Steven H. Kroft, MD; Karen L. Kaul, MD, PhD
Am J Clin Pathol. 2023;160(3):222-224.
Hospital laboratories perform hundreds, even thousands of different tests to support the care of the patients they serve. These tests vary considerably, ranging from microscopic evaluation of biopsies and resected tissues by pathologists, with manual preparation and staining of these slides by trained histotechnologists, to high-volume analyses performed on automated instruments. In the microbiology lab, samples of blood, tissues, and body fluids are cultured on various media to determine whether a microbe is present—and, if so, what it is and how to treat it. Other sections of the lab detect and quantify specific cells, proteins, hormones, electrolytes, DNA, and countless other molecules present in the human body, all aimed at detecting, categorizing, and monitoring human disease. Some of these tests are manual and complicated, while others involve the nearly hands-off analysis of a sample on an automated instrument using a US Food and Drug Administration (FDA)–approved kit. Laboratory professionals carefully oversee the testing processes and interpretation of results; they are focused on the highest quality and safety for patients.
All testing procedures in clinical laboratories are carefully assessed and monitored under the oversight of the Clinical Laboratory Improvement Amendments (CLIA) of 1988. This legislation, published in the Federal Register, outlines procedures to ensure the safety and accuracy of clinical laboratory testing in the United States.[1,2] The CLIA is overseen by the Centers for Medicare & Medicaid Services (CMS) but primarily administered by a few “deemed-status” accrediting organizations, including the Joint Commission and the College of American Pathologists. These organizations provide an extensive and detailed framework for laboratory quality assurance. For example, the College of American Pathologists publishes 21 different specialty-specific checklists, each ranging up to 130 pages, outlining appropriate tasks to validate new tests and ensure ongoing quality of test results. Inspectors make unannounced visits to clinical laboratories every second year to comb through the records, procedures, and data the labs collect. Additionally, unknown samples for each analyte tested in the lab are distributed multiple times per year (proficiency testing); the lab analyzes these samples and returns the results so that they can be compared with results from other labs across the country. Results are published and returned to participants so that procedural corrections and adjustments can be made, if necessary. Repeated errors will require that the lab cease performance of that test, and these failures may be reported to CMS; labs that repeatedly fail proficiency testing can neither perform nor bill for such tests.
Regardless of the analyte being measured, under CLIA regulations laboratories carefully monitor the quality of the entire testing process, beginning from the collection of the appropriate sample from the patient, through the analytic steps, and concluding with proper reporting of the results. The FDA approval of a kit addresses only a portion of the testing process—specifically, the manufacturer’s claims regarding the performance of the kit itself. Many quality and safety issues occur during the pre- and postanalytic processes that are not addressed by FDA approval of a test kit.[2]
Laboratories can choose to use a non–FDA-approved method when no FDA-approved assay is available to meet a particular clinical need. Sometimes, the information needed is the result of a complex process that does not use a kit, such as microscopic evaluation of a tissue sample or culture of a microbe. Laboratory-developed procedures (LDPs), also referred to as laboratory-developed tests, have historically often preceded the development of a kit that is later submitted for FDA review.[3] These LDPs have had an important role in the establishment of the clinical utility of new analytes and have been particularly valuable in rapidly evolving areas such as genomics, ensuring that testing can be kept up-to-date with clinical best practices. Sometimes, laboratories need to validate the use of an FDA-approved kit for a nonapproved sample type to fill a clinical need, thus generating extensive validation data to demonstrate that the kit works reliably on an alternative sample type or age range. Obtaining FDA approval is a time-consuming and costly endeavor; therefore, approval is generally sought only for the highest-volume tests and most common sample types.
In recent years, the FDA has indicated an intention to exercise increased oversight of LDPs, using the rationale that existing CLIA oversight does not adequately ensure the quality and performance of LDPs. Well-publicized examples of poorly performing LDPs in the molecular space are cited as evidence that the public is insufficiently protected under the current regulatory framework. Although controversial, the FDA believes that it has statutory authority to provide such oversight under the Medical Devices Act of 1976 and that it has been exercising enforcement discretion to date. The most recent iteration of the movement toward FDA regulation of LDPs is embodied in the Verifying Accurate Leading-Edge IVCT Development (VALID) Act of 2021, legislation that the agency strongly supports. The VALID Act creates a unified, risk-based oversight structure for both in vitro diagnostics and LDPs, newly combined in a class termed “in vitro clinical tests (IVCTs).” This new structure would be distinct from the existing medical device oversight framework under which in vitro diagnostics were previously regulated. Under the VALID Act, IVCTs deemed to be high risk would require premarket approval. Notably, the VALID Act includes a new process referred to as “technology certification,” whereby a developer would submit a representative moderate-risk IVCT for approval. Once approved, other analytes using the same processes and procedures as the submitted representative analyte would not require FDA review. Notably, VALID contains a grandfathering provision, and low-risk tests would be exempt from premarket review moving forward. All IVCTs, however, including those that are low risk or grandfathered, would need to be registered with the FDA.
