Special Report: FDA’s proposed rule to bring LDTs into the fold

After bringing up the topic many times over the past two decades, the FDA has finally done it: They’ve issued a proposed rule that would eliminate the agency’s enforcement discretion over laboratory-developed tests (LDTs), integrating LDTs into the FDA’s oversight over a period of four years. This issue has been top of mind for the FDA and most everyone in the diagnostics ecosystem for many years. The change has the potential to be a very big deal. Whether for good or bad depends on your viewpoint.

But before we get into the details, let’s start with a few (simplified) definitions. (If you already know this stuff, go ahead and skip to the next section.)

The FDA categorizes diagnostic tests as medical devices and divides them into two types. As of now, those types are regulated very differently.

In-vitro diagnostics (IVDs)are tests that are approved or authorized by the FDA on a test-by-test basis.
The two major types of IVD approvals are:
1. PMA (Pre-Market Approval): For new tests that require clinical trials and submission of scientific and clinical data.
2. 510k (Pre-Market Notification): For tests that are substantially equivalent to tests that have already been granted a PMA. The notification submission requirements are less burdensome than for a PMA.
Laboratory Developed Tests (LDTs)are designed, manufactured, and used within a single clinical laboratory. These tests are not approved on a test-by-test basis, but rather, the lab that runs the test goes through the Center for Medicare and Medicaid Services’ Clinical Laboratory Improvement Amendments (CLIA) lab approval process. CLIA and its associated regulations focus on lab quality and staffing standards.

The regulations that currently govern these tests originate in the Medical Device Amendments of 1976. At that time, the FDA established that they would use “enforcement discretion” for LDTs – which means that while each new LDT requires documentation and validation, most of them do not go through complete FDA review. (A few companies have already chosen to generate the necessary data and submit their LDT for review as an IVD.) The specific rules to oversee the process were updated with the 1988 CLIA amendments.

Why is the FDA doing this?

The short version, in the agency’s words: “the risks associated with most modern LDTs are much greater than the risks that were associated with LDTs used decades ago.”

Over the past five decades, the FDA says it has seen “an explosion in the volume, complexity, and scope” of LDTs, and that “LDTs are increasingly relying on high-tech instrumentation and software” (think AI). They note high levels of variability in tests as well as lack of necessary safeguards, and they highlight specific examples (especially in cancer and prenatal testing) where inaccurate test results caused harm.

How will the potential new rule be implemented?

The FDA outlined a four-year, five-stage process. Key points: After the publication of the final phase-out policy, high-risk tests will have to be approved within 3.5 years (but not before Oct. 1, 2027); low- and medium-risk tests will have to be approved within 4 years (but not before April 1, 2028).

What does the FDA hope to achieve with the new rule?

The agency’s press release headline says it all: They want to ensure safety and effectiveness for LDTs. By ensuring that diagnostics are more accurate, the agency also hopes to engender more confidence in diagnostics among health care professionals as well as patients.

How will the FDA be able to keep up with the massive influx of tests needing approval?

The FDA currently utilizes a third-party review program that authorizes external organizations to conduct audits and other reviews for 510k submissions. The proposed rule notes that the agency intends to enhance and expand this program.

Commentary: This gets a small mention but is a critical part of the proposed rule. For a long time, the diagnostics industry has said that the FDA review process is too long and therefore discourages innovation. This problem was on display during the early stages of the AIDS epidemic, when our understanding of the virus was expanding quickly. Tests were taking so long to go through the FDA that they were irrelevant by the time approval was granted.

What are the costs and benefits of this plan?

The FDA’s calculations:

Costs for industry: $5.9 billion, including fees to the FDA and the cost of generating clinical and scientific data for new submissions
Costs for FDA: $530 million
Benefits: $31 billion, including “a reduction in health care costs associated with unsafe or ineffective tests. . . . and from therapeutic decisions based on the results of those tests,” as well as a reduced number of lawsuits and reduced costs to health care systems

Commentary: For benefits – we believe that there will also be a reduction in the number of repeat tests. Important to note – these estimates have large error bars / high variability (the FDA’s range for potential benefits goes from $3 billion to $86 billion, for example). For a small lab, the agency estimates that the new costs may be as much as 22% of their annual sales. For a large lab, the estimates range from 5 to 20% of annual sales.

