Comparison of FibroScan-Aspartate Aminotransferase (FAST) Score and Other Non-invasive Surrogates in Predicting High-Risk Non-alcoholic Steatohepatitis Criteria

Abstract

Identification of non-alcoholic steatohepatitis (NASH) with high activity and fibrosis is a major priority in patients with non-alcoholic fatty liver disease. We validated the predictive value of the FibroScan-aspartate aminotransferase (FAST) score and other non-invasive fibrosis surrogates in predicting high-risk NASH criteria. This multicenter retrospective study recruited 251 biopsy-proven non-alcoholic fatty liver disease (NAFLD) patients (132 [52.6%] men) between 2011 and 2014. The FAST score was calculated using transient elastography data and aspartate aminotransferase (AST) levels. The NAFLD fibrosis score (NFS), fibrosis-4 index (FIB-4), and AST to platelet ratio index (APRI) were calculated using biochemical data. The area under the receiver operating characteristic curves (AUCs) of the FAST score, liver stiffness, NFS, FIB-4, and APRI were 0.752, 0.718, 0.609, 0.650, and 0.722 for NAFLD activity score (NAS) >= 5 (n = 117, 46.6%); 0.788, 0.754, 0.649, 0.701, and 0.747 for fatty liver inhibition of progression-NASH with histologic activity >= 3 (n = 202, 80.5%); 0.807, 0.806, 0.691, 0.732, and 0.760 for severe disease with activity >= 3 and/or fibrosis >= 3 (n = 132, 52.6%); and 0.714, 0.812, 0.748, 0.738, and 0.669 for NASH with NAS >= 4 and fibrosis >= 2 (n = 70, 27.9%), respectively. The FAST score had the highest AUC for the most high-risk NASH criteria, except for in predicting NAS >= 4 and fibrosis >= 2. The liver stiffness value showed consistently acceptable performance in predicting all high-risk NASH criteria. The FAST score has acceptable performance in identifying high-risk NASH. However, liver stiffness alone was not inferior to the FAST score.