Integrative landscape analysis of prognostic model biomarkers and immunogenomics of disulfidptosis-related genes in breast cancer based on LASSO and WGCNA analyses

BACKGROUND: Disulfidptosis is a novel type of programmed cell death. However, the value of disulfidptosis-related genes (DRGs) in the prediction of breast cancer prognosis is unclear. METHODS: RNA-seq data of 1231 patients, together with information on patient clinical characteristics and prognosis, were downloaded from TCGA. DRGs were identified between cancerous and non-cancerous tissues. The LASSO algorithm was used to assign half of the samples to the training set. Risk scores were used for construction of a prognostic model for risk stratification and prognosis prediction, and the clinical applicability was examined using a line diagram. The relationships between risk scores, immune cell infiltration, molecular subtypes, and responses to immunotherapy and chemotherapy were examined. RESULTS: We identified and obtained four DRG-related prognostic lncRNAs (AC009097.2, AC133552.5, YTHDF3-AS1, and AC084824.5), which were used for establishing the risk model. Longer survival was associated with low risk. The DRG-associated lncRNAs were found to independently predict patient prognosis. The AUCs under the ROCs for one-, three-, and 5-year survival in the training cohort were 0.720, 0.687, and 0.692, respectively. The model showed that the high-risk patients had reduced overall survival as well as high tumor mutation burdens. Furthermore, high-risk patients showed increased sensitivity to therapeutic drugs, including docetaxel, paclitaxel, and oxaliplatin. CONCLUSION: The risk score model was effective for predicting both prognosis and sensitivity to therapeutic drugs, suggesting its possible usefulness for the management of patients with breast cancer. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00432-023-05372-z.