Tools, Technologies and Training for Healthcare Laboratories

FDA on LDT: GED, out!

The FDA has a new Proposed Rule to assume regulation of Laboratory-Developed-Tests (LDTs). How scary is that?

FDA on LDTs: GED Out!

A close reading of the proposal in the Federal Register

Sten Westgard,MS
October 2023

A new assault on Laboratory-Developed-Tests (LDTs) began on October 3rd with the FDA asserting its authority to regulate these methods in the Federal Register Docket No. FDA-2023-N-2177 RIN 0910-A185 Proposed Rule.

For those of you who are faithful readers of Westgard Web, you can recall the times that we have been highly critical of FDA, CLIA, CMS, over multiple concerns. You might therefore be surprised that, upon review of the FDA proposal, we believe they have a point.

It comes down to “General Enforcement Discretion” (GED)

From the very start, FDA has had the authority to regulate LDTs. It just hasn’t done so. This is because of “General Enforcement Discretion.” It decided not to regulate them. But now it has decided it will.

The Proposed Rule then details all of the statutes that grant the FDA authority to regulate. There are a lot of them. It's hard to deny that the FDA is supposed to regulate medical devices.

Why were labs getting GED in the past?

“At the time of the passage of the MDA [Medical Device Act], LTDs were mostly manufactured in small volumes by laboratories that served their local communities….They also tended to employ manual techniques (and did not use automation), performed by laboratory personnel with specialized expertise; to be used and interpreted by physicians or pathologists in a single institution responsible for the patient (and who were actively involved in patient care); and to be manufactured using components legally marketed for clinical use….”

In other words, in the past LDTs were a small operation impacting small numbers of patients. So the FDA gave them a pass.

What’s changed with LDTs to warrant the end of GED?

“Many LDTs are manufactured by laboratory corporations that market the tests nationwide, as they accept specimens from patients across the country and run their LDTs in very large volumes in a single laboratory.”

There are more LDTs now and they are being used by more and more people on a much larger scale. Why is that bad?

“The risks associated with most modern LDTs are therefore much greater today than they were at the time FDA began implementing the MDA, and most LDTs today are similar to other IVDs that have not been under FDA’s general enforcement discretion.”

More LDTs lead to more risk, that’s not difficult to admit, is it? But when the LDT becomes a way to game the traditional FDA device approval pathway, that’s an extra negative:

“In fact, today, the testing industry has come to view FDA’s general enforcement discretion approach as an alternative pathway to market for test systems, such that test systems are often “launched as LDTs” with no assurance that they meet requirements under the FD&C act and its implementation regulations.”

Indeed, the FDA notes it has seen manufacturers who fail their normal premarket submissions go on to “reclassify” the methods as LDTs and take them to market.

What could go wrong with that? A loophole to exploit and no consequences to those who exploit it.

The Proposed Rule then provides a litany of LDTs that have gone wrong. Misdiagnoses, adverse events, forged results, inaccurate results, and more. And then there’s the case of Theranos, our personal favorite, which weaponized the LDT loophole to the tune of a $9 billion valuation, and subsequently proved to be a complete fraud.

In summary:
1. The FDA has always had authority to regulate LDTs
2. The current lack of regulation on LDTs has harmed the public

Legally and probably ethically, FDA has a duty to end the GED and force the laboratory community to make their LDTs avilable for premarket submission and review.

It’s all over but the shouting.

The Outcry against the FDA

There’s one statement in the Proposed Rule that stands out as purposely naive:

“FDA does not anticipate that ending the general enforcement discretion approach for LDTs would unduly impair test innovation and patient access.”

All the objections we have seen center around this issue. If laboratories are now device manufacturers, how much harder, more expensive, more delayed will it be to bring an LDT to market? Won’t innovation suffer, as some laboratories abandon their LDTs rather than go through a presumably, predictably bureaucratic labyrinth to get approval?

Without question, some labs will stop running LDTs. They won’t have the resources, the time, the money to do it. And other labs will not even start new LDTs because of those obstacles.

