Comparative efficacy and safety of anticoagulant strategies for acute coronary syndromes. Comprehensive network meta-analysis of 42 randomised trials involving 117,353 patients

International guidelines differ in strengths of recommendation for anticoagulation strategies in acute coronary syndromes (ACS). We performed a comprehensive network meta-analysis (NMA) of randomised controlled trials (RCTs) to investigate the comparative efficacy and safety of parenteral anticoagulants in ACS. MEDLINE, Cochrane, EMBASE, Google Scholar, major cardiology websites, and abstracts/presentations were searched. Six treatments were identified: 1) unfractionated heparin (UFH) + glycoprotein IIb/IIIa inhibitor (GPI) [UFH+GPI], 2) UFH±GPI, 3) bivalirudin, 4) low-molecular-weight heparins (LMWHs), 5) otamixaban, and 6) fondaparinux. Prespecified outcomes (death, myocardial infarction [MI], revascularisation, major bleeding [MB], minor bleeding, and stent thrombosis [ST]) were evaluated up to 30 days. Forty-two RCTs involving 117,353 patients were included. No significant differences in mortality rates were found among strategies. Compared to UFH+GPI, bivalirudin reduced the odds of MB but increased the odds of ST and MI. LMWHs vs bivalirudin reduced MI risk at the price of MB excess. UFH±GPI significantly increased the odds of MI vs LMWHs, of ST vs UFH+GPI, and of MB vs bivalirudin. Reduced ST risk with otamixaban vs UFH±GPI and vs bivalirudin was offset by a marked 2.5- to four-fold MB excess. Fondaparinux showed an intermediate profile. Results for ST-segment elevation MI were consistent with the overall findings. Early anticoagulant strategies for ACS differ in efficacy and safety, with UFH+GPI and LMWHs reducing ischaemic but increasing bleeding risk, and bivalirudin reducing MB but increases MI and ST. The findings support individualised therapy based on patients' bleeding and ischaemic risks.