Intravenous administration of thrombin inhibitors, such as hirudin, has been shown to decrease the frequency of coronary artery reocclusion after thrombolysis. However, recent findings in large clinical trials in patients with unstable angina and myocardial infarction have failed to demonstrate a sustained antithrombotic effect after cessation of drug treatment. These findings indicate a need for a prolonged antithrombotic regimen, preferably an orally active thrombin inhibitor. To test the hypothesis that a regimen consisting of oral thrombin inhibitor will delay or prevent the formation of occlusive clot, anesthetized dogs were given saline (n = 9) or a single dose of a novel active site low-molecular-weight thrombin inhibitor melagatran by nasogastric tube (1.5 mg/kg, n = 6; 2.5 mg/kg, n = 6), and 15 min later, a potent thrombogenic stimulus in the form of anodal current (100 microA) was applied to the intimal surface of the narrowed left anterior descending coronary artery (LAD). All saline-treated dogs developed stable thrombus, indicated by zero flow at 34 +/- 7 min after initiation of direct current. On the other hand, one of the six dogs given high-dose melagatran did not develop thrombotic occlusion of the LAD during the entire 4 h of observation. Mean time to occlusive thrombus formation in 11 other dogs was prolonged 4-5 times as compared with that in the saline-treated dogs (p < 0.001). Spontaneous thrombolysis was observed in three of 11 dogs after initial clot formation. Overall, the coronary artery was patent for 68% (low dose) and 75% (high dose) of the observation period in melagatran-treated dogs (vs. 14% of observation period in saline-treated dogs). Peak plasma concentration was 0.87 +/- 0.22 microM in dogs given low-dose and 1.38 +/- 0.30 microM in dogs given high-dose melagatran. The activated partial thromboplastin time (aPTT) increased 1.5-fold at peak plasma concentration of melagatran. These observations imply (a) thrombin generation plays a critical role in thrombus formation in narrowed coronary arteries, (b) oral melagatran prevents or delays thrombus formation, whereas the aPTT is only modestly prolonged, and (c) the thrombus formed in the presence of melagatran is prone to spontaneous lysis in this canine model of coronary thrombosis.