Refining Noninvasive Diagnostics In Nonalcoholic Fatty Liver Disease: Closing the Gap to Detect Advanced Fibrosis

See article on page 1075 Potential conflict of interest: Dr. Loomba consults for Bird Rock Bio; Celgene; Enanta; GRI Bio; Metacrine; NGM; Receptos; Sanofi; Arrowhead Research; GIR, Inc.; and Metacrine, Inc. He consults for and advises Madrigal, NGM, and Galmed. He advises Allergan; BMS; BI; Daiichi‐Sankyo, Inc.; Eli‐Lilly; Galectin; GE; Genfit; Intercept; Janssen, Inc.; Merck; Pfizer; Prometheus; Siemens; and Sirius. Dr. Schattenberg consults for and advises Gilead Sciences. He consults for Intercept Pharmaceuticals, Genfit, Medimmune, Novartis, Pfizer, and Boehringer Ingelheim. The epidemic of nonalcoholic fatty liver disease (NAFLD) has created a large demand for diagnostic tools to identify patients at risk of disease progression, including the development of end‐stage liver disease. The prevalence of NAFLD is rising worldwide, and it is estimated that approximately 1 billion individuals may be afflicted with NAFLD globally.1 Currently, the gold standard to diagnose advanced fibrosis, and therefore the reference for all noninvasive technology, is liver histology and its assessment by an experienced liver histopathologist. While training and experience can improve the interobserver variation when reading liver histology, sampling error always remains a source of concern in addition to the invasive nature of the procedure and associated adverse events including pain, infection, bleeding, and, extremely rarely, even death. Therefore, development of biomarkers of fibrosis is a major unmet need in the field to overcome these limitations. In the clinical setting, the relevant questions are (1) Is there lobular inflammation? and (2) How much of a, and which, pattern of fibrosis is present? To diagnose nonalcoholic steatohepatitis (NASH), lobular inflammation and hepatocyte ballooning are required in addition to steatosis, while the amount and pattern of fibrosis reflect the degree of injury that has accumulated over time. Although assessment of liver histology for characterization of pattern of injury and fibrosis will still require a liver biopsy assessment, noninvasive quantitative assessment of liver fibrosis may have prognostic significance and may eliminate the need for liver biopsy in the majority of individuals in the coming years. Recent studies have shown that presence of fibrosis, especially advanced fibrosis, is associated with increased risk of all‐cause and liver‐related mortality.2 Current guidelines recommend using noninvasive clinical prediction rules such as the NAFLD fibrosis score (NFS) and the FIB‐4 index to rule out patients with advanced fibrosis.3 The NFS is based on six readily available variables (age, body mass index, hyperglycemia, platelet count, albumin, and aspartate aminotransferase [AST]/alanine aminotransferase [ALT] ratio) and is calculated using an openly accessible and published formula (http://nafldscore.com/).4 However, serum ALT and AST, which are included in the NFS and other noninvasive surrogate scores, are influenced by the degree of hepatic inflammation, as reflected by an increased ALT and as ALT tends to decrease with age, making adaptions in populations >65 years of age necessary to achieve comparable specificity compared to younger patents.5 Therefore, the development of surrogate scores that are less dependent on serum ALT but include a marker that is perhaps more closely linked to hepatic fibrogenesis or fibrolysis and reflects extracellular matrix remodeling in liver tissue could be potentially helpful to improve accuracy.6 Among these, N‐terminal type III collagen propeptide (PRO‐C3), a marker of type III collagen formation that can be measured by enzyme‐linked immunosorbent assay, is emerging as a useful tool and is currently being explored in clinical trials to assess treatment response. A recent report performed a post hoc analysis of two independent clinical trials that included patients with hepatitis C virus and diabetes type 2 and found that PRO‐C3 was able to identify patients who progressed in fibrosis score on liver biopsy, suggesting that PRO.C3 could be used as a predictive marker.7 In the current issue of Hepatology, Samuel and colleagues propose the ADAPT scores, comprising age, diabetes type 2, PRO‐C3, and platelets, to detect advanced fibrosis in patients with histologically confirmed NAFLD.8 In two well‐characterized cohorts of patients with biopsy‐proven NAFLD comprising 150 patients in the derivation cohort and 277 patients in the validation cohort derived from tertiary‐care medical centres in Sydney and Nedlands (Australia), Nottingham (United Kingdom), and Kurume (Japan), the ADAPT showed areas under the receiver operating characteristics curve (AUROC) of 0.86 (95% confidence interval [CI] 0.79‐0.91) in the derivation cohort and 0.87 in the validation cohort (95% CI 0.83‐0.91) for the detection of advanced fibrosis. While these results are comparable to ALT/AST ratio or the NFS in other published cohorts,9 the substitution of ALT and AST with a marker that more closely correlates with extracellular matrix turnover is likely to produce more robust results over different cohorts and countries. Still, these results are below the accuracy of modern imaging‐based biomarkers, albeit similar to an AUROC of 0.83 using vibration‐controlled elastography (VCTE) but a significantly higher AUROC of 0.93 using magnetic resonance elastography (MRE) for detection of stage 3 and stage 4 fibrosis in NAFLD.10 A head‐to‐head comparison as well as a combination approach with ADAPT, VCTE, and the enhanced liver fibrosis panel will be helpful in assessing if one could achieve an AUROC similar to MRE. Furthermore, it is also possible to rule in stage 2 or higher fibrosis by such a combination approach to find individuals who meet the criteria for treatment in the setting of a registration trial. At this point, it seems too early to suggest that ADAPT could be used to screen an asymptomatic population with no known liver disease because the significantly lower a priori likelihood of advanced fibrosis will result in lower test performance, and further studies are needed in the context of use before making such a recommendation. In the current study, 42.6% of patients had NASH. Also, the results will have to be validated beyond Australia, Japan, and the United Kingdom. Future studies need to focus on ruling in those with stage 2 or higher fibrosis as the field is reaching a point where individuals may be ruled out for advanced fibrosis with good accuracy but a positive test does not translate into a high positive predictive value of having advanced fibrosis if confirmed. In summary, ADAPT is a compound surrogate score that will advance the available noninvasive screening options for patients with NAFLD in the future. Its strength results from the inclusion of an innovative and easily measurable marker of extracellular matrix turnover in the liver. While it may be a bit premature to propagate its use as a stand‐alone test in the general population, it seems that the stage is set to explore its accuracy in high‐risk populations, e.g., type 2 diabetic patients who are at risk for advanced fibrosis. Over the next 5 years, two major NAFLD biomarker initiatives, the Non‐Invasive BioMarkers of Metabolic Liver Disease (NIMBLE) and the Liver Investigation: Testing Marker Utility in Steatohepatitis (LITMUS) Consortium, are likely to further advance our knowledge and validate the individual as well as combined diagnostic accuracy of many of these biomarkers and panels and help refine clinical practice.