Abstract
Over the past several decades, there has been a significant increase in the number of breast cancer patients. Among the four subtypes of breast cancer, Her2-positive breast cancer is one of the most aggressive breast cancers. In this study, we screened the differentially expressed genes from The Cancer Genome Atlas-Breast cancer database and analyzed the relationship between immune cell infiltration and differentially expressed genes using weighted gene co-expression network analysis. By constructing a module-trait relationships heatmap, the red module, which had the highest correlation value with M1 macrophages, was selected. Twenty hub genes were selected based on a protein-protein interaction network. Then, four overlapping M1 macrophage infiltration-related genes (M1 MIRGs), namely CCDC69, PPP1R16B, IL21R, and FOXP3, were obtained using five machine-learning algorithms. Subsequently, nomogram models were constructed to predict the incidence of Her2-positive breast cancer patients. The outer datasets and receiver operating characteristic curve analysis were used to validate the accuracy of the four M1 MIRGs and nomogram models. The average value of the area under the curve for the nomogram models was higher than 0.75 in both the training and testing sets. After that, survival analysis showed that higher expression of CCDC69, PPP1R16B, and IL21R were associated with overall survival of Her2-positive breast cancer patients. The expression of CCDC69 and PPP1R16B could lead to more benefits than the expression of IL21R and FOXP3 for immunotherapy. Lastly, we conducted immunohistochemistry staining to validate the aforementioned results. In conclusion, we found four M1 MIRGs that may be helpful for the diagnosis, prognosis, and immunotherapy of Her2-positive breast cancer.