ORF15-07 - Assessing Different Strategies to Account for Age When using FIB-4 for the Screening of Liver Fibrosis in MASLD

Current guidelines unanimously recommend FIB-4 index as the first step to screen for liver fibrosis in individuals with obesity or type 2 diabetes. However, because age is part of the FIB-4 formula, results of this score can increase with advancing age independently of fibrosis severity. This results in higher false-positive rates and the need for age-specific cutoff recommendations in older adults. Whether alternative methods of handling age can improve diagnostic accuracy, while preserving simplicity, remains unclear. We hypothesized that age-related inflation of FIB-4 contributes to reduced specificity without improving discrimination, and assessed different strategies to reduce age-driven misclassification.
Patients evaluated at our tertiary hospital between 2018-2025 with metabolic dysfunction associated steatotic liver disease (MASLD) and complete FIB-4 and magnetic resonance elastography (MRE) data were included (n=604). Laboratory values obtained near the time of MRE were used to calculate FIB-4. MRE liver stiffness measurement served as the reference standard for clinically significant fibrosis (CSF) defined as ≥3.14 kPa and advanced liver fibrosis (AF) defined as ≥3.53 kPa. Five strategies were evaluated: (1) standard FIB-4 with a single cutoff of 1.3; (2) FIB-4 using a cutoff of 1.3 for individuals < 65 years and a cutoff of 2.0 in those ≥65 years; (3) FIB-3, removing age from the equation with newly derived cutoffs; (4) capped FIB-4, limiting age input to 65 years; and (5) log-transformed FIB-4, using logarithm of age.
A total of 604 individuals were included (age: 54±13 years; 63% female; White 76%, African American 15%, other 9%; BMI 34.7±7.0 kg/m²; 26.3% with diabetes; A1c 6.2±1.3%). Patients aged ≥65 constituted 21% of the total study cohort. The overall prevalence of CSF and AF were 17.4% and 11.5%, respectively. For CSF, AUROC values were similar across strategies, ranging from 0.76 to 0.77 (all p=NS). FIB-3 demonstrated higher sensitivity (74%, CI95 [0.64,0.82]) than standard FIB-4 (62%, CI95 [0.52,0.72]) or FIB-4 with age-specific cut-offs (49%, CI95 [0.39,0.59]), while maintaining a specificity of 70% (vs. 74% with standard FIB-4). For AF, FIB-3 again showed higher sensitivity (79%, CI95 [0.67,0.88]) compared with standard (66%, CI95 [0.53,0.77]) and age-adjusted FIB-4 (51%, CI95 [0.39,0.64]). Capped and log-transformed FIB-4 showed modest improvements in sensitivity but did not outperform FIB-3.

Conclusion:
Age modification strategies for FIB-4 yielded similar discrimination when compared against standard FIB-4 index using MRE as the reference. Removal of age preserves diagnostic performance while improving sensitivity and eliminating the need for age-specific cutoffs. These findings support further evaluation of age-independent fibrosis indices as a simplified and potentially more robust approach for MASLD fibrosis screening.