NEW YORK (Reuters Health) – A risk prediction tool for breast cancer in US Black women showed similar accuracy to current questionnaire-based models in a validation study, and seems suitable for risk stratification in primary care settings.
The tool is available here.
“We are working with neighborhood health centers in Massachusetts to examine provider acceptability of this tool,” Dr. Julie Palmer and Dr. Karin Grunebaum of Boston University School of Medicine told Reuters Health by email. “At the same time, we are developing plans for disseminating the tool to primary care provider associations across the U.S.”
“We will eventually add mammographic density to the model, which will likely increase the predictive ability for Black women who have already had at least one mammogram,” they said. “We are currently in the process of obtaining digital records of mammograms from participants in the Black Women’s Health Study so that this factor can be added to the model and validated.”
As reported in the Journal of Clinical Oncology, to develop an absolute risk model, the team estimated relative and attributable risks of breast cancer using data from Black women in three US population-based case-control studies that included 3,090 cases and 3,578 controls ages 30-69, and combined them with SEER age- and race-specific incidence rates (incorporating competing mortality).
The model was validated in data from 51,798 participants in the Black Women’s Health Study, including 1,515 who developed invasive breast cancer, and was found to be well-calibrated. Discriminatory accuracy, which reflects how well a model predicts risk for an individual woman, was similar to that of the most frequently used questionnaire-based breast cancer risk-prediction models in white women, and worked best for women under age 40.
The team also developed a second risk-prediction model on the basis of estrogen receptor (ER) – specific relative risks and attributable risks. They found that variables associated with increased relative risk in both ER+ and ER- subtypes were first-degree T2 family history of breast cancer, breast biopsy, five or more years of oral contraceptive use, earlier age at menarche, and lack of breastfeeding; bilateral oophorectomy was associated with reduced risk.
Family history of prostate cancer, lower BMI at age 18, BMI greater than 30 kg/m2 during the postmenopausal period, later age at first birth, and nulliparity were associated with increased risk of ER+, but not ER- breast cancer.
By contrast, higher parity was associated with increased risk of ER- breast cancer.
Calibration and discriminatory accuracy were almost identical in this model as in the original model.
Thus, the authors note, “We demonstrated that a risk prediction model on the basis of separate relative risks and attributable risks for ER+ and ER- breast cancers does not improve performance relative to a single-disease model.”
Drs. Palmer and Grunebaum added, “We are working with collaborators at several other institutions on developing a risk-prediction model for lung cancer in Black men and women. In addition, we are in the process of examining how well the National Cancer Institute risk-prediction model for colorectal cancer predicts risk in Black women.”
Dr. Manmeet Ahluwalia, Chief of Solid Tumor Medical Oncology, Chief Scientific Officer and Deputy Director at Miami Cancer Institute, part of Baptist Health South Florida, commented on the study in an email to Reuters Health. “In recent years, there has been a big focus on diversity and for improving disparities. We know that cancer often is a genetic disease and that there are genetic variations in Black and Hispanic women compared to white women. Hence, this tool is a big step in the right direction.”
“We will use this tool in our clinics to have more accurate modeling for Black women (rather than) tools that were developed primarily using majority white women,” he said. “The breast cancer risk- prediction model can be an added tool in the armamentarium of physicians who take care of this population.”
Like Drs. Palmer and Grunebaum, he pointed to the need for mammographic density to improve discriminatory accuracy.
Further, he noted, while a polygenic risk score (PRS) can improve the utility of prediction models, “PRS stratifies risk poorly because of the genetic variation in Black women. The addition of a race-specific PRS in the future may prove useful.”
SOURCE: https://bit.ly/2YXQ9D3 Journal of Clinical Oncology, online October 8, 2021