PDX model initiatives, collections & standards (second part) 10:45 - 12:05

Zinaida Perova¹, Csaba Halmagyi¹, Alex Follette¹, Jeremy Mason², Mauricio Martinez¹, Federico Lopez Gomez¹, Tushar Mandloi¹, Abayomi Mosaku¹, Nathalie Conte³, Ross Thorne¹, Steven Neuhauser⁴, Dale Begley⁴, Debra Krupke⁴, Terrence Meehan⁵, Carol Bult⁴, Helen Parkinson¹

¹ EMBL-EBI, Hinxton, UK
² Melio Ltd, London, UK
³ AstraZeneca, Cambridge, UK
⁴ The Jackson Laboratory, Bar Harbor, ME, USA
⁵ Kymab Ltd, Cambridge, UK

Introduction

Patient-derived cancer models (PDCMs) are a powerful oncology research platform for studying tumour biology, mechanisms of drug response and resistance and for testing personalised medicine. Distributed nature of repositories for PDCMs (xenografts, organoids and cell lines) and the use of different metadata standards for describing model’s characteristics make it difficult for researchers to identify suitable PDCM models relevant to specific cancer research questions. PDCM Finder aims to solve this problem by providing harmonized and integrated model attributes to support consistent searching across the originating resources.

Materials and methods

PDCM Finder (https://www.cancermodels.org/) builds on the success of the PDX Finder resource (https://www.pdxfinder.org/, PMID:30535239).

Critical PDCM attributes, such as diagnosis, drug names and genes, are harmonized and integrated into a cohesive ontological model based on the PDX Minimal information standard (PDX-MI, PMID: 29092942). PDX-MI has become established in the community for data exchange, adopted by the PDX providers, consortia and informatic tools integrating PDX data. We are driving the development of and promoting the use of descriptive standards to facilitate data interoperability and promote global sharing of models. We provide expertise and software components to support several worldwide consortia including PDXNet, PDMR and EurOPDX. PDCM Finder is freely available under an Apache 2.0 license (https://github.com/PDCMFinder). This work is supported by NCI U24 CA204781 01, U24 CA253539, and R01 CA089713.

Results and discussion

Users can search for models of interest by either (1) exploring molecular data summaries for models of specific cancer types, or (2) using the intuitive search and faceted filtering options of the web user interface or (3) access resource database via REST API to run their own analysis. The data includes gene expression, gene mutation, copy number alteration, cytogenetics, patient treatment and drug dosing studies.

Conclusion

PDCM Finder is a new portal that aggregates patient-derived models from 27 academic and commercial providers, enables users to search and compare over 6000 models and associated molecular data, and connects users with model providers to facilitate collaboration among researchers.

Marieke van de Ven¹, Emilie Vinolo², Enrique Javier Arenas Lahuerta³, Joaquín Arribas³, Francesca Baietti⁴, Sander Basten⁵, Andrea Bertotti⁶, Massimiliano Borsani⁷, Annette T. Byrne⁸, Carlos Caldas⁹, Robert B. Clarke¹⁰, Didier Decaudin¹¹, Zdenka Dudova¹², Wendy Greenwood⁹, Stevene de Jong¹³, Jos Jonkers¹, Daniela Krasser¹⁴, Ales Krenek¹², Luisa Lanfrancone⁷, Eleonora Leucci⁴, Elisabetta Marangoni¹¹, Gunhild Mari Maelandsmo¹⁵, Jordi Martinez-Quintanilla³, Michaela Th. Mayrhofer¹⁴, Ian Miller⁸, Fariba Nemati¹¹, Jens Henrik Norum¹⁵, Geir Frode Øy¹⁵, Hetor G. Palmer³, Helen Parkinson¹⁶, Mauro Paschetta⁶, Zinaida Perova¹⁶, Alejandro Piris-Gimenez³, Leo Price⁵, Sergio Roman-Roman¹¹, Francesca Sarno¹⁷, Laura Soucek³, Katherine Spence¹⁰, Dalibor Stuchlik¹², Livio Trusolino⁶, Luca Vezzadini¹⁸, Alberto Villanueva¹⁹, Andrea Wutte¹⁴, Enzo Medico⁶, on behalf of the EurOPDX Research Infrastructure and the EurOPDX Consortium

¹ The Netherlands Cancer Institute, 1066CX Amsterdam, The Netherlands

² seeding science SRL, 1342 Limelette, Belgium

³ Vall d’Hebron Institute of Oncology and CIBERONC, 08035 Barcelona, Spain

⁴ Katholieke Universiteit Leuven, TRACE Platform, 3000 Leuven, Belgium

⁵ Crown Bioscience Netherlands, Leiden, The Netherlands

⁶ Candiolo Cancer Institute IRCCS and Department of Oncology, University of Torino, 10060 Candiolo, Torino, Italy

⁷ European Institute of Oncology, 20139 Milan, Italy

⁸ Royal College of Surgeons in Ireland, Dublin 2, Ireland

⁹ Cancer Research UK Cambridge Institute, Cambridge Cancer Centre, University of Cambridge, Cambridge CB2 0RE, UK

