Originally Published September 1999
REGULATIONS & LEGAL AFFAIRS
FDA in 2000: A Changing Regulatory Environment
Reengineered and modernized, FDA is striding forward into the post-Kessler era. To keep up, manufacturers will need to watch closely how the agency implements long-sought-after reform programs.
Jonathan S. Kahan
No serious observer of the medical device industry could fail to note that today's regulatory climate differs significantly from that which has existed for a good part of the current decade. But while most analysts agree that industry is less troubled by regulatory aggravations than in the recent past, dealing with today's changing environment is no less challenging.
Just as a number of factors contributed to the regulatory nightmare that many companies experienced earlier in the decade, the current improved environment also has roots in a variety of events. Certainly, the resignation of Commissioner David Kessler in 1997 has been one factor in changing attitudes at FDA. During his tenure, in the early 1990s, operations at the Office of Device Evaluation (ODE) in FDA's Center for Devices and Radiological Health (CDRH) almost ground to a halt. Equally important, CDRH repeatedly demonstrated that it was not particularly sensitive to the effects that the logjam in premarket notification (510(k)) clearances and multiyear delays in premarket approvals (PMAs) had on both large and small device companies. In an industry where a product can become obsolete in a matter of months, the delays in device marketing clearances were financially devastating for many companies.
Pressed by industry to repair the collapsing premarket review process, CDRH began in 1993 to implement a series of programs similar to those already implemented at FDA's Center for Drug Evaluation and Research. These included a system for prioritizing the review of product submissions (the "triage" process), rules to permit the agency to refuse to file deficient submissions, and standards for expedited review of important new technologies. More effective than these reforms, however, was the fact that ODE resources were focused on 510(k) reviews, and that ODE reviewers were once again empowered to clear devices without asking round after round of questions to build the review file.
Although the results of CDRH's early reform efforts were promising, pressure from industry continued to mount, invigorated in part by the Republican takeover of Congress in the 1994 elections. Agency officials recognized that unless further internal reforms were undertaken, the agenda for reform would likely be set by an emerging coalition between Congress and industry activists. CDRH therefore undertook a full reanalysis of how it cleared devices to market and initiated in earnest the process that has come to be known as reengineering. More than 50 separate reengineering initiatives were considered, including reviews of long-standing programs in both ODE and the Office of Compliance at CDRH.
To some extent, FDA's efforts at reform were both preempted and codified by enactment of the FDA Modernization Act of 1997 (FDAMA). Congress made it known that this legislation was to correct "an overly complex, bureaucratic, time-consuming, and expensive regulatory system" by assisting the agency in bringing "food and drug laws into the next century."
At the same time, however, Congress recognized that the device-related provisions of FDAMA were more evolutionary than revolutionary. Many of the reforms imbedded in FDAMA were based upon CDRH's earlier reengineering initiatives, including efforts to streamline administrative procedures and to accomplish its mandate with limited resources.
The current regulatory environment has been influenced by all of industry's experiences over the past decade. But how industry will fare in this reengineered and modernized post-Kessler era remains to be seen. One way to get a glimpse of the future is to examine how well—or how poorly—the long-sought-after reform programs are now being implemented by FDA and industry. Along the way, manufacturers may be able to discern the outlines of both strategic opportunities and areas in which regulatory problems might soon develop.
Clinical Data Requirements
In passing FDAMA, Congress imposed on FDA a greater degree of accountability for the way it pursues its mandate.1
The Senate Report on FDAMA, for instance, pointed out that the legislation was necessitated in part by the heavier regulatory burdens FDA had imposed on industry. The Senate found that "FDA's requirements for clinical testing and its premarket reviews of new products have grown increasingly complex, time-consuming, and costly."2 This burden, Congress explained, is
borne directly by the public in delayed access to important new products—including life-saving medical therapies—and in higher costs. They are a growing disincentive to continued investment in the development of innovative new products and a growing incentive for American companies to move research, development, and production abroad, threatening our Nation's continued world leadership in new product development, costing American jobs, and further delaying the public's access to important new products.3
In seeking to repair FDA's costly and time-consuming medical device clearance process, Congress took on a formidable challenge. But today's question is whether FDAMA was successful in meeting that challenge. Did the reform act result in CDRH and ODE implementing programs to change the product review process for the better? The answer is a clear yes and no.
