A new model published online November 27 in the Journal of the National Cancer Institute (JNCI) more accurately estimates the risk of invasive breast cancer for African American women than the current NCI Breast Cancer Risk Assessment Tool.

The NCI tool is widely used for determining both risk for individual women and enrollment eligibility for clinical trials of breast cancer prevention agents, but was developed using data collected primarily from white women.

Researchers have long been concerned that the NCI tool may not be as precise in predicting risk for African American and other nonwhite women, and the online version of the tool currently conveys this concern in a disclaimer.

"A lot of the original work was done using white women and validated using white women, and there is a need for studies of this type to either validate or develop specific models for other groups," says Dr. Mitchell Gail, from NCI's Division of Cancer Epidemiology and Genetics, lead author of the JNCI paper and developer of the original model behind the NCI risk-assessment tool, known as the Gail model.

Researchers based the new model on data collected for the Women's Contraceptive and Reproductive Experiences (CARE) study, from 1,607 African American women diagnosed with breast cancer and 1,647 African American women without breast cancer. Information was available for all participating women on the risk factors used to compute the Gail model: age at the start of menstruation; age at first live birth; number of previous benign breast biopsies; and number of first-degree relatives (mother or sisters) with breast cancer.

This information was combined with nationwide breast cancer incidence data from NCI's SEER program and with national mortality data. The researchers then validated the new CARE model using information from 14,059 African American women enrolled in the Women's Health Initiative. The CARE model predicted that 323 women in the group would develop invasive breast cancer. The actual number who developed cancer - 350 - was not statistically different, though the CARE model did underestimate risk among women with a previous benign breast biopsy.

Overall, the CARE model "tended to produce larger estimates of absolute invasive breast cancer risk than the NCI Breast Cancer Risk Assessment Tool in African American women aged 45 years or older," state the authors.

To see how this increased estimation of risk would affect enrollment into breast cancer prevention clinical trials, the investigators compared how many African American women would have been eligible to enroll in the STAR trial using the CARE model versus the number actually enrolled using the NCI Breast Cancer Risk Assessment Tool. Enrollment in the STAR trial required a 5-year projected risk of at least 1.66 percent.

They found that while 14.5 percent of African American women screened using the NCI tool were eligible, this number rose to 30.3 percent using the CARE model. "While the average increase in 5-year risk found with the CARE model compared to the NCI tool is less than one-half percent, the proportion eligible increased by more than twofold," said Dr. Joseph Costantino, director of the National Surgical Adjuvant Breast and Bowel Project Biostatistical Center and co-author of the CARE model.

Like the NCI tool, the CARE model would not be accurate in certain subpopulations, such as women with a prior history of breast cancer or with mutations in genes such as BRCA1 or BRCA2, which greatly increase the risk of breast cancer.

But while the CARE model would still benefit from further validation, especially in women under the age of 50, the authors recommend that clinicians adopt the CARE model "for counseling African American women and for determining their eligibility for breast cancer prevention trials." The CARE model will be included in a revision of the online NCI tool.

—Sharon Reynolds