The Thrombolysis and Thrombin Inhibition in Myocardial Infarction (TIMI) 9A trial compared the efficacy and safety of intravenous hirudin with heparin as adjunctive therapy to thrombolysis and aspirin in patients with acute myocardial infarction. The primary safety end point was the occurrence of major hemorrhage or anaphylaxis. Based on experience in phase II trials, TIMI 9A used a hirudin bolus of 0.6 mg/kg followed by a fixed-dose 96-hour infusion of 0.2 mg/kg per hour. A modified weight-adjusted heparin regimen was used (5000-U bolus and infusion of 1000 U/h for patients < 80 kg or 1300 U/h for patients > or = 80 kg) with titration to a target activated partial thromboplastin time (aPTT) of 60 to 90 seconds. Because rates of hemorrhage in both treatment arms were higher than expected, randomization was suspended in TIMI 9A after 757 patients had been enrolled. Intracranial hemorrhage occurred in 1.7% of patients treated with hirudin and 1.9% of those treated with heparin (P = NS). Major spontaneous hemorrhage at a nonintracranial site occurred more frequently in hirudin--than in heparin-treated patients (7.0% versus 3.0%; P = .02), whereas major hemorrhage at instrumented sites was similar (5.2% in both hirudin and heparin groups). Patients who developed a major hemorrhage were older (P < .001) and had higher aPTT values, especially in the first 12 hours after thrombolysis (P = .001). The rate of major spontaneous hemorrhage for both heparin and hirudin in TIMI 9A was higher than that seen in TIMI 5, TIMI 6, and GUSTO 1. This was possibly a result of high levels of anticoagulation at the doses of heparin and hirudin used, low previous estimates of the hemorrhage risk at the doses of hirudin used in TIMI 9A due to the relatively small number of patients receiving that dose in earlier studies, and enrollment of patients at higher risk of hemorrhage. Because a prolonged aPTT was associated with an increased risk of major hemorrhage in both heparin- and hirudin-treated patients, it now appears important to monitor aPTT on a regular basis when using either antithrombin to identify those patients who require downward adjustment of the infusion. TIMI 9B has therefore been configured with a lower hirudin bolus (0.1 mg/kg) and infusion (0.1 mg/kg per hour) and lower heparin infusion (1000 U/h without weight adjustment). Infusions of both antithrombins will be titrated to a target aPTT of 55 to 85 seconds.