Objective: To evaluate hyaluronic acid (HA) and Hepascore as diagnostic replacements for liver biopsy in a population with mixed liver disease.
Materials and methods: The utility of HA concentration and Hepascore for staging fibrosis, detecting any fibrosis and detecting advanced fibrosis, was assessed in 73 consecutive patients, with varied liver pathologies requiring biopsy. Subgroup analyses compared utility of disease-specific and universal cut-offs for HA and Hepascore.
Results: Forty-one patients (56.2%) had liver fibrosis on biopsy. HA and Hepascore varied significantly with METAVIR stage, although ranges overlapped, precluding their use in staging fibrosis. When detecting any fibrosis (METAVIR F1-F4), HA and Hepascore had areas under the receiver operator characteristic curve of 0.63 and 0.66, respectively, and approximately two-thirds of patients were correctly categorized using optimal cut-offs. For detection of advanced fibrosis (METAVIR F3/4), HA and Hepascore had areas under the receiver operator characteristic curve of 0.81 and 0.80, respectively, and three-quarters of patients were correctly categorized using optimal cut-offs. In subgroup analysis, locally derived, disease-specific cut-offs in hepatitis C virus patients yielded greatest diagnostic efficiency, whereas the tests performed worst in cryptogenic aetiologies.
Conclusion: HA and Hepascore cannot accurately stage hepatic fibrosis in this population. Locally derived, disease-specific cut-offs for HA gave the higher diagnostic efficiency observed. Although HA and Hepascore may be useful where the disease aetiology is known, particularly in established hepatitis C virus, the high cost of false positives and false negatives are such that neither a reliable enough to replace biopsy without substantial further characterization.