Should Biocompatibility Tests be performed on Production Samples?

[DDsystems]

Hello Everyone,

I finally was able to submit my 510(k) for my class II device to a third party reviewer about a month ago. I have since received their comments back about the deficiencies. Their main point was the biocompatibility of my device. My device is a very basic device that sits on a patient's head just like swimming goggles. They said that we have to make the complete set of molds and then assemble a production piece and then send it off for biocompatibility testing. I find this hard to believe because how do companies create full production injection molds if it has not been cleared? The material I am using is Dupont Hytrel 7246 which is a well-known material and has all the safety data attached.

My question is how do I approach this? Do I need to create the molds for everything and then create parts and then send them off for biocompatibility testing? Or do I just show them that it is the SOP's that it will be tested and be required to meet a certain standard when in full production?

Does anyone have experience with this? We are a very small startup company trying to make it in the medical device world but if we have to make molds and something needs to be changed it will put us out of business.

Thanks!!

 

[Ronen E]

Hello Everyone,

I finally was able to submit my 510(k) for my class II device to a third party reviewer about a month ago. I have since received their comments back about the deficiencies. Their main point was the biocompatibility of my device. My device is a very basic device that sits on a patient's head just like swimming goggles. They said that we have to make the complete set of molds and then assemble a production piece and then send it off for biocompatibility testing. I find this hard to believe because how do companies create full production injection molds if it has not been cleared? The material I am using is Dupont Hytrel 7246 which is a well-known material and has all the safety data attached.

My question is how do I approach this? Do I need to create the molds for everything and then create parts and then send them off for biocompatibility testing? Or do I just show them that it is the SOP's that it will be tested and be required to meet a certain standard when in full production?

Does anyone have experience with this? We are a very small startup company trying to make it in the medical device world but if we have to make molds and something needs to be changed it will put us out of business.

Thanks!!

Hi,

Biocompatibility testing is one aspect of biological evaluation. Not all cases require testing, and the extent of testing required is also debateable. To streamline a submission, a documented analysis by a qualified and experienced toxicologist would be desireable. That service is offered by the larger test houses, however I don't know whether or not you've allowed for it in your budget.

Assuming that tests are required, they're supposed to be representative of the commercial device version. That means, first and foremost, the same manufacturing technologies. If the device is to be made of some injection moulded parts you have to test injection moulded parts; data covering the raw material (eg pellets) will not do, and the same can be expected for machined parts (even if they're from the same base material). On the other hand, I would argue that tested units don't necessarily have to come from the full-scale production tool; it should suffice that the pilot / small scale tool works on a similar technology, is made of similar materials and employs similar process settings (temperatures, pressures etc.). To that end one has to have access to the actual technical details to be able to make a defendable argument.

Good luck,
Ronen.

 

[yodon]

Ronen is spot on as usual.

Take a look at this guidance: https://www.fda.gov/downloads/Medic...onandGuidance/GuidanceDocuments/UCM348890.pdf

*In theory* you can possibly justify not doing testing if you can assert equivalence.

 

[rwend07]

Hi,

Biocompatibility testing is one aspect of biological evaluation. Not all cases require testing, and the extent of testing required is also debateable. To streamline a submission, a documented analysis by a qualified and experienced toxicologist would be desireable. That service is offered by the larger test houses, however I don't know whether or not you've allowed for it in your budget.

Assuming that tests are required, they're supposed to be representative of the commercial device version. That means, first and foremost, the same manufacturing technologies. If the device is to be made of some injection moulded parts you have to test injection moulded parts; data covering the raw material (eg pellets) will not do, and the same can be expected for machined parts (even if they're from the same base material). On the other hand, I would argue that tested units don't necessarily have to come from the full-scale production tool; it should suffice that the pilot / small scale tool works on a similar technology, is made of similar materials and employs similar process settings (temperatures, pressures etc.). To that end one has to have access to the actual technical details to be able to make a defendable argument.

Good luck,
Ronen.

To add to this, if you can find a molder that will make a cheaper "soft" mold that will simulate a full production tool, you may be able to get the molded parts much cheaper than needing a full production tool. With that said, something that I have never considered previously is if an aluminum mold would produce "bio-similar?" parts that could stand in for an eventual steel tooled part. Ronen, do you have any experience with this? I would guess that pressures/temperatures would likely be different between aluminum/steel tools but I have never looked closely at this issue.

 

[rwend07]

To add to this, if you can find a molder that will make a cheaper "soft" mold that will simulate a full production tool, you may be able to get the molded parts much cheaper than needing a full production tool. With that said, something that I have never considered previously is if an aluminum mold would produce "bio-similar?" parts that could stand in for an eventual steel tooled part. Ronen, do you have any experience with this? I would guess that pressures/temperatures would likely be different between aluminum/steel tools but I have never looked closely at this issue.

Just to follow up on this, you would want to find a molder that can do both the "soft" and "production" tools so they can use the same manufacturing processes on both sets of units.

 

[Ronen E]

With that said, something that I have never considered previously is if an aluminum mold would produce "bio-similar?" parts that could stand in for an eventual steel tooled part. Ronen, do you have any experience with this? I would guess that pressures/temperatures would likely be different between aluminum/steel tools but I have never looked closely at this issue.

At the macro level temperatures and pressures should be similar because they are dictated by the raw material grade and part design. At the micro level they might be somewhat different, primarily because aluminium has different heat properties from steel and additionally the cavity feeding system, maybe even gate location, type and size, may be different. Either way, that wouldn't be my main concern in that context. I wrote:

it should suffice that the pilot / small scale tool works on a similar technology, is made of similar materials and employs similar process settings

(I added the bold type emphasis now)

From a biocompatibility perspective, to do what you suggest you must be able to argue that there's no significant interaction between the cavity material and the polymer, or that the nature and magnitude of such interaction is not significantly different between an aluminium tool and a steel one. I never tried such an argument.

