Noninvasive diagnosis of fibrosis in non-alcoholic fatty liver disease: diagnostic accuracy of different scores

Abstract

Non-alcoholic fatty liver disease (NAFLD) is a spectrum of pathologies characterized by liver damage without history of excessive alcohol intake. Advanced fibrosis, generally detected by transient elastography (TE), is the most significant predictor of poor prognosis and mortality among these patients. This study aimed at assessing the accuracy of five noninvasive methods, compared to TE, for the evaluation of severity of liver fibrosis in patients with NAFLD. The cohort included 41 patients, in whom the result of TE was compared to AST/ALT ratio, BARD Score (Body Mass Index, AST/ALT ratio, diabetes), AST To Platelet Ratio Index (APRI), Fibrosis-4 Index (FIB-4 Index) and NAFLD Fibrosis Score (NFS). The severity of fibrosis, assessed by TE, was the following: F0 (absence of fibrosis): 17%, F1 (mild): 39%, F2 (moderate): 17%, F3 (advanced): 10%, F4 (cirrhosis): 17%. Performances of the diagnostic scores were: 49% for AST/ALT ratio, 68% for BARD Score, 73% for APRI, 59% and 71% for the lower and upper cut-off of FIB-4 Index, 61% and 76% for the lower and upper cut-off of NFS. Considering the scores compared to TE, AST/ALT ratio was not enough sensitive, while BARD Score had better diagnostic performance and APRI had a superior accuracy than the formers. However, FIB-4 and NFS were the most useful tests and their performance could be improved through the use of a single cut-off. These findings demonstrated that the most accurate scores, compared to TE, were NFS and FIB-4.