Reduced- Compared with Standard-Dose Direct Oral Anticoagulant for Extended Treatment of Venous Thromboembolism: A Systematic Review and Meta-Analysis

Abstract

Introduction

Reduced-dose direct oral anticoagulants (DOACs) may provide similar efficacy with less bleeding than standard dose for extended venous thromboembolism (VTE) treatment. It is unclear whether standard doses are preferable in certain subgroups.

Methods

We systematically searched MEDLINE, EMBASE, EMCARE, and Cochrane Central Register of Controlled Trials (CENTRAL) for randomized trials comparing reduced- with standard-dose DOACs for extended VTE treatment (registration number: INPLASY202550061). Outcomes included recurrent VTE, major bleeding, clinically relevant non-major bleeding (CRNMB), and mortality.

Results

Five trials (8,781 patients) were included. Reduced-dose DOACs (apixaban 2.5 mg twice daily or rivaroxaban 10 mg once daily, n = 4,395), compared to standard dose (apixaban 5 mg twice daily or rivaroxaban 20 mg once daily, n = 4,386), resulted in similar rates of recurrent VTE (1.66% vs. 1.78%; risk ratio [RR] 0.94, 95% confidence interval [CI] 0.68–1.29). Major bleeding was less frequent with reduced dose (1.16% vs. 1.96%; RR 0.62, 95% CI 0.42–0.92), as was CRNMB (5.16% vs. 7.00%; RR 0.75, 95% CI 0.63–0.88). Mortality rates were comparable (4.91% vs. 5.81%; RR 0.86, 95% CI 0.63–1.17). These results held for high-risk subgroups, including patients with recurrent VTE or active cancer, except that reduced-dose DOACs appeared to lower recurrent VTE risk in males but increase risk in females (p = 0.04). Risk of bias was rated “low” in four studies and “some concerns” in one study. Certainty of evidence was moderate for three outcomes and low for one outcome.

Conclusion

For extended VTE treatment, reduced-dose DOACs have a similar risk of recurrent VTE and a lower risk of major and CRNMB compared to standard dose, including high-risk patients. A potential interaction with sex warrants further investigation.

Keywords

direct oral anticoagulant - extended treatment - secondary prevention - venous thromboembolism - bleeding

Introduction

Patients with a history of unprovoked venous thromboembolism (VTE) or persistent risk factors are at high risk of recurrent VTE events and may require extended anticoagulation.[1] [2] Randomized controlled trials (RCTs) have shown that extended treatment with reduced- compared to standard-dose direct oral anticoagulant (DOAC) therapy is associated with similar rates of recurrent VTE and may decrease the risk of bleeding.[3] [4] [5] More recent studies have demonstrated similar outcomes with reduced-dose DOAC therapy in patients at higher risk of recurrent VTE, such as those with a history of recurrent events or active cancer.[6] [7] [8] However, it remains unclear whether certain patient subgroups, including younger individuals, patients with elevated body weights or body mass indices (BMIs), or those with preserved renal function, benefit more from continuing extended anticoagulation with standard-dose DOAC.

We conducted a systematic review and meta-analysis of randomized trials comparing reduced- and standard-dose DOAC therapy for extended VTE treatment, aiming to estimate treatment effects on recurrent VTE and bleeding, and to determine if specific patient subgroups require ongoing standard-dose therapy.

Methods

The study was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and is reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines ([Supplementary Table S1]).[9] [10] The study protocol was registered prospectively with the International Database to Register Systematic Reviews (INPLASY) before commencing literature searches (registration number: INPLASY202550061).

Eligibility Criteria

We included RCTs that enrolled adult patients (aged 18 years or older) with VTE who had completed an initial 3 to 6 months of standard-dose anticoagulation therapy and were subsequently randomized to receive extended anticoagulation with either reduced- or standard-dose DOAC for any duration. To be included, studies had to report either recurrent VTE or bleeding events.

Literature Search

A librarian-designed comprehensive electronic search of MEDLINE, EMBASE, EMCARE, and the Cochrane Central Register of Controlled Trials (CENTRAL) was conducted from inception through May 27, 2025. The search strategy incorporated medical subject headings (MeSH) and keywords related to: (“deep vein thrombosis” OR “pulmonary embolism” OR “venous thromboembolism”) AND (“direct oral anticoagulant” OR “non vitamin K oral anticoagulant” OR “apixaban” OR “dabigatran” OR “edoxaban” OR “rivaroxaban”) AND (“dose reduction” OR “drug tapering”). No language restrictions were applied during the search process. A detailed search strategy is provided in the [Supplementary Tables S2] to [S5].

