摘要：

Among women in the U.S., breast cancer is the most commonly diagnosed cancer and the second leading cause of cancer death (after lung cancer). Metabolomics, an approach to the study of small molecules, provides insight of characteristic biochemical phenotypes in disease, and facilitates the development of novel diagnostic tools. This thesis project was to investigate the metabolic signature and to identify potential biomarkers for breast cancer using metabolomics methods. GC-TOFMS and LC-TOFMS spectra were acquired in the plasma collected from 138 breast cancer patients and 76 healthy women. Multivariate and univariate statistics methods were applied to analyze the metabolic alterations in breast cancer. Of the 41 identified differential metabolites, aspartate was the most significantly reduced in breast cancer plasma samples and obtained good predictive power for distinguishing breast cancer patients from healthy controls. An established combination of 7 markers (asparagine, hypotaurine, 5-oxoproline, cysteine, aspartate, glutamate and glutamine) was found to provide even better predictive power than aspartate alone. The altered expression of aspartate was confirmed in an independent set of serum samples and 20 pairs of breast tumor tissue and its adjacent normal tissue. It was also found that the metabolic profiles of stage I and stage IV patients can be separated in the constructed OPLS-DA model. In conclusion, breast cancer exhibits profound metabolic dysregulations and potential biomarkers in breast cancer can be identified using metabolomics approach.

展开