The day many laboratorians hoped would never come arrived with a flurry of activity and ended with no certain outcome. In announcing the public meeting on laboratory-developed tests (LDT) that took place July 19 and 20, officials of the Food and Drug Administration (FDA) declared the agency intended to forgo its policy of enforcement discretion and instead begin actively exercising oversight of LDTs.
Why the sudden change? At least in part the agency seems to be responding to the explosion of direct-to-consumer (DTC) genetic tests, which have been drawing closer scrutiny from FDA as well as from Congress. In its announcement of the meeting, FDA cited risks to public health posed by changing dynamics in how LDTs are offered and by whom, highlighting the trend whereby increasingly complex tests are brought to market by commercial entities without FDA review and sometimes directly to consumers.
All sides of this debate, from FDA assertions of its right to regulate LTDs to manufacturers and professional association proposals for new regulatory regimes, have been aired publicly and privately for nearly two decades, with observers inside and outside the agency frequently noting that FDA does not have the resources to tackle formal review of every LDT that clinical labs perform. But the launch of formal discussion at the FDA public meeting and the sternness of the agency’s language in its announcement underscore the depth of concern at FDA and its desire to finally take action.
Now that agency officials have put forth a full-throttle effort to gather information and set priorities for oversight of LDTs, it’s time for laboratorians to pay even closer attention and ensure their voices get heard as FDA mulls its options, said Bill Clarke, PhD, DABCC, who represented AACC in a presentation at the public meeting. Clarke is associate professor and director of therapeutic drug monitoring and toxicology in the department of pathology at Johns Hopkins University School of Medicine in Baltimore. “This is going to happen, it’s just a question of how. We can’t just say anymore that we won’t accept regulation of LDTs,” Clarke said. “Given the current regulatory environment, there is not enough agreement among labs, patient advocates, and other groups to say that this can get stuffed back in the box. Now it’s a question of working with FDA and other stakeholders to make sure it’s done in a way that’s not obstructive to labs.”
The Spark: DTC Genetic Tests
The surge in DTC genetic tests being offered primarily via the Internet set off the new focus on LDTs at FDA. These tests, which have never been reviewed by the agency, are performed in the manufacturers’ own accredited labs as LDTs. On May 11, one month before FDA announced its 2-day meeting on LDTs, Pathway Genomics and Walgreens announced plans to sell a saliva collection kit for Pathway Genomics’ DTC Discover Your DNA genetic tests in the drug store giant’s locations nationwide. The test advertised that it would use the customer's DNA to assess risk for breast cancer, heart attack, and a host of other conditions.
Two days later, after receiving a letter from FDA requesting an explanation of why the test should not require the agency’s review, Pathway Genomics and Walgreens announced that they were putting their plans on hold.
FDA then ramped up pressure on DTC firms with five more letters in June. These “untitled letters”—not technically as severe as official warning letters, went out to 23andMe, deCODE genetics, Illumina, Navigenics, and Knome. Similar to the Pathway Genomics letter, FDA again noted it considered these companies’ products to be medical devices subject to its authority and asked the companies to explain why their tests should not require premarket review. In four of the letters, however, FDA noted that it did not consider the tests to be LDTs in the first place, for a variety of reasons.
For example, in Illumina’s case, FDA was drawing attention to an instrument, not a test, noting that “the Illumina Infinium HumanHap550 array…is labeled For Research Use Only. Yet Illumina is knowingly providing the HumanHap550 array to 23andMe and deCODE Genetics for clinical diagnostic use without FDA clearance or approval.” In the case of 23andMe and Knome, the agency pointed out that these companies’ tests use software at external laboratories to produce results.
With the exception of Navigenics, each letter pointed out subtleties about the way in which the test was performed that disqualified it from LDT status. However, the very day that FDA’s public meeting was getting underway in Hyattsville, Md., the agency issued 14 new letters to DTC genetic testing companies, this time making no mention of whether the tests could be considered LDTs. Rather, the communiqués simply stated that each of these tests “appears to meet the definition of a device as that term is defined in section 201(h) of the Federal Food Drug and Cosmetic Act.”
