Window of opportunity trial of preoperative low dose azacitidine in high-risk early-stage breast cancer

Introduction: Hypomethylating agents such as 5-azacitadine (5-aza) can derepress human endogenous retroviral (HERV) genes and potentially generate immunogenic neoepitopes while reversing epigenetic T-cell exhaustion. Although extensively studied in hematologic malignancies, the role of 5-aza as an immune stimulant in solid tumors, particularly immunotherapy-refractory luminal and high-risk breast cancers, remains underexplored. We tested this hypothesis in a window of opportunity single-institution clinical trial in patients presenting for initial multidisciplinary curative management of high-risk tumors. Objectives: The primary objective is to determine the effect of low dose 5-aza therapy on tumor infiltrating lymphocytes (TILs) in patients with high-risk early-stage breast cancer. Correlative objectives include the effects of azacitidine therapy in high-risk subjects on immune response markers. Methods: UICC BRE-04 (NCT04891068) is a single-institution preoperative “window” trial testing 5 consecutive days of low-dose subcutaneous 5-aza 50 mg/m² in patients (Age 18-99; ECOG 0-2) with stage T1b-T3 triple-negative breast cancer or ER⁺ disease with at least one adverse feature (HER2⁺, node⁺, PR⁻, or high-risk genomic profile). Pre-treatment and post-window tissues preserved in FFPE were evaluated for the primary outcome of changes in TIL infiltration (quantitative and qualitative) and for HERV gene expression determined by quantitative-PCR. Cytokine concentrations were determined from time-matched plasma specimens. Results: 27 patients have completed study treatments. At this interim report, 18 paired specimens from 9 patients have been analyzed by multiplex immunofluorescence. Overall CD8⁺ and CD4⁺ T-cell densities and subset proportions (PD-1⁻, CD45RO⁻, FOXP3⁻) were not significantly altered (all p > 0.05), however, post-treatment tumors exhibited a 45.3% reduction in the average distance between immune cells and PanCK⁺ tumor epithelial cells ( p < 0.05). Directional increases in intratumoral and stromal CD8⁺ effector subsets and stromal CD4⁺ effector T cells were observed without expansion of regulatory T cells, consistent with heterogeneous but coordinated immune engagement. qPCR HERV gene expression analysis is ongoing. No serious adverse events were observed in any patient. Conclusion: We demonstrate treatment-associated immune remodeling following a short-course of treatment with 5-aza. Specifically, effector CD8+ and CD4+ populations migrate in closer proximity to target tumor cells even though no observed effects on overall immune cell numbers were identified. Responses were heterogenous and suggest other factors may influence the effectiveness of 5-aza in immune priming. Final results from the entire enrolled cohort will be available at the time of meeting presentation.