GeneBites

Since the advent of direct-to-consumer genetic ancestry testing almost 25 years ago, the possibility of health-related genetic screenings — particularly in the form of genetic risk scores — has taken the world by storm. These polygenic risk scores are the result of analyzing the millions of miniscule changes across an individual’s entire genome, and adding up risk contributions to create a singular numerical score for a patient. This score can then be used to estimate a patient’s lifetime risk of developing conditions like breast cancer, heart disease, or type 2 diabetes.

Importantly, most polygenic risk scores have been primarily developed using data from people of European genetic ancestry, limiting both their accuracy and applicability to other diverse ancestry populations. Since there are general differences in genetic risk by ancestral population, the use of European ancestry-derived scores can cause issues in the clinic when applied to diverse patient populations.

In their recent Nature Medicine study, a collaborative group of researchers at the Broad Institute of MIT and Harvard attempt to address the issue of ensuring effective and equitable implementation of polygenic risk scores in healthcare settings. They worked with various academic medical centers across the US to implement select polygenic risk scores in patient settings using patients within the Electronic Medical Records and Genomics (eMERGE) network. 

In the paper, the researchers outlined and executed the processes required to assess polygenic risk scores across a diverse set of individuals, calibrate them based on patients’ genetic ancestry, and facilitate return of results to patients and providers. 

A key component of making risk score implementation equitable is ensuring these risk scores are accurate and precise for a genetically diverse range of individuals. Lennon and colleagues surveyed existing literature to refine polygenic risk scores, seeking those that had been validated across diverse genetic backgrounds. They also identified scores with modifiable disease risks, that is those amenable to medical intervention, screening, or individual-level lifestyle adjustments.

The team chose 10 conditions for polygenic risk score assessment: asthma, atrial fibrillation, breast cancer, chronic kidney disease, coronary heart disease, hypercholesterolemia, obesity, prostate cancer, type 1 diabetes, and type 2 diabetes.

They also describe the pipeline used to generate each individualized clinical report, which includes a qualitative framework to indicate diseases for which a high PRS was found. Among the first 2,500 individuals assessed using this pipeline, they identified 581 individuals with a high PRS for at least one of the ten conditions. Implemented equitably at scale, it’s easy to imagine how patients can take meaningful steps to improve their health having knowledge of their genetic predisposition to life-changing conditions.

In their return of risk score results, the researchers — in conjunction with academic medical center partners — crafted materials in their pipeline to inform patients and providers about what risk scores mean, how they were broadly constructed, who they’re applicable for, and additional details on actionability.

Healthcare providers and patients generally struggle to understand and communicate polygenic risk score results because of statistical obscurity and lack of training. Implementing effective disease prevention strategies post-results is a major challenge, too. Just obtaining results won’t be beneficial without efficient preventive measures or early detection strategies. While the eMERGE Network’s work offers some initial guidance, ongoing research will be crucial to ensure responsible and effective genetics-based risk assessment.

More studies are therefore required to see if using these results can help improve patient health through varied health promotion and disease prevention strategies. Most importantly, it will remain to be seen that these fancy risk scores result in improvements felt by individuals of all diverse backgrounds and in the most vulnerable groups of society. 

Edited by JP Flores (he/him)