Abstract
BACKGROUND: Sulfur metabolism plays a crucial role in the initiation and progression of cancer. Our objective is to elucidate the molecular diversity inherent in breast cancer and to develop a predictive index grounded in sulfur metabolism-related genes (SMRGs). METHODS: We acquired transcriptomic and clinical datasets pertaining to breast cancer from publicly accessible repositories. Consensus clustering analysis was used to classify SMRGs in breast cancer into molecular subtypes. To pinpoint characteristic genetic markers and assemble a risk stratification model, we executed differential expression profiling, along with univariate and Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression analyses. Supplementary assessments encompassed gene set enrichment studies, scrutiny of somatic mutations, examination of the immune milieu, and exploration of chemotherapeutic responsiveness. Additionally, a nomogram model was constructed. RESULTS: Through the categorization of SMRGs in breast neoplasms, two distinctive molecular phenotypes emerged, each harboring substantial disparities in clinical outcome. Among the nine SMRGs linked to prognosis, eight key players-notably TSTD1, SULT1A2, MPST, GOT2, ENOPH1, CSAD, CHST12, and CDO1-served as the foundation for developing a predictive risk profile dedicated to breast carcinomas. This novel signature adeptly partitioned subjects into high- and low-risk strata, wherein the latter exhibited markedly superior overall survival relative to the former. Comparative analysis unveiled that the low-risk cohort featured diminished tumoral mutational load, elevated stromal and immune indices, and reduced tumoral purity. Notably, this subgroup showcased attenuated infiltration by M2 and M0 macrophages, contrasted with augmented infiltration by B and T lymphocytes. Furthermore, heightened responsiveness to paclitaxel and docetaxel was observed in the high-risk population compared to the low-risk counterpart. Conclusively, amalgamation of the risk metric, chemotherapeutic record, disease stage, and nodal status culminated in the construction of a nomogram aimed at breast cancer management. The dysregulation of SMRGs in the signature were further validated in external cohort. CONCLUSIONS: SMRGs contribute to the heterogeneity of breast cancer, and a molecular classification and prognostic evaluation can be achieved based on SMRGs.