It’s been about two months since FDA issued its Technical Assistance (TA), and the buzz around FDA’s draft legislation has not died down. In fact, FDA has continued to promote the proposal, including a speech by Commissioner Gottlieb’s Chief of Staff, Lauren Silvas, in late September at the Precision Medicine World Conference, in which she said the proposal came “from a thoughtful exercise where we carefully considered what works well for diagnostics.” After having received many questions as to what we thought of the draft, we thought it was about time that we weighed in again.
As we noted in our original blog post (here), a legislative approach holds promise. With some of the large labs and IVD manufacturers having bought into the concept of LDT/IVD reform, it is now more likely (more so than back in 2014 with the draft guidances) that we will see sweeping changes for the IVD regulatory framework. That said, we agree with ACLA and AdvaMed that additional, significant clarity is needed before this or any other legislation is passed.
While most commenters have focused on why are there no longer moderate risk tests and how will Pre-Cert work and a few larger topics, we wanted to highlight a few smaller issues that could also present significant confusion. This underscores a key concern over any legislation that is enacted: details matter. The nuts-and-bolts of regulation will have a critical impact on the real-world impact of any legislation. A few of these issues we observed in the TA are noted below, but these same concepts would apply to any future legislation.
The proposed TA provides little to no information as to how general purpose test platforms will be regulated. Test systems, like DNA genetic analyzers, liquid chromatography, and mass spectrometry, are the hardware on which most LDTs are run. These platforms are typically analyte/test agnostic and are, therefore, currently 510(k)-exempt. If and when these instruments are bundled with reagents for a specific clinical application they typically become subject to premarket clearance or approval requirements. Test platforms are included within the new definition of an in vitro clinical test, and there appears not to be any exemption from pre-market review for these systems. There are limited premarket exemptions for components and 510(k)-exempt devices, but these exemptions specifically exclude test platforms.
So what does this mean for instrument manufacturers under the draft TA? Must they obtain premarket approval for their instrument and would the standard for approval of a generic instrument be different from a full test? It is unclear. The TA implies that test platforms may require premarket approval. In one of the final sections of the TA, it states that test developers using test platforms that “was not cleared, authorized, or approved” by FDA may continue to do so for five years after enactment of the legislation. (Section 4(f)). Beginning five years after enactment, test developers must use test platforms that comply with the Act. Again, what it will mean to comply is unclear. If the intention is that the current exemption for test platforms would satisfy the “cleared, authorized, or approved” requirement, that should be explicitly stated.
Components, Parts, and Accessories
Like test platforms, many components, parts, and accessories are currently 510(k)- exempt because they, themselves, are not the test – they are merely the building blocks that a developer could use to build a test. The TA appears to acknowledge this fact and included an exemption for “components, parts, and accessories” that are subject to further development (Section 3, Section 587A(b)). The definition of an in vitro clinical test appears to specifically carve out certain components, parts, and accessories in Section 201(ss)(1)(F). In the pertinent section of the proposed regulations, however, the definition of “components, parts, and accessories,” points to Section 201(ss)(1)(E) (not F), which according to the draft TA, is limited to software only. It is possible that this is a typographical error, but it might also be a signal that FDA would like to take a more limited approach to the types of components, parts, and accessories that would be exempt from the premarket requirements.
Research Use Only
FDA has not been fond of the way some RUO products have been promoted and used. While FDA specifically excludes other non-clinical products, such as tests for law enforcement, from the proposed requirements, there is no similar exemption for RUO products. The TA’s labeling provision (Section 2, Section 587K(d)(4)) does include an exemption from the labeling requirements and performance standards (21 C.F.R. Part 861) only for RUO products. This section also notes that FDA should modify the applicable regulations as needed. It is unclear how or why the RUO regulations would need updating if FDA’s position regarding RUOs is not changing. The limited exemption and the statement that FDA will update the regulations leave open the question of whether and how FDA may modify its position regarding RUOs. RUOs are important elements of in vitro clinical tests and it could be detrimental to the industry if FDA were to begin over-regulating RUOs.
Grandfathering may be one of the most important concepts in the entire draft. The provision is somewhat awkwardly worded, however, and limited. An LDT is exempt, under the proposal, if (1) it was developed by a CLIA-certified high-complexity lab; (2) it is “for use only within that certified laboratory;” (3) it has not been cleared or approved; (4) it was offered more than 90 days before enactment of the legislation; and (5) it has not been modified within 90 days prior to enactment. (Section 2, Section 587A(c)(2)). Criteria one and two align with FDA’s definition of an LDT. The problem with these two criteria, however, is that FDA has acknowledged that many important LDTs may not meet these narrow, strict criteria (See FDA’s draft LDT Guidance at 5-6). For example, a company that developed a test in its small development lab and then moved to a commercial lab for validation prior to commercial launch would not meet FDA’s criteria (1) and (2). We have not, however, seen FDA object to such a test that was developed and owned by a single company as not being an LDT, but according to the draft statute such a test would not be grandfathered. Criterion 5 could also be particularly far-reaching, by seemingly freezing any changes. Given the nature of laboratory assays, that criterion may be difficult to meet, and also counter-productive if it allows no flexibility to labs to make any adjustments. What would happen to these tests is an important question which should be clarified prior to enactment.
While we view the draft TA as a creative approach, significant additional clarification is needed. Even if additional clarification is added to the legislation, even more detail will be needed in the form of regulations and guidance. While these are developed, there should be a significant phase-in time for companies to come into compliance. As worded, the draft TA would take effect immediately and encompasses tests that entered the market within 90 days prior to enactment. (Section 2, Section 587A(c)(2)(A)). This could potentially remove important tests from the market as they will have inadequate time to come into compliance with the new requirements – in fact, they may not even understand what those requirements are. It is important to recall that FDA’s draft LDT guidance included phase-in requirements that lasted years, similarly the Quality System Regulation had a two-year phase in to allow for companies to come into compliance. Any proposed LDT legislation should include ample time to allow labs and manufacturers to come into compliance.
What will ultimately happen to the TA or other IVD legislation is uncertain. Clearly, a new law will not be enacted this year. Given the competing perspectives on IVD regulation and the different stakeholders, the fate of future legislation is a large question mark. But what should not be a question mark is that if legislation does emerge, the details will be critically important.