Using post‐column infused internal standard assisted quantitative metabolomics for establishing prediction models for breast cancer detection

• Published in: Rapid communications in mass spectrometry Vol. 34; no. S1; pp. e8581 - n/a
• Main Authors: Huang, Marisa, Li, Hung‐Yuan, Liao, Hsiao‐Wei, Lin, Ching‐Hung, Wang, Chin‐Yi, Kuo, Wen‐Hung, Kuo, Ching‐Hua
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01.04.2020
• Subjects: Adult; Aged; Aged, 80 and over; Breast cancer; Breast Neoplasms - diagnosis; Breast Neoplasms - metabolism; Chromatography, High Pressure Liquid - methods; Citric acid; Female; Health risk assessment; Humans; Identification methods; Lactic acid; Liquid chromatography; Mass spectrometry; Metabolites; Metabolome; Metabolomics - methods; Middle Aged; Prediction models; Prognosis; Strategy; Tandem Mass Spectrometry - methods
• ISSN: 0951-4198; 1097-0231; 1097-0231
• DOI: 10.1002/rcm.8581

Rationale Breast cancer is one of the most common cancers among women and its associated mortality is on the rise. Metabolomics is a potential strategy for breast cancer detection. The post‐column infused internal standard (PCI‐IS)‐assisted liquid chromatography/tandem mass spectrometry (LC/MS/MS) method has been demonstrated as an effective strategy for quantitative metabolomics. In this study, we evaluated the performance of targeted metabolomics with the PCI‐IS quantification method to identify women with breast cancer. Methods We used metabolite profiling to identify 17 dysregulated metabolites in breast cancer patients. Two LC/MS/MS methods in combination with the PCI‐IS strategy were developed to quantify these metabolites in plasma samples. Detection models were built through the analysis of plasma samples from 176 subjects consisting of healthy volunteers and breast cancer patients. Results Three isotope standards were selected as the PCI‐ISs for the metabolites. The accuracy was within 82.8–114.16%, except for citric acid and lactic acid at high concentration levels. The repeatability and intermediate precision were all lower than 15% relative standard deviation. We have identified several metabolites that indicate the presence of breast cancer. The area under the receiver operating characteristics (AUROC) curve, sensitivity and specificity of the linear combinations of metabolite concentrations and age with the highest AUROC were 0.940 (0.889–0.992), 88.4% and 94.2% for pre‐menopausal woman, respectively, and 0.828 (0.734–0.922), 73.5% and 85.1% for post‐menopausal women, respectively. Conclusions The targeted metabolomics with PCI‐IS quantification method successfully established prediction models for breast cancer detection. Further study is essential to validate these proposed markers.