Abstract
Breast cancer (BC) is the most prevalent cancer type in women worldwide. BC cells need more amino acids to meet the demands of rapid proliferation. The present study focused on the role of near-infrared (NIR) spectroscopy in analyzing blood plasma samples of patients with BC to differentiate them from healthy controls. The possibility of quantitative detection of 20 amino acids in BC plasma by NIR spectroscopy was investigated. A total of 180 samples (from 80 patients with BC, 30 patients with benign breast disease and 70 healthy controls) were analyzed. Canonical correlation analysis was used to analyze the relationship between clinical biochemical parameters and the amino acid metabolic profile of patients with BC. In the study, plasma glutamine, histidine, threonine, proline and phenylalanine content of patients with BC was higher than that in plasma from healthy controls (P<0.05) based on NIR spectroscopy. There was an overall correlation (r=0.935) between three clinical parameters (age, albumin and total triglyceride) and four amino acids (glutamic acid, tyrosine, valine and lysine) in patients with BC. In this study, a quantitative and qualitative model was constructed, and this was used to detect plasma amino acids of patients with BC using NIR spectroscopy with good performance. It fully demonstrated the potential and advantages of NIR spectroscopy combined with chemometrics in BC research, which can provide a novel research strategy and technical platform for the study of plasma amino acid metabolism of BC and provide basic experimental for the clinical diagnosis and treatment of BC.