LDTs and the FDA: The saga continues

In recent months, the U.S. Food and Drug Administration (FDA) has communicated more frequently with the laboratory community than it had in the previous 25 years. Specifically, we have seen a modification of prior approaches to federal regulation of laboratory-developed tests (LDTs). The changes may indicate a better understanding on the part of the agency of the way in-house tests are developed and qualified under the current regulations in 42 CFR.493

Background and issues

Both FDA and the Centers for Medicare and Medicaid Services (CMS) seem to have authority to ensure performance of LDTs. The first Citizen’s Petition (CP) against FDA authority was submitted in 1992 after the publication of an RUO/IUO Guidance that asserted FDA authority indirectly. This CP requested FDA to state unambiguously that FDA did not regulate LDTs. Two other petitions did the same in later years. The following reasons were given, and they represent most of the legal arguments for no FDA involvement in LDTs:

a.    Labs were already regulated by CLIA. If there is a problem, CMS (then HCFA) should take action, not FDA.
b.    CLIA gives CMS authority to require performance specifications for LDTs.
c.    Labs are not manufacturers, but service providers.
d.    No article was shipped in interstate commerce, a requirement for FDA’s authority.
e.    A change like this would be costly and detrimental to public health, by limiting availability of needed tests.
f.    Such a change would be substantial and require rulemaking, not simply guidance.
g.    HHS is the Agency that is responsible for both CMS and FDA, and FDA cannot assume authority over LDTs unless HHS allows it, which they have not.
h.    FDA has yet to provide scientific evidence of a public health concern that leads them to insist on changing the current system for ensuring test quality.¹

Answers to these arguments have come from FDA and from a number of interested parties. Genentech published a petition in 2008 asking FDA to stop “enforcement discretion,” develop a regulation, and enforce it for LDTs. The position agreeing with FDA would counter the arguments above:

a.    Two agencies can regulate the same entity, as long as their requirements do not contradict, so that one agency is requiring what the other agency is forbidding.
b.    The differences between what FDA will require and what CMS requires is substantial. While CMS has the requirement to demand performance specifications for tests, its requirements are not transparent to the end user of the data. Also, they are not applied until after marketing has commenced. FDA provides independent and public pre-market review of new IVDs, requires publication of performance characteristics, and regulates the design and development of IVDs.
c.    LDTs fit the FDA definition of IVD, and the laboratory using reagents and equipment to develop a test is doing what is described in the design control segment of the Quality Standard Regulation (QSR). Subsequent test assembly under a fixed protocol could be described as manufacture.
d.    An activity can be interstate commerce if it involves a substantial interest involving interstate commerce. Both the first and the 9th Federal District have found interstate commerce to exist when the ingredients of a product are shipped interstate.² Additionally, sale of medical devices carries the presumption of interstate commerce.
e.    The cost must be balanced against the risk involved in using tests which have not been fully characterized in terms of performance. FDA announced a joint Task Force with CMS which will consider QA and Inspection systems.
f.    Since IVDs are the subject of considerable regulation already and LDTs are IVDs, FDA does not believe that regulation beyond the existing is necessary.
g.    HHS has been silent on its role in the controversy.
h.    Finally, FDA has rules for pre-market review, reports post market, design and development and modification of an IVD. These rules are all applied to manufacturers, and not (at present) to labs offering LDTs. This means there are at times (e.g. HER-2) two very similar products are subject to two very different requirements. In fairness, the same products should be subject to the same regulation.³

During this period, FDA began to send untitled letters to labs offering various LDTs to inform them that their products were neither approved nor cleared and asking them to respond. This continued into 2016, and while there is no data on lab responses, it appears to have been minimal.

FDA’s two attempts to develop guidance

The IVD Multivariate Index Assays (IVDMIA) draft guidance was published in 2006. These assays, corresponding roughly to tests in the High Complexity category, must be subject to FDA review pre-market. The reaction of the lab community was negative, and instead of taking their concerns to FDA, they went to Congress. Congress responded by including LDTs in the FDA Safety and Innovation Act (FDASIA) in 2012. They required FDA to discard the IVDMIA guidance, and to start over with a new approach.

