Increased Plasma Levels of Tissue Factor Associated to Thrombomodulin Resistance in Patients with Advanced Liver Cirrhosis

Background: Despite reduced coagulation factors, it was recently described that patients with cirrhosis have an increased thrombotic tendency. Therefore, assays predicting the risk of bleeding or thrombosis are needed.

Aims: As the Tissue Factor (TF) pathway plays a major role in the initiation of the coagulation, we evaluated the equilibrium between Tissue Factor activity (TFa) and Tissue Factor Pathway Inhibitor (TFPI) and their relationship with the thrombin generation (TG) assay in 42 cirrhotic patients. TG was realized in the absence and presence of thrombomodulin (TM), since it was previously evidenced a resistance to TM in cirrhotic patients.

Methods: Citrated blood samples of 42patients with confirmed cirrhosis classified according to the Child score (A,B,C), free of familial history of thromboembolism or thrombophilia and not treated by anticoagulant were analyzed. TG was triggered by 5 pM of TF in the presence/absence of thrombomodulin (4 µm), and 4 µM of phospholipids (Cat Assay, Stago,France). TFa was determined by a home-made assay, free TFPI (fTFPI) by the Asserachrom Free TFPI (Diagnostica Stago, France). Protein C (PC) and S (PS) quantified by Staclot Protein C and S assays (Diagnostica Stago, France). Results were compared to a group of 30 healthy subjects (CT).

Results: Main results are summarized in the table. For TG assay, only the Lag-Time (LT) and the Endogenous Thrombin Potential (ETP) were indicated.

TFa was significantly (p < °0.01) increased in comparison with CT, and patients with the most severe disease (Child C) have higher levels (p < 0.001) than patients with the lowest Child score (A). In contrast, fTFPI levels were not significantly different from CT, whatever the Child score was. A significant decrease of ETP (without TM) was observed for cirrhotic patients with the lowest Child scores (A and B) in comparison with CT, whereas the lag-time (with and without TM) and ETP with TM were not different from CT. Therefore, the ratios of ETP with TM/ETP without TM increased significantly, indicating a resistance to the anticoagulant activity of TM, in relation to the severity of the disease. This resistance was in part explained by the decrease levels of PC since there was a significant negative correlation (r = - 0.35, p = 0.03) between PC levels and the ratio ETP + TM/ETP - TM. PC levels were also inversely correlated to ETP levels, in the absence (r = - 0.54, p = 0.0006) or presence of TM (r = - 0.53, p = 0.0006). PC levels were also inversely correlated to TFa levels (r = - 0.39, p = 0.01), suggesting that PC decrease could be related to a defective synthesis, but also possibly by a consumption due to the activation of the coagulation cascade by the TF. In contrast, no significant correlation were observed for PS.

Conclusion: In addition toaresistance to the anticoagulant activity of the PC pathway, cirrhotic patients with the more severe disease stage (Child C) have high levels of TF activity which is not counterbalanced by TFPI. This could contribute to the higher prevalence of thrombotic disease in these patients. The origin of TF remains to be established, but could come from Kuppfer cells in reaction to the liver injury. These results may have clinical implication for the treatment or prophylaxis of thrombosis in cirrhotic patients.