- Using data from genome-wide association studies, Massachusetts General Hospital researchers have derived polygenic risk scores that can identify individuals who are at clinically significant increased risk of certain common diseases
- In some cases, the scores identified a substantially larger fraction of the population at risk than would be identified by rare monogenic mutations
- 19.8% of 288,978 participants were at =3-fold increased risk for at least one of the five diseases studied
- Inclusion of polygenic predictors in clinical care would have important implications for risk communication to patients
- The data used to create the risk scores came primarily from people of European ancestry, and genome-wide association studies are needed for other ethnic groups
Rare gene mutations are known to confer higher risk of certain common diseases in heterozygous carriers. For example, a familial hypercholesterolemia mutation is associated with up to a 3-fold increased risk of coronary artery disease. Another example is the p.Glu508Lys missense mutation in HNF1A, which is linked to a 5-fold increased risk of type 2 diabetes.
However, for most common diseases, polygenic inheritance—involving many common gene mutations that have small effects—plays a greater role than rare monogenic mutations. In Nature Genetics, Sekar Kathiresan, MD, director of the Center for Genomic Medicine, Amit Khera, MD and colleagues reported that polygenic risk scores they derived can predict a given individual’s risk of certain common diseases, which has important implications for screening and prevention.
The researchers studied five diseases that have major public health impact: coronary artery disease, atrial fibrillation, type 2 diabetes, inflammatory bowel disease and breast cancer. For each disease, they derived several candidate genome-wide polygenic risk scores, based on statistics from genome-wide association studies (conducted primarily in people of European ancestry) and data from the 1000 Genomes Project (n=503 Europeans).
To validate the scores, the researchers used the UK Biobank, which contains aggregated genotype and phenotypic data on more than 400,000 people of British ancestry. The scores that performed best in the initial dataset from 120,280 participants were subsequently assessed in a dataset from 288,978 participants.
The researchers found that the best polygenic risk scores could identify substantially larger fractions of the population at risk of a disease than rare monogenic mutations do. For example, 8% of the people they studied had a genetic predisposition that conferred at least 3-fold increased risk of coronary artery disease. The corresponding percentage for familial hypercholesterolemia mutations is 0.4%, so the polygenic risk score identifies 20 times more people at risk.
Dr. Kathiresan’s group notes that polygenic risk scores can be readily calculated for all common diseases simultaneously, based on data from a single genotyping array. The genetic testing could be done any time after birth, long before it is possible to identify most risk factors (such as hypertension or type 2 diabetes for coronary artery disease). Thus, clinicians would have guidance about which patients need more intensive screening or preventive interventions.
Communication with patients about disease risk will warrant serious consideration, the researchers add. There is potential for confusion of relative risk and absolute risk, and the usefulness of genetic knowledge and the potential harms may vary with the disease and the patient’s stage of life. However, it may not be feasible or appropriate to withhold the information.
The researchers emphasize that most genetic studies are done in people of European ancestry. They urge the biomedical community to undertake genome-wide association studies in non-European ethnic groups so that all people will eventually have access to high-quality genetic risk prediction.
Learn more about the Center for Genomic Medicine
Refer a patient to the Mass General Heart Center