Cytokeratin-18 fragments and biomarkers of the metabolic syndrome in nonalcoholic steatohepatitis

Abstract

Nonalcoholic fatty liver disease (NAFLD) remains a leading cause of chronic liver disease. In the context of NAFLD, the presence of nonalcoholic steatohepatitis (NASH) portends an adverse prognosis with greater risk of liver fibrosis and cirrhosis. Although liver biopsy is the keystone of patient management in NAFLD, it is also increasingly clear that such evaluation has its limitations. The availability of biochemical markers of NAFLD and NASH has tremendous potential to radically alter management strategies for these conditions, as well as to monitor disease activity. Our article provides an overview of biomarker discovery and selection in the setting of NAFLD and highlights future directions in the field.

INTRODUCTION

Nonalcoholic fatty liver disease (NAFLD) is a leading cause of chronic liver disease and its incidence is rising worldwide[1-3]. The term NAFLD is used to describe a wide spectrum of fatty liver changes ranging from simple steatosis to nonalcoholic steatohepatitis (NASH)[4]. Although simple steatosis usually follows a benign course, steatohepatitis is prone to progress to hepatic fibrosis and cirrhosis leading to excess morbidity and mortality[5,6]. In this context, early identification of patients with NASH prior to the onset of advanced fibrosis would be helpful in guiding aggressive intervention. Liver biopsy remains the gold standard for obtaining an accurate diagnosis of NASH, as well as for differentiating this condition from simple steatosis. Unfortunately, biopsy is a costly and invasive diagnostic procedure, possibly subject to inter- and intra-observer variability[7,8]. Thus it is becoming increasingly clear that although liver biopsy is the keystone of patient diagnosis and management in the setting of NAFLD, such evaluation has its limitations. Clinicians should therefore use additional tools to aid clinical assessment and to enhance their ability to identify the patient at risk for NASH and liver fibrosis. Biomarker use is one such tool to better identify subjects with NASH in the context of NAFLD, and hopefully to effectively prognosticate patients with this condition.

The development of NASH biomarkers can be theoretically achieved via two different strategies: the first strategy can be defined as “knowledge-based” (deductive method based on the current knowledge of NASH pathophysiology), while the second one is more “unbiased” (inductive strategy). The “knowledge-based” approach relies on a direct understanding of the pathophysiological processes that underlie the development of NASH as well as the evolution of its sequelae. It may consist of biochemical assays aiming to assess attractive novel candidate markers informed by the biology of the disease process. For instance, the understanding of the role played by hepatocyte apoptosis[9] and insulin resistance[10,11] in the pathobiology of liver injury has enabled the development of promising biomarkers of NASH, such as caspase-cleaved cytokeratin 18 fragments or numerous different adipokines. On the other hand, the “unbiased” approach involves the use of modern techniques including proteomics, metabolomics, and bioinformatics that have allowed unbiased investigations of numerous putative markers that may be informative with regard to the various stages of NAFLD, including overt NASH and its sequelae[12].

This article provides an overview of biomarker discovery and selection in the setting of NASH starting with some “knowledge-based” biomarkers. The list of biochemical markers provided in this review is not intended to be exhaustive; rather, a brief summary of some key biomarkers is provided.

BIOMARKERS OF HEPATOCYTE APOPTOSIS

To illustrate the opportunities and challenges related to the use of “knowledge-based” biomarkers of NASH, let us consider as an example biomarkers of hepatocyte apoptosis. Growing evidence has now accrued that hepatocellular apoptosis plays a central role in NAFLD progression[9]. Interestingly, data from animal and human studies have suggested that apoptosis is prominent in NASH but not in simple steatosis[13]. Cytokeratin 18 (CK-18) is the major intermediate filament protein in the liver and one of the most prominent substrates of caspases during hepatocyte apoptosis. Apoptotic cell death of hepatocytes is associated with release of caspase-cleaved CK-18 fragments into the bloodstream[14], and several studies have demonstrated elevation of these molecules in the context of NAFLD. A pilot study by Wieckowska et al [14] was the first to measure the levels of caspase-generated CK-18 fragments in patients with NAFLD. Results showed that caspase-cleaved CK-18 fragments were significantly higher in NAFLD patients than in controls, and that levels of these molecules correlated with the presence of liver fibrosis[14]. In line with these results, the usefulness of CK-18 fragments for the diagnosis of NAFLD was subsequently confirmed in obese patients[15], as well as in pediatric populations[16]. Interestingly, caspase-generated CK-18 fragments released in NAFLD also serve as an indicator of hepatic inflammation. The increased apoptic rate as a consequence of the hepatic inflammatory response is reflected by elevation of serum CK-18 fragments that may therefore distinguish NASH from simple steatosis[17]. These results have been further confirmed; even in NAFLD patients with normal aminotransferase levels[18].