Study Selection and Data Extraction

Two independent reviewers (C.K.MC., S.C., or T.C.S.) screened titles and abstracts for potential eligibility. All potentially relevant citations were advanced to full-text review. The full texts of identified studies were assessed using the predefined inclusion and exclusion criteria. Disagreements during full-text review were resolved through consensus.

Data extraction was performed independently by two authors (C.K.M.C. and T.C.S.) with discrepancies resolved through re-examination of the source articles and consensus discussion. If more than one publication of a study was found, the publication with the most complete data was used in the analyses. Extracted data included first author, year of publication, study design, sample size, number of patients on reduced- and standard-dose DOAC, study period, and duration of follow-up. We collected detailed patient characteristics including age, sex, body weight, body mass index (BMI), presentation of index VTE event (pulmonary embolism [PE] with or without deep vein thrombosis [DVT] vs. DVT alone), prior history of VTE, presence of provoking factor for VTE, creatinine clearance (CrCl), presence of known thrombophilia, primary cancer site, presence of metastasis, cancer stage and Eastern Cooperative Oncology Group (ECOG) Performance Status Scale. Study quality was assessed using the revised Cochrane risk-of-bias tool for randomized trials (RoB 2) Tool.[11]

Outcomes

The primary outcomes included recurrent VTE, major bleeding, clinically relevant non-major bleeding (CRNMB)—both bleeding events defined by the International Society on Thrombosis and Haemostasis (ISTH) criteria[12]—and all-cause mortality. Prespecified subgroup analyses were performed: Age (<75 vs. ≥75 years or <65 vs. ≥65 years depending on the available data), sex (male vs. female), presentation of index VTE event (PE with or without DVT vs. DVT alone), prior history of VTE (present vs. absent), BMI (<30 vs. ≥30 kg/m²), known thrombophilia (present vs. absent), and CrCl (<50 mL/min vs. 50 to <80 mL/min vs. ≥80 mL/min). We assessed the quality of evidence for each outcome using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) method with the GRADEpro Guideline Development Tool (McMaster University, 2015).[13]

Statistical Analysis

Pooled effect estimates were calculated as relative risk (RR) with 95% confidence intervals (CIs) using the Mantel–Haenszel method with random effects models. We assessed heterogeneity using the Cochran's Q statistic and I ² statistic, supplemented by visual inspection of plots. Differences between subgroups were assessed using the chi-square (χ²) test, with p-values below 0.1 signifying significant differences. All analyses were performed using RevMan 5.4 (Cochrane Collaboration) and R Studio (RStudio Team, 2020).[14]

Results

The initial search yielded 1,387 studies. After excluding duplicates, 972 studies remained. Of the 26 studies selected for full-text assessment, 5 studies met eligibility criteria. Twenty-one studies were excluded based on the following reasons: Duplicate publications, review articles or editorials, trial design protocols, and irrelevance to the study question ([Supplementary Table S6]). The study selection process is illustrated in the flow diagram, following the PRISMA model ([Fig. 1]).

Fig. 1 PRISMA flow diagram of study search and selection process. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Study Characteristics

The five included studies were multicenter RCTs, with their key characteristics summarized in [Table 1].[3] [4] [6] [7] [8] The AMPLIFY-EXT, EINSTEIN CHOICE, and RENOVE trials primarily enrolled non-cancer patients, while API-CAT and EVE enrolled patients with cancer-associated thrombosis (CAT). AMPLIFY-EXT and EINSTEIN CHOICE employed three-arm designs (including placebo or aspirin comparator groups alongside reduced- and standard-dose DOAC arms, respectively), whereas the remaining trials used two-arm designs (reduced- and standard-dose DOAC). All participants had completed at least 6 months of standard-dose anticoagulation before enrollment. Three of the five trials investigated only apixaban (2.5 mg vs. 5 mg twice daily). At the same time, EINSTEIN CHOICE evaluated rivaroxaban at 10 mg versus 20 mg once daily, and RENOVE assessed both apixaban and rivaroxaban. The non-cancer trials used symptomatic recurrent VTE as the primary outcome, while the cancer trials also included incidental VTE in their assessment. Most trials followed patients for 12 months, except for RENOVE, which followed patients for up to 5 years. The risk of bias was low in four studies, and one was rated as “some concern” due to potential deviations from the intended intervention in the RoB 2 assessment. Details of the bias assessment are provided in [Supplementary Fig. S1]. No formal test for publication bias was conducted, owing to the low number of included studies.

