This is the fourth of several Cleaning Memos dealing with the selection of a worst-case product in a grouping approach for cleaning validation. In August we considered use of the pharmacopeia solubility categories, and also presented what may be a more useful system based on how “solubility bands” are defined. In September we considered issues in the use of typical evaluation systems involving calculation of a risk score that include a solubility parameter in addition to other parameters such as toxicity, potency, and operator cleaning difficulty. In October we discussed solid oral drug products based on two formulation parameters (solubility and concentration of each formulation component). If you haven’t done so, please review those before continuing. What follows are separate considerations for liquids, for small molecule APIs, and for biotech proteins.
Liquids
We’ll start this month by addressing liquid formulations, specifically liquids that are aqueous and where the excipients in the formulation are readily water soluble (such as soluble at least by the USP category “slightly soluble“). For these liquid formulations, it not necessary to consider the intensive evaluations that were covered last month for solid drug products. For those solid products, a calculation was done to establish the “Total % Insolubles” (as defined last month) to determine the “most difficult to clean” drug product. For these liquids, there are two possibilities.
Possibility 1: One or more of the drug products has an active in the “practically insoluble” or the “very slightly soluble” category. This is likely to occur if some liquids are “true” solutions and other liquids are “suspensions”. In this situation,, if there is only one product with an active in the lowest solubility category, that product is considered the “worst case”, and that product is chosen as the “representative” product in a grouping validation protocol approach. If two or more products fall into the same solubility category, one option is to use the criteria discussed in Example B in the October Cleaning Memo. That is, to “break the tie” move to a second parameter such as concentration of the active in the formulation. Another option to “break the tie” is to move to from solubility categories to specific water solubility data (such as data in mg/L or ppm). If such data exists for a reliable comparison, the active with the lower solubility is the worst case.
Possibility 2: None of the drug products has an active in the “practically insoluble” or the “very slightly soluble” category. This is more likely if all products are true solutions. In this case, realize that picking the “most difficult to clean” product is like asking “how many angels can dance in the point of a pin?” Does it really matter, or can anyone really know? However, just to “formalize” the decision-making process, one option would just to pick the product with the lowest solubility category (even though it is not “practically insoluble” or “very slightly soluble”). A second option, would be to the pick the product with the highest concentration of active. Additional consideration include lowest PDE active, or even the product most frequently made (to facilitate getting the validation runs performed in a timely manner).
On the other hand, there may be formulations that are creams or ointments. In that case, consideration of the total formulation ingredients (as done for solids) may be appropriate. However, because of the possible complexity of the cleaning process, it may be better to move to a laboratory evaluation as a more reliable way to determine the relative “difficulty of cleaning”.
Small molecule APIs
Now we’ll “switch gears” and look at what can be done for small molecule API synthesis. In that situation, it is necessary to address whether there are process materials (in addition to the API itself) that may affect the difficulty of cleaning. Examples of such materials include charcoal and metal catalysts. If it is solely the API that is of concern, and if the different APIs are cleaned with the same solvent, then it is possible to just compare the solubility of the different APIs in that one solvent at the temperature of the cleaning process. Note in this case it probably is not appropriate to compare different APIs if different cleaning solvents are used. Part of the reason is that an “apples to apples” comparison is difficult; any difference in the temperature of cleaning (at the solvent reflux, for example) further complicates the comparison. Use of different solvents also violates an essential principle for product grouping – that the products in the group be cleaned by the same cleaning procedure.
If “other” insoluble materials are present and if those same insoluble “other” materials are present in all products in the group at the same level, then it may be possible to look at the API solubility to determine the worst case. If different insoluble materials are present or the insoluble “other” materials are present at different levels, then it probably complicates the situation such that it is likely that a laboratory test procedure would be the easiest way to determine difficulty of cleaning.
Biotech proteins
A final consideration for special cases involves biotech protein drugs. In this case, we should realize that the most important issue may be the time for degradation of the protein to form readily soluble smaller fragments. It may be possible to select a worst case based on a laboratory evaluation (using an end point via SDS-PAGE comparison) for different proteins stressing the time element to determine which is the slowest in achieving “complete” degradation. This laboratory evaluation should consider equivalent cleaning procedure parameters including the ratio of protein to cleaning solution. On the other hand, since the equipment used may vary from protein to protein, the better option may be to just stay with an approach of doing cleaning validation separately on each product manufactured.
This concludes thisseries on selecting the “worst case” products in a grouping strategy. It should be obvious that in all situations the selection process depends to a significant degree on a manufacturer knowing and understanding its products and its cleaning procedures. Furthermore, I would be remiss if I did not also say that in a grouping approach, the protocol for the “most difficult to clean” product should have a residue limit that is the lowest limit (such as L3) of any product in the group. Copyright © 2023 by Cleaning Validation Technologies