Abstract PR004: Multi-omic spatial analysis reveals reshaping of tumour-immune dynamics at the transition to invasive colorectal cancer

Abstract The introduction of bowel cancer screening has led to an increase in the detection and removal of pre-malignant adenomas, however many of these would never progress to colorectal cancer (CRC) within a patient’s lifetime. This implies a key evolutionary bottleneck at the transition from adenoma to carcinoma, with immune surveillance likely playing a central role in suppressing the outgrowth of invasive cells. Here we analysed a cohort of “ca-in-ads” (adenomas with a small focus of cancer) to provide a unique snapshot of the invasive transition. We combined genomics, spatial transcriptomics, digital pathology and multiplex imaging, concurrently deriving the distribution of tumour clones, their gene expression and their interplay with the immune microenvironment. We performed multi-region low-coverage whole genome sequencing (lcWGS) and T-cell receptor sequencing (using FUME-TCRseq, a novel method developed in our lab) on archival ca-in-ads from 40 patients. For a subset of cases we performed spatial transcriptomics using the 10x Genomics Visium platform, including custom probes to detect T-cell clonotypes of interest. We applied a deep learning cell classifier to haematoxylin and eosin (H 2025 Dec 4-6; Albuquerque, NM. Philadelphia (PA): AACR; Cancer Res 2025;85 (23Suppl): Abstract nr PR004.