Serum lipidomics reveals novel biomarkers by UHPLC-HRMS to assist the discrimination between colorectal adenoma and cancer

Abstract

Purpose The mechanisms of colorectal adenoma (CA) transforming colorectal cancer (CRC) remain unclear, and there are still lacking biomarkers to distinguish CA from CRC. Hence, this study aimed to explore underlying the mechanisms of CA canceration and novel biomarkers to differentiate CA from CRC. Methods An ultrahigh performance liquid chromatography-high resolution mass spectrometry (UHPLC-HRMS) method was employed to measure global lipid profiles of 100 serum samples from CA and CRC patients. Univariate and multivariate statistical analysis were applied to select the differential lipids between two groups. Combining constituent ratio analysis and performance evaluation analysis by the receiver operating characteristic (ROC) curve explored the mechanisms of CA canceration and lipid markers to discriminate CA and CRC. Results Serum lipid profiles presented obvious differences between CA and CRC. Totally 85 differential lipids (P < 0.05 and fold change > 1.5) were discovered and mainly involved in fatty acid (FA), phosphatidylcholine (PC), and triacylglycerol (TAG) metabolism with ratio > 63%. 7 lipids served as the putative biomarkers with the area under the curve (AUC) > 0.800. Among them, a lipid panel containing four lipids was selected as the potential biomarkers with excellent performance (AUC = 0.971) and the highest selected frequency to distinguish CA from CRC based on SVM classification model. Conclusion The lipid panel could act as the potential biomarkers aiding the early diagnosis of CRC. Metabolic dysregulation of FAs, PCs, and TAGs should be involved in CA canceration, which provides new clues to understand its underlying mechanisms.