Spatial transcriptomic mapping of ER-positive breast tumors reveals immune pathway remodeling and resistance mechanisms following anti-PD-1 therapy

Estrogen-receptor-positive (ER+) breast cancers (BCs) make up approximately 70%-80% of BC cases and demonstrate low lymphocyte infiltration, mutational burden, and modest objective response rates to immune checkpoint blockade (ICB), making them attractive candidates for novel immunostimulatory combination therapies. Here, we employ CosMx Single Cell Imaging (SMI) technology to spatially profile the transcriptional landscape of ~6000 targets in 9 ER+ BC patients, both pre- and post-treatment with the anti-PD-1 antibody Cemiplimab, to identify spatially resolved mechanisms of resistance and potential therapeutic targets. After quality control, ~3.1 million cells across 18 unique tissue sections were analyzed, allowing for robust characterization of immune, stromal, and epithelial cell niches with the ability to identify rare immune cell subsets at a high resolution. Comparing relative abundances of cell types of patients pre- and post-therapy, we see a high concordance of cell-types between tissue sections of the same patient. While 8/9 patients were dominated by Luminal A/B cancer epithelial cells, one patient had purely basal cancer epithelial cells, highlighting the value in spatial transcriptomic platforms for molecular subtyping in BC patients. To define metrics of patient response to Cemiplimab therapy, we quantified changes in tumor-intrinsic inflammatory/immune pathways and leveraged spatial information to calculate changes in tumor-infiltrating lymphocytes (TILs). 3/9 patients showed increased levels of type I/II IFN signaling in the tumor epithelium, 5/9 showed no change, and 1/9 showed a sharp decrease, coinciding with a unique increase in TNF-alpha signaling via NF-κB. 6/9 patients showed increases in TILs following therapy, defined as the fraction of tumor cells ≤ 0.1 mm from the nearest CD8 T-cell. The patient with the most pronounced decrease in TILs was the same patient who showed reduced IFN signaling previously, indicating agreement between response indicators, and suggesting upregulation of TNFa signaling via NFkB as a resistance mechanism to T-cell invasion. Our early work provides a comprehensive spatially resolved analysis of the mechanisms through which ER+ BCs resist ICB, suggesting that spatially mapped transcriptional programs can uncover actionable resistance pathways, and guide the design of next-generation immunotherapeutic combinations for ER+ BC.