Abstract 178: Automating a spatial profiling workflow to explore the effects of hypoxia in the tumor microenvironment in head and neck cancer

Abstract Background: Head and neck cancer (HNC) encompasses a diverse group of malignancies originating from the mucosal surfaces of the upper aerodigestive tract. The tumor microenvironment (TME) in HNC is marked by immune cells, stromal cells, and extracellular matrix components. Hypoxia, a key feature of the TME, drives tumor growth, invasion, and metastasis by modulating gene expression related to angiogenesis, cell survival, and metabolism. Understanding the interplay between hypoxia and immune infiltrates in the TME is essential for developing new therapeutic strategies for HNC. While whole transcriptome analysis via digital spatial profiling is a powerful method for probing the TME, it is time-consuming for large cohorts. Automating the profiling workflow can reduce hands-on time and selection bias in regions of interest (ROIs), facilitating the exploration of large cohorts to identify mechanisms of action, potential drug targets, and biomarkers. Methods: We developed an optimized spatial multi-omic workflow for high-throughput spatial analysis using the GeoMx® Digital Spatial Profiler (DSP) with the Whole Transcriptome Atlas (WTA) and immunofluorescent morphology markers: SYTO82 (nuclei), CAIX (hypoxia), pan-cytokeratin (epithelium), and CD3 (T-cells). AI-based analysis (Oncotopix® Discovery) of serial section H&E and GeoMx IF images was employed to identify ROIs for GeoMx collection. Immune hot and cold selections were based on leukocyte density, while tumor/stromal interface selections focused on epithelial areas. Areas of illumination (AOI) were chosen using concentric CAIX expression gradients. Integrated analysis of digital images and the whole transcriptome was conducted to assess the TME compartments. Results: Automated ROI placement based on tumor/stroma, hypoxia, and immune infiltration, along with AI/Deep Learning-based AOI segmentation, reduced AOI selection time and improved tissue compartment enrichment accuracy across samples and tissue types. The development of hypoxia gradient-based AOI selection enabled a strategy not possible with standard DSP software. Cell phenotyping using IF morphology scans supervised cell deconvolution. DSP results showed strong correlation with patient outcomes. Conclusions: This study demonstrates that ROI-based spatial analyses can effectively explore the impact of hypoxia on immune infiltration in HNC. Automated AI-based ROI selection offers an unbiased, reproducible method for sampling relevant tumor subtypes based on hypoxia and immune infiltrate criteria. This approach provides a standardized, automated method for selecting ROIs and segmenting AOIs across diverse tissue types and pathological subtypes, enhancing throughput. Citation Format: Kyla Teplitz, David Mason, Regan Baird, Kelly Hunter, Joana Campos, Jill Brooks. Automating a spatial profiling workflow to explore the effects of hypoxia in the tumor microenvironment in head and neck cancer [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 178.