Where is the COMPASS Taking Us? Rivaroxaban, Aspirin, or Both for Stable CVD ?

Since the introduction of direct oral anticoagulants (DOACs) less than a decade ago, use of this class has expanded beyond the prevention and treatment of venous thromboembolism and stroke prevention in the setting of atrial fibrillation. Specifically, the potential role of DOACs in the secondary prevention of coronary artery disease (CAD) has been of considerable interest. Both warfarin and rivaroxaban, in conjunction with antiplatelet agents, have previously been shown to reduce recurrent events in patients with CAD.^{1, 2} So, why aren’t clinicians using these agents in combination with antiplatelet therapy more often? It doesn’t take a COMPASS to point to the answer – major bleeding. In the setting of CAD, warfarin has resulted in significant more major bleeding when given either alone or in combination with antiplatelet agents when compared to aspirin alone.  Therefore, clinicians have been reluctant to embrace the combination of an anticoagulant plus an antiplatelet agent. However, could DOACs have a role in stable CAD? The COMPASS trial aimed to find an answer.³

The Cardiovascular Outcomes for People Using Anticoagulation Strategies (COMPASS) trial was a double-blind, double-dummy, randomized controlled trial designed to evaluate whether rivaroxaban alone or in combination with ASA would be more effective than ASA alone in the secondary prevention of cardiovascular (CV) events. COMPASS was conducted at 602 centers in 33 countries and enrolled patients with a history of stable CAD and/or peripheral arterial disease (PAD) who were not at high risk of bleeding, did not have severe heart failure, and had an eGFR > 15ml/min.

Eligible participants, except those patients who entered the study immediately following coronary bypass surgery, entered a 4-week run-in phase during which they received a rivaroxaban-placebo twice daily plus ASA 100 mg administered daily. Patients who were adherent and did not experience adverse events were randomly assigned in a 1:1:1 ratio to receive rivaroxaban 2.5mg twice daily plus aspirin 100mg daily, rivaroxaban 5mg twice daily, or aspirin 100mg daily.

The trial was conducted using a 3-by-2 partial factorial design where patients completing the run-in phase were secondarily randomized to recieve the proton pump inhibitor (PPI) pantoprazole or placebo. The PPI arm of the trial is still ongoing. Following randomization, participants were seen at month 1, month 6, and then at 6-month intervals for the remainder of the trial.

The primary efficacy outcome was the composite of CV death, stroke, or MI. The primary safety outcome was major bleeding based on a modification of the International Society of Thrombosis and Hemostasis criteria, which included fatal bleeding, symptomatic bleeding in a critical area or organ, or bleeding leading to hospitalization. The secondary outcomes were: a composite of ischemic stroke, MI, acute limb ischemia, or death from coronary heart disease (CHD); the composite of ischemic stroke, MI, acute limb ischemia, or CV death; and death from any cause.  The investigators also compared the net-clinical benefit of the three treatment arms based on a composite of cardiovascular death, stroke, MI plus fatal bleeding or symptomatic bleeding into a critical organ.

There were more than 27,000  participants included in the analysis with a mean age of 68.2 years, 22% were women, and 62.1% were white. Ninety percent of participants had a history of CAD and slight more than a quarter had a history of peripheral arterial disease (PAD). Patients were followed for a mean of 23 months. The primary efficacy outcome occurred in 379 patients (4.1%) in the rivaroxaban plus aspirin (ASA) group, 448 (4.9%) in the rivaroxaban group, and 496 (5.4%) in the ASA group (HR = 0.76 [95% CI 0.66 to 0.86; P<0.001] for the comparison of rivaroxaban plus ASA with ASA alone and HR = 0.90 [95% CI 0.79 to 1.03; P=0.12] for the comparison of rivaroxaban alone with ASA alone.)  All secondary outcomes occurred in fewer participants in the rivaroxaban plus ASA group compared to ASA alone group and all were statistically significant. It is important to note the treatments were terminated early for efficacy.

The primary safety outcome of major bleeding occurred in more patients in the rivaroxaban plus ASA group than in the ASA group alone (288 patients [3.1%] vs. 170 patients [1.9%]; hazard ratio, 1.70; 95% CI, 1.40 to 2.05; P<0.001) . The excess bleeding in the rivaroxaban plus ASA group occurred most commonly in the GI tract. There was no differences between groups in terms of intracranial bleeding, fatal bleeding, or bleeding into a critical organ. Major bleeding events also occurred in more patients in the rivaroxaban group than in the ASA group (255 patients [2.8%] vs. 170 patients [1.9%]; hazard ratio, 1.51; 95% CI, 1.25 to 1.84; P<0.001). The composite net-clinical-benefit outcome favored rivaroxaban plus ASA compared to ASA alone (431 patients [4.7%] vs. 534 patients [5.9%]; hazard ratio, 0.80; 95% CI, 0.70 to 0.91; P<0.001).  

The benefits observed in this study were driven by reductions in CV death, death from CHD, and stroke. Additionally, all-cause mortality was significantly reduced in the rivaroxaban plus ASA group. On the other hand, rivaroxaban monotherapy increased major bleeding compared to ASA alone but did NOT reduce risk of major CV events.

The COMPASS trial has several strengths.  It was very large.  It included multiple centers around the globe and the population was reasonably diverse.  The study was appropriately blinded and most study outcomes were centrally adjudicated. To the investigators’ credit, a statistical adjustment for multiple comparisons was made (Hochberg-based gatekeeping procedure).  However, there are several important limitations to be considered. The early termination of the trial increases the risk that the treatment effect may be overestimated. Although not reported in the full-text publication, supplemental data reveals a significant number of patients (8.2%) were excluded after the run-in period, increasing the risk of selection bias.⁴ The modified ISTH criteria for major bleeding is a more conservative measure including patients with any bleeding leading to hospitalization, thus, capturing more bleeding events – a strength.  HOWEVER, only the most severe bleeding event was considered if patients had more than one major bleeding event during the study.  Furthermore, adjudication of fatal bleeding was left to individual investigators and, thus, some events may not have been accurately classified.

Although the results of the COMPASS trial are promising, a deeper dive into the trial leaves some questions unanswered. Although the use of the “net-clinical-benefit” outcome indicates a 1.2% absolute risk reduction, it is important to consider that the composite, which included three ischemic and two bleeding endpoints, favors the combination treatement group.  When efficacy and safety endpoints are considered separately, the combination of rivaroxaban plus ASA reduced the primary efficacy endpoint by 1.3% at the cost of a 1.2% absolute increase in major bleeding.  These differences represent a number needed to treat (NNT) of 77 and a number needed to harm (NNH) of 83. With the ongoing study arm evaluating PPI use, it is yet unknown whether GI prophylaxis perhaps impacted bleeding risk and, thus, altered the risk-to-benefit ratio.  Despite these limitations, the combination of low-dose rivaroxaban plus ASA likely will be of benefit to select, high-risk patients. Shared-decision making seems critical as the difference between benefits and risks is razor thin.  What we sorely need is a clinical decision tool that weighs both CV and bleeding risks so that we can more precisely determine who might benefit from the combination of rivaroxaban plus aspirin the most. What do you think?  Should we broadly recommend rivaroxaban plus aspirin to most patients with stable CAD or to a narrowly defined sub-set of patients?