Abstract 6212: A differential ligand-receptor network inference method to identify alterations in communication between myeloid cells and CD8 T cells in response to PancVAX neo-epitope peptide vaccine, anti-CTLA-4 and anti-PD-1 antibodies in a murine model of pancreatic ductal adenocarcinoma

Abstract Pancreatic Ductal Adenocarcinoma (PDAC) is a deadly cancer with a low tumor mutational burden that leads to the generation of few neo-epitopes derived from mutated genes that can be recognized by cytotoxic T cells. The immunosuppressive microenvironment of PDAC comprises suppressive cell populations that include myeloid cells, cancer associated fibroblast subpopulations, and regulatory T cells that reduce T cell infiltration and cytotoxic function. We have developed a murine personalized cancer vaccine to enhance cytotoxic T cell function against PDAC, PancVAX (Kinkead et al, JCI Insight 2018). Survival benefit with PancVAX was dependent on the addition of immune checkpoint inhibitor (ICI) therapy to increase the number of cytotoxic CD8 T cells infiltrating tumors and maintain their cytotoxic function (Huff et al, JCI Insight 2023). Yet, the impact of cell-cell communication on effector T cell function arising from the immunosuppressive microenvironment of PDAC remains unknown. We previously developed a ligand-receptor inference method called Domino that infers inter- and intra-cellular signaling networks in single-cell RNA-seq data (Cherry et al, Nat Biomed Eng 2021). However, this technique is limited to data from single conditions. Understanding how this combination immunotherapy alters cellular signaling networks requires new computational tools that identify changes in regulatory networks between treatment conditions. We developed capabilities for assessment of differential signaling with Domino founded on a basis of differential signal receipt. Active receptors are identified by receptor expression in recipient cells that is correlated with downstream transcription factors in the intracellular signaling cascade. Identification of causative ligands for receptor activation is assessed by a probability of interaction for ligand-receptor interactions between cell types. This probability metric considers mean expression of ligand and receptor genes as well as the number of sender cells for comparison across treatment groups. Comparing communication networks in PancVAX alone to PancVAX with ICI, PancVAX with ICI enhances cell-cell adhesion and interferon gamma signaling in cytotoxic CD8 T cells towards exhausted CD8 T cells. Macrophages in PancVAX with ICI enhance expression of immune checkpoint ligands towards CD8 T cells, despite the presence of blocking ICI antibodies. Intracellular signaling downstream of PD-1 in exhausted T cells shows diminished numbers of transcription factors linked to PD-1 signaling. This software informs prioritization of inhibitory signals for modulation in future immunotherapy regimens and can be applied generally to other tumor subtypes and treatment contexts. Citation Format: Jacob T. Mitchell, Atul Deshpande, Amanda L. Huff, Chris Cherry, Sushma Nagaraj, Kavita Krishnan, Dmitrijs Lvovs, Jennifer H. Elisseeff, Elizabeth M. Jaffee, Neeha Zaidi, Elana J. Fertig. A differential ligand-receptor network inference method to identify alterations in communication between myeloid cells and CD8 T cells in response to PancVAX neo-epitope peptide vaccine, anti-CTLA-4 and anti-PD-1 antibodies in a murine model of pancreatic ductal adenocarcinoma abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 6212.