Immune-evasion-associated prognostic gene expression score stratifies survival and immune microenvironment features...

Breast cancer is characterized by marked prognostic and immune heterogeneity, and tumor immune evasion may contribute to variable responses to immunotherapy. This study aimed to characterize immune-evasion-associated transcriptomic patterns in breast cancer and to develop an exploratory prognostic score linked to immune microenvironment features. A curated set of 182 immune-evasion-associated genes (IEGs) was analyzed in The Cancer Genome Atlas Breast Invasive Carcinoma (TCGA-BRCA) cohort. Fifteen differentially expressed IEGs distinguished two molecular clusters with different overall survival (OS), pathway-enrichment profiles, and inferred immune-cell characteristics. To construct an immune-evasion-associated prognostic score (IEPS), all 182 IEGs were screened by univariate Cox regression, and 44 OS-associated candidates were further analyzed using least absolute shrinkage and selection operator (LASSO) Cox regression, yielding a 12-gene score. Using the TCGA-derived median cutoff, high IEPS was associated with poorer OS in TCGA-BRCA and remained associated with OS in the external Gene Expression Omnibus (GEO) GSE20685 cohort; however, discrimination was modest, with external 3-, 5-, and 10-year area under the curve (AUC) values of 0.560, 0.572, and 0.581, respectively. Adding IEPS to available clinicopathological variables produced only modest performance improvement. Exploratory prediction analysis of microarray 50-gene (PAM50) subtype analysis showed that IEPS differed across molecular subtypes, indicating that the score may partly capture subtype-related transcriptomic heterogeneity. Low-IEPS tumors showed more immune-active pathway and immune-cell signatures, whereas high-IEPS tumors were associated with relatively less active antitumor immune features. Protein disulfide isomerase family A member 3 (PDIA3), one IEPS gene, was selected for preliminary functional validation; small interfering RNA (siRNA)-mediated PDIA3 knockdown reduced proliferation, clonogenic growth, and migration of breast cancer cells in vitro. Overall, IEPS may represent an exploratory transcriptome-based indicator associated with prognosis and immune-related features in breast cancer, but it should not be interpreted as a clinically validated standalone or molecular subtype-independent prognostic tool. Further independent validation and immune-relevant mechanistic studies are required, particularly to clarify the role of PDIA3 in breast cancer immune evasion.