Lab-Developed Tests vs In Vitro Diagnostics – Balancing patient safety, access, innovation, and commerce

By Thomas Maeder
No Comments

On July 19 and 20, 2010, the Food and Drug Administration held a public meeting on regulatory oversight of laboratory-developed tests (LDTs). This is considered a major step in an ongoing debate on how best to handle two different, but often overlapping, sets of diagnostic tools in a manner that best serves patient safety and public health, while recognizing the realities of clinical practice and medical product development.

Stakeholder interest was intense. The original meeting space reached capacity and registration closed within two days, prompting FDA to shift the conference to a larger venue. Nearly 650 people attended, while 650 more watched via webcast.  FDA’s sense of urgency on the matter was further suggested by the June 10 issuance of “it has come to our attention” letters to six genetic testing companies, followed by another 14 on the opening day of the conference itself.

What are the issues? Why the concern? What does it all mean, and where might the Agency go?

Once upon a time, medical diagnosis was a subtle art with simple tools—the sight, sounds, feel, even smell and taste of the patient, combined with personal and family history and the physician’s skill and experience. Even with the advent of more complex instruments, from stethoscopes and thermometers through scopes, monitors, and blood tests, the administration and interpretation of diagnostic procedures remained within the realm of medical practice, or at least within the treating physician’s extended world of medical or scientific colleagues at a hospital, with whom he enjoyed a direct collegial relationship.

Lab tests for transfusion and transplant patients are created under exceptionally stringent standards, but are so customized that they do not lend themselves to manufacturing rules – Andrea Zachary of the American Society for Histocompatibility & Immunogenetics

This situation has dramatically changed during the past 50 years. The rapid introduction of new diagnostics has greatly improved medical care, and also added new costs and scientific complexity. Two distinct paths for diagnostic tests emerged: some are still developed and produced by internal hospital analytic and pathology labs, but a growing number are purchased as packaged kits from manufacturers, or contracted out through commercial service companies.

Differing oversight

The oversight of the two routes is different. Those created in hospital labs, sometimes deprecatingly called “home brew,” fall under facility quality standards, notably the CLIA (Clinical Laboratory Improvement Amendments) regulations administered by CMS (Centers for Medicare & Medicaid Services). CLIA establishes lab requirements at five levels of stringency according to the complexity of the diagnostic tests. Higher-level CLIA certificates require regular facility inspections that evaluate physical set-up, procedures, personnel training and proficiency, quality control, and record-keeping. Labs must document their assessment of test clinical relevance, but CLIA inspectors only monitor analytic validation, relying on lab directors or clinical consultants to determine whether the tests actually yield the medically useful information they are purported to do. Other organizations, including the College of American Pathologists and the American College of Medical Genetics, provide guidelines or recommendations, and New York State has its own strict clinical lab requirements, though these all lack the nationwide consistency and authority of federal regulation.

FDA regulation evolved along a path very different from the medical practice model. Originally focused on fraudulent and mislabeled medical products, the Agency included safety considerations in the Food Drug & Cosmetic Act of 1938, and required premarket demonstration of drug safety and effectiveness following the 1962 Kefauver-Harris Amendments. Some high-risk devices and a few diagnostics were informally regulated as drugs, but only with the Medical Device Amendment of 1976 was device regulation legally formalized.

Diagnostics swept along with devices

Diagnostics were swept along with devices. In 1976, aside from the raw chemical reagents and laboratory procedures – neither within FDA’s jurisdiction – the critical components of diagnostics were the hardware used to process tests. The original Division of Clinical Laboratory Devices evolved into the current Office of In Vitro Diagnostics (OIVD). If OIVD’s placement within CDRH was serendipitous, the flexible, risk-based Quality System Regulation approach to devices lends itself well to the explosive development of new diagnostic modalities (or as well as possible, with caveats).

Drugs, devices, and biologics are different, hence there are different FDA centers with different rules. But diagnostics differ from all of them. And one diagnostic is not like another.

