As novel genomic approaches move into routine practice, health care systems are routinely encountering new challenges. Examples of the impediments faced by health care systems in these situations are limited clinician knowledge about genomics, inconsistent reimbursement policies, and the need to analyze a complex evidence base to justify the new procedure. This chapter describes the challenges and successes encountered by three programs that are currently integrating genomic approaches to human disease prevention, diagnosis, and treatment (see Box 5-1). The programs described in this chapter, which are all at different points along the translational pipeline, include a statewide cancer genomics program, a coordinated effort to provide genomic diagnoses to people with monogenic diabetes, and a small pilot program that explored the genetic basis for rare childhood diseases.
In 2002 the Health Resources and Services Administration provided funding to a small group of states, including Michigan, to perform stakeholder needs assessments and develop statewide genetics plans. As part of its needs assessment, the Michigan Department of Health and Human Services (MDHHS) engaged hundreds of stakeholders, including patients, clinicians, researchers, teachers, and other representatives to determine what was desired from a state health department in terms of genetics. One message that came through quite clearly, according to Debra Duquette, genomics coordinator at the MDHHS, was that stakeholders wanted a more comprehensive genetics program that went beyond newborn screening and examined prevalent chronic diseases such as cancer, diabetes, and cardiovascular disease. At the same time, Michigan’s cancer division received many questions about hereditary breast and ovarian cancer, which led to the realization that expertise was needed on staff to address issues regarding genomic medicine on a population health level. As a result, a full-time genomics coordinator and a part-time cancer genetics coordinator were brought on, Duquette said.
The Michigan Cancer Genomics and State Genetics Plan that emerged from this process had six discrete goals designed to improve traditional maternal and child public health genetic services as well as to create a more comprehensive agenda covering common chronic diseases with onset in adult life (see Box 5-2; Michigan Department of Community Health, 2004).
In 2003 Michigan entered into a 5-year cooperative agreement with the Office of Public Health Genomics in the Centers for Disease Control and Prevention (CDC) in an effort to integrate genomics into the chronic disease realm. In 2008 Michigan received additional funding from the CDC to implement and disseminate information about genetic tests with a strong evidence base such as those used for BRCA1 and BRCA2 screening and tests for Lynch syndrome.
In 2014 the state entered into its fifth cooperative agreement with the CDC’s Division of Cancer Prevention and Control. As part of this most recent project, Michigan is implementing education and surveillance systems for Lynch syndrome and for hereditary breast and ovarian cancer. According to Duquette, this effort aligns with Healthy People, an initiative of the Office of Disease Prevention and Health Promotion. Healthy People proposes 10-year national objectives designed to improve the health of Americans, and two of the objectives in the most recent Healthy People 20201 plan involve the use of genomics-based tools for improving the health of the overall population. The two genomics-specific goals of Healthy People 2020 are:
- To increase the proportion of women with a family history of breast and/or ovarian cancer who receive genetic counseling.
- To increase the proportion of persons with newly diagnosed colorectal cancer who receive genetic testing to identify Lynch syndrome (or familial colorectal cancer syndromes).
The Comprehensive Cancer Control Plan for Michigan for 2009–2015 included a plan to increase the availability of cancer-related genetic information to the Michigan public and decrease barriers to risk-appropriate services. To achieve these aims, the Cancer Control Plan included three implementation objectives (Michigan Cancer Consortium, 2014):
- By 2011, expand public knowledge about the impact of genetics on cancer risk and management (breast, ovarian, and colorectal cancers).
- By 2015, expand provider knowledge about the impact of genetics (breast, ovarian, and colorectal cancers) on cancer control.
- By 2015, improve genetic health care financing and access to testing and support services.
Implementing the Program
To support the implementation of these objectives, a multidisciplinary group of partners was enlisted, including federal organizations, state and local agencies, clinical practices, providers, patients, and families. One partner organization, the Michigan Cancer Consortium, is a network that includes more than 100 public and private groups working toward cancer prevention and control. The state’s cancer control plan also helped facilitate the creation of the Michigan Cancer Genetics Alliance network, which currently has about 250 members and has been an important part of the effort, Duquette said.
There are 25 health plans in place in Michigan that cover the vast majority of its 10 million residents, Duquette said. State officials have been working to educate health plan leaders on the best practices for hereditary breast and ovarian cancer. A major partner in this effort is the Michigan Association of Health Plans, a group that disseminates information to its constituents about cancer genomics policies, among other issues, in the form of a quarterly newsletter.2
The MDHHS developed specific metrics to assess genetic counseling, testing, and management policies pertaining to BRCA1 and BRCA2. When a health plan performs well on every metric, it receives an honor from the department, Duquette said. The number of plans qualifying for
awards has steadily increased, from 4 when the work began to 16, covering more than 80 percent of the population, she said. Data are collected from every board-certified or board-eligible genetic counselor in the state who is involved in BRCA counseling. Currently, the database includes information on more than 17,000 patients and is useful for addressing open questions such as how insurance coverage affects testing.
