Noninvasive fibrosis tools in NAFLD: validation of APRI, BARD, FIB-4, NAFLD fibrosis score, and Hepamet fibrosis score in a Portuguese population

Objectives: The burden of nonalcoholic fatty liver disease (NAFLD) is increasing, with an estimated prevalence in Europe of 20-30%. Although most patients present with simple steatosis, some progress to advanced fibrosis, cirrhosis, and hepatocellular carcinoma. Definite diagnosis and staging require liver biopsy, which is not feasible given the high prevalence of NAFLD. As such, several noninvasive tools have been formulated. However, to date, none have been validated in the Portuguese population. The aim of this study was to determine the diagnostic accuracy of the aspartate aminotransferase to platelet ratio (APRI), the BMI, AST/ALT ratio and Diabetes (BARD), the FIB-4 Index (FIB-4), the Hepamet fibrosis score (HFS), and the NAFLD fibrosis score (NFS) in a Portuguese population. Methods: A retrospective review of liver biopsies from two hospital centers was performed. Patients with NAFLD and no decompensated cirrhosis, liver cancer, or terminal illness were included. APRI, BARD, FIB-4, HFS, and NFS were calculated for each patient. Results A total of 121 individuals were included, of which 21.5% had advanced fibrosis (F >= 3). There was a moderate or high correlation between most tools. The negative predictive factor (NPV) and area under receiver operating curve (AUROC) were 89.9% and 0.80 for APRI, 91.8% and 0.84 for BARD, 95.7% and 0.88 for FIB-4, 96.4% and 0.88 for HFS, and 93.0% and 0.86 for NFS, respectively. Conclusion The tools analyzed had excellent performance (AUROC >= 0.80) and were adequate for ruling out advanced fibrosis (NPV >= 89.9%) in a Portuguese population. As such, they are adequate for use in clinical practice or as a part of referral and follow-up programs wherever this population is treated.