Table 1 Characteristics of included studies

Study	Study design	Intervention	Reduced dose (n)	Standard dose (n)	Mean age	Sex (male)	Inclusion criteria	Key study population characteristics, percentage	Follow-up duration
Non-cancer-associated VTE
AMPLIFY-EXT (Agnelli et al, 2013)[3]	Multicenter, randomized, double-blind, placebo-controlled study	Apixaban 2.5 mg BID vs. 5 mg BID vs. placebo	840	813	• Reduced dose: 57 • Standard dose: 56	57%	• Symptomatic • Proximal DVT or PE	• Unprovoked 92% • Provoked 8% • Previous history of VTE 13% • Known thrombophilia 4%	12 months[a]
EINSTEIN CHOICE (Weitz et al, 2017)[4]	Multicenter, randomized, double-blind, double-dummy, active-comparator, event-driven study	Rivaroxaban 10 mg OD vs. 20 mg OD vs. aspirin 100 mg OD	1,127	1,107	• Reduced dose: 59 • Standard dose: 58	55%	• Symptomatic • Proximal DVT or PE	• Unprovoked 41% • Provoked 59% • Previous history of VTE 18% • Known thrombophilia 7%[b] • Hormonal therapy 3%	12 months[a]
RENOVE (Couturaud et al, 2025)[6]	Multicenter, randomized, open-label, blinded endpoint trial	Reduced dose (apixaban 2.5 mg BID or rivaroxaban 10 mg OD) vs. standard-dose (apixaban 5 mg BID or rivaroxaban 20 mg OD)	1,383	1,385	• Reduced dose: 62 • Standard dose: 63	65%	• Symptomatic • Proximal DVT or PE • First unprovoked VTE • Multiple episodes of VTE • Persistent risk factor • Other clinical situations considered to be a high risk of recurrence (e.g., major thrombophilia, receiving antithrombotic treatment, PE at high-risk of death at the acute phase, ilio-vena cava DVT)	• Unprovoked 61% • Previous history of VTE 33% • Thrombophilia 37%[c] • Major thrombophilia 17%	37.1 months (median)
Cancer-associated VTE
EVE (McBane et al, 2024)[7]	Multicenter, randomized, double-blind trial	Apixaban 2.5 mg BID vs. apixaban 5 mg BID	179	181	• Reduced dose: 64 • Standard dose: 64	45%	• Symptomatic or incidental • Proximal or distal DVT of the lower extremity • upper extremity DVT • PE • cerebral venous sinus thrombosis • Splanchnic vein thrombosis	• Previous history of VTE 9% • Metastatic cancer 60% • Systemic cancer therapy 61%	12 months[a]
API-CAT (Mahé et al, 2025)[8]	Multicenter, randomized, double-blind, non-inferiority trial	Apixaban 2.5 mg BID vs. apixaban 5 mg BID	866	900	• Reduced dose: 67 • Standard dose: 68	43%	• Symptomatic or incidental • Proximal lower limb DVT or PE in a segmental or larger pulmonary artery	• Previous history of VTE 19% • Metastatic cancer 66% • Systemic cancer therapy 81%	12 months[a]

Abbreviations: BID, twice daily; DOAC, direct oral anticoagulant; DVT, deep vein thrombosis; OD, once daily; PE, pulmonary embolism; vs., versus; VTE, venous thromboembolism.

^a Figures represent the intended study period, as the median follow-up was not provided.

^b Thrombophilia refers to antiphospholipid antibodies, antithrombin deficiency, protein C deficiency, protein S deficiency, factor V Leiden, or prothrombin gene mutation.

^c Thrombophilia refers to antiphospholipid antibodies, antithrombin deficiency, protein C deficiency, protein S deficiency, homozygous or heterozygous factor V Leiden or prothrombin gene mutation, compound heterozygosity, and elevated factor VIII.

Baseline Characteristics of Participants

Collectively, the studies enrolled 8,781 patients, including 4,902 males (55.8%), 3,878 females (44.2%), and 1 patient with unreported sex. Treatment allocation included 4,395 patients receiving reduced-dose DOACs versus 4,386 on standard-dose regimens. The specific DOAC distributions were apixaban 2.5 mg twice daily (n = 2,510, 28.6%), apixaban 5 mg twice daily (n = 2,524, 28.7%), rivaroxaban 10 mg once daily (n = 1,885, 21.5%), and rivaroxaban 20 mg once daily (n = 1,862, 21.2%). The mean ages ranged from 56 to 63 years in non-cancer trials, compared to 64 to 68 years in cancer trials. Regarding the presentation of index VTE events, 5,574 patients (63.5%) had PE with or without DVT, compared to 3,191 patients (36.3%) with DVT alone (in either limb or an unusual site), and 16 patients (0.2%) had asymptomatic or unconfirmed index events. In non-cancer trials, unprovoked VTE accounted for 41.3% to 91.7% of cases, with known thrombophilia present in 3.8% to 37% of patients. Among the two CAT trials, the prevalence of metastatic disease ranged from 59.7% to 65.6%. Concurrent cancer therapy was administered to 218 patients (60.6%) in EVE versus 1,434 (81.2%) in API-CAT.

