NATIONAL HARBOR, Md. (GenomeWeb) — At the annual meeting of the Association for Molecular Pathology here this week, attendees gathered for a question and answer session with US Food and Drug Administration officials Alberto Gutierrez and Elizabeth Mansfield, who fielded 45 minutes of queries on the scope and details of the FDA's draft guidance on the regulation of laboratory developed tests.
AMP also kicked off the meeting with a major lobbying push, as a cadre of members descended on Capitol Hill on Wednesday to educate legislators on what they believe would be an overwhelmingly negative impact of FDA regulation on the molecular diagnostics industry and the patients it serves.
AMP has taken a clear stance that the majority of lab-developed tests should continue operating under CLIA and the Centers for Medicare and Medicaid Services and not be subject to premarket review by the FDA. Central to the group's opposition to FDA oversight is an assertion that LDTs, at least many of them, are not medical devices, but rather laboratory procedures.
The ability, especially of small labs, to develop tests in house and customize them to their own associated clinical practice is integral to effective patient care, members argued at the meeting. If every new iteration must be individually submitted and approved by the FDA, labs simply cannot and will not continue to develop and customize assays in this way. According to association members, this will negatively affect patients.
However, even amidst the association's united opposition, many meeting attendees demonstrated lingering confusion about numerous practical aspects of the FDA's proposed guidance, including how it will determine a test's risk classification, which establishes whether and how soon a lab must submit it for FDA review.
The LDT draft guidance is currently open for public comment. The agency has allowed 120 days, longer than usual, to allow extra leeway around the holiday season.
During their panel, Mansfield, director of personalized medicine, and Gutierrez, director of the FDA's Office of In Vitro Diagnostics, said that the next official public meeting to discuss the draft is tentatively scheduled for Jan. 8 or Jan. 9. The date will not be officially solidified for another few weeks, Gutierrez said.
In the meantime, both officials expressed sincere desire on the part of the agency for members of the MDx community to comment on and review the draft guidance, especially in regard to the test categories the FDA intends to continue enforcement discretion over, and specifically whether any should be added or subtracted.
A central question among both AMP members and general meeting attendees was why the FDA must deal with small labs developing and modifying tests for needy subsets of patients the same way as large labs that purportedly develop LDTs then market them much like other IVDs as a way of avoiding FDA regulation.
Use of the LDT pathway as a means to skirt regulation is one that appears to be a significant factor in the FDAs acceleration of efforts to end its enforcement discretion over these tests.
Decades ago, as the first lab developed tests — at that time often called homebrew tests — were developed, the FDA decided, based on its available resources and understanding of the types of tests being provided at that time, that it would not regulate them and instead adopt what it calls enforcement discretion over LDTs.
"What has happened since then is that there has been a dual path to get on the market in the US," Gutierrez said during his presentation. "Clearly what is happening now is that there are many tests that are being developed by companies that look more like test manufacturers than they look like traditional LDTs."
But AMP members expressed wide frustration that the activity of small, often single hospital-based labs should be lumped in with those travelling this parallel track for more questionable reasons.
LDTs, though not currently regulated by the FDA, are monitored through CLIA guidelines by CMS, and must demonstrate initial analytical validity, as well as submit to periodic reassessments. Though CLIA does not require labs to demonstrate clinical validity for tests, AMP meeting-goers argued that small hospital labs work with their clinical colleagues to define and establish the clinical usefulness of the markers or analytes they measure in an evolving medical and therapeutic landscape.
Mansfield told attendees after the session that she has sympathy in this regard. But the problem is that it is not enough for the FDA to simply assume labs are doing this and doing it appropriately. The agency requires a way to objectively distinguish tests with true clinical validity from those without, hence the need for regulation.
Gutierrez, answering a question about the burden of regulation on small labs as compared to large manufacturers, reiterated that point.
"You may be under control, but there other laboratories that nobody is looking at and they look more and more like manufacturers than they do like you... You say differentiate us, but we need some help. If you have a better way — other than saying, 'I'm good and my people are good' — you should be putting that on the table," he said.
Meanwhile, several questions suggested that fundamental confusion about the details of the FDA's proposed strategy remain widespread. One attendee questioning Gutierrez and Mansfield during the panel seemed perplexed about the FDA's risk classification strategy for LDTs.
"If one of your major concerns is the clinical validity of a test, I'm not sure I understand your risk framework," the attendee from the University of Colorado said. "By definition, any test that has been approved [by the FDA] has had its clinical validity well-vetted ... so then risk has to come from analytic validity, and if you are operating in labs with ... tests subject to CLIA regulations, analytic validity is part of our validation ... Why would tests with established validity be considered high risk?"
Gutierrez responded, "The way we look at risk is by asking 'What is the risk of an undetected erroneous result?' Whether the test is approved or not there is no perfect test out there. So if you have an erroneous result and you don't detect it, what is going to happen to the patient?"
