Polygenic scores (PGS), sometimes referred to as polygenic risk scores (PRS), are a developing risk estimate tool used to determine personalized risk for complex conditions that are influenced by both genetics and environment, such as breast cancer. Historically, utilization of PGS in genetic testing has been discriminatory and inequitable across various ancestries, which likely exacerbates racial inequities. While genetic ancestry is biologically based, it can correlate with race (a social construct); therefore, inequities in ancestry-based data add to racial health care disparities. Until 2021, PGS for breast cancer was only available to cisgender women of self-reported European ancestry due to a lack of sufficient GWAS data to identify relevant SNPs among other populations. Events of 2020, including the murder of George Floyd, sparked the country’s short-term widespread awareness of, and engagement in, addressing racial inequality. The country’s reaction, combined with increasing pressure from many individuals in the genetics field concerning the racial inequality of PGS, resulted in some changes in reporting practices of PGS. Laboratories who previously offered this testing updated their test menus; some removed PGS testing, while other laboratories released updated versions.
Despite modifications, it has been demonstrated that PGS are still not equitable across ancestries. As genetics providers, we require transparency in marketing materials, equal discriminatory power across all populations, and demonstration by genetic testing laboratories of true commitment to reduction of healthcare disparities before use of PGS can be considered equitable and able to be used across ancestries.
In November of 2022, Hughes et al. published an updated PGS for breast cancer they call “multiple-ancestry polygenic risk score” or “MA-PRS”. The authors developed a breast cancer risk assessment with greater accuracy for cisgender women of non-European ancestry by adjusting the weight given to each single nucleotide polymorphism (SNP). This MA-PRS uses 56 ancestry-informative SNP markers to determine the patient’s proportion of African, East Asian, and European ancestry. It then weighs the 92 previously identified breast cancer-associated SNPs based on the relative proportion of each ancestry.
While we acknowledge that this methodology does improve the performance of PGS in the non-European population, MA-PRS still does not perform equally across ancestral populations and therefore remains discriminatory. In particular, based on Table 2 of Hughes et al., MA-PRS does not delineate between low and high-risk scores as well for individuals who are Black/African as compared to the other ancestral categories studied. Furthermore, utilizing only three SNP-informed ancestral categories likely fails to represent many Americans.
The National Society of Genetic Counselors (NSGC) and Wand et al. have recently published a Practice Resource on PGS which argues that “equitable access to polygenic scores is threatened by differential test performance across populations, differential capacities to support population-wide delivery of genetic services, and differential resources for [PGS] education or uptake of information in a population.” Similarly, there is a new statement on clinical application of PGS published by The American College of Medical Genetics and Genomics (ACMG) and Abu-El-Haija et al., which includes the need to “improve available data sets for populations with non-European ancestry and optimize analytic methods [of PGS] so that genomic risk can be accurately and equitably identified across all human populations.” While the MA-PRS attempts to ameliorate some of these disparities, we argue that significant barriers to equal access remain.
In addition to these concerns regarding equity and access barriers related to the MA-PRS, there remains a significant question regarding the clinical utility of PGS. Currently, the National Comprehensive Cancer Network (NCCN) guidelines expressly counsel against using PGS results for clinical decision-making due to a lack of proven clinical validity. Therefore, insurance coverage for any medical management based on an elevated PGS score is highly in question. Similarly, the NSGC Practice Resource states, “clinical utility of [PGS] remains largely hypothetical, with increasing research evaluating clinical outcomes.” Furthermore, “genetic counseling about [PGS] should be framed in the broader context….[PGS] often does not capture all genetic risk.”
Considering the remaining disparity in clinical validity among populations, the complexity of PGS results interpretation, the lack of demonstrated clinical utility, and the potential lack of insurance coverage, we argue that significant work from the genetics community is still needed in order for PGS to truly be equitable and clinically useful. We acknowledge that MA-PRS are a first step towards that goal, but additional improvements need to continue.
As laboratories continue to improve or develop PGS, we ask for the following:
- Transparency by genetic testing laboratories offering PGS.
- Is this PGS performing equitably across ancestries? If marketed towards diverse patient use but without actual equal performance this could be misleading at best, and potentially harmful to patients at worst.
- Is there clinical utility currently for this PGS? Providers should not be told that PGS will help with clinical management and qualifying for high-risk cancer screenings so long as NCCN and other governing bodies recommend against such.
- Validation and equal power across all populations.
- Who can use this PGS? Given the development of PGS for use in non-European populations, there should not be movement backward. All future PGS options should be available and validated in diverse populations.
- How well does this PGS perform in diverse populations? There should be equal power and validation across all ancestral groups; it should not perform better or worse for one group over another.
- Demonstration of true commitment to inclusion and equity for patients by addressing underlying barriers.
- What research and data is this PGS based on? Eighty-four percent of GWAS participants in cancer risk studies are of European ancestry. This GWAS data has been the foundation of all genetic testing (including PGS). We encourage researchers to foster a culture of transparency and trust with underrepresented populations with goals of obtaining ancestrally diverse representative data. Therefore, allowing for development of wholly new PGS and mitigating the need to reanalyze the currently available and ancestrally limited data.
- What relationships are involved? Who are the collaborators? Bias exists in many areas of medicine; limiting that bias should be done whenever possible. Collaboration with and funding for groups specifically focused on diverse experiences, such as patient advisory boards and community-based participatory research projects, should be prioritized.
- How are other barriers or health disparities being addressed by laboratories offering PGS? Health disparities in genetics, such as access to genetic counselors or germline testing and higher rates of variants of uncertain significance for patients who are from underrepresented populations, already exist. As mentioned, although race is a social construct, disparities of testing and healthcare based on ancestry further exacerbate racial inequities. True commitment to inclusion and equity does not stop at PGS. Rather, it is necessary to address across all areas of genetics and throughout other health care specialties.
If you agree, join us and please sign this petition to register your support for transparency, validation across populations, and true commitment to inclusion and equity from PGS producing laboratories. These are the opinions of the individuals listed below, and not their institutions.
[alphabetical order]
Fatima Amir, MS, CGC
Suzy Cahn, MMSc, CGC
Tiffiney Carter, MS, CGC
Hayley Cassingham, MS, CGC
Katie Church, MS, CGC
Jeanne Devine, MS, CGC
Jennifer Eichmeyer, MS, CGC
Lauren Gima, MS, CGC
Helen Kim, MA, MS, CGC
Katie Lang, MS, CGC
Heewon Lee, MS, CGC
Kelsie McVeety, MS, LCGC
Jessica Scott, MS, CGC
Stephanie Spaulding, MGC, CGC
Melissa Truelson, MS, LCGC
Natalie Vriesen, MS, CGC
Kristin Zelley, MS, LCGC
The DNA Exchange 