The JAX Patient-Derived Xenograft Program: A Unique Resource to Advance Genome-Guided Cancer Medicine and Therapeutic Agent Testing.

e22151 Background: To address the demand for new pre-clinical models,The Jackson Laboratory (JAX), in collaboration with founding member UCDCC and 24 other medical centers, has engrafted primary and metastatic tumors from over 20 types of human cancer into NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ (NSG) mice to create a bank of >270 clinically annotated and genomically characterized Patient Derived Xenografts (PDXs). Compared to cell lines, PDX models reflect the complexity of human tumors, including clonal heterogeneity and stromal interactions. Sub-grouping PDX models using genomic properties (e.g., cancer driver gene mutations) provides a unique platform for study of new therapies and mechanisms of drug resistance. Methods: Patient tumors were implanted subcutaneously in the flank of NSG mice. The first tumor to reach the target volume was harvested and used for expansion to a second passage (P1). Tumors from P1 were cryoviably frozen and re-implanted as needed to create study cohorts for therapeutic agent testing. PDX tumor genomes were characterized for somatic mutations, gene expression, and copy number variation. Some models have been tested for drug response using Standard of Care drug treatment regimens as part of the PDX platform validation strategy. Results: The JAX PDX resource includes 270 established models and 640 models in development from 1270 implantation events. Engraftment success varies significantly by tumor type, grade and stage. Genomic analysis reveals a high degree of complexity among different models established for the same tumor type, which can be subgrouped based on mutational status and other genomic properties. Passaged tumors show strong concordance in histology and genome properties with the original patient tumor and across passages. Conclusions: The JAX program is unique among similar resources in the breadth of the cancers represented, the use of the NSG mouse as the tumor host, the depth of genomic characterization of the models, and the quality of clinical annotation. This platform is ideally suited for drug development, precision therapeutic interventions based on genomic characterization, and identification and circumvention of drug resistance.