Immune interception strategies across pre-cancers using single-cell and spatial atlases
Primary supervisor: James Reading, UCL
Secondary supervisor: Nicholas McGranahan, UCL
Project
Cancer prevention represents one of the most promising frontiers in oncology, particularly as screening programs increasingly identify pre-cancerous lesions that present unprecedented opportunities for prevention in high-risk individuals. Growing evidence points to immune evasion as a fundamental checkpoint to malignant transformation in multiple cancers [1,3]. Recent advances in transcriptomics have revealed that pre-cancerous lesions harbour distinct immune microenvironments [3]. This creates a unique therapeutic window where targeted immune interventions could prevent cancer development entirely, transforming prevention from passive surveillance to active interception [1].
Our team has discovered that the nascent tumour microenvironment drives a finely balanced CD4 T cell effector and regulatory response that determines the progression or regression of preinvasive lesions via heterogeneous cellular preinvasive immune hubs, using squamous lung cancer as a model disease (Gamble S et al, in revision). Utilising a curated multi-cancer dataset, the student will use computational analysis of single-cell spatial profiling of antigen-specific T cells and patient-derived explants in preinvasive disease to understand and rewire the T cell response for cancer interception. This computational immunology project will systematically map T cell regulatory networks across human pre-cancerous lesions to identify actionable targets for cancer interception with multi-cancer applicability. Using our comprehensive pre-cancer immune atlas spanning ~10 cancer types, the student will characterize how T cells recognize early neoplastic changes and identify conserved regulatory mechanisms that suppress anti-tumour immunity during pre-malignant progression.
The key aims include analysing single-cell and spatial transcriptomic profiles from paired normal tissue, pre-malignant lesions, and early cancers using advanced informed computational tools to identify conserved immunosuppressive pathways acting on T cells within spatially defined cell neighbourhoods across cancer types. High-confidence targets will be validated using patient-derived explant models to assess response. This will be combined with personalized neoantigen stimulation, in order to model the impact of target modulation on neoantigen vaccine response [2].
This primarily computational project offers training in cutting-edge single-cell and spatial transcriptomics, neoantigen prediction, and translational immunology. The candidate will work with state-of-the-art curated datasets, including a Xenium-based spatial transcriptomics atlas, and develop novel approaches for precision cancer prevention.
The ultimate goal is to identify targetable immune checkpoints that can synergize with neoantigen vaccination to prevent progression from pre-cancer to invasive disease. This represents a paradigm shift from surveillance-based screening toward active preventative intervention, with potential to prevent thousands of cancers annually. The ideal candidates are medical students with interests in computational immunology, cancer prevention, and precision medicine. This project positions them at the forefront of preventive oncology, combining cutting-edge computational approaches with direct clinical relevance to develop the next generation of cancer interception strategies.
References
- Reading JL, Hwang ES, Yu J, Dotto GP, Grady WM, Czerniak B, et al. Pre-cancer: From diagnosis to intervention opportunities. Cancer Cell. 2023 Apr 10;41(4):637–40.
- Reading JL, Al Bakir M, et al. Clonal driver neoantigen loss under EGFR TKI and immune selection pressures. Nature. 2025 Mar;639(8056):1052-1059.
- Mascaux C, Angelova M, Vasaturo A, Beane J, Hijazi K, Anthoine G, et al. Immune evasion before tumour invasion in early lung squamous carcinogenesis. Nature. 2019 Jul;571(7766):570–5.