Genomic medicine is defined as the routine use of genomic information about an individual as part of his or her clinical care as well as the health outcomes and policy implications of that clinical use. It is one approach that has the potential to improve the quality of health care by allowing practitioners to tailor prevention, diagnostic, and treatment strategies to individual patients (NHGRI, 2018). In recent years, research breakthroughs, technological advances, and the decreasing cost of DNA sequencing have led to the wider adoption of genomic medicine. However, as with the introduction of new technologies into health care, there are concerns that genetic and genomic testing and services will not reach all segments of the population both now and in the near future, and there remains a gap in knowledge regarding potential health care disparities2 in genomic medicine and precision health approaches, said Cathy Wicklund, a workshop co-chair, the director of the graduate program in genetic counsel-
1 This workshop was organized by an independent planning committee whose role was limited to the identification of topics and speakers. This Proceedings of a Workshop was prepared by the rapporteurs as a factual summary of the presentations and discussion that took place at the workshop. Statements, recommendations, and opinions expressed are those of individual presenters and participants, and are not endorsed or verified by the National Academies of Sciences, Engineering, and Medicine, and they should not be construed as reflecting any group consensus.
2 A health care disparity as defined by the Institute of Medicine report Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care is a difference in treatment provided to members of different groups that is not justified by the underlying health conditions or treatment preferences of patients (IOM, 2002, pp. 3–4).
ing at Northwestern University’s Feinberg School of Medicine, and the past president of the National Society of Genetic Counselors.
Two of the major obstacles to accessing genetic testing and the corresponding downstream care are the potential for high out-of-pocket costs to patients and a primary care workforce that may be unprepared to deliver genomic medicine to patients (Hauser et al., 2018). In addressing these challenges, it may be useful to explore how public and private payers design health insurance benefits and make decisions regarding the reimbursement of genetic testing and counseling, which could point to ways to help lessen any disparities in access to genomic medicine. To begin with, making genetic and genomic services accessible and operational in all primary care settings, including clinics that are under-resourced and serve diverse patient populations, may help avoid worsening health care disparities (Mikat-Stevens et al., 2015). Another issue is poor communication between health care providers and the community, which helps create an access barrier and can contribute to a fear of genetic and genomic testing among patients who are confused about and lack awareness of this aspect of medicine (Rahman et al., 2013). To address these issues and provide equitable access to genetic testing and downstream care, it may be useful to strengthen relationships between primary care providers and genetics specialists. This can lead to an increase in appropriate referrals to genetics specialists when necessary and therefore foster patients’ trust in the usefulness of genetic medicine (Rahimzadeh and Bartlett, 2014). Other potential solutions are providing better access to reliable patient and provider educational materials at the point of care and developing continuing education programs for primary care providers that emphasize genomic medicine (IOM, 2015).
On June 27, 2018, the Roundtable on Genomics and Precision Health hosted a public workshop to examine the gaps in knowledge related to access to genomic medicine and to discuss health care disparities and possible approaches to overcoming the disparate use of genomic medicine among populations.3 Workshop topics included research on access to genetic and genomic services in medically underserved areas, model programs of care for diverse patient populations, and current challenges and possible best practices for alleviating health care disparities as they relate to genomics-based approaches. The workshop brought together diverse stakeholders to offer their perspectives and participate in workshop discussions.
Disparities in access to genomic medicine is a profoundly complex issue that affects many populations, including underrepresented minorities, rural communities, medically underserved groups, and others, said Wicklund. The topic of disparities in access to genomic medicine is one that has come up repeatedly in discussions of the Roundtable on Genomics and Precision Health, she added. The workshop planning committee decided to focus the day’s discussions on exploring gaps in knowledge about health care disparities among these different populations, illuminating the challenges, and identifying possible ways to address those disparities and move the field forward instead of focusing on challenges related to research. The planning committee recognized that there is a lack of genomic data on ancestrally diverse populations in existing databases, Wicklund said, and that this creates major challenges for interpreting genetic test results. Because several research consortia are already trying to address this issue, the planning committee decided to not make it a focus for discussion during this workshop. Instead, she said, the workshop was intended to explore access to genetic and genomic services in the clinical realm and the barriers that affect a variety of populations.
