|
|
||||||||
Clinical Chemistry Forum |
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
|
Reflecting concern about the rapid proliferation of genetic tests before their safety and effectiveness had been demonstrated, the NIH-Department of Energy (NIH-DOE)1 Working Group on Ethical, Legal, and Social Implications of Human Genome Research (ELSI) convened the Task Force on Genetic Testing in 1995. The ELSI Working Group asked the Task Force to review genetic testing in the United States and, when necessary, to make recommendations to ensure the development of safe and effective genetic tests. Part of the review was accomplished by the survey already mentioned. The Task Force defined safety and effectiveness "to encompass not only the validity and utility of genetic tests, but their delivery in laboratories of assured quality, and their appropriate use by healthcare providers and consumers" [Ref. (7), p. 4]. I will discuss these components, and the recommendations of the Task Force for ensuring safety and effectiveness of newly developed genetic tests.
The need for new policies concerning genetic testing has been
recognized by the Secretary of Health and Human Services. Secretary
Shalala asked to be briefed about the progress of the Task Force, and
when its report was submitted, she created an interagency working group
in the Department of Health and Human Services to review the
recommendations of the Task Force and to consider their implementation.
One reason the Report gained such attention was that it represented the
consensus of multiple stakeholders in genetic testing (Table 2
). There were no dissenting votes for any of the
recommendations. Another reason was the effort of the Task Force to
solicit public and professional input. The Task Force had public
hearings, and published its preliminary recommendations in the
Federal Register (8). Eighty-two people responded
at least once. On the basis of those comments, the Task Force
drastically modified a number of its recommendations before issuing its
final report [(7), Appendix 1].
|
During the summer of 1998, Secretary Shalala implemented one
recommendation of the Task Force (Table 3
) by creating the Secretary's Advisory Committee on Genetic
Testing. As of January 1999, the members of this committee, which is
being administered from the Office of the Director, NIH, had not been
appointed. Over 200 nominations were received.
|
| Validity |
|---|
|
|
|---|
Two intrinsic features of genetic diseases affect clinical validity, heterogeneity and penetrance. Heterogeneity affects sensitivity, the probability that a person who will get the disease will have a positive test result. When many disease-related alleles exist at either one gene locus (allelic diversity), which is often the case, or at several loci (locus heterogeneity), current technology sometimes fails to detect all of them. Heterogeneity occurs frequently for both Mendelian and common, complex disorders. Sensitivity is further lowered for the latter because the vast majority of cases are not attributable to alleles at any gene locus, but to a combination of genetic and environmental factors, which will not be detected by any genetic test.
When penetrance is incomplete, the chance that a person with a positive
test result will develop the disease [positive predictive value
(PPV)] is lower than when penetrance is complete. Complete penetrance
is usual for Mendelian disorders for which, consequently, we can speak
of disease-causing rather than susceptibility-conferring genotypes.
Penetrance is incomplete when other genetic or environmental factors
must be present before disease manifests. Despite the strong
association of certain BRCA1 or BRCA2 alleles
with breast cancer in certain families (PPV
85%), fewer than 60% of
Ashkenazi Jewish women found by a population-based search to have any
one of three such alleles developed breast cancer by age 70
(9). Alleles at other gene loci and similar environments are
more likely to be shared by relatives than by people in the general
population.
The Task Force was concerned that genetic tests intended to predict
risk of future disease in apparently healthy people were becoming
available before adequate data on sensitivity and PPV had been
collected. It first established the three criteria shown in Table 4
, requiring an investigational stage and defining in broad terms
the types of studies that were needed.
|
Are test developers going through an investigational stage? Under Food and Drug Administration (FDA) regulations, organizations developing new medical devices must have their investigational protocols approved by an institutional review board (IRB) if human subjects are involved. Genetic testswhether marketed as services or kitsare considered by the FDA to be medical devices [Ref. (7), pp. 2930]. The FDA has not enforced this regulation for developers planning to market tests as services, which at the present time is the way most new tests are marketed. A laboratory that has developed a new test and markets it as a service needs only approval under CLIA 1988. That does not involve submitting data on clinical validity or having an approved IRB protocol. On the other hand, organizations marketing tests as kits must submit data to the FDA. They are, therefore, much more likely to comply with FDA requirements, including submission of protocols to an IRB.
