October 2011: Volume 37, Number 10
Under Construction: The Road to FDA Clearance
Should New Tests Be Cleared Based on Old Laws?
By Bill Malone
Labs work hard to deliver accurate test results at the right time to help ensure that patients receive safe and effective care. With patient safety a top priority, labs must believe that test kits they purchase from manufacturers have passed muster with the Food and Drug Administration (FDA) through the agency’s 510(k) clearance process and can be trusted. But is FDA really clearing new tests for market based on safety and effectiveness like labs and patients assume?
A new report from the Institute of Medicine (IOM) issued in July paints a picture of FDA working with one hand tied behind its back—scant legal authority to demand that most tests and other medical devices are truly safe and effective, but an obligation to ensure public health. To allow lab tests to be sold by manufacturers, FDA clearance relies on a device’s similarity to an already cleared device or one on the market before 1976 when the 510(k) law was passed—not a high enough standard, according to IOM, and cause for the 510(k) system to be scrapped and replaced.
These were not charges that either FDA or manufacturers were quite ready to hear. As test menus have grown and laboratory science become more complex, manufacturers have publicly expressed weariness and concern that FDA moves too slow or is too risk-averse. In response, FDA has promised to step up efforts to accelerate the process. Now, at the same time that industry complaints about FDA are reaching a climax, the report from the highly regarded IOM has added new, uncomfortable questions about FDA’s plans to improve a system that reviews upwards of 90% of tests and other medical devices on the market.
The IOM committee that developed the report made clear that it is not warning of an impending public health disaster. However, committee members felt they could not endorse a framework that they believed, at its core, does not focus on the safety and effectiveness of new medical devices, said committee member Steven Gutman, MD, associate director at the Blue Cross Blue Shield Association Technology Evaluation Center and former director of the Office of In Vitro Diagnostic Device Evaluation and Safety (OIVD) at FDA. “We don’t think this is some catastrophe waiting to happen. But we definitely had a great sense that looking backwards and trying to map new products to the past may have served the agency for 35 years, but that it was time for the agency to consider a regulatory path that looked forward,” he said. “What we really didn’t want to happen was in 35 more years, we would have another IOM report that said FDA was still finding things appropriate for market based on similarity to products that linked back to 1976. That seemed to us illogical.”
What is the Standard?
Unlike standards for the pharmaceutical industry, the 510(k) section of the Federal Food, Drug, and Cosmetic Act does not require device manufacturers to submit clinical trials and other evidence to FDA demonstrating safety and effectiveness for most new medical devices, including in vitro diagnostics. The act requires such evidence only for what are considered high-risk devices, for example, a test based on totally new technology. These high-risk devices go through a more rigorous review called premarket approval (PMA). FDA clears the vast majority of tests via its 510(k) program. For this path, the criterion is substantial equivalence to a predicate device—a comparison to a similar product already on the market.
Manufacturers can chose from a wide variety of already-cleared products, and in some cases multiple products, to make a case for substantial equivalence. Usually companies chose a fairly recent product. However, all products must be cleared based on a chain of similar products that can be traced back to 1976 when the statute became law.
The words safety and effectiveness do find their way into the statute governing 510(k)s, wherein a device must be “at least as safe and effective” as another legally marketed device. Yet the final bar in the law remains the previous product, and not safety or effectiveness per se, according to Gutman. “The law as it is dances around the idea of safety and effectiveness,” he said. “The IOM didn’t believe that a product being ‘as safe and effective’ as a predicate to be the same thing as showing that a new device is in fact safe and effective. We thought that it would be better if FDA simply had the authority to do what we think the public thinks that they’re doing, and make sure that new devices are truly safe and effective.”
This continuous accrual of comparisons is no substitute for a firm standard of safety and effectiveness, and marks a fatal flaw in the 510(k) program, according IOM committee chair David Challoner, MD, emeritus vice president for health affairs at the University of Florida in Gainesville. “Our careful, in-depth study confirmed that the 510(k) legislation was put in place simply to prevent a 1977 market disruption that would have resulted from making every device already on the market go through a pre-market safety and effectiveness evaluation. Now, like a vestigial appendage, it has never gone away over the last 35 years. The committee realized that the 510(k) emperor had no safety and effectiveness clothes.”
