Feature Story

Watchdog role up for grabs in genetic testing

Stigmatization fires the debate
Conflicting data on genetic testing quality

July 2001
Anne Paxton

A disease you know you will get before the age of 50. A job for which you’re disqualified based on a blood test. A child you’re afraid to conceive. To many people, these are not imaginary products of genetic testing—they have already proven quite real. And among those who have not yet been directly affected, there is widespread disquiet at the prospect of genetic testing potentially so revealing that it could foretell a person’s entire medical history.

Few would disagree that the quality of such high-stakes testing should be of paramount concern to professional organizations, regulators, and the public. Numerous state laws have already been enacted to regulate genetic testing while numerous federal advisory panels and regulatory agencies consider proposed approaches.

"There are a whole host of issues beyond the traditional scientific questions of whether a test is analytically sound and clinically makes sense. They are not unique, but are particularly highly charged with genetic testing," points out Steven Gutman, MD, director of the Division of Clinical Laboratory Devices at FDA’s Center for Devices and Radiological Health.

However, the laboratory community and federal policy-makers are far from consensus on how much additional regulation is needed to ensure genetic testing quality.

A multi-tiered "genetic action plan" proposed recently by the Food and Drug Administration under the aegis of the Secretary’s Advisory Committee on Genetic Testing, or SACGT, would greatly expand federal regulation of genetic testing. Still in draft form, it includes a premarket review template that would regulate in-house developed tests. But the College and other professional societies in molecular medicine believe the oversight they provide in voluntary partnership with federal agencies is preferable to a heavy-handed regulatory scheme.

Molecular geneticist Walter Noll, MD, chair of the CAP Molecular Pathology and Genetics Cluster and professor of pathology at Dartmouth-Hitchcock Medical Center, Lebanon, NH, warns that the genetic action plan "risks costly administrative burdens, stifling of test improvement and new test development, and restriction of patient access to care."

Concern about genetic testing has swelled over the last decade, as the number of genetic tests has mushroomed from a few tests for Duchenne muscular dystrophy and cystic fibrosis to an estimated 750 tests performed by some 400 to 500 laboratories. Alongside that growth, a series of task forces, commissions, and panels have convened to consider additional regulation of genetic testing.

Chief among them is the SACGT, chartered in 1998 to advise the secretary of Health and Human Services on medical, scientific, ethical, legal, and social issues raised by the development and use of genetic tests. The committee was asked in 1999 to assess the adequacy of oversight of genetic tests and make recommendations if warranted.

The FDA’s action plan calls for registration and listing, medical device reporting of adverse events, guidance templates and standards, a public database containing information about tests, test classification, and finally, a phased-in review. But the FDA component is only part of what SACGT is proposing. SACGT recommendations call for a multi-agency approach to genetic testing, with the FDA actually reviewing the tests, CLIA being augmented, and with CDC taking responsibility for writing the enhanced regulations, says Judith Yost, MA, MT(ASCP), director of the CLIA program for the recently renamed Center for Medicare and Medicaid Services (formerly the Health Care Financing Administration).

Last year, CDC published a Notice of Intent to add a genetic testing specialty under CLIA and to solicit comments to the remaining Clinical Laboratory Improvement Advisory Committee recommendations for CLIAfor genetic testing. The proposed rule, being prepared now for release next year, will address preanalytic, analytic, and postanalytic phases of the testing, as well as issues of informed consent, confidentiality, counseling, and clinical appropriateness of the tests.

The FDA’s Dr. Gutman said the genetic action plan is on course. "There is a lot of opportunity to continue to mold it. There is a deliberate objective to maintain flexibility and an appropriate focus in this plan. And the plan remains a high-priority new work item" for the FDA.

At the same time, there is some feeling that the climate for a large, new regulatory framework may have changed, and that the Bush administration, which put temporary holds on pending regulations from the Clinton White House, may not approve of the direction SACGT has taken. A letter jointly signed by the College, American Association for Clinical Chemistry, and several other leading laboratory groups was sent May 16 to secretary of HHS Tommy Thompson, requesting a review of former HHS secretary Donna Shalala’s directive setting out expanded regulation of genetic testing.

