August 26, 2024 | A final rule on laboratory-developed tests (LDTs) released by the U.S. Food and Drug Administration (FDA) in early May has done little to settle the longstanding debate about how LDTs ought to be regulated. The rule clarifies that in vitro diagnostic products (IVDs) are medical devices under the Federal Food, Drug, and Cosmetic Act, even when the manufacturer is a laboratory, and establishes a framework for FDA oversight of LDTs inclusive of a policy to phase out the agency's "enforcement discretion" policy.
The potential impact and path forward with the final rule and associated reclassification activities of the FDA was the topic of an informative plenary fireside chat on LDTs at last week’s Next Generation Dx Summit held in Washington D.C. Moderator for the discussion was industry veteran B. Melinda Cimler, Ph.D., CEO and founder of PandiaDx, who put the odds of the new rule waltzing into place on par with peace in the Middle East.
A day before the panel convened for their discussion, the Association for Molecular Pathology (AMP) filed a lawsuit challenging the rule, citing the far-reaching and long-lasting consequences to Assosciation members and patients. In it, AMP makes the case that disruption to the field would be substantial and harmful—points well highlighted at the summit by panelist Anna Scrimenti, associate director of public policy and advocacy for AMP.
“The change that the FDA made to the regulation is very small in a sense that it only added a half sentence to the definition of in vitro diagnostic; however, the impact was tremendously significant because... it will treat laboratory healthcare professionals as manufacturers and then enforce the medical device requirements upon all laboratory-developed testing procedures,” says Scrimenti. “AMP has long held that LDTs are not medical devices and... our members strongly believe that the ill-suited and onerous medical device pathway will have very devastating consequences for laboratory medicine and also patient care.”
Joining Cimler and Scrimenti for the dialogue were Lakshman Ramamurthy, Ph.D., vice president of regulatory affairs for GRAIL; Stefan Burde, Ph.D., director of global strategic business development IVD at TÜV SÜD, a Notified Body under the European Medical Device and In Vitro Diagnostic Regulations (IVDR); and Girish Putcha, M.D., Ph.D., principal with Precision Medicine & Diagnostics.
Enforcement discretion for LDTs, which has been FDA policy for decades, is to be phased out over a four-year period that begins May 6, 2025, Cimler says. The first phaseout will be around how a laboratory handles and evaluates complaints, adverse events, and recalls per quality system regulation requirements. These are critical components that get prioritized for review during FDA inspections, she notes.
The second year of the phaseout of enforcement discretion will be focused on compliance with labeling, registrational listing, and investigational requirements, continues Cimler. The labeling requirement for LDTs has caused a good deal of confusion, but she speculates that it refers to “anything outward facing” such as patient reports and technical information sheets.
Next month, the FDA is hosting a webinar about the labeling requirements for IVDs, including LTDs, which should add some clarity, she says. The agency has also published a lot of guidance on the labeling topic.
Starting in the third phaseout period, laboratories will be expected to be compliant with certain quality system requirements—specifically, design controls, purchasing controls, and acceptance activities, as well as the records requirement—followed by the FDA’s premarket review of high-risk IVDs and down-classification activities, which should help mitigate some of the impact, says Cimler.
In the final stretch, beginning the fourth year after release of the rule, premarket review for other low-risk devices will be in place. “Most low-risk devices really are exempt from premarket review, but that doesn’t mean you don’t have to comply with... all the other elements [of the rule],” Cimler says. “It’s just that you don’t have to submit a premarket application.”
Although LDTs are not a concept under IVDR, which applies to 27 EU member states, LDT-like tests are exempt from some of requirements of the regulation, according to Burde. Qualifying tests need to be manufactured in a single healthcare institution, or a network of hospitals located in the EU. “Any of the big U.S. labs would, by definition, not qualify for this.”
Exempt devices “still need to meet federal safety and performance requirements,” he continues, and “cannot be transferred from one legal entity to another”—including a company operating as two different legal entities in the U.S. and a Europe. These LDT-like tests also need to be manufactured under an appropriate quality management system at the lab and comply with either the international standard (ISO 15198) or the national provisions of the member state where they are located.
Additionally, exempt tests need to testify that the population it is targeting is not already being adequately addressed by an existing CE-marked device, says Burde. “As soon as there’s a CE-marked test, you have a much higher hurdle to justify that your device is in some way better.”