¹⁰ Manchester Cancer Research Centre, University of Manchester, Manchester M20 4QL, UK

¹¹ Institut Curie, PSL Research University, 75005 Paris, France

¹² Masarykova Univerzita, Institute of Computer Science, 625 00 Brno, Czech Republic

¹³ University Medical Centre Groningen, University of Groningen, 9713GZ Groningen, The Netherlands

¹⁴ Biobanking and BioMolecular resources Research Infrastructure - European Research Infrastructure Consortium, 8010 Graz, Austria

¹⁵ Oslo University Hospital, Institute for Cancer Research, 0424 Oslo, Norway

¹⁶ European Bioinformatics Institute, Wellcome Genome Campus, Hinxton, Cambridge CB10 1SD, UK

¹⁷ Hospital de Fuenlabrada, 28942 Fuenlabrada, Spain

¹⁸ Kairos3D SRL, 10134 Torino, Italy

¹⁹ Catalan Institute of Oncology ICO, Bellvitge Biomedical Research Institute IDIBELL, 08098 L’Hospitalet de Llobregat, Barcelona

Launched in 2013, the EurOPDX Consortium now gathers 18 academic research institutions throughout Europe and in the US (www.europdx.eu). The goal of the Consortium is to maximize exploitation of PDXs and other patient-derived models for cancer research by: (i) integrating institutional collections into a multicentre repository; (ii) defining common standards to improve the quality and reproducibility of oncology preclinical data; (iii) sharing models within and outside the consortium to perform collaborative precision oncology “xenopatient” trials. Building on its first successes, EurOPDX has been teaming up with other key academic and SME partners in a four-year project to build the “EurOPDX Distributed Infrastructure for Research on patient-derived Xenografts" (EDIReX project, Horizon 2020 grant no. 731105).

We will present our work towards standardisation and optimisation of biobanking, quality control and data tracking, and the performance of in vivo drug efficacy experiments, as a necessary step when setting-up this new cutting-edge Research Infrastructure offering access to PDX resources through 6 state-of-the-art installations or “nodes”. A cross-validation study was designed to experience the harmonisation of procedures in “real life” in the nodes and ensure interoperability. We will present the objectives of this study and its first results.

This comparative study goes even beyond the improvement of access provision by the EurOPDX Infrastructure. Improving reproducibility in preclinical research by investing in practical solutions as study design and biological materials remains a critical cornerstone to solve challenges such as attrition rate and costs of therapeutic drug development. We aim to improve preclinical and translational cancer research and promote innovation in oncology by integrating a European PDX repository and facilitating access to this much-needed resource for European and worldwide researchers.

Laia Monserrat¹, Claudia Yañez¹, Mónica Sánchez-Guixé², Cristina Viaplana³, Fara Brasó-Maristany^4,5, Meritxell Bellet^1,6, Sarat Chandarlapaty⁷, Lara Nonell⁸, Guillermo P Vicent⁹, Violeta Serra^1,10

¹ Experimental Therapeutics Group, Vall d’Hebron Institute of Oncology, Barcelona, Spain
² Institute for Research in Biomedicine (IRB Barcelona), The Barcelona Institute of Science and Technology, Barcelona, Spain
³ Oncology Data Science Group, Vall d'Hebron Institute of Oncology, Spain
⁴ Translational Genomics and Targeted Therapies in Solid Tumors, August Pi i Sunyer Biomedical Research Institute (IDIBAPS), Barcelona, Spain
⁵ Department of Medical Oncology, Hospital Clínic of Barcelona, Spain
⁶ Breast Cancer and Melanoma Group, Vall d’Hebron Institute of Oncology, Barcelona, Spain
⁷ Departments of Pathology and Human Oncology and Pathogenesis Program, Memorial Sloan-Kettering Cancer Center, New York, USA
⁸ Bioinformatics Unit, Vall d’Hebron Institute of Oncology, Barcelona, Spain
⁹ Molecular Biology Institute of Barcelona (IBMB-CSIC)
¹⁰ Vall d’Hebron Institute of Research

Introduction

Up to 90% of estrogen receptor-positive (ER+) breast cancers (BC) also express the androgen receptor (AR). Selective modulators of the AR (SARMs), such as RAD140, have shown clinical activity in endocrine therapy (ET)-resistance. However, the mechanism by which AR agonists block ER is not clearly understood. Moreover, the combination of a SARM plus CDK4/6-inhibitors (CDK4/6i) could be a novel therapeutic strategy for ER+/HER2- patients who do not tolerate ET. Here, we aimed to study the antitumor activity of RAD140 in a cohort of ER+/HER2- patient-derived xenografts (PDX) to identify biomarkers of response.

