Abstract
BACKGROUND: Noninvasive fibrosis markers are routinely used in patients with liver disease. Magnetic resonance elastography (MRE) is recognized as a highly accurate methodology, but a reliable blood test for fibrosis would be useful. We examined performance characteristics of the Enhanced Liver Fibrosis (ELF) Index compared to MRE in a cohort including those with HCV, HIV, and HCV/HIV. METHODS: Subjects enrolled in the Miami Adult Studies on HIV (MASH) cohort underwent MRE and blood sampling. The ELF Index was scored and receiver-operator curves constructed to determine optimal cutoff levels relative to performance characteristics. Cytokine testing was performed to identify new markers to enhance noninvasive marker development. RESULTS: The ELF Index was determined in 459 subjects; more than half were male, non-white, and HIV-infected. MRE was obtained on a subset of 283 subjects and the group that had both studies served as the basis of the receiver-operator curve analysis. At an ELF Index of > 10.633, the area under the curve for cirrhosis (Metavir F4, MRE > 4.62 kPa) was 0.986 (95% CI 0.994-0.996; p < 0.001) with a specificity of 100%. For advanced fibrosis (Metavir F3/4), an ELF cutoff of 10 was associated with poor sensitivity but high specificity (98.9%, 95% CI 96.7-99.8%) with an AUC of 0.80 (95% CI 0.749-0.845). ELF Index performance characteristics exceeded FIB-4 performance. HCV and age were associated with increased fibrosis (p < 0.05) in a multivariable model. IP-10 was found to be a promising biomarker for improvement in noninvasive prediction algorithms. CONCLUSIONS: The ELF Index was a highly sensitive and specific marker of cirrhosis, even among HIV-infected individuals, when compared with MRE. IP-10 may be a biomarker that can enhance performance characteristics further, but additional validation is required.