The Advanced Medical Technology Association (AdvaMed) and other industry groups criticized the IOM report immediately, firing back that the “510(k) process is overwhelmingly safe,” and preferring that FDA forge ahead with its existing plans to tweak the current regulatory scheme (See Box, below). Even FDA backed away from the report, which the agency itself had requested, saying the 510(k) program should be improved but not replaced. However, some in academia and medicine spoke out to agree with IOM’s conclusions, including Gregory Curfman, MD, executive editor of the New England Journal of Medicine and Rita Redberg, MD, editor of the Archives of Internal Medicine. “We believe that the IOM report is insightful, judicious, sensible, and long overdue. The 510(k) clearance process was established 35 years ago, and although it may have been a reasonable approach then, it surely is not today,” the pair wrote in an editorial in the New England Journal of Medicine (10.1056/NEJMp1109094).
A Look at FDA Plans for Improving the 510(k) Program
After a 2-year study, in January FDA announced its own plan to improve the 510(k) program through which most tests and other medical devices are cleared for marketing. The changes are intended to improve the predictability, consistency, and transparency of the 510(k) program, and to facilitate innovation while assuring safety and effectiveness, according to the agency. The 510(k) improvements include:
- Streamlining the review process for innovative, lower-risk products, through the “de novo” process.
- Publishing guidance for industry to clarify when clinical data should be submitted to increase predictability and transparency.
- Developing a network of external experts who can use their knowledge and experience to help the agency address important scientific issues regarding new medical device technologies.
- Establishing a new Center Science Council of senior FDA experts within the agency’s medical device center to assure more timely and consistent science-based decision making.
- Establishing a public database of important device information, such as medical device labeling and summaries of the basis for the FDA’s decision to clear specific devices.
- Requiring a brief description of scientific information regarding the safety and effectiveness known to the manufacturer for select higher-risk devices on a case-by-case basis through device-specific guidance.
Industry: Fix it, Don’t Ditch It
Although industry has frequently protested FDA’s 510(k) process as slow and difficult to navigate, manufacturers are setting their hopes on efforts already underway to improve the program. AdvaMed has strongly refuted the IOM report’s conclusion that FDA cannot or does not screen products for safety and effectiveness.
Despite the law’s focus on substantial equivalence, in practice FDA’s approach to 510(k) products does rest heavily on safety and effectiveness, emphasized AdvaMed’s vice president for technology and regulatory affairs, Janet Trunzo. “If you look at the kind of data that a company has to submit, it can range from bench testing, performance testing, testing against an FDA recognized standard or guidance document, all the way to clinical data,” she said. “FDA does have the authority to ask for whatever information it deems necessary to make an evaluation, and by the questions that FDA asks during the 510(k) evaluation process, it’s clear that the agency certainly does exert its authority to establish safety and effectiveness.”
Jeffrey Gibbs, an attorney at Hyman, Phelps and McNamara who represents companies on FDA-related issues, agreed, calling the IOM report “a triumph of form over substance.” The essence of FDA review is very different from the legal framework, he noted. “FDA questions and comments may be couched in terms of substantial equivalence, but even so, it’s safety or effectiveness they’re after,” he said. Gibbs is a former associate chief counsel for enforcement at FDA.
In fact, the way the 510(k) process works fits the medical device industry well, making room for continuous innovation of tests and other products, according to Trunzo. Moreover, the trajectory of incremental improvement means that substantial equivalence is almost always a comparison to a very recent device, not something on the market in 1976, she stressed. “If you think about the incremental change that occurs with 510(k)s—in that there is always this constant improvement, adding new technological features, making devices better—the idea that a company would claim equivalence to something that was a technology from 35 years ago just makes no sense.”
Manufacturers have become increasingly exasperated with what they see as confusion and delays at FDA, and are eager not to start over, but to streamline the current system, according to Gibbs. “I think everyone’s perception in industry is that the 510(k) process is becoming more difficult and unpredictable,” he said. “If everyone knew that it was going to take an extra 10 days, for example, I think people could deal with that more easily. But it’s very, very hard to predict anymore how long it will take, and that is a huge problem. And it’s not just duration—it’s how we get to that duration: unpredictability, uncertainty, unexpected questions. Those are hard to cope with for a manufacturer trying to stay competitive.”