As one example, we should examine FDA's attitude toward the clinical data required for a PMA application. Today, FDA is increasingly open to collaboration with industry as a means of establishing what data should be required. The agency is meeting with company representatives in several types of conferences mandated by FDAMA, including the "Determination" and "Agreement" meetings (sections 201 and 205, respectively), and the "Day 100" PMA meetings (section 209). Also in response to FDAMA's requirements, FDA has published several guidance documents in these areas.4-6
Together with the agency's initiatives for improving the review of investigational device exemptions (IDEs) and PMAs, these guidances have clearly helped to open the lines of communications between FDA and industry.
Even with the lines of communication opened, however, has the actual review process changed significantly? Arguably not. Although ODE now clearly recognizes that part of its mandate is to help make important new technologies available to the public as quickly as possible, the office still appears to be focused primarily on the second part of its mandate—ensuring that only safe and effective devices are allowed into the marketplace. Surely, no one in industry would argue that FDA should allow unsafe or ineffective devices onto the market. But the "zero-risk" mind-set of the Kessler era is still inherent in many ODE decisions.
When evaluating scientific issues, for example, there are often instances when a method less burdensome than a prospective, randomized, masked, multicenter clinical trial would be sufficient. Nevertheless, whenever a device raises even a slight hypothetical safety risk FDA often imposes significant clinical data requirements as a means of evaluating that risk. Such a zero-risk mentality can never really be changed by legislation, nor is the agency itself inclined to change it—no FDA reviewer wants to be held responsible for clearing a device later proved to present an unreasonable risk of injury. Instead, FDA's bureaucracy is strongly inclined to impose ever-increasing clinical data requirements so that there can be no question about whether the agency performed its job properly.
The silicone gel breast implant fiasco of the mid-1990s is an example of this paradigm. Although there was no hard evidence that such implants caused autoimmune disease, FDA essentially forced manufacturers to take their products off the market and to conduct extensive clinical studies with a 10-year follow-up. While scientific investigations now appear to have shown that FDA's fears were unfounded, thousands of women lived in fear of these diseases unnecessarily and hundreds of millions of dollars were wasted in spurious lawsuits. Hopefully, FDA has learned something from this fiasco, and some progress has been made, but it will still be difficult to change the zero-risk mentality at ODE.
510(k)s
One of the most important reengineering efforts involved the consideration of a new paradigm for the clearance of 510(k) notices. Ultimately, in March 1998, CDRH adopted the special 510(k) and abbreviated 510(k) options as alternative means for obtaining 510(k) clearance.
Abbreviated 510(k)s. The abbreviated 510(k) process permits a company to certify that its product complies with certain FDA-recognized standards or special control guidance documents rather than submitting all of the underlying data supporting substantial equivalence. FDA reportedly expected to receive 1000 abbreviated 510(k) notices annually, but CDRH reports that only 58 abbreviated 510(k) submissions had been received as of March 1999. Of the 58 received, 33 had been reviewed, and only 29 cleared. Apparently, industry is still wary of declaring and certifying conformity to a standard or guidance—possibly because of the regulatory and legal implications of filing a false certification.
As companies gain more experience with abbreviated 510(k)s, it seems likely that industry reliance on this clearance option will increase. For now, however, it appears that only companies whose devices fit squarely within a guidance document or recognized consensus standard will be inclined to adopt this alternative. To avoid missteps, it would be prudent for companies to discuss this option with the appropriate ODE branch chief prior to filing, to determine whether the abbreviated 510(k) is a viable path. By discussing the issue with FDA beforehand, companies can avoid delays in the process and reduce the possibility that FDA will question their declarations of conformity.
