In August 2015 I wrote about “route specific” PDE values as a possible basis for cleaning validation limits. By “route specific” I am referring to situations where the cleaned product is administered by one route (such as parenteral) and the next manufactured product is by a different route (such as oral). While the basic principles and/or consequences of that Cleaning Memo have not changed, I have significantly revised how that approach should be presented in this Cleaning Memo. These changes are similar to the changes proposed last month for duration specific PDE values.
First is a basic consideration. The issue of “route specific” is not something that is typically applicable to finished drug product manufacture. The reason is that generally all finished drug products manufactured on a given equipment train are administered by the same route. For example, product contact equipment used for oral drugs is not generally used for parenteral drugs. However, for API manufacture it may be the case that some APIs are used for a parenteral drug product, while other APIs are used for oral drug products.
A second consideration is that I now advocate not adjusting the PDE value itself. By that I mean that I only set one PDE for a given active, and that value is based on the ISPE Risk-MaPP definition which specifies “by any route” and is therefore protective of “all populations of administration”. I do not propose having one PDE value for parenteral routes and a different PDE value for oral routes. Note that while I fully realize there is “wiggle room” in both the Risk-MaPP document and in the EMA document for having a “route specific” PDE value, I believe now that it is better to have only one PDE covering any route of administration, and then making a route specific correction in the carryover equation.
A third consideration is most companies that sell PDE monographs sell them with one PDE value based on the “by any route of administration” principle. Most commonly that route is a parenteral route (although there may be unique situations where that is not the case).
In the past (the August 2015 Cleaning Memo) I proposed adjusting the PDE value to a higher value based on oral toxicity studies as compared to PDE values (based on parenteral toxicity studies) if the next manufactured product in a carryover calculation is administered orally. What I didn’t suggest was how to do this. However, here are several possibilities that the toxicologists in your company could address.
1. One obvious approach is to start with oral NOAEL to calculate directly an “oral PDE”. If that is possible, it can be used. However, it does not appear that most third party PDE monograph suppliers do that now.
2. A second approach is the principle in the EMA guideline which seems to allow “correction factors” if the route difference in API bioavailability is >40%. Unfortunately there is no guidance in that EMA document as to what those correction factors might be.
3. The principle in the EMA guideline could be “fleshed out” by considering a modification of the approach used in ICH Q3D (R1) dealing with PDE values for “elemental impurities”. In Appendix 2 of Q3D values for many different elements are presented for three routes of administration – oral, parenteral and inhalation. My point in discussing that Appendix is simply that PDE values have been established for those elemental impurities for different routes of administration. The approach is that the PDE values for the different routes are established by starting with NOAEL values for each specific route (which makes sense for elemental impurities such as metals).
However, in Section 3.1 of Q3D is a discussion of how to establish correction factors to derive parenteral PDE values from oral PDE values in the absence of sufficient parenteral data. Note that this is not the problem faced by the pharma industry for most APIs. In the pharma industry for cleaning validation of APIs, if we depend on third party PDE monographs in for cleaning validation purposes for APIs, we generally get PDE values for any route, and the parenteral route is the most stringent (that is, the lowest value). So I’ll first present the Q3D approach for deriving parenteral PDE values from oral PDE values, and then I’ll reverse that analysis to get correction factors for a carryover calculation if the next product is administered orally.
Here is the approach in Q3D:
“In the absence of data and/or where data are available but not considered sufficient for a safety assessment for the parenteral and or inhalation route of administration, modifying factors based on oral bioavailability were used to derive the PDE from the oral PDE:
For example, if the oral PDE is 100 mcg/day, and the oral bioavailability is only 40%, then a modifying factor of 10 is applied to the oral PDE to derive a parenteral PDA of 10 mcg/day. Note that parenteral PDE value would be the same regardless of whether the oral bioavailability was 1% or 49%.
So, how does this relate to the question of dealing with a parenteral PDA if the next product is by the oral route? One approach is to reverse the Q3D approach above and derive a “correction factor” for use in the typical carryover calculation. In this approach the PDE value by any route of administration (assumed to be parenteral or inhalation) would be used in the carryover equation, but a correction factor would be applied to the numerator to allow for an increased L4 limit. The “correction factor” (which may be called a route correction factor) applied to the carryover equation would be the same numerical value as the “modifying factor” used in Q3D.
For example, the PDE by any route of administration (assumed to be parenteral) is used unmodified in the carryover equation. If the oral bioavailability of that API is only 40% and if the next product is administered orally, then a “correction” factor of 10 is applied to numerator in the carryover equation. Note that the correction factor would be the same regardless of whether the oral bioavailability was 1% or 49%.
Note also that you could conceivably just calculate separately an “oral PDE” (using the same correction factor applied to the “parenteral PDE”), and some may decide to do that. However, the approach presented above is similar with the approach discussed last month for dealing with duration specific issues, so it may allow for a more consistent approach in how limits are calculated and presented in your documentation.
In any case, the use of the approach discussed above should be understood and agreed to by your toxicologists. Those toxicologists should also fully understand the reasons why cleaning validation scientists might need to address situation where the PDE is for parenteral routes and the actual next product is administered orally.
As a final caveat (and as most of you are aware), I am not a toxicologist or pharmacologist. However, my contention is that this approach is a reasonable utilization of scientific principles.
Copyright © 2024 by Cleaning Validation Technologies