The Preliminary Research On Discriminate Breast Cancer Patients Based On Detecting Metabolic Targets Using HPLC And Near Infrared Spectrum

Objective:This research intends to build rapid detection of breast cancer patients'quantitative model of plasma amino acids and qualitative model of its classification that recognized by near infrared spectroscopy,which could provide the experiental data for the research on metabolism of amino acid spectrum characteristics of breast cancer patients.Methods:1.Through the optimization of chromatographic conditions and different pretreatments,we will establish a pre-column-derivation HPLC method for the determination of 20 kinds of amino acid from plasma,then compare with different amino acid contents among breast cancer,breast benign disease and healthy women.2.Through comparing different optical path colorimetric cup,scanning resolution and scanning times influce on the model prediction results,we will establish the optimal detection performance quantitative model of plasma amino acid content of breast cancer patients.3.Through selecting different spectral pretreatment methods,we will establish the quickly identifying qualitative model of mammary gland disease plasma by near infrared spectral technology,and then discuss the classification results of NIR qualitative model concerning nthe relationship of plasma component content.4.We will use canonical correlation analysis by SPSS software to investigate the relationship between metabolism of plasma amino acid pattern of breast cancer and the clinical parameters.Results:1.The established HPLC method within the scope of the detection of amino acid containing quantity had good linear relationship with the peak area,which is simple,rapid and low cost;2.Through the optimized conditions,choose 1 mm long optical path than color plate,the resolution is set to 8 cm^-1,the number of scanning for64 times,and the optimal sampling background sampling effect in water,the established quantitative model 20 kinds of amino acids in plasma can be better applied to detection of unknown samples.3.The established model of near infrared spectra classification for calibrating the recognition accuracy of 100%,validation set accuracy is 98%,and the model and the component content about 20 kinds of amino acids in the plasma.4.Breast benign compared with those of healthy women and patients with breast cancer in plasma Gln,his,Thr,Pro,Phe content compared statistically significant difference?p<0.05?.Three clinical parameters of patients with breast cancer,albumin and total triglyceride?age?and four kinds of amino acid?Glu,Tyr,Val and Lys?has the overall correlation?r=0.935?Conclusion:breast cancer patients in this study plasma amino acids in near infrared quantitative and qualitative model has good performance,but for the study of breast cancer patients with plasma amino acid metabolism characteristics provide support on the experimental data,the reference can provide basic data for clinical analysis.