Oncology – Target Validation
Contents1. New tools for target validation to improve drug efficacy
2. Project Description
3. Key Deliverables of the Project
4. EFPIA Participants in the Project
5. Role of EFPIA Participants in the Project
6. Indicative duration of the project
7. Indicative total in kind contribution from the EFPIA companies
8. Indicative expectations from the “Applicant Consortium”
In the Efficacy Pillar the areas of cancer, infectious diseases and inflammation are seen as priorities for this year.
In Oncology the focus is foreseen on the following fields:
- The development, evaluation and qualification of imaging biomarkers of tumor cell proliferation and death, and of the invasive phenotype is one area with the focus to create a network of imaging centers allowing clinical validation of imaging biomarkers across multiple sites.
- Another field in the area of cancer is the search for new tools for target validation to improve drug efficacy, including improved models and integrated bioinformatics to generate testable hypotheses (systems biology).
- A third field are molecular biomarkers for the acceleration of cancer therapy development and refining of patient care focusing on the characterization of predictive, prognostic and pharmacodynamic biomarkers and the standardization of analytical methods and data retention and sharing.
There is a huge cancer burden in Europe. In 2006 there were an estimated 3,191,600 new cases of cancer diagnosed, and 1,703,000 deaths from cancer (Ferlay J., et al., Annals of Oncology, 18: 581-592, 2007). While it is widely recognised that major advances have been made both in the understanding of the disease and also in the treatment of many forms of cancer, a large number of anticancer drugs still fail due to a lack of efficacy in late stage (post-phase IIb) trials. There are a variety of reasons for these failures. One major reason is considered to be the failure to appropriately validate potential drug targets at the start of the drug discovery process.
Improvements in our understanding of the underlying biology of cancer and the development of new models for target validation is essential to support the significant advances required to improve the quality of this first key phase of drug discovery. Historically targets have been inappropriately selected or validated due to using reductionist models which do not represent the complexity of tumours in situ, which led to failure in the clinical setting. In order to improve the likelihood of success it is essential to better validate potential drug targets by:
1. Improving in vitro models of the human disease, through the development of complex, reproducible and robust models that more closely mimic the cellular organisation of tumours (e.g. in three dimensions) and the cellular heterogeneity within human malignancies
2. Cross validating, in a reciprocal way, these novels, complex in vitro models against relevant in vivo models which more closely reflect characteristics of human cancer pathology, particularly tumours arising in transgenic mice.
3. Using a systems biology-based approach to integrate and compare ‘omics data derived from the novel models and the public databases, to generate testable in silico models of the biochemical circuitry associated with potential drug targets.
The need for a collaborative approach
There are already significant efforts ongoing within academia and industry to address the development and validation of novel models to support target identification and validation. However, these efforts are often fragmentary, for example with respect to expertise within in vitro cell biology and in vivo models of cancer, and they have often lacked a pharmacological perspective. IMI offers the opportunity to integrate the innovative thinking and resources of independent experts in this field to significantly advance this field on a scale greater than the sum of each individuals own efforts. The opportunity for pre-clinical scientists from academia, SMEs and industry to come together to share already existing data and expertise to tackle these areas of pre- competitive research more efficiently is unprecedented in this arena. The focus is to develop transferable platforms to increase the efficiency with which effective medicines can be delivered to a patient population most likely to benefit from treatment. In addition to efforts conducted within the framework of this project, dissemination through such consortium will benefit the global scientific community.
The development of improved in vitro and in vivo models to support target identification and target validation with greater predictive capacity to the human disease.
