. aspirin alone was associated with greater benefit among those with heart failure vs. those without heart failure (p for interaction = 0.05)
• All-cause mortality: Rivaroxaban plus aspirin vs. aspirin alone was associated with greater benefit among those with heart failure vs. those without heart failure (p for interaction = 0.05)

COMPASS-Vascular Risk: In this analysis, subjects were classified as high-risk according to REACH scoring (2 or more vascular beds affected, history of heart failure, or renal insufficiency) or CART scoring (2 or more vascular beds affected, history of heart failure, or diabetes).

• High-risk features-REACH, MACE, acute limb ischemia, or total vascular amputation: 5.8% with rivaroxaban plus aspirin vs. 8.0% with aspirin (events prevented per 1,000 patients treated [95% CI] = 36 [21-52])
• High-risk features-CART, MACE, acute limb ischemia, or total vascular amputation: 5.6% with rivaroxaban plus aspirin vs. 7.5% with aspirin (events prevented per 1,000 patients treated [95% CI] = 33 [19-47])
• High-risk features-REACH, severe bleeding: 1.1% with rivaroxaban plus aspirin vs. 0.8% with aspirin (events caused per 1,000 patients treated [95% CI] = 3 [-4 to 9])
• High-risk features-CART, severe bleeding: 1.0% with rivaroxaban plus aspirin versus 0.8% with aspirin (events caused per 1,000 patients treated [95% CI] = 1 [-4 to 6])

COMPASS-Diabetes:

Cardiovascular death, myocardial infarction, or stroke:

• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone in diabetics = 0.74
• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone in nondiabetics = 0.77 (p for interaction = 0.77)

Major bleeding:

• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone in diabetics = 1.70
• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone in nondiabetics = 1.69 (p for interaction = 0.97)

COMPASS-Sex Differences: 22% of subjects were women and 78% were men. 

Cardiovascular death, myocardial infarction, or stroke:

• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone for women = 0.72
• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone for men = 0.76 (p for interaction = 0.75)

Major Bleeding:

• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone for women = 2.22
• Hazard ratio for rivaroxaban plus aspirin vs. aspirin alone for men = 1.60 (p for interaction = 0.19)

COMPASS-Obesity: 24% of subjects had normal body mass index (BMI), 44% were overweight, and 32% were obese.

• The primary outcome of cardiovascular death, myocardial infarction, or stroke was reduced with rivaroxaban plus aspirin vs. aspirin alone irrespective of BMI or bodyweight.
• Effects on major bleeding and net clinical benefit were also consistent and irrespective of BMI or bodyweight.

Overall cost differences (MACE): Direct cost savings of event plus procedure costs with rivaroxaban 2.5 BID plus aspirin 100 mg daily vs. aspirin 100 mg daily: mean, $682; overall, $6,144,221. Cost savings were robust across the four enrolling countries – United States, Canada, France, and Germany.

Estimation of lifetime benefit and risk: There is a wide distribution of benefit from ischemia prevention and harm from excess bleeding from the addition of rivaroxaban to aspirin.

COMPASS-Net Clinical Benefit: Net clinical benefit was defined as cardiovascular death, stroke, myocardial infarction, fatal bleeding, or symptomatic bleeding into a critical organ. Net clinical benefit was 2.5% per year in the rivaroxaban plus aspirin group vs. 3.1% per year in the aspirin alone group (HR 0.80, p = 0.0005).