Spatial multi-omics profiling of clear cell renal cell carcinoma (ccRCC) pre-and post-immunotherapy identifies new druggable targets

Introduction: Immunotherapy (IO) and VEGF targeting TKIs have transformed outcomes of patients with advanced ccRCC. However, much less is known about treatment resistance mechanisms after IO therapy. Methods: 15 patients (pts) with advanced ccRCC who had tumor samples pre- and post-treatment (Tx, IO-IO or IO-TKI) were included in this study. Tumors were profiled using the Nanostring GeoMx spatial profiling platform. 4 to 12 regions of interest (ROIs) were selected for each patient, and the ROIs were further segmented into tumor (PanCK+), immune (CD45+), and stroma (PanCk-, CD45-) compartments. Differentially expressed genes and proteins were assessed with Gene Set Enrichment Analysis (GSEA). Bulk gene expression was further deconvoluted to profile immune cell subsets using CIBERSORT. Deep learning algorithms analyzed H&E slides for angiogenic, immune gene signatures, and non-negative matrix factorization (NMF) clustering. Results: In the tumor compartment (PanCK+), several clinically druggable targets such as Axl, HER2, TROP-2, GITR, HIF-1alpha, and FGFR2 were significantly upregulated after IO. Interestingly, Trop-2 expression was upregulated in IO-IO (Log2FC 0.038, p<0.01) post-Tx but not after IO-TKI (Log2FC -0.070, p<0.01). Some patients who received IO/TKIs had decreased tumor angiogenic expression post-Tx. NMF clustering classified post-Tx as NMF4 in 6/15 patients (4 of 10 in IO-TKI cohort and 2 of 5 in IO-IO cohort). NMF cluster 3 remained post-IO in most patients.While most pre-IO samples had moderate to high immune infiltration, post-IO samples had even higher levels. Notably, CIBERSORT profiling revealed post-IO proportion of immune suppressive regulatory T cells were decreased (IO-TKI, p <0.01), while cytotoxic CD8+ T cells were increased (IO-TKI, p <0.01) with both Tx cohorts. However, resting NK cells were increased (p=0.29), while activated NK cells were decreased (p <0.01) significantly only in IO-IO treated subsets. Conclusion: Molecular gene expression changes in RCC after IO-based therapy identified clinical targets in residual tumors. NMF cluster 4 persistence, CD8+ infiltration, and reduction in Treg proportion suggests an expansion or persistence of immune cells post-IO. NMF cluster 3 and increased proportion of resting NK cells may explain the resistance of residual tumor post-IO. Given the heterogeneity in RCC, it is imperative to profile individual tumors to improve IO therapy outcomes. Ongoing analyses include IHC target validation.