SAT-674 Predicting Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) in Patient Without Diabetes, Is FIB4 Still the Best?

Abstract Disclosure: N.N. Nguyen-Jones: None. G. Arndt: None. M. Lindsay: None. S. Swartwood: None. S. Arrigain: None. R. Lopez Moscoso: None. A. Wieland: None. Introduction: With the rising prevalence of MASLD worldwide as well as emerging therapies to manage more progressive disease, it has become critical to identify patients with significant fibrosis (Stage 2 or greater) through use of noninvasive testing (NIT). The current guideline from AASLD recommends primary risk assessment using FIB-4 ≥ 1.3 as a cutoff due to its excellent negative predictive value. Our study aims to compare the ability of different NIT methods including FIB-4, AST to platelet ratio index (APRI), AST/ALT ratio, and NAFLD fibrosis score (NFS) to predict significant fibrosis on FibroscanTM (liver stiffness ≥8 kPa) due to MASLD. The study also aims to assess FIB-4 cutoff values used to rule-in and rule-out significant fibrosis in this population. Methods: A retrospective cohort study is performed on 127 non-diabetic patients with a diagnosis of MASLD at the endocrinology or hepatology clinic at the University of Colorado Medicine, who had a Fibroscan within 6 months of the calculated NITs including FIB-4, APRI, AST/ALT ratio, and NFS. Significant fibrosis on FibroscanTM is defined as a liver stiffness measurement (LSM) of greater than or equal to 8 kPa. The study uses ROC curves and estimated AUC to assess predictability of each score. Optimal cutoffs of each score are calculated using Youden index. A p-value <0.05 is used to determine significance. Results: The mean age of the subjects is 51 years, and 53% were female. Of the 127 patients, 26 (20.3%) subjects had significant fibrosis. Patients with significant fibrosis had a higher BMI (median of 36 kg/m2 vs 32 kg/m2), CAP score, and FAST score. FIB-4 (AUC; 95% CI: 0.67; 0.55-0.79), APRI, (AUC; 95% CI: 0. 65; 0.52-0.78) and NFS (AUC; 95% CI: 0.68; 0.57-0.79) are found to perform similarly well while AST/ALT ratio did not (AUC; 95% CI: 0.50; 0.37-0.63). Optimal cutoffs predicting LSM >8 kPa with FIB-4, APRI, and NFS were 1.21, 0.39, and 1.64 respectively. In this population, a FIB-4 cut point of 1.34 has a sensitivity of 42% and specificity of 78%. Of the 15 individuals with false negative using FIB-4 score cutoff of 1.34, 60% would have been correctly identified by NFS, 33% with FIB-4 cutoff of ≥ 1.21, and 33% with APRI cutoff ≥ 0.39. Discussion: FIB-4, APRI, and NFS were all found to be better than chance when assessing for significant fibrosis and perform similarly when looking at the AUC curves. Classic FIB-4 cutoffs >1.34 underperformed when compared to a FIB-4 cutoff of 1.21. This study showed NFS score cutoffs may have fewer false negative and could be better for ruling-out disease. This is different than our prior studies in diabetic populations that showed APRI outperformed FIB4 and NFS for ruling out significant disease. Future steps will look to see overall performance of NITs in patient population regardless of diabetes status to ease use of NITs when deciding who to screen for significant fibrosis with FibroscanTM. Presentation: Saturday, July 12, 2025