Sometimes I hear clients (or potential clients) get confused about what is the “worst-case” product for a protocol. This Cleaning Memo will clarify the question, and provide several answers depending on the context.
Use I
In a grouping approach for cleaning of drug products, it is generally the most “difficult to clean” drug product that is selected as the representative product to clean for the protocol. The idea here is that if the “most difficult to clean” drug product can be cleaned below a certain residue limit, then all other products (easier to clean) should also be cleaned down to that same residue value. There are several options for picking that worst-case drug product, which may include a variety of physical properties, chemical properties, lab tests, and operator experience. Note that for drug product manufacture, it may include the properties of the entire formulation as well as the properties and concentrations of each ingredient (actives and excipients). In drug substance (API) manufacture, just substitute “drug substance” (or “API”) for “drug product” and follow the same principles.
Use II
A second use in a grouping approach is to say that the “worst-case” is the residue (typically the active) that results in the lowest residue limit among all actives in the group of drug products. That residue limit is best expressed as a L3 value. While in most cases a L4 limit could be used, the use of different sampling parameters (such as the area swabbed or the volume of liquid used for extracting the swab) could give an erroneous value for the lowest residue value. Just selecting the active with the lowest ADE/PDE value may also give misleading worst-cases. The reason for this is that it is the residue level on the surface that is critical. That residue level will depend on the batch size (of the next drug product) and maximum daily dose (of the next drug product). For example, suppose Product Q has a PDE (of the active) of 8 mcg, and a batch size of 1000 kg and maximum daily dose of 500 mg, while Product R has a PDE (of the active) of 10 mcg, and a batch size of 1200 kg and maximum daily dose of 200 mg. While Product Q has the lowest PDE, the lowest residue limit should be based on Product R. The calculations below help express the relative difference based on multiplying the PDE of one product by the batch size of the other product and then dividing by the maximum daily dose of that other product.
L2 for Q = 8 mcg X 1200 kg / 200 mg = 0.48 kg
L2 for R = 10 mcg X 1000 kg / 500 mg = 0.20 kg
This means that the residue limit should be based on the calculation for Product R, and not for Product Q.
Use III
A third possibility is in selecting batch sizes to use for the most “difficult to clean” product when actually performing the cleaning validation protocols. That is, suppose that the most “difficult to clean” product (the one you are using for performing the cleaning validation protocol) is made in several batch sizes, one size being 1000 kg and a second size being 2000 kg. In this case, the three batch sizes used for the protocol should preferably be the larger batch size. Why is this the case? The most logical rationale is that, other things being equal, the larger batch size is more likely to leave higher levels of residues after a fixed cleaning process.
There may be some variations on this approach. One variation is if the batch size varies continuously between 1000 kg and 2000 kg. In this case, it may be permissible to require that the three batches all be in a suitable range, such as 85%-100% of the largest batch size. This is mainly done to deal with scheduling issues.
A second variation is to have only one batch at the maximum, while the other two batches are less than the maximum. In this situation, it is preferable that the largest batch be the first protocol run, and then to compare the residue data for the two smaller batch sizes to the residue data for the largest size, making sure that the residue data for the smaller batches is no more than 120% (for example) of the data for the larger batch (that 120% value may be used to account for normal “noise” in collected residue data).
Use IV
A fourth possible use of “worst-case” product is when, as part of the limits calculation for a given product, the subsequently manufactured products, and specifically the minimum batch size and the maximum daily dose of that next product are considered. Remember that in a carryover calculation for residue limits that the batch size is in the numerator (and therefore the “minimum” is the worst case) and that the daily dose is in the denominator (and therefore the “maximum” is the worst case). Therefore, assuming other things like shared surface area and sampling parameters are the same, it is the subsequently manufactured product with the lowest ratio of minimum batch size to maximum daily dose that will result in the lowest residue limit for the cleaned product (that is, the previous product).
A final caveat is necessary. If you typically also use calculations based on 0.001 of the minimum therapeutic dose of the active or based on 10 ppm of the active in the subsequently manufactured drug product, then some slight adjustments may be necessary in selecting a “worst-case” for uses involved in determining the lowest residue limit. However, understanding the principles behind the four uses discussed above should be a helpful guide. Also note that if the equipment is dedicated to only one drug product, then only Use III is likely to be considered.
As a review, here is when the four “uses” given above are most likely to be considered:
Also, remember that there may other situations in cleaning validation where there is a reference to a “worst-case”. For example, there may be “worst-cases” in the values for parameters in the cleaning procedure (the cleaning “process”). For example, some companies might like to challenge parameters like times and temperatures of the cleaning procedure as worst cases in their CV protocols. While that can and has been done, a better approach is to try to narrow the ranges of those parameters in the design/development studies, only challenging the extremes of those parameters in Stage I of a life-cycle validation approach. If that is done, then only the “normal” parameters are used in the validation protocols in Stage II.
Copyright © 2023 by Cleaning Validation Technologies