In the survey of biotechnology companies (BTCs) and nonprofit organizations (NPOs) completed in 1995, only 23 of 37 BTCs (62%) and 97 of 127 NPOs (76%) that were developing genetic tests had ever submitted protocols to an IRB. Only 8 of 14 BTCs (57%) and 72 of 95 NPOs (76%) using home brews for genetic tests had ever submitted protocols to an IRB (6).
To give an example of the problem: When a polymorphic mutation in Ashkenazi Jews (10) was reported to increase the risk of colon cancer, the availability of a clinical laboratory test for the mutation was announced simultaneously. The work received considerable publicity, and in the following weeks, >1000 people called the center from which the report emanated to find out about getting the test. At that point, the work had not been independently replicated. Moreover, in the original research report, the risk of colon cancer in those with the mutation (PPV) was only ~30%, twofold higher than in Ashkenazi Jews without the mutation. In subsequent studies on other Ashkenazi Jewish populations, the risk has been reported to be either nonexistent (11) or lower (12). Thus, under existing regulations, the test could be made available as a clinical laboratory service before clinical validity or utility were established.
The Task Force went even further to assure that an investigative stage
would precede marketing (Table 5
). It made explicit that tests developed commercially (as well
as those developed in state and local health departments) should submit
protocols to IRBs although companies without Federal suppport were
under no legal obligation to do so [Ref. (7), pp. 30,
334]. That IRBs have the authority to assess protocols for
scientific merit as well as protection of human subjects is made clear
by the Office of Protection of Human Subjects from Research
Risks, which governs IRBs: "(I)f a research project is so
methodologically flawed that little or no reliable information will
result, it is unethical to put subjects at risk or even to
inconvenience them through participation in such a study
(emphasis added) [Ref. (13), p. 41]. The
Task Force maintained that a protocol for genetic test development
would be flawed if it did not collect data on clinical validity and
utility in an appropriate manner.
|
| Utility |
|---|
|
|
|---|
| External Review of Data Collected under Protocols |
|---|
|
|
|---|
With the exception of genetic tests marketed as kits, for which the FDA reviews data on clinical validity (but not utility as the Task Force defines it), no regulations exist for external review of genetic tests. External review can have a marked influence on providers' decisions to use, or not use, new medical technologies. Examples include statements of professional societies, consensus development panels, and ratings by the US Preventive Services Task Force (15). The decision of health insurers on whether a specific genetic test will be included in their benefits or reimbursement packages can also influence use. The Task Force called on the FDA to consider the need for new policies on genetic tests [Ref. (7), p. 38].
Two problems need to be addressed regarding external review: the
lengthy time needed to obtain data on utility as well as, occasionally,
validity; and the large number of new genetic tests that might be
developed, which could overwhelm any review system. To remedy the first
problem, the Task Force presented several alternatives that could be
used "when preliminary data indicate a test is likely to have
validity and utility" [Table 6
and Ref. (7), p. 36]. The first option, voluntary
postmarket collection of data by test developers, leaves unanswered the
question of whether anyone external to the developing organization will
decide when sufficient data have been collected and when to
definitively review the data to decide on safety and effectiveness. The
second option, reimbursement by health insurers while the test is still
in an investigative stage, will greatly facilitate data collection
because many providers will be reluctant to order a test under
development if patients must pay for it out of pocket. The insurers
themselves could make the final determination of a test's validity and
utility, although their decision might be unduly influenced by whether
the test will lower their costs (M. Schoonmaker, submitted for
publication).
|
Unless the FDA decides to regulate genetic tests marketed as services as it does kits, which it acknowledges it has the authority to do [Ref. (7), pp. 2930], the third option, conditional premarket approval by the FDA, applies only to genetic test kits. The Task Force recommended that "(w)hen FDA considers it likely that a test kit will prove to make an important contribution to the prevention or management of the disorder, it should grant conditional premarket approval when a developer requests it. (FDA frequently clears or approves products for a limited-indication use with the requirement for postmarket studies or the expectation that claims may be extended as sufficient evidence accumulates.)" [Ref. (7), p. 36]. In return for conditional approval, developers could include a profit markup in the price of their test. They could promote the test, but would have to indicate that its safety and effectiveness were still under investigation. Informed consent would be needed, but the Task Force also recommends it for many predictive genetic tests fully approved for marketing (see below). Developers of kits would continue to collect and periodically present data to the FDA until such time as the agency either gave unconditional approval for marketing or denied it.
