Abstract 12863: Rivaroxaban for Extended Thromboprophylaxis After Hospitalization for Medical Illness: Pooled Analysis of Mortality and Major Thromboembolic Events in 16,496 Patients From the MAGELLAN and MARINER Trials

IntroductionThromboprophylaxis extended after hospital discharge in high- risk medical patients remains controversial due to uncertainty about the benefit for preventing fatal or major thromboembolic events, and the risk of major bleeding offsetting reduced non-fatal venous thromboembolism.HypothesisThis post-hoc analysis tested the hypotheses that extended thromboprophylaxis with rivaroxaban 10 mg once daily would (1) reduce all-cause mortality (ACM) or major/irreversible thromboembolic outcomes (ie symptomatic deep venous thrombosis, non-fatal PE, myocardial infarction or non-hemorrhagic stroke) with an acceptable risk of on treatment fatal or critical site bleeding (eg intracranial) (2) reduce ACM up to six weeks after hospital discharge.MethodsThe MARINER trial compared rivaroxaban begun at hospital discharge at a dose of 10 mg (CrCL ≥ 50 mL/min) or 7.5 mg (Cr Cl 30 to < 50ml/min) with placebo for 45 days (randomization stratified by renal function/rivaroxaban dose). The MAGELLAN trial compared rivaroxaban (10mg QD for 35±4 days) with enoxaparin 40mg QD for 10±4 days followed by placebo. The safety population from each study was used to derive the pooled analysis population. The data from the 10 mg dose strata in the MARINER trial were pooled with the patients who continued the outpatient phase of the MAGELLAN trial (ie patients free of thromboembolism or bleeding up to the last dose of enoxaparin).ResultsThe analysis included 16,496 patients. The treatment groups were comparable after pooling. The efficacy and safety outcomes are shown in the Table.ConclusionsThe results suggest a benefit for reducing the composite of ACM and major/irreversible thromboembolic outcomes (NNT 197), with a favorable trade-off to fatal or critical site bleeding (NNH 2,045). Extended thromboprophylaxis with rivaroxaban did not achieve a statistically significant reduction in ACM, but a potentially important benefit cannot be excluded.