FIB-4 fails to identify significant liver fibrosis in people with HIV: A large multinational screening study

Abstract

BACKGROUND AND AIMS: Steatotic liver disease (SLD) and liver fibrosis are major comorbidities in people with HIV (PWH). Guidelines recommend stepwise screening using the Fibrosis-4 (FIB-4) index followed by transient elastography (TE), yet its accuracy and the extent of FIB-4 misclassification in PWH remain uncertain. We evaluated the diagnostic performance of FIB-4 against TE, quantified missed fibrosis, and assessed whether metabolic and HIV-specific factors improve risk prediction. APPROACH AND RESULTS: We conducted a multinational study of 4917 PWH without viral hepatitis coinfection or hazardous alcohol intake undergoing TE screening across 7 centers. SLD was defined by a controlled attenuation parameter ≥275 dB/m and classified as metabolic dysfunction-associated SLD (MASLD) or metabolic dysfunction-associated alcohol-related liver disease (MetALD). Significant fibrosis [liver stiffness measurement (LSM) ≥8 kPa] was present in 12.6% of participants, advanced fibrosis (LSM ≥11 kPa) in 6.1%, and SLD in 21.7% (20.6% MASLD, 1.1% MetALD). FIB-4 showed modest accuracy for significant fibrosis (AUROC 0.69, 95% CI 0.67-0.72) and misclassified 36% of fibrosis cases as low risk (FIB-4 <1.3). Performance was poorer in MASLD than in non-MASLD (AUROC 0.60 vs. 0.76; p <0.001). Participants with false-negative FIB-4 exhibited a more metabolic phenotype, including higher BMI and steatosis. Incorporating metabolic and HIV-specific factors improved discrimination and reclassification and enabled the development of the FIB-HIV score, which outperformed FIB-4 (AUROC 0.78 vs. 0.69; p <0.001). CONCLUSIONS: In PWH, liver fibrosis is common and frequently missed by FIB-4, particularly in MASLD. TE-centered screening strategies augmented by metabolic and HIV-specific indicators may improve early fibrosis detection and risk stratification.