The Project


Current methods used in drug development lead to a large number of drugs not reaching the market due to the late discovery of cancerous effects during pre-clinical trials. Such cancers are not the result of direct genetic toxicity, as drugs that exhibit such characteristics are eliminated from consideration early in their development. They can however result from changes in gene expression or cellular phenotypes caused by mechanisms other than changes in the underlying DNA sequence. Drugs that induce such cancers are called non-genotoxic carcinogens (NGC).

At present, due to a lack of validated short-term study techniques such non-genotoxic, cancer causing drugs tend to be identified following prolonged pre-clinical studies. Using the liver, the major target organ for such drug-induced tumours, the MARCAR project aims to advance our understanding of the availability of early "biomarkers" of NCG action to help in the design of more predictive short term assessment tools to reduce the requirement for costly long-term biological testing and help to deliver safer and faster medicines to patients.


Identification of biomarkers will be achieved using techniques that, at a molecular level, can identify changes occurring as a result of exposure to drug compounds. This analysis will predict the induction of cancer in well-developed pre-clinical models similar to those currently used by pharmaceutical companies for identifying genotoxic carcinogens. This study will generate a profile of the early changes occurring in cells after exposure to known NGC. From these profiles it is anticipated that novel markers will be identified which predict drug-induced alterations which precede and predict the eventual manifestation of tumours.

State-of-the-art analytical techniques will be used to identify the cause of precancerous changes and tumours that develop as a result of exposure to NGC. MARCAR will carry out studies to find biomarkers and to identify tumours which arise spontaneously and distinguish them from those which arise as a result of exposure to an NGC. To analyse the validity of the biomarkers the results from these studies will be collated using novel bioinformatics approaches and standardised data management. The output will identify novel candidate biomarkers that can identify the causes of cancer development. These biomarkers will then be incorporated into new biological testing systems that will test their reliability in predicting the cause and onset of drug induced-cancer.

One of the main drivers behind finding early biomarkers for risk assessment is the potential to reduce animal use in drug and chemical risk assessment, i.e. the NC3Rs directive on the reduction/refinement/replacement of experimental animal use. MARCAR aims to tackle this issue by developing a method for conducting reliable short term studies which will reduce the need for long-term biological testing, and by exploring the use of non-invasive imaging techniques to monitor precancerous changes and tumor development.

Reprinted by permission from Macmillan Publishers Ltd: NATURE (Qui.J NATURE 441, 143-145), copyright 2006Objectives:

The main objectives of the project are to:

Identify early biomarkers for more reliably predicting which compounds have a potential for later cancer development

Improve the scientific basis for assessing carcinogenic potential of non-genotoxic drugs

Identify the molecular response to NGC exposure that underpins development of early exposure biomarkers

Improve drug safety and the efficiency of drug development by progressing the development of alternative research methods (such as the "3Rs" concept: reduction, refinement and replacement of animal experimentation).


The main impact of the MARCAR project will be to establish, for the first time, proof of concept that early biomarkers can reliably and robustly predict later cancer development. Validation of the biomarkers identified in the initial experiments will be a key stage in the project as early biomarkers of carcinogenicity would be immensely valuable in preclinical development of new drug compounds.

The identification of these early biomarkers will not only allow for the reduction/refinement/replacement of experimental animal use, they will also be of benefit to pharmaceutical companies through much earlier de-selection of potential drug compounds which are carcinogenic with consequent saving of time effort and money spent on promising drug candidates which eventually turn out to be carcinogenic. It will also prevent the futile efforts being pursued so that drugs that work can be developed. It will also mean fewer delays and adverse effects during late-phase drug development and much improved preclinical carcinogenicity safety assessment prior to clinical trials. Translation of early cancer biomarkers into the clinic would also improve safety for patients participating in clinical trials.



  • Identification of the interactions between genes during tumour development following exposure to a prototypical NGC. By understanding the interactions between genes key regulators of tumour development can be identified and robustly modelled in mathematical models to provide novel insights into liver cancer (Luisier et al, 2014)

  • Identification of modifications in DNA, following drug exposure, that correlate to changes in gene expression. Ultimately such work has allowed MARCAR to predict possible early biomarkers for the early stage progression of NGC driven cancer (Thomson et al 2013)

  • Development and refinement of techniques for identifying changes in DNA at a molecular level. These techniques will now be incorporated into assays for the identification of DNA modification as a result of NGC exposure (Thomson et al 2013)

  • Comprehensive analysis of the genetic mutations in liver tumours. Analysis of these tumours has revealed in some detail the molecular mechanisms underlying NGC driven mutations and a number of biomarkers which are potentially suited to discriminate spontaneous liver tumours from liver tumours induced by a known NGC have been identified.

  • Comparative analysis of the epigenetic and transcriptomic profiles between DEN/PB induced liver tumours to normal and 12wk PB treated livers. Analysis of these datasets complements the molecular mechanisms identified in NGC driven mutation studies and strengthens the selection of biomarkers which are suited to discriminate between early and late exposure to a typical NGC

  • Completion of a study into the effects of a typical NGC on liver cells. The study showed that NGC alter the functions of liver cells and cause over expression of specific genes which regulate cell processes. These genes are now being investigated as potential early biomarkers for NGC action (Klepeisz et al 2013)

  • Generation of models to assess the dependency of human nuclear receptors in cancer development after exposure to and recovery from an NGC. These models will be used to define pathways of gene regulation through the action of specific nuclear receptors and determine the extent to which they are involved in the mechanism of action of NGCs. Importantly after recovery from NGC exposure the persistence of identified biomarkers (Expression, epigenetic, proteomic and metabolomics) will be assessed using these models.

  • Substantial progress towards enabling development of safer drugs by identifying specific genes that may be able to predict potential cancer inducing properties of drugs as they progress through the drug development pipeline (Lempianen et al 2013)

  • Implementation of state of the art molecular techniques to identify changes in liver which are likely to induce cancer after exposure to drug compounds. This includes discovery of changes happening in specific parts of DNA which represent fingerprints of exposure to NGC compounds (Thomson et al 2012)

  • Development of new tools with the capacity to automatically map the relationships between biological processes at a molecular level. The tool has already been made available to the wider research community via open access (Wrodek et 2013)

  • Successful determination of the mutation status of tumors exposed to the model NGC Phenobarbital (PB). The results indicate over expression of specific genes in tumors resulting from NGC exposure. Identification of these genes may act as markers to determine if tumors have occurred as a result of exposure to an NCG or if they have occurred spontaneously.

  • Evidence that model NGC have a direct inflammatory effect on liver cells, which may favour the development of cancerous cells. These results have led to the current hypothesis that genes involved in growth selection of pre cancerous liver cells within an NGC-stressed liver, may serve as candidate biomarkers predicting the carcinogenic activity of NGC in model systems.