Harvard Medical School Researchers Develop Genetic Risk Scores for Six Common Diseases

A team of researchers from Harvard Medical School and its affiliated hospitals have developed genetic risk scores — which have the potential to aid screening and prevention in high-risk patients — for six common diseases and applied them clinically.

The study — published in Nature Medicine, a top peer-reviewed publication, on April 18 — was co-led by Harvard Medical School professors Jason L. Vassy and Matthew S. Lebo, and included researchers from HMS affiliates Brigham and Women’s Hospital and Veterans Affairs Boston Healthcare System.

The researchers analyzed the genomes of thousands of patients and developed polygenic risk scores, which use thousands of individual genes to predict the likelihood of a patient developing a given disease.

The methodology has widespread applications in detecting and preventing diseases, including common ones like atrial fibrillation, coronary artery disease, type 2 diabetes, breast cancer, prostate cancer, and colorectal cancer — the focus of the team’s study.

Vassy said polygenic risk scores could allow physicians to target the highest-risk patients for prevention and screening.

“I would say in the next 10 years, at least some polygenic risk scores of some shape and form will be regularly used,” he said.

Peter Kraft, a Harvard School of Public Health epidemiology professor and co-author, said the technique had previously been demonstrated in a research setting but was difficult to apply clinically.

“There’s no guarantee that what we saw in our research studies is going to work in a free-living human population,” he said.

Kraft said the team successfully proved the validity of their methods in a clinical setting, but still faced the challenge of helping physicians understand and implement polygenic risk scores in patient care.

“Primary care doctors in particular are just swamped these days — Covid notwithstanding — so they're not going to take a new genetics class or a new curriculum on polygenic risk scores,” Vassy said. “We wanted to be able to deliver the intervention in this research study in a very, very time-conscious way.”

Vassy said the use of polygenic risk scores in clinics also raises ethical concerns.

“The body of cohort studies in the world consists predominantly of European ancestry individuals, so these polygenic risk scores have been developed in those individuals,” Vassy said. “When you look at porting them over to cohort studies that consist of more diverse ancestry individuals, they don't perform as well.”

Anna C. F. Lewis, a research associate at Harvard's Safra Center for Ethics who co-authored the study, said while variance in the efficacy of polygenic risk scores is an “ethical problem,” the methodology should still be tested.

“The right thing to be doing is to test them and to measure how many of these concerns are in fact realized,” she said.

Kraft said enrolling more genetically diverse cohorts is the next step.

“The more diverse of a population we have to study, the better, because ultimately we would like to apply these scores and make sure they work for everybody equally well,” he said.

Correction: May 4, 2022

A previous version of this article misquoted Harvard School of Public Health epidemiology professor Peter Kraft. He said there is no guarantee what he found in research studies “is going to work in a free-living human population.” He did not say “pre-living human population.”