Extended treatment for venous thromboembolism: Is direct oral anticoagulant dose‐reduction better for all?

Treatment for acute venous thromboembolism (VTE) requires a minimum of 3 months of therapeutic anticoagulation to prevent recurrence.1 Guidelines recommend extending anticoagulation beyond this initial period in patients with unprovoked VTE or those with persisting risk factors such as active cancer.2 Following the introduction of direct oral anticoagulants (DOACs) for acute VTE management, two clinical trials investigated extended treatment with reduced-dose apixaban (2.5 mg twice daily)3 or rivaroxaban (10 mg once daily) in patients with VTE at intermediate risk of recurrence.4 While both trials showed that apixaban and rivaroxaban were more effective than placebo or aspirin, respectively, in preventing recurrent VTE, they were not designed or powered to compare reduced-dose with full-dose regimens. Although the results suggested that extended treatment with reduced-dose DOACs may be as effective and potentially safer as full-dose therapy, adequately designed and powered trials were awaited to address this important knowledge gap. In the past 2 years, three randomized controlled trials (RCTs) have consolidated the efficacy and safety of reduced-dose DOACs, including in patients at high risk of recurrent VTE. The RENOVE study investigated patients with acute symptomatic pulmonary embolism or proximal deep vein thrombosis at high risk of recurrent VTE, defined as first unprovoked VTE, recurrent VTE, or with other persistent risk factors.5 After completing 6–24 months of full-dose anticoagulation, patients were randomized to receive reduced- or full-dose apixaban or rivaroxaban. Although RENOVE did not meet formal noninferiority for reduced-dose versus full-dose DOACs, the absolute difference in 5-year cumulative outcomes favored reduced dosing with a clear net clinical benefit (recurrent VTE +0.4% vs. clinically relevant bleeding CRB, the composite outcome of major bleeding and clinically relevant non-major bleeding −5.3%). The EVE and API-CAT trials included patients with cancer-associated VTE, comparing reduced- versus full-dose apixaban for extended anticoagulation after at least 6 months of full-dose therapy.6, 7 They showed that reduced-dose apixaban was noninferior to full-dose apixaban for preventing recurrent VTE, while significantly lowering major and CRB.7 Collectively, these studies suggest a net clinical benefit for extended treatment with reduced-dose DOACs among patients at high risk of recurrent VTE including those with active cancer (Table 1).8 The key question now is whether reduced-dose DOACs should become the standard for all patients receiving extended anticoagulation for VTE, or whether certain populations warrant a more individualized approach? Moreover, are there any differences between DOACs? Proximal or distal lower extremity DVT; upper extremity DVT; PE; cerebral venous sinus thrombosis; splanchnic vein thrombosis (hepatic, portal, splenic, mesenteric, renal, or gonadal) Proximal DVT of lower limb; symptomatic or incidental PE Indication for full-dose anticoagulation other than VTE Female with HERDOO2 ≤ 1 Antiphospholipid syndrome Isolated distal DVT High bleeding risk Renal insufficiency Chronic liver disease or chronic hepatitis Dual antiplatelet therapy or aspirin >100 mg Concomitant use of strong CYP3A4 inhibitors or inducers Active cancer Pregnancy Life expectancy 12 months prior to randomization Severe liver disease (Child-Pugh B or C) Active hepatitis Mechanical heart valve prosthesis Bacterial endocarditis Use of P2Y12 inhibitors Use of strong CYP3A4 inducers or inhibitors Atrial fibrillation or other indication for full-dose anticoagulation other than VTE Active bleeding or bleeding disorder Documented anticoagulant failure Pregnancy or nursing VTE at other locations Mechanical heart valve prosthesis Bacterial endocarditis Antiphospholipid syndrome Atrial fibrillation or other indication for full-dose anticoagulation other than VTE Use of P2Y12 inhibitors Use of strong CYP3A4 inducers or inhibitors Dual antiplatelet therapy or aspirin > 100 mg High bleeding risk ECOG 3 or 4 Life expectancy 30 kg/m2 and even fewer >35 kg/m2.7 Obesity was also not a predictor of bleeding in API-CAT.11 Consistent with this, a small observational study in patients with severe obesity (BMI > 40 kg/m2 or weight >120 kg) found no signal of poorer efficacy or safety outcomes with dose-reduced apixaban during extended treatment.12 In patients with pulmonary embolism at high or intermediate-high risk, dose reduction of DOACs may also not lead to improvement in net clinical benefit and death. In a post hoc analysis of RENOVE on this subgroup of patients, recurrent VTE occurred in 3.2% of patients receiving reduced-dose versus 1.9% with full-dose DOACs (aHR 2.10, 95% CI 0.37–12.0).10 CRB was comparable (15.4% vs. 15.5%; aHR 0.77, 95% CI 0.41–1.44), and the composite of recurrence and bleeding showed no significant difference (aHR 0.87, 95% CI 0.49–1.57). All-cause mortality was also similar between reduced- and full-dose groups (15.0% vs. 13.2%; aHR 0.72, 95% CI 0.36–1.46). In contrast, among patients with intermediate-low or low-risk