Regulation of Biomarker Clinical Laboratory Tests Used in Personalized Medicine

Personalized medicine allows for the selection of therapies specific to an individual patient's own genetic makeup. To translate the information from the genome into therapeutic products, we need links between the genetic information and the clinical outcomes associated with that information. These links—or biomarkers—are the focus of intense interest in biopharmaceutical research and development today.

Paul Radensky

As part of product research and development, these biomarkers can be used to identify promising candidates in terms of predicted effectiveness or safety profiles, and these markers may be accepted as surrogate endpoints to allow for more rapid product approvals. In clinical practice, these biomarkers may be developed as diagnostic tests to help: (1) identify patients with a particular disorder or condition (diagnosis), (2) distinguish patients who are likely to benefit from treatment from those who are unlikely to benefit (treatment prediction), (3) identify patients who may be likely to experience serious adverse events (toxicity), (4) monitor therapy on an ongoing basis (safety and effectiveness), and (5) determine dosing.

As diagnostic tools, biomarkers can be developed for clinical use as in vitro tests, imaging procedures, or as other diagnostic modalities. Many novel biomarkers are being developed as "companion diagnostics" tied to specific therapies, with the regulatory clearance or approval of the biomarker combined with the clearance or approval of the therapeutic, and the intended use of the biomarker explained in the labeling for the product. Others are being developed as stand-alone diagnostics not tied to a specific therapeutic product. When developed in combination with the therapeutic, the scientific evidence required to support the approval of the biomarker and the timeline to approval are tied to the requirements and timeline of the therapeutic. However, when the biomarker is developed as a stand-alone diagnostic, the regulatory requirements and timelines are less clear. In addition, whether developed as a companion or as a stand alone, important questions about how these tests will be paid for must be considered. In the sections below, we discuss some of these issues in the context of biomarkers being developed and used as clinical laboratory tests.

CURRENT REGULATORY PATHWAYS FOR BIOMARKER CLINICAL LABORATORY TESTS

Diagnostic tests involving the examination of human specimens for providing information for the diagnosis, prevention, or treatment of any disease or the assessment of health are clinical laboratory tests that lawfully may be performed only by certified clinical laboratories (unless they have been cleared for use by patients in their homes). Clinical laboratories must be certified to perform testing under the federal Clinical Laboratory Improvement Amendments of 1988 (CLIA) and are generally required to be licensed by individual states as well. Clinical laboratories certified under CLIA and licensed under state law must meet prescribed standards governing personnel, quality systems, and proficiency and must undergo periodic inspections to assess compliance with these standards. Some laboratories are inspected by private accreditation organizations, such as the College of American Pathologists (CAP) or the Joint Commission, whose standards are more rigorous than those specified by the CLIA program.

Clinical laboratories can follow one or both of two pathways in developing and offering novel biomarker diagnostic tests: (1) the laboratory can develop a test for its own use in testing specimens sent to the laboratory (laboratory-developed tests or LDTs) or (2) the laboratory can perform a test using a test kit cleared or approved for use by the US Food and Drug Administration (FDA).

Diagnostic tests that are developed as kits by manufacturers and shipped to laboratories for performing tests clearly fit under the definition of a medical device under the Federal Food, Drug, and Cosmetic Act (FFDCA). Similarly, reagents, equipment, and supplies used by laboratories to perform diagnostic tests generally fit under the definition of medical devices. As medical devices, these test kits or components are subject to FDA clearance or approval based on an assessment of the novelty and risk involved.

Manufacturers of the lowest-risk test kits or components generally must comply with limited regulatory requirements under the FFDCA, such as registration of the establishment with the FDA, listing of the test kit or component, maintaining complaint records, and reporting adverse events to the FDA. Compliance with the FDA's quality system regulations may be required as well.

Manufacturers of moderate-risk test kits or components must comply with the same controls but also generally must submit a notice to the FDA before marketing, called a 510(k) clearance, to show that the new test (or component) has the same intended use and technological characteristics as some established test kit or component, or that any differences do not raise significant issues of safety or effectiveness (which may require support from studies, including clinical studies).

