Abstract P4-05-16: Immune landscape of glucocorticoid receptor high breast cancer

Abstract

Abstract BACKGROUND: Glucocorticoid receptor (GR) activation is pro-apoptotic in lymphoid malignancies, whereas in estrogen-receptor negative (ER-) breast cancer, GR activation is anti-apoptotic and associated with relapse through a direct transcriptional role in tumor cells. A meta-analysis of GEO dataset has shown that higher GR expression is associated with an improved relapse free survival (RFS) in ER+ breast cancer but a worse RFS in triple negative breast cancer (TNBC). The relative contributions of GR expression on tumor vs. immune cells which may influence outcomes have not been studied. We hypothesize that there is a difference in the immune landscape between GR-high and GR-low tumors. METHODS: mRNA expression data was accessed from The Cancer Genome Atlas (TCGA) and METABRIC. K-clustering classified GR gene (NR3C1) expression into “high” and “low”. Computer algorithms TIMER, CIBERSORT, Immune pathway gene score, and cytolytic activity score (CYT) were used. Single-cell sequencing data were obtained from GSE75688 and GSE114725 cohorts. RESULTS: 1904 and 1087 pts with stage I-III breast cancer had clinical and genomic data available in METABRIC and TCGA respectively. GR-high tumors had better median overall survival (mOS) [15 vs. 12yrs {HR 0.8(0.71-0.9), p<0.001}] in METABRIC, however, no significant difference observed in TCGA. GR-high tumors also had better median disease-specific survival (mDSS) [25 vs. 23.5yrs {HR 0.79 (0.67-0.93), p=0.005}] in METABRIC and [NR vs. NR {HR 0.4 (0.23-0.81), p=0.004}] in TCGA. Hormone receptor (HR) + subtype had better mOS [15.2 vs. 12.8yrs {HR 0.98(0.75-1), p=0.021}] but no significant difference in mDSS in GR-high vs. -low tumors in METABRIC; however, TCGA showed better mDSS [NR vs. NR {HR 0.4(0.22-0.92), p=0.049}]. TNBC subtype was associated with better mOS [17 vs. 7.4yrs {HR 0.66(0.48-0.9)}, p=0.0094] and mDSS [NR vs. 20yrs {HR 0.67(0.46-0.98)}, p=0.038] in GR-high vs. -low tumors in METABRIC; however, no significant differences were observed in TCGA. GR-high tumors had significantly higher number of immunomodulatory cells [B-cells, dendritic cells and mast cells] (p<0.001) and lower number of immunosuppressive cells [T regulatory cells (Treg) and T follicular helper cells] (p<0.001). There was a moderate correlation between GR-high and high CD8+T-cells (r=0.49, p=0.001) and macrophages (r=0.35, p=0.02). GR-high tumors had increased macrophages, lymphocyte infiltration and TGF-β response (p<0.001). Significantly higher expression of T-cell exhaustion genes PD-1, PD-L1, CTLA-4, IDO1 and TIM3 was observed in GR-high tumors (p<0.001). Strikingly, single cell sequencing data showed higher GR expression on immune cells (T-cells, B-cells and myeloid cells) compared to stromal or tumor cells (p<0.001), with significantly higher GR expression on CD8+T-cells compared to other immune cell subsets including Treg, CD4+T-cells, neutrophils, monocytes, dendritic cells, mast cells and macrophages (p<0.001). These findings explain why CYT was consistently elevated in GR-high tumors (p=0.001) observed both in TCGA and METABRIC. CONCLUSIONS: This study demonstrates that there are significant differences in the immune cell landscape in GR-high vs. GR-low tumors, suggesting that a higher immune cell infiltration accounts for greater anti-tumor cytolytic activity in GR high tumors. Thus, in addition to a direct transcriptional role for GR on tumor cells, contribution to outcomes by immune cells should also be acknowledged. Further studies examining the important contribution of factors causing activation/inactivation of immune cells in tumor microenvironment and their effects on clinical outcomes are warranted. Citation Format: Shipra Gandhi, Ahmed Elkhanany, Mateusz Opyrchal, Elizabeth A Repasky, Kazuaki Takabe. Immune landscape of glucocorticoid receptor high breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-05-16.