Non-vitamin K antagonist oral anticoagulants (NOACs) after transcatheter aortic valve replacement (TAVR): a network meta-analysis.

backgroundBalancing the risk of thromboembolism and bleeding after transcatheter aortic valve replacement (TAVR) remains clinically challenging. Questions regarding the efficacy and safety of non-vitamin K oral anticoagulants (NOACs) after TAVR still need to be definitively answered.

objectivesTo evaluate the efficacy and safety of NOACs after TAVR in individuals with and without indication for anticoagulation. SEARCH

methodsWe searched CENTRAL, MEDLINE, Embase, Web of Science, ClinicalTrials.gov, and WHO ICTRP on 7 October 2023 together with reference checking and citation searching to identify additional studies. SELECTION CRITERIA: We searched for randomised controlled trials (RCTs) that compared NOACs versus antiplatelet therapy or vitamin K antagonists (VKAs) after TAVR in adults with or without an indication for anticoagulation. DATA COLLECTION AND ANALYSIS: We used standard Cochrane methods and conducted random-effects pair-wise analyses and network meta-analyses (NMAs). Our primary outcomes were all-cause mortality, cardiovascular mortality, stroke, and major bleeding. We used GRADE to assess the certainty of evidence. MAIN

resultsWe included four RCTs with 4808 participants in the NMA. Of these, one compared rivaroxaban versus antiplatelet therapy in people without an indication for anticoagulation after TAVR; one compared apixaban versus antiplatelet therapy in people without an indication for anticoagulation or versus VKA in people with an indication for anticoagulation after TAVR; one compared edoxaban versus VKA in people with an indication for anticoagulation after TAVR; and one compared edoxaban with antiplatelet therapy in people without an indication for anticoagulation after TAVR. The mean age of trial participants was 81 years. Follow-up duration ranged from 6 to 18 months. Overall, we judged the risk of bias in the included trials to be low in all domains except for blinding, which was assessed as high in all four studies. No studies evaluated dabigatran. In people without an indication for anticoagulation, rivaroxaban and apixaban may increase all-cause mortality after TAVR as compared to antiplatelet therapy (rivaroxaban: risk ratio (RR) 1.67, 95% confidence interval (CI) 1.13 to 2.46; studies = 1, participants = 1644; moderate-certainty evidence; apixaban: RR 1.71, 95% CI 0.97 to 3.02; studies = 1, participants = 1049; low-certainty evidence), while edoxaban may result in little or no difference (RR 1.59, 95% CI 0.27 to 9.36; studies = 1, participants = 229; low-certainty evidence). Low-certainty evidence suggests little or no difference between rivaroxaban, apixaban, or edoxaban and antiplatelet therapy in cardiovascular mortality (rivaroxaban: RR 1.28, 95% CI 0.78 to 2.10; studies = 1, participants = 1644; apixaban: RR 1.30, 95% CI 0.64 to 2.65; studies = 1, participants = 1049; edoxaban: RR 7.44, 95% CI 0.39 to 142.38; studies = 1, participants = 229) and between rivaroxaban or edoxaban and antiplatelets in stroke (rivaroxaban: RR 1.19, 95% CI 0.71 to 2.00; studies = 1, participants = 1644; edoxaban: RR 1.06, 95% CI 0.15 to 7.42; studies = 1, participants = 229). While rivaroxaban versus antiplatelets probably increases major bleeding after TAVR (RR 1.98, 95% CI 1.07 to 3.65; studies = 1, participants = 1644; moderate-certainty evidence), there may be little or no difference between apixaban and antiplatelet therapy (RR 1.07, 95% CI 0.70 to 1.64; studies = 1, participants = 1049; low-certainty evidence). It is unclear if edoxaban has an effect on major bleeding, although the point estimate suggests increased bleeding (versus antiplatelets: RR 2.13, 95% CI 0.54 to 8.30; studies = 1, participants = 229; low-certainty evidence). In people with an indication for anticoagulation, low-certainty evidence suggests apixaban or edoxaban may result in little to no difference in our predefined primary efficacy outcomes after TAVR when compared to VKA (all-cause mortality: apixaban: RR 1.02, 95% CI 0.59 to 1.77; studies = 1, participants = 451; edoxaban: RR 0.91, 95% CI 0.69 to 1.20; studies = 1, participants = 1426; cardiovascular mortality: apixaban: RR 1.43, 95% CI 0.76 to 2.70; studies = 1, participants = 451; edoxaban: RR 1.07, 95% CI 0.72 to 1.57; studies = 1, participants = 1426; stroke: apixaban: RR 1.28, 95% CI 0.35 to 4.70; studies = 1, participants = 451; edoxaban: RR 0.83, 95% CI 0.51 to 1.34; studies = 1, participants = 1426). While apixaban may result in a similar rate of bleeding as VKA in this population, edoxaban probably increases major bleeding after TAVR in people with an indication for anticoagulation (apixaban: RR 0.90, 95% CI 0.53 to 1.54; studies = 1, participants = 451; low-certainty evidence; edoxaban: RR 1.44, 95% CI 1.08 to 1.93; studies = 1, participants = 1426; moderate-certainty evidence). AUTHORS'

conclusionsIn people without an indication for oral anticoagulation, rivaroxaban and apixaban may increase all-cause mortality when compared to antiplatelet therapy, while edoxaban may result in little or no difference. There might be little or no difference between rivaroxaban, apixaban, or edoxaban and antiplatelet therapy in cardiovascular mortality, and between rivaroxaban or edoxaban and antiplatelets in stroke. While rivaroxaban probably increases major bleeding following TAVR, there might be little or no difference between apixaban and antiplatelet therapy, and the effect of edoxaban on major bleeding remains unclear. In people with an indication for anticoagulation, apixaban and edoxaban may be as effective as VKA in preventing all-cause mortality, cardiovascular death, and stroke. Apixaban may lead to a similar rate of major bleeding as VKA in this population. However, edoxaban probably increases major bleeding following TAVR when compared to VKA. Our NMA did not show superiority of one NOAC over another for any of the primary outcomes. Head-to-head trials directly comparing NOACs against each other are required to increase the certainty of the evidence.