Biomarkers and molecular tumour classification for non-genotoxic carcinogenesis
Surrogate markers for micro- and macro-vascular hard endpoints For innovative di


Start Date
End Date
IMI1 - Call 1
Grant agreement number

IMI Funding
6 049 578
EFPIA in kind
5 155 604
1 867 557
Total Cost13 072 739


One of the key questions scientists ask during early drug development is: could a potential drug cause cancer? MARCAR project scientists set out to discover early biological indicators which could help detect some of the more indirect ways in which drugs cause tumour formation. During the project, they discovered several potential biological indicators and a wealth of knowledge which could make drugs safer and decrease the time it takes for innovative drugs to reach patients. Additionally, one of the imaging techniques developed during the project has the potential to significantly reduce the number of animals used in the early stages of drug development.

When it comes to the development of innovative medicines, one of the main challenges is cancer risk assessment, i.e. ensuring that the potential drug does not cause cancer. There are two ways in which drugs can induce cancer: by directly altering the structure of the DNA and causing genetic mutations (genotoxic), and by various other indirect mechanisms which are harder to identify (non-genotoxic). While genotoxic changes are quick and generally easy to detect during the early stages of the drug development process, non-genotoxic changes require more intensive studies that utilize a large number of laboratory animals (approximately 500 per drug), and carry a very significant financial burden (about €3 million per drug). Due to this long and laborious process, many developed compounds either do not reach the market, or carry drug label warnings for potential carcinogenic risk due to the observation of cancerous effects in preclinical laboratory (e.g. animal) studies.

The goal of the MARCAR project was to develop early biological indicators (biomarkers), for early detection of non-genotoxic carcinogenesis. This type of cancer formation is thought to be facilitated by ‘epigenetic’ changes, i.e. modifications of the DNA and proteins that surround and functionally organize the DNA in our cells. Changes in these molecules can alter the 'readability' of the DNA and so affect the activity of our genes. The MARCAR consortium set out to shed more light on these epigenetic effects, using a combination of novel and sophisticated molecular technologies. Combining their expertise in the field of biomarkers, human and rodent cancer models, imaging, molecular profiling and bioinformatics, the researchers initially focused on liver tumours, an organ that is frequently affected by non-genotoxic carcinogenesis during the preclinical safety evaluation of medicines.

Project success: New potential epigenetic biomarkers identified

Early efforts in MARCAR focused on optimising methods and technologies which allow scientists to look into the relationship between drug exposure and measurable epigenetic changes in liver cells transitioning to a tumour. Once a methodology was established, the consortium partners dug deeper to determine which specific epigenetic changes were linked with tumour growth. The goal was to identify early epigenetic biomarkers which could signal that a specific drug might cause non-genotoxic changes, and ultimately create an environment in which cancer occurs. At least two classes of such potential biomarkers were identified.

Early in the project, MARCAR scientists discovered a promising early RNA biomarker called Meg3 that originates from a very special epigenetically-regulated region of the genome. Progressive activation of Meg3 by non-genotoxic carcinogens in mice is believed to reflect reprogramming of normal adult liver cells towards a stem cell-like pluripotent state, a change which might contribute to the development of cancer in the longer-term. This biomarker is currently being evaluated with a broader range of drugs and has the potential to explain the molecular basis of species differences in responses to non-genotoxic carcinogens, a key consideration for interpreting human relevance of animal studies.

Another very important clue came from looking at the DNA bases or letters that make up the genetic code. Although basic biology teaches us that there are four basic DNA bases (A, T, C and G), multiple epigenetic modifications of the C (“cytosine”) base dramatically increase the complexity of our genes. MARCAR project scientists focused on a recently discovered modified form of cytosine called 5-hydroxymethylcytosine which is epigenetic in nature because it is involved in regulating the activity of genes, i.e. switching them on and off. They were able to show that this modified base is altered in very specific patterns by exposure to drugs which are known to cause liver cancer in rodents. On this basis, the scientists postulated that 5-hydroxymethylcytosine represents a powerful early biomarker for predicting tumour growth.

Both of these novel biomarker discoveries have the potential to underpin the design of improved preclinical safety studies that should ultimately reduce the cost, and accelerate the development of, innovative medicines.

Mouse MRI to reduce number of animals used in testing

In pre-clinical trials, in order to track tumour growth in animals, scientists have to look at the tissue directly, and this means sacrificing the animal. MARCAR project scientists developed a new method which is non-invasive and thus dramatically reduces the number of animals that need to be used for mechanistic carcinogenicity studies. The method involves a special magnetic resonance imaging (MRI) technique which allows researchers to scan mice and locate tiny tumours early in the screening process that are just 1 mm in size. They can then track the subsequent growth of the tumours by rescanning the same animals instead of having to sacrifice them every time. The method holds great value in reducing the number of animals used in drug development.  

