The role of simple and specialized non-invasive tools in predicting of metabolic dysfunction-associated fatty liver disease severity and prognosis

Abstract

Objectives

Metabolic-associated fatty liver disease (MAFLD) is a universal health concern. Detecting advanced fibrosis significantly impacts prognosis. This study designated to assess the accuracy of FIB-4, FIB-6, Agile3+, Agile4, and NAFLD fibrosis score (NFS) in predicting disease severity.

Design and Methods

Clinical, laboratory, and FibroScan findings of adult MAFLD patients were recorded. A fibrosis (F) score over 10 kPa indicates advanced fibrosis. We assessed the area under the receiver operating characteristic curve (AUROCC), along with the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy with various cutoff values. Reliability was analyzed with the intra-class correlation coefficient (ICC).

Results

Advanced fibrosis was found in 5 of 103 patients (4.85%). AUROCC values were as follows: 0.967 for Agile3+, 0.951 for FIB-6, 0.932 for NFS, 0.909 for Agile4, and 0.869 for FIB-4. The PPVs ranged from 18.18% (FIB-4) to 57.09% (NFS), followed by 75.02% (FIB-6), and 18.18% (Agile4) to 100% (Agile3+). All tools achieved acceptable NPVs above 96%. The ICC between the fibrosis score and other tools was 0.772 (95% CI: 0.696-0.834). Significant differences were noted in gamma-glutamyl transferase (p=0.039), diabetes mellitus (P=0.011), platelet count, hemoglobin A1C, alkaline phosphatase, triglycerides, fasting blood glucose, and Vitamin D levels