The Task Force dealt with the second problem by recommending that tests
defined by the proposed advisory committee on genetic testing (Table 3
)
as requiring "stringent scrutiny" be given priority for review.
"Stringent scrutiny," the Task Force report said, "is indicated
when a test has the ability to predict future inherited disease in
healthy or apparently healthy people, is likely to be used for that
purpose, and when no confirmatory test is available" [Ref.
(7), p. 11]. It called on the proposed advisory committee
on genetic testing, perhaps in conjunction with the Office of
Protection of Human Subjects from Research Risks or the FDA, to define
additional indications.
Review panels could become enmeshed in endless debate if they attempt to set cutpoints for sensitivity and PPV; these should vary depending on the particular test, its use, options for treatment, and other factors. Even for a particular test, reasonable people will differ on how much test uncertainty they can tolerate. The Task Force maintained that it was more important for external reviewers to ensure that the data have been collected and analyzed appropriately than to attempt to set cutpoints. That way individual users (providers and patients) could decide whether a test was suited to their needs. Review panels could suggest intended uses, e.g., the target f groups for specific tests [Ref. (7), p. 17].
| Laboratory Quality |
|---|
|
|
|---|
The Task Force was also concerned with the quality of information provided to practitioners or patients by clinical laboratories. Drawing on data from the 1995 survey, Cho et al. (17)(18) collected pamphlets that BTCs and NPOs gave to practitioners, patients, or both and determined how many of 10 elements each pamphlet included. Of the 115 pamphlets reviewed, only three elements were mentioned in a majority of pamphlets: who the candidates for testing were; a description of the condition; and availability of, or need for, genetic counseling. Only 10% included a statement on risks, limitations, or benefits of testing, and fewer than one-third contained a statement regarding confidentiality or need for informed consent. Although almost one-half made some statement about test performance these were usually vague, often using the term "accuracy" without indicating sensitivity, specificity, or predictive value (17). Pamphlets from NPOs were significantly more likely than those from BTCs to include information on patient rights and the need for or availability of counseling. They were significantly less likely than those from BTCs to indicate the intended purpose of the test (18).
In late 1996, the Task Force collected informational material from four laboratoriesthree BTCs and one NPOoffering testing for inherited susceptibility mutations in the BRCA1 and BRCA2 genes. They differed markedly in content. One commercial brochure said, "population screening should be offered where feasible" (19), but the brochure from the University of Pennsylvania maintained that systematic monitoring was still "essential ... so that one day we will be able to offer appropriate screening guidelines based on firm clinical data" (20). Myriad's brochure said, "Early [breast] cancer detection provides the best opportunity for reducing mortality from cancer" (21), but OncorMed's brochure pointed out, "There is no surveillance or prevention strategy which is proven to decrease the mortality associated with carrying a [BRCA] mutation" (22). Myriad's brochure also stated that, "prophylactic mastectomy does not completely eliminate the risk of breast cancer. ... However, the procedure substantially reduces the risk of breast cancer". The University of Pennsylvania's brochure accurately pointed out that, "there is very little data available as to how effective prophylactic surgery is at reducing breast cancer risk" (20).
The Task Force called for external review of proposed informational material to make sure the data were interpreted correctly and that test limitations (such as imperfect sensitivity and PPV) were indicated. It also encouraged the College of American Pathologists and the American College of Medical Geneticists, which have programs for evaluating clinical laboratories performing genetic tests, to place greater emphasis on the pre- and postanalytic phases of testing in their programs for evaluating laboratories and to seek greater input from consumers and genetic counselors on educational, psychological, and counseling issues [Ref. (7), p. 55].