The real calculus is, how many patients can we accept being harmed by continuing the process the way it is? For every new LDT, is it acceptable to allow harm to X patients? What is the permissible collateral damage of our current LDT regulations? If “innovation” means allowing patients to get the wrong diagnosis or the wrong treatment, maybe we’re thinking about it the wrong way.

This brings up the failure of the current regulations. CMS or CLIA could have imposed stricter regulations or heightened inspections or stronger enforcement of LDTs. They didn’t. Laboratories could have imposed stricter guidelines on themselves, through their membership organizations (CAP, CLMA, ADLM). They didn’t. By refusing to address the growing problem, they opened the door to the FDA.

Arguments that advocate for the current status quo – no FDA regulation, and no changes to CLIA regulation, just let the market keep doing what it’s doing – is untenable.

Without doubt, it will be more expensive and cumbersome to bring LDTs to market. If that protects the public from ineffective tests, that’s the “Do No Harm” option we have to accept.

The FDA has the law and ethics on its side. But the practical costs and the political mood will have its say. The current climate in Washington DC is not a strongly pro-regulation one. In fact, the very foundation of the administrative authority of federal agencies is coming before the Supreme Court. That may be the forces that prevent the FDA from taking over LDT regulation.

The implementation: A phased exit from GED

If the FDA gains control over LDTs, the shock won’t happen overnight. It will take a number of years, spreading out the pain to ease laboratories into the premarket submission world.

Year 1 after final FDA rule is in place
Adverse Event Reporting. It’s hard to believe this doesn’t already exist, but it doesn’t. If an LDT contributes to a patient death, labs will have 5 to 30 days to report that to the FDA. If a lab modifies and LDT to reduce a health risk, or if the test is removed from the market for the same reason, the lab has 10 days to notify the FDA. This reporting will help the FDA get a better handle on the state of the LDT market.

Year 2 after final FDA rule is in place
Laboratories that use LDTs will have to register with FDA as a medical device manufacturer. The 2024 registration fee is $7653 and this will require an annual renewal. A lot of information will be required for this registration. Laboratories will also have to properly label their LDTs (there is a lot of guidance on label requirements).

Year 3 after final FDA rule is in place
Laboratories will have to comply with current good manufacturing practices (cGMP). This is not CLIA, but there is some overlap. Labs with CAP accreditation or New York’s accreditation are closer still to compliance with cGMP. Labs must define their LDT design, such as patient needs, clinician needs, and laboratory needs. All of those decisions must be documented, and further still, labs must explain how the LDT satisfies all of those needs. Labs must also put into place Design Controls, which include verifying and validating the LDT. Most laboratories already have these protocols in place, but there will be additional demands to record who conducted the validation, when was it done, etc., and all of this must end up in a designated design history file (DHF). Further still, labs must establish Purchasing Controls, so that they can prove they have evaluated their suppliers, and that those suppliers meet quality requirements established by the laboratory.

This is a big year for documentation, as you can see, and it doesn’t end yet. Laboratories must establish Acceptance Criteria for all parts of the LDT. Like a regular diagnostic manufacturer, they will need reagent lot release criteria, and reagent lot-to-lot variation acceptability criteria, etc. Some of this already exists under CLIA, such as the requirements for monitoring with QC, but there is a new formality to the process under FDA. Everything must be documented, documented, documented.

The fun just gets funner in year 3, because the time for submissions begin. High-risk LDTs will be required to submit first. This is not scheduled to happen before October 1st, 2027.

Year 4 after the Final Rule is in place. The rest of the methods catch up with High-risk IVDs. Moderate and Low-risk LDTs (that require premarket submission) now have to submit. This is not scheduled to happen before April 1st, 2028. (Is that date chosen on purpose?)

There’s a lot more in the Proposed Rule, for example what types of tests might be excluded from regulation, even what kind of laboratories (“academic medical centers”?) might get special exclusion from the regulation. We’ll dig into it more in future articles. This is an evolving story. Expect that the Final Rule will be significantly modified from this initial Proposed Rule.

The FDA is holding a free webinar describing this Proposed Rule on October 31st, 2023 (again, date chosen for a reason?) and the public can submit comments on this rule. You can also make public comments in support or in opposition of the Proposed Rule here.