Tamoxifen is an estrogen receptor (ER) antagonist that is most commonly used for the treatment of ER-positive breast cancer. However, tamoxifen resistance remains a major cause of cancer recurrence and progression. Here, we aimed to identify hub genes implicated in the progression and prognosis of ER-positive breast cancer following tamoxifen treatment. Microarray data (GSE9893) for 155 tamoxifen-treated primary ER-positive breast cancer samples were obtained from the Gene Expression Omnibus database. In total, 1,706 differentially expressed genes, including 859 upregulated genes and 847 downregulated genes, were identified between relapse samples and relapse-free samples. Weighted correlation network analysis clustered genes into 13 modules, among which the tan and blue modules were the most significantly related to prognosis. From these two modules, we further identified and validated two prognosis-related hub genes (GRSF1 and MAPT) via survival analysis based on several publicly available datasets. High expression of GRSF1 predicted poor prognosis, whereas MAPT indicated favorable outcomes in ER-positive breast cancer. Using breast cancer cell lines and tissue samples, we confirmed that GRSF1 was significantly upregulated and MAPT was downregulated in the tamoxifen-resistant group compared with the tamoxifen-sensitive group. The prognostic value of GRSF1 and MAPT was also verified in 48 tamoxifen-treated ER-positive breast cancer patients in our hospital. Gene set enrichment analysis suggested that GRSF1 was potentially involved in RNA degradation and cell cycle pathways, while MAPT was strongly linked to immune-related signaling pathways. Taken together, our findings established novel prognostic biomarkers to predict tamoxifen sensitivity, which may facilitate individualized management of breast cancer.

Background: Turning the “cold” tumor immune microenvironment into “hot” is a critical issue in cancer treatment today. Hormone receptor-rich breast cancer (HR+ BC) was previously considered immunologically quiescent. Objective: This study aims to explore the immunomodulatory effects of endocrine therapy on HR+ BCs. Methods: The infiltrations and alterations of the tumor immune microenvironment in HR+ BCs before, after 10-14 days, and after three months of neoadjuvant endocrine therapy were computationally analyzed according to MCP-counter, CIBERSORT, xCell algorithms, and gene-set enrichment analysis (GSEA). The primary microarray data were obtained from three HR+ BC gene expression datasets (GSE20181, GSE55374, and GSE59515). Single-sample GSEA of hallmark and immune response gene sets was performed to evaluate the correlation between suspected treatment response and activated immune pathways in tumors. Results: Both immune and stromal cells were specifically recruited into the HR+ BCs who responded to the neoadjuvant endocrine therapy by letrozole. Besides the enhanced infiltrations of immunosurveillance-related cells such as CD8+ T cells, dendritic cells, and the activation of immune response-related signals, the immunosuppressive M2-like macrophages, as well as the expression of immune checkpoint genes like PDCD1, SIRPA, and some HLA genes, were also stimulated in responders. We identified four pretreatment indicators (the intrinsic luminal subtype, the estrogen response early/late pathway, and the epithelial-mesenchymal transition pathway) as potential predictors of both clinical response and the activation of the tumor immune microenvironment post letrozole. Conclusions: Neoadjuvant endocrine therapy showed a promising way to convert the immunologically “cold” HR+ BCs into “hot” tumors. This study provides new insights into the application of immunotherapy for HR+ BCs especially those who respond to endocrine therapy.