Centre for Cancer Genetic Epidemiology
Department of Public Health and Primary Care
University of Cambridge, Cambridge, UK
MedicalResearch: What is the background for this study?
What are the main findings?
Dr. Mavaddat: Recent large-scale genome wide association analyses have led to the discovery of genetic variation- called single nucleotide polymorphisms (SNPs) associated with breast cancer risk. Individually these variants confer risks that are too small to be useful for risk prediction. But when combined as a single score, called a polygenic risk score (PRS), this score may be used to stratify women according to their risk of developing breast cancer. This stratification could guide strategies for screening and prevention.
Our study was a large international collaboration involving 41 research groups from many different countries and included 33,673 breast cancer patients and 33,381 controls. We found that the genetic variants act more or less independently, and that the more risk variants a woman has the higher her risk of breast cancer. When women were ranked according to their PRS, women with scores in the top 1% had a threefold increased risk of breast cancer. This translates into an absolute risk of breast cancer of 29% by age 80. By contrast, women with the lowest 1% scores had a risk of 3.5%.
The PRS was effective in stratifying women with and without a family history of breast cancer, so that highest risk was for women with a family history and a high PRS. Finally, we showed that the PRS was better at predicting the risk of ER-positive breast cancer (potentially relevant to the application of risk stratification to chemoprevention for example, with tamoxifen, raloxifene or aromatase inhibitors).
There has been much debate as to whether genomic profiles are useful for individual risk prediction, especially in the context of the preventative strategies available at the present time. The estimates provided in this study will help inform these debates.
MedicalResearch: What should clinicians and patients take away from your report?
Dr. Mavaddat: These findings are likely to generate interest in women wanting to know their risk of breast cancer. In the medium term this should lead to a more rational approach to prevention and screening, in women with or without a family history. We also aim to put these factors together into a comprehensive risk prediction tool, building on our current risk prediction model, BOADICEA. However further research is needed, particularly in determining the best way to combine the results with other risk factors, the best approaches to delivering the test and its acceptability, before the risk prediction test can be made available in routine clinical practice.
MedicalResearch: What recommendations do you have for future research as a result of this study?
Dr. Mavaddat: The next research priority is to investigate how information from the Polygenic Risk Score Improves Breast Cancer Prediction can be combined with other risk factors, including lifestyle factors like hormone replacement therapy and BMI, breast density and previous breast benign breast disease. Studies are also underway gathering prospective data, in order to validate our findings in an independent, prospective cohort.
Although this study was very large, we would like more precise estimates of risk for people at very high or very low risk, and for women at very young ages. We also need to determine better how to calculate the risks for different tumour subtypes (for example oestrogen-receptor negative disease, or high-grade disease) and how the PRS might affect the prognosis from breast cancer.
Our results apply to women of European ancestry. We don’t yet know how the results apply to women in other populations, for example in African or Asian women.
MedicalResearch.com Interview with:, Nasim Mavaddat M.B.B.S. MPhil PhD PhD (2015). Polygenic Risk Score Improves Breast Cancer Prediction