The sequencing of the human genome and the identification of associations between specific genetic variants and diseases have led to an explosion of genomic-based diagnostic tests. These tests have the potential to direct therapeutic interventions, predict risk or onset of disease, or detect residual disease. As research progresses and an increasing number of associations are found, further tests will be developed that can aid in providing personalized treatment options for patients.
However, the adoption of genomic diagnostic tests by health care providers has been limited due to a lack of evidence regarding the clinical utility of many tests.2 Health funders and practitioners lack the data necessary to distinguish which tests can improve practice or the clinical settings in which tests will provide the greatest value. The Roundtable on Translating Genomic-Based Research for Health held a workshop in November 2010 (IOM, 2011b) to determine what evidence is needed and how it is viewed by different stakeholders in order to develop genomic diagnostic tests of clinical value.
Many workshop participants noted that evidence is lacking for the
1 The planning committee’s role was limited to planning the workshop, and the workshop summary has been prepared by the workshop rapporteurs as a factual summary of what occurred at the workshop. Statements, recommendations, and opinions expressed are those of individual presenters and participants, and are not necessarily endorsed or verified by the Institute of Medicine, and they should not be construed as reflecting any group consensus.
2 Formally, a diagnostic test confirms a specific condition, while a prognostic test predicts the possibility of developing a specific condition. This report uses “genomic diagnostic test” to refer to any genetic or genomic test used in making health care decisions.
impact of most genetic and genomic tests on health outcomes and that better mechanisms are needed to generate this evidence. “If you look at many of the evidence-based reviews in the literature, insufficient evidence is one of the predominating assessments of most of the tests currently on the market,” said Debra Leonard of Weill Cornell Medical Center in recounting that meeting.
Different stakeholders presented new models, strategies, and specific actions for generating evidence for genomic diagnostic test development. For example, workshop participants looked at whether combining evidence from different studies would generate sufficient evidence to meet stakeholder requirements for analytic validity, clinical validity, and clinical utility (these terms are defined in Chapter 2 of this report) as compared to the evidence generated from a single good clinical trial. The evidence should only need to be “adequate,” said Leonard, “not perfect.” The evidence should be strong enough to “get us to 85 percent, ‘B grade,’ certainty for these tests.”
New economic models for reimbursement were discussed in which a test’s value would be determined by its ability to direct clinical care, such as preventing the use of ineffective therapies or directing patients to therapies that improve outcomes, as opposed to the cost of performing the test. Participants discussed implementing a system that does not pay for a treatment if the treatment is not supported by prognostic or predictive tests. “This is a huge issue in the United States,” observed Leonard. Health care costs now constitute 17 percent of the U.S. gross domestic product, yet the U.S. health care system is ranked the lowest in outcomes and the most costly among developed countries (Peterson and Burton, 2007; SSAB, 2009). An ongoing global economic crisis demands fiscal responsibility, said Leonard, which is driving efforts to improve the cost-effectiveness of the health care system. “There’s a very negative approach to rationalization of health care in the United States. We need to think about how to address that.”
The previous workshop also sought to address the variance in stakeholder evidentiary requirements, specifically probing whether demonstration of safety and efficacy is enough to justify use of a new genomic test in medical practice or whether tests need to demonstrate clinical utility or cost-effectiveness instead. The Roundtable on Translating Genomic-Based Research for Health held a follow-up workshop on November 15, 2011 titled Facilitating Development and Utilization of Genome-Based Diagnostic Technologies to further explore the differences in evidence required for clinical use, regulatory oversight, guideline inclusion, coverage, and reimbursement of genomic diagnostic tests among stakeholders with the goal of clarifying a pathway for successfully bringing tests to clinical use for
the benefit of patients.3 Presenters at the workshop were asked to consider four broad issues:
1. How are the barriers to successful genomic test development viewed?
2. What are potential solutions?
3. What are the obstacles to achieving those solutions?
4. How can those obstacles be overcome?
This report summarizes the presentations and discussions that took place throughout the workshop. Chapter 2 relates two presentations which sparked extensive discussion. One presentation proposed that all genomic diagnostic tests be reviewed and approved by the U.S. Food and Drug Administration (FDA; see Box 1-1). The other observed that venture capitalists are no longer investing substantially in the development of genomic diagnostic tests because of a lack of clarity surrounding regulatory and reimbursement pathways. Though the two talks may seem only distantly related, both suggested the need for major changes in the systems used to develop, regulate, and reimburse genomic diagnostic tests.
The next four chapters present the perspectives of different stakeholders in the development of genomic diagnostic tests. Chapter 3 addresses the concerns of test developers; Chapter 4 those of patients; Chapter 5 those of payers; and Chapter 6 those of government officials. Each stakeholder group has a different set of needs and issues of importance, yet commonalities among them are apparent, such as the need to put patients and health outcomes at the center of discussion and action.
Chapter 7 summarizes the rich and extensive discussions that occurred throughout the workshop. These discussions have been organized thematically, with the identification of speakers who made specific proposals and recommendations. Collectively, the participants at the workshop charted a variety of ways to move forward in developing genomic diagnostic tests that could substantially improve human health.
3 The full statement of task can be found in Appendix C.
Decisions by FDA to clear or approve medical devices, including genomic diagnostic tests, for marketing are based on the safety and effectiveness of the product. The Medical Device Amendments of 1976 implemented a three-tier system. Class I devices are common, low-risk devices that are generally exempt from premarket evaluation by the agency. Class II devices are moderate-risk devices that are subject to premarket notification (also known as the 510(k) process), in which the sponsor must demonstrate substantial equivalence of the device to an already marketed product. Class III devices are the most complex and present the highest risk; makers of Class III devices must submit a premarket approval application demonstrating safety and effectiveness and obtain FDA approval prior to marketing.
Initially, genetic tests focused on single genes. The in vitro diagnostics industry was not very interested in developing such tests because they typically constituted a small market with poor reimbursement, according to Leonard. As a result, genetic tests were developed largely by clinical laboratories using standard molecular biology methods. These laboratory-developed tests (LDTs) tended to be based on published genotype-phenotype correlations, were developed using a set of patient and control samples, and usually were produced in small volumes. They were performed by specialists with advanced training and usually required expert interpretation. LDTs generally were and still are developed under the provisions of the Clinical Laboratory Improvement Amendments (CLIA) without clearance or approval from FDA.
Today’s genomic tests are quite different. They often are based on complex testing algorithms that encompass multiple genetic variants, genes, or gene expression patterns and, most recently, whole-exome or whole-genome sequencing, said Leonard. The results are used not only for diagnosis but for the selection of therapies, dosing decisions, prognosis, and detection of residual disease. Tests are increasingly empirical and nontransparent and rely on complex statistical methods. They often require complex software, many incorporate automated interpretation, and their clinical validity is not well understood (Wright and Kroese, 2009). Novel tests are often developed by companies and “licensed” to a laboratory, the volume and types of LDTs have grown significantly, and they are often a mechanism for the market entry of novel tests. A higher proportion come from commercial laboratories and biotechnology companies, and they often do not involve a close relationship between clinicians, pathologists, and patients. Some tests are broadly advertised and aggressively marketed to clinicians. Others are marketed directly to consumers and are available over the Internet with overnight shipping (Meyers, 2011). They can have a national or even international reach.