The US Food and Drug Administration should model its proposed regulatory guidelines for a complex subset of algorithm-based homebrew tests on the New York State Department of Health’s policy for laboratory-developed assays, speakers told the FDA at a recent public meeting.
Several speakers at a Feb. 8 FDA meeting in Gaithersburg, Md., said New York’s Clinical Laboratory Evaluation Program is a viable framework for the FDA to keep in mind when developing a “least burdensome” review process for the products, which the FDA calls in vitro diagnostic multivariate index assays.
“The US has in the New York State model an alternative premarket review process [that] is already applied successfully to in-house tests and maybe the FDA can learn from this model,” Stuart Hogarth, a research associate at the University of Cambridge’s Epidemiology for Policy Group, said at the meeting.
The meeting was convened by the FDA to discuss the regulation of IVDMIAs, which use algorithms to interpret gene and protein data to guide medical decision-making [see PGx Reporter 02-14-07]. Many of the companies whose IVDMIAs FDA has said it intends to review have already garnered approval under the New York system.
Before a laboratory-developed test that relies on biological specimens can be sold in New York State it must first be regulated by the state’s Department of Health, even if the tests were developed outside New York. On the NY DoH’s website, the agency requests “complete submissions … for molecular-based infectious agent, oncology, and genetic assays,” as well as “documentation of initial and ongoing validation” for homebrew tests.
Under CLEP, the department awards exceptions for PMAs under certain circumstances, particularly if the product is FDA approved or if the laboratory has garnered an exemption by cultivating a positive record of following the DoH’s validation protocols.
While the state health agency regulates IVDMIAs the same way as other homebrew tests, the FDA believes IVDMIAs should receive special treatment.
Historically, the agency has practiced enforcement discretion for tests developed at a single laboratory, deferring instead to CMS’ Clinical Laboratory Improvement Amendments guidelines. According to an FDA draft guidance, IVDMIAs, due to their complexity, do not fall under the category of regular homebrew tests and must therefore be cleared by the agency [see PGx Reporter 9-13-06].
However, during the meeting, a majority of speakers representing industry, interest groups, and patient advocacy organizations said the FDA’s draft guidelines are confusing, potentially illegal, and may be a barrier to innovation for diagnostics companies catering to niche disease populations.
According to Hogarth, the existence of the New York model is proof that a system of premarket approval can exist without stifling innovation. As a balance to FDA regulation, speakers recommended several “tools” the agency can use to incentivize the development of diagnostics, such as granting marketing exclusivity for “orphan diagnostics” or small volume tests, and creating a publicly accessible online databank of marketed assays.
Can NY Show FDA the Way?
One of the main concerns at the public meeting was how FDA regulations would dovetail with CLIA guidelines. It turns out that the New York system appears to be a happy medium between the FDA system and CLIA on the state level.
“Although we heard that we have a wonderful model in New York, it in fact combines the activities of CLIA and the FDA,” Patricia Charache, representing the American Society for Microbiology, told meeting participants in an unscheduled presentation. “So, New York State has had available something that we don’t have nationally, where the moving together of the policies of FDA and CLIA seems to me almost like trying to make the North rim and the South rim join at the Grand Canyon.”
While speakers highlighted New York’s policies for lab-developed tests positively during the meeting, it is unclear how it would inform a national policy specifically for IVDMIAs, particularly since, according to officials from the NY DoH’s Clinical Laboratory Evaluation Program, the division currently has no special policy for IVDMIAs.
“These types of assays are evaluated like any other assay on a case-by-case basis, and our standard submission guidelines apply,” Michele Caggana and Erasmus Schneider of the CLEP wrote in an e-mail to Pharmacogenomics Reporter this week. “While the analysis may be more complex, the criteria for clinical validity are not dissimilar from single-point assays. Using this approach, several IVDMIAs have been approved by the oncology section.
“Many discussions with regulators and laboratories are needed prior to [determining] whether the NYS model may work on a national level,” said Caggana, who is head of genetic testing, and Schneider, who is head of oncology, at CLEP.
Still, New York’s system does provide a functioning example of how FDA might coordinate its activities with CLIA to develop a broader regulatory approach for all laboratory-developed diagnostics.
The NY DoH’s review process for diagnostics comprises a system of premarket approvals, conditional exemptions, and surprise audits to ensure the efficacy and safety of tests.
On its website the DoH outlines that laboratories must get premarket approval from the department for: assays labeled “Research Use Only” and those using analyte specific reagents; FDA-approved assays for “investigational use only” and RUO assays that have been modified from their intended use; and tests developed in house.
In order to sell diagnostics in New York, the laboratory must have a qualified director with at least four years postdoctoral laboratory experience and also have expertise interpreting genetic test results. The lab must show the DoH that it is compliant with standards of practice.
