Let’s say you have previously validated a cleaning process, and that things have changed such that the limit for the active of the product cleaned has changed. This is generally a significant issue only if the result of the change is a lower limit. This change may have been caused by the addition of a new drug product on the same equipment; based on that new product being a possible “next” product in a carryover calculation, the “batch size/maximum daily dose” ratio may have changed resulting in a lower limit. Another cause might be a decrease in the dosing or HBEL values of the active of the cleaned product, again resulting in a lower limit. A third cause might be a change from the older grouping approach of separate protocols for the “hardest to clean” product and the “most toxic” product to the newer approach of one protocol for the “hardest to clean” product at the lowest limit of any product in the group. Other possible causes include a change in equipment that results in increased surface area, a change (decrease) in the batch size of a “next” product, and a change (increase) in the maximum daily dose of the next drug product. Needless to say, all these changes should be addressed as part of a change control program.
The issue for this Cleaning Memo is what should be done if my validated limit decreases. Do I have to repeat my cleaning validation protocol at that new (lower) limit? Certainly that is an option. But, there are some things you can consider that may help you avoid a complete revalidation exercise (or conversely may help convince you that you have to revalidate completely).
Here is one exercise to consider. This is based on comparing the actual validation data obtained in the prior protocol to the new, lower limit required by the change. Let’s say that the prior validation had a limit for the active of 3.0 mcg/cm2 (this value will be used for this example and all the subsequent examples), and that the change has reduced that limit to 0.7 mcg/cm2. What you can do is evaluate the actual data obtained in the prior protocol to see whether that actual data is below the new, lower limit. For this first example, let’s say Table 1 is the swab data values for the active for sampling locations A, B, C, D and E (with the LOD of the analytical method being 0.01 mcg/cm2).
| A | B | C | D | E |
|---|---|---|---|---|
| 0.12 | 0.08 | <LOD | 0.10 | 0.07 |
In this situation a reasonable conclusion might be that if the protocol had been performed at an acceptance limit of 0.7 mcg/cm2, the protocol would have passed. If subsequent yearly confirmatory protocols (subsequent to the original validation protocol), as well as routine monitoring, supports the fact that the cleaning process is in a state of control, that data assessment might be adequate to say that a new protocol is not needed. Realize that there might be other reasons why you decide to repeat the protocol with the new, lower limit. For example, you might decide that the age of the original protocol is such that it might be prudent to repeat it, or it may be that you want to make other changes that would require a new validation protocol.
On the other hand, let’s say the original swab data was as given in Table 2, with the LOD of the analytical method still being 0.01 mcg/cm2.
| A | B | C | D | E |
|---|---|---|---|---|
| 0.45 | 0.34 | 0.29 | 0.47 | 0.53 |
In this situation a reasonable conclusion might be that if the protocol had been performed with the lower limit, it would have passed. But, do I really want to operate with a cleaning process that close to failure? In that situation, I might conclude that going forward, I would want to make my cleaning process a little more robust (perhaps by increasing the washing and/or rinsing time). Success with that additional protocol would give me more confidence in the robustness of my cleaning process.
Let’s consider a third situation. If the data in either Table 1 or Table 2 were my original results, but the new, lower cleaning limit was only 0.09 mcg/cm2. Clearly in those cases my prior validated process is inadequate to demonstrate that I could effectively continue cleaning with my previously validated cleaning process. There may be several options in this situation if I want to continue with this lower limit. One would be to redesign my cleaning process so I could consistently meet this much lower acceptance limit. Depending on the level of robustness desired, I also may have to develop my analytical method to have a lower LOD/LOQ. A second option (one that I generally try to avoid) is to restrict the order of manufacture such that my limit for the cleaned product is higher. Another option might be to adjust my swab handling conditions (such as swabbing a larger area or extracting the swab with a larger volume of liquid) to increase my limit.
Here is a fourth situation. Let’s say a third set of data (given in Table 3) is my prior validation results and that the newer acceptance limit is 0.7 mcg/cm2, but that my LOD for the analytical method was 1.0 mcg/cm2 (higher LOD as compared to the first situation).
| A | B | C | D | E |
|---|---|---|---|---|
| < LOD | < LOD | < LOD | < LOD | < LOD |
Clearly in this situation, while my “true” residue levels may meet my acceptance limit, I cannot confirm that it is truly the case. My analytical method is just not robust enough to confirm that I am below the new acceptance limit of 0.7 mcg/cm2. It may be or may not be; I don’t know. In this situation, I would try to lower the LOD/LOQ of my analytical method (which would allow me to clearly establish the actual values in one or more protocol runs), or perhaps try to do some of the things mentioned in the prior example to raise the acceptance limit.
Here is a fifth situation. Let’s say a fourth set of data (given in Table 4) is my prior validation results and that the newer acceptance limit is 0.7 mcg/cm2, but that my LOD for the analytical method was 0.3 mcg/cm2 (higher LOD as compared to the first three situations but lower compared to the fourth situation) and that my LOQ for the analytical method was 0.9 mcg/cm2.
| A | B | C | D | E |
|---|---|---|---|---|
| < LOD | < LOD | < LOQ | < LOD | < LOQ |
This is another situation, while my “true” residue levels may meet my acceptance limit, I again cannot confirm that it is actually the case. Yes, the “<LOD” samples are meeting the lower acceptance limit, but the two “<LOQ” samples may have “true” values above my acceptance limit. My analytical method is just not robust enough to confirm that I am below the new acceptance limit of 0.7 mcg/cm2. In this situation (just like the fourth situation), it may be or may not be; I don’t know. In this situation, I would try to lower the LOD/LOQ of my analytical method, or perhaps try to do some of the things mentioned in the prior example to raise the acceptance limit.
The purpose of this Cleaning Memo is not to recommend one approach or another for resolving any issues in these situations. Each company should evaluate the risk as to the acceptability of any approach.