Two DOACs show similar bleeding risks in head-to-head VTE trial

In a post hoc analysis of the RENOVE trial, the direct oral anticoagulants (DOACs) rivaroxaban and apixaban demonstrate comparable risks of clinically relevant bleeding (CRB) and recurrent venous thromboembolism (VTE). These were observed across the crude, adjusted, and IPTW* analyses.

In the overall population, the 5-year cumulative incidence of CRB was 11.7 percent with rivaroxaban and 13.7 percent with apixaban. A comparison between arms yielded a crude hazard ratio (HR) of 0.85. This effect was sustained after adjusting for antiplatelet therapy, age, sex, creatinine clearance, history of cancer, and history of bleeding (adjusted HR [aHR], 0.90).

The comparable treatment effects between rivaroxaban and apixaban were sustained, be it with the full (14 percent vs 16.5 percent; aHR, 0.99) or reduced (9 percent vs 11 percent; aHR, 0.81) dose after adjusting for covariates. The IPTW analysis showed similar trends (HRs, 0.92, 0.99, and 0.82 in the overall, full-dose, and reduced-dose populations, respectively).

The 5-year cumulative incidences of major bleeding were also not significantly different between the rivaroxaban and apixaban arms in the overall (2.5 percent vs 3.9 percent; aHR, 0.82), full-dose (3.5 percent vs 4.6 percent; aHR, 0.79), and reduced-dose (1.6 percent vs 3 percent; aHR, 0.87) cohorts. These patterns were sustained in the IPTW analysis (HRs, 0.80, 0.74, and 0.96, respectively).

Similar trends were observed for recurrent VTE in both the adjusted (aHRs, 1.15, 1.08, and 1.20 for the respective overall, full-dose, and reduced-dose cohorts) and IPTW (HRs, 1.17, 0.90, and 1.45) analyses. The same goes for the composite of recurrent VTE and CRB (aHRs, 0.95, 1.01, and 0.88 [adjusted] and 0.96, 1.00, and 0.90 [IPTW]).

Rivaroxaban tied to lower risk of SAEs

However, the negative control outcome demonstrated a significantly lower rate of serious adverse events (SAEs) unrelated to recurrent VTE and bleeding in individuals who have received rivaroxaban. [ASH 2025, abstract 740]

“In the overall and full-dose cohorts, the risk of SAEs was reduced by approximately 30 percent, primarily due to infections and chronic inflammatory disease,” said Dr Francis Couturaud from the FCRIN INNOVTE** Network, Saint Etienne, France, who presented the findings at ASH 2025. “This indicates the presence of residual confounding associated with treatment channelling.”

Pragmatic equivalence of both DOACs

“In patients requiring extended anticoagulation, reduced-dose DOAC regimens lower CRB without compromising efficacy,” Couturaud said. However, there are no randomized studies comparing rivaroxaban and apixaban in the long-term setting. [N Engl J Med 2013;369:799-808; N Engl J Med 2017;376:1211-1222]

The RENOVE open-label trial compared full- and reduced-dose DOACs in patients with VTE who were at high risk of recurrence, had completed at least 6 months of full-dose treatment, and had an indication for extended anticoagulation. [Lancet 2025;405:725-735]

This post hoc analysis included 2,768 RENOVE participants. Of these, 1,385 received full-dose rivaroxaban (20 mg OD; n=755) or apixaban (5 mg BID; n=630), and 1,383 were given reduced doses (10 mg OD [n=758] or 2.5 mg BID [n=625], respectively).

Compared with the rivaroxaban arm, the apixaban arm had older participants (mean age 64.5 vs 61.1 years) and higher proportions of women (38.7 percent vs 32.1 percent), participants with high bleeding risk according to VTE-bleed score (33 percent vs 25 percent), and participants with active cancer (4.1 percent vs 1.3 percent) and previous cancer before VTE (13.1 percent vs 8 percent).

Overall, about 60 percent of participants had a first unprovoked VTE, while the remainder had multiple VTE episodes. Approximately two-thirds had creatinine clearance ≥80 mL/min. The median follow-up was 36 months.

“Taken together, the findings support the pragmatic equivalence of the two DOACs, while underscoring that non-randomized treatment allocation likely influenced some observed differences,” concluded Couturaud.