Women who learn that they have a mutation in the breast cancer gene BRCA1 face a wrenching decision. Their doctor or genetic counselor will likely tell them that women with such mutations have, on average, a 72% lifetime risk of breast cancer and a 44% risk of ovarian cancer. Given that, up to half decide to have prophylactic mastectomies, and many have ovaries removed, too.
But recent studies show a woman could receive a more individualized, accurate cancer risk estimate by factoring in other gene variants. A preprint posted last month finds that a person's "polygenic" background influences not only the disease risk conferred by a BRCA1 defect, but also risks from single gene mutations linked to colorectal cancer and heart disease. Some individuals were very likely to develop cancer or heart disease by age 75, the analysis showed, whereas in others the risk was not much greater than in a person without the high-risk mutation.
"It's pretty striking," says cardiologist and geneticist Amit Khera of Massachusetts General Hospital (MGH) in Boston, leader of the study, which is on the medRxiv preprint server. "It's become clear that there are both monogenic and polygenic [disease] drivers. The future is to assess both."
"The message is a very important one for patients and clinicians," says Teri Manolio of the National Human Genome Research Institute in Bethesda, Maryland. "Carriers of BRCA1 mutations or other pathogenic variants don't invariably develop disease, and genomics can be used to help parse carriers who are at lower risk." Others caution, however, that risk scores summing how dozens to thousands of other genetic variants interact with a single major disease gene aren't yet accurate enough to use in the clinic. The new paper "is teasing at the possibility, but there's a lot of work to be done," says Harvard University epidemiologist Peter Kraft.
MGH cardiology fellow Akl Fahed and others in Khera's group explored polygenic influences on the three important single-gene disorders in the United States: familial hypercholesterolemia, which leads to sky-high cholesterol levels and dramatically elevates risk of heart disease; Lynch syndrome, a flaw in DNA repair that brings a lifetime risk of colorectal cancer of about 60%; and inherited breast cancer, caused by variants in BRCA1 or BRCA2. They took advantage of databases that combine medical and genomic information from thousands of people, enabling researchers to tally how the many genetic variants with subtle effects modify disease risks and complex traits such as height.
Drawing on some 50,000 participants in the UK Biobank and 19,000 women tested for BRCA genes by the company Color Genomics, the team found that polygenic background strongly modified the risk of carrying a mutation in the key genes for the three disorders. For a small proportion of major disease gene carriers, other genetic variants boosted their overall risk of cancer or heart disease to about 80%, well above the average of 30% to 40% that Khera's group estimated for its study populations based on just the single disease gene mutations. (The team's monogenic disease risk predictions are lower than many other estimates for several possible reasons, Khera notes, including that the UK Biobank participants are healthier than the general population.) At the other extreme, the polygenic analysis suggested that a few people with those mutations have much lower risks than predicted by their single mutation alone, as low as 11% for colon cancer, 13% for breast cancer, and 17% for heart disease—not much higher than other people in general.
Khera's group says adding polygenic data to single-gene tests could help people decide whether to take aggressive steps to head off disease—mastectomy or removal of the ovaries for women carrying BRCA mutations or frequent colonoscopies for people with Lynch syndrome. But the new study does not include enough data for clinical decisions, says genetic epidemiologist Antonis Antoniou of the University of Cambridge in the United Kingdom. Only 116 women in the UK Biobank sample had BRCA mutations, which he notes "is an extremely small number to make inferences about risks."
Two years ago, Antoniou led a study that reported on how polygenic scores influence risks in 25,000 carriers of BRCA mutations and found nearly as wide a range of overall cancer risks. His team has incorporated those data into a breast cancer risk estimator along with factors such as family history.
The MGH study is "an important and exciting paper" that complements other work, says David Ledbetter, chief scientific officer for the Geisinger Health System in Danville, Pennsylvania. His team recently looked at 92,000 participants in an ongoing genomic medicine study called MyCode, focusing on those who carried mutations predisposing them to 11 rare disorders that affect traits such as height, weight, and cholesterol levels. Incorporating polygenic scores helped predict those traits, the group reported on 25 October in Nature Communications.
It may be a while before physicians are comfortable telling patients how genetic backgrounds modify the risk posed by a major disease gene mutation. But some companies already offer polygenic scores for cancer and other diseases, and tests that combine both kinds of information are imminent. Before insurance companies agree to pay for such tests, Ledbetter cautions, "They're going to want to see much more clinical validation"—including for minorities, because current polygenic analyses draw on data primarily from people of European ancestry.