More Questions than Answers
FDA began the public meeting by making its case that the great benefits of in vitro diagnostics, especially in advancing personalized medicine, also carry great risks, and that FDA has a responsibility to mitigate these risks to patients. Joshua Sharfstein, MD, principle deputy commissioner at FDA, cited the recent editorial in The New England Journal of Medicine by FDA commissioner Margaret Hamburg, MD, and Francis Collins, MD, PhD, director of the National Institutes for Health (NIH), which praised advancements in genetics but denounced the fact that so many new tests do not pass through FDA and instead are marketed as LDTs. “This undermines the [FDA] approval process that has been established to protect patients, fails to ensure that physicians have accurate information on which to make treatment decisions, and decreases the chances that physicians will adopt the new therapeutic-diagnostic approach,” they wrote.
While FDA took a strong stand on the need for oversight of LDTs, agency officials were equally eager to note that no definite frame work had been worked out and that they did not want to disrupt testing. “Although FDA has decided to exercise authority over LDTs, we have not made any decisions about how we will exercise that authority. That is what this two-day public meeting is about—we want to hear from you,” said Jeffrey Shuren, MD, JD, director of the Center for Devices and Radiological Health (CDRH).
In their presentations, Courtney Harper, PhD, director of the Division of Chemistry and Toxicology Devices for the Office of In Vitro Diagnostics (OIVD) and Elizabeth Mansfield, PhD, director of Personalized Medicine for OIVD, fleshed out the agency’s rationale for regulating LDTs and described the circumstances that led up to the decision. Both emphasized the “bifurcated pathway” that the agency now feels is broken. One the one side, manufacturers sell test to labs as kits that go through FDA premarket review or 501(k) clearance, and on the other side labs are developing their own tests with oversight from the Center for Medicare and Medicaid Services (CMS) or a third party accreditor, such as the College of American Pathologists (CAP). This second pathway includes all LDTs, but significantly, also includes a great level of variation in the type of tests offered and how they are marketed—the center of focus for FDA.
This concern was clear in agency’s letter to Illumina questioning its marketing of the Infinium HumanHap550 array instrument. FDA officials also linked their alarm over LDTs to the use of reagents marketed as research use only that also found their way into clinical use via LDTs.
“The types of LDTs offered 30 years ago continue to be offered today in many cases. There is still a lot of testing out there that requires expert pathologist interpretation and that is performed because there is unmet need,” said Harper. “However, there has been a change in LDTs in the United States since 30 years ago when enforcement discretion began. The volume and type of LDTs have grown exponentially. Today there is a higher proportion of LDTs in commercial labs and biotechnology companies setting themselves up as laboratories.”
Harper noted that in this commercial model outside of hospitals, there frequently is little or no clinician-pathology-patient relationship, and these tests are often broadly advertised and “aggressively marketed,” in many cases directly to consumers.
Mansfield offered some hints as to FDA’s thinking on how it can tackle LDTs without disrupting patient care, especially considering the agency’s limited resources. First, she cautioned labs to “plan for some re-assessment across the board,” suggesting that in order to give more attention to certain high-risk LDTs, other tests could be down-classified to a lower risk status and therefore require less arduous review from FDA. Another idea was to try pilot programs in which oversight of lower-risk LDTs would be exercised through third-party accreditors and coordinated under the Clinical Laboratory Improvement Amendments (CLIA) through CMS. Mansfield also promised that FDA would consider the cost to labs under any new regulation and that the agency understood that “lots of outreach and education may be needed.”
The Lab’s Perspective
A special CLN survey conducted in July shows that while even among readers there is widespread anxiety about LDT regulation, there are significant disagreements about FDA’s role and how high-risk LDTs ultimately should be dealt with (See Box, below).
To see more survey results, as well as comments submitted as part of this survey, click here.
If there is any consensus from the meeting, it’s that the greatest concern on the part of FDA and stakeholders is validation of the clinical utility of genetic testing. “Most of the focus was on genetic testing—nobody talked about immunoassays, and there was just one presentation on chromatography mass spectrometry,” noted Clarke. Another common refrain by FDA and diagnostic manufacturers was the so-called unlevel playing field—the idea that LDTs are a kind of loophole for certain companies to set up testing in their own lab and evade FDA review, while manufacturers selling kits to labs bear the burden of premarket approval or 510(k) clearance through the agency.
