Breast cancer screening is ripe for change. We need to assess a woman’s risk – not just her age
- Written by The Conversation
Australia’s BreastScreen program offers women regular mammograms (breast X-rays) based on their age. And this screening for breast cancer saves lives.
But much has changed since the program was introduced in the early 90s. Technology has developed, as has our knowledge of which groups of women might be at higher risk of breast cancer. So how we screen women for breast cancer needs to adapt.
In a recent paper, we’ve proposed a fundamental shift away from an age-based approach to a screening program that takes into account women’s risk of breast cancer.
We argue we could save more lives if screening tests and schedules were personalised based on someone’s risk.
We don’t yet know exactly how this might work in practice. We need to consult with all parties involved, including health professionals, government and women, and we need to begin Australian trials.
But here’s why we need to rethink how we screen for breast cancer in Australia.
Why does breast screening need to change?
Australia’s BreastScreen program was introduced in 1991 and offers women regular mammograms based on their age. Women aged 50–74 are targeted, but screening is available from the age of 40.
The program is key to Australia’s efforts to reduce the burden of breast cancer, providing more than a million screens each year.
Women who attend BreastScreen reduce their risk of dying from breast cancer by 49% on average.
Breast screening saves lives because it makes a big difference to find breast cancers early, before they spread to other parts of the body.
Despite this, around 75,000 Australian women are expected to die from breast cancer over the next 20 years if we continue with current approaches to breast cancer screening and management.
Who’s at high risk, and how best to target them?
International evidence confirms it is possible to identify groups of women at higher risk of breast cancer. These include:
women with denser breasts (where there’s more glandular and fibrous tissue than fatty tissue in the breasts) are more likely to develop breast cancer, and their cancers are harder to find on standard mammograms
women whose mother, sisters, grandmother or aunts have had breast or ovarian cancer, especially if there are multiple relatives and the cancers occurred at young ages
women who have been found to carry genetic mutations that lead to a higher risk of breast cancer (including women with multiple moderate risk mutations, as indicated by what’s known as a polygenic risk score).
Women in these and other high-risk groups might warrant a different form of screening. This could include screening from a younger age, screening more frequently, and offering more sensitive tests such as digital breast tomosynthesis (a 3D version of mammography), MRI or contrast-enhanced mammography (a type of mammography that uses a dye to highlight cancerous lesions).
But we don’t yet know:
how to best identify women at higher risk
which screening tests should be offered, how often and to whom
how to staff and run a risk-based screening program
how to deliver this in a cost-effective and equitable way.
The road ahead
This is what we have been working on, for Cancer Council Australia, as part of the ROSA Breast project.
This federally funded project has estimated and compared the expected outcomes and costs for a range of screening scenarios.
For each scenario we estimated the benefits (saving lives or less intense treatment) and harms (overdiagnosis and rates of investigations in women recalled for further investigation after a screening test who are found to not have breast cancer).
Of 160 potential screening scenarios we modelled, we shortlisted 19 which produced the best outcomes for women and were the most cost effective. The shortlisted scenarios tended to involve either targeted screening technologies for higher-risk women or screening technologies other than mammography for all screened women.
For example, in our estimates, making no change to the target age range or screening intervals but offering a more sensitive screening test to the 20% of women deemed to be at highest risk would save 113 lives over ten years.
Alternatively, commencing targeted screening from age 40 and offering a more sensitive screening test annually to the 20% of women at highest risk, and three-yearly screening (of the current kind) to the 30% of women at lowest risk, would save 849 lives over ten years.
However, less frequent screening of the lower risk group was expected to lead to small increases in breast cancer deaths in that group.
We also outlined 25 recommendations to put into action, and set out a five-year roadmap of how to get there. This includes:
a large scale trial to find out what is feasible, effective and affordable in Australia
making sure women at higher risk in different parts of Australia are offered suitable options regardless of where they live and who they see
better data collection and reporting to support risk-based screening
testing how we assess women for their risk of breast cancer, including whether these assessments work as intended and make sense to women from a range of backgrounds
clinical studies of screening technologies to determine the best delivery models and associated costs
ongoing engagement with groups including women, health professionals and government.
Breast cancer screening review out soon
Federal health minister Mark Butler said a review of the BreastScreen program would consider our recommendations. The results of this review are expected soon.
We’re not alone in calling for a move towards risk-based breast cancer screening. This is backed by national and international submissions to government, policy briefing documents and the Breast Cancer Network Australia.
We’ve provided an evidence-based roadmap towards better screening for breast cancer. Now is the time to commit to this journey.
We acknowledge Louiza Velentzis from the Daffodil Centre, and Paul Grogan and Deborah Bateson from the University of Sydney, who co-authored the paper mentioned in this article.
