Near‑infrared spectroscopy of plasma amino acids with chemometrics towards breast cancer discrimination

Abstract

Breast cancer(BC) is the most prevalent cancer and the second-leading cause of cancer-related death for women in the worldwide. BC cells need more amino acids to meet the demand of rapid proliferation. In this paper, we focus on the role of near‑infrared (NIR) spectroscopy to analyse blood plasma samples of breast cancer patients to differentiate healthy controls. The possibility of quantitative detection of 20 amino acids in breast cancer plasma by NIR spectroscopy was investigated for the first time in this study. 180 samples (80 BC patients, 30 benign breast disease patients and 70 healthy controls) were analysed. Canonical correlation analysis (CCA) was used to analyse the relationship between clinical biochemical parameters and amino acid metabolic profile for BC patients. In this study, plasma glutamine, histidine, threonine, proline, phenylalanine content of BC patients was higher than healthy control plasma (p < 0.05) by NIR spectroscopy. There was overall correlation (r = 0.935) between three clinical parameters(age, albumin, total triglyceride) and four amino acids (glutamic acid, tyrosine, valine and lysine) from BC patients. In this study，we have built quantitative and qualitative model which is used to detect plasma amino acids of BC patients by NIR spectroscopy has good performance. It fully demonstrated the great potential and advantages of NIR spectroscopy combined with chemometrics in breast cancer research, which can provide a new research strategy and technical platform for the study of plasma amino acid metabolism in breast cancer patients, and provide basic experimental data reference for clinical diagnosis and treatment of breast cancer patients.