In Vitro Diagnostics Are Different
The headlines in the mainstream press on the IOM report focused almost exclusively on devices like metal hip replacements. Part of this is due to the fact that there have been few in vitro diagnostic test-related failures that grabbed public attention like other products in FDA’s medical device category. However, as Alberto Gutierrez, PhD, director of OIVD at FDA explained, diagnostics are unique in the 510(k) program because the agency often uses a reference standard when comparing a new test to an old one. Rather than a point-to-point comparison, the agency can relate both the old and new device to the reference standard and determine substantial equivalence with an eye toward maintaining or improving performance.
For example, most new glucose tests are compared to the Yellow Springs Instruments’ glucose analyzer. This allows the agency to work within the legal framework of substantial equivalence, but still make sure that new products are safe and effective. For this approach to work, reference standards must be widely accepted gold standard methods recognized by the lab community, Gutierrez said, and his office uses this tactic for about half of all submissions.
Gutierrez could not comment directly on the IOM report, as his office was preparing at CLN press time for a public meeting on the report that took place September 16 in Silver Spring, Md. However, Jeffrey Shuren, MD, director of the FDA Center for Devices and Radiological Health within which OIVD operates, released comments that distanced the agency from IOM’s findings, emphasizing that the IOM recommendations are not binding and that “the public should continue to feel confident in the medical devices on the market today.”
If IOM believes the 510(k) program is so flawed, how can it also express confidence that most devices on the market are safe? Gutman said that because of the ingenuity and dedication of FDA, the agency has been able to stretch the original framework to fit the public health goals of safety and effectiveness. But the time has come for a better way. “FDA is tireless in trying to get the most out of the existing law that they can,” he said. “We think that FDA has remarkable fortitude and intelligence, and if they color a little bit outside of the lines, they might come up with something that makes more sense than looking for equivalence to a product that was marketed before. FDA needs to be able to ask the right questions. In some cases there may be more questions, in other cases there may be fewer, and maybe in some cases actual exemption from 510(k) review with more reliance on post-market controls, but always it should be the right questions with public health in mind.”
Little Chance of Congressional Action
Ultimately, the IOM report is a call to legislators, since FDA cannot change the laws behind regulation. IOM intended to start a public conversation, but did not want to micromanage the process or dictate how a new regulatory program would work, Challoner said. “We’ve run the flag up the flagpole for a public conversation. We’ll see who salutes politically as this conversation goes forward over time,” he said.
While opinions differ about the ultimate fate of 510(k), Congress has taken an interest at least in how FDA runs the program. In a July 28 hearing, the Senate Committee on Health, Education, Labor, and Pensions (HELP) took up the issue within the context of negotiating user fees for FDA. Since 2002, device manufacturers have had to pay fees to submit a product for FDA 510(k) clearance or premarket approval.
At the hearing, HELP Chair, Senator Tom Harkin (D-Iowa), rebuked industry representatives who had begun criticizing the IOM report before it was published. “I want to make sure that we have an agency that is independent and that is able to withstand the tremendous firepower of an industry that has a lot of money, and obviously wants the least amount of regulatory oversight,” he said.
On the other hand, HELP committee members also questioned whether FDA was going too far with 510(k) when it does consider safety and effectiveness, and not just equivalence. In an exchange with FDA Commissioner Margaret Hamburg, MD, Senator Richard Burr (R-N.C.) asked whether the agency actually followed the law on this point. Hamburg demurred. “At the end of the day, the goal is to ensure that products are safe and effective. Substantial equivalence is the criteria, but safety and effectiveness is the goal,” she said.
Burr noted again that the applicable statue was substantial equivalence. “Are you telling me that’s not the threshold?” he asked. “I would question whether you have the authority without a change in rules to do what you’re saying you do.”
Other members of the HELP committee repeatedly questioned whether FDA was living up to its goals of efficiency in reviewing new products, expressing concern that FDA might dampen innovation and suppress an industry that provided jobs to a foundering economy.
