Abstract 1054: Systematic establishment of robustness and standards in xenograft experiments and analysis

Abstract

Abstract Patient-Derived Xenograft (PDX) models are a valuable model system for demonstrating clinical translatability in the preclinical testing of therapeutic agents. Large PDX collections such as the NCI PDXNet Consortium are a critical resource for assays of therapeutic efficacy. However, research groups vary in their procedures for experiments and assessments, limiting conclusions about the robustness of PDX results and their mapping to patient outcomes. To determine the sensitivity of PDX assays to different procedures, the PDXNet conducted a study of temozolomide drug response for PDX models derived from three patients, performed in replicate across four PDX Development and Trial Centers (PDTCs) with all sites blinded to the procedures and results of other groups. The three PDX models were chosen based on prior treatment experiments conducted by the NCI Patient Derived Models Repository that showed that one model was sensitive, one was resistant, and one had intermediate sensitivity. Each PDTC used their internally developed SOPs, which varied in their procedures for establishment of xenografts, treatment dosing and scheduling, time period of study, and response criteria. Additionally all groups followed internal sequencing procedures to generate one exome-seq and one RNA-seq per model. Drug responses were consistent across all five testing groups, with each group identically classifying the sensitive, resistant, and intermediate models. We also trained novel statistical classifiers for assessing sensitivity, resistance, and intermediate response on 105 retrospective cohorts of mice collected among the PDTCs. In addition, we compared and benchmarked exome-seq and RNA-seq analysis pipelines developed by each individual PDTC on synthetic test data. We assessed the robustness of sequencing assessement on PDX tumors by applying the optimized pipeline across the replicate samples generated across the separate PDTC groups. Overall, our results show that drug responses of PDX models are consistent and robust across different academic centers. We are developing standardized SOPs for experimental procedures, response measurement, statistical assessment of response, and sequence analysis based on these results. All SOPs will be released publicly through protocol descriptions and shared computational workflows within the NCI Cancer Genomics Cloud. We expect these PDXNet standards to improve the use of PDX and other in vivo models in advancing cancer precision medicine. Citation Format: The NCI PDXNet Consortium, Yvonne Evrard, Jeffrey H. Chuang. Systematic establishment of robustness and standards in xenograft experiments and analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1054.