Innovation in animal models: a mouse with a human receptor

Another innovation focused on a well-characterised epilepsy drug, which causes cancer in mice, but not in humans. In order to shed more light on this discrepancy, MARCAR scientists inserted a human receptor for this drug into the body of a mouse, creating a ‘humanised mouse’ model. They wanted to see if the human receptor would work in the body of the mouse, and if it would lead to the formation of cancer; it did, but with a reduced incidence. Does that mean that the drug is carcinogenic for humans after all? It’s not so simple because the tumour growth still occurred in the body of a mouse. However, the use of this humanised mouse model could be useful in the further study of the mechanisms of how cancer develops.

For the benefit of the industry, patients and academia

The science, technology and models developed during the MARCAR project have the potential to strengthen the scientific rigour and strength of cancer risk assessment; improve drug safety; reduce animal use and reduce the time it takes for new innovative medicines reach patients. Furthermore, the biobank and database created during the project will be assets in future research.

What next?

Now that the project has drawn to a close, one of the priorities will be to share the knowledge gained with the global community. This will be done in a multitude of conferences and workshops.  

Read the interview with project coordinator

Achievements & News

MARCAR demonstrates new way of tracking tumour development over time
February 2012

IMI project MARCAR has demonstrated that magnetic resonance imaging (MRI) can be used to reliably detect liver tumours in mice when they are just 1 mm across – previously more invasive techniques were needed to pick up tumours of this size.### The findings, published in the journal Toxicological Sciences, are important for two reasons. Firstly, because MRIs are non-invasive, they can be repeated at different stages of the study, meaning that fewer animals are needed to obtain reliable results. This will therefore help to reduce the number of animals used in experiments. Secondly, the fact MRIs can be used to detect tumours at an early stage and monitor their reversibility makes them an invaluable tool in assessing the cancer risk of potential drugs. Looking to the future, the work will ultimately help MARCAR achieve its goal of discriminating between spontaneously occurring liver tumours, and tumours that represent the unwanted side effect of a drug candidate.

MARCAR tracks down biomarkers
April 2011

The goal of the MARCAR project is to identify early biological indicators (biomarkers) that can be used to predict the effects of non-genotoxic carcinogens (NGCs).### Now scientists from the MARCAR consortium have used genome-wide profiling of changes to DNA methylation to detect early markers of NGC activity in rodents. The researchers have published their findings in the journal PLoS ONE and write: ‘This study contributes to understanding the scale and nature of drug-induced epigenetic changes in an in vivo setup relevant for drug safety assessment.’
More on MARCAR:

MARCAR project results will contribute to making drugs safer for patients
Cancer risk assessment is one of the most important steps during the development of new medicines. Launched in 2010, IMI’s MARCAR project aimed to develop early biological clues – biomarkers – which could help predict which drugs might lead to tumour growth. ### In an interview with the IMI Programme Office at the end of the project, MARCAR project coordinator Jonathan Moggs of Novartis spoke about the main project achievements and explained how MARCAR project outputs will make drugs safer in the future. ‘MARCAR contributed to the scientific rigour and strength of the cancer risk assessment, by which we will ensure that safer medicines reach patients, and in some cases accelerate the time it takes for those medicines to reach patients,’ said Moggs. ‘What causes delays is having an unexpected tumour finding in pre-clinical animal studies that we cannot easily explain; this prevents us from having a good way of making an accurate cancer risk assessment for humans. MARCAR’s science, technology and models should enable those kinds of questions to be answered more quickly and more comprehensively.’
 - Read the full interview
 - Find out more about the project's achievements in the project factsheet
 - Visit the project website
(January 2017)

Participants Show participants on map

EFPIA companies
  • Bayer AG, Berlin, Germany
  • Boehringer Ingelheim International GmbH, Ingelheim, Germany
  • H. Lundbeck A/S, Valby, Denmark
  • Novartis Pharma AG, Basel, Switzerland
  • UCB Biopharma SPRL, Brussels, Belgium
Universities, research organisations, public bodies, non-profit groups
  • Eberhard Karls Universitaet Tuebingen, Tuebingen, Germany
  • Institut National De La Sante Et De La Recherche Medicale, Paris, France
  • Medizinische Universitaet Wien, Vienna, Austria
  • Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen, Reutlingen, Germany
  • University of Dundee, Dundee, United Kingdom
  • University of Edinburgh, Edinburgh, United Kingdom
Small and medium-sized enterprises (SMEs)
  • CXR Biosciences Limited, Dundee, United Kingdom


Project coordinator
jonathan Moggs
Novartis Pharma AG
Managing entity
Roland Wolf
University of Dundee