Last month we discussed options for shortening the DHT to simplify (and perhaps improve) execution of the cleaning procedure. This month we will focus on options for actually validating the maximum DHT. In this current discussion, it is assumed that the desired maximum time has already been selected (see considerations for selection of a maximum time in last month’s Cleaning Memo). So, the question is “How do we handle this in the validation protocol to be able to claim that the maximum DHT is actually validated?” I am assuming (for simplicity) in the examples discussed that there will be three runs for the protocol itself.
The most straightforward option is to have all three validation runs be at the maximum DHT. If the maximum DHT is relatively short and this can be done without unnecessarily interfering with production scheduling, it certainly should be considered. It is the “cleanest” approving for passing scrutiny from auditors or regulators. A possible downside to this approach is that while for the three validation protocols the appropriate short timing can be adhered to, later on it becomes very difficult to stay within that maximum time, resulting in a “deviation” situation. Of course, if that is a possible future problem then you probably should not have selected that very short time for the maximum in your protocol.
Another option is possible provided there is some evidence that the “difficulty of cleaning” of the product on the equipment surfaces does not change with the elapse of time (or at least with an extension of time through the maximum time you have selected for your DHT). For example, this might be the case with certain process steps of oral solid dosage form manufacturing. In that situation where one is just processing dry powder, it is not expected that the equipment becomes more “difficult to clean” with increasing elapsed time before cleaning is initiated.
Whatever the situation, there should be some written rationale or assessment that it is likely that the difficulty of cleaning does not change significantly with increasing time. This may be because the product is a dry powder (and thus evaporation does not occur). For liquid products it may be because while there is some level of evaporation during the DHT, the cleaning process has been designed with a worst-case situation in mind. That worst-case situation may be where the liquid product has been allowed to evaporate to “practical” dryness, or where in simulated lab tests the liquid product has been allowed to dry for a time equivalent to the maximum DHT (or perhaps an even slightly longer time than the maximum DHT as a challenge).
In either case, realize that in any cleaning process validation protocol there always is an actual DHT, even though it may be very short (see the April 2017 Cleaning Memo for a further elaboration on why I say this). So, if I have a rationale for the maximum DHT having no effect on the difficulty of cleaning because the manufactured product is a dry product, can I just use any DHT in my three validation runs? That is, for the protocol runs can I utilize any times as long as each is within the maximum DHT? While that is possible, it is probably preferable to use three different times. For example, if the maximum DHT were three days, I might select one run to be a one-day DHT, another run to be a two-day DHT, and another run to be a three-day DHT. That way I have at least one run at the maximum, but I also have runs at shorter times. If it were the case that the difficulty of cleaning did not change with time, then the analytical results should be more-or-less the same (within expected experimental error or variation). So, I would have additional support confirming the original assessment that difficulty of cleaning does not change with time.
How does that apply to the situation where the product is a liquid product, where there is a likelihood of it drying and becoming more difficult to clean. My recommendation would be to do the same approach for a protocol as suggested above for dry product. For example, if my maximum DHT were six days, I might select one run to be a one-day DHT, another run to be a three-day DHT, and another run to be a six-day DHT. That way I have at least one run at the maximum, but I also have runs at shorter times. If as determined in my lab testing the cleaning process was effective when the liquid product was “fully” dry, I could see one of three possible sets of residue data (given as examples below assuming an L4b carryover limit of 15.7 ppm).
| Case I | Case II | Case III | |
| One-day DHT | 1.4 ppm | 1.2 ppm | 1.2 ppm |
| Three-day DHT | 1.9 ppm | 2.4 ppm | 5.3 ppm |
| Six-day DHT | 0.9 ppm | 4.7 ppm | 12.5 ppm |
For Case I it appears that the residue level is at a very low level (compared to my limit of 15.7 ppm), which probably means that the cleaning process I designed is so robust that even at the maximum DHT my data is close to my LOQ for the analytical method. So, I am happy.
For Case II, the data is consistent with my assessment that cleaning does, in fact, become more difficult with elapsed time, which is why I have designed my cleaning process so that even at the maximum DHT my analytical results are well below my L4b limit. So, I am happy here.
For Case III the analytical results are similar to Case II except that the value at the six-day DHT is much closer to my L4b limit. With data like this, it is reasonable to require either (1) that a total of three runs be performed at that maximum DHT, or (2) that my cleaning process be improved so that the analytical data more closely resembled the results for Case II, or (3) that I shorten my maximum DHT to only about three days. With any of these three alternatives, it means a lot more work, so I am not happy. Note also that even if I obtained data in the range of 12.5 ppm for the additional two runs using a six-day DHT, my personal preference would be for future work to improve the cleaning process to obtain better data or to shorten the maximum DHT to just three days.
For clarification in the three examples above, only one data value is given. In actual fact for most real life situations, there will be multiple values in each protocol run (for different swab locations, for example). Based on that fact, additional considerations will be necessary if values for different swab locations vary significantly.
We will explore further options for extending an already validated maximum DHT next month (January 2023).
Copyright © 2022 by Cleaning Validation Technologies