Predictive role of serum procollagen III peptide and Knodell's index in survival prognosis of patients with hepatitis B virus liver cirrhosis.

Abstract

The objective of the study was to improve the accuracy of survival prognosis in patients with liver cirrhosis using procollagen III peptide (PIIIP), as a marker of inflammation and fibrogenesis, and Knodell's histologic activity index (KI) in addition to previously used prognostic factors. Five-year survival was followed in a group of 75 patients with hepatitis B virus (HBV) liver cirrhosis (patients testing anti-HBe positive and HBV-DNA negative). There were 31 patients with compensated cirrhosis and 44 with decompensated cirrhosis. The diagnostic procedure included clinical, laboratory, ultrasound and pathohistologic examination. We combined PIIIP and KI with other significant variables to achieve the highest possible sensitivity, specificity and accuracy for survival prognosis in HBV liver cirrhosis. The models were compared using ROC analysis. At the end of the five-year period of survival follow-up, there were 39 survivors and 36 patients had died (only three died from an extrahepatic cause). In the quantitative model, the discriminant canonical function (DCF) identified PIIIP, bilirubin, prothrombin time, ascites and KI as statistically significant parameters in the prognosis of five-year survival. Calculation of the score based on DCF yielded an accuracy of 89.3%. In the semiquantitative model, the analysis of variance identified PIIIP, bilirubin, albumin, pro-thrombin time, alkaline phosphatase, ascites and KI as significant variables. When PIIIP was added to the clinicohistologic diagnosis, Child-Pugh score and KI, the level of accuracy improved by 12% (from 78% to 90%), 11% (from 79% to 90%) and 10.6% (from 80% to 90.6%), respectively. When calculated with the three biochemical parameters (alkaline phosphatase, PIIIP and bilirubin) and KI identified by DCF, the accuracy was 90.6%. Combining PIIIP and KI with other prognostic parameters is useful in achieving a better precision of survival prognosis in patients with HBV liver cirrhosis.