More Accurate Method of Estimating Invasive Breast Cancer Risk in African American Women Developed
- Posted: November 27, 2007
A new model for calculating invasive breast cancer risk, called the CARE model, has been found to give better estimates of the number of breast cancers that would develop in African American women 50 to 79 years of age than an earlier model which was based primarily on data from white women. Both models were designed to be used by health care professionals and should either be used by them or in consultation with them. Researchers at the National Cancer Institute (NCI), part of the National Institutes of Health, and their collaborators report on the study methodology and results online in JNCI on November 27, 2007.
"NCI's Breast Cancer Risk Assessment Tool has been widely used for counseling women and determining eligibility for breast cancer prevention trials," said NCI Director John E. Niederhuber, M.D. "The development of the CARE model highlights the need to develop targeted tools to assess an individual woman's risk, and those tools must be based on many factors that also assure that the tool can be used in a non-discriminatory manner."
The NCI investigators worked with colleagues from the Women's Contraceptive and Reproductive Experiences (CARE) Study, the Women's Health Initiative, and the Study of Tamoxifen and Raloxifene trial (a breast cancer prevention trial) to produce and test the new model. Some members of the team had worked on both the CARE and earlier model, called BCRAT (Breast Cancer Risk Assessment Tool). Because of the higher accuracy of the CARE model for African American women, the NCI authors are now recommending its use for counseling these women regarding their risk of breast cancer.
While the BCRAT allows for projections for African American women and for women from other racial and ethnic groups, these projections are based on certain assumptions. In particular, it is assumed that the relative risk of breast cancer associated with having a specific profile of risk factors for white women applies to African American women and to women from other racial and ethnic groups as well. Because of the need to rely on these various assumptions, rather than on sufficient data from African American women and women in other racial and ethnic groups, BCRAT, which can be found on the NCI Web site at http://www.cancer.gov/bcrisktool, includes a disclaimer for African American women and for women in other groups that their projections might be inaccurate.
To develop a new model that would more accurately assess an African American woman's chance of developing breast cancer, researchers in the CARE study examined data from 1,607 African American women with invasive breast cancer and 1,637 African American women of similar ages who did not have breast cancer. The factors used in the model were age at first menstrual period, number of first degree relatives (mother or sisters) who had breast cancer, and number of previous benign breast biopsy examinations. A woman's age at the birth of her first child, a risk factor for white women, did not improve prediction in African American women and so was not included in the model. Risk was calculated by combining information on these factors with African American rates of new invasive breast cancer from NCI's Surveillance, Epidemiology and End Results Program and with national mortality data.
To test the accuracy of the model, researchers compared data in the CARE model with data from the 14,059 African American women aged 50 to 79 in the Women's Health Initiative (WHI) study who had no prior history of breast cancer. From the risk factor profiles for breast cancer that were collected at entry into the WHI, the researchers used the CARE model to estimate the number of women who would be expected to develop invasive breast cancer and found that the model predicted that 323 would be affected, close to the 350 breast cancers in African American women that actually occurred during the WHI follow up. According to Mitchell H. Gail, M.D., NCI, the lead author of this study, "The CARE model predicted the numbers of breast cancer diagnoses well overall, and in most categories."
One of the key uses of the BCRAT has been to determine eligibility criteria for a number of breast cancer prevention trials. For African American women 45 and older, the CARE model risk projections were usually higher than those from the BCRAT. To assess what the impact of using the CARE model might have been on a recently completed prevention trial, the researchers used eligibility screening data from 20,278 African American women who were examined in the Study of Tamoxifen and Raloxifene (STAR) trial between 1999 and 2004. The investigators estimated that 30.3 percent of African American women would have had significant five-year invasive breast cancer risks based on the CARE model, compared to only 14.5 percent based on BCRAT.
"African American women were both more interested in and more likely to enroll in the STAR trial compared to the earlier Breast Cancer Prevention Trial, but the recruitment process and our enrollment task would have been easier if the CARE model had been available," said Worta McCaskill-Stevens, M.D., NCI, one of the leaders of the STAR trial.
Additionally, inaccurate projections using the BCRAT could result in African American women receiving an underestimate of their breast cancer risk. As a result of this underestimate, African American women might not get counseling about actions they could take to reduce their risk. "There has been great interest in developing race- or ethnicity-specific adaptations of the BCRAT model that are based on sufficient race- or ethnicity-specific data, and the CARE data enabled us to develop the new model," said Gail.
It should be noted that the CARE model, like the BCRAT, needs to be approached with caution or avoided for certain special populations. These models should not be used for women with a previous history of breast cancer. The models tend to underestimate risk in women who have received radiation to the chest and in women who are known to carry mutations associated with increased risk of breast cancer, such as mutations in the BRCA1 and BRCA2 genes. While the CARE model has not yet been incorporated into the BCRAT on the NCI Web site, NCI plans to have the tool updated by the spring of 2008.
Reference: Gail MH, Costantino JP, Pee D, Bondy M, Newman L, Selvan M, Anderson GL, Malone KE, Marchbanks PA, McCaskill-Stevens W, Norman SA, Simon MS, Spirtas R, Ursin G, and Bernstein L. Projecting Individualized Absolute Invasive Breast Cancer Risk in African American Women. JNCI. Vol. 99, No. 23.
Author affiliations: Centers for Disease Control and Prevention, Atlanta, Ga.; Fred Hutchinson Cancer Research Center Seattle, Wash.; Information Management Services, Rockville, Md.; Karmanos Cancer Institute at Wayne State University, Detroit, Mich.; University of Michigan, Ann Arbor, MI; Division of Cancer Epidemiology and Genetics and Division of Cancer Prevention, National Cancer Institute, Bethesda, Md.; National Institute of Child Health and Human Development, NIH, Bethesda, Md.; University of Pennsylvania School of Medicine, Philadelphia, Pa.; University of Pittsburgh, Pittsburgh, Pa.; Keck School of Medicine and Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, Calif.; The University of Texas M. D. Anderson Cancer Center, Houston, Texas.
For more information about cancer, please visit the NCI Web site at www.cancer.gov, or call NCI's Cancer Information Service at 1-800-4 CANCER (1-800-422-6237).