MAPPING CANCER-ASSOCIATED CELLS IN CONTEXT: A COMPUTATIONAL APPROACH TO MYELOID AND FIBROBLAST DYNAMICS IN CANCER

Cancer is a multifaceted disease, rooted in developmental process and fuelled by inflammation as an evolutionary driver. Timing is critical: in rapidly progressing tumours such as pancreatic ductal adenocarcinoma (PDAC), inflammatory loops stem from immune dysregulation driven by IL-1β– producing tumour-associated macrophages (TAMs); conversely, in more indolent malignancies like colorectal cancer (CRC), inflammation remodels the stromal–tumour interface to favor the emerging tumour micro-environment (TME). Both pathways converge on a common outcome: malignant transformation and establishment of a TME niche that fosters tumorigenesis. To unravel these mechanisms, we combined single-cell sequencing, spatial transcriptomics, and machinelearning approaches. In the immune compartment, we delineated an early PGE₂–TNFα–IL-1β circuit in TAMs that marks initial dysregulation that led to immune suppression. In parallel, using a collection of CRC models, we identified conserved stromal–tumour interactions where cancerassociated fibroblasts (CAFs) orchestrate TME adaptation by linking oncogenic signalling with fibrotic pathways. Our findings reposition inflammation as a master regulator of TME evolution, advocating for therapies targeting inflammatory circuits in both immune and non-immune compartments to disrupt stromal–tumour synergy. This work establishes a roadmap for stroma and immune-aware oncology, emphasizing temporal inflammation dynamics as critical determinants of therapeutic vulnerability