NEWS

Cancer geneticists tackle troubling ethnic bias in studies

Multi-million efforts are underway to fill long-standing gaps in genomic data from minority groups.

Search for this author in:

A 14 year old girl with Hodgkins Lymphoma is to be treated at St. Jude in Memphis, Tennessee

African Americans, such as this teenage girl with lymphoma (left), are under-represented in genetic studies of cancer.Credit: Karen Focht/ZUMA Press

Atlanta, Georgia

When Bárbara Segarra-Vázquez’s breast cancer came roaring back last summer after a 13-year hiatus, her physicians recommended surgery and a genetic test to determine whether chemotherapy was warranted. The test results suggested that she could forgo the drugs, and she did. But a nagging doubt remains.

“They said, ‘You don’t need chemo.’ But do I, or do I not?” says Segarra-Vázquez, dean of the School of Health Professions at the University of Puerto Rico in San Juan, who is Latina. “I don’t know, because they didn’t test people like me. The validation of that test was done in white Europeans.”

Her story illustrates a long-standing bias in cancer research: most studies and genetic databases are populated mainly by data from people of European descent. This knowledge gap exacerbates disparities in cancer incidence and outcomes around the world. In the United States, for example, African American men are about twice as likely as white men to die of prostate cancer.

But researchers who study these inequities say they are encouraged by renewed interest in closing the data gap from their colleagues and funders, including the US government. The issue was unusually prominent at the annual meeting of the American Association for Cancer Research (AACR) this month in Atlanta, Georgia — one of the world’s biggest gatherings of cancer researchers.

“It’s a historical year for us working in cancer health disparities,” says Laura Fejerman, a geneticist at the University of California, San Francisco, who studies breast cancer in Latina women. “We’ve been trying to show researchers who don’t work on health disparities that this is a really important issue.”

Numbers game

Differences in cancer risk and survival are thought to be caused by a complex mix of social, economic and genetic factors. The criteria used to select participants for clinical trials are often unintentionally biased against minority ethnic groups — for example, by excluding people with certain disorders that are more common in such populations. And members of these communities are sometimes distrustful of medical researchers, a legacy of past discrimination and studies conducted without adequate consent.

Then there is the simple matter of numbers: the rarer a cancer, the harder it is to enrol enough study participants from a minority population to gather statistically meaningful data. A crop of large studies is attempting to address this problem, Fejerman says. They include an effort that the AACR launched last year to sequence the genomes of tumours from 2,020 African Americans by 2020.

Another project, announced last July and about to start recruiting participants, aims to enrol 10,000 African American men recently diagnosed with prostate cancer. The US$26.5-million effort is funded by the US National Institutes of Health and the Prostate Cancer Foundation, a charity in Santa Monica, California. Led by genetic epidemiologist Christopher Haiman of the University of Southern California in Los Angeles, it will examine not only biological features of participants’ cancers, but also characteristics of their neighbourhoods and the social stressors — such as discrimination — they have experienced.

“People tend to think that the molecular features of a tumour are everything,” says Jennifer Doherty, a cancer epidemiologist at the Huntsman Cancer Institute at the University of Utah in Salt Lake City. “But then we don’t remember that the tumour exists within a human being.”

Breaking down barriers

How genomic data is collected and labelled is also becoming more nuanced, as researchers re-examine the broad ethnic categories they have long used. The term ‘Asian’ encompasses dozens of countries with disparate lifestyles and genetic backgrounds. Similarly, ‘African American’ is used to refer to US citizens of African descent, regardless of which region of Africa their ancestors called home.

That is a mistake, oncologist Olufunmilayo Olopade of the University of Chicago in Illinois told the AACR meeting. “Africa is a huge continent and you can’t just reduce it to one monolithic population,” said Olopade, who presented her work on African Americans in Chicago and Africans in Nigeria.

Mindful of such concerns, Fejerman has expanded her genetic studies of breast cancer to include Latinas outside the United States. She is putting together a consortium of researchers to study the disease in California and Latin American countries such as Peru. “It’s great for us and for the Latinas in the US because we are learning about a genome that they share,” she says. “But we are also giving something to the populations in Latin America that don’t have the resources.”

Despite such efforts, she worries that the cancer data gap will continue to widen. That would leave under-represented populations at a disadvantage as researchers pursue precision medicine, in which treatments are tailored to a person's genome and physiology. “The problem is that [Latinas] are so behind, so low in terms of numbers compared to women of European ancestry, that it’s going to take a long time and a lot of money to make it equal,” Fejerman says.

And with a problem as multifaceted as health-care disparities, there are inevitable debates about where funding and efforts should be focused. Candace Henley, a patient advocate in Chicago, finds preliminary genetic studies of minority groups frustrating. Why spend so much time and money digging through genomes when society has yet to address known causes such as discrimination and access to health care, she asks. “If these issues are not addressed, they will continue to cause disparities,” says Henley. “We still have a long way to go.”

Nature 568, 154-155 (2019)

doi: 10.1038/d41586-019-01080-2
Nature Briefing

Sign up for the daily Nature Briefing email newsletter

Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.

Sign Up