Discussion

The attrition rate of oncology drugs in the clinic is dismal – more than 90%, worse than any other disease area. One reason for this may be inadequate pre‐clinical models that suggest targets and efficacy against tumor growth that cannot be translated into the clinic. PREDECT partners are constructing model platforms, building complexity from 2D cell culture via 3D (co‐)cultures to in vivo orthotopic humanized mouse models on the one hand, or, on the other hand, starting out with human or GEMM‐derived tumors, and deconstructing these down to 2D cell cultures. All models will be compared to each other, and primary human tumor samples, using a central TMA platform, looking at tissue heterogeneity, interaction between individual compartments, preservation of signaling pathways ‐ ultimately trying to answer the question which model is fit for which purpose, and thus delivering a set of well characterized model platforms for pre‐clinical target validation, biomarker identification, and compound characterisation.