The final panel session, moderated by Greg Feero, professor in the Department of Community and Family Medicine at Geisel School of Medicine and faculty at Maine Dartmouth Family Medicine Residency Program, examined system-level barriers to the widespread adoption of genomics in health care. Topics discussed spanned raising patient awareness of genomics, getting patients tested, laboratory concerns, delivery of results, and medical management based on the results.
PROMOTING PATIENT ENGAGEMENT IN GENOMICS
Tshaka Cunningham, chief science officer and cofounder of Polaris Genomics and executive director of the Faith-Based Genetic Research Institute (FBGRI), discussed how FBGRI is working to empower minority communities with the knowledge needed to engage in and benefit from genomics and precision medicine.
For many in minority communities, trust in medical research is lacking because of a long history of discrimination and exploitation, Cunningham said. One well-known example of this is the U.S. Public Health Service Syphilis Study at Tuskegee that ran from 1932 to 1972 and had detrimental health effects on African Americans in rural Alabama.1 Another example is Henrietta Lacks’ tumor cells, taken without her knowledge or consent to establish the HeLa cell line,2 that has been used in the research and development of numerous medical products that have earned the pharmaceutical industry billions of dollars, with no acknowledgment or compensation to her family. In addition, African American patients have experienced, and continue to experience, “countless instances of health disparities.”
To promote trust in science and medicine, FBGRI is partnering with community and faith leaders to provide sustained outreach to underserved minority communities through ongoing in-person events at local churches
___________________
1 See https://www.cdc.gov/tuskegee/index.html (accessed January 9, 2023).
2 See https://www.britannica.com/biography/Henrietta-Lacks (accessed January 9, 2023).
and other community-serving institutions, as well as online resources. He pointed out that, for many African Americans, the church has long been considered a safe space where they can receive information from their pastors, community leaders, and other trusted individuals. FBGRI deploys “honest broker communicators,” including people of color from the communities and scientists of color to discuss the health benefits of genomics and precision medicine and to address concerns about participating in research (e.g., fears about how a patient’s data might be shared or used against them). “People really respond to people who are from their community; there’s an instant trust bond there,” he said. When people have a better understanding, there is greater engagement and participation. Another strategy for building trust in communities could be ensuring that the genomics infrastructure and industry support innovators who are people of color in developing new technologies and products. This, Cunningham said, could fill a current gap and could help individuals in these communities see themselves in all aspects of genomics.
In summary, Cunningham listed three opportunities for promoting trust and engaging communities in genomics:
- Highlight the health benefits that minority patients have seen thus far that are attributable to advances in science and biomedical research.
- Provide culturally congruent “honest brokers” that can be a trusted source of scientific information for African Americans and other minorities.
- Address racism at the highest levels in academia, industry, and government to create a truly level playing field for scientists of color in the United States.
GENOMICS IN FAMILY MEDICINE PRACTICE
Philip Zazove, professor emeritus and past chair of the Department of Family Medicine at the University of Michigan, shared his perspective as a family medicine physician on the main systems issues that are affecting the uptake of genomics in primary care and offered some potential solutions.
Quality of Care
Primary care providers have, on average, 15 to 30 minutes to spend with each patient to address the reasons for their visit, Zazove said. He noted that just covering the relevant preventive services recommended by the U.S. Preventive Services Task Force takes far longer than that. Additional time is needed for patients with multiple conditions and for patients
whose health is affected by social factors that need to be addressed. When it comes to integrating genomics into the encounter, “there is no time.”
To address this, Zazove said that new paradigms of care are needed that shift the focus toward population health and that emphasize the quality of care each patient receives rather than the number of patients each provider sees in an hour. He suggested that this shift in practice could result in having more time available to incorporate genomics.
Institutional and System Factors
Zazove reiterated that when institutions and systems emphasize volume (e.g., seeing as many patients as possible) and productivity (e.g., money generated) there is little incentive for spending time addressing issues such as genomics. Another barrier is that genomic data are not widely integrated into EHR systems. There are no pop-up notifications to alert the physician that the patient is at risk for a genetic disorder or to recommend appropriate genetic tests. Zazove said it can also be very difficult for the primary care providers to get information about the patient’s genetic history and test results. He noted that genetics clinics have staff dedicated to gathering and reviewing each patient’s genetic history and test results, and family medicine practices generally do not have the resources to do this.
One solution, Zazove said, is to incorporate identification of at-risk patients and recommendations for genetic testing into the EHR. Another is to implement tools such as the Inherited Risk Evaluation Tool (InheRET),3 developed at the University of Michigan, which facilitates risk assessment and identification of patients who should be tested. He also highlighted the need to provide financial incentives for health systems to implement the infrastructure needed to support genomics in primary care.
Scope of Primary Care
Zazove said most of what family medicine providers see in their practices are multifactorial conditions for which there are not yet genomic explanations or actions to be taken (e.g., diabetes, congestive heart failure, depression, hypertension). Practices also deal with the health effect of patients’ socioeconomic and environmental conditions and the effects of having multiple diseases. On average, only about five patients out of a typical panel of 2,000, for example, would have genomic findings associated with BRCA or Lynch syndrome. “There is no real solution here,” Zazove said, other than to “recognize that what we see in family medicine and primary care is very different than what the specialists see.” What might
___________________
3 See https://www.inheret.com/Pages/Home# (accessed December 14, 2022).
work for specialist practices does not necessarily work for primary care practices, he said.
Complexity of Primary Care
“Primary care is incredibly complex [and] not well understood by most,” Zazove said. The breadth of services that primary care practitioners provide every day makes it very difficult to stay current on the many advances in the field of genomics.
Again, a solution here is to add functionality to the EHR to automatically identify at-risk patients and provide guidance on testing, he said. He also noted that residents receive training in genomics and often have more up-to-date knowledge than many faculty, and he suggested using resources such as the Society of Teachers of Family Medicine to ensure that providers stay up to date.
ISSUES FOR THE LABORATORY
Karen Kaul, chair of the Department of Pathology and Laboratory Medicine at NorthShore University HealthSystem and clinical professor of pathology at the University of Chicago Pritzker School of Medicine, said DNA- and RNA-based testing have “revolutionized” the practice of medicine. For example, somatic genomic testing of cancer cells aids diagnosis and informs treatment decisions, and germline testing has many applications, including diagnosis of genetic disorders, assessment of risk of heritable cancers, and pharmacogenomics. Molecular diagnostics have transformed microbiology, and she mentioned COVID-19 molecular diagnostics as an example. Although progress in genomics has advanced patient care, there are systemic barriers to the uptake of genomics, and Kaul discussed these challenges and potential solutions from a laboratory perspective.
Education of Clinical Providers
A challenge for the medical education system is that the field of genomics is advancing very rapidly. Medical school and residency curricula are not keeping pace, and future providers are not being sufficiently trained in these areas. Kaul offered several solutions, which she said should be implemented broadly and equitably. These included ongoing training for practicing providers, training champions in primary care (e.g., a hub-and-spoke model of expertise) and developing EHR-embedded tools that enable providers to incorporate genomic testing into routine care.
EHR Functionality
As discussed by previous speakers, genomics is complex, and greater functionality in the EHR could help overcome some of the practice barriers and support research. EHRs need to contain structured genomic data, Kaul said, and should facilitate the identification of family groupings and connections. Ordering genomic tests is also complex, and EHRs should enable ordering of the appropriate test and any necessary reflex testing and facilitate any required consent and preauthorization. The display and accessibility of genetic testing results present barriers as results are generally scanned into a PDF file that is uploaded to the EHR. Results need to be “intelligible, actionable, and ideally, make them live,” she said. For example, NorthShore University HealthSystem has built a just-in-time pop-up function into its EHR that alerts providers ordering a drug if there are relevant pharmacogenomics data available. The EHR should also alert providers to the existence of results for germline testing to eliminate unnecessary reordering of these “once-in-a-lifetime” tests. EHR vendors are working to incorporate these types of tools, she said, but needed functionality is still generally lacking.
Reimbursement of Testing
One systemic testing-related barrier that has been discussed, Kaul reiterated, is the need for reimbursement of the time spent by genetic counselors, primary care providers, or other practitioners to explain the testing and results to patients. Laboratories also face reimbursement challenges as reimbursement rates for genetic testing are often low. Tests are frequently considered to be experimental, and reimbursement is denied. Some payers only reimburse tests that are recommended by clinical practice guidelines, but these guidelines often lag behind practice. Kaul also noted that, although CPT codes have been added to the clinical laboratory fee schedule, there is no reimbursement for the interpretation of the sequencing results by laboratory professionals.
