This article has been updated to clarify the ACMG's position on preimplantation polygenic risk score testing.
SALT LAKE CITY – Members of the American College of Medical Genetics and Genomics professional practice and guidelines committee advised against the use of polygenic risk score testing at the present time to select embryos for implantation in a new policy statement about clinical applications of PRS testing.
According to the ACMG "points to consider" statement, published in Genetics in Medicine on Wednesday, preimplantation PRS testing "is not appropriate for clinical use and should not be offered as direct-to-consumer testing at this time."
The statement, presented here at the ACMG annual meeting during a session on Wednesday by first author Aya Abu-El-Haija, a medical geneticist at Boston Children's Hospital and Harvard Medical School, comes at a time when several companies in the US — including Genomic Prediction, Orchid Health, and MyOme — are either already offering or conducting studies on preimplantation genetic testing for polygenic disorders (PGT-P).
ACMG members are currently working on two separate policy statements that will elaborate on the use of PRS testing for embryo selection, according to Patrick Turley, a professor of economics and director of the Behavioral and Health Genomics Center at the University of Southern California, who presented his work on this topic during the same session.
Turley, whose team recently published results of a survey of the US public on polygenic screening of embryos that found more favorable views than expected, said that there is "a disconnect between scientific groups and the general public" on the use of PRS in embryos that needs to be addressed. "We should have had this conversation five years ago," he added.
Prospective parents opting to have their embryos tested, and selecting them based on the results, may not fully understand the risks and benefits, and the approach could be used to select for nonclinical traits including those that have been part of eugenic policies in the past. "We need to figure this out fast," he said, pointing to the work of the Polygenic Embryo ELSI Research Group.
Overall, the authors of the new ACMG statement remained skeptical of the clinical use of PRS testing in general, pointing to limitations of the scores and a lack of evidence that they actually have clinical utility and improve patients' outcomes. "At this time, the ACMG advocates against clinical implementation of PRS testing unless the provider and patient have a clear understanding of the limitations of the testing and applicability to the specific patient, including how the results will be used to guide evidence-based clinical care," they wrote.
Specifically, they pointed out that PRS don't provide a diagnosis but merely a relative genetic risk of developing a disease that doesn't account for other risk factors, such as environmental risks or monogenic risk from highly penetrant variants.
Furthermore, most PRS models to date are based on genome-wide association studies that were conducted in populations of European ancestry and may not be applicable to other populations. Also, there is some evidence that PRS results may differ by sex, for example in cardiovascular disease or psychiatric disorders, they noted.
There is a dire need for studies showing improved clinical outcomes from using PRS to guide clinical management, they wrote, and cost-effectiveness studies are lacking. While it might make sense to use the same strategies for PRS risk that are used for managing other types of risks for certain diseases — for example, measuring fasting glucose for diabetes risk — other strategies, such as prophylactic surgery for cancer risk, maybe be inappropriate for PRS risk. "Currently, there are no clinical guidelines available for the use of PRS," the authors wrote.
A separate ACMG statement on laboratories' perspectives in the development of PRS is currently in press with Genetics in Medicine and was presented during the meeting by lead author Honey Reddi, a professor of pathology at the Medical College of Wisconsin.
Among other points, she stressed that the limitations and uncertainties of PRS need to be clearly indicated in clinical reports. PRS models may improve over time, she noted, so polygenic risks may need to be updated for individual patients in the future. Looking forward, PRS, which are currently based on single-nucleotide variants only, may also include epigenetic information.