What is the expectation for Biocompatibility Testing when switching Suppliers?

[rwend07]

What is the expectation for biocompatibility testing when switching suppliers? If you switch suppliers of a silicone component (Class VI --> Class VI) that was part of a biocomp test, would it be expected to re-do the testing to verify that the newly sourced component meets previous test results? This would seem like an extremely costly requirement if testing needed to be re-done for every change in material.

 

[JeantheBigone]

Because FDA generally wants to see testing on final finished product, my initial reaction is that if the manufacturing process is not identical, meaning the exact same raw material from the exact same source is used to produce the exact same component using the exact same equipment, you'll need new biocompatibility testing.

However, in making changes of any kind to a device,it's about performing a risk analysis. What does the original 10993-1 say? Was biocompatibility testing required and why? In other words, I would start with ISO 10993-1.

You may also wish to consult
http://www.fda.gov/ucm/groups/fdagov-public/@fdagov-meddev-gen/documents/document/ucm514771.pdf

 

[Marcelo]

Testing is an output of the biological evaluation/risk analysis process. You will ned to perform the process again. See also ISO 10993-1, ISO TR 15499 and the document "Use of International Standard ISO 10993-1, "Biological evaluation of medical devices - Part 1: Evaluation and testing within a risk management process".

 

[rwend07]

Because FDA generally wants to see testing on final finished product, my initial reaction is that if the manufacturing process is not identical, meaning the exact same raw material from the exact same source is used to produce the exact same component using the exact same equipment, you'll need new biocompatibility testing.

However, in making changes of any kind to a device,it's about performing a risk analysis. What does the original 10993-1 say? Was biocompatibility testing required and why? In other words, I would start with ISO 10993-1.

You may also wish to consult
http://www.fda.gov/ucm/groups/fdagov-public/@fdagov-meddev-gen/documents/document/ucm514771.pdf

Thank you both for the responses. Interestingly, after re-reviewing the "When to submit 510(k)" guidance, it appears to point towards not needing to re-do the assessment if a risk analysis determines there is no new concern for biocompatibility. See C4 below:

"The answer to C4 may be no if a knowledgeable individual reviews the differences in chemical composition or physical properties and determines that the change is minor enough that there is no new concern about biocompatibility"

If there is no additive difference between the two suppliers and the material is Class VI at both, I would argue the new supplier does not raise any new concern about biocompatibility. Am I reading this wrong?

 

[Marcelo]

"The answer to C4 may be no if a knowledgeable individual reviews the differences in chemical composition or physical properties and determines that the change is minor enough that there is no new concern about biocompatibility"

If there is no additive difference between the two suppliers and the material is Class VI at both, I would argue the new supplier does not raise any new concern about biocompatibility. Am I reading this wrong?

Please note that the document "Deciding When to Submit a 510(k) for a Change to an Existing Device"is focused on the need to submit a new 510k (as per the name), not if a new/revised biological evaluation is needed. The need for a new/revised biological evaluation is discussed on the three documents I mentioned.

 

[JeantheBigone]

I hope I didn't confuse anyone with the link to the Guidance Document for when to submit a new 510(k). My point was that if you want to be very cautious, you may wish to review that document and confirm that no new 510(k) is needed.

 

[rwend07]

I hope I didn't confuse anyone with the link to the Guidance Document for when to submit a new 510(k). My point was that if you want to be very cautious, you may wish to review that document and confirm that no new 510(k) is needed.

I understood your use in including the document. I still feel like the statement I included implies that a risk assessment could conclude that no new biocompatibility testing would be needed for a 1:1 supplier change, and I haven't seen anything in 10993 or its FDA guidance that directly says that a change in supplier = new biocomp testing. For example, from the FDA guidance on 10993:

"Another common change that might impact biocompatibility is a change in resin supplier. For example, if the new resin supplier does not remove all processing solvents (some of which may be known toxic compounds, such as formaldehyde), the final manufactured device could cause unexpected toxicities (e.g., cytotoxicity, irritation, sensitization, genotoxicity) that were not seen with devices manufactured from the original resin."

Key word being "might". If both suppliers have a USP Class VI cert, wouldn't that imply that any toxic compounds left over after manufacture are equivalent and would not be expected to cause a change in toxicities? Also, couldn't you verify through a master file that there are not any differences in processing solvents that would cause new concerns for biocompatiblity?

Thanks again for both your help on this. I know that generally it has been standard to re-do biocomp testing, but I haven't seen a direct link to anything that says it is necessary even if there is a valid risk assessment that states there are no new questions of toxicities. Working with startups, re-doing biocompatibility testing can be 100% cost prohibitive to a simple 1:1 supply switch so it is quite frustrating that it is expected.

 

[Marcelo]

I haven't seen anything in 10993 or its FDA guidance that directly says that a change in supplier = new biocomp testing

As I mentioned before, testing is an output of the biological evaluation process per ISO 10993. You might not need testing if you conclude that testing is not needed after applying the biological evaluation process together with the risk management process as per ISO 10993-1.

Key word being "might". If both suppliers have a USP Class VI cert, wouldn't that imply that any toxic compounds left over after manufacture are equivalent and would not be expected to cause a change in toxicities? Also, couldn't you verify through a master file that there are not any differences in processing solvents that would cause new concerns for biocompatiblity?

This is a questions that needs to be evaluated during the biological evaluation. Also, you don't "imply" things, you conclude after an analysis.

has been standard to re-do biocomp testing

Because a lot of people don't understand/use erroneously the biological evaluation process of ISO 10993. People take a look at the table in Annex 1 and concludes that testing needs to be done. The current version of the standard was written explicitly to make it clear that the table is the output of the analysis, so you "might"need to test/retest depending on the analysis, not the other way around.

 

[Ronen E]

Good discussion!

Sorry for coming late to the party, but I think I can make the two ends meet...

The best / most qualified way to run the biological risk evaluation process as per 10993 (in non-trivial cases) is with the support of a qualified and experienced toxicologist. The accredited test houses always have them on board, so although not cheap I would recommend engaging with such a test house for that purpose. That doesn't necessarily mean any testing will have to be done. If indeed the analysis output will be no-testing, this route will cost just a fraction of the typical biocomp testing engagement.

I believe that the "knowledgeable individual" that the guidance referred to was that toxicologist...

The analysis may include examination of master files (if available - not so easy to get formulation info from suppliers) and other documentation, and/or "equivalence testing" which is aimed at showing that the materials from the 2 different sources are equivalent from a biological reaction perspective. Such testing is usually simpler, cheaper and quicker than the full-blast biocomp test set (if required).

To wrap up: I agree that the starting point is re-evaluating according to 10993 & FDA guidance. I also agree that it may well end up in a paper risk evaluation. The important bit is to get a qualified and experienced toxicologist involved in that evaluation, so as to give it validity and hopefully reduce actual re-testing to the minimum.

Side note: USP class VI (and that entire classification, for that matter) is outdated in the context of biological evaluation for medical devices. It was phased out when the FDA endorsed ISO 10993, years and years ago. I'm not sure why it stayed around for so long, but it seems to occasionally show up when medical devices design is being discussed.

 

[Marcelo]

I believe that the "knowledgeable individual" that the guidance referred to was that toxicologist...

Yes, it is a toxicologist. And you got the point, that's in fact a big part of the problem, because usually manufacturers do not have a toxicologist on board and do no contract one to do the biological evaluation.