| Provider Education |
|---|
|
|
|---|
The Task Force made several recommendations to rectify the problem of inappropriate ordering of, and counseling about, tests. One of its principles was that "People being offered testing must understand that testing is voluntary. Their informed consent should be obtained". To accomplish informed consent, "healthcare providers must describe the features of the genetic test, including potential consequences, to potential test recipients" [Ref. (7), p. 12]. The Task Force emphasized this point: "The responsibility for providing information to the individual lies with the referring provider, not with the laboratory performing the test" [Ref. (7), p. 12]. To improve the chance that physicians will convey accurate information the Task Force recommended that, "Hospitals and managed care organizations ... should request evidence of competence before permitting providers to order predictive genetic tests defined as needing stringent scrutiny or to counsel about them. Periodic, systematic medical record review, with feedback to providers, should also be used to ensure appropriate use of genetic tests" [Ref. (7), p. 67]. To improve the chances that physicians of the future would have a better grasp of genetics, the Task Force commented, "It will improve the likelihood that genetics will be covered in [medical school and specialty training] curricula if relevant genetics questions are included in general licensure and specialty board certification examinations, and if correctly answering a proportion of the genetics questions is needed to attain a passing score" [Ref. (7), p. 66].
| Appropriate Use of Genetic Tests by Consumers |
|---|
|
|
|---|
| Conclusions |
|---|
|
|
|---|
For obvious reasons, commercial test developers will resist government regulations that might impede the speed with which their products reach the market or restrict the size of those markets. From the problems encountered thus far, I cannot conclude that the public is well-served by a laissez-faire policy. In one area in particular, a simple change in Federal policy could correct many of the problems: the FDA could regulate tests marketed as services the way it markets other medical devices, including tests marketed as kits. The Task Force could not reach consensus on this recommendation, but a majority agreed that the FDA's proposed regulation on analyte specific reagents (since issued)requiring manufacturers of such reagents to register with the FDA and to follow good manufacturing practicesdid not address the problem (26).
Despite the publicity given to new tests, it is not evident that the FDA would immediately be flooded by a host of applications or that it would have to deal with a tremendous backlog. If this were to happen, the Agency could set priorities for review by using criteria for "stringent scrutiny" as suggested by the Task Force. I do not think there will be a deluge. Because of the difficulties of establishing new genetic tests with current technology; relatively few companies are investing in it (6). With the development of oligonucleotide chips and other advances, the technology is likely to change, and more manufacturers may be drawn in. One could argue that the FDA could wait until then, particularly because chip technology lends itself to kit development more than the current technology, which often involves some sequencing. If chips were marketed as kits, they would be covered by current FDA policy. However, many people are being exposed to genetic tests marketed as services, and even chips may be marketed that way. If the FDA is to live up to its commitment to assure that medical devices, which include genetic tests whether marketed as services or kits, are safe and effective it will need to extend its regulatory authority. If this policy change were to be made, commercial developers will discover, I believe, that the market for their tests will expand as the doubts that consumers and providers have about their benefits and risks are removed.
| Footnotes |
|---|
The work reported here was supported by the National Human Genome Research Institute, NIH (grant HG-00026).
1 Nonstandard abbreviations: DOE, US Department of Energy; ELSI, Ethical, Legal, and Social Implications of Human Genome Research; PPV, positive predictive value; FDA, Food and Drug Administration; IRB, institutional review board; BTC, biotechnology company; and NPO, nonprofit organization. ![]()
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
L. A. Barclay and K. S. Markel Discrimination and stigmatization in work organizations: A multiple level framework for research on genetic testing Human Relations, June 1, 2007; 60(6): 953 - 980. [Abstract] [PDF] |
||||
![]() |
H. H. Ezzeldin and R. B. Diasio Genetic testing in cancer therapeutics. Clin. Cancer Res., July 15, 2006; 12(14): 4137 - 4141. [Full Text] [PDF] |
||||
![]() |
S. E. Waisbren, S. Albers, S. Amato, M. Ampola, T. G. Brewster, L. Demmer, R. B. Eaton, R. Greenstein, M. Korson, C. Larson, et al. Effect of Expanded Newborn Screening for Biochemical Genetic Disorders on Child Outcomes and Parental Stress JAMA, November 19, 2003; 290(19): 2564 - 2572. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. O. Keku, T. Rakhra-Burris, and R. Millikan Gene Testing: What the Health Professional Needs to Know J. Nutr., November 1, 2003; 133(11): 3754S - 3757. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. S. Young and D. G.B. Leonard Issues in Genetic Testing Clin. Chem., June 1, 1999; 45(6): 915 - 926. [Full Text] [PDF] |
||||
![]() |
D. S. Young Issues in Genetic Testing Clin. Chem., May 1, 1999; 45(5): 725 - 725. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |