Study Questions Accuracy of Genetic Risk Models Used to Determine Testing

Many breast cancer patients who have a genetic predisposition for the disease, but who do not have an obvious family link to breast cancer, are not considered candidates for genetic testing, new research shows.

The study challenges the accuracy of models used to determine who should be offered testing for the breast cancer gene mutation BRCA, finding the models to be poor predictors of risk for women with few female relatives over the age of 45.

The models assess risk based on known family history of breast cancer.

"You have to have a family to have a family history," researcher Jeffrey Weitzel, MD, of the City of Hope Cancer Center in Duarte, Calif., tells WebMD.

He added that women with few older female relatives on either their mother's or father's sides were not well served by the available models.
Mutation Uncommon

Inherited gene mutations play a role in about 5% of all breast cancers. But women with BRCA1 or BRCA2 mutations have about a 50% to 85% lifetime risk of developing the disease, compared with a roughly 10% lifetime risk among the general population of women.

BRCAcarriers are also much more likely to develop breast cancers before the age of 50, and their risk for ovarian cancer is also greatly increased.

Identifying women with BRCA mutations is a proven strategy for reducing the risk of death from breast or ovarian cancer. But since the mutations are rare, testing all women or even testing all breast cancer patients may not be practical.

Most women who develop breast cancer before the age of 40 are considered for genetic testing, but clinicians often rely on the prediction models to determine whether women diagnosed in their 40s should be screened if they have no known family link to the disease.

Weitzel and colleagues assessed the accuracy of the prediction models in their study appearing in the June 20 issue of TheJournal of the American Medical Association.

A total of 1,543 breast cancer patients enrolled in a cancer registry between 1997 and 2007 were included in the study. Of these patients, 306 were diagnosed before age 50 and reported no known first- or second-degree relatives with breast or ovarian cancers.

When these women were questioned further about their family history, half of them (153) were determined by researchers to have too few close female relatives to accurately assess risk, defined as fewer than two females relatives age 45 or older on both their mother's and father's sides of the family.

Genetic testing revealed that these women had a 14% risk of having a mutation, compared with a 5% risk among women with a larger number of female relatives who survived to their mid-40s without developing breast cancer.

"The likelihood of finding a mutation was almost three times greater [among the women with fewer older female relatives]," Weitzel says. "We are clearly missing women who could benefit from knowing their genetic risk."
Understanding the Limitations

Cancer geneticist and gynecologist Noah D. Kauff, MD, of Memorial-Sloan Kettering Cancer Center, agrees.

"We have to understand the models and their limitations if we are going to use them," he says. "They likely still have a role in helping us identify patients who would benefit from genetic testing, but it is important that they are not used in isolation."

Breast cancer patients with BRCA mutations are generally treated more aggressively; they are at increased risk for another breast cancer and ovarian cancer.

The American Cancer Society now recommends annual breast screening with magnetic resonance imaging (MRI) along with mammography for high-risk women, as determined by existing family-risk models.

Kauff tells WebMD that the recommendation illustrates the importance of finding better ways to identify women at increased risk.

"If these models are incomplete for testing, they are obviously incomplete for making decisions about who should be screened and treated," he says.

source:www.webmd.com