Medical Device Daily Washington Editor
FDA held a webinar yesterday to deal with questions about the series of proposed changes to the 510(k) clearance process, and one development of interest is that FDA might not let industry know of its intent to rescind a 510(k) despite that any devices using it as a predicate may have to come off the market as a consequence of that rescission. Another potentially important development is that according to one member of the management team at the Center for Devices and Radiological Health, data found in patent applications may constitute information a sponsor should “reasonably“ know about, which could be boot-strapped into a requirement to disclose such information to the agency in 510(k) submissions.
Regarding 510(k) rescissions, Christy Foreman, director of the Office of Science and Engineering Labs at CDRH, said “our intent is to revisit that issue and consider additional regulations we need to clarify the process.“ However, she said that FDA has “not yet established whether that would be a public process or a behind-the-scenes process.“
The complicating factor behind this is that a rescission of a 510(k) that has served as a predicate for other applications could force the sponsors of those products to pull their products from the market with little or no warning. Alberto Gutierrez, director of the Office of In-Vitro Diagnostics at CDRH, said later in the call, “when a 510(k) is rescinded, those devices don't have clearance anymore and therefore are being marketed illegally.“
Gutierrez continued: “They have to either come back in with a 510(k) if they think they can provide some kind of update or a way to mitigate whatever problem was found, but as they are, they would be illegally marketed.“
The issue arose again later in the webinar, and Foreman offered more detail. She said, “we also have to look at the basis of the decision“ to rescind, giving the example of a device filing that included fraudulent information. “Many of our rescissions have been related to fraud,“ she commented, adding that a decision to rescind would not necessarily migrate to the downstream devices assuming “those three subsequent devices do perform as intended.“
Any safety issues related to the predicate “would probably go to all those devices that exhibit the same issue,“ Foreman observed, but the unanswered question is that of whether CDRH would give itself time to review data for those downstream devices if it felt the need to rescind the predicate was urgent.
Jonathan Sackner-Bernstein, associate center director of CDRH for post-market operations, said during a discussion of the proposed requirement of more exhaustive data submissions for 510(k)s that “there were often surprises that came up as a product that came up under review“ or in post-market experience, “so by asking for as much data and information as a manufacturer would reasonably be expected to know, we believe we can avoid some of those surprises.“ He gave the example of data found in patent applications.
Sackner-Bernstein later clarified that his comment “was not intended to say we want patent applications submitted. Rather it was pointing out that [such information would appear in] a document that the sponsor had in their possession that disclosed study results that were not made available to us at the time of submission.“ He said, “certainly they would know about studies they performed in support of a patent application. That's what we're presenting as a recommendation in these reports, that this kind of information would come“ under a requirement for full disclosure.
As for the searchable online database for 510(k)s, Jeff Shuren, MD, director of CDRH, said the availability of such data would allow patients, doctors and industry to “see why we made a particular decision“ and that such a database would “also include pictures or a schematic of a device.“ He assured industry, however, “we're not proposing that these will be blueprints . . . but instead a schematic or a photograph of the device.“ The database would also include labeling to help sponsors determine the appropriateness of that device as a predicate.
Regarding how the CDRH action might affect 510(k) applications made to the Center for Biologics Evaluation and Research, Shuren said, “this pertains only to the CDRH program.“
On the subject of a recent study said to have demonstrated that the 510(k) program has not produced a conspicuous number of failed devices, Shuren said the study “tried to look at one particular measure,“ namely recalls, which he characterized as “a limited safety measure. It's helpful,“ he said, but remarked, “we're seeing that the program is becoming challenged.“
“The wheels are starting to rattle on the cart,“ Shuren claimed, adding that while the percentage of recalls “might not be a very high number,“ the question for FDA is “have we done a perfect job on safety? No.“
Shuren also said that changes to the third-party 510(k) review program are intended “to make optimal use of the third-party program“ and to address the fact that “the quality of the third-party reviews was highly variable.“ He said that almost half the applications that went through a third party ended up with an FDA review to obtain additional information. The idea here, he said, is to more clearly determine which technologies should and should not go through third parties.
Mark McCarty, 703-268-5690