This study used untargeted lipidomics to analyze the characteristics of lipid metabolism in the serum of women with antiphospholipid syndrome. Twenty female patients with APS and 20 healthy controls were recruited to this study. Untargeted lipidomics with liquid chromatography-tandem mass spectrometry was used to profile serum lipids. Correlation analyses were performed on the significant differential lipids. Least absolute shrinkage and selection operator (LASSO) regression and binary logistic regression analyses were performed to find possible biomarkers. Finally, receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance of biomarkers. The serum lipid profile of women with APS was significantly different from that of healthy controls. After statistical analysis and database matching, 12 lipids were found to be significantly different between the two groups, with the most obvious differences being triacylglycerols and phosphatidylcholines. LASSO regression and binary logistic regression models revealed that PC (17:0/22:6) and ACar (17:3) may be serum biomarkers of APS. ROC curve analysis shows that PC (17:0/22:6) and ACar (17:3) combined gave an area under the curve of 0.865 for the diagnosis of APS. PC (17:0/22:6) and ACar (17:3) could be used as biomarkers of APS.
Keywords: antiphospholipid syndrome; lipidomics; phosphatidylcholines; triacylglycerols.