The VALID Act failed to pass in 2022 and appears to have lost momentum in Congress. In the absence of imminent legislation, the FDA has announced its intention to proceed with implementing LDP oversight through rulemaking, within the existing construct of medical device regulation. One challenge in planning any such implementation is the fact that no good data are available on the number of discrete LDPs being performed in laboratories across the country. It is of obvious interest, therefore, for both the FDA and the laboratory community to understand the scope of the task that the agency intends to take on. In this issue of the American Journal of Clinical Pathology, Rychert et al set out to do just that.[4] Specifically, these authors inventoried all the tests ordered in an academic health care system, categorized by whether the tests were performed with FDA-cleared kits, LDPs, or standard methods. The last category, which is infrequently discussed, consists of tests performed using universally applied standard protocols; one example of a standard method is manual cell counting of a body fluid. Although the goals of this study were simple, the results have potentially far-reaching implications.
The first striking result of this analysis was that the vast majority of laboratory tests ordered in an academic health system (93.9%) are performed using FDA-cleared kits. Only 3.9% of tests were accomplished using LDPs, with standard methods making up the remainder. Equally striking, though, was that 45% of the of the total assays performed in the lab were LDTs, required for that 3.9% of tests. Although the laboratory that provided the testing in the study of Rychert et al undoubtedly performs more LDPs in house than most other academic health system laboratories, the overall percentage of LDP-based tests ordered is likely similar in other academic centers, whether performed internally or referred to an external reference laboratory. Also notable in this study was that although molecular diagnostic assays are often viewed as prototypical of tests in need of increased regulatory oversight, noninfectious disease molecular tests (eg, assays for oncology and prenatal screening) represented just 8.8% of total LDP tests performed and 0.3% of total tests. Commonly ordered LDPs were heavily skewed toward assays performed using mass spectrometry, often for the monitoring or management of immunosuppressed patients. Also represented in the top 20 LDPs ordered are flow cytometry and chromosome analysis, widely performed and well-described tests that are essential to classification and management of hematolymphoid malignancies. These LDPs are critical for routine patient care.
Based on their data, Rychert et al[4] imply that it is perhaps overkill to create an entirely new regulatory infrastructure to regulate a tiny proportion of the tests actually ordered to take care of patients. It is also reasonable to assume that more LDPs are ordered in academic health care systems, which have specialized services and patient populations, than in the community at large. Further, because so many of the assays used to perform that small fraction of total lab tests are LDPs (45% of assays in the present study), the administrative effort to accomplish this oversight and the potential regulatory burden on laboratories (academic health system laboratories in particular) seems at first blush to be grossly disproportionate to the actual magnitude of the perceived problem. It may be misleading, however, to consider the proportion of test orders that are LDPs to be a surrogate for the scale of the issue. Specifically, Rychert et al do not take into account the relative impact of LDPs in diagnosis and management. In other words, the tests performed using LDPs may be (and likely are) disproportionately important in driving outcomes, potentially both positive and negative, depending on the quality and clinical validity of the LDPs being employed. This theoretical clinical impact metric is presumably directly related to the risk categories described in VALID or likely to be incorporated into FDA rulemaking outside of the VALID Act.
The debate about FDA regulation of LDPs revolves around access, innovation, and safety. If the cost of developing low-volume, highly complex assays to support cutting-edge clinical applications becomes too great (for example, because of excessive regulatory burdens), health system laboratories may choose to send more and more testing to external reference laboratories or even choose not to offer certain testing options to ordering clinicians, potentially delaying appropriate care or necessitating the implementation of therapies without complete information about the biology of the disease being managed. With respect to innovation, the argument can be made that those who are embedded in specialized care-giving teams are best equipped to identify unmet laboratory needs in providing optimized, precision care to patients, and that removing that decision-making to reference laboratories that function at a distance from the bedside might slow down critical advances in laboratory medicine. In contrast, the FDA and the proponents of VALID argue that the safety of the public is inadequately protected under existing CLIA oversight and that the agency has not only the authority but also a fundamental obligation under its charter to provide more stringent oversight of LDPs. The FDA further believes that this oversight can be accomplished without adding undue regulatory burden to clinical laboratories by focusing on the highest-risk tests. In regard to this point, it is important to emphasize that although VALID provides a general framework for oversight, the exact regulatory toll it would take on laboratories remains unclear, pending FDA rulemaking. The devil, as they say, is in the details. Opposition voices believe that the CLIA framework (as it currently exists or with modifications) is adequate to ensure that laboratory testing is of high quality. After all, it was CLIA sanctions that halted the clinical testing performed by Theranos in 2016.[5]
Finally, wherever one falls in the debate over proper oversight of LDPs, the authors of this commentary are confident that we all have the same goals: high-quality, safe, cost-effective laboratory care for our patients. If we all stay grounded in this premise, we can hopefully reach an end point that satisfies all stakeholders.
References
Am J Clin Pathol. 2023;160(3):222-224. © 2023 American Society for Clinical Pathology
Leave a Reply