Comparative Evaluation of Non-Invasive Analytical Scores (APRI, FIB-4, GPR, AAR) with Actual Fibrosis Stages

View at publisher
Authors: Yasin Sönmez, Ömer Faruk Ertuğrul, Emrah Aydemir
Cited by: 0
References: 17

Abstract

Liver fibrosis constitutes a key pathological process underlying the progression of chronic liver diseases. Accurate staging of fibrosis is essential for prognostic evaluation, treatment planning, and clinical monitoring. Although liver biopsy remains the historical reference standard for fibrosis assessment, its invasive nature, risk of complications, and sampling variability have driven the development of non-invasive diagnostic alternatives. In this study, the diagnostic performance of four analytical indices—AST to Platelet Ratio Index (APRI), Fibrosis-4 (FIB-4), Gammaglutamyl transpeptidase to Platelet Ratio (GPR), and AST to ALT Ratio (AAR)—was evaluated using data from 3,106 patients with histologically confirmed fibrosis stages (F0–F6). Each score was calculated from routine laboratory parameters, and their diagnostic accuracy was assessed through correlation analysis and receiver operating characteristic (ROC) curve evaluation for clinically significant fibrosis thresholds (≥F2, ≥F3, F6). Results indicated that all scores increased with advancing fibrosis stages and exhibited significant positive correlations with histological fibrosis (ρ = 0.315–0.373). Among the indices, the GPR score achieved the highest diagnostic performance for detecting advanced fibrosis (AUC = 0.750), while APRI demonstrated the strongest performance for identifying significant fibrosis (AUC = 0.672). These findings support the clinical utility of simple, low-cost, and non-invasive biochemical scores—particularly GPR and APRI—as valuable tools for screening and monitoring patients at risk of advanced liver fibrosis

Fields of Study & Subjects

Liver Disease Diagnosis and Treatment Liver Disease and Transplantation Liver physiology and pathology

Publication details

Journal / source
International Informatics and Software Engineering Conference (IISEC)
Pages
82–85
Publication date
5 February 2026
Language
EN