Background Adjunctive therapy for thrombolysis in acute myocardial infarction consists of platelet inhibition with aspirin and thrombin inhibition with heparin. Thrombin inhibition may be improved by the use of hirudin as indicated by experimental and phase II clinical studies. The randomized, double-blind phase III r-Hirudin for Improvement of Thrombolysis study (HIT III) compared a recombinant hirudin (HBW 023) with heparin. The primary end point was the incidence of death or reinfarction.Methods and Results Seven thousand patients with acute myocardial infarction and a duration of symptoms of less than 6 hours were to be randomized to receive intravenous heparin (70 IU/kg body wt bolus and 15 IU x kg sup -1 x h sup -1) or hirudin (0.4 mg/kg body wt bolus and 0.15 mg x kg sup -1 x h sup -1) infused over 48 to 72 hours and adjusted to an activated partial thromboplastin time of 2 to 3.5 times baseline values. In a pilot phase, 1000 patients receiving front-loaded alteplase for thrombolysis were to be recruited by 93 German centers. After enrollment of 302 patients, the trial was stopped after an increased rate of intracranial bleeding was observed in the hirudin group (5 of 148, 3.4%) compared with the heparin group (0 of 154). The overall stroke rate was 3.4% in the hirudin group and 1.3% in the heparin group. Other major bleeding occurred in five versus three patients and ventricular rupture occurred in three versus one patient in the hirudin and heparin groups, respectively. There were 19 in-hospital deaths, with 13 of them from the hirudin group.Conclusions Although the number of patients was too small for a definite benefit-risk assessment, at the dosage tested, hirudin in combination with front-loaded alteplase and aspirin may be associated with an increased rate of intracranial hemorrhage. Our findings are consistent with the observations of the GUSTO-II and TIMI-9 trials, where higher doses of another recombinant hirudin were used. Therefore, the therapeutic range of hirudin as an adjunct to thrombolysis may be smaller than previously thought, and reappraisal of dose finding should be considered. (Circulation. 1994;90:1638-1642.)