Stakeholders Ask FDA to Educate Labs on Agency Thinking, Terminology Before Finalizing LDT Guidance

NEW YORK (GenomeWeb) – During the second day of a public workshop on the regulation of laboratory-developed tests (LDTs), industry stakeholders asked the US Food and Drug Administration to produce several publications to better explain how they can meet the agency's requirements.

Specifically, workshop attendees requested FDA publish a paper explaining how it gauges the risk a test poses to public health; put out a document outlining the areas where the agency will harmonize its regulations with existing requirements under the Centers for Medicare & Medicaid Services; and issue another, revised LDT draft guidance before publishing a final version.

"It's my understanding that the entire framework is predicated on risk classification," Amy Miller, executive VP of the Personalized Medicine Coalition, told FDA. Noting that labs are unaccustomed to FDA processes and terminologies, Miller and others at the meeting asked FDA for more clarification on how it plans to classify LDTs into high- (Class III), moderate- (Class II), and low-risk (Class I) categories. These risk classifications are critical for labs to understand, since a test's risk category maps to the premarket requirements it will have to meet.

For example, under FDA's draft oversight framework published in October, LDTs deemed to be Class I devices will not need to be submitted for premarket review or meet quality system regulations, but developers will have to notify the agency and perhaps register the tests. Meanwhile, over the first two years after finalizing the LDT guidance, the FDA plans to focus on bringing companion diagnostics and other high-risk/Class III LDTs under its oversight. Toward the tail end of its nine-year phased-in framework, the agency will focus on moderate-risk tests. Generally, LDTs in the Class II and Class III risk categories will have to garner premarket approval or clearance from the agency. 

"Although some … think it's easy and obvious that this test fits into this box … I think there's a lot of confusion about what [test] has the highest risk," Miller said. "And there is a lot of confusion about that line, where the end of highest risk ends and the beginning of moderate risk [begins]."

"Laboratories are very new to FDA's vision," she reminded the agency. The PMC issued a statement recently requesting the agency publish draft guidance documents clarifying the definition of risk and how FDA regulations will be harmonized with the Clinical Laboratory Improvement Amendments — the standards that labs have had to historically meet under the Centers for Medicare & Medicaid Services. Finally, the PMC wants FDA to incorporate stakeholders' suggestions in a second draft LDT guidance before finalizing it.  

The FDA is planning to issue guidelines on the types of LDTs that fall into its three-tier risk classification system, and garner input from expert panels to determine the classification for a particular test. But before that, lab industry players need to understand risk in the way the agency thinks about it, Andrew Fish, executive director at the diagnostic manufacturers' industry group AdvaMedDx, said at the meeting. As a first step, even before issuing draft guidance on risk classifications for LDTs, Fish recommended the agency publish a "document laying out the way in which you think risk should be assessed generally."

The word "risk" is "like the word 'strategy,'" he told FDA. "Everyone has a different idea of what strategy is. So, if you sit down and discuss strategy and you don't know what strategy means, then you can have a long conversation and not get anywhere."

Another area of concern for members of the lab industry is that they will have to meet FDA's quality systems regulations (QSRs) on top of quality requirements they are already subject to under CLIA. "In the draft LDT guidance, the FDA proposes to require a vast variety of laboratories to comply with QSRs but provides scarce details on how those QSRs will be applied to LDTs generally and to different laboratories more specifically," Sheila Walcoff, CEO of consulting firm Goldbug Strategies, told FDA.

At the meeting, Walcoff was speaking on behalf of the Coalition for 21st Century Medicine, which represents the interests of diagnostics firms and venture capital firms. "While the coalition understands and appreciates the need to enable flexibility in how a lab might choose to comply with FDA's QSRs, labs are already subject to existing quality systems regulations and often seek additional certifications," Walcoff said.

She further noted that the coalition believes that the FDA should first resolve questions unaddressed by its draft guidance regarding how labs can comply with its regulations – such as the QSR issue – before finalizing its framework. "It is essential that the FDA harmonize the QSR requirements with CLIA requirements at a more granular level, to prevent duplicate efforts and to ease the regulatory burden."

FDA is in the process of doing just this, according to Andrew Hoofnagle, assistant professor at the University of Washington's department of laboratory medicine in the diabetes and obesity center of excellence. According to Hoofnagle, FDA has conferred with CMS and come up with a crosswalk table that the agencies have yet to share publicly. "That's the first step as we try to begin to understand how the quality system regulations are going to affect the practice of medicine and clinical laboratories," he said. "The biggest … difference is design control."

According to FDA guidance on the topic, design controls are procedures that test developers use to guide the development and design of their devices. Through this systematic process, if deficiencies in the design of a particular test can be corrected early in its development, then it "increase[s] the likel[ihood] that the design transferred to production will translate into a device that is appropriate for its intended use," the FDA states.

"That's pretty complicated especially for LDTs that are already being used in the practice of medicine," Hoofnagle noted. "How to take design control for something that is already designed, produced, and implemented retrospectively is pretty tricky."

Many other questions regarding how labs can fit their procedures into FDA requirements came up during the first day of the meeting. For example, stakeholders questioned whether labeling an LDT with an intended use would mean they would have to police whether doctors were using that test off label. Many speakers asked FDA to clarify the specific instances when a lab would have to submit a premarket application for a test that is already on the market with FDA's blessing, but has been tweaked or updated in some way – a common practice within labs.

Before the FDA can regulate LDTs at all, however, it needs to catalog how many tests are available and for what purpose. The American Clinical Laboratory Association has estimated that there are some 10,000 labs in the US providing more than 100,000 LDT services. The FDA, however, has maintained that the actual number of LDTs on the market is currently unknown, necessitating a more formal survey. For this purpose, the FDA has asked labs to notify it about all the LDTs they provide, regardless of their risk class.

A number of speakers objected to the notification requirement, noting that for small labs with limited staff, having to provide the requisite information even on a handful of tests would be burdensome. In this regard, once again, there might be an opportunity for FDA and CMS to work together, several stakeholders noted at the meeting, since labs already provide to CMS a lot of the information FDA is requesting on LDTs for notification purposes.

Additionally, in developing an LDT repository, the FDA can draw on information and best practices from several existing registries tracking marketed genetic tests, such as NIH's Genetic Testing Registry. Since its launch in 2012, 431 labs worldwide (more than half of which are from the US) have voluntarily registered information on more than 25,000 tests. Of those, only 16 GTR-listed tests have FDA approval or clearance.

Wendy Rubinstein, director of the GTR, highlighted the importance of "explicit data collection" where the registry contains the conditions under which a test is performed. "The types of the data [included in the registry] really have to be relevant to the risk to the patients and based on the test results and how they're acted upon or not acted upon," she said. "Those are the important [data] elements."

But again, in the context of how LDTs are developed and performed, the FDA will need to specify when labs need to make a new notification on an already listed test. "What happens to a test that's tweaked?" Rubinstein posited. "That really pertains to how you define a test … and then you understand which elements of tweaking change it, and mandate a new notification."

Coverage of day one of FDA's public hearing on its planned regulation of LDTs can be found here.