FDA published two drafts in late 2014, one addressing the general regulatory approach and the other addressing reporting and notification/registration. The framework guidance re-defined LDT, and continued to require either pre-market clearance or approval in high and very high risk categories. The draft developed the concept that the test developer was also responsible for demonstrating the clinical validity of the test—that is, the effect of test result on the health of patients. The draft received more than 1,400 comments, and FDA continued to claim that it would produce a final guidance last year. FDA continued to assure the lab community that a final draft was in progress, but in November, FDA stated that it was not going to finalize the guidance. The reasons given were the irreconcilable differences among many comments, such that it was impossible to reach consensus. With such disagreement, the preferred route would be through Congress.⁴

FDA’s discussion paper

On January 13, 2017, FDA signaled that, although its second draft guidance was as dead as the IVDMIA draft, the agency still had something to say about LDTs. The discussion paper repeats some of the discarded draft guidance⁵ but adds some new possibilities:

a.    Grandfathering: LDTs currently being offered would not be required to do pre-submissions. They would be expected to maintain complaint files and comply with post marketing reporting. FDA would not demand public availability of performance data for these products, but would “suggest.” (This change alone would reduce FDA workload significantly.)
b.    Explicit description of performance data—it can include both bench and clinical data.
c.    A discussion of difference between current LDT QA and FDA QSR. CMS/FDA QA joint taskforce was referenced. Differences found in design control and vendor qualification, many similarities. Possibility of non-FDA inspection to continue was noted.
d.    Novel approach to significant changes in product post-marketing. FDA only reviews the protocol for validating changes.
e.    Third party review of LDT 510(k)s should be considered.
f.    Public availability of performance data required.
g.    Comparison between FDA Clinical Validity and CMS Clinical Utility: intended use of the data: CMS Economic Effect, FDA Health Care effect. (A distinction without a difference?)

So, what next?

FDA, like any agency in our new administration, will not be encouraged to issue new regulations or guidances unless clearly required by Congress. The anticipation is that FDA will not continue to issue untitled letters to laboratories offering LDTs unless there appears to be a real public health issue.

Additionally, the 20th Century Cures Act⁶ urges the Agency to be very serious about facilitating new medical device development, especially in oncology. The new law gives the Office of Device Evaluation (ODE )and Office of In Vitro Diagnostics and Radiological Health many tasks that will require development of guidance or regulation in the coming months. It is unlikely that FDA will continue to pursue laboratories, but it is possible that CMS and FDA would agree to cooperate in certain laboratory QA activities.

NOTES

1. The issue of the necessity for change was also explored. In 2015, FDA published 20 cases which illustrated its concern with LDTs. The report was descriptive only, and was not scientific proof or even a rationale for FDA concern. Some of the cases were not easily related to a testing error. No one considered these to be convincing evidence that the current CMS system needed revision. (But a number of the respondents to the 2014 draft guidance indicated that there were elements of FDA oversight hat would be an improvement for labs.)
2. Baker V US, 932F2d 813, 814-5 (9th Cir., 1991) and US vs Article of Food…Coco Rico, Inc., 752 F2d 11, 14 (1st Cir., 1985). Cited in the first petition response, 1998.
3. For example, the 2008 Genentech CP pointed out that they had developed and tested Herceptin using companion diagnostics for HER2 in tumor tissue. These tests had been subject to extensive FDA scrutiny , and had published test performance. However, laboratories were offering LDTs for HER2 without such extensive scrutiny pre marketing, and Genentech had no assurance that these tests would in fact indicate use of Herceptin in treatment.
4. The absence of FDA representatives at a September 2016 hearing of the Senate Committee on Health, Education Labor and Pension “Laboratory Testing in the ERA of Precision Medicine” may have been a signal of this withdrawal.
5. For example, risk-based test categorization, and exemptions for some LDTs.
6. Twentieth Century Cures Act, Title III, Subtitle F, Section 3051ff.

Anna Longwell, JD, MBA, MS, is principal of the Palo Alto law firm, Longwell and Associates, which specializes in Food and Drug law. The firm has expertise in U.S. FDA expectations, regulation and law, affecting the development and ultimate marketing of new medical products, drugs, devices and biologics.  Prior to establishing the firm, Ms. Longwell served as VP of Regulatory Affairs for Becton Dickinson Medical. She has also been a visiting lecturer in food and drug law at the University of Santa Clara School of Law; a visiting lecturer in food law at the Institute of Agribusiness, University of Santa Clara School of Business; a visiting lecturer in regulatory topics at the Haas School of Business, UC Berkeley, and the UC Santa Cruz Extension; and an instructor for the Food and Drug Law Institute (FDLI) internship program at Catholic University, Washington DC.