Protein profiling predicts the response to anthracycline and taxanes based neo-adjuvant chemotherapy in breast cancer

Abstract

Neo-adjuvant chemotherapy for breast cancer substantially benefits patients who achieve pathological response. However, clinical or pathological response information can only be obtained a period of time after chemotherapy. The identification of novel bio-markers or the application of new technique that can be used to predict treatment response before chemotherapy would allow therapy to be tailored on an individual patient basis. The purpose of this study is to identify the chemo-sensitivity and chemo-resistance related proteins using antibody microarray profiling, and to develop a multi-protein predictive model for breast cancer. Total protein was extracted from core needle biopsy samples obtained from 15 patients before treatment with neo-adjuvant TA (combination of taxanes and anthracycline) chemotherapy. Protein profiling was analyzed by antibody microarray. 10 patients were used as training set to develop the predictive model using the software PAM(prediction analysis of microarray). Another 5 patients were used as a validation set to test the model. In cross-validation, the molecular predictive model showed an accuracy of 90%, in independent validation, the model classified the cases with an accuracy of 80%. In conclusion, the proteomic predictive model has the potential to predict pathological response to neo-adjuvant TA chemotherapy.