Opportunity for public comment?

Yes, FDA will accept comments through December 1.

Are any tests exempt from this proposed rule?

A few:

Human leukocyte antigen (HLA) tests, which are used to match organ and tissue donors and recipients
Tests used exclusively for forensics (think police)
Tests used exclusively for public health surveillance purposes
“1976-Type LDTs”: Non-automated tests performed manually by laboratory personnel with specialized expertise

Overall Commentary:

The use of diagnostics requires trust and confidence. Physicians and patients need to believe that the test will be accurate. Virtually every other product in the physician-initiated clinical health care system goes through the FDA for approval. Diagnostics, however, have been different.

While the FDA’s position is itself not a surprise, there were some surprises in the proposal. Many expected that many tests (especially Class 1 or low-risk tests) would be grandfathered and not have to go through any FDA approval process. But the FDA estimates that “approximately 50% of IVDs offered as LDTs would not require premarket review.”

On the flip side, many expected that the transition time would be much shorter – perhaps as short as 18 months (consistent with the transition of COVID tests currently under EUAs to 510K approval). (More on this below.)

Challenges: This change will cause disruption on multiple levels. First, the scramble to get existing LDT tests approved / re-approved over the next few years. There are A LOT of tests (FDA estimates 80,000 from 1,200 labs). Are there enough samples for all of these trials? Are there enough statisticians to process all of these submissions? Is the agency really ready for the onslaught, even with those additional third-party reviewers at the ready?

And maybe even more importantly, will the FDA be able to keep up with new applications? We have no doubt that they want to do so – they want to be responsive, and they want to learn new technologies and evaluate them fast and fairly. But can they really do it? If they can’t, then this process squashes the ability of diagnostics companies and labs to adapt to new findings. (We would hesitate to say that it stops all innovation, but it certainly creates a barrier.)

In addition, we have to acknowledge that this change will likely cause the demise of many small labs. In some cases, that’s a good thing: There are labs out there that don’t have adequate quality-control systems and somehow got through a CLIA inspection or did not bother with the process at all. These labs hurt all of us, and they’re part of the reason why this new rule is being proposed. (How many are there? We wish we knew.) But mostly, forced lab closure is a bad thing. There are many good labs that have only a few tests, which they run efficiently. But they don’t have the people, resources, or expertise to put together an FDA filing.

For larger, established labs, we believe that transition over 3.5 to 4 years is reasonable. But we would be even happier if they would add a year or two to everything – 6 years for low-to-moderate risk and a round 5 years for high risk. For smaller labs, the FDA suggested that there might be an alternative where these labs would have up to 10 years to comply.

All that said, we believe that this will ultimately be positive both for the diagnostics industry and for patients. And before you start throwing virtual tomatoes – for better or worse, the FDA imprimatur is the gold standard in the industry today. The more tests that have it, the more reliable the industry looks as a whole.

In addition, we hope that if this new rule is implemented, payors will take note of the change and reimburse tests appropriately. We are not naïve – this will not be a magic wand that just broadly increases payments – but we believe that it has the potential to encourage payors to treat diagnostic tests like other parts of the healthcare system and pay for their value.

What is our perfect world? It’s one in which diagnostics has its own Center at the FDA, as devices, biologics, and drugs do now. We would call it CADER – Center for Advanced Diagnostics and Research. Is it likely to happen? No. But, as the FDA reiterated, diagnostics determine 70% of medical decisions – that is enough to justify that we stand on our own.

Thanks to Professor Michael Donovan, Biomedical Diagnostics lead at the College of Health Solutions at Arizona State University, for his feedback on our review of the FDA proposal. We note, however, that the commentary above is ours alone and does not represent any institution.

Authors:

MARA G. ASPINALL AND LIZ RUARK

OCT 5, 2023

reposted with permission