Abstract 2080: Whole transcriptome single cell spatial imaging of tumor vs normal tissue microarray

Abstract Tissue microarrays (TMAs) are widely used for high-throughput research but present challenges due to variations in tissue fixation, morphology, and cellular composition, which can impact staining, RNA/protein preservation, and segmentation accuracy. Analyzing diverse tissue types, especially when comparing normal and tumor samples, requires advanced imaging and computational tools. The CosMx® Spatial Molecular Imager (SMI) addresses these challenges by enabling accurate definition of single cells with multi-tissue AI segmentation models and high-resolution data extraction from complex TMA samples. We employed the CosMx SMI with the CosMx Human Whole Transcriptome (WTX) panel, which targets approximately 19,000 RNA targets, to analyze human FFPE tissue microarrays. Two FFPE TMAs—one with normal and the other with cancerous tissue—were constructed with cores from multiple organs including breast, skin, lung, colon, liver, prostate, and brain. RNA was detected in its native tissue context, and hematoxylin and eosin (H&E) staining provided histological reference for spatial transcriptomics. Uniform Manifold Approximation and Projection (UMAP) and Leiden clustering were used for dimensionality reduction and cell type identification based on marker gene expression. Differential expression (DE) and pathway analyses identified molecular signatures and biological processes across tissue types. CosMx SMI achieved high cell counts per sample and precise segmentation, enabling reliable analysis of diverse tissue types. H&E staining on the same slide enhanced AI algorithm training, improving tissue structure identification and segmentation accuracy. We identified distinct cell types and tissue-specific biomarkers. DE and pathway analysis revealed potential therapeutic targets and biomarkers, such as PSA, PSMA, AR, PTEN, and ERG fusion in prostate cancer, along with key pathways like AR signaling and DNA damage repair. Transcript profiles localized to specific cell types, demonstrating the platform's robustness in spatial transcriptomics. CosMx SMI significantly enhances single-cell transcriptome studies, providing detailed molecular insights across diverse tissue types. This platform is a powerful tool for tissue analysis, biomarker discovery, and spatial genomics, supporting a wide range of molecular and cellular research applications. -Bruker Confidential- Citation Format: Wei Yang, Isabel Lee, Michael Patrick, Yi Cui, Shanshan He, Martin Shelton, Haiyan Zhai, Michael Rhodes, Joseph Beechem. Whole transcriptome single cell spatial imaging of tumor vs normal tissue microarray [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 2080.