Novel immune-related prognostic model and nomogram for breast cancer based on ssGSEA

This study aimed to construct an immune-related prognostic model and a nomogram to predict the 1-, 3-, and 5-year overall survival (OS) of breast cancer patients. We applied single-sample gene set enrichment analysis to classify 1,053 breast cancer samples from The Cancer Genome Atlas (TCGA) database into high and low immune cell infiltration clusters. In cluster construction and validation, the R packages “GSVA,” “hclust,” “ESTIMATE,” and “CIBERSORT” and GSEA software were utilized. ImmPort, univariate Cox regression analysis, and Venn analysis were then used to identify 42 prognostic immune-related genes. Eventually, the genes TAPBPL, RAC2, IL27RA, ULBP2, PSMB8, SOCS3, NFKBIE, IGLV6-57, CXCL1, IGHD, AIMP1, and CXCL13 were chosen for model construction utilizing least absolute shrinkage and selection operator regression analysis. The Kaplan–Meier curves of both the training and validation sets indicated that the overall survival of patients in the low-risk group was superior to that of patients in the high-risk group (p < .05). The areas under curves (AUCs) of the model at 1, 3, and 5 years were, respectively, .697, .710, and .675 for the training set and .930, .688, and .712 for the validation set. Regarding clinicopathologic characteristics, breast cancer-related genes, and tumor mutational burden, effective differentiation was achieved between high-risk and low-risk groups. A nomogram integrating the risk model and clinicopathologic factors was constructed using the “rms” R software package. The nomogram’s 1-, 3-, and 5-year AUCs were .828, .783, and .751, respectively. Overall, our study developed an immune-related model and a nomogram that could reliably predict OS for breast cancer patients, and offered insights into tumor immune and pathological mechanisms.