LAUSR

Based on recent studies, we want to provoke thoughts on how to further optimize research on antiplatelet treatment in patients on oral anticoagulation (OAC). One in three patients suffering from atrial fibrillation (AF) develops concomitant coronary artery disease. Hence, these patients are likely to undergo percutaneous coronary intervention (PCI). AF often leads to an indication for OAC, and after PCI, a dual antiplatelet therapy (DAPT) consisting of aspirin and a P2Y12 inhibitor is recommended. These overlapping recommendations led to many patients with AF after PCI and/or acute coronary syndrome (ACS) being on “triple therapy” (TT) consisting of OAC and DAPT. In TT, both non-vitamin-K-anticoagulants (NOACs) and vitaminK-antagonists (VKA) are used. TT is thought to be effective in preventing thrombotic events, but known to be associated with risks of bleeding. This is of interest, as AF patients are often elderly and, therefore, at greater risk for bleeding complications. Many physicians and scientists questioned the value of aspirin in TT and “dual therapy” (DT) consisting of only OAC and P2Y12 is increasingly used. Recently, another large RCT compared DT vs. TT in AF patients after PCI/ACS. While the previous trials consistently showed decreased bleeding rates in DT, none of these trials was powered to assess thrombotic complications. We aimed to perform a trial sequential analysis and meta-analyze the available evidence. Our search strategy involved screening Medline, Embase, Google Scholar, and Cochrane Register of Controlled Trial. The search was conducted on March 17th in 2019. Two independent reviewers assessed and extracted available data (B.W.&M.L). We included randomized controlled trials investigating humans > 18 years suffering from AF and being treated with either DT or TT. The most recent report was included. The primary endpoint was (definitive or combined definitive/probable) stent thrombosis. Heterogeneity was assessed using the I2 statistic. Pooled event rates were obtained and combined in a meta-analysis, and odds ratios were calculated using a random-effects model (Mantel–Haenszel). Trial sequential analysis was performed assuming two-side testing, type-I error of 5%, power of 80%, and a risk reduction for stent thrombosis of 33%. Most patients received clopidogrel (ranging from 88 to 100%) or ticagrelor (ranging from 0 to 12%) as P2Y12. We included four RCTs in the final meta-analysis [1–4]. We compared all patients on DT vs. TT (both on VKA and on NOAC) from AUGUSTUS trial. Trial sequential analysis revealed that 42% of the required patient number (10073 of 23872) is available. The z-curve did not cross boundaries. Stent thrombosis was statistically not dissimilar (OR 1.50 95% CI 0.96–2.36; p = 0.08; I2 0%, Fig. 1) between DT and TT. Rates of any bleeding (OR 0.44 95% CI 0.38–0.52; p < 0.001; I2 10%) were lower in DT vs. TT. Both rates of minor bleeding (OR 0.49 95% CI 0.39–0.61; p < 0.001 I2 0%) and major bleeding (OR 0.51 95% CI 0.39–0.67; p < 0.001 I2 0%; Fig. 2) were lower in DT compared to TT. Rates of myocardial infarction were not dissimilar between DT and TT (OR 1.18 95% CI 0.93–1.50; p = 0.17; 10%). There were no differences in mortality rates (OR 0.97 95% CI 0.68–1.37; p = 0.85; I2 51%). Sensitivity analysis excluding the WOEST study revealed a trend towards higher rates of stent thrombosis (OR 1.60 95% CI 1.01–2.54; p = 0.04; I2 0%) in DT compared to TT, whereas rates of any (OR * Christian Jung [email protected]