Patient-derived xenografts (PDX) models are an important oncology research tool for the study of tumours. They are created by implanting tumours from cancer patients into a ‘transplant-compliant’ mouse host.
PDX are thought to better reflect the heterogeneity of tumours than cell lines or transgenic animal models. This heterogeneity is usually what causes problems in clinical trials, as some patients respond well to a trial, while others do not.
PDX models help to better plan for clinical trials in patients by identifying and targeting subpopulations known to respond best to a drug and excluding those that do not. By selecting trial responders, the chances of success and saving massive amounts of money on failed trials are maximised. This is why pharma is interested in using PDX models and why PDX model selection is so crucial for designing a therapy to a non-responder to standard care.
Model selection is, however, a challenge. The description of the data about PDX models is not consistently annotated since to date no standard has been agreed, substantially limiting their findability. An agreed standard on PDX model definition will help the community to be able to compare different models for their selection and reuse. Meehan et al. have recently published a PDX models minimal information standard (PDX-MI) that defines clinical attributes of the patient’s tumour, the implantation processes, passaging and quality assurance.
Such a standard will facilitate the PDX model data exchange and the discoverability of models from distributed sources, promoting their research reproducibility. The Repositive PDX platform is already using the PDX-MI standard to facilitate the exchange of metadata about models between vendors and pharma users. We have also built internally on the standard to accommodate vendor and pharma specific needs and requirements. For example, we use PDX-MI attribute 'Tumor ID' in conjunction with an additional Repositive-specific Model ID to cater for multiple vendors having the same 'Tumor ID'.
Repositive welcomes this much needed standard and is actively working with it to provide a better service for its users. We look forward to actively contributing to this effort as we continue developing our platform.
Acknowledgement: Sam Shelton, Craig Smith and Anaid Diaz contributed feedback and content to this entry.