Material and Methods

Hormone-dependence was evaluated in 30 PDXs using ovariectomized (OVX) vs non OVX mice. Alterations in ESR1 and PIK3CA genes were determined with the MSK-IMPACTTM targeted exome panel and RNA-sequencing analysis. Pharmacodynamic experiments (PD) were conducted in 35 PDX models after 5-days treatment with the selective ER degrader fulvestrant, RAD140 or dihydrotestosterone (DHT). Protein levels of AR, ER and Ki67 were assessed by immunohistochemistry. Modulation of cell-cycle proteins by AR agonists was evaluated by Western blot. In vivo antitumor activity of RAD140 alone or in combination with palbociclib was tested in comparison to fulvestrant plus palbociclib.

Results and Discussion

Ten out of thirty ER+/HER2- PDXs harbored ESR1-mutations or fusions and were hormone-independent in vivo. PD experiments showed a Ki67-antiproliferative response with both RAD140 and DHT in 10 out of 35 models. We confirmed sensitivity to RAD140 monotherapy in 4 out of 7 Ki67-responsive models tested. Interestingly, AR agonists decreased E2F1, cyclin A/E2 protein levels in RAD140-sensitive but not resistant-PDX. RAD140 showed equal antitumor activity than fulvestrant when combined with palbociclib in 7 out of 8 models tested and superior activity in an ESR1-mutant model.

Conclusion

RAD140 exhibits antitumor activity in hormone-independent and ESR1-altered ER+/HER2- PDX models as single agent or in combination with CDK4/6i. Mechanistically, AR agonists decrease proliferation as shown by a decrease in Ki67 and S/G2-cell cycle cyclin levels.

Tom Van Nyen^1,2, Mélanie Planque^3,4, Lara Rizzotto⁵, Wout De Wispelaere¹, Hugo M. Horlings⁶, Ben Davidson⁷, Reuven Agami², Sarah-Maria Fendt^3,4, Daniela Annibali^1,2* and Frédéric Amant^1,8,9*

¹ Gynecological Oncology Laboratory, Department of Oncology, KU Leuven and Leuven Cancer Institute (LKI), 3000 Leuven, Belgium
² Division of Oncogenomics, Oncode Institute, The Netherlands Cancer Institute, 1066 CX, Amsterdam, The Netherlands
³ Laboratory of Cellular Metabolism and Metabolic Regulation, VIB-KU Leuven Center for Cancer Biology, VIB, Herestraat 49, 3000 Leuven, Belgium
⁴ Laboratory of Cellular Metabolism and Metabolic Regulation, Department of Oncology, KU Leuven and Leuven Cancer Institute (LKI), Herestraat 49, 3000 Leuven, Belgium
⁵ TRACE PDX Platform, Department of Oncology, KU Leuven and Leuven Cancer Institute (LKI), 3000 Leuven, Belgium
⁶ Division of Molecular Pathology, The Netherlands Cancer Institute, 1066CX, Amsterdam, the Netherlands
⁷ University of Oslo, Faculty of Medicine, Institute of Clinical Medicine, N-0316 Oslo, Norway; Department of Pathology, Oslo University Hospital, Norwegian Radium Hospital, N-0310 Oslo, Norway
⁸ Department of Obstetrics and Gynecology, University Hospitals Leuven and Department of Oncology, 3000 Leuven, Belgium
⁹ Centre for Gynecologic Oncology Amsterdam (CGOA), Antoni Van Leeuwenhoek-Netherlands Cancer Institute (AvL-NKI), University Medical Center (UMC), Amsterdam, The Netherlands

Introduction

Resistance to platinum-based chemotherapy represents a major clinical and societal challenge for the management of many cancer patients, particularly those with epithelial ovarian cancer. Although the majority initially respond, a high proportion experiences several response-relapse-response events during the course of the disease, until tumors become resistant. Despite improved knowledge of the molecular determinants of platinum resistance, lack of clinical applicability limits exploitation of many potential targets, leaving patients with limited options. Serine biosynthesis has been linked to cancer growth and poor prognosis in various cancer types, however its role in platinum-resistant ovarian cancer has not been investigated before.

Material and Methods

Phosphoglycerate dehydrogenase (PHGDH) expression has been determined in matched biopsies from ovarian cancer patients collected longitudinally at diagnosis and after platinum treatment. Metabolomic and molecular analyses have been performed on the isogenic ovarian cancer cell lines A2780 (wt/cis). Results have been confirmed in sensitive, resistant and matched ovarian cancer PDX models and PDX-derived organoid cultures.

Results and Discussion

We found that a subgroup of resistant tumors decreases phosphoglycerate dehydrogenase (PHGDH) expression at relapse after platinum-based chemotherapy. Mechanistically, we observe that this phenomenon is accompanied by a specific oxidized nicotinamide adenine dinucleotide (NAD+) regenerating phenotype, which helps tumor cells in sustaining Poly (ADP-ribose) polymerase (PARP) activity under platinum treatment. Consequently, combining carboplatin and PARP or NAD+ synthesis inhibitors affects the growth of resistant models with decreased serine synthesis activity.

Conclusion

Our findings have strong potential clinical applicability, because they identify alterations in serine and NAD+ metabolism as actionable vulnerabilities in a subgroup of platinum resistant ovarian cancers, and provide a rationale to test novel combinatorial therapeutic approaches to target resistance.