Spatial multiomic profiling to identify de-differentiated tumor phenotypes and exhausted T cell niches in rhabdoid and sarcomatoid RCC.

4557 Background: Rhabdoid and sarcomatoid (S/R) differentiation are high-grade dedifferentiated states that can arise across renal cell carcinoma (RCC) subtypes. Although associated with poor patient prognosis, S/R RCC paradoxically exhibits increased sensitivity to immune checkpoint blockade. The mechanisms underlying this aggressive, yet immunotherapy-responsive phenotype remain unclear. Spatial multi-omic profiling offers a powerful framework to interrogate tumor-intrinsic and microenvironmental programs that account for this phenotype. Here, we applied spatial multi-omic analyses to treatment-naïve clear cell RCC specimens from seven patients, enabling direct comparison of sarcomatoid, rhabdoid, and clear cell regions captured within the same tissue sections. Methods: Hematoxylin and eosin (H&E)–stained sections from seven FFPE patient samples were reviewed and annotated by a genitourinary pathologist to identify discrete regions of sarcomatoid, rhabdoid, or clear cell morphology. Annotated FFPE samples were processed by Singular Genomics using the G4X spatial multi-omic platform, which enables quantification of 315 RNA transcripts, 14 protein targets, and fluorescent H&E imaging. Results: Initial spatial multi-omic profiling resolved distinct tumor regions with clear cell biology alongside areas exhibiting rhabdoid or sarcomatoid features. Comparative analysis of immune infiltrates across these regions revealed enrichment of both T cells and B cells within rhabdoid regions. In addition, T cells localized to rhabdoid and sarcomatoid regions displayed distinct phenotypic states compared with those in clear cell regions. Differential gene expression (DGE) analysis of T cells from rhabdoid and sarcomatoid regions relative to clear cell RCC demonstrated upregulation of activation markers ( CD69 , IL2RA ), exhaustion markers ( LAG3 , PDCD1 ), and cytotoxic genes ( GZMA , GZMK , NKG7 ). Parallel analysis of tumor cell transcriptional programs further highlighted distinct tumor biology across histologic regions. Sarcomatoid tumor cells exhibited increased expression of genes associated with cellular stress, inflammation, and epithelial–mesenchymal transition (EMT), including IGFBP7 and FOS . In contrast, rhabdoid tumor cells showed upregulation of antigen-presentation machinery ( HLA-A , TAP1/2 ) along with elevated expression of EMT-associated genes ( MET , VCAM1 , and CD44) . Conclusions: Spatial multi-omic analysis of S/R RCC revealed distinct immune and transcriptional programs across sarcomatoid, rhabdoid and clear cell regions. T cells within S/R regions exhibited activated and exhausted phenotypes, while tumor cell characterization of S/R cells reveals upregulation of stress, inflammatory and antigen-presentation machinery, features that may underlie the sensitivity of S/R RCC to immune checkpoint inhibitors.