Share this:
Each year, more than one million women have biopsies that show noncancerous changes in the breast, known as benign breast disease (BBD). Even though the changes may not require treatment, studies over the years at Mayo Clinic and elsewhere have found that not all cases of BBD are the same, and some will go on to develop breast cancer.
“The real challenge has been determining women’s individual risk for developing breast cancer in order to provide them with the best possible intervention,” says Mark Sherman, M.D., an epidemiologist and laboratory medicine and pathology researcher at Mayo Clinic in Florida.
Building on several decades of breast cancer research at Mayo Clinic, an interdisciplinary team aims to predict which women with BBD are at risk for breast cancer. The team, led by breast surgeon Amy Degnim, M.D., in Rochester, and Dr. Sherman, recently received a $3.1 million grant from the National Cancer Institute, a division of the National Institutes of Health to develop a breast cancer risk prediction model to help guide clinical care. The model will take into account demographic factors, as well as recent research about features of breast tissue that may heighten the risk for cancer.
Researchers at Mayo Clinic published the first report on a cohort of 9,000 women just over two decades ago, supporting previous studies that had stratified patients with BBD into high, medium and low risk categories. In the last 15 years, screening techniques have become more highly targeted, with the use of mammograms, MRI, and radiologically guided needle biopsies. They’re now catching more details in the breast changes. Just as significantly, new information has emerged about how characteristics of breast tissue, such as density, are relevant to risk. And recent studies have suggested when breast lobules, the ducts that make milk, don’t shrink as a woman ages, the risk for cancer increases.
Using information from more than 7,000 patients at Mayo Clinic in Rochester, the next-generation risk model will incorporate a wide range of risk factors, including demographics, samples from radiologically-guided needle biopsies, information about breast density, molecular biomarkers, and breast lobules. Partnering with a team from Karmanos Cancer Institute that has an established cohort of nearly 4,000 African-American women, the study will also provide an understanding of ethnoracial breast cancer risk factors. The model will rely on big data and sophisticated machine learning techniques to generate risk predictions following a BBD diagnosis.
Ultimately, the model may provide women with highly individualized options. Those at high risk may consider preventive treatment, while those at low risk may avoid unnecessary tests.
“Even though we’ve been able to generalize predictions for patients, we know not everyone has the same degree of risk,” Dr. Sherman says. “Our goal is to achieve individualized risk prediction for a better targeted approach to care.”
This article originally appeared on the Advancing the Science blog.
