173 Deciphering the tumor microenvironment of renal cell carcinoma: integrating single-cell sequencing and In Situ spatial transcriptomics for comprehensive characterization of tumor-infiltrating T cells

BackgroundRenal cell carcinoma (RCC) is a prevalent and aggressive form of kidney cancer, with variable response rates to checkpoint blockade immunotherapies. While high T cell infiltration often indicates a favorable response to immunotherapies, this correlation is inconsistent in RCC. To accurately identify RCC patients who will benefit from immunotherapies, additional markers are needed to differentiate between subsets of tumor-infiltrating T cells. Recent advances in single-cell sequencing and subcellular spatial transcriptomics have revolutionized cancer immunology, enabling the analysis of individual cells and their interactions within the tumor microenvironment.MethodsIn this study, we used comprehensively analyzed the molecular profiles of individual cells in RCC to understand the interplay between tumor cells and neighboring immune cells. Utilizing patient-matched peripheral blood mononuclear cells, cryopreserved dissociated tumor, and adjacent normal cells, we performed single-cell RNA, antibody-derived tag, and TCR sequencing on the blood and dissociated cells (n=10 patients). 10x Genomics Xenium in situ spatial transcriptomics with single-cell resolution was conducted using a custom designed gene panel targeting kidney, tumor specific genes and various immune related genes with emphasis on those relevant to the analysis of T cells, which was guided by single-cell RNA sequencing data. Formalin-fixed paraffin-embedded samples from tumor and adjacent normal tissue were subjected to in situ spatial transcriptomics (n=4 patients).ResultsBy aligning single-cell RNA sequencing and in situ spatial transcriptomics with single-cell level resolution, our findings revealed highly heterogeneous yet distinct T cell subpopulations within renal cell tumors compared to adjacent normal tissues. Notably, tumor-infiltrating CD8 and CD4 T cells expressing CXCL13, which is often associated with T cell exhaustion program, demonstrated a highly organized spatial distribution. These cells were positioned in close proximity to tumor cells, suggesting antigen-specific engagement. Furthermore, we tracked corresponding T cell subsets and clones in patient-matched blood samples, assessing their phenotypes as potential biomarkers for tumor-reactive T cells.ConclusionsThrough the integration of single-cell multi-omics analyses and single-cell in situ spatial transcriptomics, we addressed critical questions concerning tumor heterogeneity, spatial organization, and the dynamic