Too Many Black Women Die from Breast Cancer. Why?
BU Slone Epidemiology Center researchers look for answers
Breast cancer is not color-blind. Although it strikes women (and less commonly, men) of every age and race, black women are more likely than white women to die of breast cancer. Why?
Researchers have been aware of the disparity for decades, but it has resisted easy explanation. And today, the stakes are higher than ever. That’s because breast cancer rates in black women are rising, even as they are dropping for white women. White women were about 15 percent more likely to be diagnosed with breast cancer than black women 15 years ago, probably because they were more likely to be screened, typically had children later, and took more postmenopausal hormone supplements, which feed the types of breast cancer more common in white women. Now, as many women have chosen to cut out those postmenopausal hormones, the lines have converged, with women of both races equally likely to develop the disease. But on average, black women are diagnosed at a younger age (median age is 58 for black women, 62 for white women) and die at a younger age than white women (median age is 62 for black women, 69 for white women).
Researchers at BU’s Slone Epidemiology Center are building on a two-decade legacy of black women’s health research to understand the breast cancer survival disparity. With support from the National Cancer Institute, they are combining expertise in genetics, biology, and epidemiology and the social factors that influence health to search for answers that will illuminate how breast cancer takes hold and proliferates in the body—and how we might be able to prevent and treat it more effectively, not just in black women, but in everyone.
“Black women have a 42 percent higher rate of death from breast cancer than white women or most other groups in the United States, and we don’t know the reasons for that,” says Julie Palmer (SPH’85), a School of Public Health professor of epidemiology and Slone Center associate director. “What we do know is that it can’t be attributed to poor care or coming to diagnosis later. There’s clearly something going on that’s different.”
A letter in the mail
In 1995, a woman named Vi Brown got a letter in the mail. “Dear Friend,” it began, “We are delighted to inform you that the US National Institutes of Health has recently provided funds for the largest ever long-term study of the health of African-American women. There has been little previous research, and this study will provide much-needed information on causes of illness.” On the left was the seal of the Boston University School of Medicine, on the right, the shield of Howard University College of Medicine, and between them, in bold, double-underlined type, the name of the new project: The Black Women’s Health Study (BWHS).
Folded inside the letter was a 14-page survey dense with questions about the recipient’s medical history (Had she every been diagnosed with cancer? Heart disease?), her life story (Did she have any children? If so, had she breastfed them?), and lifestyle (Did she smoke? Drink? Eat vegetables?). The letter went out to more than 100,000 women, and Brown was among the 59,000 women who returned it.
Then in her mid-30s, she had actually read about the study in Essence magazine. “I never knew back then that they had not done a comprehensive health study on African-American women,” she says. “I thought, well, gee, why not?” Brown, who now works as a business and engineering consultant and owns her own company, reached out to the organizers, and the letter arrived shortly after.
When she volunteered, she had never had any serious health problems. But just two years later, she was diagnosed with early stage breast cancer. She is now celebrating nearly 20 years of being cancer-free, and is encouraged by the move toward personalized medicine informed by a patient’s gender, race, and background. “The more we know, the more specific we can be in making sure the person is getting the right treatment,” Brown says.
Today, the study, the nation’s largest and longest running examination of black women’s health, is still going, led by Palmer, Slone Center senior epidemiologist Lynn Rosenberg (GRS’65), an SPH professor of epidemiology, and Lucile Adams-Campbell of Georgetown University’s Lombardi Comprehensive Cancer Center. It has far outlived its original five-year commission, and researchers keep up with their volunteers via new surveys that go out every two years. The BWHS has also teamed up with three other such projects under the umbrella of the AMBER Consortium (African American Breast Cancer Epidemiology and Risk), which is supported by a grant from the National Cancer Institute. Combined, the projects include some 5,000 African American women with breast cancer, plus 10,000 healthy control volunteers, giving researchers more possibilities of revealing potentially subtle effects influencing black women’s risk of getting, and prospect of surviving, breast cancer.
Doctors once thought that all breast cancers were the same. We now know that breast cancer is actually a multiplicity of diseases, manifesting in different tumor types that are driven by diverse genetic defects. These varied breast cancers can be roughly divided into two classes: estrogen receptor positive (ER-positive) breast cancers, which feed on estrogen and can be knocked down by antiestrogen therapies like tamoxifen, and ER-negative tumors, which don’t respond to estrogen. (ER-negative cancers include a rarer class of tumors called triple-negative cancers, which lack receptors for progesterone and the HER2 protein as well as estrogen.) Women diagnosed with ER-negative breast cancer, treated with the blunt weapons of chemotherapy and radiation, are less likely to survive than those with ER-positive tumors.
For reasons scientists are still trying to understand, black women are more likely than women of other races to be diagnosed with difficult-to-treat ER-negative breast cancer. “There’s a big difference in the prevalence of those types of tumors in black women and white women with breast cancer,” says Palmer. “In black women, the cancer is twice as likely to be ER-negative than in all other groups.” This partly explains the survival disparity, but it raises another question: why are black women more likely to be diagnosed with ER-negative cancer in the first place?
