115P Preclinical models for cancer immunotherapy: A systematic review of ex vivo and in vitro platforms

<b>Background</b><br/>Cancer immunotherapy (CI) has shown remarkable clinical results in solid tumours, highlighting the need for robust in vitro and ex vivo models to improve efficacy and elucidate mechanisms of resistance. In vitro and ex vivo models represent a critical tool to advance CI. We conducted a systematic review to: 1) Map the use of preclinical models across cancer types, immune contexts, and immunotherapies, evaluating their degree of transferability 2) Provide a comprehensive dataset of studies employing these platforms to guide researchers.<br/><b><br/>Methods</b><br/>Adhering to PRISMA 2020, we searched PubMed (2000–2025) for studies in vitro or ex vivo models of solid tumours tested in an immuno-oncology context. The main extracted domains included model type, immune cell representation, type of immunotherapy assessed and relative transferability. Additional information (culture setup, validation, species) was collected for contextual comparison and practical usefulness. Extracted data were summarized into structured tables, with a complete dataset generated as supplementary material.<br/><b><br/>Results</b><br/>We included 75 studies spanning six main experimental model types: patient derived organoids (PDOs), 3D spheroid, 2D cell lines, Tumour tissue slices, Tumour-on-a-chip, and other types across 23 cancer types with a prevalence of breast, lung and colorectal cancers. The main immune cell type used were T lymphocytes (i.e. TILs and CAR-T cells) followed by NK cells. Models investigated different immunotherapy regimen such as Immune Checkpoints inhibitors, particularly anti-PD-1/PD-L1 antibodies, followed by Adoptive Cell Therapies such as CAR-T and NK cell-based therapies. While 2D models are suitable for initial screenings, 3D models (spheroids and PDOs) provide higher fidelity by recapitulating tumour microenvironment; tissue slices excel at preserving native tumour architecture, and tumour-on-a-chip systems are the most versatile for modelling dynamic immune processes.<br/><b><br/>Conclusions</b><br/>We have generated a curated catalogue of models for CI that provides researchers with a valuable resource to advance experimental immunotherapy and preclinical testing.<br/>