Access to POEMs and Essential Evidence Plus will no longer be included in CMA membership as of Dec. 1, 2023.
Clinical Question
For patients with venous thromboembolism, is extended anticoagulation more beneficial than a limited duration of anticoagulation?
Bottom line
This study generally supports extended anticoagulation for 6 months to 2 years for patients with unprovoked venous thromboembolism (VTE), especially unprovoked pulmonary embolism (PE). There are several caveats, however. The studies of warfarin were largely independently funded, while those of direct oral anticoagulant (DOAs) were industry funded—and there is good evidence that industry-funded studies tend to find more favorable outcomes. It is also difficult to know which subgroup of patients benefits the most. A randomized trial with 4 arms (short-term vs long-term anticoagulation, DOA vs warfarin) is needed to definitively answer this question. 1a-
Reference
Study design: Meta-analysis (randomized controlled trials)
Funding: Self-funded or unfunded
Setting: Outpatient (any)
Synopsis
The most recent American College of Chest Physicians guidelines for antithrombotic therapy for VTE (Chest 2016;149(2):315-352) recommend extended anticoagulation for patients with an initial unprovoked proximal deep vein thrombosis (DVT) or PE and a low/moderate bleeding risk, as well as in those with a second unprovoked episode of VTE and a low/moderate bleeding risk. This meta-analysis identified 16 studies with 12,458 patients who were randomized to receive warfarin or a DOA for 3 months to 6 months or for a more extended period of anticoagulation (range = 3 to 25 additional months). Most patients had an unprovoked VTE: 7 of 16 studies included only patients with unprovoked VTE, and 4 other studies' populations were more than 50% patients with unprovoked VTE. The percentage with PE as the index event varied from 0 to 100% (4 studies), but those studies typically had 20% to 40% of patients with PE. Results were stratified by whether the anticoagulant was a vitamin K antagonist like warfarin or a DOA. Overall mortality was not decreased with long-term warfarin, but was decreased with long-term DOA (relative risk [RR] 0.48, 95% CI 0.27 - 0.86). However, the absolute risk difference was tiny (pooled values of 0.42% vs 0.77%; P < .05; number needed to treat [NNT] = 285 over 1 to 3 years' follow-up). The same pattern was seen for VTE-related mortality: no difference for warfarin and a reduction with a DOA (RR 0.36, 0.15 - 0.89). VTE recurrence was reduced with both warfarin and DOA (RR 0.21 - 0.22 for DOA and warfarin), while major bleeding was increased (RR 2.7 for warfarin and 1.6 for DOA). The net clinical benefit that incorporated recurrent VTE and major bleeding was reduced for both warfarin treatment (RR 0.46, 0.30 - 0.72; NNT = 16) and DOA (RR 0.25, 0.16 - 0.39; NNT = 15).
Reviewer
Mark H. Ebell, MD, MS
Professor
University of Georgia
Athens, GA