Reply to: CHAI and MACK‐3 as noninvasive indices for nonalcoholic steatohepatitis

Abstract

We greatly appreciate the letter from Polyzos and Kountouras 1 and their comments in relation to our article. 2 It is interesting to note that Polyzos et al have reported the use of a similar combination of homeostatic model assessment of insulin resistance (HOMA-IR), aspartate aminotransferase (AST) and cytokeratin-18 (CK-18), called CHAI, for the diagnosis of nonalcoholic steatohepatitis (NASH). 3 While it is widely known that HOMA-IR, AST and CK-18 are indicators of insulin resistance, hepatocyte injury and apoptosis, respectively, and reasonable to expect these markers to be elevated in more severe nonalcoholic fatty liver disease (NAFLD), the development of an acceptable algorithm for diagnostic use is not as straightforward. We have evaluated AST and CK-18 as individual tests previously and have found them to be fair and poor, respectively, for the diagnosis of NASH. 4 alanine aminotransferase (ALT) performed poorly too in that study, and we agree with Polyzos and Kountouras that AST is a better marker of more severe NAFLD than ALT. Although Polyzos et al 3 and Boursier et al 5 have used the same markers in their algorithm, several differences are noteworthy, including the different definition of NASH between the two studies, the different diagnostic goals, the different methods in developing the algorithm, and the different composition of histological severity in their study populations. These differences largely underlie the better diagnostic accuracy seen in the algorithm developed by Boursier et al.We agree with Polyzos and Kountouras that further studies in different populations are needed to validate the algorithm. Although we found combination of hoMa, Ast, CK18 (MACK-3) to be good for diagnosing fibrotic NASH, we did not find it to be superior to the NAFLD fibrosis score (NFS) and Fibrosis-4 index (FIB4). Instead, we found MACK-3 to be superior to NFS and FIB4 for the diagnosis of active NASH and similar to NFS and FIB4 for diagnosis of fibrosis stage ≥F2. We believe that the difference could be owing to the different proportion of active NASH patients with ≥F2 fibrosis between the two study populations.This further highlights the need for further studies in different populations. The strength of MACK-3 lies in its ability to rule out fibrotic NASH given its excellent sensitivity and negative predictive value that were consistently shown in both studies. 2,5 With this test, the need for liver biopsy can hopefully be significantly reduced when screening for fibrotic NASH among the very large population of NAFLD patients.