Extraction of Urinary Metabolite-Derived Biomarker Candidate for Breast Cancer

Urine is more than 90% water, and can contain more than 3000 metabolites. Consequently, extraction of biomarker candidates from urine for measuring subtle fluctuations in the quantities of metabolites in healthy individuals and cancer patients is very difficult. A variety of technologies, such as liquid chromatography/mass spectrometry (LC/MS), for analysis of urinary metabolites and identification of substances that show fluctuations in cancer patients compared with healthy individuals was used for this purpose. The substances detected by LC/MS were evaluated using machine learning, multivariate analysis to reduce and visualize parameters, and mass spectrometry for structural estimation of metabolites with unknown structures. One of the important biomarker candidates for breast cancer was determined to be 2-amino-4-methyl-1,1-pentanediol by an MS/MS spectrum. A one-dimensional linear discriminant analysis shows that this compound is estimated to be a promising biomarker for breast cancer. Urinary metabolites are found to be biomarkers for breast cancer.