Subtype-specific genetic drivers of immune evasion in breast cancer

Immune evasion is a hallmark of cancer and a driver of therapeutic resistance. Although immunotherapy is effective in highly immunogenic cancers, its efficacy in breast cancer (BC) remains limited. We aimed to determine the prognostic relevance of immune-related gene signatures across distinct BC subtypes.

We used transcriptomic and clinical data from three independent cohorts [Gene Expression Omnibus (GEO), The Cancer Genome Atlas, and GSE96058]. We analyzed 106 genes associated with the evasion of immune destruction (EID) and 182 genes involved in the evasion of killing by cytotoxic T lymphocytes (ECTL). Expression of signatures was stratified by BC subtypes. Cox regression and Kaplan-Meier curves were used to assess survival, with false discovery rate (FDR) correction ensuring statistical robustness.

High expression of the ECTL signature was significantly associated with improved overall survival (OS) in basal BC patients [hazard ratio (HR) 0.25, 95% confidence interval (CI) 0.16-0.4, P = 8.6e-10, FDR <1%], and similar trends appeared in human epidermal growth factor receptor 2 (HER2)-positive BC. For EID genes, high expression also correlated with favorable OS in basal BC (HR 0.3, 95% CI 0.18-0.49, P = 3.4e-7). Gene-level analyses revealed CXCL10, CXCL9, CXCR4, and JAK3 as robust predictors of OS in basal BC, validated across independent datasets. This four-gene signature demonstrated strong predictive power for response to immune checkpoint inhibitors (ICIs) (area under the curve = 0.722, P = 1.3E-07). In HER2-positive BC, a three-gene signature (IL2RG, CD3E, CD3G) was consistently prognostic (HR 0.41, 95% CI 0.24-0.71, P = 0.0011, GEO) across independent datasets.

We identified subtype-specific signatures that predict survival and immunotherapy response, providing clinically actionable biomarkers.