Continued versus Discontinued Oral Anticoagulant Therapy for Unprovoked Venous Thromboembolism: A Target Trial Emulation and Evidence Synthesis

Highlights

• Continuing oral anticoagulants (OACs) beyond initial 90-day treatment in unprovoked VTE patients markedly reduces recurrent VTE and mortality but increases major bleeding risk.
• The net clinical benefit of extended OAC use remains robust across duration categories, OAC types (warfarin, direct OACs), and real-world clinical settings.
• Real-world target trial emulation using large US databases supports individualized risk-benefit assessment favoring extended anticoagulation in many patients with unprovoked VTE.
• Prior randomized trials and cohort studies align with contemporary findings emphasizing the need for nuanced clinical decision-making on treatment duration.

Background

Unprovoked venous thromboembolism (VTE), which occurs without reversible risk factors, poses a substantial clinical challenge due to its heightened risk of recurrence and associated morbidity and mortality. Oral anticoagulants (OACs), including vitamin K antagonists like warfarin and direct oral anticoagulants (DOACs), are the cornerstone for VTE treatment. Standard initial treatment durations commonly span 3 to 6 months. However, the optimal duration of anticoagulation remains debated, given the competing risks of recurrent VTE versus major bleeding complications. Extended anticoagulation may prevent recurrent events but predisposes patients to bleeding, necessitating evidence to guide patient-centered therapeutic decisions.

Recent observational data combined with controlled trials provide new insights into balancing these risks among unprovoked VTE patients in routine clinical practice, particularly regarding anticoagulation beyond 3 months. This review synthesizes the newest evidence, including a large-scale target trial emulation study from US claims data, and summarizes earlier landmark randomized controlled trials and cohort analyses pertinent to duration and safety of OAC therapy.

Key Content

Chronological Evidence Development and Trials on OAC Duration

The 2001 Warfarin Optimal Duration Italian Trial was an important randomized controlled trial comparing 3 months versus 12 months of warfarin therapy in patients with idiopathic deep venous thrombosis (DVT). Extended therapy reduced recurrence during active treatment but did not sustain benefit post-therapy cessation, with similar long-term recurrence rates between groups (N Engl J Med 2001;345:165-9). Importantly, extended anticoagulation was associated with increased risk of non-fatal major bleeding, highlighting the dilemma of continuing anticoagulation indefinitely.

The understanding that unprovoked VTE imparts a particularly high risk of recurrence led to clinical guidelines recommending individualized decisions on indefinite anticoagulation based on risk factors, bleeding risk, and patient preferences. Related cohort studies, such as fibrinolytic parameter evaluations, found limited predictive yield for recurrent events, emphasizing clinical risk assessment over biomarkers (Thromb Haemost 2001;85:390-4).

The real-world application of thrombophilia testing has been variable, with recent large registry data from Italy indicating its limited impact on treatment selection or outcomes, except for conditions such as antiphospholipid syndrome, where vitamin K antagonists remain preferred over DOACs (Blood Transfus 2021;19:244-252). DOACs have increasingly replaced warfarin in practice due to better safety profiles and ease of use.

Target Trial Emulation Study: Methods and Findings (Lin et al., BMJ 2025)

A landmark US study by Lin et al. (2025) used a target trial emulation approach leveraging two large representative datasets: Optum Clinformatics Data Mart and Medicare claims. The study included 30,554 propensity-matched pairs of adults newly diagnosed with unprovoked VTE who initiated OAC therapy (warfarin or DOACs) within 30 days of event and continued treatment for ≥90 days. Patients were categorized into those who continued anticoagulation longer versus discontinued therapy (defined as no prescription refill within 30 days after 90 days).

Primary outcomes examined were recurrent VTE hospitalization (effectiveness) and major bleeding (safety). Secondary outcomes included a composite net clinical benefit measure (recurrent VTE and bleeding) and all-cause mortality. Stratified analyses were conducted by duration of initial OAC treatment (ranging from 90 to ≥1080 days).

