GMXV Meeting - Questions and Comments
On November 8-9, 2023, NHGRI sponsored its 15th Genomic Medicine meeting, Genomic Medicine XV: Genomics and Population Screening, in Bethesda MD. The following questions were answered during the meeting.
Effectiveness Models of PRS that use them in isolation, not considering the scores in the context of non-genetic risk factors as a part of genetically informed risk assessment, may not reflect the value of PRS. Careful evaluation of the populations of reference that gave rise to the PRSs used in a clinical population and considering any discordance in the population that gave rise to the PRS and the clinical population it has been implemented in... would be key to any empirically grounded evaluation of the clinical utility and value of a given PRS. If the PRS used in a cost effectiveness model was drawn from developmental PRS vs. a validated PRS would also make a large difference in how said PRS would perform.
Thank you for your comment.
Can the panel address the social structural ramifications of screening based on race and/or genetic ancestry?
Recognizing that genetic variants track imperfectly with self-identified race or background genetic ancestry, screening based on these factors will inevitably miss some persons at risk. Ideally, everyone would have genomic data available and screening would be based on actual variant data rather than these imperfect proxies. Until that occurs, however, if it ever does, there are cases for which targeting screening to certain groups may maximize efficiency and minimize cost, just as some screening is currently targeted to people with certain environmental exposures or family histories.
Can the panel address the impact on the kidney of long-term use of hypertensive meds, where in our effort to treat hypertension (coupled with disparities in access to foods, safe contexts to adopt lifestyle modifications that mitigate the BP control)? If we increase screening and even control BP better for those with genetic risk factors for CKD, how do we address disparities in transplant access and after care — ethically should screening come before our ability to discern clinically high-risk people? Or address the disparities on the back-end of the genetic result?
Long-term use of antihypertensive medications has repeatedly and overwhelmingly been shown to preserve renal function, regardless of other lifestyle modifications. Such modifications also provide additional benefits in multiple realms of health and well-being. The issue of whether individuals who are screened and found to have a manageable risk can then access the care they need to reduce that risk is a critical issue in the U.S. It underlies many of the health disparities we see in renal disease and other adverse outcomes of hypertension. Given the large numbers of un- or under-insured people in the U.S., disproportionately affecting those carrying some of the genetic risk factors for CKD, it is something that those who plan population screening programs have to think through in advance. Fortunately, many interventions to lower blood pressure, including medication and diet/lifestyle changes, are extremely low cost so the key issue will be ensuring that participants have access to follow-up care.
Scientific communications researchers will be key - in agreement with Dr. Nathanson.
Thank you for your comment.
FYI The PRS paper from CISNET that I mentioned https://pubmed.ncbi.nlm.nih.gov/32853342/
Thank you for your comment.
If we base our approaches on self-report, we need to ensure it is indeed ascription vs. assignment, also - there are socio-structural determinants of how people ascribe that have nothing to do with genetic ancestry and thus the two can't be conflated - (race/ethnicity and genetic ancestry).
Thank you for your comment.
I’m curious to hear panelists’ thoughts about the role that private companies have in pop screening implementation. For example, Helix is establishing pop screening programs in health systems throughout the country and establish a Helix Research Network.
Helix is working to prove the potential sustainability of the “sequence once, query often” model that pairs well with population screening. Private (laboratory) companies can play a role by developing products and services that solve shorter term problems that health systems, payers (public and private), employers and pharmaceutical companies have to solve, and in doing so, build an infrastructure that can support, de-risk, and de-burden population screening programs. It must involve systems-level thinking, focus on patient-centered value and outcomes, and pay attention to equity concerns to ensure that any program is built on real value. It will likely take private efforts to test different implementation models. Ideally, they would track implementation and patient outcomes and publish the results; this has been a stumbling block, but private companies such as Helix have published in the past.
Thank you for these presentations. Can the presenters address the social and actuarial consequences of EHR availability of genetic information?
The question seems to address the potential for genetic discrimination if Protected Health Information such as this becomes known. This had been a concern even before genetic variant information was available (based on family history), and we now have well over 10 years’ experience with placing genetic information in the EHR. There is as yet no empiric evidence that this information is being used in a way to negatively impact individuals in a social or insurance context, but lack of evidence of harm is not evidence of lack of harm. GINA would provide protection for health insurance and employment for genetic information placed in the EHR from population screening. However, those who remain concerned about genetic-based discrimination may still avoid genetic testing even if a family member has a known pathogenic variant. They can at least be advised to seek phenotypic screening (such as echo/ECG for cardiac conditions, mammography/colonoscopy for cancer screening) to manage their potential risk.
Just a comment on "motivating' people to engage in health behaviors. I think we need to consider carefully the way in which we approach this and consider our own biases. When we say that patients or families are not motivated that is harmful, as it comes from a very paternalistic view, can perpetuate ableism, and disregards the history of the harms that have been incurred by high-risk populations who bear the burden of disease in this country. Consider a few things, what are the patient and family goals? How does your patient/family define health? How can we elevate the voices of those who have been underrepresented and minoritized in science and healthcare? Consider how stating that patients/families are unmotivated is in fact blaming people for their illness and operates under the false assumption that all people have equitable access to healthy behaviors and treatment.
Thank you for your comment.
All of the presentations have significance to oppressed communities that are not federally recognized. How can the nuances of the AI/AN Research experts and the concern for respecting them, be extended to other URM groups?
It is important to remember that biological kinship relations do not stop at the end of a community’s legal jurisdiction — 80% of AI/AN people, for example, reside in urban and suburban areas, yet this is where much recruitment occurs. Another important issue is being careful and cognizant of community, regardless of legal recognition and norms and regulations about consent and data sharing. Regarding the research process, having a statement about group risk and group harm and how the study cannot protect against harms due to group inferential statements should be considered for inclusion as part of the informed consent process. Researchers should be cautious about the use of population descriptors and community-level demographics to avoid reifying race as biological constructs. Transparency, research accountability, and benefit sharing with URM communities should be integral to the research process. Researchers should also carefully consider how the benefits of research can be operationalized in a way that can be meaningful to the URM communities in question, as well as being transparent, especially in cases where genomic data becomes the commercial property of testing companies. Finally, there have been advancements made to data sharing and governance policies and models related to URMs, and there continues to be research in this process; using these new policies and digital data tools can ensure responsiveness to URM communities’ concerns.
There are many harms associated with the concept of genetic similarity, can those be addressed as well?
Potential harms from an excessive focus on genetic similarity, or genetic relatedness among individuals, include misrepresenting or misclassifying populations, or de-emphasizing differences across people when diversity is both informative and advantageous. There has been recent guidance about best practices for reporting and resources on understanding human genetic variation, such as the recent NASEM report on population descriptors and the NIH Curriculum Development Supplement Series. Ultimately, harms associated with genetic similarity should be described, quantified, and studied to determine both net benefit and approaches for reducing harms in an evidence-to-decision framework.
Last updated: January 18, 2024