Regulatory Issues
There are also regulatory concerns for laboratories that conduct genomic tests. The number of FDA-approved molecular diagnostics for microbiology is increasing, Kaul said, which she said helps to combat the type of fraud discussed by Quinn. However, most genetics and cancer testing are still done using tests developed by each individual laboratory. Clinical laboratories, and laboratory-developed tests (LDTs), are regulated by CMS under the Clinical Laboratory Improvement Amendments (CLIA)
Act. The level of FDA oversight of LDTs has long been debated, Kaul said. Most recently, the Verifying Accurate Leading-Edge IVCT Development Act of 2022 would require that all laboratory testing be reviewed by FDA, which Kaul said could increase the cost of LDTs (which she noted are already expensive) and reduce patient access. She explained that FDA regulates laboratory test kits, while CLIA assesses the quality of the entire laboratory testing process (which might include a kit) and having FDA review LDTs would be redundant with CLIA. Discussions are ongoing, and she said that regulatory issues need to be considered as genomic testing becomes more widely implemented in routine patient care.
THE COMPLEX IMPLICATIONS OF GENETIC TESTING
Kara Maxwell, assistant professor of medicine at the Perelman School of Medicine of the University of Pennsylvania, discussed several broad system-level barriers to the uptake of genomics using cancer genetics as a case example. Management of patients with identified genetic risks has become increasingly complex and, she said, “There is not a one-size-fits-all approach.”
Patient Genetic Knowledge and Testing Venues
Patients can receive genetic test results that indicate an increased risk for cancer in a variety of care delivery settings, Maxwell said, and it is important to recognize that the level of knowledge the patient comes away with varies with the setting.
In a traditional genetic testing setting for cancer risk, patients receive pretest counseling, relevant tests are ordered, and results are explained in posttest counseling. This is facilitated by a genetic counselor, and possibly a cancer genetics physician, which Maxwell noted are providers that have the most knowledge to impart in this area. A patient might also be tested at the point of care. For example, the patient’s oncologist orders the test and delivers the results, possibly with the assistance of a genetic counselor.
Findings of a mutation associated with increased risk of cancer can also be found incidentally, such as germline mutations identified during reproductive planning in association with a prenatal geneticist. Tumor testing ordered by an oncologist and intended to inform treatment decisions can also turn up incidental findings. Incidental findings of increased cancer risk are also returned to participants in research studies, with no particular provider associated to impart information about the findings. Finally, direct-to-consumer genetic testing presents new challenges for imparting knowledge as it provides results directly to the individual who was tested.
Medical Management of Genetic Testing Results Is Complex
Genetic testing for cancer risk is not simply about assessing if a patient is at increased risk; it is also about what can be done to mitigate that risk, Maxwell said. Medical management of cancer risk associated with a genetic mutation varies in complexity over the life span and is relative to the patient’s overall health status. For example, Maxwell said that a woman who is a BRCA mutation carrier, but who is otherwise healthy, might continue to have breast cancer screenings across her lifetime but would also be considered for surgical interventions such as risk-reducing salpingo-oophorectomy between ages 35 and 45. Preimplantation genetic diagnosis would be relevant for persons of childbearing potential with known genetic cancer risks. Pancreatic cancer screening would be most relevant for those over age 55 who are at increased risk for hereditary pancreatic cancer. Urgent decisions regarding possible cancer treatment must be made at all ages. Discussions of addressing barriers should consider not just barriers to testing, but to providing appropriate care following testing.
Genomics Expertise Varies Among Clinical Providers
Patients can receive care for their genetic concerns from a range of providers with varying levels of genetic expertise. Expertise stems not only from training but from clinical experience. For example, Maxwell said that a genetic counselor is trained in genetic risk and counseling, and between 5 and 50 percent of their patients will have a genetic mutation (depending on the context of their practice). An oncologist is trained in the clinical management of cancer, and depending on the tumor type, between 1 and 20 percent or more of their patients will have a genetic mutation. Primary care physicians are trained in clinical management focused on risk reduction. They see patients with a wide range of conditions, and generally less than 1 percent of the patients they see will have a genetic mutation (can be higher depending on the ethnicity of the practice population). It is not realistic to expect that a primary care provider would have the same level of expertise as a genetic counselor because they do not see as many patients with genetic mutations in their practice. “We need to rely on the levels of expertise that people have, and make systems fit their expertise, not the other way around,” Maxwell said.
DISCUSSION
Telemedicine
Panelists discussed using telemedicine to facilitate access to genetic services. One challenge for telemedicine is the need for genetic counselors to be licensed in the state where the patient is. While not easy, Maxwell said, it is possible for genetic counselors to be licensed in all states. She noted that all counselors in the telegenetics program at the University of Pennsylvania are licensed in all 50 states. Telehealth has enabled many people to get care they could not have received otherwise, Zazove said. He noted that his primary care practice in Michigan is 35 minutes from the Ohio border and many of their patients come from Ohio; however, the practice cannot conduct telehealth visits with patients in Ohio, which presents a problem.