Outcomes

The prespecified outcomes of this review are presented in the summary of findings ([Table 2]).

Table 2 Summary of findings

Outcomes	Number of participants (studies), follow-up	Certainty of the evidence (GRADE)	Relative effect (95% CI)	Anticipated absolute effects
				Risk with standard-dose DOAC	Risk difference with reduced-dose DOAC
Recurrent VTE	8,781 (5 RCTs)	⨁⨁⨁◯ Moderate[a]	RR 0.94 (0.68–1.29)	18 per 1,000	1 fewer per 1,000 (6 fewer to 5 more)
Major bleeding	8,781 (5 RCTs)	⨁⨁⨁◯ Moderate[a]	RR 0.62 (0.42–0.92)	20 per 1,000	7 fewer per 1,000 (11 fewer to 2 fewer)
Clinically relevant non-major bleeding	8,781 (5 RCTs)	⨁⨁⨁◯ Moderate[a]	RR 0.75 (0.63–0.88)	70 per 1,000	17 fewer per 1,000 (26 fewer to 8 fewer)
Mortality	8,781 (5 RCTs)	⨁⨁◯◯ Low[b]	RR 0.86 (0.63–1.17)	58 per 1,000	8 fewer per 1,000 (22 fewer to 10 more)

Abbreviations: CI, confidence interval; DOAC, direct oral anticoagulant; GRADE, Grading of Recommendations Assessment, Development, and Evaluation; RCT, randomized controlled trial; RR, risk ratio; VTE, venous thromboembolism.

The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

GRADE Working Group grades of evidence

High certainty: We are very confident that the true effect lies close to that of the estimate of the effect.

Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.

Very low certainty: We have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

Explanations

^a Inconsistency because of uncertainty of treatment effect (−1).

^b Inconsistency because of uncertainty of treatment effect (−2).

Recurrent Venous Thromboembolism

The meta-analysis of all five trials showed similar recurrent VTE rates between reduced-dose and standard-dose DOACs (1.66% vs. 1.78%, risk ratio [RR] 0.94, 95% CI 0.68–1.29; p = 0.70; [Fig. 2A]). Similar effectiveness was seen in both non-cancer trials (1.37% vs. 1.39%; RR 0.98, 95% CI 0.65–1.48) and cancer trials (2.58% vs. 2.96%, RR 0.87, 95% CI 0.53–1.45; p-value for subgroup differences =0.71; [Supplementary Fig. S2]). The quality of evidence was moderate.

Fig. 2 (A) Recurrent VTE. (B) Major bleeding. (C) Clinically relevant nonmajor bleeding. (D) Mortality. Forest plots of the effects of reduced-dose versus standard-dose DOACs. CI, confidence interval; DOAC, direct oral anticoagulant; M–H, Mantel–Haenszel; VTE, venous thromboembolism.

Major Bleeding

The meta-analysis of all five trials showed that reduced-dose DOAC decreased the risk of major bleeding compared to the standard dose (1.16% vs. 1.96%, RR 0.62, 95% CI 0.42–0.92; p = 0.02; [Fig. 2B]). A similar effect size was observed in both non-cancer trials (0.66% vs. 1.36%, RR 0.54, 95% CI 0.27–1.06) and cancer trials (2.78% vs. 3.79%, RR 0.73, 95% CI 0.46–1.17; p-value for subgroup differences =0.47; [Supplementary Fig. S3]). The certainty of the evidence was moderate.

Clinically Relevant Non-Major Bleeding

The meta-analysis of all five studies showed a reduction in CRNMB risk with reduced-dose compared to standard-dose DOACs (5.16% vs. 7.00%, RR 0.75, 95% CI 0.63–0.88; p = 0.0006; [Fig. 2C]). Similar effect sizes were observed in both non-cancer trials (3.91% vs. 5.51%, RR 0.71, 95% CI 0.57–0.89) and cancer trials (9.19% vs. 11.56%, RR 0.80, 95% CI 0.62–1.02), with p-value for subgroup differences =0.52 ([Supplementary Fig. S4]). The overall certainty of the evidence was moderate.

Mortality

The meta-analysis of all five studies showed similar all-cause mortality between reduced-dose and standard-dose DOACs (4.91% vs. 5.81%, RR 0.86, 95% CI 0.63–1.17; p = 0.35; [Fig. 2D]). Mortality rates were not different between dosing strategies in either non-cancer populations (1.31% vs. 2.00%, RR 0.69, 95% CI 0.31–1.53) or cancer populations (16.46% vs. 17.48%, RR 0.94, 95% CI 0.78–1.14; p-value for subgroup differences =0.46; [Supplementary Fig. S5]). The evidence certainty was low.