The College, AACC, and the others are particularly concerned about the role proposed for the FDA. The joint letter said: "Under the directive, laboratories developing and conducting in-house genetic tests would essentially be considered ’manufacturers’ and subject to direct FDA oversight—thus forcing them to meet specified regulatory requirements beyond the existing CLIA standards." The proposal to register all genetic testing laboratories is redundant, while other proposals such as developing a test categorization process may become burdensome and costly, says Phillip Bongiorno, assistant director for public health and scientific affairs in the CAP Division of Government and Professional Affairs.

The letter also said the action plan may be targeting the wrong issue. "Our review of the situation indicates that the majority of concerns in this area pertain to the use of the data (i.e., use of data by health plans and/or employers) as well as a need for more physician and patient education (e.g., improper informed consent, tests ordered for the wrong indication, and patients not fully understanding their test results), rather than in the quality of the laboratory procedure."

The FDA template was actually developed in cooperation with the professional organizations and was intended to augment the existing clinical laboratory inspection process administered by HCFA under CLIA, says Debra G.B. Leonard, MD, PhD, past president of the Association for Molecular Pathology and director of the molecular pathology laboratory at the University of Pennsylvania Hospital, Philadelphia.

But, speaking at the May SACGT meeting, she said making use of the FDA review template mandatory for all in-house developed tests by all clinical laboratories in the United States "will be an administrative nightmare for both the laboratory and the reviewing agency."

The College, in fact, has convened a working group that is exploring how CAP’s existing inspection and accreditation requirements in molecular medicine might be modified to meet the quality assurance concerns of the SACGT and other groups.

For its part, the FDA is looking for cooperative solutions and is treading carefully, Dr. Gutman maintains. At least three independent advisory panels, he recalls, have expressed concern about the status quo but only one has posited a role for the FDA. "The other two are more nebulous or silent, so my perception is there is interest in seeing incremental change."

"The FDA is very anxious to find ways it can help," he says. "We’re certainly anxious to do that gently and not to chill technology. We are very cognizant of the delicate balance that is present here, and from an agency standpoint we are in an exploratory mode." Downplaying the regulatory scheme that is being contemplated, he adds, "Probably the most intrusive thing we’ve put on the table that we are actually seriously entertaining is registration and listing."

"That’s clearly a change from where we’ve been. But we’re very interested in interacting with professional societies to see what kind of synergisms are available, and how we can be here to help," he stresses. "It’s not our issue to create problems for genetic laboratories." He declines to say that FDA regulation is inevitable and says even the proposed action plan continues to be "vetted."

"Anything is possible," he says.

The mere task of agreeing on a definition of genetic testing has proved difficult for the SACGT. "Genetic testing is a term we use to describe lots of different kinds of testing, depending on who’s talking about it," says Dr. Noll. "Some people would include in genetic testing anything that has to do with DNA or RNA analysis, extending to identification of microorganisms."

Key among the definitional issues is a distinction between tests for inherited mutations versus acquired ones. "What they really are concerned about is heritable disorders, not somatic genetic changes," Dr. Noll says. "But the SACGT has allowed itself to include somatic mutations as well, and that is making the discussion unnecessarily complicated."

"Some of the tests we do identify mutations that are present in all cells of a person’s body and can be passed to the next generation," he says. "These tests are quite distinct from those that test for mutations that may accumulate in clusters of cells in one of our tissues and cause those cells to develop into a cancer. That’s a whole different kettle of fish, because you can’t pass these mutations on to your children, they tell no information about your parents or siblings, and there’s nothing there of interest to employers or insurers."

The Clinical Laboratory Improvement Advisory Committee has been one entity suggesting an all-embracing definition. "They felt that since the technology is the same, whether the mutation is inherited or acquired, why shouldn’t a lot of the laboratory standards be the same?" says Wayne W. Grody, MD, PhD, chair of the CAP Molecular Pathology Committee and the CAP/ACMG Biochemical and Molecular Genetics Resource Committee. "They have a point, but although the technical issues are the same, the pre- and postanalytical issues are vastly different. That’s why we wanted to keep them separate."