Another challenge noted was that patients often need in-person physical evaluations. Maxwell said this is an infrequent barrier as many conditions do not require a physical exam. She emphasized that the occasional need for a physical exam should not stop the deployment of telegenetics care, especially to rural areas. Zazove agreed, and said only 5 to 10 percent of the physical exams they conduct reveal an abnormal finding. Further, once a patient has had an exam, and testing has been ordered, follow-up counseling can be done via telemedicine. Kaul and Cunningham reiterated the importance of telemedicine for genetic counseling, highlighting concerns about the shortage of genetic counselors. Telemedicine can be used for both pretest counseling and posttest review of results, Kaul said, though the training pipeline of genetic counselors will need attention.
Direct-to-Consumer Genetic Testing
Direct-to-consumer genetic testing is extremely popular, Irons said, and people are presenting at their primary care practice with concerning results in hand (e.g., a cancer gene mutation) wanting to know what to do. Or alternatively, they receive results that could instigate preventative action, but they do not tell their provider and later need treatment for an advanced-stage disease.
Maxwell expressed her concern that people (including some physicians) do not understand direct-to-consumer genetic testing, and misinformation abounds. Test panels vary, and people often do not understand which genetic variants are included in the test they took, and which are not, or that the lack of a finding (“being negative”) does not mean lack of risk. Maxwell shared her concern that regulation of direct-to-consumer genetic
tests is limited.4 As a provider, “If you haven’t seen a report, you have no idea what your patient has been tested for,” she said.
Zazove said he commonly sees patients who are worried because they took a genetic test and they “have the gene.” In the vast majority of cases, the findings are not significant, but they cause a lot of anxiety. He suggested there could be a clearinghouse where people could easily get free information about their test results. Cunningham described similar personal experiences, and said better education about genetic testing is needed. He mentioned that the UK National Health Service (NHS) has a genomics educational program for providers.5 He also reiterated the need for more genetic counselors and the increased use of telemedicine for counseling.
Irons recalled a prior roundtable workshop on direct-to-consumer testing6 and relayed a story shared by an individual who found out she was a BRCA1 mutation carrier on a Friday afternoon and “spent the weekend with Dr. Google” because she could not get in touch with her doctor. By Monday she was extremely anxious, but her physician referred her to a genetic counselor who was able to give her information about her results. Engaging Google and health systems could be one strategy for ensuring that helpful information about genetic testing—and what to do when you get your results—is available to people searching for it online, Irons suggested.
Maxwell said that the University of Pennsylvania does have information posted online, as do other research institutions. Unfortunately, she added, it is not possible to ensure that the people searching get to the most reliable websites. Kaul cautioned that providing more information on the Internet also does not eliminate the risk that a person will assume they are not at risk because they are “negative.” A participant suggested that direct-to-consumer testing is “an opportunity for people to enter into genetics,” and presents an opportunity not only for patient education and engagement but also for provider education as well. Feero stated, “In many ways [consumer genomics] forced conversations about accuracy, clinical validity, [and] clinical utility that needed to happen,” as well as the regulation of genetic testing and how data moves around in the system.
___________________
4 For more information on FDA oversight of direct-to-consumer genetic testing see https://www.fda.gov/medical-devices/in-vitro-diagnostics/direct-consumer-tests (accessed January 4, 2023).
5 For more information on the Health Education England Genomics Education Programme see https://www.genomicseducation.hee.nhs.uk/ (accessed January 4, 2023).
6 See https://nap.nationalacademies.org/catalog/25713/exploring-the-current-landscape-of-consumer-genomics-proceedings-of-a (accessed December 14, 2022).
Secondary Uses of Genetic Data
Concerns about secondary uses of individuals’ genetic data for research or commercial product development were discussed. In his moderator role, Feero pointed out that the Geisinger Health System entered into a research partnership with Regeneron that he said was novel and very transparent. He noted that secondary uses are often not as transparent.
Cunningham emphasized the need for diversity in clinical studies and in biobanking of samples for future uses. A look back at the history of genetic research shows that much of it was not conducted with diversity goals in mind, he said, and the clinical studies done were often not representative of the population. As a result, variants of significance for different populations, especially minority populations, are not known, and many of the genetic tests on the market are therefore not broadly applicable. The NIH All of Us program emphasizes diversity in its data collection, as does the UK Biobank, and he called upon the pharmaceutical research community to commit resources to building a more inclusive data set. Maxwell highlighted the foresight of the Department of Veterans Affairs (VA) health care system in establishing a diverse DNA biobank in 2010, the Million Veteran Program. Thus far, this program has done single nucleotide polymorphism genotyping and whole exome sequencing for over 800,000 veterans, and within the program, 25 percent of participants are self-identified racial minorities, she said. Cunningham agreed, adding that VA was very intentional in designing an outreach and recruitment strategy for minority veterans. He added that the lesson learned here is “If you really put in the time and effort, you can see the results.”
