In the ideal word we would prefer to treat all products manufactured in a given equipment train as part of just one group. But since we don’t live in an ideal world, there are some situations where we have to make tradeoffs. One situation involves the manufacture of traditional products and highly hazardous product on the same equipment. By “traditional”, I mean products with relatively high PDE values; by “highly hazardous” I mean products with relatively low PDE values. I have written about this issue in the past (see Cleaning Memo of February 2020), where other thing being equal, it probably is best to make those two types of products on separate (distinct) equipment trains. So, the issue for this current Cleaning Memo can we group all the products together as one group, select the “most difficult to clean” product, and for that worst case product perform three protocol runs with the limit for the active at the lowest limit of any activity of any product in the group (see the Cleaning Memo for May 2017 for this strategy)?
Certainly that can be done, but it may lead to complications that we may not want to deal with. Here is an illustration with some examples based on the manufacture of six products (three traditional and three highly hazardous) made on the same equipment train.
Product | PDE (µg/day) | Relative Difficulty of Cleaning Ranking* | Max. Daily Dose** | L4b*** |
A | 20 | 6 | 1 | 5.00 |
B | 50 | 1 | 4 | 16.7 |
C | 40 | 2 | 2 | 10.0 |
D | 3 | 4 | 1 | 0.75 |
E | 1 | 5 | 3 | 0.25 |
F | 2 | 3 | 1 | 0.50 |
*With “1” being the most difficult to clean and “6” being the easiest to clean.
** Max. Daily Dose is number of tablets taken per day.
*** L4b is the limit in ppm in the extracted swab sample, assuming every other product could be the next product, and with batch sizes, tablet weights and shared area being the same.
Using the approach of “most difficult to clean at lowest limit”, I would perform a validation protocol using Product B at a limit of 0.25 ppm. If that is practical and readily achievable, that would be my approach. But, what if the analytical method for the active in Product B was only validated to measure down to 2.00 ppm? In that case I would have to either revise the method to measure down to at least 0.25 ppm (and preferably below that), or I would have to explore another option (to be discussed). There may be other situations where I would not like the approach of treating all six products as one group. What all those situations are don’t need to be covered here; but if you face one, you might consider “Plan B”.
So, what is “Plan B”? This involves forming two groups. As shown in Table 2, we have created Group I with products A, B and C, and then Group II with products D, E and F. All six products are made on the same equipment train.
Group | Product | PDE (µg/day) | Relative Difficulty of Cleaning Ranking* | Max. Daily Dose** | L4b*** |
I | A | 20 | 6 | 1 | 5.00 |
B | 50 | 1 | 4 | 16.7 | |
C | 40 | 2 | 2 | 10.0 | |
II | D | 3 | 4 | 1 | 0.75 |
E | 1 | 5 | 3 | 0.25 | |
F | 2 | 3 | 1 | 0.50 |
*With “1” being the most difficult to clean and “6” being the easiest to clean.
** Max. Daily Dose is number of tablets taken per day.
*** L4b is the limit in ppm in the extracted swab sample, assuming every other product could be the next product and with batch sizes, tablet weights and shared area being the same.
In this situation involving two separate groups, the most difficult to clean product for Group I is Product B, but the lowest limit for that product is now 5.00 ppm. Therefore we would be able to actually measure that residue in the equipment with the analytical method for the active in Product B, which has a LOQ of 2.00 ppm.
We would do the same thing for the most difficult to clean product in Group II. That product is Product F; and the limit for the active in Product F would be the lowest limit of any product in Group II, which would be 0.25 ppm. If the analytical method for the active in Product F could measure down to 0.25 ppm, I have resolved my dilemma. (Note that if I could not analytically measure the active in Product F at 0.25 ppm, I would have to consider still a different “Plan B” for the products in Group II.)
It should be carefully noted in the situation of two groups that the limit for each product does not change just because there are now two groups. It is still the case that the limit for each product should consider every other next product manufactured on the same equipment as a possible next product in a typical carryover calculation. It is important to carefully consider both (a) what the appropriate conditions for grouping of products might be and (b) how to matrix all the products on shared equipment for an applicable carryover calculation. Said differently, in the example given it is not appropriate to only do limit calculations for the three products in Group I or Group II just using the three products in that given group as the next product in a carryover calculation.
In the discussion of the analytical method, we based the analytical method capability on the LOQ of the method. While some may prefer to base it on the LOD of the analytical method, in essence that creates what is in essence a “pass/fail” protocol (the residue is either detected or it is not). This is a possible short term approach, but it is not desirable in the long term.
The second example presented assumed that the analytical method for Product B was insufficient to measure down to 0.25 ppm. But what if I could actually measure Product B down to a limit of 0.25 ppm? Another reason some might prefer two groups in this situation is because of increased regulatory concern about the treatment of highly hazardous products. For that reason, some companies might want to make sure at least one highly hazardous product is treated more robustly in terms of yearly cleaning verification protocols (essentially a repeat of one run of the original validation protocol). In addressing that concern, some might think this is just getting back to the old concept of doing one validation protocol on most difficult to clean product at its limit and one cleaning validation protocol on the “most toxic” product (meaning the product with the lowest limit). However, forming two groups and within each group selecting the most difficult to clean product and setting residue limits at the lowest limit in any product in that group is not the same as the older concept.
While grouping of products can make for a more efficient cleaning validation program, it may also present some challenges to implement consistently. The issues discussed here may illustrate the idea that grouping can become so complex that it becomes complicated and difficult to implement. So, a well thought-out and documented grouping approach is critical to success in implementation.