One of the questions I get asked a lot is whether a client needs to hold up their 510(k) or PMA until they have all the reagent stability data they want (usually 12 to 24 months). The associated question is whether FDA will accept accelerated stability data in an application.
The answer to the first question varies a bit depending on which part of the Agency the application is being reviewed by and what type of stability data we’re talking about.
For open-vial stability studies it’s important to have all of your data in your application. CBER will ask for open-vial stability studies to be performed at the end of the claimed reagent shelf-life. For example, if the claim is that a reagent can sit on the shelf for 12 months, than CBER wants to see open-vial data on a kit that’s at least a year old. I haven’t had that request from CDRH yet. Since open-vial studies typically only look at a relatively short period of time (a few weeks at most) these studies usually don’t present too much of an issue.
Expiration dating on kits needs to be based on a well-controlled, real-time stability study. Studies typically include at least three lots of reagents. These studies need to select reasonable time points for testing that will demonstrate product stability over time.
The time between testing points is a choice to be made for each product. Choosing longer time points carries a risk. The expiration dating will be based on the last time point the product met specifications. If that earlier time point was say, 4 months, than the expiration date moves all the way back to four months (actually three since it’s typical to short dating by a month to ensure product performance) since there is no way to demonstrate where the product stopped meeting specifications.
In my experience, FDA has not accepted accelerated data in lieu of real-time data. They will review accelerated data if it is submitted but I’ve found that it creates more questions than it answers and I typically recommend clients not include it in their applications. Accelerated stability data may be helpful for development or project teams to anticipate risks but it should not be considered a substitute for real-time studies.
If the studies are at the six-month point but you ultimately want 18 months data, it will be necessary to submit the existing data as well as the protocol used to generate the data. FDA will review both and assuming they accept the studies as appropriately rigorous, dating can be updated without a new application or input from the FDA.