Identification of molecular subtypes and prognostic features of breast cancer based on TGF-β signaling-related genes

Abstract

Abstract Purpose:The TGF-β signaling system is well understood to be involved in numerous cancer progressions, including cell invasion, epithelial-mesenchymal transition, and immunosuppression. Immune checkpoint inhibitors (ICIs) and TGF-β targeting drugs offer great promise as cancer therapies. However, the role of TGF-β in prognostic categorization and breast cancer (BC) treatment is unknown. Methods: First, we used data from The Cancer Genome Atlas database to create a predictive model that included five TGF-signaling-related genes (TSRGs). GSE161529 dataset from the Gene Expression Omnibus collection was collected to perform single-cell analysis to further describe these TSRGs. Furthermore, based on five TSRGs, an unsupervised algorithm was used to stratify two groups of immunity and overall survival (OS) in BC patients. More research was conducted on the differences in pharmaceutical therapy and tumor microenvironment among different patient groups and clusters. Results: The predictive model for BC contains five TSRGs: FUT8, IFNG, ID3, KLF10, and PARD6A, with single-cell analysis revealing that IFNG is primarily expressed in CD8+ T cells. Consensus clustering separated BC patients into two clusters, with cluster B having a longer OS and a better prognosis. Immunoassays revealed larger amounts of cluster B immune checkpoints and immune cells, implying that they would respond better to ICIs. Conclusion: Our findings supports accurate prognostic classification and efficient individual care strategies for BC patients by highlighting the possible significance of TGF-β signaling pathway for BC prognosis.