Abstract 6279: Identification of novel secreted proteins modulating immunotherapy efficacy through spatial transcriptomics models

Abstract Background: Secreted proteins, derived from around 1900 human protein-coding genes, play a crucial role in mediating intercellular communication. Despite their significance, the downstream signaling effects of most secreted proteins in shaping antitumor immunity remain largely unexplored, except for extensively studied cytokines. Spatial transcriptomics (ST), which profiles gene expression in the spatial context of intact tissues, enables us to investigate spatial relationships between secreted proteins and their downstream targets. This technology provides insights into the activity and function of secreted proteins within the tumor microenvironment. Objective: The current study aims to leverage target gene signatures trained from ST data to identify novel secreted proteins modulating the efficacy of immunotherapy. Results: We collected hundreds of tumor ST samples across over twenty cancer types. Using Moran’s I correlation, we assessed the spatial relationships between secreted proteins and potential target genes. After filtering out low-quality targets, approximately 1, 200 secreted protein signaling signatures were retained. Based on these data, we developed SecAct, a multiple regression framework to predict secreted protein signaling activity. Next, SecAct’s reliability was validated using clinical trial data from cytokine-blocking treatments. We observed that the inferred activity reduction of secreted proteins significantly correlates with patients' clinical responses in both tumor and inflammatory disease settings, demonstrating SecAct’s clinical relevance. Subsequently, we applied SecAct to analyze forty-two cancer immunotherapy cohorts treated with immune checkpoint blockades, identifying secreted proteins whose activity is strongly associated with therapeutic efficacy. This prioritized list includes both established cancer immunotherapy regulators and novel candidates without previously known roles in cancer. Several new high-ranked regulators were further validated through in vivo mouse models. Conclusion: Our framework provides a data-driven screening ground for identifying novel regulators of antitumor immunity, potentially guiding the development of new immunotherapies. Citation Format: Beibei Ru, Lanqi Gong, Emily Yang, Kenneth Aldape, Lalage Wakefield, Peng Jiang. Identification of novel secreted proteins modulating immunotherapy efficacy through spatial transcriptomics models [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 6279.