Show me the Evidence
Let’s Talk About Risk: Personalizing Breast Cancer Risk Prediction
Improved genetic profiling and risk communication are enabling targeted breast-cancer prediction, detection and treatment for those who need it most
At a Glance
Who: Dr. Jacques Simard, research centre of the Centre hospitalier universitaire de Québec and Laval University
Issue: In Canada, one out of nine women will develop breast cancer during her lifetime. At present, predictive genetic testing exists only for mutations in the BRCA1 and BRCA2 genes, which account for just 5% of all breast cancers.
Projects: Dr. Simard leads the CIHR Team in Familial Risks of Breast Cancer, which includes more than two dozen Canadian and international researchers. The team, which is part of the international Collaborative Oncological Gene-environment Study (COGS), is developing a personalized risk stratification system to identify women who will most benefit from earlier screening, detection and targeted treatment of breast cancer. The team has also evaluated different strategies for sharing risk information with patients.
Research Evidence: In March 2013, COGS released the largest-ever genetic association study in cancer. The study, involving more than 100,000 women worldwide, identified 49 new genetic markers, or genetic “spelling mistakes,” related to breast cancer. In addition, the CIHR team has found that the risk communication format most commonly used by clinicians is the least preferred and understood by breast cancer patients.
Evidence in Action: The CIHR team’s risk communication research has contributed to making BOADICEA, the world’s primary risk prediction model, more patient-friendly through the incorporation of a risk curve graphic to its online risk communication package. The COGS study, meanwhile, provides the CIHR team with genetic information for the development of a new, broad-based genetic profile screening test for breast cancer, one that will enable improved personalized breast cancer screening.
Sources: Michailidou, Kyriaki, et al. “Large-scale genotyping identifies 41 new loci associated with breast cancer risk,” Nature Genetics 45, 4 (2013): 353–61. Dorval, Michel, et al. “A focus group study on breast cancer risk presentation: one format does not fit all,” European Journal of Human Genetics 21, 7 (2013): 719–24.
The big white chest freezer in Dr. Jacques Simard’s lab at the research centre of the Centre hospitalier universitaire de Québec doesn’t look that different from a freezer you would find in the home. It’s what’s inside it that makes it unique: invisible clues for the more accurate prediction of millions of women’s long-term breast cancer risk. The freezer preserves the DNA samples of 20,000 women from Africa, Asia, Australia, Europe and North America. What links them is that all of the women have mutations in either one or both of the breast-cancer susceptibility genes BRCA1 or BRCA2.
This wealth of material represents just one-seventh of the total number of samples collected as part of the global study that Dr. Simard is involved in with hundreds of colleagues. These researchers are mining a broad base of dozens of genetic markers to create significantly more accurate personalized breast cancer risk prediction models, work that is forging a new era in breast cancer screening.1
“All women have an intrinsic genetic risk of developing breast cancer,” says Dr. Simard, Canada Research Chair in Oncogenetics. “The current challenge is to create a breast cancer prediction model incorporating a detailed genetic profile coupled with other non-genetic risk factors. This will enable health professionals to personalize each woman’s risk within a risk stratification framework.”
Researchers believe that personalized risk stratification – determining a patient’s risk of developing a specific disease – can help clinicians make better decisions about which patients would benefit from further testing or preventive treatment. In the case of breast cancer, it will help identify the women most likely to develop the disease, therefore enabling more successful, targeted screening, prevention measures and treatment earlier in women’s lives.
The CIHR Team in Familial Risks of Breast Cancer, led by Dr. Simard, is also improving the way that breast cancer risk results are shared with patients. He feels that this risk communication work is just as important as the genetic research itself. “If you have a bad genetic profile but you can’t translate this information in a useful way, then health professionals won’t really understand it, and women won’t understand how it can help improve their health,” he says.
The long reach of breast cancer
Breast cancer affects one out of nine Canadian women in her lifetime, and every year about 23,000 Canadian women are diagnosed with breast cancer. More than 5,000 women die of this disease each year.2
In the early 1990’s, Dr. Simard’s genetic sleuthing contributed to the discovery of the BRCA1 (BReast CAncer gene one) and BRCA2 mutations and the full gene structure of the latter. These are the first genes clearly linked to a significantly higher risk of breast and ovarian cancer. In 1996, Myriad Genetics released the first publicly available test for BRCA1 and BRCA2. Since then, this genetic screening has enabled hundreds of thousands of women with a familial history of early onset breast cancer to be tested for the mutations.
These two genes represent just a beginning in the use of genetic information for improved breast cancer prediction, according to Dr. Simard. Only about one in four hundred women in the general population carries either the BRCA1 or BRCA2 genetic mutations.3 Even among those who do, the predicted breast cancer risk (calculated by also factoring in family history and lifestyle) ranges dramatically from a modest 30% chance to a much more cautionary 90% chance.
Dr. Simard and his colleagues see room for improvement. With support from the Genomics and Personalized Health Genome Canada /CIHR partnership, they are focused on extending the genetic basis of breast cancer risk prediction from two genes, to a more detailed and accurate genetic profile based on BRCA1 and BRCA2, along with dozens of other genetic markers linked to breast cancer.
