Metabolic analysis of early nonalcoholic fatty liver disease in humans using liquid chromatography-mass spectrometry

• Main Authors: Cheng Hu, Tao Wang, Xiaoyu Zhuang, Qiaoli Sun, Xiaochun Wang, Hui Lin, Mingli Feng, Jiaqi Zhang, Qin Cao, Yuanye Jiang
• Format: Article
• Language: English
• Published: BMC 2021-04-01
• Series: Journal of Translational Medicine
• Subjects: NAFLD; Early diagnosis; Metabolomics; LC–MS/MS; Biomarkers
• Online Access: https://doi.org/10.1186/s12967-021-02820-7

Abstract Background Nonalcoholic fatty liver disease (NAFLD) is a common metabolic disease that affects 20–30% of individuals worldwide. Liver puncture remains the gold standard for the diagnosis of liver diseases despite limitations regarding invasive nature and sample variability. It is of great clinical significance to find noninvasive biomarkers to detect and predict NAFLD. Objective The aims of this study were to identify potential serum markers in individuals with early-stage NAFLD and to advance the mechanistic understanding of this disease using a high-throughput mass spectrometry-based untargeted metabolomics approach. Methods One hundred and twelve patients with early-stage NAFLD aged 18–55 were recruited according to the guidelines. The control group included 112 healthy participants. The demographic, anthropometric, clinical and laboratory data of all participants were systematically collected. Serum samples were obtained after an overnight fast. The comprehensive serum metabolomic analysis was performed by ultra-performance liquid chromatography-Orbitrap mass spectrometry. The resultant data was processed by Compound Discover and SIMCA-P software to validate the potential biomarkers. Significantly altered metabolites were evaluated by variable importance in projection value (VIP > 1) and ANOVA (p < 0.01). Pathway analysis was performed using MetaboAnalyst 4.0. Results The liver function test of early NAFLD patients showed no statistical differences to control group (p > 0.05). However, obvious differences in blood lipids were observed between subjects with NAFLD and controls (p < 0.001). In total, 55 metabolites showed significant changes in experimental group were identified. The area under curve (AUC) values deduced by receiver operating curve (ROC) analysis indicated that these newly identified biomarkers have high predictability and reliability. Of these, 15 metabolites with AUC greater than 0.9 were of great diagnostic value in early NAFLD patients. Conclusion In this study, a total of 15 serum metabolites were found to strongly associate with early NAFLD. These biomarkers may have great clinical significance in the early diagnosis of NAFLD, as well as to follow response to therapeutic interventions.