Abstract
Recent data suggest that bivalirudin, a reversible direct thrombin inhibitor, may be noninferior to heparins (unfractionated heparin/low molecular weight heparin) in providing protection against cardiovascular events, with significantly fewer bleeding complications. Whether this advantage is consistent has not been fully defined. We evaluated cardiac outcomes with bivalirudin vs the heparins in management of acute coronary syndromes (ACS), including patients undergoing percutaneous coronary interventions (PCI). Formal computer-aided searches of electronic databases (MEDLINE, PubMed, Cochrane Controlled Trials Registry) were performed by scrutiny of the reference lists of trials and review articles, abstracts, meeting proceedings, and the manufacturers of direct thrombin inhibitors. Five randomized controlled trials (BAT, 1995; CACHET, 2002; REPLACE-2, 2003; REPLACE-1, 2004; and ACUITY, 2006) comparing bivalirudin to the heparins in patients with ACS, including patients undergoing PCI, were identified. The meta-analysis consisted of 25 457 patients (bivalirudin, 15 077; heparins, 10 380). The primary safety end point was major bleeding, defined as intracranial, intraocular, or retroperitoneal hemorrhage; clinically overt blood loss leading to a hemoglobin drop exceeding 3 g/dL (or 10% of hematocrit) and transfusion of 2 or more units of whole blood or packed red blood cells. The combined relative risks (RR) across all of the studies and the 95% confidence intervals of death, myocardial infarction (MI), and revascularization (bivalirudin vs heparins) were computed using the Mantel-Haenszel fixed-effect model, whereas the random-effect model was used for major bleeding. A 2-sided α error < .05 was considered to be significant. There were no significant differences in patient characteristics between the 2 groups. Compared to the heparins, the risk of death, MI, revascularization, and composite ischemic end points were similar with bivalirudin monotherapy. However, the risk of major bleeding was significantly lower with bivalirudin use (RR = 0.553; 95% CI = 0.402-0.761;