Breast cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death among Hispanic women in the U.S. Yet the NCI's Breast Cancer Risk Assessment Tool (BCRAT), which helps providers calculate risk in individual patients, has no specific model for Hispanic women and currently underestimates their risk of developing breast cancer.
We aim to change that with the first-ever Hispanic Risk Model (HRM) for breast cancer, based exclusively on data from Hispanic women, both U.S.- and foreign-born.
Details of the HRM are published in the Journal of the National Cancer Institute and will be incorporated into the BCRAT following feedback from scientific and clinical experts (2017; https://doi.org/10.1093/jnci/djw215).
The BCRAT already includes breast cancer risk models for White, African-American, and Asian-American women.
Building & Validating the Risk Model
We started with data from the San Francisco Bay Area Breast Cancer Study, which included 1,086 Hispanic women who developed breast cancer between 1995 and 2002 and 1,411 women who did not have breast cancer. Nearly 1,000 of the women were born in the U.S., and 1,500 were born in other countries. We also included breast cancer incidence and mortality data from the California Cancer Registry and NCI's Surveillance, Epidemiology, and End Results program.
To validate the HRM, we used independent data from the Women's Health Initiative and the Four-Corners Breast Cancer Study. The HRM was well-calibrated for U.S.-born Hispanic women and accurately predicted the number of breast cancers in this group (observed/expected [O/E] ratio=1.07, 95% confidence interval [CI]=0.81 to 1.40). The HRM slightly overestimated the number of breast cancers among foreign-born Hispanic women in the WHI (O/E ratio=0.66, 95% CI=0.41 to 1.07), although this overestimation was not statistically significant.
In regard to discriminatory accuracy, the nativity-specific concordance statistics (AUC) of the HRM in the WHI cohort were an AUC of 0.564 (95% CI=0.485 to 0.644) for U.S.-born Hispanic women and an AUC of 0.625 (95% CI=0.487 to 0.764) for foreign-born Hispanic women.
Who Should Use the HRM?
Once the HRM is incorporated into the BCRAT, any health care provider or researcher will be able to use it to help estimate breast cancer risk in female Hispanic patients. Because we built the HRM using data from Hispanic women in California, who are mostly of Mexican and Central American descent, the HRM will be most accurate for these women. We have not yet assessed the performance of the HRM in other Hispanic women (e.g., Puerto Ricans or Cubans).
The HRM, like the BCRAT, should not be used to predict breast cancer risk for women who have a history of breast cancer, those who have an inherited genetic mutation known to cause breast cancer (e.g., BRCA1 and BRCA2), or for those who received radiation therapy on the chest for other types of cancers.
Clinical Implications
The development of the HRM is good news for Hispanic women. They are the largest racial/ethnic minority group in the U.S., so it's important the NCI's risk assessment tool include a model specifically for these women.
The BCRAT asks health care providers to enter information about the patient's age, race, family history of breast cancer, and other risk factors, including:
* when the patient started her first menstrual period;
* how old she was when she gave birth to her first child;
* whether she has first-degree relatives with breast cancer; and
* whether she has had a breast biopsy for benign breast disease.
The HRM includes data about all of these risk factors from Hispanic women. It also includes data from Hispanic women born in the U.S. and those born in other countries. This is an important distinction since Hispanic women born in the U.S. have a higher risk of developing breast cancer.
All of this information will be added to the BCRAT to increase its accuracy. Now, when providers enter "Hispanic" into the BCRAT, a box pops up that says "Assessments for Hispanic women are subject to greater uncertainty than those for White and African-American women." Once the HRM is added to a prediction tool, that box will likely go away, and the tool will provide more accurate risk calculations for Hispanic women. This will help providers make better decisions about when and how to screen these women for breast cancer.
MATTHEW P. BANEGAS, PHD, MPH, is lead author and researcher, Kaiser Permanente Center for Health Research. KATIE DEMING, MD, is a radiation oncologist, Kaiser Permanente Northwest, Portland, Ore.