Background: The ELF test (score calculated from markers of fibrosis: hyaluronic acid, procollagen III amino-terminal peptide, and tissue inhibitor of matrix metalloproteinase 1) is a non-invasive blood test used to estimate fibrosis stage in chronic liver disease. Since NAFLD patients with advanced fibrosis are at the highest risk for adverse long-term outcomes, the aim is to assess performance of ELF for identification of advanced fibrosis in patients with NAFLD.

Methods: Patients with biopsy-proven NAFLD were included. ELF scores were calculated using an ADVIA Centaur XP analyzer.

Results: There were 409 patients with NAFLD: 47 ± 12.9 years, 30% male, 70% white, 34% type 2 diabetes (T2DM), 56% hyperlipidemia, 54% hypertension, mean (SD) BMI 41.0 ± 10.0 kg/m2, ALT 48.4 ± 42.0 U/L, mean ELF score was 9.0 ± 1.3. Of the study cohort, 23% patients (N=93) had advanced fibrosis. Patients with advanced fibrosis had significantly higher ELF: 10.2 ± 1.3 vs. 8.6 ± 1.0, p<0.0001. There was a correlation between ELF and liver enzymes: rho=0.23 for ALT, 0.34 for AST (p<0.0001). ELF negatively correlated with BMI: rho = -0.28 (p<0.0001). Across all NAFLD patients, ELF had good performance for identifying patients with advanced fibrosis: area under curve (AUC) = 0.83 (0.79 - 0.87). The performance of ELF was similar in NAFLD patients with [AUC = 0.81 (0.73 - 0.87)] and without [AUC = 0.83 (0.77 - 0.87)] T2DM. With the commonly used cut-off of ELF ≥9.8 for prediction of advanced fibrosis, observed accuracy was as follows: sensitivity 62.4% (51.7% - 72.2%), specificity 89.6% (85.6% - 92.7%) positive predictive value (PPV) 63.7% (55.1% - 71.6%), negative predictive value (NPV) 89.0% (86.1% - 91.3%). Cut-off of 11.3 returned almost perfect specificity 99.0% (97.3% - 99.8%) with a sensitivity of 22.6% (14.6% - 32.4%), PPV 87.5% (68.1% - 95.8%), NPV 81.3% (79.6% - 82.9%).

Conclusions: ELF test performs well in identifying high risk NAFLD patients with or without T2DM.


Z. Younossi: Consultant; Self; AbbVie Inc., Bristol-Myers Squibb, Gilead Sciences, Inc., Intercept Pharmaceuticals, Inc., Novo Nordisk Inc., Terns and Viking. S. Felix: None. T. Jeffers: None. E. Younossi: None. F. Nader: None. H. Pham: None. A. Afendy: None. R. Cable: None. A. Racila: None. Z. Younoszai: None. B.P. Lam: None. P. Golabi: None. L. Henry: None. S. Clement: None. M. Stepanova: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at