Gene network-based and ensemble modeling-based selection of tumor-associated antigens with a predicted low risk of tissue damage for targeted immunotherapy

• DOI: 10.1136/jitc-2023-008104
• Publication Date: 2024-05-10T03:05:32Z
• Authors: Christopher Lischer; Martin Eberhardt; Cindy Flamann; Johannes Berges; Esther Güse; Anja Wessely; Adrian Weich; Jimmy Retzlaff; Jan Dörrie; Niels Schaft; Manuel Wiesinger; Johannes März; Beatrice Schuler-...; Harald Knorr; Shailendra Gupta; Krishna Pal Singh; Gerold Schuler; Markus Vincent Heppt; Elias Andreas Tho...; Nadine D van Kleef; Julian J Freen-va...; Annelies W Turksma; Olaf Wolkenhauer; Bettina Hohberger; Carola Berking; Heiko Bruns; Julio Vera
• Keywords: 0303 health sciences; 03 medical and health sciences; Antigens, Neoplasm; Neoplasms. Tumors. Oncology. Including cancer and carcinogens; Humans; Basic Tumor Immunology; Gene Regulatory Networks; Immunotherapy; RC254-282

Tumor-associated antigens and their derived peptides constitute an opportunity to design off-the-shelf mainline or adjuvant anti-cancer immunotherapies for a broad array of patients. A performant rational antigen selection pipeline would lay the foundation immunotherapy trials with potential enhance treatment, tremendously benefiting patients suffering from rare, understudied cancers. We present experimentally validated, data-driven computational that selects ranks in multipronged approach. In addition minimizing risk immune-related adverse events by selecting based on expression profile tumor biopsies healthy tissues, we incorporated network analysis-derived indispensability index modeling results, candidate immunogenicity predictions machine learning ensemble model relying peptide physicochemical characteristics. study uveal melanoma, Human Leukocyte Antigen (HLA) docking simulations experimental quantification peptide-major histocompatibility complex binding affinities confirmed our approach discriminates between high-binding low-binding affinity performance similar established methodologies. Blinded validation experiments autologous T-cells yielded stimulation-induced interferon-γ secretion cytotoxic activity despite high interdonor variability. Dissecting score contribution tested revealed induce cytotoxicity but unsuitable due tissue damage instability were properly discarded pipeline. this study, demonstrate feasibility de novo capacity anti-tumor immune response predicted low damage. On translation clinic, supports fast turn-around validation, example, adoptive T-cell transfer preparations, both generalized personalized antigen-directed settings.