Although CK-18 fragments can be used to identify patients with NASH and these molecules are currently being incorporated into multimarker schemes[15], their routine use in patients with suspected NAFLD depends on the answers to several questions: Are the sensitivity and specificity of these tests similar to those obtained with liver biopsy? Where is the test being performed (Gastroenterology Department, Physician’s Office, Outpatient Unit)? What discrimination limits should be used? Is the test being performed for diagnosis or for prognosis? Answers to these questions are part of the scientific evaluation process that is critical for assessing whether the information gained from CK-18 fragments is worth its cost to the healthcare system. Such answers require the performance of large studies to evaluate the relationship of caspase-cleaved CK-18 fragments with histological phenotypes of interest in the liver and, when applicable, the conduction of clinical trials to relate caspase-cleaved CK-18 fragments to disease risk and to therapeutic responses in patients with NAFLD.

BIOMARKERS OF THE METABOLIC SYNDROME

Although the pathogenesis of NAFLD is clearly multifactorial, this condition is currently considered as the hepatic manifestation of the metabolic syndrome. Comprehensive assessments of the role played by insulin resistance in the pathogenesis of NAFLD have also been published recently[19-21]. The present review will not attempt to replicate these texts. On the other hand, we aim to provide here a concise overview of biomarkers of the metabolic syndrome (leptin, adiponectin, resistin, soluble RAGE) in the setting of NAFLD, including a display of the evidence linking them to NASH.

Leptin is a peptide hormone, mainly produced by adipocytes, that plays a central role in the regulation of body weight[22]. In human liver, leptin has been shown to attenuate a number of insulin-induced activities ultimately resulting in insulin resistance[23]. Furthermore, a proinflammatory and profibrogenic activity of leptin has been reported[24]. Since leptin has been linked to metabolic abnormalities and insulin resistance, its potential role in NAFLD has been the focus of much investigation. Compared with controls, significantly higher levels of leptin have been observed in patients with NAFLD[25,26] and in those with NASH[27,28]. However, there are inconsistencies in published literature, with some authors showing an unaltered level of leptin in the setting of NAFLD[29]. In addition, serum leptin levels showed no correlation with liver fibrosis[30].

Adiponectin is an adipose tissue-specific protein whose receptors are expressed in several cell types including hepatocytes[31]. Main functions of this molecule comprise of the downregulation of inflammatory processes, the promotion of lipolysis, and the prevention of lipid accumulation[32]. Adiponectin is known to exert antifibrogenic and antiestrogenic effects in the liver[33]. Hypoadiponectinemia has been suggested as contributing to insulin resistance and the metabolic syndrome[34], and several studies have reported decreased adiponectin levels in patients with liver steatosis as compared with controls[35-37]. Of note, some authors have also demonstrated that lower adiponectin levels are associated with more extensive necroinflammation in the setting of NAFLD and that they may contribute to the development of NASH[38,39]. However, controversy surrounds the role of decreased adiponectin level as a reliable laboratory marker of NASH[40].

Resistin is a 10 kDa protein of 94 amino acids highly expressed in the adipose tissue. It is a major determinant of hepatic insulin resistance induced by high-fat diet in animal models[41]. Although human data regarding the role of this adipokine in insulin sensitivity and the metabolic syndrome are controversial[42], preliminary evidence seems to suggest a potential role of this adipokine in the pathogenesis of NAFLD. Pagano et al [43] have initially shown that patients with NAFLD are characterized by higher serum resistin levels in association with the NASH score, an index that takes into account necrosis, inflammation, and fibrosis in liver biopsies. However, no correlation was found between insulin resistance and hepatic steatosis score[43]. These findings were recently replicated by Jiang et al[44], who showed that serum resistin levels were significantly elevated in patients with NAFLD compared to controls. In contrast, Cho et al [40] failed to show altered levels of this molecule in Korean male patients.

Besides classical adipokines, levels of soluble receptor for advanced glycation endproducts (sRAGE) have been recently linked to insulin resistance and several components of the metabolic syndrome[45]. Recently, Yilmaz et al[46] have investigated concentrations of sRAGE across the spectrum of NAFLD. Levels of sRAGE were significantly lower in patients with definite NASH and borderline NASH compared to controls. Interestingly, levels of sRAGE were significantly and inversely correlated with serum aminotransferase, indicating that lower concentrations of sRAGE are associated with the most severe forms of NAFLD[46].

Altogether, our investigations indicate that a flurry of case-control studies relating biomarkers of the metabolic syndrome to NAFLD and NASH have been conducted in the past decade. Most of the available data are on adipokines, and much less is known about soluble RAGE. Of note, most studies to date also have been carried out in groups of people of Caucasian ancestry, and there are few data on African populations or African Americans. Furthermore, matching of cases and controls limits comparisons across sex and age demographics. Finally, the use of different assays across the studies also makes it difficult to define cut-off points. Additional larger studies of more diverse populations, including a full range of potential confounding variables, would be helpful at this juncture.