A publicly available declaration about the device is also a necessity, he adds. For the highest-risk Class D devices, it is essential also to maintain documentation about the manufacturing facility and process, as well as design and performance data, and to carry out post-market surveillance. Further, exempted devices must not be manufactured on an “industrial scale,” although the term is not specifically defined.
To regulators in the EU, it matters not if a lab in the U.S.—be it manufacturing FDA-approved tests, or those certified by the Clinical Laboratory Improvement Amendments (CLIA) of the Centers for Medicare and Medicaid Services (CMS) or the Clinical Laboratory Evaluation Program (CLEP) of New York State—as soon as it renders a result for a European patient it must comply fully with the IVDR, says Burde. “Exemption doesn’t apply because you’re not located in the EU, and the member states can basically say we don’t like what you’re doing, so cease and desist.”
Putcha, for one, says he personally welcomes the “clarity” on LDT regulation that might come by way of legislation or litigation in the U.S. “The reality, as usual, is that there is probably truth on both sides of this debate.”
Five concerns make his short list of deficiencies with the current framework, says Putcha, the first being the lack of “publicly accessible, independent, transparent, rigorous assessment of the safety and effectiveness of these [laboratory-developed] tests,” referring to both their analytical and clinical validity. Second, “there is not publicly available, accessible, transparent resource for systemically reporting and cataloguing performance issues and patient safety concerns for such tests.” The harms of such tests become known only if they are “obvious, egregious, or frequent.”
Third, Putcha continues, there is not “publicly accessible, transparent, rigorous assessment of the quality of the laboratory performing such tests,” and fourth, “the marketing claims for such tests are for all intents and purposes unregulated. And finally, there is no incentive—and I would argue a very strong disincentive—for innovators to pursue FDA authorization.”
Putcha says the situation is “somewhat analogous to allowing a drug manufacturer to market and sell a generic drug at the same time or even before the equivalent branded drug is launched.” He notes that “no one is arguing that this would be good for innovation in the drug industry, even when we as a diagnostics industry generally have far weaker reimbursement and IP [intellectual property] protection than drugs do.”
At the same time, “there are definitely reasonable and potentially significant concerns about the rule, including the FDA’s capacity for and timeliness of reviews, the time and cost of these reviews for test developers, and the resulting impact on patient access and public health,” Putcha says. “Frankly, I think both sides are really guessing at these impacts.”
Other unknowns are “how willing and able the FDA is to use all the tools in its arsenal to, at its discretion, mitigate these concerns,” he says. “The more general concern is if this is even possible to discuss in good faith in the middle of an election year given the current political climate, the function or dysfunction of Congress, and the significant self-interests that exist.”
The rationale for FDA regulation of LDTs was taken up by Ramamurthy, who worked for the FDA’s Center for Devices and Radiological Health between 2007 and 2013 and served as chair the LDT Oversight Steering Committee. Conversations about how to properly oversee LDTs were already underway when he arrived at the agency, he says.
Historically, it made sense for the FDA to have enforcement discretion on LDTs because they were laboratory tests quickly being put together by a community hospital or regional center to address an immediate need of some sort, Ramamurthy says, citing the Zika virus as an example. But LDTs today may serve much larger areas—the tests can often be ordered online and process mailed-in samples—making risks more difficult to manage.
“I think the FDA is trying to keep up with the times of how technology has evolved,” he says. “We all care about the same thing, which is the patients.”
It was long the case that Europe was viewed as the easier place to certify an IVD but with the passage of IVDR “the pendulum... swung” in favor of going through the regulatory rigors in the U.S. first, says Cimler, noting the FDA’s global influence on harmonization efforts. But, as Burde points out, it may be more a matter of different than easier to navigate the agency’s regulatory approach.
The FDA might be able to pick up some lessons from what has happened under the IVDR in Europe. One unique aspect of the European approach is the idea of “sampling” such that devices not in one of the high-risk classes get put in a group with similar devices that get collectively analyzed by a Notified Body and issued a certificate, says Burde. With legislation or rulemaking, the FDA might borrow the tactic to review LDTs in batches to certify the device-making competency of labs.
In the EU, ongoing review begins the year after the initial sampling exercise and might involve analyzing a different test from the same group. The tactic “reduces the [regulatory] burden somewhat,” he says.