Pilot tools may be much cheaper than full scale production tools through means other than build materials, or more generally - through elements that clearly don't affect biocompatibility. One of the more obvious is single cavity vs. multi-cavity (though a very pedantic reviewer could argue that melt behaviour, and thus finished part properties might be altered). Another is automation - for example, production tools usually call for automatic ejection, while for limited volumes semi-automatic or fully manual ejection tools can be built at a significantly reduced cost.

Low cost tool makers are usually experts in tool cost reduction and could offer a lot of ideas if asked to minimise the total tool cost. The client's job would then be to evaluate such solutions and ensure that they don't disqualify the resulting parts for biocompatibility tests (if required) and other product validation purposes.

 

[Ronen E]

Just to follow up on this, you would want to find a molder that can do both the "soft" and "production" tools so they can use the same manufacturing processes on both sets of units.

I would argue to the contrary. If it's a matter of mere modest scale-up (say, move from 2 cavities to 4 or 8), it would make sense to stick wth the same tool maker. However, if we're talking mass production where final finished parts need to be extra-cheap and tools need to be high-end 32- or 64-cavity precision machines, and the goal of the pilot tool is minimum cost (to keep a start-up company breathing), and usually short build lead-time, these are two very different types of expertise. I would look for a separate expert for each of those tasks, and would insist on receiving good documentation of process parameters in the first phase so that I could provide that input to the moulder on the 2nd phase.

I do strongly recommend looking for tool makers who are also capable of moulding in-house (at least for break-in and fine-tuning purposes) or willing and able to take responsibility for moulding at a nearby partner that they know well.

 

[rwend07]

I would argue to the contrary. If it's a matter of mere modest scale-up (say, move from 2 cavities to 4 or 8), it would make sense to stick wth the same tool maker. However, if we're talking mass production where final finished parts need to be extra-cheap and tools need to be high-end 32- or 64-cavity precision machines, and the goal of the pilot tool is minimum cost (to keep a start-up company breathing), and usually short build lead-time, these are two very different types of expertise. I would look for a separate expert for each of those tasks, and would insist on receiving good documentation of process parameters in the first phase so that I could provide that input to the moulder on the 2nd phase.

I do strongly recommend looking for tool makers who are also capable of moulding in-house (at least for break-in and fine-tuning purposes) or willing and able to take responsibility for moulding at a nearby partner that they know well.

Either way, DDsystems, you should try to be realistic about your anticipated sales and when you will be upgrading your tools and use that to guide your selection of a shop. If you expect a step between your initial tool and a large scale production tool (i.e. 32/64 cavity) then try to find a shop that is efficient at both soft molds and early production molds as Ronen eluded to. If you think you will be making a quick, big step, then consider using 2 different shops. Also, keep detailed documentation of the processes as Ronen has recommended.

You can get tools produced for pretty darn cheap to get parts (much cheaper than biocomp) so hopefully you can get some parts made to get your regulatory through. Another option that Ronen eluded to would be to go with a true prototype lab (protolabs, goproto, etc), document the entire process, then have a true production tool maker try to copy the parameters of the prototype molder. I've never tried this method and haven't discussed it with someone that has, but I can't see why you wouldn't have a solid argument for why your biocomp on the prototype parts shouldn't be satisfactory for the production parts. Just make sure you try to copy everything down to the mold release.

 

[Ronen E]

If you expect a step between your initial tool and a large scale production tool (i.e. 32/64 cavity)...

It's very true that one doesn't normally go from 1 cavity to 32/64 in one leap. However, even with an in-between phase (say, 1st gen 1/2 cavity, 2nd gen 4/8 cavity, 3rd gen 32/64/128 cavity) I would look for a toolmaker that specialises in prototype tooling for the 1st gen and then look for a production tooling specialist for the heavy-lifting production tools. If the end goal is 32 cavities you want to do your "experimental" work (eg 4 cavity tool) with a high-end toolmaker rather than with a prototype tool expert - in my opinion.

For complicated parts that never go beyond 4 cavity I might consider going high-end from the beginning but that might not be a viable option for a startup company. It's a bit difficult to prescribe the best course of action when we know so little about the situation.

Another option that Ronen eluded to would be to go with a true prototype lab (protolabs, goproto, etc), document the entire process, then have a true production tool maker try to copy the parameters of the prototype molder. I've never tried this method and haven't discussed it with someone that has, but I can't see why you wouldn't have a solid argument for why your biocomp on the prototype parts shouldn't be satisfactory for the production parts. Just make sure you try to copy everything down to the mold release.

I don't know the labs you named or others around you, but my own experience indicates that this is not a good path though at face value it looks very tempting. In my experience most development labs either think in low-count prototype terms and never go to proper injection moulding, or they actually outsource tooling and moulding to tool makers / moulders that specialise in cheap tools and short runs (same as my suggestion). If one is lucky they actually take responsibility and drive the necessary results out of their suppliers; if not, you find yourself chasing those sub-suppliers (that actually owe you nothing because you're not their client, plus the occasional culture/language gap) and ending up compromising any combination of project outcomes, project schedule and project budget.

 

[Access2hc]

Ronen is spot-on..

all these fall under the need to have a 'production equivalent' in trying to meet the 'design validation' clause.

firstly the scope would need to be determined - perhaps not every part needs to be considered for biocompatibility?

your toolmaker/molder should understand that it's part of a medical device. The price will go up, but this is also the price of being in the medical device industry.

Hope it helps

Cheers,
Ee Bin
@Access2HC