According to Gibbs, there’s little chance Congress will tackle new laws for the 510(k) process anytime soon. “It is a call to legislators, which of course means that it’s a call to nothing,” he said. “The chance of legislators agreeing on how this ought to be handled is nil. I don’t think there is any enthusiasm in Congress to revisit the 510(k) process.”
In the wake of the IOM report, Gibbs suggested that labs and test manufacturers take a hard look at FDA’s ongoing plan for 510(k) reform, as well as a deluge of draft guidance documents issued by the agency recently (See Box, below). “Companies are going to have to work harder than they have in years to try to keep up with the recent flood of draft guidance documents and make sure that comments are submitted, whether it’s by individual companies or trade associations,” he said.
FDA Guidance Aims to Keep Research Products Out of Clinical Labs
Will It Prevent Labs from Performing Some Lab-Developed Tests?
At the same time that FDA has been working on its long-term plan to improve the 510(k) review process, the agency has kept the lab community busy with a recent flurry of draft guidance documents. Among the most controversial has been that on research use only (ROU)-labeled products, which include reagents, instruments, software, and test kits.
Unless FDA’s draft guidance on RUO products is changed, many clinical labs would find themselves unable to obtain the products they need for some of their own laboratory-developed tests (LDT), industry and lab associations are warning. The recent draft guidance makes clear that, not only are RUO-labeled products inappropriate for clinical testing, but also that FDA will go after manufacturers who knowingly sell RUO-labeled products for clinical use.
Generally, labs use analyte-specific reagents (ASR) or general purpose reagents, along with FDA-cleared instruments, to assemble their LDTs. However, when an FDA-cleared product is not yet available and is a necessary ingredient in an LDT, labs have become accustomed to using products labeled “for research use only.” RUO products have never been sanctioned for use in clinical use, since they undergo no regulatory quality or safety checks. However, the new draft guidance signals a potential crackdown on manufacturers who supply RUO products, ending any toleration for this practice.
According to comments submitted to FDA from the Association for Molecular Pathology (AMP), the new draft guidance threatens labs’ ability to continue offering many tests, including most instruments, software, and reagents for sequencing assays. The guidance document also could dramatically limit access to high-resolution human leukocyte antigen (HLA) testing. In addition, newborn screening programs and hepatitis C genotyping could be shut down, AMP warned.
According to FDA, the guidance has not changed the rules, and only clarifies long-standing policy that labs and manufacturers should already be aware of. “If you look at the draft, it’s almost identical to what was put out in the early 1990s,” said Alberto Gutierrez, PhD, director of the FDA Office for In Vitro Diagnostic Safety and Evaluation. “We do point out in this draft that there is a regulation on the books that allows FDA essentially to take it a step further in the sense that if we think that a manufacturer is labeling something as RUO but knowingly selling it to a laboratory that is using it for diagnostics, then we believe that the manufacturer is responsible for that use.”
Gutierrez cautioned labs that RUO products are not supposed to be used clinically for a reason. “Labs producing LDTs based on RUO reagents have a reason to be concerned that their test is standing on very shaky ground because they have no idea if those reagents and instruments are well-validated, they have no idea whether the manufacturer is making changes or not and how that will affect their test, and therefore they have no real control over their LDTs,” he said. “Purchasing an RUO comes with no guarantees. I know labs feel that they can validate an RUO, therefore they know what it is, but I can tell you that validating your reagents is a very difficult thing to do. Most manufacturers don’t rely on testing alone to validate their reagents, and they also have contracts with their suppliers and audit them to make sure they’re producing them under good manufacturing processes and other controls. Most laboratories do not have any other controls than raw material acceptance testing with their RUO starting products. I think labs should be worried that they could easily lose control of their laboratory and provide erroneous results to their patients.”
FDA sees the clinical use of RUO products as a way around proper regulation for some manufacturers, especially those that put out complete test kits. Gutierrez suggested that labs talk to their suppliers and find alternative solutions to using RUO products. Manufacturers have the option of developing components of their RUO kits under ASR and general purpose reagent rules to follow a legitimate regulatory pathway, he said.
The draft guidance, “Draft Guidance for Industry and FDA Staff—Commercially Distributed In Vitro Diagnostic Products Labeled for Research Use Only or Investigational Use Only: Frequently Asked Questions,” is available from the FDA website.