2018 was a banner year for rivaroxaban clinical trials, with the results of COMPASS,1,2 COMMANDER HF,3 NAVIGATE ESUS,4 MARINER,5 and SELECT-D6 announced. The first three studies were designed to reduce arterial events, while the latter targeted venous thromboembolism (VTE). In total, more than 28,000 study participants were exposed to rivaroxaban in doses ranging from 2.5 mg twice daily (bid) to 15 mg bid. The goal of this review is to briefly summarize the trial results and attempt to put them into context (Table). Categorization of a study as a “success” or “failure” is based solely on achievement of the study’s primary outcome.
Summary of Large Rivaroxaban Clinical Trials Reported in 2018
| RCT . | Study Population . | Rivaroxaban . | Comparator . | Primary Efficacy Outcome . | Rivaroxaban Event Rate . | Comparator Event Rate . | HR (95% CI) . |
|---|---|---|---|---|---|---|---|
| COMPASS (n=24,824) | Stable multivessel CAD or PAD or both and age > 65 years plus ≥ 2 vascular beds involved or 2 additional risk factors | 2.5 mg bid + aspirin 100 mg od (n=8,313) or | aspirin 100 mg od (n=8,261) | stroke, MI, cardiovascular death | 4% | 6% | 0.74 (0.65- 0.86) |
| 5 mg bid alone (n=8,250) | 5% | 6% | 0.89 (0.78-1.02) | ||||
| COMMANDER HF (n=5,022) | Chronic HF LVEF < 40% and CAD plus BNP ≥ 200 pg/mL or NT-proBNP ≥ 800 pg/mL and worsening HF within 21 days | 2.5 mg bid (n=2,507) plus standard care (aspirin and DAPT permitted) | Placebo (n=2,515) plus standard care (aspirin and DAPT permitted) | All-cause death, MI, stroke | 25.0% | 26.2% | 0.94 (0.84-1.05) |
| NAVIGATE ESUS (n=7,213) | Ischemic CVA (7 days-6 months) and age > 49 years; no lacunae; no extracranial/ intracranial stenosis > 50%; no embolic source + 20 hrs EKG; if age 50-59 years, also had to have at least 1 additional risk factor | 15 mg od (n=3,609) | Aspirin 100 mg (n=3,604) | First recurrent stroke or systemic embolism | 5.1% | 4.8% | 1.07 (0.87-1.33) |
| MARINER (n=12,019) | age ≥ 40 years, plus hospitalized for 3-10 days plus IMPROVE score ≥ 4 or a score of 2 or 3, plus D-dimer ≥ 2 × upper limit of normal plus received in hospital LMWH or UFH | 10 mg od (n=6,007) for 45d post discharge. 7.5 mg if creatine clearance 30-50 mL/min | Placebo (n=6,012) | Symptomatic VTE or VTE death | 0.83% | 1.10% | 0.76 (0.52-1.09) |
| SELECT-D (n=406) | Active cancer with symptomatic or asymptomatic VTE | 15 mg bid (n=203) × 3 weeks, then 20 mg od | Open-label dalteparin 200 IU/kg od × 1 mo then 150 IU/kg od | VTE recurrence | 4% | 11% | 0.43 (0.19-0.99) |
| RCT . | Study Population . | Rivaroxaban . | Comparator . | Primary Efficacy Outcome . | Rivaroxaban Event Rate . | Comparator Event Rate . | HR (95% CI) . |
|---|---|---|---|---|---|---|---|
| COMPASS (n=24,824) | Stable multivessel CAD or PAD or both and age > 65 years plus ≥ 2 vascular beds involved or 2 additional risk factors | 2.5 mg bid + aspirin 100 mg od (n=8,313) or | aspirin 100 mg od (n=8,261) | stroke, MI, cardiovascular death | 4% | 6% | 0.74 (0.65- 0.86) |
| 5 mg bid alone (n=8,250) | 5% | 6% | 0.89 (0.78-1.02) | ||||
| COMMANDER HF (n=5,022) | Chronic HF LVEF < 40% and CAD plus BNP ≥ 200 pg/mL or NT-proBNP ≥ 800 pg/mL and worsening HF within 21 days | 2.5 mg bid (n=2,507) plus standard care (aspirin and DAPT permitted) | Placebo (n=2,515) plus standard care (aspirin and DAPT permitted) | All-cause death, MI, stroke | 25.0% | 26.2% | 0.94 (0.84-1.05) |
| NAVIGATE ESUS (n=7,213) | Ischemic CVA (7 days-6 months) and age > 49 years; no lacunae; no extracranial/ intracranial stenosis > 50%; no embolic source + 20 hrs EKG; if age 50-59 years, also had to have at least 1 additional risk factor | 15 mg od (n=3,609) | Aspirin 100 mg (n=3,604) | First recurrent stroke or systemic embolism | 5.1% | 4.8% | 1.07 (0.87-1.33) |
| MARINER (n=12,019) | age ≥ 40 years, plus hospitalized for 3-10 days plus IMPROVE score ≥ 4 or a score of 2 or 3, plus D-dimer ≥ 2 × upper limit of normal plus received in hospital LMWH or UFH | 10 mg od (n=6,007) for 45d post discharge. 7.5 mg if creatine clearance 30-50 mL/min | Placebo (n=6,012) | Symptomatic VTE or VTE death | 0.83% | 1.10% | 0.76 (0.52-1.09) |
| SELECT-D (n=406) | Active cancer with symptomatic or asymptomatic VTE | 15 mg bid (n=203) × 3 weeks, then 20 mg od | Open-label dalteparin 200 IU/kg od × 1 mo then 150 IU/kg od | VTE recurrence | 4% | 11% | 0.43 (0.19-0.99) |
Abbreviations: bid, twice daily; BNP, brain natriuretic peptide; CAD, coronary artery disease; DAPT, dual antiplatelet therapy; HF, heart failure; LMWH, low-molecular-weight heparin; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NT-proBNP, N-terminal pro-brain natriuretic peptide; od, once daily; PAD, peripheral arterial disease; UFH, unfractionated heparin; VTE, venous thromboembolism.
