Benefit-Risk Assessment of Rivaroxaban for Extended Thromboprophylaxis After Hospitalization for Medical Illness

Abstract

Background Venous thromboembolism (VTE) often occurs after hospitalization in medically ill patients, but the population benefit-risk of extended thromboprophylaxis remains uncertain. Methods and Results The MARINER (Medically Ill Patient Assessment of Rivaroxaban Versus Placebo in Reducing Post-Discharge Venous Thrombo-Embolism Risk) study (NCT02111564) was a randomized double-blind trial that compared thromboprophylaxis with rivaroxaban 10 mg daily versus placebo for 45 days after hospital discharge in medically ill patients with a creatinine clearance ≥50 mL/min. The benefit-risk balance in this population was quantified by calculating the between-treatment rate differences in efficacy and safety end points per 10 000 patients treated. Clinical characteristics of the study population were consistent with a hospitalized medical population at risk for VTE. Treating 10 000 patients with rivaroxaban resulted in 32.5 fewer symptomatic VTE and VTE-related deaths but was associated with 8 additional major bleeding events. The treatment benefit was driven by the prevention of nonfatal symptomatic VTE (26 fewer events). There was no between-treatment difference in the composite of critical site or fatal bleeding. Conclusions Extending thromboprophylaxis with rivaroxaban for 45 days after hospitalization provides a positive benefit-risk balance in medically ill patients at risk for VTE who are not at high risk for bleeding. Registration URL: https://ichgcp.net/clinical-trials-registry/NCT02111564" title="See in ClinicalTrials.gov">NCT02111564.

Keywords: benefit–risk; medically ill; rivaroxaban; thromboprophylaxis; venous thromboembolism.

Figures

Figure 1. Risk differences per 10 000 patients and 95% CI for pairs of benefit–risk outcomes.

ACM indicates all‐cause mortality; CV, cardiovascular; MI, myocardial infarction; PE, pulmonary embolism; and VTE, venous thromboembolism.

Figure 2. Risk differences per 10 000 patients and 95% CI for key outcomes.

CV indicates cardiovascular; DVT, deep vein thrombosis; Hgb, hemoglobin; PE, pulmonary embolism; and VTE, venous thromboembolism.

Figure 3. Benefit–risk balance for key efficacy and safety outcomes over time.

A, Cumulative rate difference per 10 000 patients: Pair 1, primary efficacy outcome vs major bleeding. B, Cumulative rate difference per 10 000 patients: Pair 3, nonfatal PE, MI, nonhemorrhagic stroke, and CV death vs critical site bleeding and fatal bleeding. CV indicates cardiovascular; MI, myocardial infarction; and PE, pulmonary embolism.