COMPARISON OF SEROLOGICAL MODELS OF LIVER FIBROSIS AGAINST TRANSIENT ELASTOGRAPHY BY FIBROSCAN® IN PATIENTS WITH NON-ALCOHOLIC FATTY LIVER DISEASE

Abstract

Introduction and Objectives: Liver fibrosis is the most important prognostic factor in nonalcoholic fatty liver disease (NAFLD). The study's objective is to compare the serological models of liver fibrosis (NAFLD-FS, FIB-4, BARD, APRI and AST/ALT) against transient elastography by FibroScan® in patients with NAFLD. Materials and Methods: Observational, retrolective and cross-sectional study of records of patients diagnosed with liver steatosis by FibroScan® without significant alcohol consumption. A Pearson's correlation and heat maps were used for the correlation between results of FibroScan® and the serological models of liver fibrosis. ROC curves were analyzed to compare the serological models against FibroScan® as the gold standard for clinically significant liver fibrosis. Results: Data from 976 files were collected, with a prevalence of 63% of liver steatosis by FibroScan® (CAP >232 dB/min) and 1.74% of significant liver fibrosis (LSM >7.0 kPa). In patients with NAFLD, a low positive correlation of NAFLD-FS (r=0.291; p<0.001) and BARD (r=0.021; p<0.001) and a very low positive correlation of APRI (r=0.184; p<0.001) with clinically significant liver fibrosis was reported. No correlation was observed with FIB-4 (r=-0.003; p=0.943) or with the AST/ALT ratio (r=-0.039; p=0.336). The NAFLD-FS reported an area under the curve (AUC) of 0.838 (95%CI 0.76-0.91) and the APRI of 0.797 (95%CI 0.68-0.92) compared to FibroScan® for clinically significant liver fibrosis (Figure 1). Discussion: Liver biopsy is an invasive method and the gold standard for evaluating liver fibrosis; however, it is not exempt of complications. Transient elastography by FibroScan® is a non-invasive and validated method but with limited availability and accessibility. Serological models are widely available and can be easily used in daily practice. In a previous study, the NAFLD-FS reported an AUC of 0.72 (95% CI 0.60-0.83) compared against liver biopsy, which is comparable to the AUC reported in this study against FibroScan®. Conclusions: The NAFLD-FS is the serological model for liver fibrosis with the best AUC and correlation with transient elastography in patients with NAFLD and is proposed as an evaluation method in places where FibroScan® or liver biopsy is not available.The authors declare that there is no conflict of interest.