Causes and consequences of portal vein thrombosis in 1,243 patients with cirrhosis: results of a longitudinal study.

Abstract

In cirrhosis, portal vein thrombosis (PVT) could be a cause or a consequence of the progression of liver disease. We analyzed data from a prospective trial of ultrasound screening for hepatocellular carcinoma in order to identify risk factors for and the impact of PVT in patients with cirrhosis. In all, 1,243 adults with cirrhosis without PVT were enrolled from 43 liver units in France and Belgium between June 2000 and March 2006. The mean follow-up was 47 months. Doppler ultrasonography was used to check the portal vein. Progression of liver disease was defined by the development of: ascites, hepatic encephalopathy, variceal bleeding, prothrombin <45%, serum bilirubin >45 μmol/L, albumin <28 g/L, and/or creatinine >115 μmol/L. G20210A prothrombin and factor V gene mutations were assessed in sera stored at three large centers. The 5-year cumulative incidence of PVT was 10.7%. PVT was mostly partial and varied over time. The development of PVT was independently associated with baseline esophageal varices (P = 0.01) and prothrombin time (P = 0.002), but not with disease progression before PVT, or prothrombotic mutations. Disease progression was independently associated with baseline age (hazard ratio [HR] 1.55; 95% confidence interval [CI]: 1.11-2.17), body mass index (HR 1.40; 95% CI: 1.01-1.95), prothrombin time (HR 0.79; 95% CI: 0.70-0.90), serum albumin (HR 0.97; 95% CI: 0.94-0.99), and esophageal varices (HR 1.70; 95% CI: 1.21-2.38) but not with the prior development of PVT (HR 1.32; 95% CI: 0.68-2.65). In patients with cirrhosis, the development of PVT is associated with the severity of liver disease at baseline, but does not follow a recent progression of liver disease. There is no evidence that the development of PVT is responsible for further progression of liver disease.