FDA-regulated IVDs, like medical devices, are subject to various pre- and post-market requirements, and are classified based on risk. Fifty percent of IVDs are Class I, including erythrocyte sedimentation rate tests, differential culture media, and other low risk tests that are exempt from premarket clearance, requiring only registration, general controls, record-keeping, and post-market reporting. Forty-two percent are Class II – e.g. factor deficiency, antimicrobial susceptibility, or thyroid stimulating hormone tests – that require 510(k)s and special controls such as performance standards. Class III diagnostics that require PMAs, eight percent of the current universe, are high-risk tests whose results directly or strongly influence patient management of a serious disease, e.g. cancer screening tests, companion diagnostics, hepatitis B or C and HIV tests, and continuous glucose monitors.

There are dangers of improperly validated tests, but there are more dangers from not having the tests – Paula Revell of the American Society for Microbiology

Technically, FDA has authority over all diagnostics, as defined under 21 CFR 809.3: “In vitro diagnostic products are those reagents, instruments, and systems intended for use in diagnosis of disease or other conditions, including a determination of the state of health, in order to cure, mitigate, treat, or prevent disease or its sequelae.” As a practical matter, FDA has exercised “enforcement discretion,” a revocable hands-off attitude that has been used in other areas, as with certain health claims about foods, or allowing entry of limited quantities of some foreign drugs for personal use.

The FDA stance on diagnostics began to shift in the 1990s.  The “ASR rule,” a set of three 1997 regulations on analyte specific reagents, sought to ensure the quality of key components of finished IVDs or LDTs such as antibodies, nucleic acid sequences, receptor proteins, and ligands. A 2006 draft guidance on the regulation of In Vitro Multivariate Index Assays (IVDMIAs), revised in 2007 but never finalized, sought greater control over complex tests that used proprietary algorithms to analyze multiple variables in the diagnosis or prediction of disease. Here, too, the concern was a lack of scientific or methodological transparency, meaning that health care providers were unable to evaluate the clinical validity of tests they employed.

At the same time, IVD manufacturers objected to the erosion of the level playing field. Tests with similar risks, or even the same test, can enter clinical practice through either the LDT or IVD route. Clinicians may not differentiate between the two, though the journey along the FDA-regulated pathway is longer, steeper, and more costly than that for LDTs, some of which are now offered beyond the walls of individual institutions through what seems to some an alternative, below-the-radar pathway with lower costs and less certain guarantees.

FDA oversight of LDTs to change

The July meeting opened with remarks by CDRH director Jeffrey Shuren and Principal Deputy Commissioner Joshua Sharfstein, who stated that FDA regulatory oversight of LDTs would change, though just how remains unknown. Nearly 70 stakeholder presentations and four panel discussions made clear that the journey would be neither straight nor swift.

The most contentious discussions fell within a few areas: whether the FDA regulation of products or the CLIA regulation of laboratories is a more appropriate model; whether FDA procedures would stifle innovation, limit patient access to needed tests, and raise barriers for tests developed for small populations or to meet transient emergency needs; how diagnostic tests should be risk-stratified; and whether, if FDA does claim oversight of a vast range of LDTs, the Agency has the necessary resources.

Need to create inventory of all available diagnostic tests

Speakers and attendees generally agreed that an invaluable first step would be the creation of an inventory of what diagnostic tests currently exist. The NIH has a voluntary genetic test registry, but no broader, mandatory requirement is in place to identify the thousands of tests in use. Knowledge of what these tests are, of the evidence supporting their clinical utility, and the sort of post-market surveillance of adverse events required under FDA regulations, are distributed in the peer-reviewed literature and in the minds of lab personnel.

There was general agreement that duplicative regulation of identical tests by multiple organizations is wasteful. Some maintained that the CLIA regulation of clinical laboratories is adequate, and that the LDT model fosters a valuable close working relationship between health care providers and the pathologists and clinical lab staff who help interpret diagnostic test results. On the other hand, FDA history is riddled with examples where the clinical judgment of practitioners, lacking the broad statistical evidence available to FDA, failed to perceive serious problems only discernible on a population-wide basis.

The agility of labs to respond to urgent or immediate needs was a major topic of discussion. Col. Alan Magill of the Walter Reed Army Institute of Research pointed out that virtually all diagnostics used with tropical diseases are not and never will be FDA approved. These aim at very small markets and would vanish overnight if any obstacle were put in the road.

Andrea Zachary of the American Society for Histocompatibility & Immunogenetics stated that lab tests for transfusion and transplant patients are created under exceptionally stringent standards, but are so customized that they do not lend themselves to manufacturing rules. Others pointed out the episodic, emergency need for certain tests, whether to look at health risks to workers cleaning up Gulf oil spills or at the World Trade Center site, or for public health surveillance of emerging pathogens or food-borne diseases.

