Breast Cancer Risk Underestimated for Blacks, Study Says
Wednesday, November 28, 2007
The formula that doctors use to calculate a woman's risk of breast cancer underestimates the danger for black women most of the time and especially for those age 50 and older -- the age when they are most likely to benefit from screening tests and protective drugs, according to the first major reassessment of the widely used tool.
"We've been concerned about the assumptions we had to make for African American women and other racial and ethnic groups for some time," said Mitchell H. Gail, a biostatistician at the National Cancer Institute who led the reevaluation of the formula he himself developed. "It turns out that we have been underestimating the risk for African American women."
The advance could have broad implications for many black women, prompting them to reconsider the danger they face from a disease that is women's leading type of cancer and second-leading cancer killer. That could translate into more women undergoing mammograms and other examinations to detect the disease in its earliest, most treatable stages; taking drugs such as tamoxifen to reduce their risk; and signing up for studies to identify better warning signs or risk-reducing medicines.
"This could very much change the way we counsel African American women," said Nancy E. Davidson, a breast cancer expert who heads the American Society of Clinical Oncology. "It will make women better attuned to their personal risk and more eligible for standard interventions, as well as for trials to improve prevention or detection."
The new findings, published online yesterday by the Journal of the National Cancer Institute, are the latest revelation about how breast cancer and other diseases can affect racial groups differently. A growing body of evidence suggests that breast cancer tends to be much more aggressive and deadly among black women, which could help explain why they are more likely to die from it even though fewer of them get it. More than 19,000 African American women are diagnosed with breast cancer each year, and nearly 6,000 die from it.
"This is extremely significant," said Lovell A. Jones, director of the Center for Research on Minority Health at the University of Texas M.D. Anderson Cancer Center. "This is emblematic of a broader problem, which is: We tend to make the assumption that one size fits all. One size does not fit all."
The research examined the Breast Cancer Risk Assessment Tool, more commonly known as the Gail model. Doctors use the model to calculate a woman's risk by plugging in variables such as the age at which she started having her period or had her first child and whether a mother or sister has had the disease.
Because the model was based largely on data collected from about 240,000 white women, Gail and his colleagues decided to try to develop a more accurate alternative using data collected more recently on more than 3,200 black women, including more than 1,600 who had breast cancer.
The researchers then tested the new version and showed that it would have accurately predicted how many African American women in the federal government's Women's Health Initiative would have developed breast cancer. Next, the team compared the new model to the old one by using both to assess data collected from 20,278 African American women who were screened for eligibility for a landmark breast cancer-prevention trial: the Study of Tamoxifen and Raloxifene (STAR) trial, which compared tamoxifen to the newer drug raloxifene.
The accuracy of the two methods varied depending on each woman's particular risk factors, and the old model slightly overestimated the risk for some younger women. But overall the old model underestimated the risk in at least 90 percent of all scenarios, particularly for older women, the team found.
"Because of this model, there are women whose breast cancer could have been detected earlier and maybe treated earlier who were not given that opportunity," Jones said. "If someone is told they are not at risk, they walk away and say to themselves, 'I don't have to worry about it.' I have had women tell me how frustrated and angry they are because they walked into a doctor's office and were given a clean bill of health only to be diagnosed with breast cancer within a very short period of time."
For example, the old model calculated that a 50-year-old black woman who started having her period at age 14, had her first baby at age 32 and had a mother and sister who had breast cancer has a 1.53 percent risk of getting breast cancer within five years. The new model puts her risk at 2.26 percent, pushing her over the line into being eligible for the STAR trial.