Special 510(k)s. The special 510(k) initiative is designed to streamline the clearance process for device modifications by allowing greater reliance on the design control procedures of FDA's quality system regulation (QSR). The special 510(k) allows companies to declare conformity to the design control requirements, thereby relieving ODE reviewers from the need to analyze every aspect of a device modification.
The special 510(k) has had a better track record than the abbreviated 510(k), with more than 222 submissions logged as of March 1999. Of these, 187 had been processed, with 165 receiving clearance. Special 510(k) filings have actually been reviewed and cleared more quickly than traditional 510(k)s: 27 days for the special filing versus 112 days for traditional premarket notifications. The record for abbreviated 510(k)s is not quite so good, with an average total review time of 72 days, but it is still better than the standard 510(k).
The special 510(k) option is one that every device company should consider. There is little downside to using this option. The worst that can happen is that FDA may reject the company's application to use the special route, in which case the company must file the supporting data that would have been required for a traditional 510(k). On the other hand, this route has the potential to substantially decrease the time required for clearance and is much less labor-intensive than a traditional 510(k).
To determine whether this option is the correct one for the modification under consideration, the company should read the published paradigm and ensure that all criteria have been met. When in doubt, a brief call to the ODE branch chief with jurisdiction over the device might be helpful.
Premarket Approval
FDA clearly understands that the PMA process has not been working well for many years. Over the past few years, partly in response to significant congressional and industry criticism, FDA has given high priority to reengineering the Class III approval process.
Product Development Protocols (PDPs). One of the key elements in FDA's reengineering of the Class III approval process is the rebirth of PDPs. The PDP process is a child of the Medical Device Amendments of 1976—a child most in industry considered stillborn. Until this reengineering initiative, not one PDP had ever been approved and completed for a Class III device. Considered an alternative to PMAs at its inception, the PDP process was simply never promoted by FDA.
After much publicity and the issuance of detailed guidance documents on how to navigate the PDP process, FDA had completed only five PDPs as of April 1999. It is still questionable whether PDPs will ever be a viable alternative to PMAs for important new Class III devices.
The PDP process is unlikely to be suited for novel devices whose safety and efficacy cannot be determined by recognized protocols. The process forces FDA and its advisory panels to reach an early agreement about appropriate study designs and endpoints for demonstrating safety and efficacy. Such a concept may work well for devices such as pacemakers and automatic implantable cardioverter defibrillators, where the study designs are well accepted. But for new devices with little literature support or prior approval history, the PDP process is ill suited and will likely delay rather than speed up the approval process.
Again, before jumping too quickly into a PDP, discussion with FDA is highly recommended. The company must have a clear understanding of the process and a well-thought-out regulatory and clinical strategy before embarking on the PDP ship.
Modular PMAs. There is one ODE reform program that has industry support and real potential to speed up approval of important new technologies. CDRH has instituted a program for modular review of PMA applications, and a guidance has been issued that advises industry on how FDA intends to implement the program.7
Under this program, rather than waiting until clinical studies are completed to begin FDA review of a PMA application, various modules of the application can be submitted to ODE for review while the supporting clinical study is nearing completion. For example, the device description module, the report of prior investigations module, and the manufacturing module could be undergoing review while the sponsor finishes the clinical study.
Although this program has the potential to expedite PMA approvals, the jury is still out on how much faster a modular PMA can be approved as compared with a traditional PMA. The modular PMA requires ODE to track each module, leading to somewhat more administrative work on the part of the reviewing division. Also, a number of the ODE divisions are implementing differing, and sometimes confusing, procedures for modular PMA reviews. It is probably fair to say that there are a number of critics of the modular review initiative within ODE.
Nevertheless, this program has full industry support. It is hoped that with the help and coordination of the PMA staff in ODE's program operations office, properly handled modular reviews will expedite the PMA review process which, even with reform, continues to be extremely slow in many cases.
