RESUMO
INTRODUCTION AND OBJECTIVES: Liver fibrosis is an important prognosis marker in non-alcoholic fatty liver disease (NAFLD). Biopsy has been considered the gold-standard method for measuring liver fibrosis; however, it is an invasive procedure. Non-invasive diagnostic tools have been developed, such as clinical scores and magnetic resonance elastography (MRE), which is the most accurate non-invasive method to determine liver fibrosis. Thus, the aim was to determine the NAFLD Fibrosis Score (NFS) and the Fibrosis-4 Score (FIB-4) cut-off points that best identify NAFLD patients at risk for developing liver fibrosis. PATIENTS AND METHODS: Single-center cross-sectional study with prospective recruitment of NAFLD (training-cohort) and MAFLD (validation-cohort) patients undergoing MRE. The NFS and the FIB-4 cut-off points that best-differentiated patients with fibrosis, using the MRE as the standard method, were determined. RESULTS: Two cohorts were analyzed, a training cohort that included the initial 183 patients with NAFLD and a validation cohort that included 289 patients. In the training cohort, 60.1% had mild steatosis and 11.5% had liver fibrosis ≥ F1 by MRE. ROC curves were developed for FIB-4 and NFS, and the cut-off points chosen were 1.505 (sensitivity=85% and specificity=86%) for FIB-4 and -0.835 (sensitivity=100% and specificity=70%) for NFS, showing greater specificity than the cut-off points currently used (51% and 76%, respectively). The two cohorts exhibited similar characteristics and similar sensitivity and specificity results for the chosen cut-off points. CONCLUSIONS: This study has shown cut-off points with greater specificity and excellent sensitivity to guide the indication for further liver evaluation by MRE in NAFLD patients.