Past & Ongoing projects
Please find below a description of some of the ongoing projects within the Consortium:
Harmonisation of procedures for the establishment, characterisation and use of PDX models, as well as for quality controls, is highly needed in order to optimise the sharing of models across laboratories and improve the reproducibility of preclinical results. Our work on standards is currently focused on three aspects:
- In November 2014, hosted by the Vall d'Hebron Institute of Oncology in Barcelona, we dedicated a workshop to defining common standards for the validation and characterisation of PDX models. First, our preliminary conclusions on the minimal information required to be made available for PDX models were since then taken over by Terry Meehan and colleagues at the EMBL-EBI ) in order to reach a global agreement. Involving 25 institutes worldwide, an article is under revision and these principles will be the basis of our future repository of models. In addition, we reached an agreement to generalise SNP based fingerprinting of the models as part of quality controls, in order to ensure the genealogy of a sample and its relatives.
- A bottleneck in the whole process of sharing models is the different health status of the animal facilities. In addition, the microbiological status of PDX models directly influences experimental parameters and imposes a risk to the personnel. We have initiated discussions between the representatives of animal facilities in the different institutions member of the Consortium, with the objective to reach a common procedure for health monitoring.
- The Laboratory Assistant Suite (LAS, Demo available here) is a laboratory information management system developed by the University of Torino in collaboration with the engineering school of the Politecnico di Torino in Italy (Baralis E et al. 2012; Fiori A et al. 2014). LAS aims at reducing operator-dependent error in PDX biobanking and in vivo data generation, tracking and storage, by minimising manual data entry through the use of an electronic system based on systematic barcoding. LAS has already been set-up by four different EurOPDX members. A distributed LAS system will permit direct sharing of PDX biobanking and in vivo data across EurOPDX centres.
COMMON DATABASE: the EurOPDX cBioPortal:
In collaboration with the Elemento lab at Weill Cornell Medical College and the Schultz lab at the Memorial Sloan Kettering Cancer Center (New York, NY, USA), as well as the University of Torino (laboratories of Translational Cancer Medicine and Oncogenomics), Ales Krenek and coll. at Masaryk University (Czech Republic) deployed an instance of the cBioPortal dedicated to EurOPDX data visualisation.
The cBioPortal allows users an exploration of multidimensional cancer genomic data (http://www.cbioportal.org/, Gao et al. Sci. Signal. 2013, Cerami et al. Cancer Discov. 2012). In this pilot stage of the EurOPDX cBioPortal (available at here) 5 colorectal or breast cancer PDX models cohorts (or "studies") from 4 laboratories of the consortium are included, and it is our aim to gradually include PDX data from the whole consortium. A study summary is available for each of them, and one can also exploit the cBioPortal tools for browsing DNA copy-number data, mRNA expression data and mutation data for a single study, or across studies by querying for instance series of genes such as EGFR, ERBB2, ERBB3, EGF, EREG, KRAS.
A "Xenopatient Discovery Platform for the integrated Systems based Identification of Predictive Biomarkers for Targeted Therapies in Metastatic Colorectal Cancer", is led by Dr. Annette Byrne at RCSI, in collaboration with EurOPDX members from the University of Turin and KU Leuven, as well as other partners (further information here). The ColoForetell project has received funding from Science Foundation Ireland, and commenced in December 2014.
Breast cancer pharmacogenomics:
Under the lead of Dr. Alejandra Bruna and Prof. Carlos Caldas at the CRUK Cambridge Institute (CRUK-CI), a pipeline to perform high-throughput single and combination drug screening assays in a feasible and cost-effective manner has been developed, by combining in vivo maintenance of human derived samples from breast cancer (BC), as low passaged PDXs, with ex vivo short-term cultures (PDX cells or PDCs) (Bruna et al. Cell 2016, link to the article). This PDX_PDC integrated platform was comprehensively characterized, which showed that it can successfully recapitulate patient-specific treatment and rationally predicted targeted treatment responses.
The authors used a large biobank of highly annotated BC PDX samples to conduct the study. This large biobank is being maintained and expanded at the CRUK-CI and is currently one of the largest BC PDX biobanks in the world in the academic setting. It is currently composed of n=93 models, from which some were initially established in other EurOPDX centres: Institut Curie (Paris, France, PI: Dr. Elisabetta Marangoni), the Vall d'Hebron Institute of Oncology (Barcelona, Spain, PI: Dr. Violeta Serra) and the Netherlands Cancer Institute (Amsterdam, The Netherlands, PI: Dr. Jos Jonkers). This initiative led by the Caldas laboratory therefore represents the first biobank of EurOPDX models with extensive molecular and drug response annotation that can be easily browsed in a public web portal developed in Cambridge (link).
This publication also supports one of the EurOPDX main objectives of biobanking and sharing large numbers of PDX models for easy access to the scientific community, and generalising the use of state-of-the-art PDX-derived in vitro models.