Package 1: Development of complex in vitro cellular models for the validation of potential drug targets and their cross-validation with well characterized in vivo models of pathology (80% of the resources should be applied against this package)
In vitro and in vivo models are essential to the initiation of drug discovery process as they are used to validate new therapeutic targets and rank novel therapeutic agents with respect to efficacy prior to progression to other preclinical studies. However, cancers are often highly heterogeneous at the molecular and cellular level and display variable clinical responses to therapies. The complex context in which a potential drug target operates within a biochemical circuitry requires that in vitro models for target validation reflect, as closely as possible, the three dimensional organisation of tumours and aspects of their cellular heterogeneity, for example with respect to host-tumour cell interactions. Whilst there are some prior examples of the successful use of predictive reductionist models for target validation, there are many instances where hypotheses supported by simplistic, reductionist cellular models (such as homogeneous cell lines growing as monolayers on plastic) have failed in the clinic.
Recent advances in this field include the development of:
(i) 3D cultures on extra cellular matrices
(ii) Primary short term explants
(iii) Tissue slices
The material for these assays has also been derived from relevant animal models – see below. These models clearly advance closer to the human disease state, however they require further characterisation and validation before they can be used in a decision making capacity during the drug discovery process.
Complex in vitro models representing the cellular heterogeneity of human tumours should permit analyses of the heterogeneity of response following target inhibition. Recently, although somewhat controversially, it has been suggested that residual disease may be due to cells expressing some properties of normal stem cells. The development of complex in vitro models to permit analysis of a heterogeneous response to target inhibition, including analysis of molecular markers representative of a stem cell signature, would be pertinent.
In developing an in vitro model(s) suitable to support target validation and later compound ranking, it is essential to consider aspects such as the relevance to the disease, validation with a variety of mechanistic agents and gold standards, stability, robustness, reproducibility and manipulability of models. Therefore validation with challenges such as the selective use of RNA interference techniques and the use of drugs and chemicals should also be addressed. Innovative methods to permit ‘omics analyses of the tumour cells grown in complex models may have to be developed.
Early validation of potential drug targets using complex in vitro models requires comparison with appropriate in vivo models, where the aspects of complexity provided by host-tumour interactions are present. The in vivo models should have characteristics closely representative of human tumours, for example defined by histology and ‘omics methods. A dynamic reciprocity of investigations between in vivo and in vitro models is an essential element of the programme.
A major issue is the availability of in vivo models that 1) represent the complexity of the human disease, 2) display key molecular genotypes and phenotypes closely reflecting the disease, 3) support investigation of host-tumour interactions and 4) which enable the successful translation of hypotheses from pre-clinical drug discovery into man.
Transgenic mouse models of cancer, and some fresh tumour implant models to immune-deprived mice, are emerging which may fulfil this role more appropriately than in vivo xenograft models.
(i) Transgenic mouse models. As new transgenic mouse models emerge which more closely represent the histopathology, molecular pathology and other characteristics (e.g. patterns of metastases) of the major human tumour pathologies, their utility in target validation should be examined. The use of transgenic models for large scale pharmacological testing is recognised as largely impractical but their use in proof of principle experiments for drug testing and in establishing primary cultures and/or tissue slices and fragments is attractive.
(ii) Primary human tumour explants. These have been investigated but their advantage offered over xenograft models is not yet fully understood, and their success is likely to depend on further validation, in the same ways as described for transgenic models.
Applicants are invited to bring forward innovative, complex in vitro and complementary in vivo models which address the key objective of target validation, balancing complexity with the practical requirements necessary to support novel drug discovery programmes at the target validation stage.
Package 2: Integrated bioinformatics of multivariate data in order to generate testable hypotheses (20% of effort).
The data accumulated from genomic and proteomic analyses of clinical human tumour samples, held in both public and industrial hands, should be used to validate the models described above. In addition, the data generated in the models themselves should permit the modelling and testing of the interactions between multiple signalling pathways, the activity of transcription factors, changes in intermediary metabolism and the impact of host-tumour interactions on the molecular circuitry of normal and tumour cells. This should provide hypotheses regarding the nature of nodal points that drive malignancy and the position of a potential drug target in this circuitry. Indicators suggesting those proteins or pathways to which certain malignant pathologies become “addicted”in order to survive, proliferate and metastasise, will suggest strategies of intervention. Testable hypotheses may also emerge implicating “synthetic lethal” strategies, targeting more than one locus in order to collapse a network unique to a tumour cell. Clearly validation of such an approach will be required using appropriate cellular models where perturbation of key pathways by chemical tools, dominant negatives or RNAi strategies is relatively facile.