No just “black” or just “white”
Slone epidemiologist Edward Ruiz-Narváez, an SPH associate professor of epidemiology, is using genetic data collected by the BWHS and other studies within the AMBER Consortium to search for genetic changes that could help explain the difference.
“We’re interested to know what parts of the genome are conferring higher risk,” says Ruiz-Narváez. Although whole-genome sequencing gets lots of media attention, it remains expensive and time-consuming, so Ruiz-Narváez turns to shrewd shortcuts that can point to areas of the genome that may hold answers.
One of those shortcuts exploits the fact that no one is really just “black” or just “white.”
“We are humans; we are a mix of different populations,” he says. Using a technique called admixture mapping, he identifies areas of the genome that have relatively more or less African ancestry. He then compares that ancestry map with breast cancer risk. His target: parts of the genome that tip toward both African ancestry and heightened breast cancer risk. These areas might contain genetic variations that are responsible for cancer types that are more frequent in black women.
“We are screening the whole genome,” says Ruiz-Narváez. “We are not making any a priori hypothesis about where in that genome are variants affecting risk of breast cancer.”
Ruiz-Narváez’s admixture mapping work has already uncovered two such regions. When he and his colleagues examined one of those regions letter-by-letter, they found a one-letter change that seems to increase breast cancer risk by almost 50 percent and is 14 times more common in people with African ancestry than in those with European roots. The research was published in the journal Frontiers in Genetics in September 2016.
One advantage of ancestry mapping is that it is an agnostic approach. “We are screening the whole genome,” says Ruiz-Narváez. “We are not making any a priori hypothesis about where in that genome are variants affecting risk of breast cancer.”
But the opposite approach—taking what’s known about cancer biology and using it to take an educated look at particular parts of the genome—is also paying off. Last year, Ruiz-Narváez, Palmer, and a team of colleagues examined 26 genes that work together to transmit signals from a cell’s outer membrane to breast cancer–related genes in its nucleus. The trail led them to gene called FGF1 and a single letter change found only in people with African ancestry that seems to dramatically hike the risk of getting ER-negative breast cancer. This finding, published in Breast Cancer Research and Treatment in 2016, is a tantalizing start, Ruiz-Narváez says, but will need to be replicated before he and his colleagues are confident that the mutation is as important as it looks.
Genetic differences may partly explain why black women with breast cancer are more likely to die than women of other races, but they are not the whole answer. “Genes don’t act in a vacuum,” says Ruiz-Narváez. The environment changes the way genes are expressed—meaning that genetic destiny is sometimes open to revision.
Researchers are tapping the BWHS data to uncover which environmental factors might be most potent at tipping the breast cancer odds. Last year, a team led by Slone epidemiologist Kimberly Bertrand (SPH’05), MED’s Dahod Breast Cancer Assistant Professor of Medicine, found that carrying extra fat around the middle, getting pregnant for the first time later in life, having more children, and forgoing breastfeeding were all linked with a higher risk of getting ER-negative breast cancer before age 45. Those cancers tend to be particularly aggressive, hit younger women, and are less likely to be found on routine mammograms, which are typically done only after age 40.
Bertrand is especially interested in how breastfeeding may reduce the risk of getting this type of cancer. She says that breastfeeding may be an especially powerful barrier against ER-negative cancers: in the study, younger women who had breastfed even a little were about 40 percent less likely to be diagnosed with ER-negative cancer than women who had never breastfed. But black women are less likely to breastfeed than women of other races, notes Bertrand, and this may be because black women have different cultural attitudes toward breastfeeding and may have less access to the support they need to make breastfeeding work.
“They may be more likely to work lower-income jobs, where they don’t have a maternity leave, and they may not have workplace policies that support breastfeeding,” she says. Health care workers may also assume that new mothers who are African American are less likely to breastfeed, and so offer them less encouragement postpartum: fewer visits from a lactation consultant, for instance, and more baby bottles and cans of formula.
But Bertrand sees the news as encouraging. “I think it points to potentially modifiable risk factors for breast cancer in black women. Breastfeeding, adiposity: these are things that, in theory, can be changed,” she says, even as she acknowledges that making those changes is not easy.
Not every risk factor is changeable. The strongest predictors of getting breast cancer are sex and age, after all. But Bertrand is also studying a third factor, called breast density (the ratio of fibrous and glandular tissue in the breast relative to fat in the breast), that could help women better understand their risk of getting breast cancer. “Women who have very high breast density have four to six times the risk of breast cancer compared to women with primarily fatty breasts,” she says. But those figures come from studies of white women, and not much is known about how they translate to black women. Using digital mammograms from women in the BWHS, she is looking to see if the relationship between breast density and cancer risk extends to black women.
Bertrand hopes that her work will help inform better risk assessment for black women. Risk assessment isn’t just fortune-telling: it can help women make informed choices about when to be screened and qualify them for studies of experimental treatments that are open only to high-risk women. But today’s standard risk models are based on data from white women, and they fall down when applied to black women—possibly because the cancer subtypes that are frequent in black women are less common in white patients.