Subgroup Analyses on Recurrent Venous Thromboembolism

Age

Three trials (n = 5,303) reported the risk of recurrent VTE stratified by age (cutoff at 75 years).[3] [4] [8] There was no heterogeneity of treatment effect by age: <75 years (1.51% vs. 2.22%, RR 0.68, 95% CI 0.44–1.05) or ≥75 years (2.56% vs. 1.43%, RR 1.67, 95% CI 0.57–4.85), p-value for subgroup differences = 0.13 ([Supplementary Fig. S6]).

When analyzed using a cutoff at 65 years of age (three trials, n = 6,384), there was no heterogeneity for recurrent VTE: <65 years (1.10% vs. 1.44%, RR 0.67, 95% CI 0.18–2.45) and ≥65 years (1.91% vs. 1.46%, RR 1.28, 95% CI 0.49–3.36; p-value for subgroup differences = 0.43; [Supplementary Fig. S7]).[3] [4] [6]

Sex

Four trials (n = 7,800) reported the risk of recurrent VTE according to sex.[3] [4] [6] [8] Male patients, compared with female patients, appeared to experience a greater benefit with reduced-dose DOACs with respect to preventing recurrent VTE (1.44% vs. 2.19%, RR 0.67, 95% CI 0.43–1.05; 1.88% vs. 1.31%, RR 1.41, 95% CI 0.81–2.45; p-value for subgroup differences = 0.04; [Fig. 3]).

Fig. 3 Forest plots of recurrent VTE stratified by sex. CI, confidence interval; M–H, Mantel–Haenszel; VTE, venous thromboembolism.

Index Presentation of Venous Thromboembolism

Four trials (n = 7,787) reported recurrent VTE according to the index location of the index event.[3] [4] [6] [8] The analysis revealed no significant heterogeneity on index event type: PE with or without DVT (1.87% vs. 1.90%, RR 0.98, 95% CI 0.63–1.53) or isolated DVT (1.23% vs. 1.66%, RR 0.74, 95% CI 0.40–1.37; p-value for subgroup differences = 0.47; [Supplementary Fig. S8]).

Prior History of Venous Thromboembolism

Three trials (n = 6,145) reported recurrent VTE based on a prior history of VTE.[4] [6] [8] There was no heterogeneity for recurrent VTE: Patients without prior VTE (1.68% vs. 1.64%, RR 1.02, 95% CI 0.65–1.59) and those with prior VTE (1.44% vs. 2.41%, RR 0.60, 95% CI 0.28–1.27; p-value for subgroup differences = 0.24; [Supplementary Fig. S9]).

Body Mass Index

Three trials (n = 6,143) reported recurrent VTE according to BMI.[4] [6] [8] There was no heterogeneity for recurrent VTE by BMI category: Non-obese (BMI <30 kg/m²: 1.55% vs. 2.09%, RR 0.75, 95% CI 0.48–1.17) or obese (BMI ≥30 kg/m²: 1.79% vs. 1.26%, RR 1.36, 95% CI 0.27–6.84; p-value for subgroup differences = 0.48; [Supplementary Fig. S10]).

Presence of Known Thrombophilia

Two trials (n = 3,396) that included 584 patients with known thrombophilia (defined as antiphospholipid antibodies, antithrombin, protein C, or protein S deficiency, factor V Leiden, or prothrombin gene mutation) were included in this analysis.[4] [6] There was no heterogeneity of treatment effect for recurrent VTE by thrombophilia status: Patients with known thrombophilia (2.42% vs. 0.68%, RR 2.72, 95% CI 0.26–28.76) and patients without thrombophilia (1.13% vs. 1.58%, RR 0.72, 95% CI 0.38–1.36; p-value for subgroup differences = 0.28; [Supplementary Fig. S11]).

Kidney Function

Four trials (n = 7,552) reported recurrent VTE according to baseline kidney function.[3] [4] [6] [8] Although the trials used slightly different cutoff values for categories of kidney impairment, we standardized the analysis by grouping patients into moderate-to-severe impairment (CrCl less than or equal to 50 mL/min), mild impairment (CrCl between 50 and 80 mL/min), and normal kidney function (CrCl greater than or equal to 80 mL/min). There was no heterogeneity in treatment effect across baseline kidney function categories: Moderate-to-severe impairment (2.43% vs. 1.57%, RR 1.57, 95% CI 0.38–6.52), mild impairment (1.97% vs. 2.22%, RR 0.88, 95% CI 0.49–1.58), and normal kidney function (1.44% vs. 1.61%, RR 0.89, 95% CI 0.54–1.47; p-value for subgroup differences = 0.75; [Supplementary Fig. S12]).

