In the largest study of its kind, a research team has meshed extensive genome data on more than 50,000 people with their electronic health records and identified potential new disease-causing genes. The data further suggest that about one in 250 people may harbor a gene variant that puts them at risk for heart attacks and strokes, yet aren’t receiving adequate treatment.
Several projects in Europe and the United States have amassed DNA from huge numbers of people and coupled the data with clinical information to search for links between genetic mutations and diseases and traits. But these studies have so far looked for common genetic markers, not the very rare variants that can have a much larger influence on disease risk. And not all of these studies are sharing the DNA results with participants.
To find rare disease variants and integrate patients’ DNA results into health care, Geisinger Health Care System in Danville, Pennsylvania, and Regeneron Pharmaceuticals in Tarrytown, New York, sequenced the exome—the 1% of the human genome that codes for proteins—of 50,726 Geisinger patients. The volunteers—mostly rural Pennsylvanians, 98% of whom had European ancestry and a median age of 61 years—had agreed to share their electronic medical records for a long-term health study.
Like previous studies, the researchers found that most individuals had a few mutations that likely disable one copy of a gene—there was a median of 21 rare predicted “loss of function” variants per person. Some are potentially of medical interest because the analysis of the patient medical records suggest that they influence some blood marker or other trait associated with a disease.
And 3.5% of participants, slightly more than the 2% expected based on smaller studies, had mutations in 76 genes that are clearly linked to disease, such as the BRCA1 breast cancer risk gene and others for heart disease, the team reports today in Science. These individuals are being informed of those results so they can adjust their health care if needed. “They’re things that are potentially life threatening,” says Michael Murray, director of clinical genomics for Geisinger.
The researchers then dug deeper into the clinical impact of three genes already known to contribute to abnormally high cholesterol levels, which can make a person prone to heart attacks and strokes at an early age. As expected, the 229 people with these variants for the disorder familial hypercholesterolemia had high “bad” cholesterol and were more likely than the average person to have had coronary artery disease. But only 58% were on statins, the standard drug for this disease, and fewer than half of these patients were getting an appropriate dose, the team reports in a second paper in Science.
The results suggest that it may make sense to screen the entire population for these cholesterol-raising variants, says cardiovascular disease geneticist Daniel Roden of Vanderbilt University in Nashville, Tennessee. “There’s a huge amount of interest in identifying these people.”
One limitation of the study is that it didn’t nail down any new disease genes that could lead to drug targets. That’s because the gene-disabling variants were rare—often found in only one or two individuals—and it takes more to make a statistically significant connection with a disease risk. “Even with 50,000, the numbers are very, very small,” Roden says.
Firming up the role, if any, of these variants will take approaches such as a precision medicine study championed by President Barack Obama that aims to enroll at least 1 million volunteers, and the Department of Veterans Affairs’s Million Veteran Program. These projects will also help fill out the role of rare disease variants in minorities—people missing from the current study.