In-hospital bleeding in the era of newer P2Y12 inhibitors in real-life patients with acute coronary syndrome: results from the GReek AntiPlatElet (GRAPE) registry

Abstract

Purpose: Data on in-hospital bleeding events with the concurrent use of clopidogrel, prasugrel and ticagrelor in real-life practice are lacking. Methods: In the context of GReek AntiPlatelet rEgistry (GRAPE), a prospective, observational, multicenter cohort study conducted in in 8 PCI-capable hospitals in Greece, we investigated in-hospital bleeding events (Bleeding Academic Research Consortium -BARC classification) in ACS patients undergoing percutaneous coronary intervention (PCI). Results: Between January and December 2012, 1280 patients were included in analysis (82.8% men, 22.7% diabetics, 53.5% admitted with ST elevation myocardial infarction (STEMI), with a mean age of 61.8±12.4 years). Overall, in-hospital bleeding BARC type 1, 2 3a and 3b occurred in 54 (4.2%), 31 (2.4%), 21 (1.6%) and 5 (0.4%) patients respectively. No fatal bleeding event occurred. In the subgroup of patients who were initially treated either with clopidogrel (N=478), prasugrel (N=92) or ticagrelor (N=228) without switching thereafter, multivariate analysis (backward logistic regression) revealed age≥75 years and recent/recurrent gastrointestinal (GI) bleeding as predictive factors of in-hospital bleeding BARC of any type (OR=2.14, 1.21-3.80 95% CIs, p=0.009 and OR=6.48, 1.13-37.2 95% CIs, p=0.04 respectively). In 215 propensity-matched pairs of patients, rate of in-hospital bleeding BARC of any type did not differ significantly between clopidogrel vs prasugrel or ticagrelor -treated patients (3.7% vs 7.9%, p=0.098). Conclusions: In a real-world PCI-treated ACS population, older age and previous GI bleeding are associated with an increased risk of low to moderate severity in-hospital bleeding events. A trend towards more bleeding with the newer agents is apparent.