Molecular Mediators of Mammographic Density

When breast tissue undergoes x-ray (i.e. a mammogram), the fatty component appears black whilst the areas containing the glandular tissue appear white. Mammograms with a large proportion of white are said to have high mammographic density (high MD). This is largely due to large amounts of supporting material in the breast called stroma.
Over recent years it has become apparent that having high MD confers a significant increased risk for the development of breast cancer, although the mechanism which confers this increased risk is not understood. High MD is a common finding in the mammograms of women in the UK and globally. Therefore, interventions targeted towards reducing MD have the potential to benefit large numbers of women.
Recently it has been shown that the anti-oestrogen drug tamoxifen can reduce the risk of developing breast cancer when given to women with a strong family history. This reduction in risk corresponds to a reduction in MD. In addition, those women who show no reduction in breast density with tamoxifen treatment show no associated reduction in breast cancer risk, suggesting that the protective effect of this drug is mediated through reduction of MD.
This project aims to build on these key observations to investigate how MD can be modulated. Thus we aim to treat breast stromal cells isolated from women with high MD with tamoxifen and analyse the pathways that are altered by treatment and that may therefore be involved in controlling MD.
At Barts Cancer Institute we routinely consent patients undergoing surgery for donation of breast tissue not needed for diagnosis to be used in research. We will use these samples in our research to investigate the mechanisms by which tamoxifen exerts its protective effect. Stromal cells will be treated with tamoxifen and the effect on a series of molecules potentially involved will be assessed. We also will carry out a global analysis on the treated cells as this could discover alterations that had not previously been considered. We will test any molecules identified for their involvement in controlling stromal function and MD, and generate a 'signature' of the most important and relevant molecules that will then be tested on biopsies taken from women who have or have not been treated with tamoxifen, as part of preventive clinical trials. The work will be carried out by a trainee pathologist who already has experience of dealing with tissue samples and therefore is ideally placed to undertake the project.
The importance of this project is that by understanding the pathways contributing to high MD we will better understand the factors leading to development of breast cancer and this will help devise new ways of preventing it. Therefore we anticipate the results of this research will particularly benefit women who are at high risk of developing breast cancer.
The results from this research may also assist doctors in predicting which of these high risk women will respond well to treatments such as tamoxifen, that are aimed at reducing MD, and critically, identify those women who will not respond. This will mean that they will only be prescribed to women who are likely to have a good response to the treatment. Those that are unlikely to respond well will be spared some of the treatment associated side effects.
This work has relevance beyond the field of breast cancer since the mechanisms leading to high MD are likely to have much in common with other diseases characterised by excessive accumulation of stroma - a process known as fibrosis. Therefore the findings from this research may also result in new treatments and preventative strategies for these patients suffering from these diseases.

Aim:
To identify pathways modified by tamoxifen to better understand the biology of mammographic density (MD), provide markers to predict a protective response to tamoxifen and ultimately identify new, potentially targetable, molecules that could be used to reduce breast cancer risk.

Objectives:
(i) To use in-vitro models of primary human breast fibroblasts to dissect the effect of tamoxifen on stromal cell function, using a candidate gene and genome-wide approach.
Primary fibroblasts will be isolated from human breast tissue of high MD and treated with tamoxifen (+/- oestradiol), or vehicle control. Fibroblast function will be analysed by: measuring proliferation; analysing key candidate molecules including LOX enzymes, collagen I, fibronectin and SMA by RT-PCR and western blotting (WB); TGF-beta expression and signalling by ELISA and WB; evidence of adipocyte differentiation (expression of leptin and CD36), and the expression and role of ER receptors.
A genome-wide approach will also be taken using RNA-Seq (Illumina HiSeq facility) with a minimum of 5 matched treated: control samples.

(ii) Establish the impact of the biological pathways identified in (i) on fibroblast-epithelial cross-talk using co-culture systems
2D and 3D systems will be used to assess the functional significance of pathways identified on normal and pre-malignant epithelial cell proliferation and invasion using knock-down and antibody inhibition approaches.

(iii) Assess the power of the biological readout generated in (i) and (ii) to predict response to tamoxifen in terms of 1) change in MD and 2) risk of breast cancer.
We will access tissue from the IBIS-1 Preventive and MD Trial and the Manchester Prevention Study, which provide data on MD response and breast cancer risk.

The project will provide broad laboratory experience with exposure to bioinformatics analysis and the results generated may identify new markers of tamoxifen response and novel targets for prevention

The findings of this research have the potential to have a major impact on the management of individuals at high risk of developing breast cancer and could result in changes in the organisation of screening programmes in the UK and abroad. In addition, these findings have the potential to impact outside the breast cancer field, in the management of non-malignant fibrotic diseases.
High mammographic density (MD) is a major independent risk factor for breast cancer, highly heritable and common globally. Thus interventions targeted at reducing MD have the potential to have a major impact on the health of millions of women. Gaining an improved understanding of the biological pathways contributing to the density-associated risk will allow researchers to make significant progress in understanding the genetic basis and epidemiology of MD, and identify populations of women at particular risk of developing breast cancer and those likely to respond to risk reducing treatments such as tamoxifen.
Such insights will not only have a profound impact for individuals responsible for managing high risk women but will also inform NHS and government health policy-makers, within the UK and abroad, and may result in changes in breast cancer screening programmes. The results from this work could make a significant contribution to the PROCAS study (which our collaborators, Stephen Duffy and Anthony Howell, are involved with) which aims to predict the risk of developing breast cancer at screening.
Measurement of MD may form an integral part of early breast cancer screening and personalised schedules may be developed based on an individual's perceived risk. Thus women with high MD may warrant a higher frequency of screening. Individualised screening programmes based on percentage MD and common genetic variations have already been demonstrated to be effective by a group in Sweden [1].
NHS policy makers such as the National Institute for Clinical Excellence in the UK, may recommend that women with particularly high MD should benefit from prophylactic treatment with tamoxifen in order to reduce their breast cancer risk. Being able to predict whether high risk women will be protected by treatments such as tamoxifen will be an important component of these strategies. Therefore, it will be crucial for women in the UK to be aware of the importance of MD measurement and potential treatment implications of the result. Thus GP practices and breast cancer charities may also be indirectly affected by our research, as they will have an important role in engaging women and counselling them about MD measurement.
As well as improving our ability to predict a woman's risk of breast cancer, understanding the way in which tamoxifen acts to reduce MD could have a profound impact on the development of new preventative treatments and the rationalisation of such treatments only to those likely to have a good response.
The findings from our genome-wide investigation into the stromal response to tamoxifen will highlight key genes involved in mediating this response. Thus it may be possible to predict individuals who will respond based on their gene expression. This could mean that individuals who are deemed unlikely to respond based on their genetic 'signature' being offered alternative treatments, saving resources and sparing them undesirable side effects associated with tamoxifen treatment. Drug companies in the commercial sector may seek to exploit these findings to develop novel treatments which can reduce MD via similar mechanisms but potentially have a better side effect profile.
The insights from our research into how tamoxifen modifies pro-fibrogenic pathways may also have the potential to impact on the management of patients with non-malignant fibrotic disorders, in terms of identifying those likely to respond to treatment and the development of new, more effective treatments for these conditions.

1. Darabi H et al, Breast Cancer Res 2012 7;14(1):R25