Spatial protein profiling of irradiated ER positive breast cancer lymph nodes suggests nodal irradiation modulates immune function

Background: We hypothesized that the systemic benefit of regional nodal irradiation (RNI) for breast cancer derives from immune stimulation in involved nodes. We compared involved lymph nodes (ILN) from patients treated on a prospective trial of pre-operative and post-operative conventional (2 Gy/fx) versus hypofractionated (2.67 Gy/fx) RNI regimens (SAPHIRE NCT02912312). Methods: 17 patients ILNs with ER+, pN+ BC were Digitally Spatially Profiled with immune-related proteins. 6 post-op cases had no neoadjuvant chemotherapy (NACT) (untreated controls). All others had NACT +/- RNI. 4 pre-operative RNI short-course (RNI_NACT-2.67Gy/fx); 5 pre-op standard RNI (RNI_NACT-2 Gy/fx); 3 post-op RNI after NACT. 5 Regions of Interest (ROIs) were selected and categorized as “Tumor” (PanCK+) or “Immune” (CD45+) compartments. Cell density and spatial analysis were assessed in selected cases. Pearson's correlations, T-Test and KS tests were performed on protein expression in all ROIs, FDR < 0.1, P value < 0.05. Signals with Log2 differences between -1 and 1 were analyzed using adjusted LMM for patient-level variation, and PCA of ROIs by compartment was performed. Results: Nearest neighbor analysis shows increasing median distance between tumor and immune cells from untreated, to NACT, and NACT+RNI, and no differences between short and standard pre-operative RNI. Proximity analysis reveals the distribution of CD3+ cells from the CK+ cells changes with increasing therapy, favoring frequency of closer immune cells, although the average number of cells around each CK+ cell diminishes with increasing therapies. CK+ cell density remains stable, suggesting migration rather than cell death. Differences in expression profiles of immune ROIs suggest more immune signaling in pre-op short. Considering all RT, 27 differentially expressed proteins (DEP) were significantly upregulated in Post tumor ROIs. None were significantly downregulated. After correction by LMM, 11 markers were up in Post cases and none downregulated. We found lower expression of SMA , CD4, CD45RO, CD14, CD11c, TIM-3, and B7-H3 in the Pre nodes. No significant DEPs remain in immune ROIs after LMM correction. The first two principal components accounted for 45.8% of the total variance, with PC1 largely separating samples based on Pre RNI. Pre-short vs Pre-standard showed 9 increased and 0 decreased proteins in Pre-short including CD27, Sting, CD40, CD11c, CD25. Conclusions: Pre-operative radiation contributes to Immune microenvironment remodeling, complementing neoadjuvant chemotherapy in ILN, without altering immune cell density. Different immune signals in irradiated ILN suggest a less immunosuppressive phenotype in RNI, and potential immune activation in Pre-short RNI. Small sample size limits conclusions, immune studies will follow to optimize radiation regimens.