Construction of a Breast Cancer Predictive Nomogram Based on Diverse Cell Death Methods and Reveal Tumor Microenvironment Characterization

ABSTRACT This study aimed to develop a robust predictive model and nomogram for breast cancer (BC) based on genes associated with diverse cell death methods. A prognostic model was constructed using the LASSO Cox method, incorporating twelve genes (CREB3L1, SFRP1, SHARPIN, AIFM1, IL‑18, CD24, EDA2R, CRIP1, XBP1, BCL2A1, NKX3‑1, and NME5). BC patients were classified into high‑risk and low‑risk subgroups, with the low‑risk subgroup showing superior survival, and this prognostic value was validated in an independent external cohort. A nomogram was also developed and confirmed as a reliable independent predictor of outcome. Enrichment analyses suggested a link between patient risk and immune response. The low‑risk subgroup exhibited a higher tumor microenvironment (TME) score. Patients in the high‑risk group showed improved responses to lapatinib, BI‑2536, OSI‑027, and SB505124, whereas those in the low‑risk subgroup had better sensitivity to axitinib, epirubicin, fulvestrant, and olaparib. Additionally, CD24 overexpression in BC cell lines promoted proliferation and migration, and inhibited apoptosis. These findings contribute to personalized treatment strategies and help elucidate the tumor microenvironment characteristics of BC patients.