Potential Hepatic Lipid Markers Associated with Nonalcoholic Steatohepatitis and Fibrosis in Morbid Obesity Patients

Abstract

This study investigated differences in lipidomic profile features in nonalcoholic steatohepatitis (NASH) between mild and significant liver fibrosis cases among patients with morbid obesity. Wedge liver biopsy was performed during sleeve gastrectomy and significant liver fibrosis was defined as a fibrosis score ≥ 2. We selected patients with NASH with non/mild fibrosis (stage F0-F1; n = 30) and NASH with significant fibrosis (stage F2-F4; n = 30). The results of the liver tissue lipidomic analysis revealed that the fold changes of triglyceride (TG) (52:6); cholesterol ester (CE) (20:1); phosphatidylcholine (PC) (38:0) and (50:8); phosphatidic acid (PA) (40:4); phosphatidylinositol (PI) (49:4); phosphatidylglycerol (PG) (40:2); and sphingomyelin (SM) (35:0) and (37:0) were significantly lower in patients with NASH with F2-F4 than those with NASH with F0-F1 (p < 0.05). However, the fold changes of PC (42:4) were relatively higher in patients with NASH with stage 2-4 fibrosis (p < 0.05). Moreover, predictive models incorporating serum markers levels, ultrasonographic studies, and levels of specific lipid components [PC (42:4) and PG (40:2)] yielded the highest area under receiver operating curve (0.941), suggesting a potential correlation between NASH fibrosis stages and liver lipid accumulation among specific lipid species subclasses. This study demonstrated that the concentrations of particular lipid species in the liver correlate with NASH fibrosis stages and may indicate hepatic steatosis regression or progression in patients with morbid obesity.

Keywords: lipid profile analysis; liver fibrosis; nonalcoholic steatohepatitis.

Figure 1

Lipidomic identification of liver lipid subclass (A) Distribution of significant lipid subclasses TG, SM, PG, PA, PE, PC, PI, CE; (B) Normalized relative abundance presented as median ± quartile of selected lipid species grouped by F0–1, F2–4. TG: triglyceride; SM: sphingomyelin; PG: phosphatidylglycerol; PA: phosphatidic acid; CE: cholesterol ester; PE: phosphatidylethanolamine; PC: phosphatidylcholine; PI: phosphatidylinositol; F0–1, grade 0–1 fibrosis; F2–4, grade 2–4 fibrosis.

Figure 2

Differentiation of the significant lipid species in two groups. (A) The volcano plot illustrates the fold change in x axis and p value < 0.05 in y axis. The red dots indicate the fold change over 1 between NASH F0–1 and F2–4, while the blue dots represent the negative fold change.; (B) Heat maps of the F0–1/F2–4 for significant change lipids, with rows describing lipid species and columns representing NASH fibrosis samples.; (C) Normalized relative abundance presented as median ± quartile of the 9 significant lipid species (p < 0.05) in NASH F0–1 vs. F2–4.

Figure 3

PLS–DA analysis and correlations between significant lipid species and clinical characteristics (A) PLS-DA plot analysis of liver lipidomic differentiating patients with F0–1/F2–4. (B) Volcano plot of lipid metabolites and clinical characteristics of NASH. X axis shows fold changes for NASH F2–4/F0–1, y axis represents p value cut off point (0.05). (C) Correlation matrix calculated using Pearson correlation, indicating 13 significant lipid metabolites and key baseline characteristics. Colour corresponds to coefficient value between −0.6 and 0.6.

Figure 4

Risk of lipid metabolite with significant versus non/mild fibrosis Odds ratio of each significant lipid species calculated using univariate logistic regression analysis, depicted in forest plot. Metabolites labelled with * have a statistically significant p value (p < 0.05).

Figure 5

Area under receiver operating curve of models Lipid metabolites with adjusted p value < 0.05 were applied to statistical model. APRI, LSM, US fibrosis score, PC (42:4) and PG (40:2) were applied to predictive model.