Arterial Endpoints
Success – COMPASS: This trial showed that for patients 65 years or older with stable coronary disease or peripheral artery disease, very-low-dose rivaroxaban (2.5 mg bid) plus aspirin (100 mg once daily [od]) reduced the risk of the composite outcome of myocardial infarction, stroke, or cardiovascular death (HR, 0.74; 95% CI, 0.65-0.86; p<0.0001).1 Similarly, in the peripheral artery disease subgroup, the prespecified outcome of major adverse limb events was reduced by nearly half in the very-low-dose rivaroxaban plus aspirin arm (HR, 0.54; 95% CI, 0.35-0.82).2
Not surprisingly, the risk of bleeding was increased in both rivaroxaban arms compared to aspirin alone despite a 30-day run-in phase with aspirin, likely selecting out the patients at highest risk for bleeding. It is also worth noting that the trial was terminated early for efficacy which has raised concern that the bleeding rates are underestimated.
Failure – COMMANDER HF and NAVIGATE ESUS: The COMMANDER HF trial showed that the same dose of rivaroxaban that was successful in COMPASS (2.5 mg bid) did not prevent the composite primary outcome of all-cause death, myocardial infarction, or stroke in patients with recently exacerbated chronic heart failure compared with standard care alone.3 Single or dual antiplatelet therapy were allowed in both arms (98% of study participants were taking at least one antiplatelet agent at baseline).
In NAVIGATE ESUS, rivaroxaban 15 mg od did not reduce the risk of first recurrent stroke or systemic embolism in patients 49 years or older with ischemic stroke of unknown origin (within 7 days to 6 months) compared with aspirin.4 The enrolled patients had presumed cerebral embolism without evidence of arterial stenosis (>50%), lacunae, or an identified cardioembolic source. These findings contrast with the efficacy of the same dose of rivaroxaban in reducing the risk of stroke due to atrial fibrillation (in patients with creatine clearance 30-49 mL/min).7 It is worth noting that the trial was terminated early due to lack of benefit and increased bleeding, including life-threatening or fatal bleeding in the rivaroxaban arm. Please see a full analysis of this study in the Year’s Best article by Dr. Stephan Moll.
Venous Endpoints
Success – SELECT-D: This trial showed that rivaroxaban was non-inferior to low-molecular-weight heparin for treatment of patients with cancer-associated thrombosis.6 SELECT-D was powered as a pilot study, which limits conclusions. However, this result together with the results of the HOKUSAI trial8,9 are slowly shifting the treatment of selected cancer patients toward direct oral anticoagulants (DOACs).
Failure – MARINER: This trial showed that rivaroxaban 10 mg daily for 45 days after discharge for medical patients 40 years or older did not reduce the composite of symptomatic VTE or thrombotic death compared to placebo.5 Interestingly, this result occurred despite the use of a high IMPROVE score and elevated D-dimer as enrollment criteria in an attempt to target patients at higher baseline risk for VTE. The overall event rate was very low, and similar to the arterial trials, the risk of major bleeding was increased in the rivaroxaban arm.
Lessons Learned
The first lesson learned by these trials is one we already knew: Anticoagulant therapy reduces thrombotic events, but excessive bleeding can completely overshadow this benefit. This is the primary reason why older trials using warfarin following ischemic stroke10 or warfarin combined with aspirin for secondary prevention of cardiovascular events11 failed. COMPASS was able to tip the scales in favor of benefit because it used a very low dose of rivaroxaban. In contrast, the reduction of symptomatic thromboembolism provided by low-dose rivaroxaban in MARINER was not enough to compensate for bleeding, despite the attempt to enroll a study population at higher risk for VTE.
The next lesson learned is the target must be right. COMMANDER HF and NAVIGATE ESUS showed that while thrombin generation may contribute to adverse events in these study populations, it is unlikely to be the primary mechanism. Ongoing trials with other DOACs in similar patient groups should support or refute this hypothesis.12,13
Lastly, anticoagulant therapy is not one-size-fits-all. For example, for secondary prevention of cardiovascular events, very-low-dose rivaroxaban (combined with aspirin) seems to be the correct dose, but 20 mg is needed to prevent stroke secondary to atrial fibrillation (15 mg if creatine clearance is 30-49 mL/min).7 With additional doses now entering the market, there is a real danger of clinicians ordering the wrong dose for a given indication. Clinicians tend to defer to lower doses thinking (correctly) they carry a lower risk of bleeding. However, depending on the patient population, the efficacy will also be substantially reduced if the wrong dose is selected.
References
Competing Interests
Dr. Linkins has received data adjudication fees for the MARINER trial.
