BACKGROUND: Whether myocardial reperfusion obtained with thrombolysis or primary angioplasty is associated with a similar recovery of function and with the same risk of hemorrhagic infarction is unknown. We evaluated the effects of mechanical and pharmacological reperfusion (with or without a plasma lytic state) on infarct size, myocardial hemorrhage, and left ventricular (LV) function in a canine model. METHODS AND RESULTS: Six groups of six dogs were subjected to balloon occlusion of the left anterior descending coronary artery (LAD) followed by 2 hours of reperfusion. The study had a two-by-three factorial design with two occlusion periods (90 and 240 minutes) and three different reperfusion strategies (placebo, 0.4 mg/kg recombinant tissue plasminogen activator, and 40 microg/kg recombinant staphylokinase). In a seventh control group, LAD occlusion was maintained without reperfusion. All dogs received aspirin and heparin. A systemic lytic state was present in staphylokinase-treated dogs. Planimetry of LV slices showed larger infarcts (percent of area at risk) and more hemorrhage (percent of IA) after 240 minutes of occlusion than after 90 minutes of occlusion (54+/-17% versus 37+/-18% and 52+/-27% versus 29+/-27%, respectively; P<.01 for both comparisons), with no significant difference among treatments. Hemorrhage was not observed in the control group without reperfusion. LV angiography showed no differences in global and regional LV function between mechanical and pharmacological reperfusion. CONCLUSIONS: In this experimental model, hemorrhagic infarctions of similar extent were observed after both pharmacological and mechanical reperfusion. The extent of hemorrhage was increased by the delay in reperfusion but not by the presence of a lytic state.