Lab groups shot back that for laboratorians developing their own tests, laboratory accreditation, backed up by CLIA, already works to ensure high-quality testing, and as a whole, is not directly comparable to FDA regulation of test kits. “A good point was made that while manufacturers have to deal with the premarket review process, as a lab that develops LDTs, we have to deal with the laboratory accreditation process, often for the hospital and then also for laboratory,” said Clarke. “So in a lot of ways, they’re comparing apples and oranges, and I think some folks did a good job of pointing that out. At the same time, others really wanted to hammer the point that everyone should play under the same rules regardless, and this became an ongoing theme of the meeting.” In addition to Clarke’s presentation, AACC submitted comments to the agency, which are available at www.aacc.org/gov.
The critique by FDA and some diagnostic manufacturers has been that lab accreditation does not go far enough when it comes to LDTs, particularly that only analytical validity is required, leaving gaps in clinical validity that would be covered by formal FDA review. Many at the meeting took issue with this portrayal.
“Some laboratorians pointed out that we are already regulated by CMS through CLIA, and we are required as medical directors to ensure the clinical validity of the test,” said Clarke. “The FDA stated position was that labs making LDTs only have to ensure the analytical validity of these tests, and I’m not sure that that’s entirely the case. In terms of the things that are spelled out under accreditation—proficiency testing, quality systems—the focus is certainly on the analytical piece of testing. We want to make sure that we get the right answer out. However, the clinical validity is still addressed, from my view, in that the medical director of the lab still has to function as a clinical consultant, and part of that is evaluating clinical validity. I think that’s an important point not to be lost, that it’s part of our job to function as clinical consultants.”
Probably the biggest concern for many in the lab is that more FDA regulation of LDTs, on top of those in place from CMS through accreditors, will make important LDTs too expensive or burdensome to perform, ultimately affecting patient care.
More regulation from FDA could have unintended consequences and could easily go too far, forcing labs to restrict what they do, said Lawrence Silverman, PhD, DABCC, DABMG, professor of pathology and scientific director of molecular diagnostics at University of Virginia Health Sciences Center in Charlottesville, Va. “If you tell me that I’m going to have to go through some kind of regulatory procedure that turns out to be onerous in order to do LDTs, that’s going to restrict what I do, because I’m not going to set these tests up,” he said. “For some of the tests it’s just not going to make any sense.”
Silverman has worked on the Clinical Laboratory Advisory Committee and been actively involved with FDA and other agencies in efforts to deal with new challenges posed by genetic testing. “We have a very complex playing field right now: on one hand the promise of the new technology we have out there, on the other hand the concern that many people in the private and public sector have about technology going wild, and we have to balance that, especially considering the fact that our regulatory agencies are overwhelmed,” he said.
Undercover GAO Report Adds Pressure
As if the controversy over LDTs—and especially the DTC market—wasn’t hot enough, a Congressional hearing on DTC tests held only days after the FDA meeting could fuel new scrutiny from Washington and the public. The House Committee on Energy and Commerce Subcommittee on Oversight and Investigations held a hearing July 22 on “Direct-To-Consumer Genetic Testing and the Consequences to Public Health” in which the chair, Representative Bart Stupak (D-Mich.), at one point characterized some DTCs as “snake oil.” (Available online.)
One rather embarrassing piece of testimony for the DTC genetic testing companies came from the Government Accountability Office (GAO), which simultaneously released its report, “Direct-to-Consumer Genetic Tests: Misleading Test Results Are Further Complicated by Deceptive Marketing and Other Questionable Practices.” (Available online.) As part of GAO’s testimony, the committee watched a YouTube video with excerpts from phone calls to DTC companies by undercover GAO investigators posing as customers . In the video, the undercover investigators received bizarre advice from DTC company representatives, including that the company could provide supplements to “repair damaged DNA” or predict in which sport that individual’s children would excel.