For manufacturers, there is no downside to pursuing a modular review strategy. However, companies should be certain that they are capable of meeting the deadlines for submission of each module agreed upon with FDA and that they can respond promptly to FDA questions on each module. To avoid getting bogged down in a paperwork morass when using the modular process, increased coordination and open and frequent communication with ODE review staff are necessary. On the other hand, the potential for a significantly faster approval time probably makes this option one that should be considered for every PMA.
Less Burden?
In FDAMA, Congress directed FDA to consider, when analyzing clinical data requirements in consultation with the manufacturer, the "least burdensome appropriate means" of evaluating device effectiveness. Similarly, when FDA is considering whether devices with differing technological characteristics are substantially equivalent, the agency may only request information necessary to make that determination, and it shall consider the "least burdensome means" of demonstrating substantial equivalence.8
Notwithstanding these provisions, it is fair to say that there is no clear consensus within CDRH on how to apply the least burdensome standard in 510(k) or PMA reviews. Often, even when less burdensome alternatives are available, FDA's position is that ODE cannot compromise its principles of good science to accept a lesser level of valid scientific evidence. The well-controlled, prospective, double-blinded, randomized trial is still ODE's gold standard.
Lip service is often given to other levels of valid scientific evidence, but there are few instances where FDA has agreed to a lesser proof of safety and effectiveness. It is possible to persuade ODE that randomization or double blinding is not appropriate for a specific study or that a retrospective analysis of clinical data may be valid, but the sponsor will clearly have an uphill fight, especially within certain ODE divisions.
ODE should focus more on the risks and benefits presented by the device. Where a device presents little or no risk, for example, FDA should be more flexible about what it requires as proof of effectiveness. Where there is no device risk, and the effectiveness data are generally good but weak in some areas, FDA should consider approving the device with labeling to advise clinicians about the strengths and weaknesses of the clinical data. Such an approval might then be followed by a postmarket study designed to further evaluate any remaining questions of efficacy. This approach was used to some extent in the recent approvals of the Eclipse Surgical Technologies and PLC Medical Systems lasers for transmyocardial revascularization. But ODE could rely on such an approach much more heavily than it currently does.
It appears that FDA will be publishing a draft guidance document on what constitutes "least burdensome means" sometime in the not-too-distant future. The various device trade associations have been working to persuade FDA that the "least burdensome means" provision should be implemented by ODE in a truly meaningful way. But for now, "least burdensome" means whatever the reviewing division says it means, with very little room for appeal.
Third-Party Reviews
Another FDAMA directive worthy of discussion is the third-party review provision. Partly at Congress's direction, FDA has spent a great deal of time putting into place a mechanism for third-party review of 510(k) notifications.9
FDAMA includes detailed provisions on the accreditation of these third-party reviewers, and the mutual recognition agreements (MRAs) negotiated with the European Union envision third-party reviews of a limited subgroup of 510(k) submissions.
This significant effort by FDA, however, may take quite some time to bear fruit. The agency is reluctant to give up any of its power to approve high-profile (and sometimes high-risk) life-sustaining or life-supporting devices. Thus, under FDAMA, third parties are barred from reviewing Class III devices, Class II devices that are intended to be permanently implanted or are life-sustaining or life-supporting, or Class II devices that require the submission of clinical data for 510(k) clearance (with some exceptions).
Consequently, third-party review at ODE is essentially dead in its tracks, and it is unlikely that this program will ever offer a viable alternative to standard 510(k) reviews. To resuscitate this program and to give it real value, Congress and FDA should open the program to all Class III devices and data-supported Class II 510(k) submissions. Industry would likely flock to such a program and would be willing to pay money to obtain third-party review if faster clearance and approval could be obtained.
Because of the inherent weakness of the third-party review initiative, it is unlikely that this aspect of international harmonization will progress in any meaningful way for years to come. Manufacturers probably will, and should, continue to focus on obtaining appropriate approvals and clearances in their respective countries. There is little incentive, if any, for a manufacturer to pursue third-party review on either side of the Atlantic at this time.
Trends to Come
FDA is now under new management with the ascension of Jane Henney, a former deputy commissioner, to the office of commissioner. CDRH is similarly under new leadership with the appointment of David Feigal as CDRH director.
