Biomarkers for the risk stratification of non-alcoholic fatty liver disease (NAFLD) in individuals with obesity

Abstract

Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disease with a diverse histological spectrum, ranging from benign simple steatosis (SS) to the more severe non-alcoholic steatohepatitis (NASH). NASH significantly increases the risk of serious liver complications such as cirrhosis and hepatocellular carcinoma and is a leading cause of liver transplantation. Despite its growing prevalence, there is a lack of non-invasive diagnosis nor approved pharmacotherapy. Liver biopsy, the gold standard for diagnosing and staging NAFLD, is fraught with challenges including invasiveness, sampling error, high costs, risk of complications, patient reluctance, and variability in pathologist interpretations. To address this, several studies have explored various non-invasive approaches, including biomarkers, imaging techniques, and scoring systems, to diagnose and stage NAFLD. The combination of multiple non-invasive tools has shown promise in improving the accuracy of diagnosing and staging NAFLD.

NAFLD's prevalence is notably higher in obese individuals with metabolic syndrome, indicating metabolic dysregulation is the primary driver in the onset and progression of NAFLD. As a result, international experts suggest renaming NAFLD to Metabolic Dysfunction Associated Steatotic Liver Disease (MASLD) to fully reflect its aetiology. Several targeted and untargeted metabolomic studies have identified specific changes in amino acids, glutathione metabolism, and lipid profiles across various cohorts, offering valuable insights into the pathophysiology and disrupted metabolic pathways in NAFLD and NASH. However, there are large variations and inconsistencies due to limited sample size, differences in study population, and varying diagnostic criteria. Furthermore, there is a lack of studies focusing on Asian populations with biopsy-proven NAFLD.

This study aimed to leverage our unique clinical cohort of obese Chinese, which covers the full spectrum of NAFLD, to systematically search for transcripts, metabolites, and lipid species associated with different stages of NAFLD and evaluate their potential in predicting NAFLD and NASH risks in obese individual, as well as a potential target for therapeutic treatments and biomarker development.

A total of 263 metabolites and 550 lipid species were detected in serum samples from 250 NAFLD patients. Differential analysis and pathway enrichment analyses revealed the progressive patterns in metabolic mechanisms during the transition from normal liver to SS and to NASH, including N-palmitoyltaurine, tridecylic acid, and branched-chain amino acid signalling pathways. The co-expression network showed a distinct correlation between different triglyceride and phosphatidylcholine species at different levels of disease severity. Multiple machine learning models were constructed to differentiate NAFLD stages and showed significant predictive improvements over traditional clinical metrics. The model integrating metabolites, lipid species, and routine clinical parameters achieved an area under the curve (AUC) of 0.954 for distinguishing NAFLD and normal liver, and an AUC of 0.915 between SS and NASH, compared to routine clinical parameters with an AUC of 0.882 and 0.785, respectively. RNA sequencing shows Trem2, Fabp4, Mmp9 are significantly upregulated between healthy liver and NAFLD. RT-qPCR performed on murine liver fed with high-fat diet (HFD) showed significant upregulation for Trem2 but not for Trem1 and Mmp9 when compared to control group. In-depth literature review was performed for the selected genes to understand their mechanism and application.