Introduction The enhanced liver fibrosis (ELF) test comprises a panel of biomarkers of liver fibrosis shown to accurately assess fibrosis and predict clinical outcomes in a range of chronic liver diseases. This is the first study of performance of ELF in patients with chronic hepatitis B (CHB) using a single aetiology cohort, and we compare its performance with an alternative method of assessing fibrosis, transient elastography (TE).
Aim We aimed to compare the performance of ELF with transient elastography (TE) in a cohort of patients with CHB, in the identification of liver fibrosis using liver biopsy as the reference standard.
Method 188 patients with CHB were recruited consecutively at a single Italian centre. TE, serum sampling and liver biopsy were performed on the same day. ELF tests were performed in one batch at a central laboratory using thawed samples previously stored at −20°C. Biopsies were assessed by one pathologist using the Ishak staging system. Diagnostic performance of ELF and TE for detection of histological stages of liver fibrosis was assessed and area under receiver operator characteristic curves (AUROC) calculated. Only those in whom TE was successfully performed were included. Different fibrosis levels were assessed, from any fibrosis (0 vs 1–6) to cirrhosis (0–4 vs 5, 6 and 0–5 vs 6).
Results Patients were treatment-naive, median age was 47 years. 78% were e-antigen negative. Biopsies reported mild/moderate fibrosis (Ishak 0–1) in 21%, moderate fibrosis (2–3) 41% and severe fibrosis/cirrhosis (4–6) 38%. ELF and TE demonstrated good performance in identifying fibrosis (Abstract P23 table 1—representative fibrosis stages are shown, further data available). Both modalities showed similar performance in identifying any fibrosis and minimal fibrosis, with TE performing better in identifying moderate and severe fibrosis, and cirrhosis.
Conclusion In untreated patients with CHB with moderate and severe fibrosis or cirrhosis, TE correlated more closely with histological staging than ELF. The relatively modest performance of ELF in detecting severe fibrosis compared to previous studies may be attributable to disease aetiology or prolonged sample storage at −20°C. Further validation of TE and ELF in CHB should include analysis at the time of sampling and evaluation of the prognostic performance for clinical outcomes as well as histology.