Kyeong Joo Jung

Accurate prognostic assessment of prostate cancer (PCa) requires an integrated understanding of tissue morphology-encompassing cell structure, glandular architecture, and tissue organization-and the immune environment. We present Prostate-TriMod , a novel tri-modal histology dataset designed to integrate high-resolution visual morphology with spatial tissue maps, immune infiltration patterns, and clinical outcomes. This dataset, generated from the Cell DIVE™ multiplexed imaging platform, consists of three synchronized modalities: (1) multiscale virtual H&E tiles (224px, 256px, 512px, and 2040px) providing visual morphological context, (2) spatial tissue maps identifying cancerous/non-cancerous epithelial cells, stroma and immune cell populations (via TOPAZ and CAT models), and (3) text captions generated from single-cell data and patterns. The dataset includes comprehensive clinical annotations, including Grade Groups and biochemical recurrence (BCR) status. By providing high-fidelity alignment between visual features, spatial tissue maps, and textual descriptions, Prostate-TriMod empowers the development of advanced multimodal AI frameworks. We expect this resource to support reuse in multimodal representation learning, spatial analysis, and benchmarking studies that link histology morphology and immune context to clinical outcomes in prostate cancer.

Prostate cancer is classified as “immune-cold” due to limited infiltration of immune cells and no clear correlation between immune cells and clinical outcomes. However, immune cells are found in prostate cancers and the spatial relationships between these immune cells and cancer cells/glands have not been investigated, partly due to a lack of automated tools that classify both cancerous cells/glands. In this paper, we have developed an end-to-end tool (TOPAZ: T issue O rganization identification using s PA tial proteomics) that combines multiplexed single-cell protein data with histology images to: 1) predict cancerous versus non-cancerous epithelial cells using a Gaussian-mixture model; 2) predict cancerous/non-cancerous gland type using a principal curve estimation. Using TOPAZ to assign cancer and non-cancerous labels to cells and glands, we extracted multiscale spatial features from the classification results—including immune dense-region geometrical features and cell-to-gland distances— and correlated the features with risk of biochemical recurrence and cancer grade. Tissue-microarrays containing 753 cores from 217 prostate cancer patients underwent multiplexed immunofluorescent imaging (Cell DIVE, Leica) for epithelial cell markers (panCK26, S6, NaKATPase), basal cell markers (p63, CK5), a cancer cell marker (AMACR), and T cell markers (CD3, CD4, CD8, FOXP3, CD68). Cancerous/non-cancerous cell classification from TOPAZ achieved 82% sensitivity and 99% specificity against expert annotation, and the pipeline further predicted cancerous/non-cancerous glands without manual threshold tuning. Regulatory-T-cell and helper-T-cell percentages decreased, and macrophage percentage increased with grade increase (P < 0.05). When the median distance from cancerous gland centroids to the nearest regulatory or helper T-cell exceeded approximately 50 µm, the hazard of biochemical recurrence doubled (log-rank P < 0.01). The open-source Shiny app TOPAZ ( https://chunglab.bmi.osumc.edu/TOPAZ ) packages the workflow, predicting individual cell types and gland shapes. By combining probabilistic cell typing with gland-shape modeling, TOPAZ yields interpretable multiscale spatial features linked to prognosis and is released as an open web app for unrestricted use. Author Summary: Spatial distribution of cancerous cells/glands and immune cell distribution has not been considered as a prognosticator in prostate cancer. In addition, automated tools that can quantify and integrate these distributions are lacking. We combined high-dimensional single-cell protein measurements with histology images to map gland structure in prostate cancer tissue. Our web-based tool, TOPAZ ( https://chunglab.bmi.osumc.edu/TOPAZ ) predicts whether each epithelial cell and gland in virtual H&E image is cancerous or not. Once the predictions are made, a spatial analysis workflow helps quantify spatial features of immune cells relative to the glands and correlate with recurrence risk and grades. Across 753 tissue cores from 217 prostate cancer patients, helper and regulatory-T-cells located more than about 50 µm away from cancerous epithelial glands were associated with a higher risk of biochemical recurrence. The pipeline provides new insights for researchers and pathologists into prostate cancer progression and biochemical recurrence through integration of spatial location of cancer glands and immune cells.

Supplementary Table 1 shows the list of antibodies used in this study. Supplementary Table 2 shows the input data and variables used for computational modeling. Supplementary Table 3 shows parameter estimates and confidence intervals for modeled cell populations. Supplementary Table 4 shows patient demographics, tumor characteristics, and treatment regimens.

Supplementary Figure S1 shows PD-1 blockade therapy combined with NK cell depletion improves tumor control and increases CD8⁺ T cell infiltration. Supplementary Figure S2 shows the combination of NK cell depletion and ICB is most effective when NK cells are depleted at an early stage. Supplementary Figure S3 shows NKRT accumulation is tumor-site specific. Supplementary Figure S4 shows phenotypic changes in CD4⁺ T cells upon NK cell depletion. Supplementary Figure S5 shows NKRTs are antigen-specific and exhibit enhanced tumor-controlling efficacy when combined with ICB. Supplementary Figure S6 shows NKRT formation depends on the timing of NK cell depletion. Supplementary Figure S7 shows NK cell depletion delays CD8⁺ T cell exhaustion in day 15 MC38 tumors. Supplementary Figure S8 shows NK cell depletion delays CD8⁺ T cell exhaustion in day 20 MC38 tumors. Supplementary Figure S9 shows NK cell depletion improves tumor control in the ACT setting. Supplementary Figure S10 shows distinct gene expression profiles in CD8⁺ T cells following NK cell depletion. Supplementary Figure S11 shows volcano plots of differentially expressed genes under specific conditions. Supplementary Figure S12 shows elevated p-STAT3 and p-STAT5 levels in CD8⁺ T cells from NK cell-depleted tumors. Supplementary Figure S13 shows cytokines play a critical role in NKRT formation and anti-tumor activity. Supplementary Figure S14 shows computational model fits of tumor growth curves and CD8⁺ T cell dynamics. Supplementary Figure S15 shows phenotypic variation in tumor-infiltrating NK cells across tumor models. Supplementary Figure S16 shows expression patterns of tissue-resident markers in CD8⁺ T cells and NK cell proportions in responders versus non-responders to immunotherapy.

Immune checkpoint blockers (ICB) targeting the PD-1/PD-L1 axis represent established therapies for many cancers. However, resistance occurs in most patients due to complex immune-suppressive mechanisms in the tumor microenvironment. NK cells can play effector roles in tumor control, but their impact on T-cell dysfunction and ICB efficacy remains controversial. Through genetic and antibody-mediated NK cell depletion, we found that a subset of tumor-associated NK cells plays a negative role in ICB sensitivity; they further impede CD8+ T-cell differentiation toward a CD69+ BCL2+ EOMES+ GZMB+ TIM3− GITR− phenotype. Mechanistically, the retinoic acid receptor α–dependent differentiation program in CD8+ T cells is hindered by tumor-infiltrating NK cells via competition for IFNα and IL-2. Finally, we observed that lower frequencies of NK cells correlate with better clinical responses to ICBs in patients with cancer. These findings suggest potential avenues for enhancing CD8+ T cell–centered immunotherapy by targeting regulatory NK cells. Significance: Although NK cells are traditionally viewed as antitumor effectors, our study uncovers their unexpected suppressive role in CD8+ T cell–based immunotherapy. By competing for cytokines, they disrupt retinoic acid receptor α–driven CD8+ T-cell differentiation and limit ICB efficacy. Clinically, reduced NK cell presence is associated with an enhanced immunotherapy response. See related article by Pozniak et al. p. XX