Tom Waits once sang of a character named Small Change, who apparently "got rained on with his own .38," and device makers might sometimes have felt as though small changes to their devices didn't always work out so well, either.
Waits is still among us, however, and the final 510(k) changes guidance is out. This effort to overwrite the K97 memo was no mean feat, although it was perhaps predictable that it would be a tough slog. After all, this guidance is easily one of the five most important guidances where FDA's regulation of medical devices is concerned, if only because it affects thousands of applications each year.
So what's the net effect of this guidance? Let's take a look.
To combo ... or not
One of the wins for industry was that the agency acknowledged that the Quality Systems Regulations were adequate to control for some changes, a proposition that drove quite a bit of debate in connection with the 2011 attempt to rewrite the K97. Just how the provisions of this guidance will apply to combination products is not clear. The 510(k) changes final says its terms don't necessarily answer all the questions that may be unique to combo products, but that the general principles do apply.
It's easy to imagine that combo products won't fare as well under this guidance. After all, a change to the design of the plunger in a pre-filled syringe might seem to someone at the agency as changing how much of the drug in that syringe is delivered, no matter how trivial that design change might be. On the other hand, combo products really are a different genus, and only time will tell if there's an unjustified difference in treatment.
See your dromedary, raise you a jumbo jet
To revisit one of the more amusing episodes revolving around the 510(k) program, those who were watching all this back in 2011 no doubt remember the report by the Institute of Medicine regarding the program. This multi-million dollar study fed a remark by one IOM member that a device maker could use the program to turn a camel into a Boeing 747, but like the camel-to-jumbo jet wisecrack, the IOM report proved to be too large a hump to get over, in part because it offered no alternative to the 510(k) program.
Nonetheless, the IOM got one thing right: The report said the concept of the predicate device was pretty much indispensable to any replacement for the 510(k) program, which seems to suggest that if you did the substantial equivalence thing all over again, you'd end up with something that looks a lot like the 510(k) program. So much for the benefits of brain trusts.
Around that time, though, there was this business of mashing together the terms "intended use" and "indications for use," a proposal the FDA floated in response to purported confusion. There may have some disingenuous arguments by device makers about the respective roles of the two terms, but you don't have to be Albert Einstein to know the difference (although someone tried to sell a webinar on that point last year). One suspects this discussion was less about definitions and more about circumstances in which a change in, or an addition to, the indications for use seemed to stray outside a device's intended use.
Thankfully, the final guidance laid this intended-use/indications-for-use imbroglio to rest (although the 2011 draft took a run at the same problem). Still, device makers might want to make note of the passage in the guidance in which the FDA says that changes to a device's indications for use give the agency a bigger case of heartburn than anything else in connection with device labels.
Hold 'em or fold 'em?
The outcome of the indication-for-use problem is that any significant rewrite of or addition to the existing indication(s) for use would likely trigger a need for a new 510(k) application, but a sponsor's rescission of one indication for use from a label with more than one indication will not trigger such a need. Are we suspicious if we see this latter provision as a concession the agency planned when writing the 2016 draft?
After all, the removal of an indication for use might mean nothing more than that nobody is using the device for that indication. Alternately, if the device maker is yanking that indication because of adverse events, those events would presumably be know to both the manufacturer and the agency, in which case one has to assume the FDA would be fine with the removal of that indication.
Other than the bargaining chip theory, the only explanation I can think of for requiring a new 510(k) in this scenario is that the FDA was worried a device maker might pull an indication to avoid a medical device reporting mess, or worse. But that sounds like regulating to the lowest common denominator, never a popular move outside of Silver Spring, Md.
There are other items in the final guidance that confirm the agency is not as militant as it was six years ago, including that the device maker has to compare its new device configuration only to the immediate antecedent. The software 510(k) changes guidance also concedes what seemed a major sticking point in the draft in that the agency said the addition of a security patch does not automatically trigger a need for a new 510(k).
The problem with that concession is that anyone who read the software draft guidance knew that idea was a dead duck, too, so we again have this question of whether this was anything more than a bargaining chip. Still, it's a good day for the 510(k) program, particularly since things in the European Union – including the handling of their version of the substantial equivalence debate – are in a real messy state right now, leading one to conclude there might be a lot more class II device applications finding their way to the FDA over these next few years.