FDA Oversight of Laboratory-Developed Tests: Where Are We Now?

On November 18, 2016, the US Food and Drug Administration (FDA) announced it would not issue a final guidance for laboratory-developed tests (LDTs).¹  The draft guidance for LDTs, released in 2014, had generated vigorous criticisms and praise, with stakeholders putting great effort into analyzing, influencing, and outlining the best way forward. There were congressional hearings, agency hearings, and many proposals from different perspectives.²  Some organizations supported greater FDA oversight of all LDTs, others challenged the FDA's legal authority, and many sought out a middle ground.

The FDA announced its legal authority over LDTs decades ago, categorizing LDTs as medical devices and clinical laboratories as manufacturers, but the agency maintained that it would “exercise enforcement discretion.” The 1997 Federal Register states, “clinical laboratories that develop such tests are acting as manufacturers of medical devices and are subject to FDA jurisdiction under the act.” The FDA went on to acknowledge that LDTs have “contributed to enhanced standards of medical care in many circumstances and that significant regulatory changes in this area could have negative effects on the public health.”¹³ This was a shot across the bow reminding laboratories that the agency could regulate but chose not to, at least at the time.

Even before the 2014 draft guidance for LDTs, many laboratorians saw the FDA as encroaching on the practice of laboratory medicine. The Washington Legal Foundation, a public interest group, filed citizen petitions with the FDA in 1992 and 2006 arguing that the FDA had no legal authority to regulate LDTs.³  After the 2014 draft guidance, the legal attack on the FDA's legal authority was joined by the American Clinical Laboratory Association and legal heavyweights Paul Clement and Laurence Tribe, who argued that the FDA's enabling statute does not grant the FDA the authority to regulate LDTs because, among other reasons, LDTs are part of the practice of medicine.⁴ 

Others argued a slightly different legal point focusing on procedural grounds. The FDA, they maintained, might have the authority to regulate LDTs, but it could not proceed through simply issuing guidance documents. Instead, the FDA was obligated under the Administrative Procedure Act to promulgate new rules through notice and comment rulemaking, and should withdraw the guidance documents and begin again.⁵ 

Manufacturers of in vitro devices (IVDs) supported FDA regulation of LDTs because they felt it would even the playing field and bring needed oversight. For example, a senior vice president at a large pharmaceutical company, after acknowledging the advantages of LDTs' ability to adapt to the rapid progress in molecular biology and technology, commented at an FDA public meeting: “The lack of controls and regulation of LDTs carries risk, and it also creates disadvantages for those manufacturers seeking IVDs.” In particular, the commentator was concerned that “LDTs aren't manufactured under GMPs [good manufacturing practices]. They have little clinical validation, but are widely used for clinical diagnosis. [This creates an uneven playing field.] The double standard penalizes both patients and clinicians who really don't understand that IVDs and LDTs are not equal.”⁶ 

Clinical laboratories performing LDTs opposed FDA regulation in an area that was traditionally and, they argued, adequately regulated by the Centers for Medicare & Medicaid Services under the Clinical Laboratory Improvement Amendments of 1988 (CLIA). Concerns centered around the increased (and unnecessary) administrative burden, cost, and impediment to access and innovation. Premarket approval is notoriously expensive, and even streamlined procedures like 510(k) clearance for a device, which requires “only” registration and notification to the FDA of equivalent devices, is a significant administrative burden. Lengthening and increasing the cost of entry into new testing, such as next-generation sequencing (NGS), would decrease innovation, lessen access, increase test cost, and dampen the innate nimbleness of LDTs.

