Introduction The enhanced liver fibrosis (ELF) test, a panel of liver fibrosis biomarkers, accurately assesses fibrosis in a range of chronic liver disease (CLD) aetiologies. We evaluated concordance of ELF and transient elastogrpahy (TE) in assessing fibrosis in a cohort of mixed aetiology CLD in our clinic.
Methods Patients who had undergone ELF testing and TE within 1-year for investigation of CLD were identified. Data derived thresholds for ELF were used to identify moderate fibrosis and cirrhosis with 90% sensitivity and specificity respectively. TE thresholds were based on a study of patients with mixed viral aetiology CLD with the same sensitivities/specificities.1 Valid TE criteria: success rate ≥60%, median stiffness IQR ≤30%. ELF test was included in routine blood testing. TE was performed in 8 min on average by an experienced nurse. Concordance with ELF and TE classification was calculated. In addition patients who had undergone ELF testing or TE with liver biopsy within 2 years were identified. Histological fibrosis severity was evaluated and agreement with ELF/TE calculated.
Results Of 110 consecutive patients, 99 had ELF/TE within 1-year. Median age 50 (22–80). Aetiology: HCV 46%, HBV 25%, unknown 17%, fat 7%, PBC 1%, HBV/HCV 1%, α-1 antitrypsin deficiency 1%, normal 1%. No ELF tests failed. TE failed in 11%, and valid results obtained in 66% and ELF/TE concordance analysis based on these. Correlation between ELF and TE was 0.6. The Abstract PWE-272 table 1 shows distribution of patients for mild, moderate-severe fibrosis and cirrhosis. Agreement between ELF/TE; κ=0.18. Prediction of moderate fibrosis, κ=0.14, with 34% discrepancy. Prediction of cirrhosis, κ=0.49, with 17% discrepancy. Analysis of the notable mild TE/moderate ELF group shows a negative skew of TE (mean 4.13, median 4.30) and a positive skew of ELF (mean 8.44, median 8.37). Clinical correlation of discrepant cases for cirrhosis indicates consistency with ELF in 13% (mod TE/cirrhosis ELF), and 67% (mod ELF/cirrhosis TE). Kappas for agreement with histology for moderate fibrosis and cirrhosis were, for ELF: 0.19, 0.71 (n=16); for valid TE: 0.13, 0.79 (n=11).
Conclusion In a cohort of mixed aetiology CLD, ELF was more reliable than TE in generating a result with a failure rate of 34% for TE. Agreement between ELF and TE increased with fibrosis severity. The mild/moderate discordance suggests a need to review thresholds. When selecting non-invasive tests for use in busy clinics failure rate of the test and time taken to obtain a result should be considered.
Competing interests P Trembling: None declared, M Cheung: None declared, S Tanwar: None declared, W Rosenberg Grant/Research Support from: Siemens Healthcare Diagnostics.
Reference 1. Degos F, Perez P, Roche B, et al. Diagnostic accuracy of FibroScan and comparison to liver fibrosis biomarkers in chronic viral hepatitis: a multicenter prospective study (the FIBROSTIC study). J Hepatol 2010;53:1013–21.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.