Noninvasive markers of liver fibrosis: on-treatment changes of serum markers predict the outcome of antifibrotic therapy.

Aim: The utility of noninvasive serum markers to longitudinally monitor liver fibrosis is not established.

Methods: A total of 70 patients with chronic hepatitis C who had previously failed antiviral therapy were randomized to receive pegylated interferon with or without silymarin for 24 months. Enhanced Liver Fibrosis (ELF) tests (hyularonic acid, terminal peptide of procollagen III, tissue inhibitor of matrix metaloproteinase-1) were performed on patient sera obtained before, during and at the end of the study (0, 12, 24 months) and liver histology obtained before and at the end of the study.

Results: At 24 months, absolute changes in Ishak fibrosis stage and ELF ranged from -4 to +4 and from -2.41 to +2.68, respectively. Absolute changes in ELF at 12 months were significantly associated with changes in both ELF and histology at 24 months. A model combining both baseline ELF and change of ELF at 12 months could predict the 24-month ELF (R=0.609, P<1×10), a decrease in ELF at 24 months [area under the curve (AUC): 0.80-0.85] and an increase in ELF at 24 months (AUC: 0.81-0.85). Furthermore, a model combining both baseline histologic stage and ELF together with the change of ELF at 12 months could predict 24-month histology (R=0.601, P<1×10, AUC: 0.88-0.92), histologic fibrosis regression (AUC: 0.81-0.84) and progression (AUC: 0.86-0.91).

Conclusion: Our observations suggest that a change in the serum marker ELF predicts changes in liver fibrosis over a longer period. These data support the use of ELF as a surrogate marker of liver fibrosis evolution in monitoring antifibrotic treatments, thus permitting 'response-guided' therapy by the early identification of patients who will benefit from prolonged treatment.