Background: We noted the presence of plasma fibrin degradation products in patients treated with recombinant human tumor necrosis factor (TNF) in a phase I trial.
Purpose: To further define this observation, we investigated the effects of TNF on the fibrinolytic system in patients entered in the same trial.
Methods: In the 14 patients studied, fibrinolytic parameters were measured by analyzing blood samples for tissue plasminogen activator and inhibitor at 0, 1, 2, 4, 6, and 18-24 hours after initiation of TNF treatment. We used a chromogenic substrate method to determine activity of plasminogen activator and its inhibitor and an enzyme-linked immunosorbent assay (ELISA) to determine levels of antigen (tissue-type plasminogen activator). Molecular weight was determined by zymographic assay.
Results: TNF treatment was associated with tissue-type plasminogen activator induction within 1 hour of TNF initiation. The plasminogen activator produced was consistent with tissue-type plasminogen activator derived from endothelium as evidenced by molecular weight analysis and ELISA. Moreover, induction of plasminogen activator inhibitor occurred following the release of tissue-type plasminogen activator, and our data suggest a dose-response effect for TNF. At high doses (i.e., 200 and 240 micrograms/m2), there was a more rapid and prolonged release of plasminogen activator inhibitor, which had an inverse relationship with the level of antigenic tissue-type plasminogen activator. Zymographic analysis showed urokinase-type plasminogen activator activity in 13 of 14 patients. In three patients, simultaneous measurements of white blood cells and tissue-type plasminogen activator revealed a temporal association between the TNF-associated rapid granulocytopenia at 30 minutes after TNF initiation and release of tissue-type plasminogen activator antigen.
Conclusions: The results suggest a positive association between TNF and rapid induction of plasminogen activator activity that is consistent with an endothelial product. It is possible that, at high doses, TNF may interact directly with vascular endothelium, leading to rapid and prolonged production of plasminogen activator inhibitor. There was a dose-response effect between TNF and release of tissue-type plasminogen activator. The release of tissue-type plasminogen activator was preceded by granulocytopenia, which may indicate an association between a proposed TNF-induced granulocyte-endothelial interaction in vivo and release of tissue-type plasminogen activator.
Implications: These findings demonstrating the effects of TNF on the fibrinolytic system can be analyzed further in experimental systems to determine the implications for use of this agent as a biological response modifier in cancer therapy.