Individualized Treatment Of Breast Cancer Based On Mass Spectrometry Data Analysis

In recent years, incidence of cancer has been a rising trend in China, among which, thebreast cancer has become a disease with highest death rate in female tumors. Although thecure rate of early breast cancer is high, due to the big development gap between urban andrural medical levels and the weak consciousness of disease prevention and early detection,most patients are diagnosed at their later period of breast cancer. Therefore, while promotingearly breast cancer detection, early diagnosis and early treatment, it’s also necessary topromote the treatment level of breast cancer and find out a more effective treatment programand strategy.Currently, the clinical and pathological features of breast cancer, including tumor size,lymph node status and grade malignancy, are usually used as a basis for treatment plan.However, breast cancer is a molecularly heterogeneous tumor, which cannot be fullyexplained by general pathologic factors. As a result, insufficient or excessive treatment oftenoccurs. Nevertheless, protein is encoded by the gene and can reflect a highly dynamic andaccurate biological state. It is an important source to excavate potential biomarkers and able toadapt to clinical needs.In this paper, we analyze the features of breast cancer in the protein molecular level, tofind evaluation indices to accurately predict prognostic risk, and thus assist individualizedtreatment by two aspects of pathological factors and proteomic.By virtue of the clinical breast cancer data samples provided by Zhejiang CancerHospital, this paper analyzes the relationship between pathological factors and breast cancerprognosis. It compares the pathological factors and overall survivals of patients withtriple-negative breast cancer and non-triple-negative breast cancer with mathematical statisticmethod, analyzes the pathological factors affecting prognosis of patients with triple-negativebreast cancer and analyzes the relationship between pathological factors and survival timewith Cox regression model. Experimental results match with the prognostic factorssummarized by College of American Pathologists, which provide a reference for treatmentand prognosis of breast cancers, especially triple-negative breast cancer.Valuable biomarkers will be selected by analyzing protein mass spectrometric data.Adopting clustering analysis and other mature algorithm frameworks, this paper, firstly,compares protein mass spectrometric data between triple-negative and non-triple-negative breast cancers and selects the protein peak with significant difference; then, compares thedeath or recurrence samples with non-recurrence samples and selectively analyzes the proteinmarkers that affect prognosis of triple-negative breast cancers. With markers classificationtests selected jointly, the results can be used to accurately predict new clinical indexes ofbreast cancer prognosis and be helpful for development of individualized treatment programs.In the end, the paper analyzes the expression of TNM staging in protein massspectrometry. Samples will be grouped according to TNM staging and differences betweenthe groups will be analyzed qualitatively through classification results. Experiments show thatthe differences of tumor size and lymph node status can be reflected on protein massspectrometry. Regarding this, we can find out these particular markers by conductinglarge-scale sample analysis, which will be of great importance to clinical treatment of breastcancer.