Major findings included:

• Recurrent VTE: Continued OAC treatment was strongly protective with adjusted hazard ratio (HR) 0.19 (95% CI 0.13-0.29), indicating an 81% risk reduction compared to discontinuation.
• Major bleeding: Continued treatment increased bleeding risk (HR 1.75, 95% CI 1.52-2.02), confirming the known trade-off.
• Mortality: Mortality rates were significantly lower with continued anticoagulation (HR 0.74, 95% CI 0.69-0.79), a novel and clinically important observation.
• Net Clinical Benefit: Despite bleeding risk, the composite outcome favored continued therapy (HR 0.39, 95% CI 0.36-0.42), consistent across OAC types and treatment durations.

This comprehensive analysis, conducted in a broad real-world population including older adults (mean age 73.9), demonstrates that extended anticoagulation yields a strong net benefit by reducing potentially fatal recurrent VTE and mortality despite bleeding risk.

Comparative Insights on Anticoagulant Types and Real-World Practice

Subanalysis in the Lin study indicated consistent net benefit across warfarin and direct OACs, suggesting class effects. This complements the increasing adoption of DOACs shown in earlier registry data (START2 Registry), where DOACs were more often prescribed in thrombophilic patients with favorable safety and compliance profiles.

Notably, thrombophilia testing, frequently performed despite guideline discouragement, showed limited impact on treatment duration but influenced anticoagulant selection, particularly in antiphospholipid syndrome where warfarin predominates.

Safety Considerations and Bleeding Risk Assessment

Major bleeding remains a critical limitation of prolonged anticoagulation. The Lin et al. study quantified an approximately 1.75-fold increase in major bleeding with continuation beyond 90 days, translating to an absolute adjusted rate difference per 1000 person years of +4.78. However, the overall net clinical benefit favored continuation due to the substantially greater reduction in recurrent VTE and mortality. This underscores the importance of bleeding risk stratification using validated scores and patient-specific factors to optimize management.

Expert Commentary

In aggregate, the evolving evidence from clinical trials, registries, and large-scale cohort emulations supports extended anticoagulation in unprovoked VTE patients beyond the initial treatment phase of 3 months, especially in those at low bleeding risk. The Lin et al. target trial emulation provides a high-quality observational analogue to an RCT design using large real-world data, broadening applicability and addressing limitations of prior trials with smaller samples or select populations.

Though extended therapy clearly reduces recurrent VTE and mortality, the trade-off with increased bleeding requires individualized risk assessment. Incorporation of clinical risk scores, patient preferences, and comprehensive thrombophilia and bleeding work-up remains paramount.

The shift towards DOACs has improved safety and ease of use, supporting longer therapy durations. However, subgroups such as patients with antiphospholipid syndrome still require tailored approaches favoring warfarin.

Unresolved areas include optimal duration beyond several years and the role of biomarkers or imaging in recurrence risk stratification. Additionally, patient adherence, cost-effectiveness, and quality-of-life impacts warrant further investigation.

Conclusion

The balance of evidence indicates that, for patients with unprovoked VTE, continuing oral anticoagulant therapy beyond an initial 90-day period significantly reduces recurrent thrombotic events and mortality despite increased bleeding risk, resulting in a net clinical benefit. This benefit is consistent across anticoagulant types and extended treatment durations, supported by robust real-world data and clinical trial evidence. Careful patient selection and bleeding risk assessment remain essential to optimize individualized treatment strategies. Future research should focus on refining risk prediction, optimizing duration tailored to patient risk profiles, and clarifying the impact of emerging anticoagulant agents.

References

• Lin KJ, Kim DH, Singer DE, Zhang Y, Cervone A, Kehoe AR, Bykov K. Continued versus discontinued oral anticoagulant treatment for unprovoked venous thromboembolism: target trial emulation. BMJ. 2025 Nov 12;391:e084380. doi: 10.1136/bmj-2025-084380. PMID: 41224478; PMCID: PMC12607009.
• Palareti G, et al. Three months versus one year of oral anticoagulant therapy for idiopathic deep venous thrombosis. N Engl J Med. 2001 Jul 19;345(3):165-9. doi: 10.1056/NEJM200107193450302. PMID: 11463010.
• Martinelli I, et al. Fibrinolytic variables in patients with recurrent venous thrombosis: a prospective cohort study. Thromb Haemost. 2001 Mar;85(3):390-4. PMID: 11307802.
• Delluc A, et al. Thrombophilia testing in the real-world clinical setting of thrombosis centres taking part in the Italian Start 2-Register. Blood Transfus. 2021 May;19(3):244-252. doi: 10.2450/2021.0262-20. PMID: 33539283.