Dr. Grody, who is professor in the Departments of Pathology and Laboratory Medicine and Pediatrics at the UCLA Medical Center and director of its diagnostic molecular pathology laboratory, says the College would prefer to have a genetic counseling clause in the definition to be sure that genetic testing is distinguished from other tests. "We want to differentiate it so they wouldn’t have it lumped in with tests for tumor markers, or even serum cholesterol, which in a way is a genetic test up to a point. Whenever you do a cholesterol test you’re essentially diagnosing a condition, and it’s even predictive because of the increased risk of coronary disease, but it hasn’t been done with informed consent, and we didn’t want tests like that subject to the same guidelines."

In Dr. Noll’s view, much of the pressure to regulate stems from a combination of genetic testing’s powerful predictive ability and the unrelated fact that most of the tests are in-house developed tests, often termed "home-brew."

In a clinical chemistry laboratory, he says, "virtually all the tests are done with FDA-approved reagent kits on FDA-approved instrumentation, and it’s often together in one package. So we have to do the tests carefully, but basically a lot of the responsibility for saying that this is an appropriate, well-designed test that gives you a certain kind of information is taken for us by the FDA."

In genetic testing, however—as in other laboratory areas such as flow cytometry, immunohistochemistry, and hematopathology tests for leukemia and lymphoma—most of the reagents are non-FDA approved; they are developed in-house or adapted from the literature or from a colleague at another institution, Dr. Noll says.

At the Center for Medicare and Medicaid Services, Yost says that while in-house tests are already regulated under CLIA, "we do not do the clinical validity evaluations that the FDA would ordinarily do on test review. The CLIA requirements become more stringent when the test does not have FDA approval. The high-complexity category kicks in and the test has to be validated analytically."

The proposal that laboratories performing in-house genetic tests register and list with the FDA is a way of incrementally strengthening oversight of genetic tests, says Dr. Gutman, who believes the template is general enough not to be intrusive. "It does not attack the broad universe of tests all at once."

This approach reflects the FDA’s inclination to regulate genetic testing but not tightly control it. The agency has already adopted controls over the active ingredients (analyte-specific reagents) laboratories use to perform genetic tests. "The FDA did not really want to go too far down the road of directly regulating laboratories when it put out the analyte-specific regulation," says Vince Stine, director of government affairs for the American Association for Clinical Chemistry. "The purpose was to ensure some minimum level of oversight, to ensure those facilities doing home-brew tests were using reagents that met certain criteria."

Adds Dr. Leonard: "They hoped that companies, rather than taking their entire kit through the FDA, could simply register the analytes and reagents they’re marketing. But there hasn’t been a great boom in the development of analyte-specific reagents."

Included in the newly proposed FDA template would be the test’s intended use, indications for its use (conditions for which it is performed and the purpose, such as diagnostic or prenatal testing), the method category, methodology, examples of test results, analytical validity, quality control procedures, clinical validity, clinical interpretation, assay limitations, and clinical utility.

The FDA’s idea for a template emerged after the SACGT tried and abandoned a scheme for categorizing genetic tests as "high scrutiny" and "low scrutiny," Dr. Noll says. "The template is simply what any responsible laboratory person does when one brings a test in house—make sure it has all the elements, it’s done on the right people, with the right samples, the right test procedures, validation, and good reporting."

In fact, most of the points are already in the CAP checklists for the Laboratory Accreditation Program, and many people engaged in discussions with SACGT members urged that the FDA "not apply the template to every home-brew test out there, but just put it out as a guide or a tool that professional societies could use in looking at genetic testing," he says.

One of the reasons for starting to think of separate categories, he explains, is the feeling that laboratory directors should have a background in genetics. Many people in the field are asking if a generally educated pathologist without special qualifications is qualified to run a molecular genetic testing laboratory. "What’s happening is we are certainly including genetics in our training programs, and a new specialty certification in molecular genetic pathology has been jointly developed by the American Board of Pathology and the American Board of Medical Genetics," Dr. Noll reports.

