Biocompatibility Testing - Component Level or Assembled Medical Device?

[kiwi.kczt]

Hi Covers, I've been struggling to find a right answer for this question.

We are manufacturing a Class IIa non-sterile device. The device is made up of approximately 6 components, out of which 2 are patient-contacting. To make the complete device, these components are assembled with medical UV-cure adhesive. The cured adhesive would not come in direct contact with the patient.

For biocompatibility testing, do we send in the patient-contacting components in the final finished form (injection and compression molded), or do we send the assembled device, from which the test lab would remove the required components for testing?

Thanks in advance!

Keith

 

[GStough]

Hi Covers, I've been struggling to find a right answer for this question.

We are manufacturing a Class IIa non-sterile device. The device is made up of approximately 6 components, out of which 2 are patient-contacting. To make the complete device, these components are assembled with medical UV-cure adhesive. The cured adhesive would not come in direct contact with the patient.

For biocompatibility testing, do we send in the patient-contacting components in the final finished form (injection and compression molded), or do we send the assembled device, from which the test lab would remove the required components for testing?

Thanks in advance!

Keith

Hi Keith,

Just curious....would it be feasible to send both versions to ensure you've covered everything?

 

[yodon]

I would suggest you talk to the lab and get their inputs. My understanding is that they take your thing and soak it in various solutions so if you need to exclude the non-patient-contacting parts, you'd have to have some kind of strategy and they should be able to work with you on that. The test labs require that you send production components since production processes may have an effect. Oh, and there are requirements for the amount of material provided. Again, they should be able to work with you on all that.

 

[Mark Meer]

In my view you have 2 options:

1. Send the 2 patient-contacting materials (without the UV-cure adhesive), and then document a justification as to why the presence of the UV-cure adhesive does not affect the results of the testing. This may be difficult to do, unless there is some precedent (e.g. research, comparable devices on the market) or some other evidence you can point to.

2. Either send them the components with the UV-cure adhesive attached and instruct them to remove it prior to testing, or remove it yourself prior to sending to the test lab and document somewhere that this was done in the samples sent to the lab.

I also agree with yodon that it's probably a good idea to ask the lab's input.

MM.

 

[Ronen E]

Assuming that we are talking about ISO 10993-1:2009 (and derivatives) -

For some reason most people seem to jump right to the test table and the tests they think apply, then dive into the implementation details, while the standard speaks a lot and in detail about "an evaluation" (not necessarily testing) and the risk management process context. Sadly this seems to be aligned with current test labs practice - they seem just too happy to get on to testing with minimally "ticking the boxes" that they have to according to the standard. It's sad because their original role in the standard is one of thought, analysis, insight and guidance.

Whatever testing is eventually carried out, it should be based on a thorough and wide analysis and consideration of all relevant available information. This analysis should be carried out by knowledgeable, qualified and experienced experts (the most trivial interpretation of which being the lab's expert toxicologists). As part of this process all the above issues and options should be taken into account and the optimal path forward should be recommended.

All the above should be then elaborated in the biological evaluation report.

 

[Mark Meer]

For some reason most people seem to jump right to the test table and the tests they think apply, then dive into the implementation details, while the standard speaks a lot and in detail about "an evaluation" (not necessarily testing) and the risk management process context.

I agree. But, similar to the way that "voluntary" standards are pretty much required at this point, the reason companies adopt this approach is that it's the surest path to passing regulatory hurdles without too much headache or uncertainty.

If a Notified Body or FDA 510(k) reviewer sees that you applied the table in Annex A of ISO 10993-1 to determine the appropriate tests, and executed them, this is a much more certain approach than justifying why you didn't apply the tests.

I wish it were otherwise, but given the uncertainty of what individual auditors/reviewers will accept as justification, just doing the tests is often the simplest path.

 

[Marcelo]

I wish it were otherwise, but given the uncertainty of what individual auditors/reviewers will accept as justification, just doing the tests is often the simplest path.

Until it's not enough and then another problem is created. I do understand why so many manufacturers prefer to "let's pass the audit"instead of "let's do the right thing" approach - and most even think that "pass the audit" means they are doing the right thing due to lack of knowledge, but 99% of the time it's easier to to the right thing (and it will be always in conformity, while the "pass the audit" solution may pass one or more time, but it may also not).

 

[Ronen E]

this is a much more certain approach than justifying why you didn't apply the tests.

The standard doesn't speak of "justifying why you didn't test". It speaks of collating and analysing all the relevant available information to reach relevant conclusions about biological safety. When this is not done, why shouldn't a reviewer object?...

Indsutry should not complain - it consistently seeks the minimal effort path (and below) so no wonder regulators adopt a mechanistic (rather than a thoughtful) approach. This trend goes well beyond biocompatibility.