Consider the challenges and questions that still need to be answered regarding disparities, said Vence Bonham, the other workshop co-chair and a senior advisor to the director on genomics and health disparities at the National Human Genome Research Institute. One desired outcome from this workshop, he said, will be the ability to identify needed research in this area. Prior to the workshop—on June 21, 2018—a Twitter chat4 was held to help identify some of the themes that might be important for the workshop discussions, Bonham said. The questions posed during the Twitter chat are shown in Box 1-1, and a summary of the discussion can be found in Appendix A.
Twitter participants identified several themes related to the challenges and barriers patients face when trying to access genetic services, including social and language barriers, provider training and education, workforce limitations, patient awareness, and privacy and potential discrimination, particularly as it pertains to insurance coverage, Bonham said.
When asked to identify approaches to increasing access to genetic services in medically underserved and minority communities, participants in the Twitter chat suggested that technologies such as telemedicine could help, as would building trust, education, and engagement with the public; increased insurance coverage for genomic testing; and the involvement of
4 Roundtable on Genomics and Precision Health Twitter Chat, #genomicsdisparities, http://www.nationalacademies.org/hmd/Activities/Research/GenomicBasedResearch/2018-JUN-27/twitter-chat.aspx (accessed August 16, 2018).
public health agencies. Twitter chat participants also suggested that partnerships among health care systems, community-based organizations, and community health workers could lead to better access to genomic services, as would efforts to increase health literacy. Drawing on lessons learned from the adoption of new technologies in other areas of medicine, participants emphasized the importance of collaboration and partnership, data collection, communication, and fostering a diverse workforce as potential opportunities for increasing access to genetic services, Bonham said.
This Proceedings of a Workshop summarizes the presentations and discussions that took place at the workshop. The day began with a keynote lecture from Otis Brawley, the chief medical and scientific officer at the American Cancer Society and a professor of hematology, medical oncology, medicine, and epidemiology at Emory University. Dr. Brawley’s presentation is summarized in this introductory chapter. Chapter 2 examines some of the barriers to accessing genomic and genetic services as experienced by individuals, patient advocates, and a provider who serves a rural area. Chapter 3 addresses the role of health systems in delivering equitable access to genomic medicine. Next, Chapter 4 explores ways in which providers can make genomic medicine more accessible to their patients. Chapter 5 discusses some innovative solutions and models of success for delivering genomic medicine to all, including medically underserved populations. Chapter 6 identifies unmet needs for alleviating health care disparities in
genomic medicine and describes some potential action items proposed by individual workshop speakers and participants.
Brawley began by defining genetics as the study of heredity—how the characteristics of living things are transmitted from one generation to the next (Advisory Committee on Health Research, 2002). Genomics is the study of genes, their functions, and interactions. The main difference between these two terms, he said, is that “genetics scrutinizes the function and composition of a single gene, whereas genomics addresses all genes and their interrelationships in order to identify their influences on growth and development of an organism.” It is common today in the discussion of genetics and genomics, he added, to hear about various mutations affecting the risk of developing a disease, as in the case of BRCA1 and BRCA2 mutations and the associated risk of cancer. What may not be appreciated, however, is that genomics is now helping to define sub-types of disease as well as patients’ response to treatment, Brawley said.
The study of genomics has led researchers to identify numerous molecular targets that underlie cancer’s growth, some of which can be targeted by drugs and are the basis of precision medicine, Brawley said. In the mid-19th century, the German pathologist Rudolf Ludwig Karl Virchow provided the first definition of cancer as a disease of uncontrolled cell growth and the subsequent spread of cells which then interfere with bodily functions. Today’s view of cancer still includes the histological changes that Virchow observed under a microscope, but it also takes into account the biologic behaviors of cancer cells and their genomic signatures as well. For example, Brawley said, as recently as 30 years ago non-small cell lung cancer was considered, based on its histology, to be one disease, but researchers today estimate that it may be as many as 80 different types of lung cancer, as defined by genomics.
