Learnaboutgmp Community

Cleanning validation by rinsing sampling

Hi to all, and thanks in advance for your answers.

So, I have to validate the cleaning process of the injectables/oftalmics area in the facility where I work. My bet is to use rinse sampling and TOC analysis.

It is going to be a tough work, since no cleaning procedure is close from especificity (“rinse with abundant water until it seems clean”).

Anyway, after all that gets done, I need to validate recovery of the sampling metod. As far as I know, it is easier to test recovery with swab sampling (althoug more variable), since I only need to spike some coupon of the same material of the equipment, and then sample and test recovery. But what happens when you have 4 kind of materials in the equipment (steel for containers and needles, silicon for hoses, glass for temporary containers, and even PET for transportation containers)? Do i have to test recovery on every material or only in the “worst case”? Also, how do I test recovery on a hose (or in any close system, like pipes)? Do I still to have to run a test recovery on them if all hoses are “product specific”?

Thanks for your advice

Dear AiC,

manufacturing injectables and ophtalmics, rinse sampling and TOC is a good strategy, but it makes no sense to even think about a cleaning validation before the cleaning procedures are not well developed and written down. As an advice, and taking into account that very few facilities can equip every water tap with a flowmeter, the cleaning procedure can specify rinsing times (“rinse, with purified water valve fully open, for two minutes covering the entire surface”).
As you said, the recovery test has to be carried out on every type of surface, but for chemical cleaning is is common practice to disregard surfaces that contribute less tha 1% to the total surface, thus skipping o-rings, gaskets, etc. and only requiring visual cleanliness for them. The dedicated parts (e.g. dedicated silicon hoses) can be left out from the sampled areas, but, in your specific case, may require some consideration because ophtalmics often include thickeners (cellulose ethers) which can be hard to remove and make a good growth medium for bacteria.

Dear AiC,

manufacturing injectables and ophtalmics, rinse sampling and TOC is a good strategy, but it makes no sense to even think about a cleaning validation before the cleaning procedures are not well developed and written down. As an advice, and taking into account that not every facility has flowmeters in every water tap, the cleaning procedure can specify rinsing times (“rinse, with purified water valve fully open, for two minutes covering the entire surface”).
As you said, the recovery test has to be carried out on every type of surface, but for chemical cleaning it is common practice to disregard surfaces that contribute with less than 1% to the total surface, thus skipping o-rings, gaskets, etc. and only requiring visual cleanliness for them. The product specific parts (e.g. dedicated silicon hoses) can be left out from the cleaning validation, but, in your specific case, may require some consideration because ophtalmics often include thickeners (cellulose ethers) which can be hard to remove and make a good growth medium for bacteria. For different types of material, it may be difficult to determine beforehand which is the “worst case” without any recovery tests.
Recovery testing for rinse sampling is essentially the same as for swab sampling, only that in many cases it is necessary to spike equipment surfaces and not coupons, and thus your recovery test may not cover contaminations above the acceptable maximum (nobody will let you contaminate the equipment “on purpose”, even if 85% of the spiked contamination is then removed with the sampling). By the way, a recovery test in the equipment is far more representative for the “real life” sampling after the cleaning, than a recovery test carried out comfortably sitting in front of a coupon in the lab. Be careful to keep the same sampling volume for both the recovery test and the cleaning test. Needles may be spiked with a solution of the target contaminant in an volatile solvent, letting a known and precisely measured amount of the solution dry in the needle and then sampling it by rinsing with a known amount of water.

Best regards

Alfred

Dear Alfred:

First of all, thanks for your answer.

Yeah, seems pretty pointless to validate a process that isnt correctly defined, especially in this example, as it is becoming a “chicken or egg” dilema: They say that they are going to change the SOPs once we validate, but we need correct SOPs before we can do that. What I plan to do, is to observe that at least the actual SOPs are followed, then standardize parameters, help write the new SOPs, and then validate them. As you say, there is no flowmeter in each water tap (in fact there is no flowmeter at all), so I was planning to specify rinsing times and valve conditions, while using a graduated or volume-known recipient to see how much time it takes to fill in those conditions and calculate an aproximate flow, in case it is needed.

In the case of the recovery studies, I have some experience on open surfaces, but not in closed ones, and I was wondering how I were supposed to spike that surfaces. The volatile solvent seems like a good idea. Any way, I dont think that not including the silicon hoses in the studies may be wrong, since, as you say, there a lot thickeners and even oils in the formulations. The original purpose on having dedicated hoses was to counter this and minimise the residues effect, but since the actual cleanning SOPs are not very specific in this aspect and even they may be not correct, probably there could be some cuantities of residues left after cleanning. I want to correct this by changing the actual SOP, and then test both (new and actual) to see what is cleanning better.

And thanks again.

AiC

Dear AiC,

the “hen and egg” problem with your SOPs is neither new nor uncommon, because very few companies take the time and effort to develop sound cleaning procedures. A possible approach is to tune your sampling and analysis methods, and then go for “prevalidation testing”, which is nothing different to try and see what you get. At least in our case, the experience was that for new and state of the art equipment, it does not matter too much how it is cleaned, even for very insoluble actives, and nearly every cleaning method proves to be robust. For older equipment with a lot of gaskets, o-rings, inflatable seals and so on, developing a robust cleaning procedure is a daunting task. You have, at least, the advantage that most of your actives should be quite soluble…

Best regards

Alfred

Dear Alfred.

One last question. Since most of the volatile solvents available are organic, it wouldnt be a problem using them to dilute samples to spike surfaces that are going to be analysed by TOC?.

Dear AiC,

there should be no problem with the organic solvents, as long as you use chromatographic quality. We have used Ethanol from Merck (111727) without problems, but in any case conducted a blank spiking and sampling to substract bias levels of organic carbon (from the sampling water, environment, materials, etc.). Don’t get anxious and let the spiked surface dry well (at least 30’, depending on volatility of the spiking solvent) without using compressed air, nitrogen, etc., because you may alterate your TOC levels.

Best regards

Alfred