High TRIM28 Expression Defines an Aggressive, Immune-Cold Phenotype with Worse Survival Outcomes in ERα-Positive Breast Cancer

Background: Although ERα-positive breast cancer represents approximately 70% of all breast cancer diagnoses worldwide, specific prognostic biomarkers for this subtype that are capable of stratifying patients remain limited. TRIM28 (KAP1/TIF1β), which is a multifunctional E3 ubiquitin ligase and transcriptional coregulator of ERα, has been shown to play oncogenic roles in multiple malignancies. However, its prognostic significance in ERα-positive breast cancer subtype has not been explored across independent patient cohorts. Methods: Multiple bioinformatics tools were employed to assess TRIM28 mRNA expression and prognostic significance across independent patient cohorts totaling over 4000 patients. The Kaplan–Meier plotter was used to examine associations between TRIM28 expression and survival outcomes in ERα-positive breast cancer. Multivariate Cox proportional hazards regression was performed in the METABRIC dataset (n = 1356) to confirm independent prognostic value, with sensitivity analyses using alternative TRIM28 expression cutoffs and PAM50 subtype-specific subgroup analyses. Independent validation was performed using multivariate Cox regression in the TCGA-BRCA Firehose Legacy ERα-positive cohort (n = 372). Gene Set Cancer Analysis (GSCA) was used to investigate pathways associated with TRIM28-correlated genes, and Breast Cancer Gene-Expression Miner (Bc-GenExminer) was used to assess immune cell infiltration. Results: TRIM28 expression is significantly elevated in ERα-positive breast cancer compared to normal breast tissue. High TRIM28 expression (defined as the upper quartile, ≥75th percentile, of TRIM28 expression) is independently associated with worse overall survival after adjustment for tumour size, tumour mutational burden (TMB), hormone therapy status, and age, with stratification on histologic grade (HR = 1.21, 95% CI: 1.03–1.41, p = 0.0194; METABRIC, n = 1356). This finding was independently validated using multivariate Cox proportional hazards regression in the TCGA-BRCA Firehose Legacy ERα-positive cohort (n = 372; events = 56), in which high TRIM28 expression remained independently associated with worse OS after adjustment for age, T-stage, and TMB (HR = 2.02, 95% CI: 1.10–3.71, p = 0.024). PAM50 subtype-specific analyses within METABRIC additionally confirmed an independent association of high TRIM28 with worse OS in Luminal A (HR = 1.37, 95% CI: 1.08–1.74, p = 0.009), with a directionally consistent but non-significant trend in Luminal B (HR = 1.21, 95% CI: 0.93–1.58, p = 0.155). The prognostic effect was robust across alternative TRIM28 expression cutoffs. TRIM28 expression positively correlates with tumour size, histologic grade, Nottingham Prognostic Index, and TMB, and negatively correlates with immune cell infiltration and is significantly lower in patients receiving hormone therapy. Genes co-expressed with TRIM28 are enriched in cell cycle and DNA damage response activation signatures and RAS/MAPK and RTK pathway inhibition signatures. Conclusions: TRIM28 is an independent prognostic biomarker in ERα-positive breast cancer, validated across multiple independent cohorts. These findings nominate TRIM28 as a priority candidate for prospective clinical validation and targeted experimental investigation in ERα-positive breast cancer.