Although the GAO admitted that it did not conduct a scientific study, “but instead documented observations that could be made by any consumer,” those observations proved to be provocative and extreme examples of less-than-solid science and misleading advertising. GAO bought 10 tests each from four DTC companies, and then sent two DNA samples from five donors to each company. One sample from each donor was accompanied by factual background information, such as age or ethnicity, and another with false information. “GAO’s fictitious consumers received test results that are misleading and of little or no practical use,” the report said. The donors “often received disease risk predictions that varied across the four companies, indicating that identical DNA samples yield contradictory results…GAO’s donors also received DNA-based disease predictions that conflicted with their actual medical conditions.” The report emphasized that the most “disturbing” finding was that one of the companies told a GAO donor that her above-average risk prediction for breast cancer could be taken to mean that she is “in the high risk of pretty much getting” the disease, which GAO said its experts “found to be ‘horrifying’ because it implies the test is diagnostic.”
Clearing a Path Forward
Among the myriad of proposals set forth by speakers making public comments at the FDA meeting, one that appears to be gaining traction both within FDA and among laboratorians would take the shape of tighter oversight of low- and moderate-risk LDTs by lab accreditors and full FDA review only for high-risk LDTs. CAP has proposed a plan that would require a lab to submit analytical and clinical validation studies to the lab’s CMS-deemed accreditor, such as CAP, before the lab could offer anything but a low-risk LDT. Currently, such validation studies are only reviewed during a lab’s inspection, and the lab may be offering an LDT before an inspection occurs.
Taking advantage of lab accreditors who already know and operate within labs could allow to the agency to focus its resources more narrowly, while at the same time avoiding an undue burden on labs, said Gail Vance, MD, FCAP, who represented CAP at a moderated panel discussion during the FDA public meeting. “The accreditors are in the lab all the time—they know the labs, they communicate with the labs, and they also perform a continuous monitoring function,” she said. “I would argue that the FDA approval does not necessarily assure good test performance. So it’s not the end-all of ensuring patient safety. Monitoring is a key component of the CAP proposal for the oversight of LDTs. For example, FDA oversight includes post-market surveillance, but I think that probably occurs episodically at best. Rather, accreditation is an automatic, ongoing system that monitors laboratory testing on an established schedule, and that function is critical.”
The CAP proposal hinges on two controversial elements: how risk should be defined, and how clinical validity would be determined by the accreditor. Under CAP’s proposal, LDTs deemed high-risk and requiring FDA review are those that “predict risk of, progression of, or patient eligibility for a specific therapy to treat a disease associated with significant morbidity or mortality; and, the test methodology uses proprietary algorithms or computations such that the test result cannot be tied to the methods used or inter-laboratory comparisons cannot be performed.”
This high-risk category would include so-called in-vitro diagnostic multivariate index assays (IVDMIA), for which FDA released a guidance in 2007 that spells out the need for the agency’s approval. However, certain other tests that do not fit the IVDMIA category but which are performed only by a single lab might also fall into a high-risk level under the CAP proposal’s rubrics, said Vance. “We see high-risk as when the test methodology is linked to the test results in a non-obvious, non-transparent way,” she said. “This includes when there is the application of a proprietary algorithm, but it’s also, we think, those tests that are performed only in a single laboratory, so that there is neither a commonality of knowledge, nor the transparency of performance, and there may not be proficiency testing (PT) products available, because there is no incentive to develop PT for a single test in a single lab.” However, she admitted that FDA is likely to take issue with not automatically putting companion diagnostics in this high-risk echelon.
Vance also acknowledged that defining clinical validity can be a grey area, but that extant resources can be put to use for this purpose, such as recommendations from Evaluation of Genomic Applications in Practice and Prevention (EGAPP) Working Group. The EGAPP Working Group was established in 2005 with support from the Centers for Disease Control and Prevention (CDC) to systematically assess evidence on the validity and utility of emerging genetic tests.
CAP has also hired an outside consultant to collect evidence-based data on defining clinical validation and how a lab would sufficiently demonstrate that it has performed adequate clinical validation.
As FDA takes commentary from the meeting and other proposals into consideration over the next several months, CAP, AACC, and other organizations have pushed for the agency to work within CMS’s accreditation framework for the majority of LDTs that are well-understood and arguably low-risk. “I think the FDA is now grasping the reality of the situation. There are thousands and thousands of these LDTs,” said Vance. “I hope FDA can see the value of the accreditation system, and that what needs to happen is for FDA, CMS, and other agencies to talk to one another more frequently and combine efforts in some respects.”