Difference between surviving and thriving
The neighborhood a person lives in, the kind of work she does, and her ability to make ends meet all affect her cancer risk, diagnosis, and outcome, as well, points out Slone epidemiologist Traci Bethea (SPH’11), a MED assistant professor of medicine. A person who has to travel miles by bus to reach the nearest grocery store may not be able to follow her doctor’s advice to fill her plate with leafy greens. A breast cancer patient choosing between paying for her apartment and paying for her prescriptions may forgo critical medicine. “These are complicated issues that can’t be solved solely in the clinical space,” Bethea says.
Affluence doesn’t necessarily buy good health, though. Combining data from the BWHS with neighborhood information from the US Census and the American Community Survey, Bethea, along with Palmer, Rosenberg, and Slone epidemiologist Yvette Cozier (SPH’94,’04), an SPH assistant professor of epidemiology, have uncovered a far more complicated picture, one that says where you live may be more important than what you earn. “For the BWHS we have a wide variability in education and also income,” says Bethea, which is uncommon among studies of black women’s health. “We have crossover: women of low socioeconomic status in neighborhoods of high socioeconomic status,” and vice versa. “Looking at the overlaps and contradictions is really important.”
“If we learn more about prevention, about behaviors, about survivorship, and giving people the tools to live healthier, and happier,” Bethea says, “we believe that will make a return not just for one person and her family but for the entire population.”
The result: when it comes to health, your neighbors’ income, employment, and education levels may be more important than your own. Their findings, published online in Ethnicity & Disease in April 2016, “suggest you can’t really earn your way out of where you’re living,” says Bethea. Black Americans “are more likely to live in disadvantaged neighborhoods regardless of their educational attainment or income,” she and her coauthors write, so understanding the connections between residential community and health is a key part of addressing health disparities.
More than 1,800 of the original 59,000 BWHS volunteers have been diagnosed with breast cancer since the study began in 1995. Bethea is particularly interested in what happens after that diagnosis: the factors that make the difference between living well with cancer and struggling to carry the physical, spiritual, and financial weight of the disease. In the coming years, she hopes to use the BWHS data to develop advice that will help breast cancer survivors live longer and healthier lives and reduce the odds of having their cancer recur.
Last year, Palmer and a team of researchers that includes Bethea and Bertrand took their first “after breast cancer” look at the BWHS data. They found a strong link between breast cancer mortality and type 2 diabetes, which also takes a disproportionate toll on black women. (African Americans are almost 75 percent more likely than non-Hispanic whites to have diabetes.) Women who had been diagnosed with diabetes at least five years prior to their breast cancer diagnosis were almost twice as likely to die of the breast cancer.
“Diabetes causes inflammation, and it has metabolic effects that could help tumors grow,” Palmer says. Although the exact biological connections are not yet clearly mapped, the researchers think that high rates of type 2 diabetes in black women could be a key factor in the breast cancer survival gap—meaning that preventing and managing diabetes could help save lives not just from diabetes, but from breast cancer, too.
“If we learn more about prevention, about behaviors, about survivorship, and giving people the tools to live healthier, and happier,” Bethea says, “we believe that will make a return not just for one person and her family, but for the entire population.”
Kate Becker can be reached at firstname.lastname@example.org.
Breast Cancer Research
April 24, 2017
Zebrafish Cancer Genetics Illuminate Human Breast Cancers
April 24, 2017
Dioxins Point to Targets for Treating Breast Cancer
April 24, 2017
Patient Navigators Improve Delivery of Care for Breast Cancer Patients
No exploration of the long dispute over whether abortion is implicated in breast cancer? Here’s a correlation waiting to be explored: black women have abortions at five times the rate of white non-Hispanic women, according to a 2014 Atlantic article: https://www.theatlantic.com/health/archive/2014/09/abortions-racial-gap/380251/
Have you thought to also explore the correlation between the access to and quality of sexual education, family planning, and prenatal care for black women in comparison to white women? I feel like that disparity might help determine why there’s such a disparity in abortion. Just a thought.
It’s not a dispute. It’s been proven multiple times there is no sufficient evidence to prove there is a correlation between abortion and breast cancer. The one study that found there was a correlation between breast cancer and abortion has been discredited a long time ago. https://www.cancer.org/cancer/cancer-causes/medical-treatments/abortion-and-breast-cancer-risk.html
The Boston Globe did an engaging article some years ago “How Racism Hurts–Literally” 7/15/2007. Societal suffering from stigma and discrimination has impact not only on telomeres but certainly on immunity …and chronic inflammatory cascades. We have to engage the biopsychosocial impacts if we are ever to understand epidemiological findings and then attend to much needed structural issues of policies and practices that contribute to dis-ease in our human populations.
Good article, lousy headline. Even one person of any race or gender dying of breast cancer is “too many”. A headline with “too many” about any harmful thing implies that there is a number that is just right, which is absurd. “Black women disproportionally afflicted with breast cancer” would have been a better summary.
As a participant in the BWHS, I’m always encouraged to read the analysis of all the data you collect. I recently had a mammogram, possibly my 10 in the last 6 years. For the first time I received a letter stating I had dense breast tissue. In the future I’d like to participate in any mammogram study.