Extended treatment of venous thromboembolism with reduced-dose versus full-dose direct oral anticoagulants in patients at high risk of recurrence: a non-inferiority, multicentre, randomised, open-label, blinded endpoint trial

Background In patients with venous thromboembolism at high risk of recurrence for whom extended treatment with direct oral anticoagulants has been indicated, the optimal dose is unknown. We aimed to assess efficacy and safety of reduced-dose versus full-dose direct oral anticoagulants in patients in whom extended anticoagulation has been indicated. Methods RENOVE was a non-inferiority, investigator-initiated, multicentre, randomised, open-label, blinded endpoint trial done in 47 hospitals in France. Ambulatory patients aged 18 years or older with acute symptomatic venous thromboembolism (pulmonary embolism or proximal deep vein thrombosis) who had received 6-24 uninterrupted months of full-dose anticoagulation and for whom extended anticoagulation has been indicated were eligible. Eligible participants were categorised as having either a first unprovoked venous thromboembolism, recurrent venous thromboembolism, presence of persistent risk factors, or other clinical situations considered to be a high risk of recurrence. Participants were randomly assigned (1:1) to receive oral treatment with either a reduced dose of apixaban (2<middle dot>5 mg twice daily) or rivaroxaban (10 mg once daily) or a full dose of apixaban (5 mg twice daily) or rivaroxaban (20 mg once daily) using a centralised randomisation procedure with an interactive web response system. The sequence generation method was a computerised random number generator and was balanced by blocks of different sizes. Randomisation was stratified by centre, type of direct oral anticoagulant, and antiplatelet drug. Physicians and participants were unmasked to treatment allocation; recurrent venous thromboembolism, clinically relevant bleeding, and all-cause death were adjudicated by an independent committee blinded to treatment allocation. The primary outcome was symptomatic recurrent venous thromboembolism, including recurrent fatal or non-fatal pulmonary embolism or isolated proximal deep vein thrombosis (non-inferiority hypothesis 90% power to exclude a hazard ratio [HR] of 1<middle dot>7). The primary outcome and first two secondary outcomes were included in a hierarchical testing procedure. This trial is registered with ClinicalTrials.gov, NCT03285438. Findings From Nov 2, 2017, to July 6, 2022, 2768 patients were enrolled and randomly assigned to the reduced-dose group (n=1383) or the full-dose group (n=1385). 970 (35<middle dot>0%) participants were female, 1797 (65<middle dot>0%) were male, and one (<0<middle dot>1%) had sex not reported. Median follow-up was 37<middle dot>1 months (IQR 24<middle dot>0-48<middle dot>3). Recurrent venous thromboembolism occurred in 19 of 1383 patients in the reduced-dose group (5-year cumulative incidence 2<middle dot>2% [95% CI 1<middle dot>1-3<middle dot>3]) versus 15 of 1385 patients in the full-dose group (5-year cumulative incidence 1<middle dot>8% [0<middle dot>8-2<middle dot>7]; adjusted HR 1<middle dot>32 [95% CI 0<middle dot>67-2<middle dot>60]; absolute difference 0<middle dot>40% [95% CI -1<middle dot>05 to 1<middle dot>85]; p=0<middle dot>23 for non-inferiority). Major or clinically relevant bleeding occurred in 96 patients in the reduced-dose group (5-year cumulative incidence 9<middle dot>9% [95% CI 7<middle dot>7-12<middle dot>1]) and 154 patients in the full-dose group (5-year cumulative incidence 15<middle dot>2% [12<middle dot>8-17<middle dot>6]; adjusted HR 0<middle dot>61 [95% CI 0<middle dot>48-0<middle dot>79]). 1136 (82<middle dot>1%) of 1383 patients in the reduced- dose group and 1150 (83<middle dot>0%) of 1385 in the full-dose group had an adverse event; 374 (27<middle dot>0%) patients in the reduced- dose group and 420 (30<middle dot>3%) in the full-dose group has a serious adverse event. 35 (5-year cumulative incidence 4<middle dot>3% [95% CI 2<middle dot>6-6<middle dot>0]) patients in the reduced-dose group and 54 (5-year cumulative incidence 6<middle dot>1% [4<middle dot>3-8<middle dot>0]) patients in the full-dose group died during the study period. Interpretation In patients with venous thromboembolism requiring extended anticoagulation, reduction of the direct oral anticoagulant dose did not meet the non-inferiority criteria. However, the low recurrence rates in both groups and substantial reduction of clinically relevant bleeding with the reduced dose could support this regimen as an option. Further research will be needed to identify subgroups for whom the anticoagulation dose should not be reduced.