Abstract 7258: Enhancing TIL efficacy in NSCLC through epigenetic reprogramming and computational modeling

Abstract Novel therapeutic protocols are desperately needed to treat non-small cell lung cancer (NSCLC), the deadliest cancer world-wide. Recently, utilizing a patient’s tumor-infiltrating lymphocytes (TIL) is proving to be a viable and highly personalized approach. Despite its promise, many obstacles remain, such as refining expansion protocols for better in vivo TIL proliferative potential, improving T cell ability for homing-to and targeting tumor cells upon re-infusion, and avoiding IL-2/lymphodepletion related side effects. We hypothesize that inhibiting epigenetic enzyme EZH2 (EZH2i) will improve both TIL expansion and infusion outcomes in NSCLC patients. Yet, there are likely mechanisms of T cell suppression that EZH2i may not overcome. To address and identify these mechanisms, stochastic mathematical modeling of gene signaling in cancer has risen as a useful tool. Thus, we further hypothesize that modern sequencing data and mathematical biology can discover alternative mechanisms of T cell suppression that will suggest novel targets for TIL-combination treatment. To test these hypotheses, we have tested the EZH2 inhibitor valemetostat with anti-PD1 in a murine model of non-small cell lung cancer. We observed tumor regression and strong IFN-gamma T cell responses in combination treated animals. We also observed up-regulation of both major histocompatibility complexes, greater pro-T cell cytokine signaling, and myeloid populations became more tumor-eliminating. To translate these findings to TILs, we are currently securing NSCLC samples from our Biospecimen Core, establishing patient-derived tumoroids, and expanding TILs ex vivo. TILs will be expanded with/without EZH2 inhibitor valemetostat (Val) under “young” protocols. At ex vivo endpoint TILs will be assessed for Val-driven differences in viability, differentiation state, and reactivity via flow cytometry. To complement or studies with an in silico approach, we are utilizing libraries of publicly available NSCLC data to perform bioinformatic analysis and implemented equation learning techniques to infer NSCLC tumor signaling networks from clinical data between immunotherapy responders and non-responders. Differentiation patterns will be assessed as a function of pseudo-time-course data to inform predictions of signal transduction pathways. Lastly, mathematical phenotype control theory will be used to uncover potential targets within the tumor to boost TIL response in NSCLC. Given that advanced stage lung cancers remains refractory to most treatments, determining methods to improve precision medicine options such as TIL therapy is an important goal for the field. Our blending of classic bench science with bio-informatics has the potential to deepen our understanding of TIL therapy and to identify new way to improve therapeutic responses. Funding: T32 CA165990 (DRP), R01 CA237643 (CFB), R01 HL170193 (CFB), P30 CA177558 (Markey Shared Resources). Citation Format: Daniel R. Plaugher, Avery R. Childress, Christian M. Gosser, Dave-Preston Esoe, Xuilong Song, Jinpeng Liu, Christine F. Brainson. Enhancing TIL efficacy in NSCLC through epigenetic reprogramming and computational modeling [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 7258.