Skip to main content
Premium Trial:

Request an Annual Quote

Large Study Shows Polygenic SNP Score Can Predict Breast Cancer Risk

NEW YORK (GenomeWeb) – An international team, led by researchers at the University of Cambridge and the Institute of Cancer Research, London, has shown that a 77-SNP polygenic risk prediction method can stratify breast cancer risk among women both with and without a family history of the disease.

In the study, published in the Journal of the National Cancer Institute late yesterday, the group used about 33,500 breast cancer cases and 33,500 controls to develop and test a risk prediction scoring strategy incorporating 77 breast cancer-associated SNPs that were identified in previous GWAS.

None of these polymorphisms are strong predictors of breast cancer risk on their own, but when the researchers combined them into a polygenic scoring strategy, they found that there was a significant correlation between patients' score and their breast cancer risk.

A woman with a polygenic risk score in the top 20 percent, for example, was 1.8 times more likely to develop breast cancer than the average woman, the authors reported. Someone in the top one percent, meanwhile, would be more than three times more likely to develop breast cancer than the average woman.

"This type of testing could fit alongside other standard risk measures, such as family history and body mass index, to improve our ability to target the best preventive treatments and advice to those women most likely to benefit from them," Montserrat Garcia-Closas, a study co-leader and professor of epidemiology at the Institute of Cancer Research, London, said in a statement.

To develop their polygenic score method, the research team evaluated the presence or absence of the 77 SNPs in a cohort of almost 90,000 women from a total of 41 separate studies using microarrays. They then divided the cohort into percentiles and analyzed the stratification of risk among these subsets.

According to the investigators, the odds ratios they observed were similar to what would be expected under a polygenic multiplicative scenario. Women in the lowest one percent of the polygenic score distribution had about a third the risk of women in the middle quintile. And those in the highest one percent of the distribution had a little more than three times the odds ratio of those in that middle group.

The team also looked at differences in the predictive strength of their polygenic scoring among ER-positive and ER-negative patients and observed that the strategy was even more predictive in ER-positive women than it was in the cohort overall.

Finally, the researchers validated their findings using a separate cohort of patients whose samples had not been used for the initial discovery of any of the polygenic score's 77 SNPs. In these new patients the odds ratios stratified similarly, except in the 60 to 80 percent and 90 to 95 percent categories, in which the ratios were slightly higher than in the initial cohort.

According to the study authors, the 77-SNP scoring system was significantly better at defining risk than previous panels that used fewer markers, such as a 10-SNP score reported by researchers from the University of Oxford in 2011 in JAMA.

Importantly, the study also found that the polygenic risk score could predict breast cancer risk both in women with and without a family history of the disease, suggesting that if used clinically, it could improve upon the accuracy of current risk prediction methods.

"There's still work to be done to determine how tests like this could complement other risk factors, such as age, lifestyle, and family history, but it's a major step in the right direction that will hopefully see genetic risk prediction become part of routine breast screening in the years to come," said Douglas Easton, another study co-leader and director of the Centre for Cancer Genetic Epidemiology at the University of Cambridge.

While the team was not able to consider the power of their polygenic risk score relative to lifestyle and environmental risk factors, they wrote that the study provides precise empirical estimates of the level of risk stratification and should help inform debate on the public health utility of incorporating genetic factors into screening strategies for breast cancer in the general public.

In a separate study, also published in JNCI prior to yesterday's larger analysis, a group led by some of the Mayo Clinic researchers involved in the larger effort evaluated whether the polygenic risk score was an independent risk factor apart from other clinical variables using three studies including 1,643 case patients and 2,397 controls in all.

That study compared subjects' five-year risk prediction calculated using a clinical risk factor model — which incorporates breast density, family history, and other factors like age and ethnicity — to a combined approach that added the 76 SNPs into the same model. According the authors, the addition of genetics improved the performance significantly.

"This genetic risk factor adds valuable information to what we already know can affect a woman's chances of developing breast cancer," study co-author and Mayo Clinic epidemiologist Celine Vachon said in a statement.

"We are currently developing a test based on these results, and though it isn't ready for clinical use yet, I think that within the next few years we will be using this approach for better personalized screening and prevention strategies for our patients," she added.