Commentary on “Cytokeratin 18, a Marker of Cell Death, is Increased in Children With Suspected Nonalcoholic Fatty Liver Disease”

Abstract

To the Editor: In a recent study published in the Journal, Vos et al (1) assessed the usefulness of plasma caspase–generated cytokeratin 18 (CK18) fragment (CK18-ASP396) as a novel biochemical marker for nonalcoholic fatty liver disease (NAFLD) in a pediatric cohort. Results showed that CK18-ASP396 levels were significantly higher in children with clinically suspected NAFLD compared with those of obese children without suspected NAFLD and healthy comparison lean children. The authors concluded that their results supported the potential utility of this biochemical test for the diagnosis of NAFLD in a pediatric setting. We believe that this article raises a number of issues. The clinical definition of NAFLD used in the study by Vos et al (1) does not agree with other standard biopsy-proven definitions (2) and actually describes patients with chronic transaminases elevation of unknown origin. Although abdominal ultrasound is widely used for screening asymptomatic children with an incidental elevation of liver enzymes, ultrasound cannot establish reliably the diagnosis of nonalcoholic steatohepatitis or the degree of hepatic fat accumulation (3). The statistical analysis of the data demonstrating an association of CK18-ASP396 with NAFLD is likely to be underpowered and prone to type I error because of the arbitrary nature of clinical classification of patients. We therefore wonder whether the authors' conclusions would be the same if children with clinically suspected NAFLD who did not undergo biopsy (n = 14) would be analyzed separately from those with biopsy-proven nonalcoholic steatohepatitis (“group IIIa” in the study of Vos et al, n = 6). Accordingly, it is possible that elevated levels of CK18-ASP396 in the entire group of 20 patients classified as having NAFLD would be chiefly driven by the values observed in the 6 children with definite nonalcoholic steatohepatitis. Another issue concerns the classification of obese children without transaminases elevation as comparison subjects without suspected NAFLD. In this regard, it should be noted that most pediatric NAFLD has been reported in children with elevated body mass index (4), and the absence of transaminases elevation does not exclude per se the presence of NAFLD (5). For a better understanding of the pathophysiological link between CK18-ASP396 and NAFLD in pediatric patients, we should be diligent in using definitions that are consistent with the guidelines. We believe that the authors' findings are biased by several methodological issues that cast a shadow on the impact of their conclusions. However, we also believe that further exploration of CK18-ASP396 in pediatric NAFLD is of extreme importance to lessen the number of biopsies performed and to provide a reliable noninvasive method to diagnose NAFLD in children.