A prognostic risk model based on programmed cell death genes for breast cancer and its potential clinical application

This study aims to identify programmed cell death (PCD)-associated genes linked to breast cancer prognosis for the construction of a prognostic model. Transcriptomic and clinical information was imported from the The Cancer Genome Atlas (TCGA) and GEO databases. Modulatory genes related to 18 types of PCD were evaluated. Furthermore, the TCGA and GEO datasets were employed as the training and validation datasets, respectively. A risk score prognostic model based on PCD-related genes was generated via univariate, LASSO, and multivariate Cox regression analyses. Further, differences in drug sensitivity, tumor mutation burden (TMB), immune-related pathways and cell infiltration, and immune checkpoints were compared between high-risk and low-risk cohorts to elucidate the clinical applicability of the model. Lastly, the model gene's expression was verified by RT-PCR. The data revealed 1480 PCD-related genes from the TCGA breast cancer gene expression matrix. Differential expression analysis identified 186 differentially expressed PCD genes. Furthermore, a prognostic model was generated according to the risk scores using multivariate Cox regression. The model comprised 8 PCD-related genes (BRSK2, CD24, IFNG, LAMB3, PDX1, PMAIP1, SLC7A11, and TRIML2). Moreover, breast cancer cases were divided into high-risk and low-risk cohorts per the median risk score. The results indicated that low-risk patients had better prognoses, and the model showed good predictive performance in the GEO validation cohort. The area under the curve values were 0.819, 0.731, and 0.674 for the nomogram's 3, 5, and 8 years overall survival, respectively. Functional enrichment analysis revealed that the prognostic model was markedly linked with the modulation of the immune microenvironment and tumor progression in breast cancer. Immune infiltration assessment revealed that low-risk patients had increased activity in immune-related pathways and infiltration of immune cells. In addition, the low-risk cohort had lower TMB and elevated immune checkpoint-related levels. Correlation analysis between immune checkpoints and risk scores indicated that low-risk patients were more responsive to immunotherapy. Drug sensitivity analysis showed variations in the IC50 values between the risk cohorts, suggesting potential variations in drug efficacy across different risk cohorts. Gene expression was verified by RT-PCR.A prognostic risk model for breast cancer based on 8 PCD-related genes was constructed and its predictive value was validated. The established model may provide novel biomarkers and effective therapeutic targets for breast cancer diagnosis and treatment.