Deep spatial analysis of the tumor-immune interface in localized prostate cancer.

377 Background: Prostate cancer (PC) has been historically considered immunologically “cold” due to low immune infiltration and a predominantly immunosuppressive tumor microenvironment (TME). However, this dogma is largely based on immune profiling studies of metastatic castration-resistant prostate cancer (mCRPC) and may not apply to earlier stages of disease. Furthermore, the composition and organization of the immune microenvironment in primary PC remains poorly defined, especially in relation to Gleason score, the gold standard of pathologic PC grading. A deeper understanding of the localized PC immune TME is necessary to elucidate the mechanisms of immunosuppression and identify targets for immunotherapy, especially in high Gleason (HG) PC. We hypothesized that the localized PC immune TME is heterogeneous with distinct immunosuppressive features and spatial organization in HG versus low Gleason (LG) PC. Methods: We analyzed 29 radical prostatectomy samples (15 LG and 14 HG PC) using a 40-marker immunofluorescence panel to profile immune and tumor cells at single-cell resolution. Immune subsets were quantified and spatially mapped to evaluate immune infiltration, regulatory populations (tumor-promoting T-regulatory cells (Tregs) and M2-like macrophages), and interactions between cell types. Results: HG PC exhibited significantly increased total immune cell infiltration and tertiary lymphoid structures (TLSs) relative to LG PC. Advancing Gleason grade was associated with increasing T cell infiltration and M2-like macrophage density. There were more Tregs in the stroma and TLSs of HG PC compared to LG PC. M2-like density was increased in the stroma in HG PC relative to LG PC. TLSs harbored higher proportions of exhausted CD4 and CD8 T cells and these exhausted T cells were found closer to Tregs in TLSs compared to their surrounding tumor and stroma. TLSs were also found to have higher ratios of M2-like to M1-like macrophages compared to surrounding tumor and stroma. Conclusions: Our findings demonstrate significantly more immune cell infiltration in HG vs LG PC, challenging the notion that PC is immunologically “cold” across all stages of disease. HG PC exhibits elevated levels of Tregs and M2-like macrophages in specific TME compartments, highlighting the distinct spatial organization of immunosuppressive populations. We observed more TLSs in HG PC, which contain exhausted T cells in proximity to Tregs and high M2/M1 macrophage ratios, suggesting that TLSs may paradoxically serve as local hubs of immunosuppression in high-grade disease. Overall, these data provide a mechanistic rationale for immunotherapeutic targeting of the TLS-rich immunosuppressive landscape of HG PC to improve clinical outcomes in the neoadjuvant setting.