Randomized Evaluation of PCI Linking Angiomax to Reduced Clinical Events - REPLACE 2


The REPLACE 2 trial compared the use of bivalirudin during elective or urgent percutaneous coronary intervention (PCI) with: 1) unfractionated heparin with planned glycoprotein (GP) IIb/IIIa inhibitor use, and 2) unfractionated heparin alone, based on historical controls from the EPISTENT and ESPRIT trial.


The primary hypothesis was noninferiority of bivalirudin compared with heparin, plus planned GP IIb/IIIa inhibitor use during PCI for the quadruple composite endpoint of 30-day death, myocardial infarction (MI), urgent revascularization or in-hospital major hemorrhage. The second hypothesis was superiority of bivalirudin compared with heparin alone, based on historical controls from the EPISTENT and ESPRIT trials for the triple endpoint of death, MI, or urgent revascularization.

Study Design

Study Design:

Patients Enrolled: 6010.
Mean Follow Up: 30 days.
Mean Patient Age: mean age 62.6 years
Female: 25%

Patient Populations:

Age >21 years and planned to undergo PCI with an approved device.


Primary or rescue PCI for acute MI; UFH within 6 hours; low molecular weight heparin within 8 hours; platelet count <100,000.

Primary Endpoints:

Quadruple composite of death, MI, severe ischemia requiring urgent revascularization by 30 days, or in-hospital major bleed

Secondary Endpoints:

Triple composite of death, MI, severe ischemia requiring urgent revascularization by 30 days

Drug/Procedures Used:

Unfractionated heparin (UFH)(65 U/kg bolus) with planned glycoprotein (GP)IIb/IIIa inhibitor use (abciximab or eptifibatide) during PCI compared with bivalirudin (0.75 mg/kg bolus and 1.75 mg/kg/hr infusion during the PCI) and provisional GP IIb/IIIa use (abciximab or eptifibatide) during PCI.

Concomitant Medications:

Aspirin; clopidogrel pretreatment at investigator discretion.

Principal Findings:

The quadruple endpoint (death/MI/urgent revascularization at 30 days or in-hospital major hemorrhage) occurred in 10.0% of patients in the UFH + GP IIb/IIIa arm compared with 9.2% in the bivalirudin arm (p=0.32, OR=0.917, 95% CI 0.772-1.089). The rate of the triple endpoint (death/MI/urgent revascularization) was 7.1% in the UFH+GP IIb/IIIa arm compared with 7.6% in the bivalirudin arm (p=0.40, OR=1.088). The trial satisfied the superiority comparison between bivalirudin and UFH alone based on historical controls. There were no significant differences between the UFH+GP IIb/IIIa arm and the bivalirudin arm in the individual endpoints of death (0.4% vs 0.2%), MI (6.2% vs 7.0%), or urgent revascularization (1.4% vs 1.2), but there was a significant 41% relative reduction in major bleeding (4.1% vs 2.4%, p<0.001). Provisional use of GP IIb/IIIa inhibitor occurred in 7.2% of patients in the bivalirudin arm.


Among low to moderate risk patients undergoing PCI, the REPLACE-2 trial met the prespecified hypothesis of non-inferiority of the quadruple endpoint comparing bivalirudin to UFH + GP IIb/IIIa inhibitor during elective or urgent PCI, as well as superiority of bivalirudin compared with UFH alone based on historical controls from the EPISTENT and ESPRIT trials. The non-inferiority established between the bivalirudin arm and the UFH + GP IIb/IIIa arm in the quadruple composite endpoint was primarily driven by a reduction in major bleeding. The definition of major bleeding used in this trial differed from that used in other trials which is TIMI major bleeding. When the TIMI major bleeding criteria are applied, the rates of bleeding do not differ as discussed in the editorial by Dr. Elliott Antman. Also as discussed by Dr. Antman, the composite endpoint in this trial combines efficacy and safety data (death, MI, and revascularization are lumped together with safety, bleeding). As a result of safety and efficacy being lumped together, if the safety signal outweighs the efficacy signal, drugs that are more effiacious can be considered to be inferior. As a class, the GPIIbIIIa inhibitors have been associated with improved mortality in meta analyses whereas the direct thrombin inhibitors have not. The ACTs in this trial were 325 seconds in the heparin arm, which are higher than those (260-280 secs) used in other trials.


JAMA 2003; 289:853-863. Presented at AHA 2002 by Michael Lincoff, M.D.

Clinical Topics: Anticoagulation Management, Dyslipidemia, Heart Failure and Cardiomyopathies, Invasive Cardiovascular Angiography and Intervention, Lipid Metabolism, Nonstatins, Heart Failure and Cardiac Biomarkers

Keywords: Myocardial Infarction, Coronary Disease, Heparin, Immunoglobulin Fab Fragments, Platelet Membrane Glycoprotein IIb, Hirudins, Percutaneous Coronary Intervention, Fatty Acids, Omega-3, Recombinant Proteins, Peptide Fragments, Platelet Glycoprotein GPIIb-IIIa Complex

< Back to Listings