Non-invasive Evaluation of Non-alcoholic Fatty Liver Disease Using Biochemical and Genetic Markers

Background and aims: Non-alcoholic fatty liver disease (NAFLD) is one of the most common chronic liver diseases worldwide. It encloses a wide disease spectrum from simple steatosis to non-alcoholic steatohepatitis (NASH). While simple steatosis is thought to be benign, NASH may progress to end-stage liver disease and hepatocellular carcinoma. Traditionally, the diagnosis of NAFLD and, particularly, NASH relies on liver biopsy. It is an invasive procedure with poor acceptance and risk of major complications such as hemorrhage. Non-invasive evaluations of NAFLD and NASH are urgently needed. In this study, we tested the performance of different biochemical and genetic markers in the diagnosis and monitoring of NAFLD and NASH.;Methods: This study included 147 patients with biopsy-proven NAFLD (hospital cohort), 51 of whom also had per protocol follow-up liver biopsies 3 years later. In addition, 922 subjects from a population screening project (community cohort) underwent proton-magnetic resonance spectroscopy to determine intrahepatic triglyceride content (IHTG). NAFLD was diagnosed when IHTG was over 5%. Subjects with IHTG less than 5% and 5 other subjects with normal liver histology served as controls. Furthermore, 154 NAFLD subjects from the community cohort were enrolled in a prospective single-blinded trial comparing a community-based lifestyle intervention programme (n=77) and standard care (n=77). Cytokeratin-18 II (CK-18), adipocyte fatty acid binding protein (AFABP) and fibroblast growth factor 21 (FGF21) were measured by enzyme-linked immunosorbent assays. Patatin-like phospholipase domain containing 3 (PNPLA3) rs738409 gene polymorphism was determined by TaqManRTM SNP Genotyping Assay.;Results: CK-18 (including apoptosis marker CK-18 M30 and 2 total cell death markers CK-18 M65 and CK-18 M65ED) and FGF21 had high accuracy in diagnosing NAFLD (area under the receiver-operating characteristics curves [AUROC] 0.84-0.94) and moderate accuracy in diagnosing NASH (AUROC 0.66-0.71). AFABP only had moderate accuracy in diagnosing NAFLD (AUROC 0.63) and NASH (AUROC 0.63). Combined application of CK-18 M30 and FGF21 using a 2-step approach further improved the negative predictive value and positive predictive value to around 80%. Changes of M30 and M65ED had high AUROC of over 0.8 in predicting disease progression in the 51 patients who underwent paired liver biopsies. Changes in AFABP and FGF21 did not correlate with disease progression.;The PNPLA3 rs738409 GG genotype was associated with 2-fold increase in the risk of NAFLD independent of dietary pattern in the community. The GG genotype was also associated with more severe histological damage in hospital NAFLD patients. The community-based lifestyle intervention programme was sustainable and effective. Subjects with allele G were more sensitive to the programme. Patients with III GG genotype had an additional 6% absolute reduction in IHTG compared with those with CC genotype from lifestyle intervention. This reduction was accompanied with greater reduction in body weight, body mass index and total cholesterol.;Conclusion: Biomarkers CK-18 M30/M65/M65ED and FGF21 have high accuracy in diagnosing NAFLD and moderate accuracy in diagnosing NASH. A two-step approach combining CK-18 M30 and FGF21 further improves the accuracy in diagnosing NASH. Changes in CK-18 M30 and M65ED have high accuracy in predicting disease progression and may be used for serial monitoring. The GG genotype in PNPLA3 rs738409 is associated with increased risk of NAFLD independent of dietary pattern. Those patients with GG genotype were more sensitive to lifestyle intervention and thus should be encouraged to participate in such programmes.