ABSTRACT
Objective To characterize the biomarkers of urine samples for early diagnosis of colorectal cancer(CRC)using proton nuclear magnetic resonance (1H-NMR)combined with pattern recognition.Methods 400 MHz 1H-NMR was used to test the urine samples obtained from 23 patients with Ⅰ/Ⅱ stage CRC,40 healthy controls (HC)and 18 patients with esophageal cancer (EC). Pattern recognition through orthogonal partial least squares-discriminant analysis (OPLS-DA)was applied on 1H-NMR data to find urine metabolic differences between CRC and HC.Results OPLS-DA could effectively determine HC,patients withⅠ/Ⅱstage CRC and patients with esophageal cancer.Compared with HC,early stage CRC had significant decreases of choline,isocitric acid,lactamine,phenylalan, cysteine,creatinine,aspartic acid,hippurate acid,methylamine,dimethyl sulfone,and increases of acetoacetate,glutamine,glycocyamine,cis-aconitate, trans-aconitate,homocycteine in the urine samples.Conclusion Urine metabonomics based on NMRIndicates that glucose metabolism,amino acid metabolism,choline metabolism,energy metabolism and intestinal microflora are disturbance in colorectal cancer patients,which provide valuable metabolic information on the molecular level for early diagnosis of colorectal cancer.