Similarly, a genomic test called Oncotype Dx looks for mutations in an array of 21 genes involved in estrogen receptor-positive breast cancer, with the results predicting the aggressiveness of a cancer and serving as a treatment-guiding profile. “We are now using genomics to actually determine how someone should be treated for lung cancer, breast cancer, and other diseases,” Brawley said. The results of the Oncotype Dx test, he said, could potentially inform 70,000 American women per year that they would not benefit from treatment with chemotherapy, thus sparing them its inherent dangers and potentially severe side effects (Sparano et al., 2018). The Oncotype Dx test, Brawley said, costs about $4,000. Other genomic tests
similar to Oncotype Dx are also expensive, which means that some cancer patients do not have the option of taking these tests.
Overall, the American health care system is inefficient, Brawley said. Some individuals consume too many of its resources—which can result in inefficiencies and sometimes even adverse effects—while others consume too few resources, which can lead to disparities in outcomes and overall health. Specifically, one issue with an inefficient health care system is that individuals who have poor access to health care tend to have worse clinical outcomes, and some may even die as a result of receiving no care or poor-quality care, Brawley said.
Given this challenge, one of the important questions is how to provide high-quality care to people who often do not receive it, Brawley said. At the American Cancer Society, he and his colleagues examine that question in terms of what would happen if the state-of-the-art in cancer care was applied to the entire U.S. population. For instance, college-educated Americans have a much lower death rate from cancer than non-college-educated Americans, Brawley said (Singh and Jemal, 2017). One in four cancer deaths (or approximately 152,000 individuals each year) could be avoided if all Americans, college-educated or not, received the same high-quality treatment, Brawley said.
On the topic of disparities in cancer deaths, Brawley noted that there has been a 49 percent decline in age-adjusted breast cancer death rates since 1981 but that this reduction has not been uniform across the United States (American Cancer Society, 2017). In fact, the decline has been less than 30 percent in Alabama, Alaska, Arkansas, Georgia, Idaho, Mississippi, Oklahoma, Tennessee, Texas, Utah, and West Virginia, Brawley said. Similarly, while there has been a 39 percent decline in age-adjusted colorectal cancer death rates since 1989 for the United States overall, the rates in Alabama, Arkansas, Georgia, Louisiana, Mississippi, Oklahoma, and West Virginia have declined by less than 30 percent. The differences, he said, are not defined by race; instead, those who have seen the largest declines are those who have healthy habits and health insurance; who have access to care, prevention, state-of-the-art diagnostics, and treatments; and who know how to take advantage of the available options for treatment.
Over the past 40 years, Brawley said, the study of health disparities has become an academic discipline that has gone from looking specifically at minority health to looking more broadly at health equity. Populations, he explained, can be defined or characterized by several features, including gender, race, ethnicity, culture, area of geographic origin, socioeconomic status, and current location (e.g., urban or rural). The study of minority health issues grew out of the civil rights movement of the 1950s and 1960s and from the realization that there were racial differences in cancer death rates and survival (Fontaine et al., 1972; Howard et al., 1992) as well as in
the occurrence of other diseases, such as sickle cell anemia, hypertension, and diabetes.
Race, a concept first promulgated by Carl Linnaeus in 1735, is defined today by the U.S. Office of Management and Budget (OMB) 2 years before each decennial census. It changes over time—President Obama would have been classified as white in 1970 and African American in 1980, for example—and is now a sociopolitical characterization and not a biological one, said Brawley. Race is rejected by the anthropological community as non-scientific and is not a good surrogate for genetics or genomics, Brawley continued. Clinical trials should strive for genetic and genomic diversity and move past using race, he said.
