BACKGROUND: The efficacy and safety of direct oral anticoagulants (DOACs) and vitamin K antagonists (VKAs) during extended anticoagulation for a VTE remains largely unknown, especially in terms of potential survival benefit. The goal of this study was to assess the effects of VKAs and DOACs on overall mortality and VTE-related mortality, as well as VTE recurrence and safety.
METHODS: PubMed, EMBASE, and the Cochrane Library were searched from January 1990 through September 2018 for randomized controlled trials evaluating the effect of extended anticoagulants as secondary prevention for VTE compared with placebo. The primary outcome was the specific effects of standard-intensity VKAs and DOACs on overall mortality.
RESULTS: Sixteen studies (12,458 patients) were included. DOACs were associated with a reduction in overall (risk ratio [RR], 0.48; 95% CI, 0.27-0.86; P = .01) and VTE-related (RR, 0.36; 95% CI, 0.15-0.89; P = .03) mortality, whereas VKAs were not (P > .50). Although VKAs and DOACs similarly prevented recurrent VTE, only VKAs were associated with an increased risk of major bleeding (RR, 2.67; 95% CI, 1.28-5.60; P < .01), resulting in an improved net clinical benefit for DOACs (RR, 0.25 [95% CI, 0.16-0.39; P < .01] vs 0.46 [95% CI, 0.30-0.72; P < .01]; Pinteraction = .05).
CONCLUSIONS: DOACs for extended anticoagulation were associated with a significant reduction in overall mortality compared with observation alone.
TRIAL REGISTRY: PROSPERO; No.: CRD42018088739; URL: https://www.crd.york.ac.uk/prospero/.
The authors note that this meta-analysis does not include studies in which VKAs are compared directly with DOACs. The VKA studies are older and the rates of events for both the VKA and placebo are much higher than in the DOAC studies, so clearly the populations are different. Soe, there are problems with this indirect comparison. With that caveat, it is worth knowing that DOACs are probably safer in the long term.
The objective of this systematic review and meta analysis of 16 randomized controlled trials (RCTs) was done to determine the safety and efficacy (mortality, recurrence) of extended use of vitamin K antagonists (VKAs) and direct oral anticoagulants (DOACS) for patients with venous thromboembolism. The trials compared each anticoagulant to placebo, observation, ASA and infrequently to each other. The RCTs may have preselected patients at low risk of bleeding. Nonetheless, this systematic review is useful as it showed that the DOACs, and not VKAs were associated with a statistically significant reduced risk of mortality and VTE related mortality compared with observation. Both reduced the risk of VTE recurrence but VKAs were associated with a high risk of bleeding. The systematic review is not directive regarding an optimal DOAC, dose, or patient subgroup-it does confirm the efficacy of extended use of anticoagulants for VTE.
This paper is timely and information on survival is quite important.
Meta-analysis adds more emphasis to the role of DOACs in extended VTE treatment. It will be surprising to many that VKAs do not reduce mortality, but most of us already know they do have higher bleeding rates than DOACs. A major limitation of this study: only WODIT, Amplify-ext, and Einstein Choice had a significant number of events (recurrent VTE/mortality), making I squared (heterogeneity) values of dubious validity. All who treat patients with VTE should review this paper.
In this meta-analysis, the authors analyze available data on extended use of VKAs and DOACs. They attempt to perform two Forrest plots - one for VKA, one for DOACs - and then say that any evaluation comparing these two should be done in head-to-head trials. They also point out that 3 of the trials they cite ARE actually head-to-head trials, yet they have parsed them out and analyzed them separately. If indeed this was the question they were trying to answer, they should have included only trials that directly compared the same populations. This would make for a smaller study, but one that is much more accurate. Also, patient population descriptions are missing. I believe the authors would agree that a population of routine medical discharges and a population of postoperative cancer surgery patients would be quite different and should be in table 1. It is not necessary for my fellow hospitalists to know this.
This meta-analysis shows a survival benefit from extended DOAC anticoagulation beyond 3 months after a VTE event (PE or DVT), where previous analysis failed to show improved mortality. This is likely due to exclusion of biased studies and inclusion of recent larger trials. Interestingly, the number of VTE-related deaths only partly explain the reduction in overall mortality. In the absence of detectable CV mortality, unrecognized and presumably preventable PE may be contributing to death in the face of confounding comorbidities. Vitamin K anticoagulation (VKA) was not associated with a reduction in overall mortality, but both DOAC and VKA were associated with decreased relative risk for recurrent VTE events. VKA was associated with increased relative risk for major bleed (RR 2.67), and this risk was not present in the DOAC studies. Fatal bleeding was infrequent, with no significant increase with either DOAC or VKA.
Important summary information with significant drug policy implications.
I`m not an expert in meta-analysis methods but assuming this was done appropriately, the finding of improved overall mortality might be a game-changer for me. It would push me more on the side of extended anticoagulation with DOACs for most patients with unprovoked or modestly provoked clot. Previously, I had thought there was sort of a trade-off between more bleeding and less clotting without a difference in overall mortality.