Much work has been done to reduce barriers for people who were not able to receive BRCA testing because of inadequate insurance coverage, Duquette added. In 2008, approximately 22 percent of the population receiving genetic counseling were not able to get BRCA testing because of inadequate insurance coverage, she said, but by 2014 that percentage had decreased to just over 8 percent.
There are vast differences in coverage policies for BRCA testing and counseling among private and public payers (Wang et al., 2011). Many of the health plans in Michigan previously covered counseling and testing solely for BRCA1 and BRCA2; however, several payers have moved away from gene-specific coverage to indication-specific policies, Duquette said. Indication-specific testing refers to genetic screening performed in response to specific clinical indications that fall into categories such as pediatric, preconceptional/prenatal, and adult-onset conditions (Pletcher et al., 2007). However, implementers, such as state health departments, could benefit from returning to the original system where reimbursement policies on genetic testing and counseling were categorized in a gene-specific manner, Duquette said.
A bottom-up approach to genomic medicine has taken place in the field of diabetes research, said Toni Pollin, associate professor in the Department of Medicine and the Department of Epidemiology and Public Health at the University of Maryland School of Medicine. In 2006 a genetic test was conducted on a sample from a young girl who had been diagnosed with type 1 diabetes at 1 month of age and was reliant on an insulin pump. The test results surprisingly indicated that she did not have type 1 diabetes, but instead possessed a rare mutation in a potassium channel subunit that was the cause of her illness. After she started a new treatment regime at the Clinical Research Center of the University of
Chicago, her islet cells started to produce their own insulin within a few weeks, and she became completely free of her insulin pump, Pollin said.
Type 1 and type 2 diabetes, the most common forms of the illness, are polygenic, meaning that multiple genes contribute to the risk of developing the disease (National Diabetes Information Clearinghouse, 2007). The young girl described by Pollin did not have type 1 or type 2 diabetes, but instead had an illness that is classified as monogenic diabetes. Monogenic diabetes encompasses rare forms of the illness that are caused by a mutation in a single gene, and they can be overcome in certain cases with high doses of sulfonylureas, drugs used to stimulate the release of insulin from the pancreas. The two forms of monogenic diabetes—neonatal diabetes mellitus and maturity-onset diabetes of the young (MODY)—account for 1 to 5 percent of all diabetes cases in young people (National Diabetes Information Clearinghouse, 2007).
MODY is often misdiagnosed as either type 1 or type 2 diabetes, Pollin said, but such misdiagnosis can be avoided with genetic testing. The majority of MODY cases are caused by defects in genes that code for transcription factors active in pancreatic beta cells or by defects in glucokinase, which is required for the phosphorylation of glucose to glucose-6-phosphate, she said. Once the disease is properly diagnosed, many MODY patients can transition from insulin to sulfonylureas without loss of glucose control (Shepherd et al., 2009). The proper diagnosis and transition away from insulin is important for patients and their families, Pollin said, because it eases the therapeutic burden and improves quality of life.
The SEARCH for Diabetes in Youth study, a multicenter examination of the etiology of diabetes in young people, found that monogenic diabetes is underdiagnosed (Pihoker et al., 2013). Although only a small percentage of patients have monogenic diabetes, this group still represents hundreds of thousands of people in the United States, Pollin said. Furthermore, MODY mutations occur across all minority groups, so populations at high risk for childhood obesity, including Hispanics and African Americans, may be particularly underdiagnosed, Pollin said.
MODY presents a compelling opportunity to implement genomic medicine, Pollin said, but the field faces a lack of awareness—many people have never heard of these forms of diabetes, and clinical overlap can lead to trouble differentiating between the types. Some algorithms for diagnosis are available, but no perfect approach exists, Pollin remarked. Additional challenges include the high cost and complexity of the current tests, intellectual property concerns, and limited professional
society guidance, she said. Finally, patients and physicians alike are largely unaware of how much a proper diagnosis can change a patient’s life. For all these reasons, Pollin said, it is important to assemble a strong evidence base that demonstrates why diagnosing MODY is beneficial.
Improving the Evidence Base
The Personalized Diabetes Medicine Program at the University of Maryland, a component of the IGNITE project, is aimed at strengthening the evidence base on genomic approaches for detecting MODY. Researchers are trying to identify patients who may have monogenic diabetes. Pollin described their approach, which begins with a patient questionnaire that clarifies several aspects of family history and initial diagnosis. The results of the questionnaire are used to determine if further clinical workup or sequencing is needed. If a patient is referred for sequencing, he or she is tested for a panel of 40 known monogenic diabetes genes. In those patients where sequencing reveals a pathogenic variant, test results are added to the electronic health record (EHR), and customized treatment begins, with genetic counseling and testing for family members. If a variant of unknown significance is found, further family and functional studies are performed. Finally, the impacts are evaluated through patient- and provider-reported outcomes.
The program is interested in disseminating its findings, especially to genetic counselors and clinical geneticists, in order to facilitate accurate diagnoses as early as possible. In order to efficiently translate their research findings to clinical care, the Personalized Diabetes Medicine Program staff members are liaising with hospital administrators at the University of Maryland, Pollin said.
