Skip to main content
Premium Trial:

Request an Annual Quote

Broad Institute, Mass General Develop Risk Score Aimed at Preventing Early Heart Attacks


NEW YORK (360Dx) – A new polygenic score developed at Massachusetts General Hospital and the Broad Institute detects more people at high risk of early heart attack than current clinical blood tests, and could spot heart disease risk for many who are unaware of it.

The genome-wide polygenic score aggregates information from millions of common polymorphisms and identifies those who are genetically predisposed to coronary disease and having a heart attack, according to the researchers who worked on it.

Sekar Kathiresan, the director of the Center for Human Genetic Research at Mass General, who participated in developing the score, said that patients could use it to trigger life-saving change in behavior. "We've shown that there are actions you can take to mitigate that risk."  

One of those activities is to take statin medication and the other is to adhere to a healthy lifestyle, he said. "We've found that DNA is not destiny when it comes to heart attack because you do have control. On the other hand, for people who don't take the test, the first symptom could be a heart attack."

A polygenic score is a number that determines risk based on a quantitative distribution; a higher number signifies higher risk. It operates in a similar way to blood-based clinical chemistry tests that help a physician determine risk by measuring a patient's level of cholesterol. But that's where the polygenic risk score's similarities with existing tests end.

People who score high on a monogenic risk test could also be identified through standard testing, but individuals identified at high risk based on the polygenic risk score are not being picked up by the standard risk factors, such as high levels of low-density lipoproteins (LDLs), he noted.

While the monogenic model identifies 2 percent of patients with early heart attack, the polygenic model identifies "close to 20 percent," he said. The polygenic risk score is "relevant in many, many more people."

People who score high on the polygenic scale tend to have a slightly higher LDL than patients that are not at high risk, and they tend to have somewhat higher blood pressure, diabetes levels, and weight, he said. However, while each parameter can be slightly higher than normal, none stands out. The researchers have found that that these patients are generally unaware of the risk.

The research, described recently on the bioRxiv preprint server, and previously in the New England Journal of Medicine and other journals, could lead to a diagnostic screening test used by physicians or consumers within two years, Kathiresan said. He added that he and his colleagues are considering options to determine the best path to commercialization including collaborating with large test labs, specialty test labs, genetic test providers, or direct-to-consumer test companies.

In their study published in the New England Journal of Medicine, the researchers concluded that both genetic and lifestyle factors contributed to individual-level risk of coronary artery disease. "The extent to which increased genetic risk can be offset by a healthy lifestyle is unknown," they said.

They used a polygenic score based on DNA polymorphisms to quantify the genetic risk for coronary artery disease in 4,260 participants in the cross-sectional BioImage Study for whom genotype and covariate data were available and in three prospective cohorts — 7,814 participants in the Atherosclerosis Risk in Communities study; 21,222 in the Women's Genome Health Study; and 22,389 in the Malmö Diet and Cancer Study. They also determined adherence to a healthy lifestyle among the participants using a scoring system consisting of four factors — no current smoking, no obesity, regular physical activity, and a healthy diet.

In recent work, the researchers developed a genome-wide polygenic score that aggregates information from 6.6 million common polymorphisms and showed that the score can identify people with a fourfold increased risk for coronary disease.

"This is a very important body of work that is still unfortunately stuck in the research compartment, but which should be available to all interested individuals," Eric Topol, director of the Scripps Translational Science Institute, who was not involved in the development of the polygenic risk score, said in an interview. "Using inexpensive array genotyping, any person can find out his or her genetic risk of heart disease, and if high, we now know from [Kathiresan's] work, that this can be highly [and] favorably influenced by adopting a healthy lifestyle."

Kathiresan said that his team's work has been focused for more than a decade on understanding why some men under the age of 50 and women under the age of 60 get heart attacks.

"It is well known that if you have a heart attack at a young age, it's more likely to be related to genes," Kathiresan said. "Genetics plays a larger role in early heart attack."

Two genetic models describe why people are predisposed to early heart attack, he said. In the first, a monogenic model, a single gene mutation is sufficient to cause early heart attack. Researchers have been working on this kind of model for up to 40 years, he said, and they have found that it is successful in identifying about 2 percent of patients with early heart attack. These patients' genetic mutations are linked to high levels of bad cholesterol created by LDL.

"That leaves a large fraction of people with early heart attack who remained unexplained," Kathiresan said, adding, "We've figured out that some of the rest of that genetic risk is reflected by the additive effect of many DNA variants. Over the past decade or so, we've tracked down all those DNA variants that contribute to risk, and in the most recent work, we have been able to aggregate those variants into a polygenic risk score."

"The new insight we've had is that if you are on the high end of the polygenic score compared to everybody else, you have quite high levels of risk, and they often exceed that of the monogenic model," Kathiresan said.

He said that with his colleagues he is exploring a few routes to a commercial test, which appears to be within reach. A key factor is its potentially affordable price, he noted. "You don't need to do sequencing on people to get to a polygenic risk score," he said. "You need a genotyping array, such as what some of the consumer genetic companies, including 23andMe, are using."

As a result, the test could be made available for less than $100, he said. But while that is likely to be affordable, getting reimbursement for patients who take the test would require its endorsement by influential groups such as the American Heart Association. Should a consumer testing company decide to take it to market, it would need to obtain FDA clearance, he added.   

"Ordering a test is one thing, but getting it reimbursed by insurance companies exposes a different set of issues," he said, and that's why its price is important.

Topol noted that the Scripps Institute provides a free app — My Gene Rank — developed through a research initiative "to help people make these data and interpretation available, once they undergo 23andMe or equivalent genotyping. Furthermore, instead of willy-nilly statin use, the same genotyping [used in the polygenic risk score] can help people decide whether statins will likely be beneficial. All this genomic information is independent of classical risk factors like family history, high blood pressure, smoking, etc.," he said.

And while the polygenic risk score works well in identifying early heart attacks, it also can be applied to test older patients, Kathiresan said. "The same factors play a role in garden variety heart attack as well" in men and women who have reached an age where heart attacks tend to occur most often, he said.

Kathiresan also noted that his team has shown that, apart from heart attack, the test is useful in determining risk for breast cancer and severe obesity. The approach will provide a framework that can become a general basis for risk stratification across a range of diseases, he said.