There’s nothing quite like a high-profile legal case to make the powers that be question whether regulation is tight enough. For Europe’s medical device sector, the trigger for the recently applied MDR and IVDR frameworks was the PIP (Poly Implant Prothese) breast implant scandal; in the US, where regulation has traditionally been stricter and changes are few and far between, a case of criminal fraud known as the ‘Theranos scandal’ was a catalyst for the FDA to move past simply questioning whether it should regulate laboratory testing and start to actually draft the legislation that would allow it to do so.
Theranos was a health technology company that’s value soared to $9bn in the 2010s after investors poured money in to back the promise of a pinprick blood test, which could detect more than 200 biomarkers, including those indicative of diabetes and cancer. In 2015, two whistle-blowers at Theranos exposed the company for the inaccuracy of its testing, which was later found to have misdiagnosed patients with a range of diseases. By 2018, Theranos was no more and the FDA introduced the first iteration of the VALID Act (Verifying Accurate Leading-edge IVCT Development Act), which would force laboratories to meet the same strict validation requirements as IVD manufacturers before they could provide them to doctors and patients.
Despite the medical world’s indignation over Theranos, the legislation failed to gain the required support until its second iteration in 2020, which – although it enjoyed bipartisan support – was shelved due to the Covid-19 pandemic. Introduced again in its third form in 2021, the Act got a boost from an October 2020 paper in The New England Journal of Medicine, in which Jeffrey Shuren and Timothy Stenzel of the FDA looked at 125 Emergency Use Authorization (EUA) requests from laboratories and found that 82 had design or validation problems, with several being denied authorisation. Although it was omitted from the recently passed Medical Device User Fee Agreement Bill (MDUFA V), many believe the VALID Act could hitch a ride on another legislative vehicle before the year is up.
Find a path to validation
Shannon Bennett has witnessed all these events during his 23 years working in clinical testing laboratories run by Mayo Clinic, before moving into his role as director of regulatory affairs for the company’s department of laboratory medicine and pathology. “About five years ago now, my role shifted more firmly into the regulatory affairs realm to coincide with VALID,” he says. “The position I’m in now was essentially created to help Mayo Clinic navigate FDA regulation of laboratory tests.”
“It’s not an even playing field, IVDs and lab-developed tests (LDTs) are not regulated in the same way. But, in lab, we would say that’s because the risk profile is very different.”
At the heart of the debate is the fact that in-vitro diagnostics (IVDs) and lab-developed tests (LDTs) broadly speaking perform the same duty of testing for analytes or biomarkers that can give an indication of health and can help diagnose diseases. In fact, all that’s required for an LDT to lose its regulatory discretion exemption and be reclassified as an IVD is for a single component in a test to be developed outside of the development laboratory. Given how easy it is to blur the lines between the two, it’s unsurprising that many of those within the IVD manufacturing space see the VALID Act as “a long-overdue modernisation of the law for all diagnostic tests”. Those were the words of Scott Whitaker, president and CEO of the Advanced Medical Technology Association (AdvaMed) – a major advocacy group for medical device and IVD manufacturers – when the act was reintroduced to the House and Senate in June 2021. Currently, laboratories are regulated by the Centers for Medicare and Medicaid Services (CMS) under the Clinical Laboratory Improvement Amendments (CLIA), and undergo an inspection every two years to maintain their CLIA certification, but don’t have to provide evidence of validity for their tests to any higher power.
Bennett acknowledges that LDTs have a privileged position as far as the diagnostic market goes, but argues there’s an important distinction to make that justifies it. “It’s not an even playing field,” he says. “IVDs and lab-developed tests (LDTs) are not regulated in the same way. But, in lab, we would say that’s because the risk profile is very different.” Indeed, in LDTs, the conditions for each lab remain the same; for an IVD manufacturer, devices must be developed to sell to hundreds of laboratories all over the world, with different climates, staff and equipment, all of which introduce variables that can add risk to the process of testing. “Running a test in humid Florida is different to very dry Minnesota in the wintertime,” Bennett explains as an example.
That’s only the technical argument for why LDTs shouldn’t be overseen with the same rigour as IVDs too. Bennett believes the application of LDTs creates another key distinction between them and IVDs. “What drives our test development is the doctor across the street calling me up and saying, ‘I have this biomarker that I need to test for to treat my cancer patients’, and then we’ll scramble to develop that test and validate it,” he says. “At Mayo Clinic, we have a reputation for developing a lot of highly esoteric tests, so a lot of rare disease testing. If we don’t develop a given test, no one probably will. It’s not cost-efficient for a large IVD manufacturer to make a test kit that they’ll only sell a couple hundred of.”
