The COMPASS (Cardiovascular Outcomes for People Using Anticoagulation Strategies) trial results were highly anticipated; the results were commensurate with expectations set by the news of its premature termination for 'overwhelming efficacy.' So what have we learned from this important study?

The dosage of 2.5 mg twice-daily dose of rivaroxaban, lower than the one currently approved for stroke prevention in atrial fibrillation, may have hit the sweet spot. It balances efficacy and safety by favorably affecting the residual risk in patients with established stable cardiovascular disease on an appropriate background medical regimen.

COMPASS investigators randomized 27,395 patients from 602 centers in 33 countries to one of the three treatments arms following a run-in phase: a) rivaroxaban 2.5 mg twice daily with aspirin 100 mg once daily; b) rivaroxaban 5 mg twice daily alone; and c) only aspirin 100 mg once daily.1 The primary outcome was a composite of cardiovascular death, stroke, or nonfatal myocardial infarction. About 90% participants had known coronary artery disease and 27% history of peripheral artery disease. Randomization was stratified by center and use of proton pump inhibitor (PPI) at baseline. Eligible patients were also randomized to PPI or matching placebo (separate trial, results awaited).

The 30-day run-in phase exposed participants to rivaroxaban placebo (twice daily) and once a day aspirin, off all non-study anticoagulants or aspirin therapy.2 This ensured medication compliance and absence of aspirin associated bleeding. Approximately 8% of the 28,275 non-coronary artery bypass grafting participants included in the study were excluded during this run-in phase (~70% for adherence concerns and ~5% for adverse events that included 23 deaths and 3 major bleeding-related exclusions). Overall, there were 25 participants who had major bleeds during the run-in phase.2 Skeptics often raise these points as a testament of selection bias and decreased generalizability of the trial findings. In my opinion, they simply encourage us to carefully assess our patients' compliance with existing medications prior to adding a new and probably more expensive twice a day drug.

A mean follow-up of 23 months was completed, prior to an early termination of this part of the trial by an independent DSMB for superiority of the rivaroxaban and aspirin arm. The primary outcome was significantly lower in the rivaroxaban plus aspirin group compared with the aspirin-only group (4.1% vs. 5.4%; HR = 0.76, 95% CI: 0.66–0.86; p < 0.001).1 This translates to an absolute risk reduction of 1.3%, a relative risk reduction of 24% and number needed to treat of 76. Although lower, this endpoint was not statistically significant with rivaroxaban alone compared with aspirin (4.9% vs. 5.4%; HR = 0.90, 95% CI: 0.79–1.03; p = 0.12). All-cause mortality with the rivaroxaban and aspirin combination was reduced by 0.7% compared to aspiring alone (HR = 0.82, 95% CI: 0.71–0.66; p = 0.01). All-cause death rates were 3.4%, 4.0%, and 4.1% in the rivaroxaban plus aspirin, rivaroxaban and aspirin-only groups, respectively. Statistically significant reductions in CV mortality and coronary heart disease mortality were present. A symmetric reduction in each component of the primary outcome measure was also seen. The secondary endpoints, comprising of ischemic stroke, MI, acute limb ischemia, and CV death or coronary heart disease death, were lower on rivaroxaban plus aspirin regimen compared with aspirin alone, although all failed to reach statistical significance after adjustment for multiple comparisons. This issue of multiplicity of inferences, although present in most clinical trials, requires effective handling to avoid unsubstantiated claim(s) for effectiveness of a drug therapy that could arise due to inflated false positive conclusions. Methods to control for this type I error rate need to be employed. Triggered by an FDA letter, in August 2014 the multiple testing strategy was revised (from Dunnett procedure to Hochberg test) to ensure the control of the family-wise type I error for testing of both primary and secondary efficacy variables. The threshold p value using the Hochberg procedure for each of the above comparisons was 0.0025.

Major bleeding rate was significantly higher in the rivaroxaban plus aspirin arm than in the aspirin alone group (3.1% vs. 1.9%; HR = 1.70, 95% CI: 1.40–2.05; p < 0.001).1 The same was observed in the rivaroxaban only arm compared to aspirin alone (2.8% vs. 1.9%; HR = 1.51, 95% CI: 1.25–1.84; p < 0.001). Most excess bleeds were gastrointestinal. However, the rate of the 'net clinical benefit,' comprising of CV death, stroke, MI, fatal or symptomatic bleeding into a critical organ, was 4.7% on rivaroxaban plus aspirin and 5.9% on aspirin alone (HR = 0.80, 95% CI: 0.70–0.91; p < 0.001). It was not significantly lowered with rivaroxaban alone (5.5%).

It could be claimed that early termination of the trial may overestimate the treatment effect and underestimate the potential harm (bleeding) of the study drugs. Moreover, as an early termination of both antithrombotic groups for efficacy was not anticipated, a strategy for formal testing of secondary outcomes at the interim analysis was not pre-specified. Despite these limitations, the COMPASS trial marks an important milestone in our search for effective secondary prevention strategy for patients with established cardiovascular disease. As Dr. Braunwald effectively articulated in his editorial accompanying the late-breaking trial publication in the October 5, 2017 issue of the New England Journal of Medicine, COMPASS findings need to be viewed as a continuum to secondary prevention trials involving low-dose rivaroxaban in ACS (Anti-Xa Therapy to Lower Cardiovascular Events in Addition to Standard Therapy in Subjects with Acute Coronary Syndrome–Thrombolysis in Myocardial Infarction 51 [ATLAS ACS 2–TIMI 51]) or low-dose ticagrelor (Prevention of Cardiovascular Events in Patients with Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin–Thrombolysis in Myocardial Infarction 54 [PEGASUS-TIMI 54]) and direct thrombin inhibitor drug vorapaxar (Thrombin Receptor Antagonist in Secondary Prevention of Atherothrombotic Ischemic Events–Thrombolysis in Myocardial Infarction 50 [TRA 2°P–TIMI 50]) in stable cardiovascular disease.3 Trials comparing these strategies to establish a personalized regimen for patients may be forthcoming. Towards this objective, perhaps the COMPASS trial will serve as our compass towards navigating the labyrinth of paths leading to an ideal secondary prevention regimen for patients with heart and vascular diseases.

While the FDA's regulatory verdict on low-dose rivaroxaban is awaited, open label drug therapy is currently being offered by the trial sponsors to COMPASS trial participants at no cost.


  1. Eikelboom JW, Connolly SJ, Bosch J, et al. Rivaroxaban with or without aspirin in stable cardiovascular disease. N Engl J Med 2017;377:1319-30.
  2. Bosch J, Eikelboom JW, Connolly SJ, et al. Rationale, design and baseline characteristics of participants in the cardiovascular outcomes for people using anticoagulation strategies (COMPASS) trail. Can J Cardiol 2017;33:1027-35.
  3. Braunwald E. An important step for thrombocardiology. N Engl J Med 2017;377:1387-8.

Keywords: Anticoagulants, Aspirin, Atrial Fibrillation, Secondary Prevention, Acute Coronary Syndrome, Proton Pump Inhibitors, Purinergic P2Y Receptor Antagonists, Medication Adherence, Peripheral Arterial Disease, Receptors, Drug, Coronary Vessels, Pyridines, Lactones, Myocardial Infarction, Stroke, Coronary Disease, Receptors, Thrombin, Aneurysm, Microvessels, Vascular Diseases, Angina, Stable

< Back to Listings