Identifying Models of Success at the Systems Level
In addition to the VA’s Million Veteran Program, panelists discussed whether there are other examples of successful integration of genomics into clinical care at the systems level that could be shared. Zazove suggested the InheRET program is one such example, which was discussed above. Zazove and Cunningham said they were unaware of other examples at the health systems level, although there likely were some. Maxwell mentioned the Flatiron7 model of engaging multiple health care systems in integrating their oncology data for real-world clinical trials. She noted that there are institutional review board challenges to be addressed when combining data across health care systems. There are also barriers to sharing data across health care systems, and she again raised the issue of genetic and genomic
___________________
7 Flatiron is a company that aims to learn from the experiences of individuals with cancer. See https://flatiron.com/about-us/ (accessed February 6, 2023).
data being stored in Portable Document Format (PDF) form. She suggested that natural language processing and artificial intelligence technologies could be used to extract data in PDFs and integrate it directly into the EHR in a format that could facilitate sharing across systems.
Opportunities for Incorporating Genomics in Clinical Care
Panelists and members of the roundtable suggested a variety of actions that could help promote the adoption of genomics in health care and precision medicine.
- Enabling multistate telemedicine licensing. Maxwell said state lines should not be barriers for telegenetics and suggested developing a road map to guide practices through the process of obtaining licensure across multiple states.
- Increasing genetic and genomic literacy. Cunningham highlighted the need for “genomics education for all,” emphasizing that educational initiatives should be equitably implemented. He suggested that the National Academies could have a role in this, and that federal agencies should support educational initiatives. One approach, for example, is the training of community ambassadors to serve as honest broker communicators about genomics.
- Supporting physicians in implementing genomics in practice. Zazove said that health systems need to support the implementation of genomics in clinical care by increasing genomic functionality in the EHR and by adopting models of care that allow physicians to spend the time needed to equitably incorporate genomics into patient care and help patients get the services they need. This could also enable systems to reach out to undeserved communities who typically do not benefit from genomics advances.
- Partnering with patients and families. A participant said health systems should engage patients in an advisory capacity on the development of genomics programs and population sequencing efforts. She emphasized the value of patients sharing their stories, and the importance of developing “a true reciprocal relationship” that provides value back to patients and communities. Cunningham agreed that community participation is extremely valuable, but he noted again the need to build trust as concerns persist about how data could be misused. It is important to provide tangible benefits back to the community, and there can be a role for compensation. He mentioned public benefit corporations as a potential model, explaining that participants are awarded shares for their data contributions and can profit financially if their data are ever sold. Goto
- said that family organizations are valuable resources that should be engaged. She described how she worked with the Alaska state affiliate of the national Family Voices organization to develop a curriculum to train navigators in rare care. Partnerships with these organizations could help to disseminate information about genetics and genomics to families in communities. She also noted the need to understand the everyday issues families face, such as helping aging parents navigate the health system and the growing population of children who are cared for by grandparents. Even the routine task of using a kiosk to check in for laboratory testing can be challenging.
- Expanding the availability of information on rapid actions to take when patients receive testing results. Irons pointed out that the American College of Medical Genetics and Genomics (ACMG) developed ACTion (ACT) Sheets decades ago to advise clinicians on immediate actions to take when a genetic condition is identified through newborn screening.8 ACT Sheets are also now available for genomic findings from other types of genetic testing. She suggested that this type of model could be adapted to cover incidental findings patients receive from direct-to-consumer products or other genetic tests. These could be useful to both the patient and their physician. Martin said ACMG has issued ACT Sheets for some secondary findings and has worked in partnership with the ClinGen actionability groups. She said that synergy between these groups is important as it minimizes duplication of efforts. Maxwell agreed and said that the experts who serve on the individual ClinGen panels are passionate about their disease area and would likely be willing to help and could easily map out the rapid actions to be taken when someone receives a genetic testing result.
___________________
8 For more information on ACMG ACT Sheets see https://www.acmg.net/ACMG/Medical-Genetics-Practice-Resources/ACT_Sheets_and_Algorithms.aspx (accessed January 4, 2023).
This page intentionally left blank.
