1
Introduction
The field of genomics has expanded greatly since the first sequence of the human genome was published a decade ago. According to workshop chair Debra Leonard of Weill Cornell Medical College, the hoped-for outcomes of the human genome project were to understand human genetic variations and their relationship to health and disease; to predict disease risks for prevention and earlier treatment; to refine disease diagnosis by understanding the underlying genetic variance and molecular mechanisms; and to use that information to create better treatments and improve the health and health outcomes of the U.S. population. Over the past 10 years, scientists have developed genomic tests based on identified gene-disease associations that can predict the response of an individual patient to a drug intervention or the risk of developing Alzheimer’s disease. However, much of the evidence surrounding the clinical value and utility of these tests has not been sufficient enough for clinical practitioners to broadly embrace many of these in practice.
A major impediment to the integration of these genomic tests into routine health care is the lack of an adequate evidence base linking the use of genomic diagnostic tests to health outcomes. Since these new technologies are beginning to play an increased role in clinical decision-making and the management of disease, the Institute of Medicine’s Roundtable on Translating Genomic-Based Research for Health hosted a public workshop on
November 17, 2010, to explore issues related to this lack of evidence.1 Various stakeholders, including regulators and policymakers, payers, healthcare providers, researchers, funders, and evidence-based review groups, were invited to share their perspectives on the strengths and limitations of the evidence being generated to assess the clinical validity and utility of genomic diagnostic tests. Specifically, panelists were asked to address the following:
-
What evidence is required by stakeholders (e.g., for decisions regarding clearance, use, and reimbursement)?
-
How is evidence currently being generated?
-
Are there innovative and efficient ways to generate high-quality evidence?
-
How can the barriers to generating this evidence be overcome?
Early genetic tests, Leonard explained, were focused on single genes. The market was limited and reimbursement was poor, and the in vitro diagnostics industry was therefore not very interested in developing genetic tests. Instead, genetic tests for the diagnosis of disease were generally developed as needed by clinical laboratories. These were based on published genotype–phenotype correlations and were developed using standard molecular biology methods and sets of patient and control samples. The Clinical Laboratory Improvement Amendments (CLIA) (42 U.S.C. 263a) allows such practices without the need for receiving device clearance from the U.S. Food and Drug Administration (FDA). However, Leonard said, there were concerns about the quality of these tests, the potential for harm to patients, the clinical validity and utility of the tests, and the relatively expensive cost.
Genetic tests are still in use today, Leonard said, but the focus has shifted to genomic tests, which are complex testing algorithms of multiple genetic variants, multiple genes, or gene expression patterns. Genomic tests are used for diagnosis as well as for therapeutic selection, dosing, prognosis, and residual disease detection. However, the majority of these tests have insufficient clinical validity and utility data, and there is currently little evidence of improved health outcomes from their use (Table 1-1). The increasing role of genomic tests in clinical decision-making has led to
TABLE 1-1 Evidence-Based Review of Select Genomic Tests
Genetic Test |
Reviewed by |
Conclusion |
Thrombophilia tests |
AHRQ/EGAPP |
No direct evidence for improved outcomes |
HER2 testing in breast cancer |
AHRQ |
Weak evidence relating test result to treatment outcomes |
Gene expression profiles for breast cancer |
AHRQ |
High quality retrospective clinical utility data for Oncotype DX |
UGT1A1 genotyping for colorectal cancer patients |
EGAPP |
Insufficient evidence for or against testing |
Genetic testing for hereditary nonpolyposis colorectal cancer (HNPCC) |
EGAPP |
Limited evidence that mismatch repair gene mutations cause family members to have increased screening |
CYP450 for non-psychotic depression |
EGAPP |
Paucity of good quality evidence that testing is useful for selective serotonin reuptake inhibitor (SSRI) outcomes |
Genomic tests for ovarian cancer |
EGAPP |
No evidence that tests affect outcomes in asymptomatic women |
Abbreviations: AHRQ (Agency for Healthcare Research and Quality), EGAPP (Evaluation of Genomic Applications in Practice and Prevention). SOURCE: Adapted from Leonard, IOM workshop presentation on November 17, 2010. |
ongoing discussions at FDA regarding the appropriate level of regulatory oversight for genetic and genomic tests.2
The barriers to evidence generation have been discussed in many venues (summarized in Box 1-1). The goal of this workshop, Leonard said, is to look beyond these barriers to define the evidence needed and the mecha-
2 |
As further discussed in Chapter 2, other stakeholder groups including payers, evidence-based review groups, providers, professional societies, and patient groups have also initiated discussions on the utility of genetic and genomic tests in clinical decision making. |
BOX 1-1 Speaker’s Perspectives on Barriers to the Collection of Clinical Validity and Utility Data for Genomic Diagnostic Tests
|
nisms to obtain it so that the promise of the human genome project and genomic diagnostic testing can be fully realized.
The report that follows summarizes the presentations and discussions by the expert panelists. Chapter 2 provides the different stakeholder perspectives on the type and level of evidence needed for decision making. Approaches for evidence generation are discussed in Chapter 3. Chapters 4 and 5 examine ways to overcome the barriers to evidence generation and strategies for moving forward. Final remarks are provided in Chapter 6, and the workshop agenda, biographical sketches of the panelists, and list of attendees are included in the appendixes.