Manufacturing complications
When it comes to the practical aspects of the FDA regulating LDTs, individual laboratories would have to seek approval for each test before they can run it, providing evidence for its efficacy, in essence forcing them to jump through the same hoops that IVD companies do. “Many laboratories today in the US don’t submit their test information to anybody,” says Bennett. “Going from that environment to the rigour of an FDA submission is a huge jump. If VALID passes, or it doesn’t and FDA starts regulating anyway, clinical laboratories are going to need a lot of education and, frankly, handholding.” There’s also the question of capability, as the theoretical questions posed by the VALID Act, are do labs need to have their tests validated externally and is the FDA the regulatory body for the job? Bennett is doubtful, at least of the second question. “They have this list of bad LDTs, but when you push them on it and ask for examples, they have a really hard time articulating what those bad tests are,” he says. “We certainly see FDA cleared or approved test kits that have problems, so clearly FDA oversight isn’t a panacea to fix all lab-testing issues.”
One problem Bennett sees arising when it comes to the FDA creating guidance for labs is the way the agency determines a validated test. “A big point of contention is what constitutes clinical validity,” he says. “I would read a published journal article from a well-respected journal saying this analyte is associated with this disease and there’s my clinical validity. The FDA has been very clear that a single journal article is not sufficient.” Part of the difference in opinion comes down to the role that test plays. For the FDA, an inaccurate test creates patient risk, but Bennett argues the agency has a tendency to interpret test results independently of the care process and that’s very different to the supporting role that laboratory diagnostics play in patient care. “Test results are not interpreted in a vacuum,” he says. “There’s a physician who sees how the patient is presenting and orders additional supportive tests, so it’s unlikely that one test result is going to drive the entirety of the patient’s treatment.” Bennett and his peers have argued this point strongly throughout the negotiations, but until the FDA drafts the relevant guidance documents, the lab-testing sector won’t know what validity requirements will look like.
“There’s a physician who sees how the patient is presenting and orders […] supportive tests, so it’s unlikely that one test result is going to drive the entirety of the patient’s treatment. ”
Time and money
Beyond the issue of validity, there’s also the time frame and the cost of approval to consider, and Bennett says elevating both of those significantly could ultimately harm patients. “I’ve heard colleagues, even from large academic medical centres, say they develop 100 new tests a year and maybe they’ll only be able to do 20, and those other 80 they’ll either send out to a large commercial laboratory or they just won’t offer that test,” he says. “If the test can go to a local laboratory and you can get a result the next day, that’s really good for the patient. If you have to send it out to a reference laboratory and it takes a week to get that result, that could have some significant patient impact, especially when you think about cancer treatments.” Bennett adds that the FDA has made some concessions to labs in this area, with low-volume tests expected to be exempt from regulatory oversight, but he says the agency’s definition of low volume will have an impact on how many tests certain labs develop, especially in the area of rare diseases. “Right now it’s set at five,” he says. “Everyone agrees five is too low. If that number is 100 or 500 for instance, then a lot of those low incidence tests will be exempt from submission and that will help a lot.”
Although critical aspects of the regulatory process that would result from the VALID Act have yet to be decided, one of which would be a laboratory diagnostics user fee agreement – LDUFA perhaps – Bennett says although negotiations with the FDA and other stakeholders have led to several changes in favour of laboratories, he still has concerns about the transition. Firstly, he wonders whether the five-year transition period specified in the current legislative document – three of which he believes it will take for the FDA to write the necessary guidance documents – might not be long enough. “That only leaves two years for labs to move into compliance, and going from being a CLIA-regulated laboratory to being an FDA-regulated laboratory is a huge, huge jump. I’m a little worried that’s not going to be sufficient time.” His second concern is with the FDA’s ability to set up the comprehensive test information system (CTIS) required to drive submissions and adverse event reporting, especially when it could be flooded with submissions right away due to the rules on grandfathered tests laid out in the VALID Act.
“The FDA is not an IT company,” he quips. “I’m a little bit nervous about their ability to get CTIS up and running. It’s estimated there’s something like 100,000 LDTs on the market right now. Granted most of them would be grandfathered, but if even a percentage of them start having test modifications that would need to be submitted, I’m afraid that FDA is going to get slammed and not be able to handle it.” Citing the volume of EUA submissions the FDA had to get through over the past few years of the pandemic, Bennett foresees such a situation occurring again but much worse.
“For Covid EUAs they got a couple of thousand and it brought the agency to their knees,” he says. “Now they’re going to get tens of thousands of submissions in fairly short order. It’s going to be rough for them.”