In combination with non-genetic risk factors, this genetic profile will create a tiered ranking of each woman’s risk. “In this way we can identify younger women, particularly those aged 35 to 49, having a significantly higher risk than average, and for whom it would be beneficial to have access to screening and risk reduction strategies earlier in their lives,” says Dr. Simard.
In March 2013, the international Collaborative Oncological Gene-environment Study (COGS), a large-scale genotyping project, announced a much more detailed mapping of the landscape of the genetics of breast cancer susceptibility.4 The researchers – including Dr. Simard and other members of the CIHR Team in Familial Risks of Breast Cancer – scanned and compared the entire genomes of more than 100,000 women, half of whom were healthy and half who had had breast cancer. They looked for tiny genetic differences, called single nucleotide polymorphisms, or SNPs (pronounced “snips”), at 200,000 DNA locations.
The big picture via international collaboration
The international Collaborative Oncological Gene-environment Study (COGS) is a consortium of more than 160 research groups from universities, hospitals and government labs in more than 40 countries. COGS’ overarching goal is to combine large study groups and the latest, fastest genetic screening technologies to usher in a new era in the high-precision identification of individual risks of breast, ovarian and prostate cancer.
Through this massive genetic association study, the largest-ever for any cancer, the researchers identified 49 new SNPs, or genetic markers, associated with the risk of developing breast cancer, almost tripling the known number.
“Already this finding demonstrates the usefulness of this genetic marker approach,” says Dr. Simard. “If this genetic information is combined with other disease risk factors such as age, family and reproductive history, as well as breast density, we think we would be able to identify the 5% of women who have a one-in-three chance of developing breast cancer.”
Communicating the risk of breast cancer
From a clinical perspective, identifying the women who are at risk is not enough. The statistics involved in breast cancer risk prediction models are often baffling to both physicians and the women who must make possibly crucial decisions based on complex statistical probabilities.5 As a result, the interdisciplinary CIHR team is also focused on finding the best ways to share breast cancer risk information with Canadian women.
A recent CIHR team study revealed that the way in which researchers generally communicate breast cancer risk information to one another is the least accessible and most confusing for patients.6
The study asked more than 100 Canadian women in three cities, all of whom had received breast cancer risk counselling, to comment on five different forms of risk communication. These ranged from a numerical format involving columns of numbers (the mode preferred by researchers) to a colour-coded scale (risk curve) ranging from low risk on the left, to high risk on the right. The study participants rejected the numerical format preferred by researchers. “If you’re giving this to a person who has no medical experience ... this would be too freaky for them,” noted one participant.
“When it comes to breast cancer risk communication, we need to be sensitive to women’s individual emotional and cognitive differences,” concludes Dr. Michel Dorval, a researcher with Laval University’s Faculty of Pharmacy and the study’s lead author. “If we want to maximize understanding of risk communication we need to use a combination of approaches.”
This communications research has already altered the way results from BOADICEA, the world’s primary breast cancer risk prediction model, are communicated. Previously, BOADICEA’s website communicated a woman’s risk prediction results only in numerical format, the one least preferred by patients. BOADICEA now also includes a more accessible risk curve graphic of the same results.
Evidence in Action: Helping high-risk women get breast cancer screening
In 2011, Cancer Care Ontario became the first agency in Canada to offer a stratified risk management approach, offering advanced screening for Ontario women deemed at high risk. This includes being a BRCA1 or BRCA2 mutation carrier, or being assessed with a greater than 25% lifetime risk of developing breast cancer based on the BOADICEA risk prediction model.7
Photo courtesy of Dr. Jacques Simard
“Often in our work we don’t see rapid results, but in this case there was a very rapid impact,” says Dr. Dorval, who leads the CIHR Team in Familial Risk of Breast Cancer’s psychosocial research component.
Back in Dr. Simard’s freezer, the DNA samples are playing a key role in a second large-scale breast cancer association study involving more than 140,000 participants and probing 570,000 SNPs, or genetic markers. Says Dr. Simard: “Future findings will only improve on the new information we’ve already discovered about common genetic markers, including those acting as modifiers of BRCA1- and BRCA2-associated breast cancer risk, which are currently being incorporated into the BOADICEA model to improve the accuracy of personalized risk prediction.”
For More Information:
- Footnote 1
Hall, P., and D. Easton. “Breast cancer screening: time to target women at risk,” British Journal of Cancer 108, 11 (2013): 2202–04. doi:10.1038/bjc.2013.257.
- Footnote 2
- Footnote 3
- Footnote 4
Michailidou, Kyriaki, et al. “Large-scale genotyping identifies 41 new loci associated with breast cancer risk,” Nature Genetics 45, 4 (2013): 353–61.
- Footnote 5
Gaudet, Mia, et al. “Identification of a BRCA2-Specific Modifier Locus at 6p24 Related to Breast Cancer Risk,” PLoS Genetics 9, 3 (2013): e1003173. doi:10.1371/journal.pgen.1003173.
- Footnote 6
Dorval, Michel, et al. “A focus group study on breast cancer risk presentation: one format does not fit all,” European Journal of Human Genetics 21, 7 (2013): 719–24.
- Footnote 7
- Date modified: