Direct-to-Consumer Polygenic Risk Scores Move Towards FDA Regulation

In medicine, we have been discussing the promise of polygenic risk scores for some time now. Polygenic risk scores determine whether a person’s risk for a common condition, like coronary artery disease or depression, is higher or lower than others.

The hope is to use polygenic risk scores and typical risk predictors (such as age and family history) to identify people who would benefit from enhanced prevention and screening and, in turn, improve overall health outcomes.

But we aren’t there yet. More research is needed before polygenic risk scores will be adopted into everyday medical practice. There isn’t enough proof yet that they improve health outcomes, and inequities in genomic research mean current polygenic risk scores may not apply equally to people with all ancestral backgrounds.

Polygenic risk scores are available to consumers through direct-to-consumer genetic testing companies. People can also upload raw genetic data from a previous DNA test to a website for fast polygenic risk score calculations.

Experts are concerned about the accuracy and risks of these unregulated health tests. The FDA has recently moved to regulate some polygenic risk scores, which may help ease some of these concerns.

Calculating polygenic risk scores

Many common health problems, like type 2 diabetes, cancer, and arthritis, are caused by a mix of genetics and environmental risk factors. Polygenic risk scores shed light on how your genes contribute to your risk of developing common and chronic diseases. 

Polygenic risk scores, which are condition- or disease-specific, are based on genome-wide association studies (GWAS) data. They analyze many genetic differences (or variants) in a person’s DNA spread across the genome.

Some variants increase the risk of disease, while others decrease it. Each variant has a small impact. Combining these variants in a polygenic sum, can predict a person’s genetic risk for a specific disease.

Common genetic variants are small differences in our genes. Some contribute to traits like height or eye color. Others can impact our risk for diseases and other health conditions. Everyone has a unique set of genetic variants that make you, you. 

Large genome-wide studies have mainly focused on people of European descent, so predictive risk scores based on these studies may not be accurate for people from diverse populations. With a push for more diversity in genomic data, this is changing.

Complex algorithms form the basis of polygenic risk score calculations. Currently, no standard way exists to report on polygenic risk score results. Some scores provide an absolute risk of developing a disease. Other scores compare your risk to the average person or show where you stand in the population. 

If you aren’t familiar with polygenic risk scores, this dynamic explainer is an excellent descriptive visualization tool.

Communicating about health risks can be difficult. More work is needed to find better ways to communicate polygenic risk score results clearly and avoid misunderstandings.

Polygenic risk scores in practice

Researchers have developed, studied, and published many polygenic scores since the first one was created in 2007. Polygenic risk scores are not yet standard in clinical settings, so there are no formal guidelines for their use with patients. In 2023, the American College of Medical Genetics and Genomics (ACMG) issued a statement addressing the challenges and limitations of polygenic risk scores. The goal was to help healthcare providers better understand this new technology. 

In a clear and concise list of points to consider, the ACMG states:

• Polygenic risk scores do not provide a diagnosis; therefore, high-risk predictions do not guarantee disease development.
• People with a low predicted risk can still face an increased risk. 
• Because of a lack of data, polygenic risk scores may be a poor tool for people from some ancestral populations (e.g., non-Europeans). 
• Limited evidence supports the use of polygenic risk scores in clinical practice.
• Before testing, patients should be aware of test limitations and how the result can impact medical management.

The ACMG made these points for clinical tests, but they are also crucial for direct-to-consumer polygenic risk score tests. With DTC tests, consumers order the test and review the results on their own. This means consumers need to be able to read and understand test materials, like the report and supporting documents, without the help of a healthcare provider. This can only happen if the DTC test manufacturer clearly communicates the essential points.

FDA oversight

The FDA does not regulate wellness products that promote a healthy lifestyle and do not have medical implications. Current DTC polygenic risk score reports recommend people make healthy choices, such as not smoking and eating well, to reduce their risk for a disease; they don’t make medical recommendations or diagnoses. 

The FDA has not been regulating these tests, likely because they consider them low-risk wellness products. But this means nobody is making sure DTC polygenic risk score tests are accurate. And nobody is checking that the test materials are clear and inclusive for consumers. Until now.

In August 2023, a UK company called GENinCode submitted a 510(k) premarket notification to the FDA seeking clearance for their new CARDIO inCode-Score (CIC-Score) polygenic test.

CIC-Score can help assess a person’s risk of developing heart disease. Upon review, the FDA decided this is the first polygenic risk score test to have evidence-based medical implications–a first of its kind. This triggered the FDA to ask GENinCode to update their 510(k) submission to a De Novo Classification Request in November 2023.

The De Novo request will pave a new regulatory path for polygenic risk scores. If successful, GENinCode’s CIC-Score will become the predicate for, no doubt, a flood of DTC polygenic risk scores seeking FDA clearance to follow.

This move by the FDA is precisely what experts have been calling for. In August 2023, Jacob Sherkow and colleagues published an article in JAMA cautioning about the risks of unregulated at-home polygenic risk score tests and calling on the FDA to intervene and regulate these tests.

The authors discussed how consumers might misinterpret or misuse polygenic risk score results without guidance from a healthcare provider. This could lead to negative consequences.

The FDA dealt with similar issues when they cracked down on 23andMe‘s direct-to-consumer testing in 2013. This crackdown resulted in a new FDA regulation for Genetic Health Risk (GHR) Tests (21 CFR 866.5950). The FDA is addressing similar concerns about public safety and well-being regarding polygenic risk scores.

Regulating polygenic risk scores

Within the regulatory process, the FDA could enhance polygenic risk scores’ analytical and clinical validity— their accuracy, reliability, and ability to predict health outcomes. 

The FDA also emphasizes the importance of clear and effective test materials for consumers, requiring evidence that the average person can understand polygenic risk score reports and related materials.

One of the special controls in the GHR regulation, which is likely to be part of the polygenic risk score protocol as well, is a user comprehension study designed to show that naive users can understand the test information with a high level of comprehension. User comprehension studies for GHR products focus on five topical domains: test purpose, limitations, next steps, results, and other factors. The FDA also asks manufacturers to review potential comprehension issues related to race and ethnicity. This critical step helps to alleviate consumers’ (and their caregivers’) misinterpretations.

What’s next for polygenic risk scores?

To propel high-quality and medically actionable polygenic risk scores toward the clinic, researchers are focusing on five main areas:

• Increasing equity of polygenic risk scores across diverse populations.
• Building and validating combined scores, which include genetic factors, non-genetic factors (such as age, sex, family history, and lifestyle factors), and even biomarkers (such as cholesterol or blood pressure) into a single actionable metric.
• Developing evidence-based clinical guidelines.
• Defining user-informed standardized ways to report polygenic risk scores results clearly. 
• Prospectively studying polygenic risk score-guided health outcomes.

Chronic and common conditions affect many people. In the future, polygenic risk scores will aim to improve health outcomes by driving more effective preventive strategies, earlier detection, timely intervention, and targeted treatments.