Subgroup Analyses for Clinically Relevant Bleeding

While our protocol also intended to investigate the effect of treatment on both major and CRNMB in subgroups, all included studies exclusively reported the composite of clinically relevant bleeding (combining major bleed and CRNMB) for each subset.[3] [4] [6] [8] We therefore adapted our subgroup analyses to this consistent endpoint. The effects of reduced- compared with standard-dose DOACs were consistent across all subgroups examined ([Supplementary Figs. S13]–[S19]).

Discussion

This systematic review and meta-analysis provide the best available estimates of treatment effect with reduced- compared with standard-dose DOACs for extended VTE treatment in patients at high risk for recurrent VTE, including those with CAT. Reduced-dose DOACs were associated with similar rates of recurrent VTE and lower rates of major and CRNMB, with no difference in mortality. Specifically, reduced-dose DOAC, compared to standard-dose regimens, achieved absolute risk reductions of 0.8% (95% CI 0.16–1.14%) for major bleeding and 1.84% (95% CI 0.84–2.59%) for CRNMB. These reductions correspond to 7 fewer major bleeds and 17 fewer CRNBM events per 1,000 patients treated. Efficacy and safety were consistent across subgroups, except that reduced- compared with standard-dose DOAC appeared to reduce the risk of recurrent VTE in males while increasing the risk in females.

Three of the four trials that contributed to subgroup interaction between DOAC dose and sex showed a similar pattern of higher risk of recurrent VTE with reduced- versus standard-dose DOAC in females. However, the finding of an interaction between DOAC dose and sex for recurrent VTE should be interpreted with caution because it is unexpected, lacks a plausible biological explanation, and was only nominally statistically significant, which suggests it might be a chance finding.

In a previous meta-analysis of the AMPLIFY-EXT and EINSTEIN CHOICE trials, Vasanthamohan et al[5] reported that, for extended VTE treatment, efficacy was maintained with reduced doses of apixaban or rivaroxaban, and there was a trend for less bleeding compared to standard doses. Our analysis, which includes data from three subsequent trials (adding 4,894 patients), not only confirms the efficacy findings but also demonstrates that major and CRNMB events are significantly reduced. Importantly, these findings were consistent in cancer patients with VTE, who are not only at higher risk for recurrent VTE than those without cancer, but also at higher risk for bleeding. Therefore, reduced-dose DOACs offer the potential for enhanced safety in terms of the bleeding risk across the spectrum of patients requiring extended VTE treatment.

The effectiveness of reduced-dose DOACs for extended VTE treatment is likely due to two factors. First, the risk of recurrent VTE is highest during the first 3 to 6 months after the acute event and declines thereafter.[15] [16] Consequently, full-intensity DOAC therapy is unnecessary for secondary prevention thereafter. This contrasts with warfarin, where previous trials demonstrated inferior efficacy of lower-intensity regimens—a reflection of its dose–response antithrombotic effect.[17] [18] [19] Second, pharmacodynamic studies indicate that reduced doses of both apixaban and rivaroxaban are effective in inhibiting thrombin generation.[20] [21] [22]

Despite emerging evidence supporting the use of reduced-dose DOACs for extended VTE treatment, they remain underutilized, with real-world data indicating that only 9% to 13% of eligible patients receive reduced-dose regimens.[23] [24] [25] This underuse may reflect clinicians' uncertainty about the bleeding risk reduction and previously insufficient efficacy data of reduced-dose DOACs. Our meta-analysis corroborates evidence that the DOAC dose should be reduced for extended VTE treatment in most patients, including those with cancer.[26] [27] [28] [29]

The strengths of our analyses include the use of robust methodology and the presentation of estimates of treatment effects based on all available randomized clinical trial evidence. However, several limitations warrant consideration. First, four of the five trials were limited to outcomes at 1 year. However, as suggested by the recent trial reported by Couturaud and colleagues,[6] the efficacy and safety of reduced-dose DOACs are likely to persist for a longer period. Second, subgroup analyses for certain populations (particularly patients with thrombophilia or obesity) were limited by small numbers of participants, which may compromise confidence in the evidence. Third, the results may not be generalizable to groups that were underrepresented in the trials, such as patients taking concomitant potent CYP3A4 or P-glycoprotein inhibitors or inducers, which may influence DOAC levels, or patients with severe kidney or hepatic impairment, which may adversely affect DOAC clearance or metabolism. These possibilities require further research. Finally, the findings apply to apixaban and rivaroxaban and cannot be extrapolated to other DOACs.