pulmonary embolism, reduced-dose DOACs were associated with a 48% reduction in bleeding and 43% improvement in net clinical benefit. This supports individualized dosing based on pulmonary severity rather than a uniform reduced-dose strategy. RENOVE also evaluated reduced-dose DOACs in a subanalysis of patients with thrombophilia.5 Approximately 37% had non-major thrombophilia (i.e., isolated heterozygous Factor V Leiden or prothrombin G20210A mutations, elevated Factor VIII), and 17% had “major thrombophilia” (i.e., protein C, S or antithrombin deficiency, homozygous Factor V Leiden or prothrombin G20210A mutation, combined thrombophilia, or presence of antiphospholipid antibodies). In non-major thrombophilia, recurrent VTE occurred in 4.1% of patients receiving reduced-dose versus 0% with full-dose DOACs (aHR 11.7, 95% CI 0.37–375) while, interestingly, there were no recurrent VTE in the 218 patients with major thrombophilia, irrespective of DOAC dose group. CRB was comparable (7.5% vs. 6.8%; aHR 1.14, 95% CI 0.40–3.22) for non-major thrombophilia but reduced-dose DOAC showed a beneficial trend for major thrombophilia (6.4% vs. 18.6%; aHR 0.35, 95% CI 0.12–1.00), which also determined the net clinical benefit (non-major thrombophilia: aHR 1.46, 95% CI 0.56–3.83, major thrombophilia: aHR 0.35, 95% CI 0.12–1.00). Thrombophilia was not considered in EVE or API-CAT, and patients with antiphospholipid syndrome were excluded from both RENOVE and API-CAT. Hence, evidence for the efficacy and safety of reduced-dose DOACs in patients with various types of thrombophilia remains uncertain, as numbers were quite small and observed differences were not statistically significant. The exclusion criteria of the trials highlight important patient populations that were underrepresented or not evaluated (Table 1). Both RENOVE and API-CAT only enrolled patients with pulmonary embolism or (proximal) deep vein thrombosis.5, 7 In contrast, EVE included a broader spectrum of VTE such as upper extremity deep vein thrombosis, cerebral venous sinus thrombosis, and splanchnic vein thrombosis, although these only accounted for a small proportion of participants (61/360 patients included 16.9%, cerebral or splanchnic vein thrombosis 6.7%).6 Hence, the efficacy and safety of reduced-dose DOACs for extended treatment of VTE at unusual sites remain uncertain in the absence of separate studies.13 Similarly, certain cancer subtypes, particularly hematologic or intracranial malignancies, were infrequently included in EVE and API-CAT. Although extrapolating the observed benefits of reduced-dose DOACs to these populations may be clinically reasonable, the lack of evidence introduces uncertainty, highlighting the need for dedicated studies in these underrepresented subgroups. Growing evidence has demonstrated the efficacy and improved safety profile of reduced-dose DOAC for extended treatment of VTE. Post hoc analyses of existing trials provide valuable insights to help clinicians individualize anticoagulation with DOAC dosing during extended VTE treatment. However, these findings should be interpreted with caution as such analyses are inherently hypothesis-generating, often limited by insufficient statistical power, multiple comparisons, and potential selection bias. The evidence for reduced-dose DOACs remains limited for certain populations including patients with obesity (≥30 kg/m2), thrombophilia, unusual site VTE, and specific types of cancer (Table 2). In these populations, the balance between thrombotic and bleeding risks may differ. Therefore, individualized treatment decisions and shared decision-making should guide the approach to extended anticoagulation for these patients, and DOAC dose-reduction should not be uniformly applied to all. Patients with (proximal) DVT or PE, with and without active cancer Patients with unusual site VTE (cerebral veins, splanchnic veins, Budd-Chiari syndrome, and other sites) Patients with BMI < 30 kg/m2 Patients with obesity (BMI ≥ 30 kg/m2) Patients with intermediate-low- or low-risk pulmonary embolism Patients with high- or intermediate high-risk pulmonary embolism Patients without thrombophilia Patients with non-majora or major thrombophilia,b or antiphospholipid syndromec Patients with solid cancers Patients with hematologic and intracranial (primary or secondary) malignancies Mandy N. Lauw: Conceptualization; writing—original draft. Saskia Middeldorp: Writing—review and editing. Marc Carrier: Writing—review and editing. Dr. Mandy N. Lauw has received research funding from LeoPharma and honoraria from Viatris, Inari Stryker, AbbVie, Amgen, BMS, and Novo Nordisk for advisory or educational activities. All honoraria are paid to her institution. Prof. Saskia Middeldorp reports honoraria from AstraZeneca, Alveron, Bayer, Hemab, Norgine, Sanofi, Sirius, VarmX, and Viatris for advisory or data safety monitoring boards or educational activities. All honoraria are paid to her institution. Prof. Marc Carrier has received research funding from Pfizer and honoraria from Anthos, Regeneron, Bayer, Pfizer, BMS, and Leo Pharma. All honoraria are paid to his institution. This research received no funding. Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.