1. Disseminated intravascular coagulation frequently accompanies Gram-negative sepsis and may contribute to widespread deposition of microthrombi. Besides the endotoxin-induced activation of coagulation, an important role for the fibrinolytic system has been postulated. The precise mechanisms underlying these fibrinolytic changes during endotoxaemia are not known but have been suggested to be mediated directly by cytokines or secondary to thrombin generation.
2. In the present study we have delineated in detail the fibrinolytic response to a bolus injection of endotoxin in non-human primates and analysed the contribution of cytokines and thrombin generation to the endotoxin-induced release of tissue-type plasminogen activator and plasminogen activator inhibitor 1. Chimpanzees received a bolus injection of endotoxin alone or in combination with blocking monoclonal antibodies directed against tumour necrosis factor or interleukin 6 or in combination with pentoxifylline. Furthermore, to assess the effect of coagulation activation on the activation of fibrinolysis, another group of chimpanzees received endotoxin in combination with either anti-tissue factor antibodies or recombinant hirudin.
3. Infusion of endotoxin induced a rapid increase in plasminogen activator activity and tissue-type plasminogen activator antigen levels and subsequent plasmin generation, reaching peak levels 2 h after endotoxin administration. Plasminogen activator inhibitor 1 levels remained constant for the first 2 h, after which time a steep increase was observed. Plasminogen activator activity and plasmin generation decreased simultaneously with the rise in plasminogen activator inhibitor 1 levels. Fibrinolytic activity remained suppressed during the remainder of the study owing to sustained increased levels of plasminogen activator inhibitor 1. The administration of pentoxifylline strongly attenuated the release of tissue-type plasminogen activator and plasminogen activator inhibitor 1, whereas the antitumour necrosis factor antibodies blocked the fibrinolytic response entirely. In contrast, interleukin 6-neutralizing antibodies did not affect the fibrinolytic response. Although endotoxin-induced generation of thrombin was completely prevented by the administration of tissue factor-neutralizing antibodies or by hirudin, no effect on the fibrinolytic response was observed. 4. In conclusion, the present study demonstrates that, after initial activation during endotoxaemia, fibrinolysis is inhibited by high levels of plasminogen activator inhibitor 1 and shows the pivotal role of tumour necrosis factor in the release of tissue-type plasminogen activator and plasminogen activator inhibitor 1 during endotoxaemia, which appears to be regulated independently of coagulation activation.