Conclusion

This systematic review and meta-analysis reinforce the rationale for using reduced-dose DOACs for extended VTE treatment to lower the risks of major and clinically relevant non-major bleeding, even in higher-risk populations and those with CAT. Prospective studies are necessary to clarify the observed interaction between DOAC dose and sex and to assess dose-reduction strategies in populations that are not well-represented in the randomized trials.

Conflict of Interest

S.C. has served as a consultant and/or received speaker fees from AstraZeneca, Pfizer, Fresenius Kabi, Leo Pharma, and Servier.

V.B. has received an unrestricted education grant from Pfizer, a research grant from Inari and the Canadian Institutes of Health Research.

R.W. has served as a consultant and/or received speaker fees from Boehringer Ingelheim, BMS, Pfizer, Daiichi Sankyo, and Leo Pharma.

J.I.W. has served as a consultant and received honoraria from Bayer, BMS, Boehringer Ingelheim, Daiichi Sankyo, Johnson & Johnson, Pfizer, and Servier.

J.W.E. has received grants and/or honoraria from Anthos, Bayer, BI, BMS, Daiichi Sankyo, Ionis, Janssen, Merck, Pfizer, and USV.

C.K.M.C., T.C.S., Q-L.Y., N.C.C., and J-A.P. have no conflict of interest to declare.

Contributors' Statement

C.K.M.C.: Conceptualization, data curation, formal analysis, methodology, writing–original draft, writing–review and editing. S.C.: Conceptualization, methodology, writing–original draft, writing–review and editing. T.C.S.: Data curation, formal analysis, methodology, writing–original draft, writing–review and editing. Q-L.Y.: Supervision, writing–review and editing. N.C.C.: Supervision, writing–review and editing. V.B.: Supervision, writing–review and editing. J-A.P.: Supervision, writing–review and editing. R.W.: Supervision, writing–review and editing. J.I.W.: Supervision, writing–review and editing. J.W.E.: Conceptualization, methodology, validation, writing–review and editing.

