Performance of fibrosis-4 and NAFLD fibrosis score for screening liver fibrosis in Indian individuals with type 2 diabetes: a prospective cross-sectional study.

BACKGROUND Blood-based non-invasive tests (NITs) in patients with T2DM/MASLD help in assessing the stage of hepatic fibrosis, but there is sparse data in the Indian population.

METHODS Fibrosis-4 (FIB-4) and NAFLD fibrosis score (NFS) were calculated for all individuals with T2DM with MASLD and compared to liver stiffness measurement (LSM) values obtained by vibration-controlled transient elastography (VCTE) with cut-offs of ≥9.7 kPa (rule-in cut-off for advanced fibrosis) and <6 kPa (as a rule-out cut-off for significant fibrosis).

RESULTS The prospective study included 789 patients (238 females) with T2DM and MASLD, aged 53.3 ± 10.3 years. The LSM values were ≥9.7 in 102 (12.9%). Results for the whole group were as follows: The optimal rule-in cut-off for FIB-4 was 1.68 (AUROC 0.69, 95% CI 0.637-0.746, sensitivity 64.7%, specificity 65.2%). The optimal rule-in cut-off of NFS for advanced fibrosis was -0.389 (AUROC 0.619, 95% CI 0.560 to 0.678, sensitivity 61.8%, specificity 57.4%). Rule-out cut-offs also had poor AUROCs (0.586 for FIB-4 and 0.570 for NFS). The diagnostic accuracy of the earlier suggested rule-in (for advanced fibrosis) cut-offs of FIB-4 (>2.67) was 81.3% (sensitivity 22.1%, specificity 92.2%). The diagnostic accuracy of NFS (>0.675) for advanced fibrosis was 79.4% (sensitivity 21.3% and specificity 90.1%). The diagnostic accuracy of the rule-out cut-offs was 52.7% for FIB-4 (at value < 1.3) and 53.8% for NFS (at value < -1.455).

CONCLUSION The performance of FIB-4 and NFS to predict or rule out advanced fibrosis was poor in Indian patients with type 2 diabetes mellitus and MASLD.