SEARCH

SEARCH BY CITATION

Aliment Pharmacol Ther 2011; 34: 1202–1216

Summary

Background  Performance of non-invasive fibrosis biomarkers may be influenced by aetiology of chronic liver disease (CLD) and the stages of hepatic fibrosis, but large-scale studies are pending.

Aim  To investigate the effect of aetiogy and stages of hepatic fibrosis on the performance of fibrosis biomarkers.

Methods  A total of 2411 patients with compensated CLD (HCV = 75.1%, HBV = 10.5%, NASH = 7.9%, HIV/HCV = 6.5%) were consecutively enrolled in 9 centres. APRI, Forns’index, Lok index, AST-to-ALT ratio, Fib-4, platelets and Fibrotest-Fibrosure were tested against liver biopsy, considered the gold standard. The effect of the stages of hepatic fibrosis to diagnose significant fibrosis and cirrhosis (≥F2 and F4 respectively) was investigated through difference between advanced and non-advanced fibrosis stages (DANA). Performance was expressed as observed area under the ROC curve (ObAUROC) and AUROC adjusted for DANA (AdjAUROC).

Results  Performance of APRI and Fibrotest-Fibrosure was higher than other biomarkers. In all aetiologies, AdjAUROC was higher than ObAUROC. APRI showed its best performance in HCV monoinfected cases, with an AdjAUROC of 0.77 and 0.83 for ≥F2 and F4 respectively. In HBV and non-alcoholic steatohepatitis (NASH) patients, its performance was poor (AdjAUROC <0.70). Performance of Fibrotest-Fibrosure was good in all aetiologies for both ≥F2 and F4 (AdjAUROC >0.73), except for ≥F2 in NASH (AdjAUROC = 0.64). Performance of all biomarkers was reduced in HCV cases with normal ALT.

Conclusions  Aetiology is a major factor influencing the performance of liver fibrosis biomarkers. Even after correction for DANA, APRI and Fibrotest-Fibrosure exhibit the best performance. However, liver biopsy is not replaceable, especially to diagnose ≥F2 and in HCV carriers with normal ALT.