Getting to this point took much longer than anticipated, Burde adds. The initial goal was to implement IVDR in five years and “we’re now seven years down the road and we have another three years to go before the final deadline.” He therefore urges regulators in the U.S. to bake some flexibility into their rulemaking in terms of their timelines for achieving change.
The current regulatory environment is overall a bit more challenging in the EU than the U.S., in Ramamurthy’s opinion. He points to the fact that the EU is comprised of 27 countries with an uneven level of subject matter expertise and, unlike the FDA’s new LDT enforcement policy, makes no exceptions for tests essential for the diagnosis of rare diseases. Unless the situation changes, “it is going to get very difficult to launch in Europe.”
With hundreds of thousands of LDTs now on the market in the U.S., it’s legitimate to ask if the FDA has the capacity to implement the new rule or can “reclassify and down-classify their way out of that,” says Putcha. Inadequate personnel was a matter highlighted by both the AMP and the America Clinical Laboratory Association in a parallel lawsuit it filed weeks after the final rule was published back in May.
The question about whether the rule can be implemented in four years is also worth exploring, he adds, given the unexpected delays seen with IVDR. “We’re guessing at implications, which of course are significant for industry and for patients most especially, so it’s hard not to pause just a little bit.”
It has been estimated that about 5,000 genetic tests alone come to market each year that the FDA would have to regulate, says Scrimenti. Especially troubling to AMP members are that the continued areas of enforcement discretion are insufficient to mitigate the harms, she adds, referencing the unmet need exemption that requires labs to be part of the same integrated healthcare system. “For rare disease patients, myself included, all of [our] specialists are in different integrated healthcare systems that use completely different labs.”
Also at risk is patient access to testing by public health laboratories that perform newborn screenings, she continues. Despite the “very narrow path” to continued enforcement discretion, if one lab puts an LDT through the FDA’s premarket approval (PMA) pathway for an unmet need, all other LDTs throughout the country addressing the same unmet need would then also have to go through the PMA process—at an estimated $4 million or more per test.
Yes, a clinical need would no longer be unmet by the FDA’s standard if an LDT gets approved by the agency, Scrimenti says, in answering a challenge from Ramamurthy. But it would put “many more burdens on labs at large... [and] significantly disrupt patient access to care.” Further, FDA authorization of one test does not mean everyone across the country has access to it.
Speaking from his old FDA training, Ramamurthy says, “there are too many tests ... [and] we don’t know how many of them are doing things right.” But there is room for negotiation between the various schools of thought, he adds, citing a suggestion years ago by FDA Commissioner Robert M. Califf. He proposed that the agency accept randomized real-world data from on-market LDTs along with a small amount of validation data.
LDTs are “already regulated under CLIA and there are several duplicative requirements in the final rule that labs are already doing every single day, and they are adhering to proficiency testing [requirements],” counters Scrimenti. Admittedly, CLIA has not been substantially updated since its enactment in 1988, “but it has successfully run for nearly 40 years.”
Pilots are already underway “to expand and enhance current regulations to reflect laboratory medicine as it is today... CLIA being the floor we can build on” without causing undue disruptions to the entire lab community, she adds.
Variability in a testing is a reality whether the oversight body is the CLIA, College of American Pathologists (CAP), or the FDA, says Cimler. Even for products approved through the agency’s PMA pathway, such as PSA tests for prostate cancer, individuals are routinely instructed to get initial and subsequent rechecks done in the same lab for this reason.
The problem boils down to a lack of standardization when it comes to evaluating lab performance, she says. Even CAP accreditation has not succeeded in leveling the playing field, noting how proficiency testing might inconsistently involve performing tests on three, 30, or 300 specimens.
CLIA does not require clinical validation of LDTs, says Ramamurthy. In fact, during his time at the FDA, he was handed files on proficiency testing to review by colleagues at CMS who lacked the subject matter training to do so on their own.
CLEP, the mandatory pathway to run LDTs in New York state, is harder in some respects than the FDA regulatory options, says Cimler. “But I don’t see anyone suing New York state. That’s somehow acceptable; why is the... [FDA’s] more generic process that we all know and understand not?”
Scrimenti’s response is that “the historical position has been to modernize CLIA.” As it is, all tests performed at labs are overseen by a board-certified pathologist, and the labs are subject to inspections. “We are very worried about the FDA’s bandwidth considering during one year of COVID under... emergency use authorization [EUA] flexibilities, they saw just under 3,700 COVID-19 tests” and the agency was completely overwhelmed. “Some of our members still haven’t received their EUAs for their COVID-19 submissions.”
