Reocclusion of the coronary artery occurs after thrombolytic therapy of acute myocardial infarction despite routine use of the anticoagulant heparin. However, heparin is inhibited by platelet activation, which is greatly enhanced in this setting. Consequently, it is unclear whether thrombin induces acute reocclusion. To address this possibility, we examined the effect of argatroban [MCI9038, (2R,4R)-4-methyl-1-[N alpha-(3-methyl-1,2,3,4-tetrahydro-8- quinolinesulfonyl)-L-arginyl]-2-piperidinecarboxylic acid], a specific thrombin inhibitor, on the response to tissue-type plasminogen activator in a closed-chest canine model of coronary thrombosis. MCI9038 prolonged the thrombin time and shortened the time to reperfusion (28 +/- 2 min vs. 59 +/- 7 min in controls; mean +/- SEM, n = 5, P less than 0.01). At the highest dose, 2.5 mg/kg per hr, complete reocclusion was prevented in four of the five experimental animals, whereas reocclusion occurred in all five controls. However, reperfusion was complicated by cycles of decrease flow, which were abolished by the thromboxane A2 antagonist, GR32191. GR32191 at 1 mg/kg combined with MCI9038 at 0.5 mg/kg per hr prevented reocclusion, whereas, at these doses, either drug alone was without effect. In addition, thromboxane A2 biosynthesis, determined as excretion of its metabolite 2,3-dinor-thromboxane B2, was increased after reperfusion at all doses of MCI9038. These data demonstrate that thrombin impairs thrombolysis induced by tissue-type plasminogen activator in vivo and induces acute reocclusion. Furthermore, the response to thrombin inhibition may be impaired by continued formation of thromboxane A2.