14 PDX models of pmCRC were established, including 9 matched PD3D/PDX models, and treatment response to 17 SoC and targeted therapies was monitored, which resulted in drug-dependent inter- and intra-patient specific growth inhibition. Analysis of the molecular data of the models and patient tissue resulted in the identification of predictive biomarkers for treatment with 5-FU, irinotecan, trametinib, erlotinib, cetuximab and avastin, oxaliplatin, selumetinib, docetaxel and everolimus. Conclusions: The establishment of a preclinical drug testing platform based on in vitro and in vivo matched pmCRC models, molecularly characterized by multi-omics technology, facilitated the identification of predictive biomarkers for treatment response, as well as cancer relevant signatures for effective therapies, ready for validation in clinical cohorts.