• References

• 1 Ortel TL, Neumann I, Ageno W. et al. American Society of Hematology 2020 guidelines for management of venous thromboembolism: Treatment of deep vein thrombosis and pulmonary embolism. Blood Adv 2020; 4 (19) 4693-4738
• 2 Stevens SM, Woller SC, Kreuziger LB. et al. Antithrombotic therapy for VTE disease: Second update of the CHEST guideline and expert panel report. Chest 2021; 160 (06) e545-e608
• 3 Agnelli G, Buller HR, Cohen A. et al; AMPLIFY-EXT Investigators. Apixaban for extended treatment of venous thromboembolism. N Engl J Med 2013; 368 (08) 699-708
• 4 Weitz JI, Lensing AWA, Prins MH. et al; EINSTEIN CHOICE Investigators. Rivaroxaban or aspirin for extended treatment of venous thromboembolism. N Engl J Med 2017; 376 (13) 1211-1222
• 5 Vasanthamohan L, Boonyawat K, Chai-Adisaksopha C, Crowther M. Reduced-dose direct oral anticoagulants in the extended treatment of venous thromboembolism: A systematic review and meta-analysis. J Thromb Haemost 2018; 16 (07) 1288-1295
• 6 Couturaud F, Schmidt J, Sanchez O. et al; RENOVE Investigators. 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. Lancet 2025; 405 (10480): 725-735
• 7 McBane II RD, Loprinzi CL, Zemla T. et al; EVE trial investigators. Extending venous thromboembolism secondary prevention with apixaban in cancer patients. The EVE trial. J Thromb Haemost 2024; 22 (06) 1704-1714
• 8 Mahé I, Carrier M, Mayeur D. et al; API-CAT Investigators. Extended reduced-dose apixaban for cancer-associated venous thromboembolism. N Engl J Med 2025; 392 (14) 1363-1373
• 9 Higgins JPT, Thomas J, Chandler J. et al, eds. Cochrane Handbook for Systematic Reviews of Interventions, version 6.5 (updated August 2024). Cochrane, 2024. Accessed July 15, 2025 at https://www.cochrane.org/authors/handbooks-and-manuals/handbook
• 10 Moher D, Liberati A, Tetzlaff J, Altman DG. PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med 2009; 6 (07) e1000097
• 11 Sterne JAC, Savović J, Page MJ. et al. RoB 2: A revised tool for assessing risk of bias in randomised trials. BMJ 2019; 366: l4898
• 12 Kaatz S, Ahmad D, Spyropoulos AC, Schulman S. Subcommittee on Control of Anticoagulation. Definition of clinically relevant non-major bleeding in studies of anticoagulants in atrial fibrillation and venous thromboembolic disease in non-surgical patients: Communication from the SSC of the ISTH. J Thromb Haemost 2015; 13 (11) 2119-2126
• 13 Chünemann H, Brożek J, Guyatt G, Oxman A. eds. GRADE Handbook for Grading Quality of Evidence and Strength of Recommendations (updated October 2013). The GRADE Working Group, 2013. Accessed July 15, 2025 at https://gdt.gradepro.org/app/handbook/handbook.html
• 14 R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, 2024. Accessed February 3, 2026 at: https://www.R-project.org/
• 15 Raskob GE, Gallus AS, Sanders P. et al. Early time courses of recurrent thromboembolism and bleeding during apixaban or enoxaparin/warfarin therapy. A sub-analysis of the AMPLIFY trial. Thromb Haemost 2016; 115 (04) 809-816
• 16 Douketis JD, Foster GA, Crowther MA, Prins MH, Ginsberg JS. Clinical risk factors and timing of recurrent venous thromboembolism during the initial 3 months of anticoagulant therapy. Arch Intern Med 2000; 160 (22) 3431-3436
• 17 Ridker PM, Goldhaber SZ, Danielson E. et al; PREVENT Investigators. Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism. N Engl J Med 2003; 348 (15) 1425-1434
• 18 Kearon C, Ginsberg JS, Kovacs MJ. et al; Extended Low-Intensity Anticoagulation for Thrombo-Embolism Investigators. Comparison of low-intensity warfarin therapy with conventional-intensity warfarin therapy for long-term prevention of recurrent venous thromboembolism. N Engl J Med 2003; 349 (07) 631-639
• 19 Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association, American College of Cardiology Foundation. American Heart Association/American College of Cardiology Foundation guide to warfarin therapy. Circulation 2003; 107 (12) 1692-1711
• 20 Samama MM, Martinoli JL, LeFlem L. et al. Assessment of laboratory assays to measure rivaroxaban–an oral, direct factor Xa inhibitor. Thromb Haemost 2010; 103 (04) 815-825
• 21 Graff J, von Hentig N, Misselwitz F. et al. Effects of the oral, direct factor xa inhibitor rivaroxaban on platelet-induced thrombin generation and prothrombinase activity. J Clin Pharmacol 2007; 47 (11) 1398-1407
• 22 Yeo M, Lee SR, Choi EK. et al. Plasma apixaban concentrations and thrombin generation assay parameters in response to dose reduction for atrial fibrillation. Br J Clin Pharmacol 2024; 90 (12) 3221-3231
• 23 Savvari P, Skiadas I, Mavrokefalou E. et al; VICTORIA Study Group. Α 6-month, multicenter, observational study investigating the treatment of venous thromboembolism in Greece (VICTORIA study). Thromb J 2025; 23 (01) 71
• 24 Antonucci E, Migliaccio L, Abbattista M. et al; START POST VTE Investigators. Treatment decision-making of secondary prevention after venous thromboembolism: Data from the real-life START2-POST-VTE register. Clin Appl Thromb Hemost 2020; 26: 1076029620945792
• 25 Palareti G, Antonucci E, Legnani C. et al; START2 Register Investigators. Bleeding and thrombotic complications during treatment with direct oral anticoagulants or vitamin K antagonists in venous thromboembolic patients included in the prospective, observational START2-register. BMJ Open 2020; 10 (11) e040449
• 26 Lyman GH, Carrier M, Ay C. et al. American Society of Hematology 2021 guidelines for management of venous thromboembolism: Prevention and treatment in patients with cancer. Blood Adv 2021; 5 (04) 927-974
• 27 Key NS, Khorana AA, Kuderer NM. et al. Venous thromboembolism prophylaxis and treatment in patients with cancer: ASCO guideline update. J Clin Oncol 2023; 41 (16) 3063-3071
• 28 Alikhan R, Gomez K, Maraveyas A, Noble S, Young A, Thomas M. British Society for Haematology. Cancer-associated venous thrombosis in adults (second edition): A British Society for Haematology Guideline. Br J Haematol 2024; 205 (01) 71-87
• 29 Farge D, Frere C, Connors JM. et al; International Initiative on Thrombosis and Cancer (ITAC) advisory panel. 2022 international clinical practice guidelines for the treatment and prophylaxis of venous thromboembolism in patients with cancer, including patients with COVID-19. Lancet Oncol 2022; 23 (07) e334-e347

Publication History

Received: 16 December 2025

Accepted: 26 January 2026

Accepted Manuscript online:
30 January 2026

Article published online:
11 February 2026