Under the final rule, the FDA is expected to see a 140% increase in PMA workload and another 40% rise in de novo and 510(k) applications, she reports. “If they struggled with just under 3,700 COVID tests in one year with flexibilities, how are they going to handle the massive influx of LDT submissions?”
The big worry for Putcha is that two “parallel universes” currently exist. Under CLIA, an IVD will get “nowhere near the level of scrutiny” as a device with the exact same intended use and patient population going through FDA approval process, he says. Unlike the FDA, CLIA also does not speak to software that has changed a great deal since 1988, let alone advances in machine learning and artificial intelligence. Understandably, most labs don’t have the personnel and expertise in those areas.
The challenge here is figuring out a way to harmonize the two regulatory frameworks and distinguish when each channel is appropriate, Putcha says. “I think there is a role for LDTs, and I always will; the question is what is that role.”
While enforcement discretion can be a deterrent to innovate, FDA authorization can be an inducement by protecting and financially rewarding innovation, he continues. A key example would be Exact Sciences, whose market capitalization was greater than that of the top 25 LDT companies due to a single, FDA-approved, Medicare-reimbursable product (Cologuard). “That doesn’t exist when you can claim whatever you want and there is no oversight of those claims.”
Ramamurthy, who describes himself as “a bit of a regulatory geek,” says he is confident that the FDA is well equipped for phaseout of enforcement discretion with LDTs. After receiving 6,000 comments on the proposed rule from the community at large, the agency provided carveouts for rare diseases and LDTs already approved by CLEP, he notes.
The agency should be congratulated for its sensible and logical approach to meeting its public health mission and being an “honest broker of the dialogue,” says Ramamurthy. The next best move would be to start using real-world data on LDTs that have already entered the market and ensuring the data is transparent and traceable. In that way, labs wouldn’t need to generate data completely de novo on LDTs needing to go through the PMA or 510(k) process—and concerns about the FDA’s bandwidth might be eased. Regulating LDTs from scratch would “break the public health system in some sense.”
Cimler adds that she has “always found the FDA [to be] collaborative,” citing the agency work with New York State Department of Health to get Memorial Sloan Kettering’s tumor profiling test authorized. But uncertainty over test reimbursement can be unnerving—notably one instance when CMS was posturing that a companion diagnostic should have gone through the FDA process to be a reimbursable test. Labs panicked and started preparing and submitting their pre-submissions with the agency, only to backtrack when they learn that local coverage (whereby such tests were deemed reasonable and necessary by a Medicare Administrative Contractor) had become available, she says.
“Local Coverage [Determination] is de facto national coverage, so what do you need the FDA for?” Putcha rhetorically asks, returning to the issue of disincentives. With an FDA-approved product, reimbursement is closely tied to the label on a device. Under the Molecular Diagnostic Services Program (MolDX) of the CMS, however, labs can more easily get test coverage outside of the label. “How does that even make sense?”
As for the concerns voiced about FDA capacity and expertise to implement the new rule, he says, consider the alternatives. “What is CMS’ capacity and expertise if we were to throw this back to CLIA? What is CAP’s capacity and expertise if we throw this back to CAP? What is New York state’s?”
Harmonizing the regulatory frameworks “would be in the best interest of everybody,” including labs, patients, and investors, Putcha says. “But it will require some real honest and difficult conversations that to be fair we have been trying to have for the past three-plus decades, and we haven’t succeeded so far.”
The real-world situation for labs isn’t so rosy in the interim, according to Scrimenti. “It is difficult to reassure our members,” she says, given all the uncertainty. Many labs—including reference labs and labs at academic medical centers and in large urban settings—have already spent significant amounts of time determining which LDTs they are going to keep and let go and they are also cutting personnel.
Labs are also vacillating about whether to keep tests that were in place prior to May 6, 2024, which will need to be updated to account for new scientific information. That means “the final rule is having a significant impact on innovation, and it hasn’t even gotten to stage one,” Scrimenti says. Since labs don’t have the luxury of putting off critical business decisions, they’re making them now and “some of them are looking at whether or not they should remain open.”
LDTs marketed prior to May 6 are exempt from the final rule. However, she notes, the FDA has given itself the power to pull an LDT off the market, regardless of when it first made its entry.