LC‐MS‐based urine metabolomics analyses for early diagnosis of medulloblastoma

Abstract

(MB) is the most common type of brain cancer in pediatric patients. In present study, we applied liquid chromatography ‐ mass spectrometry (LC‐MS) based metabolomics to analyze a cohort including 425 urine samples (118 healthy controls, 112 medulloblastoma patients, 33 patients with brain malignant cancers and 53 patients with brain benign diseases, 29 MB patients with surgery after a week and 80 MB patients with surgery after a month). Differential metabolites were analyzed to investigate if urine metabolites could differentiate MB from non‐MB. Apparent separation was observed between MB urine samples and healthy controls, brain benign diseases, other brain malignant tumor. The AUC in discovery group respectively were 0.94, 0.938, 0.808, and the AUC of external validation dataset were 0.913, 0.889, 0.822. According to the comparisons, 14 metabolites were found to be the specific metabolites of MB. Among 14 metabolites, two metabolites were validated in postoperative samples and with high AUC values (0.858 compare to non‐MB). Our data showed that urine metabolomics could be used for MB diagnosis and would provide candidates for the potential biomarkers of MB.Support or Funding InformationNational Basic Research Program of China (No. 2013CB530805, 2014CBA02005)