AbstractBackground:The groundbreaking SWOG S1801 and NADINA trials herald the use of neoadjuvant immune checkpoint blockade (NICB) as the standard of care for stage III-IV melanomas. The shift from radiographic to pathologic response scoring offers an unprecedented window of opportunity to interrogate ‘on-treatment’ biospecimens for the immune mechanisms of response and resistance at curative-intent surgeries. To date, single-cell/bulk RNA studies have demonstrated that TCF7+ stem-like CD8+ T cells and tertiary lymphoid structures (TLS) are positive predictors of ICB response. However, how these immune cells organize and communicate within the spatial context of the neoadjuvant tumor microenvironment remains poorly understood.Methods:We assembled a retrospective cohort of 68 patients with stage 3 melanoma: 35 NICB (Ipilimumab-nivolumab, nivolumab-relatlimab, nivolumab) (14 complete response, 5 partial response, 14 non-response) and 33 treatment-naive (TxN). We dissected the spatial molecular architecture of all tumor beds (n= 68) and 10 uninvolved LNs (5 post-PD-1/5 TxN) using integrative multiplexed error-robust fluorescence in situ hybridization (MERFISH, 305 genes), Vectra multiplex IF and RNAscope multiomics in 38 whole slides and 3 TMAs (77 cores). We also performed single-cell-FFPE-seq using matched blocks (n=12) as an orthogonal benchmark for the MERFISH data.Results:To analyze these high-dimensional datasets, we built a computational framework that includes 1) cell-typing with scFFPEseq validation, 2) novel spatial quantification methods to compute receptor-ligand (R-L) interactions that account for cell-cell distance and chemical signaling, 3) a graph-based clustering that quantifies germinal centers (GCs)/TLS, B-cell follicles and other spatial topography, and 4) MERFISH/ scRNA imputation via generative AI. Here, we showed that the quantity and size of spatially resolved GC/TLS and B-cell follicles surrounded by TCF7+ stem-like CD4+ and CD8 + T-cells, plasma cells, Tregs and myeloids are associated with positive response to neoadjuvant ICB. Our R-L analysis further revealed the critical importance of CXCR4-CXCL12 between TCF7+ T-cells and M2-myeloids and CCR7-CCL19/21 between TCF7+ T-cells in organizing the immune hubs in NICB response. Lastly, we incorporated scFFPEseq and MERFISH data into a unified latent embedding by environmental variational inference (ENVI) allowing us to impute melanoma phylogenetics, geographically accentuated hypoxic foci, and immune hubs within the spatial MERFISH images.Conclusions:We leveraged cutting-edge spatial -omics technologies and novel computational methods to resolve the immunologic hallmarks of NICB response in metastatic melanoma. We believe our approach will usher in a new paradigm for the investigation of other cancer biospecimens in the new era of standard of care neoadjuvant immunotherapy.Citation Format:Zichao Liu, Xiaofei Song, Jodi Balasi, Wei-Shen Chen, Jiang He, Justin He, Jonathan Nguyen, Carlos Moran-Segura, Joseph Johnson, Chaomei Zhang, Jane Messina, Zena Sayegh, Nan Sun, Douglas Marchion, Sean Yoder, Vernon K. Sondak, Jeffrey H. Chuang, Pei-Ling Chen. Spatially resolved immunologic hallmarks of response to neoadjuvant immune checkpoint blockade in metastatic melanoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5809.