Covering the Test Panel
A payer advisory panel working with the Personalized Diabetes Medicine Program indicated that payers are primarily interested in covering those genetic tests with demonstrated clinical utility, Pollin said. An interesting conundrum has taken place with payers over the number of genes on the monogenic diabetes panel, Pollin said. Payers often will only want to cover the subset of genes on a panel for which the clinical utility has already been clearly demonstrated. For example, if 5 of 40 genes have strong evidence to support clinical use, payers do not want to cover the costs of testing the other 35 genes, she said. In response to the
objection that the full panel costs no more than a partial panel, the payers have indicated that they do not want to see the results of the other 35. It is important to consider that the additional genes on the panel may strengthen the evidence base and lead to improved testing and interventions, Pollin said.
Génome Québec recently launched the Integrated Clinical Genomic Centre in Pediatrics in collaboration with Centre Hospitalier Universitaire Sainte-Justine, said Marc LePage, president and chief operating officer of Génome Québec. The pediatric genomics center is the first of its kind in Canada and is attempting to overcome diagnostic challenges in children with rare genetic diseases. Pediatric genetic disorders affect approximately 500,000 children in Canada, and many of the genes that cause these disorders are unknown because gene-discovery studies are especially challenging with limited patient samples. Of those 500,000 affected children, 50 percent do not receive a diagnosis and 40 percent receive an incorrect diagnosis, LePage said.
In an effort to improve the diagnosis of rare genetic diseases, the pediatric genomics center developed a pilot project in which the center provided the sequencing capacity to examine a small cohort of children with undiagnosed illnesses, LePage said. Researchers carried out exome sequencing in 96 children, which resulted in a molecular diagnosis for 37 percent and a tentative diagnosis that required further confirmation for an additional 15 percent. This represents a major step forward because a diagnosis often means better clinical care for these patients, LePage said.
Now that the pilot project is finished, LePage and his colleagues are envisioning much larger studies that would tackle genomics issues in cardiology, oncology, neurology, and other fields, with Génome Québec providing the centralized sequencing and clinical accreditation. The pilot research project on rare diseases gave Génome Québec the initial momentum and small-scale experience needed to take on a bigger challenge, LePage said.
The biggest issue facing Génome Québec during the pilot projects, he said, is the funding gap that exists between research and the health care system. Research agencies are hesitant to fund projects that are in the clinic, and vice versa, LePage said. It is important to make further
inroads into the clinical arena, he said, to bring the results of research to patients and providers. “That is our challenge in this field right now.”
A National Bioresource Network
In addition to its support of pilot projects in genomics, Génome Québec is working to expand national bioresources. In 2007 a databank known as CARTaGENE was created to collect and store biological material specific to Québec (see Chapter 3). The CARTaGENE bioresource is now one of five regional projects that are part of the Canadian Partnership for Tomorrow Project (CPTP), a pan-Canadian network that stores clinical information and biological samples from approximately 300,000 people (Borugian et al., 2010). CARTaGENE was initially designed as an ongoing prospective investigation of the environmental, lifestyle, and genetic influences on cancer, but its reach is spreading to other chronic diseases now, LePage said. The CARTaGENE cohort consists of male and female participants ranging in age from 40 to 69, which is the demographic at the highest risk for developing chronic disorders (Awadalla et al., 2013). Participants will be followed on a long-term basis in order to better understand the influence of genetics and environment on health and disease.
The principles of implementation science can be used to help build a common framework for incorporating genomics into clinical practice by drawing from other disciplines such as the management sciences, organizational design, and engineering, said Brian Mittman of Kaiser Permanente Research. In some cases, patient care will be customized and unique; in other cases, it will be routine. If a generic framework similar to a standard operating procedure existed, the appropriate customizations and individualization could be applied as needed, Mittman said.
Given the inertia of clinical care, leaders of health care systems must make challenging decisions between wholesale change and implementing single applications. Alexandra Shields of Harvard Medical School and Massachusetts General Hospital supported the idea of thinking about one application at a time. Shields cautioned against placing genome-wide sequencing results directly into patients’ EHRs with the idea that they
will become useful someday because that approach does not have the infrastructure or resources to support it. However, she also noted that infrastructure and resources could be standardized moving forward, so that the investigation of genomic applications becomes routinized. This process could encompass patient education, ethical issues associated with consents and the return of results, and follow-up genetic counseling. Such supports could “realize the benefits of genomic medicine that apply to all different cases,” Shields said.
Developing completely unique methods for implementing every genomic application is a challenge, said Robert McDonough of Aetna. He emphasized the value of a common framework for evaluating genomic tests and their implementation. From a payer’s perspective, he observed, an overarching framework to deal with the plethora of tests would make it easier to consider the unique aspects of an individual test. A process similar to this is occurring in oncology, said Jane Perlmutter of the Gemini Group, where the mentality is shifting away from thinking of cancer by organ site and more toward classifying cancer by the genetic mutations.
This page intentionally left blank.