The highest-risk test kits or components require approval of a premarket approval application (PMA) before commercial use, which involves conducting clinical trials to show that the test is accurate and reliable. Major modifications to these highest-risk tests, such as modifying labeling for a new intended use, also will require new clinical trials for approval of the modification.

In contrast, diagnostic tests that are developed as LDTs have historically been regulated by CLIA and state law and do not follow an FDA pathway. The CLIA pathway requires establishing the performance specifications of the test. This includes the analytical validity of the assay—how well the test measures what the laboratory says it measures—and may include clinical performance characteristics, such as sensitivity, specificity, and predictive values. Compliance with CLIA requirements also involves maintaining quality systems aimed at ensuring the laboratory results are accurate and reliable. Laboratories offering LDTs must make available a clinical consultant to help the treating physician in ordering appropriate tests, to ensure that test reports include pertinent information for specific patient interpretation, and to provide consultation on matters related to the quality of the test results reported and their interpretation concerning specific patient conditions.

IN VITRO DIAGNOSTIC MULTIVARIATE INDEX ASSAYS

For many years, the FDA maintained it had the authority to regulate LDTs as medical devices but would use enforcement discretion not to require laboratories to obtain clearance or approval for LDTs as long as certain requirements were met (including a prescribed disclaimer to be included in reports of tests developed using FDA-cleared or approved analyte specific reagents). In 2006, however, the FDA announced its intention to regulate, as medical devices, one segment of LDTs, which the Agency described as in vitro diagnostic multivariate index assays (IVDMIAs). IVDMIAs had never been described before the draft guidance was published in September 2006, and the definition was revised in a subsequent draft guidance released in July 2007, so the precise definition remains unclear. The focus of the draft IVDMIA guidance appears to be diagnostic tests comprising two or more underlying variables that are combined using an interpretation function, such as a mathematical algorithm, to produce a patient-specific result. In its draft guidance, the FDA identifies the following examples as fitting under its definition of IVDMIA:

• a gene expression profiling assay for breast cancer prognosis

• a test integrating quantitative results from multiple immuno-assays to obtain a qualitative score that predicts a person's risk of developing a disease or condition

• a test integrating a patient's age, sex, and genotype of multiple genes to predict risk of or to diagnose a disease or condition.

The release of the draft IVDMIA guidance generated substantial confusion and controversy because many in the laboratory community believed this represented an important reversal of long-standing FDA policy of not requiring clearance or approval of LDTs as medical devices. Concerns were also raised that FDA was imposing new regulatory requirements on clinical laboratories that are regulated under CLIA and state laboratory licensure laws through a guidance process, which is inconsistent with the nonbinding status of guidance documents. At the same time, most stakeholders agree that these advanced technology assays may require more oversight than is provided under the existing framework under CLIA and state laws.

The Secretary's Advisory Committee for Genetics, Health, and Society (SACGHS) issued a report this spring identifying gaps in the regulatory oversight framework for genetic tests, including the kinds of biomarker tests discussed here. Among many recommendations, SACGHS has called for creating a registry to allow for the collection of information about these tests, which can help promote transparency and accountability for the truthfulness of claims about biomarker clinical laboratory tests. In addition, a registry could allow for data collection to inform appropriate regulatory pathways for new tests. In addition, a few legislative proposals have been introduced in Congress that also may lead to the development of a new oversight framework for these tests.

Discussions continue among laboratories, medical device manufacturers, other stakeholders, and regulatory authorities at the FDA and the Department of Health and Human Services. Principal issues that remain to be worked out include: (1) which agencies are best equipped to oversee the development and performance of these tests, (2) what standard of evidence should apply to support intended use claims that are meaningful to physicians and patients, (3) how a new oversight framework can be developed and implemented without stifling innovation and delaying the introduction of useful biomarker diagnostics, and (4) how these tests will be reimbursed. Developing answers to these questions will not be an easy task, but the stakes are too high not to meet the challenge.

Paul Radensky, MD, JD, is a partner at McDermott, Will & Emery, Washington, DC, 202.756.8794, pradensky@mwe.com