Validation of the accuracy of the FAST(™) score for detecting patients with at-risk nonalcoholic steatohepatitis (NASH) in a North American cohort and comparison to other non-invasive algorithms

BACKGROUND AND AIMS: Management of patients with NASH who are at elevated risk of progressing to complications of cirrhosis (at-risk NASH) would be enhanced by an accurate, noninvasive diagnostic test. The new FAST(™) score, a combination of FibroScan(®) parameters liver stiffness measurement (LSM) and controlled attenuation parameter (CAP) and aspartate aminotransferase (AST), has shown good diagnostic accuracy for at-risk NASH (area-under-the-Receiver-Operating-Characteristic [AUROC] = 0.80) in European cohorts. We aimed to validate the FAST(™) score in a North American cohort and show how its diagnostic accuracy might vary by patient mix. We also compared the diagnostic performance of FAST(™) to other non-invasive algorithms for the diagnosis of at-risk NASH. METHODS: We studied adults with biopsy-proven non-alcoholic fatty liver disease (NAFLD) from the multicenter NASH Clinical Research Network (CRN) Adult Database 2 (DB2) cohort study. At-risk-NASH was histologically defined as definite NASH with a NAFLD Activity Score (NAS) ≥ 4 with at least 1 point in each category and a fibrosis stage ≥ 2. We used the Echosens(®) formula for FAST(™) from LSM (kPa), CAP (dB/m), and AST (U/L), and the FAST(™)-based Rule-Out (FAST(™) ≤ 0.35, sensitivity = 90%) and Rule-In (FAST(™) ≥ 0.67, specificity = 90%) zones. We determined the following diagnostic performance measures: AUROC, sensitivity (Se), specificity (Sp), positive predictive value (PPV), and negative predictive value (NPV); these were calculated for the total sample and by subgroups of patients and by FibroScan(®) exam features. We also compared the at-risk NASH diagnostic performance of FAST(™) to other non-invasive algorithms: NAFLD fibrosis score (NFS), Fibrosis-4 (FIB-4) index, and AST to platelet ratio index (APRI). RESULTS: The NASH CRN population of 585 patients was 62% female, 79% white, 14% Hispanic, and 73% obese; the mean age was 51 years. The mean (SD) AST and ALT were 50 (37) U/L and 66 (45) U/L, respectively. 214 (37%) had at-risk NASH. The AUROC of FAST(™) for at-risk NASH in the NASH CRN study population was 0.81 (95% CI: 0.77, 0.84. Using FAST(™)-based cut-offs, 35% of patients were ruled-out with corresponding NPV = 0.90 and 27% of patients were ruled-in with corresponding PPV = 0.69. The diagnostic accuracy of FAST(™) was higher in non-whites vs. whites (AUROC: 0.91 vs 0.78; p = 0.001), and in patients with a normal BMI vs. BMI > 35 kg/m(2) (AUROC: 0.94 vs 0.78, p = 0.008). No differences were observed by other patient characteristics or FibroScan(®) exam features. The FAST(™) score had higher diagnostic accuracy than other non-invasive algorithms for the diagnosis of at-risk NASH (AUROC for NFS, FIB-4, and APRI 0.67, 0.73, 0.74, respectively). CONCLUSION: We validated the FAST(™) score for the diagnosis of at-risk NASH in a large, multi-racial population in North America, with a prevalence of at-risk NASH of 37%. Diagnostic performance varies by subgroups of NASH patients defined by race and obesity. FAST(™) performed better than other non-invasive algorithms for the diagnosis of at-risk NASH.