Oral rivaroxaban versus enoxaparin with vitamin K antagonist for the treatment of symptomatic venous thromboembolism in patients with cancer (EINSTEIN-DVT and EINSTEIN-PE): a pooled subgroup analysis of two randomised controlled trials

Summary Background Patients with venous thromboembolism and cancer have a substantial risk of recurrent venous thromboembolism and bleeding during anticoagulant therapy. Although monotherapy with low-molecular-weight heparin is recommended in these patients, in clinical practice many patients with venous thromboembolism and cancer do not receive this treatment. We aimed to assess the efficacy and safety of a single-drug regimen with oral rivaroxaban compared with enoxaparin followed by vitamin K antagonists, in the subgroup of patients with cancer enrolled in the EINSTEIN-DVT and EINSTEIN-PE randomised controlled trials. Methods We did a subgroup analysis of patients with active cancer (either at baseline or diagnosed during the study), a history of cancer, or no cancer who were enrolled in the EINSTEIN-DVT and EINSTEIN-PE trials. Eligible patients with deep-vein thrombosis (EINSTEIN-DVT) or pulmonary embolism (EINSTEIN-PE) were randomly assigned in a 1:1 ratio to receive rivaroxaban (15 mg twice daily for 21 days, followed by 20 mg once daily) or standard therapy (enoxaparin 1·0 mg/kg twice daily and warfarin or acenocoumarol; international normalised ratio 2·0–3·0). Randomisation with a computerised voice-response system was stratified according to country and intended treatment duration (3, 6, or 12 months). The prespecified primary efficacy and safety outcomes of both the trials and this subanalysis were symptomatic recurrent venous thromboembolism and clinically relevant bleeding, respectively. We did efficacy and mortality analyses in the intention-to-treat population, and bleeding analyses for time spent receiving treatment plus 2 days in the safety population (all patients who received at least one dose of study drug). The EINSTEIN-DVT and EINSTEIN-PE studies are registered at ClinicalTrials.gov, numbers NCT00440193 and NCT00439777. Findings In patients with active cancer (diagnosed at baseline or during treatment), recurrent venous thromboembolism occurred in 16 (5%) of 354 patients allocated to rivaroxaban and 20 (7%) of 301 patients allocated to enoxaparin and vitamin K antagonist (hazard ratio [HR] 0·67, 95% CI 0·35 to 1·30). Clinically relevant bleeding occurred in 48 (14%) of 353 patients receiving rivaroxaban and in 49 (16%) of 298 patients receiving standard therapy (HR 0·80, 95% CI 0·54 to 1·20). Major bleeding occurred in eight (2%) of 353 patients receiving rivaroxaban and in 15 (5%) of 298 patients receiving standard therapy (HR 0·42, 95% CI 0·18 to 0·99). The overall frequency of recurrent venous thromboembolism in patients with only a history of cancer (five [2%] of 233 patients in the rivaroxaban group vs five [2%] of 236 in the standard therapy group; HR 0·98, 95% CI 0·28–3·43) was similar to that of patients without cancer (65 [2%] of 3563 vs 70 [2%] of 3594, respectively; HR 0·93, 95% CI 0·66–1·30), but the frequency was increased in patients with active cancer at baseline (six [2%] of 258 vs eight [4%] of 204, respectively; HR 0·62, 95% CI 0·21–1·79) and most markedly increased in patients whose diagnosis of cancer was made during the study (ten [10%] of 96 vs 12 [12%] of 97, respectively; HR 0·80, 95% CI 0·34–1·88). The overall frequency of major bleeding in patients with only a history of cancer (one [ vs four [2%] patients in the standard therapy group; HR 0·23, 95% CI 0·03–2·06) was similar to that of patients without cancer (31 [1%] vs 53 [1%], respectively; HR 0·58, 95% CI 0·37–0·91), but was increased in patients with active cancer at baseline (five [2%] vs eight [4%], respectively; HR 0·47, 95% CI 0·15–1·45) and was highest in those with cancer diagnosed during the study (three [3%] vs seven [7%], respectively; HR 0·33, 95% CI 0·08–1·31). Interpretation In patients with active cancer and venous thromboembolism, rivaroxaban had similar efficacy to prevent recurrence of venous thromboembolism and reduced the number major bleeding events compared with treatment with enoxaparin and a vitamin K antagonist, although there was no difference between groups for clinically relevant bleeding. Based on these results, a head-to-head comparison of rivaroxaban with long-term low-molecular-weight heparin in patients with cancer is warranted. Funding Bayer HealthCare Pharmaceuticals and Janssen Research & Development.