From a practical standpoint, it is too early to tell how these new leaders will affect device regulatory issues. During her confirmation hearings, Henney promised Congress that she would seek to fully implement the FDAMA reforms as intended by Congress. To date, it does not appear that she intends to go back on that promise, nor does it appear she is poised to change direction significantly from that of previous FDA management.
To a certain extent, Feigal is an unknown when it comes to devices. As deputy director of the Center for Biologics Evaluation and Research (CBER), he was intimately involved in the tissue reference group, which regularly evaluates whether certain tissue products should be regulated as human tissue, devices, or biologic products. He certainly gained a familiarity with device law and regulation through those reviews and through his involvement with numerous products regulated by CBER under the device authorities, such as certain in vitro diagnostics. Industry is hopeful that Feigal will continue the CDRH reengineering and reform efforts and not seek to impose the principles of biologic product review and approval onto the device process.
The new FDA and CDRH managers must face many important issues, which must be dealt with in both the long and short run. With the ever-increasing complexity of medical devices, training and retaining trained personnel are critical. Training resources must be utilized so that CDRH is ahead of, and not behind, the technology learning curve.
CDRH and FDA are also going to have to deal with a new world in which the transfer of information is close to instantaneous. On one level, CDRH has done a splendid job of maintaining Web-site and e-mail communications with industry, an effort that should be a model for the rest of the U.S. government. On another level, in the advertising and promotion area, CDRH and FDA are still struggling to deal with Web-site and Internet communications related to products approved overseas and not in the United States. How the agency deals with this information revolution will be critical to industry marketing and sales practices in the next century.
Conclusion
Despite the valiant efforts of Congress and CDRH leadership to streamline the device clearance and approval process, much remains to be done. Certainly, the lack of resources will continue to plague CDRH as it does the rest of FDA. However, user fees, as proposed by the Clinton administration, are not the answer.
CDRH and ODE have made significant strides by streamlining the review process, empowering reviewers to make decisions, and exempting low-risk devices from premarket review. Even greater accomplishments can be made in the forthcoming millennium through a rational application of the least burdensome means standard, a greater emphasis on risk/benefit analysis in device reviews, and an expansion of the third-party review process. Exploring the potential of these initiatives should enable FDA to make even greater strides while avoiding the imposition of destructive user fees on an industry composed mostly of small and relatively fragile companies.
References
1. H. Rep. No. 43, 105th Cong. 1st sess. 2 (1997).
2. S. Rep. no. 43, 105th Cong. 1st sess. 2 (1997).
3. S. Rep. no. 43, 105th Cong. 1st sess. 2 (1997), p 7.
4. "Guidance on IDE Policies and Procedures" (Rockville, MD: Office of Device Evaluation [ODE], Center for Devices and Radiological Health [CDRH], FDA, 1998).
5. "Early Collaboration Meetings under the FDA Modernization Act (FDAMA), Guidance for Industry and CDRH Staff" (Rockville, MD: CDRH, FDA, 1998) rpt. in FDA Modernization Guidances: Guidance Documents and Federal Register Notices Pertaining to the FDA Modernization Act of 1997 [FDA Modernization Guidances] (Santa Monica, CA: Canon Communications, 1998), pp 21–26.
6. "Guidance on PMA Interactive Procedures for Day-100 Meetings and Subsequent Deficiencies for Use by CDRH and Industry" (Rockville, MD: ODE, CDRH, FDA, 1998), rpt. in FDA Modernization Guidances (Santa Monica, CA: Canon Communications, 1998), pp 155–163.
7. "Guidances for the Medical Device Industry on PMA Shell Development and Modular Review" (Rockville, MD: ODE, CDRH, FDA, 1998).
8. The FDA Modernization Act of 1997 (S.830), Public Law 105–115 [FDAMA], section 205.
9. FDAMA, section 210.
Jonathan S. Kahan is a partner in the law firm of Hogan & Hartson (Washington, DC) and a member of the Medical Device Executive Portfolio editorial advisory board.