Mansfield further clarified that "risk of something not being clinically valid is completely different from the way we risk-classify devices for what type of submission they need to make.".
Essentially, whether and when a test must be submitted for FDA review depends on its potential risk to patients. This has nothing to do with a test's actual analytical or clinical validity, but rather what its indication is, as well as associated questions like what patient population it is directed towards and how it is or would be used, and therefore who it could hurt and how badly if something went wrong.
"Regarding clinical validity, we are worried some of these tests don't have the meaning ascribed to them, or don't have meaning at all," Mansfield said. "Risk classification determines how hard we are going to look at a test because a wrong result could cause serious harm."
Several other session attendees expressed anxiety about the fact that the FDA's draft guidance does not define exactly which tests will be regulated and which can continue to operate under enforcement discretion.
The draft guidance outlines three risk classes for LDTs, with differing requirements and timelines. For Class II and III tests, moderate- and high-risk, respectively, developers will be required to register, list, report adverse events for, and submit the test for 510(k) clearance and premarket review during a multi-stage phase-in period.
After the guidance is finalized, labs with LDTs have six months to notify the FDA what their tests are. Then, one year after the LDT guidance is finalized, labs offering the highest-risk assays — such as tests with the same indication as FDA-approved companion diagnostics, screening tools for asymptomatic patients, and high-risk infectious disease diagnostics — will be required to submit for premarket review. Additional high-risk tests will follow, and then moderate risk tests will be brought in only after about five years.
For Class I devices meanwhile, and for LDTs which meet an unmet medical need regardless of risk, the FDA will continue to practice enforcement discretion, leaving oversight responsibilities to CMS.
Meeting attendees, however, expressed confusion and skepticism about which tests might fall under the unmet medical need category. Many, in predicting that FDA regulation will prevent small labs from offering life-saving custom assays — for example to match cancer patients to targeted treatments — seem to be assuming that few tests, even those that arguably serve unmet needs, will be spared.
During a press conference yesterday following the FDA panel, Aaron Bossler, director of the University of Iowa Hospital's Molecular Pathology lab and co-chair of AMP's economic affairs committee, cited BRAF testing as an example.
Bossler said that though an FDA-approved assay is available as a companion diagnostic to guide treatment with BRAF-inhibiting drugs, the approved assay is limited because it only addresses the most common mutation known to confer sensitivity to the drug.
In fact, two BRAF mutation assays have been approved by the FDA to guide treatment with targeted drugs. The FDA approved two drugs from GlaxoSmithKline, Tafinlar (dabrafenib) and Mekinist (trametinib) alongside the THxID BRAF test from BioMérieux in 2013. BioMérieux's PCR-based BRAF test determines whether a patient's melanoma cells have the V600E or V600K mutation in the BRAF gene.
Prior to this, the FDA also approved Roche's cobas 4800 BRAF V600 Mutation Test in August 2011 as a companion diagnostic to Zelboraf (vemurafenib).
"We now know there are many other mutations that can be targeted using this intervention," Bossler said. "My lab and others have recognized that we need to be able to provide accurate testing to identify those mutations [so] we continue to use our laboratory-developed procedures because we can detect additional mutations, and provide more accurate information to clinicians so they can provide better care to patients."
In reality, from the FDA's draft guidance as it stands, it's not clear whether small labs offering their own evolving panels for cancer mutation testing (or other tests that might fall under the unmet need category) will or will not have to submit their tests for approval or clearance.
The FDA has said it plans to issue a separate guidance in the next year and a half clarifying the criteria for determining the risk classification of an LDT, but in anticipation of this better explanation, anxiety among testing labs is clearly high.
Elaine Lyon, president of the AMP board of directors and medical director of the genetics division at ARUP Laboratories, estimated during the press conference that the FDA's requirements, as understood from its draft guidance, will mean that ARUP will have to cease significant numbers of its tests.
Lyon told GenomeWeb that ARUP runs just under 1,500 LDTs in total, with new tests coming online all the time. The lab estimates that at least 100 if not hundreds could be considered high-risk.
According to Lyon, ARUP's analysis predicts that 95 percent of its potential high-risk tests, which she stressed could also be of very high benefit to the patient, may need to be eliminated. Up to 62 percent of moderate-risk tests could meet the same fate.
However, the precise number of high-risk or moderate-risk tests being performed by any lab cannot be determined until the guidance is finalized and labs begin the first step of the FDA's proposed process under which they notify the agency of their tests, and the FDA determines their risk category.
The FDA has maintained that, according to its legal mandate, if tests pose a high or moderate risk, do not serve an unmet medical need or fall into another category under enforcement discretion, and cannot fulfil the FDA's requirements for submission or garner clearance and approval based on demonstrated clinical validity, the agency cannot allow them to be sold.