Background
Phase 2 open-label RCT (SELECT-D). 406 patients with active cancer and confirmed symptomatic or incidental VTE (DVT or PE). Compared rivaroxaban (oral, no parenteral lead-in) to dalteparin (standard-of-care LMWH per CLOT). Pilot/phase 2 design — not powered for definitive efficacy, but provided key signal for GI bleeding risk with rivaroxaban in cancer patients. Duration 6 months. Published 2018 concurrent with Hokusai-VTE Cancer.
Interventions and follow up
Arm A: Rivaroxaban 15 mg twice daily × 3 weeks, then 20 mg once daily for 6 month
Arm B: Dalteparin 200 IU/kg once daily × 1 month, then 150 IU/kg once daily × 5 months (6 months total)
Primary endpoint: Recurrent VTE at 6 months (descriptive; no formal hypothesis test due to phase 2 design)
mFollow up: 6 month
Arm B: Dalteparin 200 IU/kg once daily × 1 month, then 150 IU/kg once daily × 5 months (6 months total)
Primary endpoint: Recurrent VTE at 6 months (descriptive; no formal hypothesis test due to phase 2 design)
mFollow up: 6 month
Results
Recurrent VTE: 4% (rivaroxaban) vs 11% (dalteparin); HR 0.43, 95% CI 0.19–0.99 — P=.04
Major bleeding: 6% vs 4%; HR 1.83, 95% CI 0.68–4.96 (non-significant)
Clinically relevant non-major bleeding (CRNMB): 13% vs 4%; HR 3.76, 95% CI 1.63–8.69
Composite major + CRNMB: 19% vs 8%; HR 2.67 (P<.001)
Major bleeding: 6% vs 4%; HR 1.83, 95% CI 0.68–4.96 (non-significant)
Clinically relevant non-major bleeding (CRNMB): 13% vs 4%; HR 3.76, 95% CI 1.63–8.69
Composite major + CRNMB: 19% vs 8%; HR 2.67 (P<.001)
Adverse events
Main adverse events: Clinically relevant non-major bleeding markedly higher with rivaroxaban (13% vs 4%; HR 3.76), largely GI/GU in origin. Major bleeding numerically higher (6% vs 4%) but not statistically significant. Most CRNMB events occurred in patients with GI or GU cancers (colorectal, gastric, bladder, renal). No fatal bleeds. Quality of life not formally assessed.
Conclusions
Rivaroxaban reduced VTE recurrence by ~57% vs dalteparin in cancer patients but was associated with substantially higher clinically relevant bleeding, particularly in GI/GU malignancies, supporting selective DOAC use in cancer-associated VTE.
Key Limitations
Key Limitations: Phase 2 trial — underpowered for formal efficacy or safety comparisons. Open-label design. The CRNMB excess is a concerning safety signal but the trial was not designed to quantify it definitively. Predominantly concentrated in luminal GI/GU cancers. No stratification by cancer type at randomization. The 6-month primary analysis is short for cancer patients who may need extended anticoagulation. Sample size (n=406) limits subgroup reliability.
Clinical Context
SELECT-D contributed to the evidence base for using DOACs in cancer-VTE but highlighted GI bleeding risk, particularly for rivaroxaban. Subsequent CARAVAGGIO trial (apixaban, n=1,170, Phase 3) showed no significant major bleeding increase vs dalteparin, making apixaban the preferred DOAC for cancer-VTE. NCCN and ASCO guidelines list apixaban as the preferred DOAC; rivaroxaban/edoxaban are alternatives, with caution in GI/GU malignancies.
References