Genetics do, however, correlate with family lineage. The sickle cell trait, for example, is a disease of people of Mediterranean and sub-Saharan African origin, even though many think of it as a disease that affects blacks. Similarly, cystic fibrosis mutations occur in people of Northern European origin, while alcohol dehydrogenase deficiency tends to occur in people of East Asian origin. Approximately 12 percent of those living along the border between Thailand and Malaysia have a genetic predisposition to a skin reaction called Stevens-Johnson Syndrome when given the anti-seizure drug carbamazepine, Brawley said. This reaction can be fatal, so identifying the individuals with this mutation can save lives; at the same time, however, the situation raises the issue of whether population profiling and genetic testing is ethical and acceptable. Caution should be exercised with regard to population profiling, Brawley said, but in this particular instance, it may be appropriate in order to save lives. The word profiling has a very negative connotation in some populations, Brawley said, and perhaps there is a need to improve the terminology.
Genomic information may help to define populations at risk of disease, Brawley said, and it may also help predict the responses of a disease to specific treatments; as such, it is an important tool in precision medicine. For example, individuals with type 2 diabetes who respond to the drug metformin have a different genomic profile from those who do not respond to the drug (Luizon et al., 2016). “Genomics is a scientific objective way to define or categorize populations,” Brawley said, adding that the goal for medicine is for people to become personalized from a fingerprint of their biologic makeup and not from racial typecasting.
The practice of “race medicine,” which is based on the assumption that diseases affect people of different races in different ways, was used in the 19th and early 20th century to justify disastrous experiments such as the Tuskegee syphilis study (Reynolds, 1997), Brawley said. Because of this misguided view, the word “genetic” is a loaded term and carries baggage for certain groups of people in the United States, he said. The concept of biological differences led to the idea that disparities in outcomes exist because
treatments tested in whites do not work in African Americans. While some believed that disparities exist because African Americans were not receiving the best available treatments, the majority opinion was that treatments needed to be tested in African Americans, Brawley said. As a result, the National Institutes of Health Revitalization Act of 19935 required that minorities be included in federally sponsored clinical trials and that there be a valid subset analysis of the differences among races in Phase III clinical trials funded after 1995.
The legislative requirement for valid subset analysis is an issue, Brawley said, because that type of analysis has well-established limitations (e.g., false positives/negatives). The phrase “valid subset analysis” is also ethically problematic, Brawley said, because it implies that race is a biological characterization and that disparities result from these genetic differences. The law also stipulates that researchers “encourage” minority participation in clinical trials, disregarding patient autonomy and the right for patients to make their own medical decisions. In addition, experts in clinical trial design often avoid subset analysis because it can produce misleading results and requires oversampling the subset. “This is an ethical issue because a clinical trial involves putting people at risk,” he explained, “so you are going to put more of the minority population at risk to get a valid subset analysis.”
One example of how subset analysis can produce misleading results is the analysis of the data from the original randomized controlled clinical trial of tamoxifen. Though tamoxifen is one of the most effective drugs for treating breast cancer, Brawley said that a subset analysis of those original data would have concluded that tamoxifen does not work nearly as well in African American women as it does in white women. In fact, he said, it is likely that the U.S. Food and Drug Administration (FDA) would not have approved the use of tamoxifen in African American women based on this analysis. Today, however, research has shown that tamoxifen only works in estrogen receptor-positive breast cancer, revealing that tamoxifen is an effective drug for women with the estrogen-receptor-positive disease regardless of race. What we can learn from this study, Brawley said, is that it is more beneficial to examine a patient’s genetic signature than their race.
On a final note, Brawley reiterated that disparities in access to health care are not just about race, but rather have to do with several socioeconomic factors, including where people live. “How can we provide adequate, high-quality care to include preventive services to populations that so often do not receive them?” Answering this question and creating an
5 More information on the National Institutes of Health Revitalization Act of 1993 can be found at https://orwh.od.nih.gov/sites/orwh/files/docs/NIH-Revitalization-Act-1993.pdf (accessed September 28, 2018).
equitable health care system in the United States, he said, will first require getting Americans to realize that there are large groups of people being left behind in terms of their medical care. “Until we change that, we are not going to change the health equity problem,” Brawley said.
This page intentionally left blank.