Unlike under CLIA, which some during FDA’s meeting pointed out, only require labs to analytically validate tests – the New York system requires labs to submit data supporting both analytical and clinical claims.
DoH approval is not needed if the test is FDA-approved, is an assay with “a record of reliability and clinical validity” using “a standardized protocol that is universally applied in labs;” is an unmodified “investigational use only” test, or has been granted an “exemption” by the department.
In order to be considered for an exemption, a laboratory must have a consistent record of having followed the department’s validation protocol, and “have a sufficient number of representative methods approved through the full validation review process,” the NY DoH states.
According to the department, the laboratory may receive a limited exemption by submitting a standardized protocol for method validation, which includes “a description of general principles and laboratory-specific practices for assay development and initial validation, an algorithm for assay validation using reference assays and/or clinical findings, and laboratory-specific protocols for on-going validation, including quality control procedures and quality assurance indicators.”
Additionally, the department conducts “random audits” of exempt assays by requesting additional data, which could be as comprehensive as a full validation package. “If these submissions are found to be deficient so as not to support performance claims or there is evidence that the approved validation protocol is no longer being followed, any approvals granted under the exemption will be re-examined,” the state health agency warns.
“To date, no full exemptions have been granted,” Caggana and Schneider said. “One laboratory has been granted a limited exemption where submission requirements are abbreviated. In genetics there are two laboratories that submit report language, the test-specific [Standard Operating Procedure Manual], reports, and validation. No laboratories have a full exemption under the genetics or oncology categories.”
At the meeting, Hogarth particularly highlighted the New York system as evidence that FDA’s regulation of lab-developed tests doesn’t have to stifle innovation.
“Clearly there is a concern that FDA regulation of in-house tests may become a block to innovation,” he said. “Companies like Quest, LabCorp, and Genomic Health are at the leading edge of diagnostic innovation. The fact that they are NY-licensed would suggest that pre-market review of in-house tests may not be a major block in innovation,” Hogarth said.
“We have very little data on the connection between regulation and innovation,” he added.
Incentives to Balance Regulation
“The US has in the New York State model an alternative premarket review process [that] is already applied successfully to in-house tests and maybe the FDA can learn from this model.”
At the meeting, some suggested that the FDA should regulate all home-brew tests, not just IVDMIAs. Others take an “if it isn’t broken, don’t fix it” approach in support of keeping oversight in the hands of CLIA.
Others felt that the FDA should make it voluntary for companies to pursue FDA approval, so the policy doesn’t penalize diagnostic shops in rare disease markets that are dependent on venture funding and can’t afford lengthy and costly development timelines [see PGx Reporter 02-14-07].
Members of the Coalition for 21st Century Medicine, an alliance formed after the FDA released its IVDMIA draft guidance, feels that the agency should strengthen existing CLIA regs before pursuing its proposed guidelines.
In a recent interview with Pharmacogenomics Reporter, Paul Radensky, an advisor to the coalition, said that the existence of state-level regulation further supports the argument against FDA oversight of IVDMIAs (see Q&A in this issue). If the agency does go forward with finalizing its draft guidance, low-volume tests should be exempt from having to get PMAs, the group insists.
Hogarth was one of the presenters who advocated for FDA to take a “comprehensive approach” in its policy toward all homebrew tests. According to the researcher, the FDA’s current practice of exercising enforcement discretion “creates a situation where the only people who can perform the test are those with a vested interest in its promotion.
“This creates anxiety that in order to recoup their R&D, companies may make strong clinical claims of their tests at a stage when the evidence base is still developing,” he said.
“The point is not that all companies making IVDMIAs are bad players making dangerous tests,” Hogarth said. “The point is that without independent evaluation by the FDA there is no way for doctors and patients to distinguish good from bad.”
In contrast, Jonathan Cohen, president of 20/20 Gene Systems, said that regulation of diagnostics need to be balanced with incentives for companies to develop such products. 20/20 Gene Systems is a member of the Coalition for 21st Century Medicine.
Cohen cited a 2004 report that found that 3 percent of life sciences venture capital leads to diagnostics. “I have no reason to believe that [number] has increased. … It’s hard to believe that it could be any harder, but perhaps with these guidelines it could be,” he said.
Instead, the FDA can use portions of its orphan drug system as a way to regulate and protect small-volume diagnostics. “The orphan drug model is a success for FDA, where a combination of exclusivity, tax credits, and grants have created a robust market where one did not exist,” Cohen said. “In part, that is applicable to diagnostics today.”
Additionally, several speakers suggested that the FDA should develop an online, publicly accessible registry of marketed tests, which would provide published data on the products.
“This database concept, or the registry, is really where we want to go, at least for now for the next few years,” Cohen said. “FDA is the right agency to manage that process and empower the marketplace to pick the winners.”