Relationship between pre-TIPS liver perfusion by the portal vein and the incidence of post-TIPS chronic hepatic encephalopathy.

Abstract

In the present study we evaluated the predictive value of pretransjugular intrahepatic portosystemic shunt (TIPS) portal perfusion as assessed by Doppler ultrasonography for the onset of chronic encephalopathy after TIPS. A total of 231 cirrhotic patients were followed-up prospectively after TIPS placement. The pattern of intrahepatic portal flow was assessed before TIPS. Patients were divided into two groups according to Doppler findings. Group 1 comprised patients with prograde portal flow (n = 200), whereas group 2 comprised those with loss of portal perfusion (hepatofugal or back-and-forth flow or portal vein thrombosis; n = 31). The presence of chronic encephalopathy during a median follow-up of 32 months was prospectively recorded. The prognostic value of the following parameters for the onset of chronic recurrent encephalopathy after TIPS was evaluated: age, presence of encephalopathy before TIPS, alcoholism, Pugh score, and loss of portal perfusion before TIPS. The independent prognostic value of each variable was tested with a multiple logistic regression analysis. The two groups were comparable in terms of age, incidence of prior episodes of hepatic encephalopathy, and portacaval gradient before and after the procedure; however, liver failure was more severe in patients in group 2 (Pugh score: 9.2 +/- 1.9 vs 10.3 +/- 1.7). The 3-yr survival was identical for both groups; 25% of the 200 patients in group 1 developed chronic encephalopathy as compared to 6% of the 31 patients in group 2 (p = 0.03). Multiple logistic regression analysis demonstrated that loss of portal perfusion and age >65 yr were the only independent predictors of the onset of post-TIPS chronic encephalopathy (odds ratios 0.24 and 1.98, respectively). Cirrhotic patients with loss of portal perfusion before TIPS were protected against post-TIPS chronic hepatic encephalopathy despite a more severe liver dysfunction at baseline. The only other independent predictive factor for the onset of this complication was age.