Usefulness of a novel serum proteome-derived index FI-PRO (fibrosis-protein) in the prediction of fibrosis in chronic hepatitis C

Background Liver biopsy is an imperfect standard for the assessment of chronic hepatitis C liver fibrosis. In this study, the diagnostic role of proteome-derived protein markers and the usefulness of a protein-based index were assessed. Methods Characteristics, clinical biochemistry, and protein markers of patients with chronic hepatitis C from a study (n=62) and validation group (n=73) were statistically assessed according to fibrosis severity. Multivariate models were built using linear discriminant analysis for the prediction of minor fibrosis (F0-F1), moderate fibrosis (F2-F3), and cirrhosis (F4). The best model was validated and diagnostic performance was compared with the aspartate aminotransferase-to-platelet ratio index based on their receiver operator characteristic curves. Results Statistical analysis resulted in significant outcomes for both clinical and protein markers. The best multivariate model was based on four protein markers: alpha-2-macroglobulin (A2M), haptoglobin, hemopexin, and galectin-3-binding protein. A2M and hemopexin were the primary predictors according to this model. A novel index A2M/hemopexin [fibrosis-protein (FI-PRO) index] showed a diagnostic performance rate of 0.80-0.92 for the detection of significant fibrosis (F2-F4) and advanced fibrosis (F3-F4) in the validation group, which was better compared with aspartate aminotransferase-to-platelet ratio index. FI-PRO had an overall positive predictive value of 86% for significant fibrosis and a negative predictive value of at least 90% for advanced fibrosis. Conclusion Proteome-derived protein markers were successfully implemented in clinical diagnosis of hepatitis C fibrosis, which resulted in the FI-PRO index. The efficiency and usability of FI-PRO should be validated in large-scale, prospective studies. Eur J Gastroenterol Hepatol 23: 701-710 (C) 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins.