GLP (Good Laboratory Practices) Requirement for Sterility Validations

[MarkinCalifornia]

I am trying to determine when the FDA requires testing to be performed using Good Laboratory Practices (GLP). Specifically, does the FDA require GLP for sterilization validation studies of medical devices? There is some debate within my organization that the scope of 21 CFR Part 58 does not include sterilization validation because the last sentence of scope includes “safety data” and some believe that the sterilization validation, in general, is not safety related (see regulation reprinted below). Practically speaking, when looking at the delta between what you get with and without GLP, there is not anything there that we haven't done anyway (e.g., protocol review and approval, data review) with the exception of the observation of the lab work, which in our case is limited to sterility testing and the counting of samples submitted. On the other hand, one can also make the argument that sterilization validation is safety related (e.g., high product bioburden can impact patient safety) and therefore the GLP requirement is applicable.

I would certainly appreciate others experience on when GLP is required. It would be particularly helpful to know if anyone has obtained FDA approvals on medical devices without a fully compliant GLP sterility validation. Obviously, the conservative approach would be do the study under GLP. However, I don't want to waste company time and resources if it really isn't needed.

Thanks for your help.

Scope of 21 CFR Part 58:
§ 58.1 Scope.
(a) This part prescribes good laboratory practices for conducting nonclinical laboratory studies that support or are intended to support applications for research or marketing permits for products regulated by the Food and Drug Administration, including food and color additives, animal food additives, human and animal drugs, medical devices for human use, biological products, and electronic products. Compliance with this part is intended to assure the quality and integrity of the safety data filed pursuant to sections 406, 408, 409, 502, 503, 505, 506, 510, 512–516, 518–520, 721, and 801 of the Federal Food, Drug, and Cosmetic Act and sections 351 and 354–360F of the Public Health Service Act

 

[Sam Lazzara]

In my experience, I have never performed lab testing related to sterilization validation per GLPs. And in my experience, which includes mostly 510(k) submissions, FDA has never questioned it. Perhaps it is because the microbiology test methods used are typically standard USP (compendial) methods and the submissions are not PMA submissions. I would be interested to hear if anyone has heard of FDA questioning the GLP-ness of this type of testing to support IDE/PMA submissions.

If you want to get a test lab's perspective (with lots more experience than me) I suggest calling Nelson Labs in Salt Lake City, UT.

ASIDE:

For biocompatibility testing that will support regulatory submissions (US, Europe, wherever) I always recommend going with GLPs since some agencies will likely require it, or at least ask fewer questions. This will pay off in the long run.

I have many times, after IDE submissions, seen FDA ask questions about the GLP-ness of (non-biocompatibility) highly customized animal studies performed to show product safety/performance. In most cases, we explained that the studies were performed in the spirit of GLPs (study director, pre-approved protocols, trained personnel, calibrated equipment, raw data review, etc.) and explained what we were missing that wouild be required for a full GLP study and claimed these deviations did not impact the validity of the study results. More recently I am hearing rumors that FDA may not be accepting these rationales. If you go this route, it is a good idea to place a GLP Statement in the protocol so the rationale is provided before the study commences.

So far, for engineering non-clinical bench testing to support IDE submissions, I have not yet seen FDA ask about GLPs (hmmmmm...maybe once). In any case, the same "spirit of GLPs" argument can perhaps be employed.