Abstract SY47-01: Single-cell mapping of tumor ecosystem reveals chemotherapy-induced myeloid-driven spatial exhaustion via NECTIN2-TIGIT ligand-receptor interaction which can be targeted to unleash anti-tumor immunity in ovarian cancer

Abstract Tumorigenesis and cancer progression are dependent on the ability of malignant cells to escape the recognition and attack by the host's immune system, involving a complex spatial interplay of cell populations. T cell exhaustion is a process in which T cells become hypofunctional upon chronic antigen stimulation; a subset of these cells can be reinvigorated by immune checkpoint blockade therapies (ICB). Cancer immunotherapies using ICB have been widely adopted in the clinical setting. However, ovarian cancer is one of the diseases where ICB therapies have failed in unselected patient populations. Anti-tumor immunity plays a critical role in high-grade serous ovarian cancer (HGSC) the most aggressive gynecological cancer. Yet little is known how the tumor ecosystem is modulated by standard chemotherapy. Recent developments in spatial technologies have enabled the detailed investigation of cellular phenotypes within their local tumor microenvironment (TME) niches. It has been demonstrated that spatial localization dictates the cellular phenotypes, and that spatial interactions play a crucial role in how tumor cells evade immune attack. Importantly, the efficacy of ICBs is dependent on the spatial interactions between neighboring cells with distinct activation states, which are reflected in the multicellular organization of the tumors. Previously, my lab has identified a phenotypically distinct TME characterized by increased spatial immunosurveillance and unique prognostic implications for BRCA1/2-deficient HGSCs (Launonen et al. Nat Comm 2021). Consistently, we demonstrated that responses to the combination of ICB and Poly-ADP Ribose Polymerase (PARP) inhibitors are governed by a distinct functional and spatial immune microenvironment (Färkkilä et al Nat Comm 2020). To uncover the spatiotemporal effects of chemotherapy on the tumor ecosystem in HGSC, we incorporate multi-omics profiling using genomics, scRNA-seq, high-dimensional whole section tissue imaging and spatial transcriptomics from 117 patient samples capturing over 25 million single-cells, and 316 spatial transcriptomic domains captured by spatial molecular profiling. For genomic stratification, we utilized whole genome sequencing, bulk RNA sequencing and AI-driven scoring for homologous-recombination deficiency (HRD) with ovaHRDscar, an algorithm which we developed and that is currently clinically implemented and utilized for centralized testing of HRD in Finland (Perez-Villatoro NPJ Prec Oncology 2022). Detailed analysis of the single-cell and spatial tumor topographies revealed increasingly versatile immune cell states forming spatiotemporally dynamic microcommunities in response to chemotherapy. We discover networks of interconnected myeloid cells termed “Myelonets” as a previously uncharacterized spatial domain in the tumor ecosystem. Myelonets ranging from tens to thousands of cells in size show distinct functional and spatial characteristics. Importantly, we demonstrate that Myelonets spatially drive the CD8+ T cell exhaustion post-chemotherapy, which associates with poor chemoresponses of the patients. Moreover, we show that M1/M2 polarization at the tumor-stroma interface is associated with CD8+ T cell exhaustion and exclusion, both correlating with poor chemoresponse. Single-cell and spatial transcriptomics reveal prominent myeloid-T-cell interaction occurring specifically via signaling via NECTIN2-TIGIT ligand-receptor interactions induced by chemotherapy. To functionally investigate therapeutic targeting of ICB and combination therapies in HGSC, we established a unique patient-derived immunocompetent culture (iPDC) model cultured on a novel physiologically relevant patient- derived matrix. We validate our functional immuno-oncology platform using histological, genomic and single- cell analysis, and show that the iPDC responses significantly correlate with the patient’s clinical responses and progression free survival, confirming that the iPDCs serve as a unique, high-throughput and functional platform for testing immunotherapy combinations. Finally, using our iPDC platform we successfully target the immunosuppressive NECTIN2-TIGIT interactions, and show that the CD8+T-cells samples can be reinvigorated using ICB post chemotherapy. Importantly, using single-cell analyses, we show that high tumor tissue NECTIN2-TIGIT signaling serves as a predictive biomarker for responses to anti-TIGIT therapies. Our findings, based on a unique patient cohort, unprecedentedly large single-cell spatial datasets, and functional models underscore the intricate co-evolution of TME topography during chemotherapy. This co-evolution exerts profound effects on niche-specific anti-tumor immunity, revealing mechanisms of immune evasion and potential therapeutic targets. Our discovery on the spatial myeloid-driven CD8+T-cell exhaustion unlocks new immunotherapeutic strategies to enhance anti-tumor immunity in HGSC. Citation Format: Anniina Farkkila. Single-cell mapping of tumor ecosystem reveals chemotherapy-induced myeloid-driven spatial exhaustion via NECTIN2-TIGIT ligand-receptor interaction which can be targeted to unleash anti-tumor immunity in ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr SY47-01.