Background: Advancing fibrosis is regarded as the most important factor when stratifying patients with chronic hepatitis C for retreatment.
Goals: (1) To compare the performance of 10 biomarkers of fibrosis, including patented tests, among patients with chronic hepatitis C and treatment failure; and (2) to assess the impact on biomarker performance of using 2 different assays of hyaluronic acid (HA).
Study: For 80 patients, liver histology (Metavir) was compared with biomarker scores using sera obtained within 6 months of liver biopsy (indirect biomarkers: AST:ALT ratio, APRI, Forns index, FIB-4, Fibrometer V3G; direct biomarkers: ELF, Fibrospect II, Hyaluronic acid-HA, Fibrometer V2G, Hepascore). Direct biomarker scores were calculated using 2 validated assays for HA (ELISA and radiometric).
Results: Using the ELISA assay for HA to calculate the direct panels, all 10 of the biomarkers exhibited comparable overall discriminatory performance (unweighted Obuchowski measure, ordROC 0.92-0.94, P-value>0.05) except AST:ALT ratio and APRI (ordROC 0.86-0.88, P-value<0.05). For the detection of moderate (F2-4) and advanced (F3-4) fibrosis, the AUROC of Fibrometer 2G were significantly higher than AST:ALT ratio and APRI but none of the other biomarkers. Good correlation was observed between the 2 HA assays (intraclass correlation coefficient=0.873) with the ELISA assay exhibiting superior diagnostic performance (ordROC 0.92 vs. 0.88, P-value=0.003). Importantly, the performance of many of the direct biomarkers at their diagnostic thresholds was heavily influenced by the choice of HA assay.
Conclusions: Although many biomarkers exhibited good diagnostic performance for the detection of advancing fibrosis, our results indicate that diagnostic performance may be significantly affected by the selection of individual component assays.