This response is part of a general view that professional organizations are the best qualified to oversee genetic testing quality. "Yes, the technology is moving incredibly fast, but I don’t know why that should scare us so much," Dr. Grody says. "Our view is that as long as you have accredited laboratories with board-certified directors who know the field, it’s no different than a surgeon assessing the latest operative techniques and whether to use them. We don’t have the FDA looking over his shoulder in the operating room. We credit the board-certified surgeon with the expertise to decide. I feel it’s the same for board-certified geneticists and molecular pathologists. They know the field, they can assess the clinical utility of the tests and know how to validate them on a series of normal and mutant specimens before putting them into practice."

Dr. Leonard agrees. "None of us is making products to sell for someone else to use," she says. "For me this is my medical practice. I don’t see patients. Developing these in-house tests, interpreting them, figuring out how to use them—that’s my medical practice."

Professional laboratory organizations are also concerned that FDA regulation of this practice may stifle innovation. "A lot of times home-brew tests are on the cutting edge," AACC’s Stine says. "Typically what happens over time, as things become more standardized and the field is more developed, the manufacturers start developing kits, and people may move from home-brew to the kits. It’s an ongoing process—that’s how the private sector works."

The registration process the FDA is proposing could be unwieldy if it has to be followed for every test and for different applications, he says. "There are a lot of unanswered questions. What happens during the review period? Do you stop doing the test? The FDA may be thinking just a little more regulation here and there won’t be a problem, but laboratories aren’t getting paid more now, and they won’t get paid for this."

Dr. Leonard doesn’t think the template itself is the primary problem; rather, she fears the unnecessary delays the regulatory process will bring. "My concern is, having looked at New York State molecular diagnostic colleagues who have to submit information to the state for review before actually starting tests, they never get answers back. They can never start their tests unless they start them without the approval."

She also believes the FDA lacks the staff to oversee genetic testing. "It’s a matter of expertise as well. The FDA doesn’t have a lot of geneticists to look at genetic tests and the information that would be submitted and assess whether they’re working properly," she says.

As an alternative, the College is working to develop a framework that would preemptively address the issues that trouble the FDA. Says Dr. Noll: "In some cases the public’s concerns are probably being adequately addressed, and in some cases not. What the College is doing right now is saying we can talk about those things and perhaps make changes." The working group is looking at everything the CAP does in the Laboratory Accreditation Program and extracting that which is relevant to genetics. "It’s there, and you can say it is. But it’s hard to go directly to the information we want that is relevant. We’re going to pull all that out and put it on the table to compare with what SACGT, the CDC, and CLIAC have been saying is desirable."

Confidentiality and informed consent will be part of the framework. "The College is strongly in favor of the strictest medical record confidentiality, but we would not be in favor of anything that puts impediments in the ability of doctors to talk to one another about patients," Dr. Noll says. "We have to be careful about how the provisions are written so there are not inadvertent effects."

For example, New Hampshire’s law proscribing discrimination based on genetic testing, he says, was written too broadly. "People view it as a good law because it does protect people against employment discrimination, but it doesn’t distinguish between my communicating with a colleague and my talking to an insurance company."

Similarly, though the College supports informed consent requirements, they should not be required for the use of materials like leftover blood for research, he emphasizes. "CAP thinks there are adequate protections in place without requiring specific informed consent to use these materials in research."

The College’s working group is hoping to redirect a genetic action plan that, under a worst-case scenario, might impose on laboratories a costly and cumbersome process that manufacturers themselves have done everything to avoid. "I know for a fact that there isn’t a single test kit manufacturer in the queue at FDA to have any kind of a kit that deals with heritable disease testing for approval," Dr. Noll said at the May SACGT meeting.

Dr. Gutman agrees that the potential regulatory burden is a real concern. "There actually isn’t an intent to have the usual review process apply here," he says, noting that applying standards such as good manufacturing practices would "in a single fell swoop, close all genetics labs in this country."

At the upcoming August meeting of the SACGT, the FDA plans to continue discussing these issues, then finalize the template in the fall. "We at the agency have constructed something that is as light a touch as we think we can get by with," Dr. Gutman says. But, he concedes, "[The FDA] has a long history of review that is pretty data-intense and pretty labeling-intense. So I’m frankly open to alternatives."

Anne Paxton is a freelance writer in Seattle.