Integration of biological characteristics of pancreatic cancer PDX and CDX models with annotations of the donor patients

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy with a five-year survival rate of approximately 12%, underscoring the urgent need for improved therapeutic strategies. A deeper understanding of PDAC biology is essential and relies on preclinical model systems such as patient-derived xenografts (PDX) and cell-derived xenografts (CDX). However, the extent to which these models reflect the primary tumors and the respective donor patients’ disease course remains elusive. In this study, we characterized clinical, histological, and genomic data from the KFO5002 cohort, including 119 patients, 71 primary tumors, 43 PDX models, and 8 CDX models. All data was integrated and analyzed using a private space of the open source bioinformatical platform cBioPortal. Genomic comparisons revealed that PDX models showed minimal de novo mutations, while CDX models largely preserved the mutational profiles of their corresponding PDX tumors. Analysis of PDX engraftment demonstrated accelerated tumor growth across increasing mouse generations. The grading of primary and PDX tumors showed a positive correlation between primary tumors and first-generation PDX models and remained stable across subsequent generations. Immunohistochemical analyses indicated higher proliferative capacity in most PDX tumors compared to primary tumors and suggested a predominant classical PDAC subtype in most of the PDX tumors. CDX models displayed heterogeneous morphology, mutational landscapes, and proliferative capacities. Drug response assays using gemcitabine, 5-fluorouracil, and paclitaxel revealed variable treatment sensitivities. Although no significant correlation between CDX drug response and patient survival was observed, clinical survival data aligned with established treatment guidelines, highlighting the relevance of the KFO5002 data set. Overall, PDX and CDX models represent valuable but limited tools for PDAC research. Their complementary use, alongside emerging approaches such as patient-derived organoids, may better capture tumor heterogeneity and therapeutic response. This study further demonstrates the utility of cBioPortal as a suitable platform for integrated translational cancer research.