This in silico work will better describe tumour signalling networks as they exist in complex model systems and should lead to improved drug targeting strategies. Additionally, it has the potential to indicate potential mechanism of drug resistance because of redundancies in pathways, and to improve the alignment of tumour models to human disease. It should better develop patient stratification hypotheses in oncology by generating ideas regarding potential biomarkers of drug sensitivity.
The key project goal is to develop robust tools and approaches to improve target validation and thus the attrition profile for oncology products, preventing unnecessary progression of targets with a low chance of success.
Package 1: Development of novel in vitro and in vivo models of disease with improved predictive capacity for target validation
- Development and validation of the next generation of in vitro and in vivo models with greater predictive capacity for the clinic.
- Alignment of models with molecular profiles obtained from high quality human tumour samples to assess their relevance and applicability.
- Establishment of the limits of manipulability of the new, complex model systems for target validation and drug testing using new methods and technologies.
Package 2: Integrated bioinformatics of multivariate data in order to generate testable hypotheses (20% of effort)
- Systems biology descriptions of tumour biochemical circuitry in novel models (compared to human tumours) capable of describing the context of novel targets and generating hypotheses to be tested in models of target validation
- Ability to more effectively identify and validate targets and to identify successful drug candidates reducing the number of studies required and increasing probability of success.
- Ability to design smaller, stratified clinical studies which deliver early signals of efficacy
- Access to validated standardised models across Europe
- Integrated approach to target validation and possible patient stratification across academia and industry
- Access to a well managed extensive sample (and data) collection of tumour tissue from animal models and patients
- Standardised and validated protocols and data analysis methods across sites.
AstraZeneca, Bayer Healthcare, Boehringer-Ingelheim, Novartis, Orion, Pfizer, F. Hoffmann - La Roche AG, Servier, Sigma-Tau, Wyeth. in the Project
The EFPIA participants will contribute:
- Pre-clinical models including cell lines, transgenic animals and associated ‘omics data.
- Pharmacology data from pre-clinical studies.
- Supplies of NCEs and marketed compounds.
- Biochemical assays that could be replicated for model development and validation and know-how in assay development (including high technology platforms such a High Content Screening imaging).
- Know-how on the development of complex cellular models.
- Tumour and surrogate tissue samples and the associated clinical data. Data from clinical studies exploring efficacy endpoints using both single agents and combinations and either NCEs or marketed drugs.
- Omic data on specific human pathologies.
- Clinical supplies of registered agents.
- Regulatory authority contacts and interactions.
- Active participation by working with the applicant consortium and supervising the EFPIA participant funded positions accordingly to achieve the key deliverables.
- Know-how in statistical analysis of genomic and clinical study data.
- General preclinical and clinical oncology expertise.
- Expertise in bioinformatics, systems biology and algorithms for modelling the perturbation of complex systems
- Know-how in sample and data management.
The indicative duration of the project is 5 years.
The provisional estimate for the EFPIA in-kind contribution to this project is EUR 8 million.
The Applicant Consortium should aim to bring forward innovative approaches to address all the major objectives outlined in the call. In summary these are :
To develop high quality and complimentary platforms to support improved target identification, target validation and potential patient stratification by developing novel in vitro and in vivo pre-clinical models reflecting the complexity and heterogeneity of human tumours, demonstration of the models’ equivalence to human pathology and demonstrating their performance in target validation using appropriate positive negative controls.
Establish a systems biology description of tumours and their models which are capable of validating and identifying novel targets and generating hypotheses to be tested in models of target validation.