ABSTRACT
PURPOSE@#For early detection of breast cancer, tests with high sensitivity and specificity are needed. Metabolomics, the study of chemical processes involving metabolites, can be used to identify diagnostic biomarkers for a variety of types of cancers. In this study we identified biomarkers of breast cancer by profiling urinary metabolites.@*METHODS@#We performed metabolite profiling of 30 urine samples from 14 patients with breast cancer and 16 healthy controls by liquid chromatography–mass spectrometry. An orthogonal partial least squares-discriminant analysis (OPLS-DA), Student's t-test, and receiver operating characteristic (ROC) analysis were performed to identify metabolites that were potential diagnostic biomarkers for breast cancer.@*RESULTS@#The OPLS-DA showed clear separation between the two groups. Of the 95 metabolites detected, 24 potential biomarkers were identified by Student's t-test. A ROC analysis showed that concentrations of N-(2-furoyl) glycine, histidine, and D-tagarose were significantly higher (area under the ROC curve [AUC] >0.7) and those of trigonellinamide, L-galacto-2-heptulose, creatinine, and xanthine were significantly lower (AUC ≥0.8) in the patients with breast cancer than in the healthy controls.@*CONCLUSION@#Measurement of the concentrations of urinary metabolites can be used to screen for early breast cancer. We plan to explore diagnostic biomarkers of breast cancer in blood and urine further in a larger study.