A “Biomarker Biopsy” for the Diagnosis of NASH: Promises from CK-18 Fragments

Abstract

To the Editor: We read with great interest the recent article by Younossi and coworkers [1] who developed a diagnostic biomarker set (NASH DiagnosticsTM) for the diagnosis of nonalcoholic steatohepatitis (NASH). The authors investigated a total of 101 patients who had liver biopsies because of suspected NASH. Subsequently, all patients were tested with enzymelinked immunosorbent assay (ELISA)-based assays for a number of potential biomarkers of liver inflammation, including adiponectin, resistin, insulin, glucose, TNF-alpha, IL-6, IL-8, cytokeratin CK-18 (M65 antigen), and caspasecleaved CK-18 (M30 antigen). Interestingly, results showed that levels of M30 antigen predicted histological NASH with 70% sensitivity and 83.7% specificity. Moreover, intact CK18 (M65) yielded 63.6% sensitivity and 89.4% specificity for the diagnosis of NASH. The findings of Younossi et al. [1] confirm and expand the current knowledge on the potential clinical usefulness of different forms of CK18 in patient sera (M30 antigen for apoptosis and M65 antigen for necrosis) in NASH, as previously reported by our [2] and other groups [3]. It is our opinion that two important points in the work by Younossi et al. merit consideration. Firstly, although the authors focused exclusively on the development of a biomarker set, they did not examine whether the diagnostic yield of their panel could be improved by the concomitant use of noninvasive radiological modalities. Although ultrasound, computerized tomography, and magnetic resonance are unable to distinguish NASH or the stage of fibrosis if used alone, it is worth noting that we have previously shown [2] that the combination of M30 antigen measurements and computerized tomography yielded a sensitivity and specificity for the diagnosis of definitive NASH of 86.1% and 83.3%, respectively. These yields were higher than those provided by biomarkers alone [2]. Similarly, the combination of M65 antigen and computerized tomography yielded a sensitivity and specificity for NASH of 83.3% and 83.3%, respectively [2]. Altogether, these data suggest that the evaluation of a biomarker panel in combination with CT could have a greater diagnostic utility for the identification of patients with definitive NASH than the use of biomarkers alone. Another point that should be considered in the study of Younossi and coworkers is that liver biopsy was classified as “no presence of fatty liver disease”, “simple fatty liver”, “steatosis with nonspecific inflammation”, and “NASH”. Notably, patients with steatosis and nonspecific inflammation were excluded from the analysis [1]. The classification system used by Younossi et al. is not in line with that previously proposed by the Nonalcoholic Steatohepatitis Clinical Research Network [4], which distinguishes between definite NASH, borderline NASH, and not NASH (simple fatty liver). Thus, patients with borderline NASH were excluded from the study of Younossi et al. [1], although a considerable number of patients with nonalcoholic fatty liver disease in clinical practice are actually OBES SURG (2008) 18:1507–1508 DOI 10.1007/s11695-008-9639-z