Tailoring breast cancer screening to individual patient risk

Professor Fiona Gilbert is using funding from our Early Detection Programme Award to assess the most effective approaches for breast screening. In this story, Fiona tells us how she hopes to develop prediction models to tailor breast screening programmes to individual patient risk.

Fiona, who is Head of Department of Radiology at the University of Cambridge, is working together with Professor Paul Pharoah, from the Centre for Cancer Genetic Epidemiology. Their research is focusing on women with particularly dense breast tissue, a population whose mammograms have reduced sensitivity and who may be at increased risk of developing breast cancer.

“Mammography has been considered the gold standard for breast cancer screening, however it’s well known to be less sensitive in women with dense breast tissue.

Studies have shown that dense breast tissue can mask cancers, making mammograms more difficult to read. Cancers when they are found in these women tend to be larger.

We wish to test a range of imaging techniques to assess their feasibility for women who may be at increased risk due to dense breasts, in addition to standard mammography in a screening setting.

The screening technologies we will be exploring include Abbreviated Magnetic Resonance Imaging, a more rapid test with the key MRI images; Contrast Enhanced Mammography, which combines iodinated contrast agents with conventional mammography to identify areas of increased blood supply, which can be an indicator of cancer;  and Whole Breast Ultrasound, which involves a large curved transducer that fits comfortably over the breast to allow the entire breast to be scanned by ultrasound.

Our study is exploring whether Abbreviated Magnetic Resonance Imaging, Contrast Enhanced Mammography, or Whole Breast Ultrasound is better as a supplemental breast cancer screening test to mammography for women with dense breasts. 

Women have a wide range of breast densities – those with the densest breasts have a higher proportion of glandular and fibrous tissue, whilst those with the least dense breasts have higher relative proportions of fat.

There are two problems connected with dense breasts: dense tissue can mask or hide cancers reducing the likelihood of detection by mammographic screening, and may also increase the risk of developing breast cancer.

We want to use an integrated approach to identify which of these women would benefit from additional imaging and to try and establish the amount of density where women would benefit from additional imaging.

Study to involve six screening centres across the UK

With our recent funds from CRUK’s Early Detection Committee, we will first randomly assign upfront women undergoing breast screening at six UK centres to one of four screening arms – Abbreviated Magnetic Resonance Imaging, Contrast Enhanced Spectral Mammography, Whole Breast Ultrasound, and standard mammography.

We will then invite women found to have particularly dense breasts to take part in the study and we’ll allocate them to their upfront screening arm.

We hope to start enrolment in the summer and altogether plan to recruit 12,000 women with dense breasts over two years, with 3,000 in each arm. So far, the breast screening centres signed up for the study include Cambridge, Leeds, Manchester, Nottingham, and Cheltenham, with a sixth centre yet to be identified.

We aim to discover which of the four modalities identifies the greatest number of breast cancers in women with dense breasts.

As well as the number of cancers identified, it will also be important to take into consideration the stage of cancers found, since, to have an impact, screening tools need to identify cancers when they can be effectively treated.

As part of the study, we’ll also be testing a range of automated tools for assessing breast density to determine what the most appropriate might be for incorporating into breast screening programmes.

Since it isn’t known at which numerical score women might actually benefit from additional imaging, we want to establish which density score would serve as the cut off at which it could be effective to offer additional imaging tests.

Helping predict breast cancer risk by validating the BOADICEA prediction model

Working with joint lead Professor Paul Pharoah from the Centre for Cancer Genetic Epidemiology at the University of Cambridge, we also plan to use a subset of women from our study to validate the BOADICEA model (Breast and Ovarian Analysis of Disease Incidence and Carrier Estimation Algorithm).

The University of Cambridge developed the computer programme to calculate the risks of breast and ovarian cancer, analysing single letter variations in DNA sequence, lifestyle and hormonal risk factors, and breast density.

For a group of women in the study we’ll calculate their risk of cancer using the BOADICEA model and then validate it against the actual number of cancers detected. Ultimately, if we can show that a prediction tool works, we could use it to stratify the population to decide which women should be invited for screening.

Advice to applicants: budget carefully upfront and get policy makers on board

My advice to anyone applying for these awards is to plan your work and think upfront about your costings really carefully as you will not have the chance to amend them in the second stage of the application process.

One of the strengths of our proposal was that we’d already had the support from the NHS Breast Screening Programme who agreed to allow us to use their centres for recruiting patients.

We did lots of groundwork to make sure screening policy makers believed the modalities we were investigating could be of value. Such preparation is critical because without their support and the possibility of introducing new approaches to screening in clinical practice, it would be a complete waste of charitable money to undertake the research.

I would also recommend getting in touch with CRUK regardless of whether your research involves discovery, pre-clinical or translational science. When the CRUK early detection funding team made contact to say that they were interested in funding research in our area, it came as a bit of a surprise as we were under the impression they only funded basic science. We have since discovered that they are really supportive of clinical work. ”