The College of American Pathologists and other medical societies sought to negotiate a middle road. Relying on the FDA's existing 3-tier device framework, these nongovernmental pathologist groups acknowledged that some FDA oversight was appropriate for high-risk LDTs, but maintained that most LDTs could be adequately regulated through CLIA with some modifications.⁷ 

The FDA's current view, revised and certainly more informed and nuanced than the 2014 draft guidance, can be distilled from its January 2017 discussion paper issued on Friday, January 13, as a synthesis of received comments. The FDA's stated intention for this discussion paper was to encourage public discussion, allow for a possible legislative solution, and “attempt to balance patient protection with continued access and innovation.”²  Pursuing its public health mission, the FDA laid out its concern: “Inaccurate or false test results, or accurate measurements with an invalid claim regarding the test results' relationship to a disease, can lead to substantial patient harms.” Although the FDA's views in this 2017 discussion paper have been tempered compared with its 2015 parade of horribles presented in “The Public Health Evidence for FDA Oversight of Laboratory Developed Tests: 20 Case Studies,”⁸  the FDA remains concerned that “different test developers are treated differently,” because this can “discourage innovation by making it difficult for high quality test developers to compete with poorer performing counterparts.”

Indeed, the FDA's current approach to laboratory testing is a confusing patchwork, likely the result of historical evolution. Distinguishing between LDTs, IVDs, and direct-to-consumer tests (DTCs; like 23andMe) is difficult. It considers who orders the test (physician versus consumer) and who interprets the test (physician versus nonexperts testing IVDs), subjecting kits to premarket review but allowing enforcement discretion for LDTs. This is incongruent with the typical device-focused, risk-based approach the FDA uses in regulating medical devices. Many pathologists will respond by arguing that LDTs are not medical devices (but perhaps DTCs or IVDs are), so we return to the beginning question about whether LDTs fit under the statute's definition of device and whether LDTs fall within the FDA's enabling statute.

Following the 2017 discussion paper, the FDA regulation of LDTs may be on hold, but it is unlikely that the issue is over. Many patient groups, oncologists, and other physicians, consumer groups, pharmaceutical companies, IVD manufacturers, and insurers supported and continue to support the FDA's risk-based approach to LDT oversight. Given this support and the political climate that may preclude any imminent direct regulation of LDTs, the FDA will likely pursue other means. For example, in September 2016 the FDA issued a safety communication to patients and physicians that warned against using ovarian cancer screening tests.⁹  Another tactic the FDA may use is to narrow what it views as a legitimate LDT while expanding what it considers to be DTC testing, perhaps requiring more intensive hands-on physician involvement for a test to qualify as an LDT. Interestingly, LDTs are not specifically defined by statute.

The procedural details of how the FDA will proceed remain murky, but its discussion paper indicates that it will move forward and that it does not consider the burden of additional regulation to be too great for clinical laboratories. The discussion paper implies a 510(k) clearance system is not overly burdensome for clinical laboratories and outlines that:

Related to the FDA's emphasis on “well-curated databases,” the FDA released 2 new draft guidance documents in the summer of 2016 that address NGS in the hopes of creating “a flexible and adaptive regulatory approach to the oversight of next generation sequencing (NGS)-based tests as part of the Precision Medicine Initiative (PMI).”^10,11  Both documents address standardization for NGS. One of the draft guidance documents proposes FDA recognition for public genetic variant databases to support clinical validity claims of NGS-based diagnostics and allow public databases to provide “valid scientific evidence to support the clinical validity of genotype-phenotype relationships in FDA's regulatory review of NGS-based tests.” This draft guidance acknowledges the importance of aggregating gene variant data and laments that many of the data are not public and available. The draft guidance emphasizes and encourages standardization and public access to data, and it signals the FDA's intention to regulate NGS/LDTs. Hopefully, this represents a step toward reducing the problem of so much clinical and genomic information being siloed by institution, disease, researcher, or enterprise. Technology like cloud computing can ease data sharing,¹²  but current clinical and research practices hinder the sharing of data, and significant legal and regulatory barriers continue to exist. To address these, the FDA should continue to work with other government agencies and industry/clinical/academic stakeholders to develop a coordinated approach.

In summary, the FDA's view has evolved since the initial draft guidance. Its repeatedly stated intention to “balance patient protection with continued access and innovation” is commendable. Let all of us continue our efforts to ensure that regulators understand the importance of LDTs in ensuring access and innovation, and let us all urge a coordinated approach.