Richard Naples of BD, which manufactures regulated IVD kits, concurred that one size does not fit all diagnostic tests and that there must be flexibility commensurate with population size and clinical need.

“There are dangers of improperly validated tests,” said Paula Revell of the American Society for Microbiology, “but there are more dangers from not having the tests.” This balance between patient access and the responsiveness of tests to public health needs against the desire to have consistent, high standards of analytic and clinical reliability were the most difficult challenge in the meeting. From FDA’s perspective, marketed manufactured tests fall under the Agency’s jurisdiction and must meet certain requirements.  The opportunistic expansion by some laboratories and companies of LDTs into what look a lot like a commercially distributed IVD is a difficult phenomenon to manage.

One size does not fit all diagnostic tests; there must be flexibility commensurate with population size and clinical need – Richard Naples of BD

Defining risk for diagnostics

If risk is the criterion for differentiating the type and degree of appropriate regulation, how is risk defined? Risk can be stratified according to the type of disease, and the risk-benefit profile of the medical intervention that follows a test, but a single diagnostic test is rarely the sole determiner of a life-or-death treatment. As representatives of the American Association for Clinical Chemistry and the College of American Pathologists pointed out, one should also factor in the complexity of the test, the ability of hospital labs or clinicians independently to gauge clinical validity, and the overall clinical context in which tests are employed.

Presenters from laboratories and smaller, innovative commercial entities voiced concern over the potentially prohibitive costs of meeting CDRH requirements in a market where profit margins tend to be slim. There was also much worry that CDRH, already struggling to meet its current work load with existing resources, would be overwhelmed by the onslaught of thousands of new diagnostic tests. Suggestions for handling this challenge included grandfathering in relatively simple, low-risk tests that are well understood and widely accepted parts of current clinical practice (as pre-amendment devices were grandfathered in 1976), making extensive use of external review panels, or simply taking a more thorough and reasoned approach to the division of diagnostics between LDTs overseen under CLIA and IVDs regulated by FDA, but upgrading higher-risk LDTs to IVD status while down-classifying lower-risk IVDs.

Direct-to-consumer diagnostic kits

Considerable discussion hovered around direct-to-consumer marketed diagnostic tests, notably genomic screening services that purport to provide personal disease risk profiles.  Though some presenters – mostly those who made and offered such testing services – defended the value of the information they provided, most commentators were critical of the sometimes questionable lab oversight and the lack of reliable medical guidance to help consumers understand the results they received.  Arguably the most egregious problems with DTC genomics tests – some grotesquely highlighted in a 2006 GAO study on “nutrigenetic testing”—are more properly a target for FTC scrutiny than for the FDA.

In the closing panel discussion at the end of two days, moderator Joan Scott of the Genetics Public Policy Center at Johns Hopkins asked for suggestions on next steps.  “More workshops like this,” replied Elizabeth Kato of AHRQ, while other panelists suggesting letting FDA take the next step, further defining the questions, identifying stakeholders to help further flesh out the answers, and issuing a draft guidance that would elicit the next round of discussion.


  1. Webcast of the FDA/CDRH Public Meeting: Oversight of Laboratory Developed Tests (LDTs), July 19-20, 2010.
  2. Individual stakeholder presentations at the meeting: – docketDetail?R=FDA-2010
  3. PowerPoint presentations by FDA speakers can be requested by sending an e-mail to Katherine Serrano in OIVD Device Evaluation and Safety at
  4. FDA Overview of IVD Regulation:
  5. Current CLIA regulations:
  6. May 2010 AHRQ Technology Assessment, “Quality, Regulation and Clinical Utility of Laboratory-developed Molecular Tests”:
  7. July 2007 FDA Draft Guidance on IVDMIAs:
  8. September 2007 FDA Guidance: FAQs on Analyte Specific Reagents (ASRs):
  9. April 2008 Final SACGHS report on the oversight of genetic testing:
  10. 2006 Testimony before the Special Committee on Aging, U.S. Senate, on Nutrigenetic Testing:

About The Author

Leave a Reply

Your email address will not be published. Required fields are marked *