MISCELLANEOUS “KNOWLEDGE-BASED” BIOMARKERS

By using the “knowledge-based” approach, numerous other potential biomarkers of NAFLD and NASH have been explored in pilot cross-sectional studies. Angiopoietin-like protein 3 (ANGPTL3) is a liver-derived plasma protein that modulates plasma triglyceride clearance[47], which has been recently investigated in the setting of NAFLD[48]. Levels of ANGPTL3 have been found to be higher in patients with NASH than in those with simple fatty liver. Nonetheless, no association with histological staging and pathological characteristics of NAFLD was seen[48]. Pentraxin 3 (PTX3) is an acute-phase reactant that reflects the tissue inflammatory response[49]. It has been recently demonstrated that plasma PTX3 levels may differentiate NASH patients from non-NASH subjects, and that higher plasma PTX3 levels are associated with severe stages of hepatic fibrosis[50]. However, it is unknown whether any of these biochemical markers are involved in the causal chain of NAFLD progression, mediating the effects of other risk factors (e.g. insulin resistance or inflammation) or whether they merely reflect the presence of NAFLD. Additional basic and clinical research will likely shed more light on these issues.

PROTEOMIC APPROACHES

Proteomic approaches to the identification of NASH biomarkers rely principally on the unbiased comparative analysis of protein expression in normal and diseased liver tissues to identify aberrantly expressed proteins that may represent new biomarkers, as well as direct serum protein profiling[12]. Proteomics methodologies include the isolation, identification, and quantification of proteins in biosamples by adsorption of proteins to activated surfaces (matrix-assisted laser desorption-ionization technology), or via peptide ionization procedures and mass spectroscopy. Mass spectrometry can yield a comprehensive profile of peptides and proteins in biosamples without the need for initial protein separation, thereby facilitating biomarker identification with reduced sample requirements and a high throughput[12].

Only two studies to date have examined the NAFLD proteome. Younossi et al [51] investigated, by means of surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF), serum protein profiles of different subtypes of NAFLD and identified twelve significantly different protein peaks across the study groups. Using a similar SELDI-TOF technique, Trak-Smayra et al[52] have searched for serum markers of steatosis and NASH in obese nonalcoholic patient candidates for bariatric surgery. The authors identified two peaks of 7558.4 and 7924.2 m/z that may distinguish between NASH patients and controls. Such peaks were identified as being the double charged ions of hemoglobin-alpha and hemoglobin-beta. There were also three peaks, the intensity of which increased significantly according to severity of liver lesions (steatosis and NASH), but no association was seen with either liver function tests or metabolic parameters[52].

Although further research of the NAFLD/NASH proteome is certainly needed, the global analysis of protein expression represents an important paradigm shift from the traditional single-molecule approach to the evaluation of protein networks. The future availability of rapid, high-throughput analytical platforms are likely to facilitate molecular phenotyping of different subtypes (simple fatty liver, borderline NASH, definitive NASH) in the spectrum of NAFLD.

CONCLUSION

The need to carry out a biopsy to distinguish NASH from simple steatosis has impeded research to develop strategies for interventions to treat steatohepatitis. In this context, the overall expectation of a NASH biomarker is to enhance the ability of the clinician to optimally manage the patient without the use of liver biopsy[53]. Theoretically, biomarkers may provide a powerful approach in the understanding of the spectrum of NAFLD, with potential applications in several areas including screening, diagnosis, prognosis, and therapeutic monitoring. Advances in proteomics, metabolomics, and bioinformatics have revolutionized unbiased investigations of several biomarkers that may be informative with regard to the various stages of NAFLD, including NASH and its potential sequelae (advanced fibrosis, hepatocellular carcinoma, end-stage liver disease). Obviously, a crucial prerequisite for the clinical use of biomarkers is elucidation of analytical features, standardization of analytical methods, assessment of performance characteristics, and demonstration of cost-effectiveness. A new biomarker of NASH will be of clinical value only if it is reproducibly obtained in a standardized fashion, it is easy to interpret by clinicians, and if it has high sensitivity and high specificity for identifying NASH, and for distinguishing this condition from benign simple fatty liver. Establishing the prognostic utility of a biomarker in the setting of NAFLD to identify patients that may progress to NASH and fibrosis is more challenging because it requires a prospective design, and serial liver biopsies are presently the gold standard. Although there is evidence to suggest that currently available biomarkers (and their combination) have an increasing ability to distinguish “case” (NASH) from “noncase” (not-NASH) in cross-sectional studies, the step ahead in this field is to differentiate by the use of a biochemical marker “those who will develop fibrosis and end-stage liver disease” from “those who will not” in longitudinal investigations. Hopefully, the ongoing research in NASH biomarker development will also mandate a systematic organization of data that may facilitate the online sharing of biomarker metadata among researchers. Over the next years, technological advances will likely facilitate the use of multimarker profiling to replace the gold standard for the diagnosis of NASH, liver biopsy, with a “biomarker biopsy” [54].