Loading...
The FDA recently approved a new 2.5-mg dose of rivaroxaban, to be used with aspirin to lower risk for major adverse cardiovascular (CV) events in patients with chronic coronary artery disease (CAD) or peripheral artery disease (PAD). To review potential benefits and harms of this intervention, it's worth revisiting COMPASS, the randomized trial upon which approval was based (NEJM JW Gen Med Oct 1 2017 and N Engl J Med 2017; 377:1319).
COMPASS included 27,000 patients with histories of symptomatic or revascularized CAD or symptomatic PAD. CAD patients younger than 65 were required to have atherosclerosis in a second vascular bed or at least two of the following risk factors: glomerular filtration rate (GFR) <60 mL/minute/1.73 m2, diabetes, smoking, heart failure, or previous stroke. Exclusion criteria included high risk for bleeding, serious comorbidities, GFR <15 mL/minute/1.73 m2, severe heart failure, and need for dual antiplatelet therapy. Patients received rivaroxaban (2.5 mg twice daily) plus aspirin (100 mg daily), rivaroxaban alone (5 mg twice daily), or aspirin alone.
The trial was terminated at 2 years, when the primary outcome (i.e., myocardial infarction, stroke, or CV death) had occurred in 4.1% of patients receiving combined therapy, 4.9% of those taking rivaroxaban alone, and 5.4% of those taking aspirin alone. Combined therapy was significantly superior to aspirin alone (P<0.001) but not rivaroxaban alone (P=0.12). However, major bleeding events were significantly more common with combined therapy than with aspirin alone (3.1% vs. 1.9%). Of this 1.2 percentage-point difference, 0.3% was fatal bleeding or nonfatal symptomatic bleeding into a critical organ. All-cause mortality was lower with combined therapy than with aspirin alone (3.4% vs. 4.1%; P=0.01).
In sum, during 2 years of treatment in COMPASS, the incidence of major CV events was 1.3 percentage points lower with rivaroxaban plus aspirin than with aspirin alone, but the incidence of major bleeding was 1.2 percentage points higher. Because most serious bleeding was nonfatal gastrointestinal bleeding, the authors concluded that combined therapy conferred a decisive net benefit. In our view, the benefit–harm tradeoff is close enough that clinicians should offer combined therapy only to patients who would have met criteria for enrollment in this trial; taking liberties with other patient populations could risk more harm than benefit.
In discussions with patients, clinicians should attempt to elicit patients' views about the CV and bleeding tradeoffs. It is also reasonable to frame outcomes as “numbers needed to treat”: In this study, about 100 patients were treated for 2 years to confer net CV benefit for 1 patient, and 7 fewer deaths occurred per 1000 treated patients. Finally, the drug cost is not trivial: If the new dose is priced similarly to other doses of rivaroxaban, the retail cost will be approximately US$10,000 per patient during 2 years. Assuming the aforementioned number needed to treat, the drug cost would be about $1 million per one net CV benefit.