Improving Breast Cancer Risk Assessment

Breast Cancer is the second leading cause of death in women and is often difficult to predict. A recent study identified changes in DNA methylation at 250 specific sites in white blood cells that were predictive of breast cancer in a prospective study cohort. Interestingly, when the methylation levels of just five of the sites were used to predict breast cancer, they did so with a greater accuracy than the current leading method of prediction. We aimed to validate these results by examining five of the DNA methylation sites previously identified using a different cohort of samples and a newly developed method. The white blood cell genomic DNA samples we used in our research were provided by a prospective study conducted by the National Cancer Institute. They were collected from 300 women who eventually developed breast cancer (cases) and 300 women who did not (controls). We received two DNA samples isolated from each individual, one from 1-2 years prior to diagnosis and another 4-7 years prior to diagnosis, with comparable time points for the control cases, amounting to 1200 total samples. Our approach, massively parallel amplicon sequencing of bisulfite treated DNA, has been proven to be robust at detecting DNA methylation levels at single nucleotide resolution. We measured DNA methylation rates at each of the five sites in all 1200 samples using next-generation sequencing. Statistical analysis of the data will assess the accuracy of using these white blood cell DNA methylation sites to assess an individual’s breast cancer risk.

Additional faculty collaborators include Susan Sturgeon, School of Public Health and David Sela, Department of Food Science.