Thrombus extension in patients with venous thromboembolism is due to the accretion of fibrin onto existing thrombi. Extension is promoted by both circulating and thrombus-bound thrombin, which convert fibrinogen to fibrin. Heparin is an effective antithrombotic agent, but it requires continuous administration to achieve persistent inhibition of thrombus extension. Heparin is highly effective in inhibiting fluid phase thrombin, but is a relatively ineffective inhibitor of thrombus- bound thrombin. Hirudin, unlike heparin, inactivates both circulating and fibrin-bound thrombin and, therefore, has the potential to prevent thrombus extension even after a short course of treatment. The aim of this study was to evaluate the time course of the accretion of new fibrin onto preexisting rabbit jugular vein thrombi after a 3-hour infusion of saline, heparin, and hirudin. Heparin and recombinant (r)- hirudin (CGP 39399) were infused at doses that doubled the activated partial thromboplastin time (aPTT). At the end of the 3-hour infusions in rabbits treated with saline, heparin (0.75 mg/kg), or r-hirudin (1.25 mg/kg), accretion of 125I-fibrinogen was 59 +/- 5 micrograms, 34 +/- 4 micrograms, and 21 +/- 2 micrograms, respectively (heparin and r- hirudin v saline, P less than .01; r-hirudin v heparin, P less than .01). Three hours after the end of the infusions, the accreted 125I- fibrinogen in the saline-, heparin-, and hirudin-treated animals was 89 +/- 6 micrograms, 51 +/- 7 micrograms, and 23 +/- 3 micrograms, respectively; 9 hours after the end of the infusions, the accreted 125I- fibrinogen was 112 +/- 9 micrograms, 82 +/- 7 micrograms, and 25 +/- 3 micrograms, respectively. aPTT and thrombin clotting time (TCT) returned to the baseline value 90 minutes after the end of heparin or r- hirudin infusion. During in vitro experiments, human fibrin clots previously incubated in human plasma containing r-hirudin did not promote fibrinopeptide A (FPA) generation when washed and then incubated in human plasma in the absence of thrombin inhibitors. This persistent inhibition of FPA production was not observed after incubation in human plasma of human plasma clots preincubated with heparin. We conclude that heparin is effective in inhibiting thrombus extension while it is present in the circulation, but that this effect is rapidly lost after its plasma clearance. In contrast, the antithrombotic activity of r-hirudin is sustained beyond its plasma clearance, presumably because of its ability to inactivate thrombus- bound thrombin. Our findings indicate that r-hirudin might be an effective antithrombotic agent even when used for short periods.