The FDA device center's upcoming proposals for reforming its pre-market policies will directly address widespread industry concern that reviews are increasingly unpredictable and opaque, CDRH Director Jeff Shuren told an industry audience last week.
CDRH Pre-Market Reform Proposals Expected Within Weeks
Article preview from "The Gray Sheet" - June 28, 2010
The FDA device center's upcoming proposals for reforming its pre-market policies will directly address widespread industry concern that reviews are increasingly unpredictable and opaque, CDRH Director Jeff Shuren told an industry audience last week.
Two reports, on 510(k) reform and on making better use of science in regulatory decision-making, will be released within the next several weeks, Shuren affirmed June 22 at a CDRH "Town Hall" meeting in Woburn, Mass.
The reports, to be issued simultaneously, are the source of significant anxiety among device companies. But Shuren stressed that industry should view the initiatives as an opportunity for improvement of currently imperfect review processes.
"In terms of what are we doing on the issue of consistency in decision-making, you are going to see what our findings and recommendations are in these reports that will be issued in the coming weeks," Shuren said.
"If we held this meeting in a few weeks, we would probably be talking about a bunch of other things," he added.
As it was, industry speakers at the town hall unleashed a litany of complaints, relaying tales of last-minute, unexpected data requests from reviewers in the midst of 510(k) reviews, experiences with agency staffers who appeared unwilling to meet face to face, and a general lack of transparency about evolving review standards.
Company reps portrayed the matters as problems that have reached a fever pitch during the past year or two.
"In my experience, FDA reviewers have always been challenging and thorough," said Maha Sallam, president of neuro-imaging device firm VuEssence. However, Sallam added, "in recent years, the questions raised by the agency have become epidemic and disproportionate to the submission in hand."
Shuren acknowledged problems with the review process, including instances of individual reviewers' scientific priorities holding up a submission more than necessary. He implored companies to elevate complaints to branch chiefs, division heads or other supervisors if they feel a reviewer is not handling an application appropriately, noting that the center has recently updated its internal protocol for resolving disputes.
"Sometimes companies are fearful to do that, because they feel like reviewers will take it out on them," he said. However, "if you do not raise it up through the ranks, people will not necessarily know about it. As a result, you may be allowing staff to have more control ... than may be appropriate."
Shuren said he is putting increased emphasis on reviewer training within the center and making it clear to managers that they need to take more control of their reviewers when necessary. Further, he is changing the structure of the organization when needed.
But Shuren also noted that the technologies reviewed under the 510(k) process are rapidly growing in complexity, causing staffers to struggle to come up with suitable review strategies without asking for a full PMA application.
"The question is, is the current 510(k) program appropriately designed to account for all these evolutions and revolutions in science that we are seeing?" Shuren told reporters following the meeting.
Though not providing many details, he said the upcoming draft reports would explain how CDRH hopes to make 510(k) and other pre-market requirements more predictable so that surprises during the review process are minimized.
New Standards Can Mean Process Complications
One underlying issue that will be addressed, particularly in the report on science and regulatory decision making, is how to implement new review standards in a predictable and transparent manner.
Several industry speakers complained of receiving specific advice on clinical trial endpoints during non-binding pre-submission meetings with the agency, and then being told late in the review process that the standards had changed.
"To now move the goal posts at the end of the game ... simply because opinions change within the FDA about optimal study design, is extraordinary," argued Eric Bornstein, chief scientific officer for Nomir Medical Technologies, in describing his firm's frustrations with the 510(k) clearance process for a laser-based toe fungus treatment.
Shuren stressed the steep challenge of implementing changes based on new science in a manner that best protects patient safety while not discouraging innovation.
"What do you do with technology that is under development? What do you do with technology you just approved two days ago? These are fundamental questions" that CDRH is trying to answer more satisfactorily, Shuren said.
510(k) Predicates And De Novos
More specifically, the 510(k) report will try to tackle the predicate device system and CDRH's process for allowing companies to rely on certain marketed devices as comparators for substantial equivalence determinations.
Multiple speakers highlighted inconsistencies with CDRH's approach to predicates. Reviewers are unfairly restricting companies from relying on certain predicates without providing a clear and transparent rationale, they argued.
