Abstract
iStock/Wavebreak
Many gene variants, one genetic profile—that's the approach Mayo Clinic scientists are using to personalize breast cancer prediction. The relevant gene variants here are common variants that contribute little to a person's overall risk of developing breast cancer—at least when they are considered individually. In combination, however, these variants constitute a powerful risk factor, one that offers clinicians as much guidance as breast density or family history.
What's more, the combination approach does not merely duplicate the predictions based on existing breast cancer risk models. It provides an independent risk factor. Accordingly, it could be integrated with existing models, improving their predictive powers.
These findings appeared April 2 in the Journal of the National Cancer Institute (JNCI), in an article entitled, “The Contributions of Breast Density and Common Genetic Variation to Breast Cancer Risk.” The article describes how the Mayo Clinic scientists combined 76 common genetic variants to create a single risk factor, a polygenic risk score (PRS). The scientists essentially added up information on 76 SNPs from 33,673 breast cancer patients and 33,381 health subjects.
The additive approach is needed with common genetic variants because each one contributes little to breast cancer risk, unlike the well-known BRCA1 and BRCA2 mutations. These mutations greatly increase risk, but they account for less than 5% of all breast cancers. The common genetic variants could have broader relevance.
To assess whether common variants offered additional predictive value, the Mayo Clinic scientists looked for overlap between the polygenic risk score (PRS) and the Breast Imaging Reporting and Data System (BI-RADS), a breast density measure that is routinely assessed in the clinical setting. Using logistic regression models, the scientists found that the PRS and BI-RADS were independent risk factors within three clinical studies. In all, these studies involved 1,643 breast cancer patients and 2,397 healthy subjects.
The scientists also incorporated the PRS odds ratio into the Breast Cancer Surveillance Consortium (BCSC) risk-prediction model, which uses breast density, family history of breast cancer, history of breast biopsy, age, and ethnicity to calculate breast cancer risk. Using area-under-the-curve (AUC) statistics, the scientists compared the five-year risk prediction calculated using the BCSC model with and without the genetic information.
“Relative to those with scattered fibroglandular densities and average PRS (2nd quartile), women with extreme density and highest quartile PRS had 2.7-fold increased risk, while those with low density and PRS had reduced risk,” wrote authors of the JNCI article. “PRS added independent information (P < .001) to the [Breast Cancer Surveillance Consortium] model and improved discriminatory accuracy from AUC = 0.66 to AUC = 0.69.”
“This genetic risk factor adds valuable information to what we already know can affect a woman's chances of developing breast cancer,” said study co-author Celine Vachon, Ph.D., an epidemiologist at Mayo Clinic. “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.”
In the current study, integrating the PRS with the BCSC helped to place 11% of women who eventually got cancer into a higher risk category where they would have been likely to benefit from interventions such as MRIs, chemoprevention, or even prophylactic mastectomies. Still, even though the BCSC-PRS model was well calibrated in case-control data, the study's authors noted that independent cohort data will be needed to test calibration in the general population.
“There have been a lot of common genetic variants associated with cancers, not just for breast cancer but also for ovarian cancer and prostate cancer, but so far we haven't seen these being used in clinical practice,” added Dr. Vachon. “In the future, these factors are going to be helpful in defining who is at highest and lowest risk of cancer and help both patients and clinicians make better decisions about their care.”