Abstract 5063: Hepatomap: An interactive shiny app for spatial transcriptomics insights in hepatocellular carcinoma

Abstract Introduction: Hepatocellular carcinoma (HCC) is a heterogeneous cancer where spatial context and cellular microenvironments influence disease progression and therapeutic responses. Spatial Transcriptomics (ST) enhances immune analysis by mapping gene expression spatially, revealing cell organization and interactions. However, interpreting ST data in HCC is challenging due to high data dimensionality, unestablished bioinformatics pipelines, and fragmented knowledge in the literature. We introduce HepatoMap, a code-free interactive tool designed to streamline ST analysis for HCC research. It enables in situ visualization of expression patterns for curated HCC-associated signatures, identification of cellular interactions through cell type annotation, and uncovers molecular heterogeneity in HCC samples. Methods: A cohort of 13 HCC patients included baseline and radiotherapy-nivolumab treated samples (with 9 pairs), 2 super responders (SRs), 2 progressors (PRs), and 9 non responders (NRs). HepatoMap (R shiny v1.9.1) supports Visium data in Seurat object and spot annotation text files. Gene enrichment scores were computed via Seurat’s AddModuleScore. It includes 2 cell type annotation methods: MCP counter (marker-based) and SPOTlight (reference-based deconvolution method, with 2 pre-trained models built on TICAtlas and GSE149614 single-cell RNA references). It also provides a curated list of HCC-associated modules (cancer stem cell (CSC), driver genes), and allows users-upload gene sets , where Seurat’s AddModuleScore is used to compute enrichment scores. ggplot2 (v.3.5.1) is used for box plots, pie charts, and tissue visualization; ComplexHeatmap (v2.10.0) is for heat maps. The tool offers multi-tissue visualization for cross-sample comparison and downloadable publication-ready graphics. Results: To characterize immune response, the SPOTlight models revealed an increased T cells in treated SRs, with TICAtlas model providing finer resolution by identifying CD4 effector memory cells. SPOTlight offers resolution flexibility but assumes reference cell types represent the entire microenvironment, which may be limiting. MCP counter complements this by computing scores for predefined gene markers, revealing B cell increases in SRs and identifying additional cell types like endothelial cells and neutrophils for more comprehensive immune profiling. Additionally, contrasting to literature, SRs showed elevated CSC scores in tumor epithelium upon treatment, while a decrease was seen in NRs and PRs after treatment, adding depth to immune cell analysis. Conclusion: We presented HepatoMap, a tool streamlining ST analysis for HCC, uncovering molecular heterogeneity and cell-type dynamics through robust cell type annotation and tumor-specific profiling. The tool is adaptable to ST technologies like Xenium and Stereo-seq, expanding its relevance in spatial immunology. Citation Format: Wei Lin Tang, Teresa J. Tye, Menaka Priyadharsani Rajapakse, Solomonraj Wilson, Olaf RÖtzschke, David Tai, Darren Wan-Teck Lim, Juha P. Väyrynen, Joe Poh Sheng Yeong, Mai Chan Lau. Hepatomap: An interactive shiny app for spatial transcriptomics insights in hepatocellular carcinoma [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 5063.