"We have experienced the situation where FDA reviewers take us aside and say, 'we are not letting you compare to that predicate anymore," said Chas Burr, graduate program director for health products regulation at Regis College. "The predicate remains on the market, remains legally marketed, which means per law we can compare to it. And yet we are being told, in the corridors, by reviewers, 'you can't do that.'"
Arthur Rankis, a device consultant, provided several specific examples, including instrument sets sold for orthopedic implant procedures, which are re-sterilized and reused by hospitals, as per labeling.
The "common" sterilization cycle lasts four minutes, according to Rankis. Yet, "There are many sets in the marketplace [for] which a bio-sterilization cycle takes longer," he noted.
"If you submit a 510(k) today with a cycle of more than four minutes, you will be blocked," he asserted. "The fact that there are many predicate devices on the market makes no difference."
He charged that "FDA needs to be aware that their policies and decisions result in very significant competitive advantages to some companies and undermine the commercial viability of others."
FDA officials have previously suggested alternatives for improving the predicate process, including more explicitly restricting use of outdated predicates and publishing more comprehensive review synopses for 510(k)-cleared products as a model for future submissions.
Shuren affirmed at the meeting that CDRH is working on posting detailed 510(k) 3 summaries online for all device categories in a manner similar to that of the in vitro diagnostics office.
But some industry reps suggested that the underlying process of relying on historical precedent to determine pre-market submission requirements for a new product is inherently flawed.
"Innovative devices should be regulated based on the actual levels of risk they present and not on whether they fall into a risk category based on legal precedent," said Paul LaViolette, a venture partner at SV LifeSciences and former chief operating officer at Boston Scientific.
Burr described the need for a long-term shift from "history-based" risk classification to "rule-based" classification, where a series of product characteristics (e.g., degree of invasiveness, whether it has local or systemic effects, etc.) are considered according to pre-defined classification rules for each new device submission. This would follow the model of the Global Harmonization Task Force's approach, summarized in a 2006 4 guidance document.
The current law "is not providing the philosophic basis, based on risk, that would allow true flexibility and transparency," Burr maintained.
Another issue that will be dealt with in the 510(k) report is the de novo review process, Shuren affirmed.
On paper, the de novo process offers some of the flexibility that companies are seeking, as it provides CDRH a route to require less data for low-to-moderate-risk products even if no appropriate predicate exists.
However, industry says de novo reviews have become increasingly unpalatable.
Zvi Laden, a principle with Boston MedTech Advisors, presented results from his group's recent analysis of de novo submissions.
According to the 5 study, there were 54 de novo approvals from 1998 through 2009. The average review duration for the 41 applications through 2006 was 240 days. But, for the 13 applications approved since 2007, the average duration was 482 days.
"Such a long period significantly exceeds FDA's commitment for the duration of the review process of even most panel-track, pre-market approval devices," Ladin said.
Easier Changes Will Take Effect This Year
Shuren stressed that no changes to device review processes would take place until the public has a chance to comment on the upcoming drafts.
The less controversial proposals, based on public feedback, will be implemented sooner, before the end of 2010, he explained. "The other recommendations that are very controversial, and would benefit from additional input - we will put that over to the Institute of Medicine," Shuren said.
IOM, which is performing a parallel review of the 510(k) program, will issue its report in 2011.
During a press conference, Shuren would not venture to say which issues are likely to be dealt with this year, and which would be passed to IOM. "We have not even put out what we are going to do," he noted. "The only thing I am very clear about [is] we are not recommending to get rid of the 510(k) program."
- David Filmore
Purchase this article online as a PDF and receive it immediately via email. Questions? Call (800) 332-2181. 100% Satisfaction Guaranteed.
Sign up for a complimentary 30-day trial of "The Gray Sheet" There is no obligation and no credit card needed. Click here to sign up and get your first issue.Plus:
To find out about more about more about Elsevier Business Intelligence's medical device publications and databases, multi-user access and/or advertising with Medical Devices Today, please contact Kristy Kennedy at (480) 985-9512
Comments