Meeting protocol requirements for bioburden is generally not a significant problem provided the final rinse is with hot Purified Water (PW) or hot Water for Injection (WFI). It is not likely that such a water final rinse would contribute to bioburden, and it is likely that the hot temperature would significantly reduce any vegetative microorganisms present in the equipment at the end of the cleaning process. This general rule also applies to situations where alcohol (such as 70% IPA) is applied to the equipment surfaces as a final step. Note that if tap water were the final rinse, then that certainly would be of greater concern. But those issues are not the main focus of this Cleaning Memo. The focus is whether limits for yeasts/molds should be included in our testing for cleaning process protocols.
The rationale for having limits for yeasts/molds is that there are USP <1111> recommended standards for yeasts/molds in pharmaceutical drug products. USP <1111> provides maximum values for aerobic bacteria (Total Aerobic Microbial Count, or TAMC) and for yeasts/molds (Total Combined Yeast and Mold Count, or TCYMC). For all drug product forms (such as oral solids, oral liquids and topicals), there is a consistent one log difference between the recommended value for TAMC and the recommended value for TCYMC. For example, for solid orals (such as tablets), the value for TAMC is 103 CFU/gram (1,000 CFU/gram), while the value for TCYMC is 102 CFU/gram (100 CFU/gram). If that same ratio for drug products were applied to cleaning validation samples, then we should consider applying a factor of 1/10 to the bacterial count to derive a safe level for yeasts/molds. Using that analogy, if the swab limit were 50 CFU/swab, then the limit for yeasts/molds would be only 5 CFU/swab. Ideally, this would involve taking two samples from each (adjacent) location, and using different media and different incubation temperatures for the two different microbial concerns (bacteria vs. yeasts/molds).
An argument could be made against that one-log difference approach. It would go something like this: The typical limit used for cleaning validation purposes is much lower than what would be needed to provide levels in the next manufactured product near or above the <1111> recommended values; therefore it is not necessary to apply the one log reduction to the TAMC to get an acceptable value for TCYMC. I can understand the logic there, but I’m not sure all would fully support such an approach.
Here is an alternate possible approach that might be acceptable and easier to implement, particularly if in my specific facility I consistently get values below 5 CFU/swab. With such low total counts, I reduce the likelihood of fast growing bacteria ‘inhibiting” the colony formation of yeasts/molds. Therefore, (with your company’s microbiologist “buying into” this) you could just do a Total Viable Count (including both bacteria and yeasts/molds). For ease of calculation for this example, I am going to use an acceptance criterion of 50 CFU/swab for total count. Then I set my yeast/mold count at one log smaller, or at 5 CFU/swab. Therefore, if my total count (bacteria and yeasts/molds) were 5 CFU/swab or less, then I can safely assume that I have acceptable levels of both bacteria and yeasts/molds. But, if my total counts were 10 CFU/swab, I would characterize the colonies to determine if they were yeasts/molds. This might be something that could be simply done by a visual inspection by a trained microbiologist as to the physical appearance of the colonies, or I might have to do more. In any case, if the number of colonies that were yeasts/molds could be clearly determined to be 5 CFU/swab or less, then I could establish that I was meeting my acceptance limit of 5 CFU/swab. This technique might involve incubating on soybean-casein digest media at a temperature intermediate between that typically used for bacteria and the temperature typically used for yeasts/molds. However, if acceptable, it simplifies the nature of testing. Note that if I consistently got results of 20 CFU/swab or more, it may not be as helpful. However, if I were to consistently get results above 20 CFU/swab, I should probably look at my cleaning process to see if it could be improved to consistently get value below 5 CFU/swab.
In applying this approach, remember that for cleanroom monitoring there is generally not a distinction in terms of total counts between bacteria and molds/yeasts; it is the total combined count that is relevant. Certainly knowing whether it is a bacterium or a yeast/mold could be important for other reasons, but for routine monitoring it is the total count that is important. Also, remember that if you consistently get values below 5 CFU/swab (with a swab being 25 cm2), you probably have an acceptable program and you probably have other areas where your efforts could be spent for the continuous improvement of your cleaning validation program.