Identifying Novel Biomarkers In Predicting Response Of Breast Cancer To Neoadjuvant Chemotherapy By Microarray Gene Expression Profiling

Background:Neoadjuvant chemotherapy was widely used as a standard procedure in patients with locally advanced breast cancer. About 10-35% breast cancer patients were resistant to neoadjuvant chemotherapy.The research of neoadjuvant chemotherapy predictive markers can help selecting the patients who might benefit from neoadjuvant chemotherapy, and keep those patients who are resistant to chemotherapy from unnessary preoperative treatment. Many large clinical trails showed patients with hormone responsive diseases are resistant to chemotherapy, while sensitive to endocrine therapy. In clinical practice, however, chemotherapy is still an important adjuvant treatment strategy for hormonal receptor positive patients, accompanying with harmful side effects.Objective:The aim of this study was to explore the gene expression related with ER of locally advanced breast cancer and to identify novel biomarkers that can be used for characterization and prediction of response to neoadjuvant chemotherapy using microarray gene expression profiling.Methods:Pretherapeutic tumorous tissues were obtained from 110 patients with locally advanced breast cancer. All patients received preoperative chemotherapy by the regimen of paclitaxel and carboplatin. Standardized surgery was performed after an interval of approximately 4 cycles. The response to chemotherapy was evaluated according to RECIST standards.In the 55 training cases, gene expression profiling of breast cancer were obtained by Human Genome Gene Chip Plus U133 2.0 Array. By using Significance Analysis of Microarrays (SAM), a set of discriminating genes was identified between pCR versus residul disease.Hierachical Clustering Analysis was applied to predict the response of the breast cancer. Gene Ontology and DAVID 2008 Functional Annotation Bioinformatics Microarray Analysis tools were used to gene functional category clustering analysis. As ER related biomarkers, the GATA3,TFF1 and TFF3 were performed further analysis in the validation set by IHC.Results:In the training set, the yield of total RNA was insufficient to assay in 16 patients and eight chips failed the quality check (QC) process. Therefore, gene expression profiles from 31 samples were available for this study. The overall pCR rate in the 86 patients was 19.7%(n=17,6 cases of pCR in training set and 11 pCR in validation set).6 patients as pCR and 25 patients as non-pCR showed significant different expression levels for 231 genes (Fold Change>2.0) by Significance Analysis of Microarrays (SAM).The list of differential expreesion genes included humoral immune response, protein-DNA complex assembly, transmembrane receptor protein tyrosine kinase signaling pathway, PARP signaling pathway and response to hormone stimulus. Setting fold change>3 resulted in 20 differentially expressed genes, among these 20 genes, TFF1,ESR1,GATA3,TFF3 were found as ER-related genes. In the further study by Real-time PCR, TFF1 and GATA3 showed significant higher expression in patients with non-pCR (P<0.05), which is consistent with the results of microarray analysis.In the 55 independent validation cases, in univariate analysis including clinical variables and ER-related genes, ER, PR, GATA3,TFF1 and TFF3 were all significantly associated with pCR (Pearson x 2,P<0.001,P=0.013, P<0.001,P<0.001, P<0.001).Only ER (P=0.016) remained significant in a logistic regression model.In spearman analysis, these 5 ER-related genes were significantly correlated with each other (P<0.05).To further analyse the predictive value of ER combined with ER-related genes, we found that the pCR rates was as high as.80% (4/5)when these 5 factors were all negative. In contrast, these 5 factors are all positive were found in 7 of 9 PD/SD patients.Conclusions:The present study suggested the possibility that microarray gene expression profiling may be usefμl in predicting response of locally advanced breast cancer to neoadjuvant chemotherapy and helpful in individualized therapy. ER related genes(ER pathway) may provide more predictive value of response to neoadjuvant chemotherapy than ER status alone. To test ER related factors can help comprehensively understanding the activity of ER pathway.