Coagulation in liver failure : the role of thromboelastometry and fibrinogen

Abstract

Liver failure is undeniably accompanied by distortions of the coagulation system. The modern theory of coagulation in patients with compensated liver disease (acute or chronic) states that the coagulation system is in a rebalanced state, also displaying prothrombotic tendencies particularly towards the end stages. Conventional coagulation tests are proved inappropriate in the evaluation of bleeding or thrombotic risks in these patients. Viscoelastic tests, such as thromboelastometry, are used clinically as point-of-care testing to determine the balance of coagulation and have proven particularly effective in the treatment of critical situations with bleeding difficult to control. Also, there have been reports that viscoelastic testing might be used for prognosis purposes in liver cirrhosis. A thromboelastometric characterization of an unselected population with an indication for liver transplantation was assumed to be useful. Additionally, a clarification of blood coagulation abnormalities in conjunction with the liver dysfunction following massive liver resections was needed.

The aim of studies 1 and 2 was to characterize coagulation status in patients with liver conditions using thromboelastometry. In study 1 we included patients with stable chronic liver disease with an indication for liver transplantation. Aside from a viscoelastic description, the primary aim of study 1 was to assess thromboelastometry usefulness in assessing the prognosis in these patients and eventually be used to improve the existing scoring systems. In general, the patients were normo- and hypocoagulable with no significant signs of hypercoagulability expressed through thromboelastometry. No relation could be found between thromboelastometry and the Child-Pugh and MELD scores. The conclusion of study 1 was that we could not find any use for thromboelastometry in evaluating the gravity of the stable chronic liver disease with an indication for liver transplantation. In study 2 we included patients who underwent hemihepatectomies and extended hemihepatectomies. Using thromboelastometry we found that postoperatively the coagulation system was rebalanced apparently in the same manner as in chronic liver disease and that the bleeding risk signalled by PT-INR was groundless. We also found that the postoperative trends of plasma fibrinogen concentration were related to the size of resection.

The findings in study 2 made us hypothesize that a deficiency in synthesis due to the loss of liver mass may be the factor which affects the plasma fibrinogen concentrations postoperatively. To explore this hypothesis, we needed to quantitatively assess the de novo synthesis of fibrinogen (and albumin) in these patients. To this end, in study 3, we adjusted the isotope tracer flooding dose technique to enable repetitive measurements in longitudinal studies. The technique described in study 3 was then used in study 4 to determine the postoperative synthesis rates of fibrinogen and albumin in patients undergoing hemihepatectomies and pancreatectomies (as controls). An extraordinary capacity of the liver to increase fibrinogen synthesis on postoperative day one was discovered. In this study, we found that the loss of liver mass through elective hemihepatectomies did not incapacitate a sharp increase in the synthesis rates of fibrinogen compared to pancreatectomies. Plasma fibrinogen concentrations may therefore reflect the dynamic changes in synthesis related to fibrinogen utilization. The postoperative synthesis course maintained effective fibrinogen levels for balanced coagulation. Albumin synthesis rates confirmed good posthepatectomy synthetic function in the remnant liver tissue and eventually in the regenerated liver tissue.