Matrixing Criteria for 24 APIs

Hello people! here, a bunch of doubts! i apologize in advance.
I am collecting all the information I can, to carry out a cleaning validation on a milling machine (Jet Mill). My goal is to perform the Worst Case Rating (WCR)
In it, 24 different APIs are processed.
So far I have Ld50 data but I find it very difficult to find the ADE or PDE.
Is there a database on this?
The first big question is, should I work with ADE?
Can I save the situation by going to other criteria like ld50?

For now I have all the LD50 and the solubilities of each API, and I grouped them as follows, giving them values:
According to solubility in water: soluble 1, slightly soluble 2, Insoluble 3.
Then, by type of cleaning, which includes similarity between APIs and their form of cleaning and also, for their solubility in different solvents, namely:

  • Class I water soluble substances. - Class II methanol soluble substances. - Class III acetone soluble substances. - Class IV separate class for special substances with defined solubility.

Once this is resolved, I will start to add the values obtained from the production operators through questionnaires, to determine how difficult it is to clean each one of them and give them values

My idea is to perform a matrixing and according to results, determine how many cleaning procedures must we do, and which API will be the chosen for carry out validation and its MACO or limits requirements based on ADE (hope can i use ld50).

I would like to know if my approach is correct.

Another questions is, after performing such a lot of work and validation. every time the equipment is used (assuming that the validation has already been carried out successfully), is it necessary in the usual practice to carry out some analytical verification ?, or is it sufficient for the cleaning procedure alone?

As I see the panorama, and from the previous results, I think I will have to make at least 4 different validations.
Any suggestions will be very welcome.
Any doubts about what is raised do not hesitate to ask me!
Thank you!

Where to start! Yes, this is wrong! Only the ADe or PDE (permissable daily exposure, a form of toxicity) is acceptable in calculating your carryover acceptance criteria. I get my ADE or PDEs from
You MUST be compliant with EU Eudralex’s chapters 3 and 5. Ignorance is no excuse!

LD50 is used in DEVELOPING a cleaning procedure, thus your classifications are meaningless. All the cleaning validations you have performed are meaningless too.

Remember it is your ‘cleaning procedure’ that is being evaluated for its effectiveness and the yardstick are your residuals (API, cleaning agents, bioburden…). Thus, for each separate cleaning validation you should have a cleaning validation protocol and final report.

Thank you Boomer. This ADE or PDE, are only applicable for pH. Eur. ? Is
the same criterion for USP regulations?

USP doesn’t have any regulations. While the FDA guidance is pre-internet (and is derived from their regulations), their interpretation has changed so that the reference to safety is now toxicology.

I ‘REALLY’ wish they would upgrade this reference!

ok, meanwhile, suppose I have all the ADE available to work, the concept about matrixing is correct in your opinion?
and I have another question.
Is it worth using different type of pre-rinsing with different APIS and then a general procedure?
For example: a rinse with ethanol for, suppose 10 APIs, another pre-rinse with nitric acid for 5 APIs, and another pre-rinse with alkaline solution for 9 APIs, and after these pre-rinses (well defined, with times and temperatures of course), perform a single general cleaning procedure with detergent for the 24 APIs.
In this way, it would perform a validation only, both to look for the tracer (a single worst case considering the 24 APIs) and to look for detergent residues.

Matrixing is OK. Usually, it is part of your risk analysis.

Using other API’s is not a good idea since to me they have a physiological effect. Chelates like EDTA and Citric Acid are commonly used by cleaning agent manufacturers (they are Food Grade). Detergents like SDS are used too.

Categorize your cleaning procedures (include work instructions), dosage forms, and the APIs. This will tell you the number of cleaning validations you will have to perform.

Dear Boomer, thank you for your response. Here the doubt is as follows, the equipment is used with 24 APIs and the doubt that arises from a colleague’s question is to know if it is possible to perform different rinses according to the solubilities of the API in the first instance, and then yes, to establish a process of general cleaning for all APIs, and in this way, perform a single validation work. if this is valid and working, I would be performing only one validation and not four (all four validations arise from my analysis).
I do not agree with this assessment of my colleague, but maybe it is valid … We are short time and proceeding in this way would be much less work. I hope you understand what I am trying to explain. Please, any questions just write to me. A warm greeting and thanks for your dedication.

He is right in the term of using your solubilities for the analytical method, but he is wrong in terms of assessing risk. It is the risk analysis that determines the target molecule and the worst case (most toxic).

I am too late for reply but it’s ok. I am explaining the sequence for selection of worst case out of 24 API. First you have to group API base on cleaning solvent. The same solvent API are in one group. It’s mean all different cleaning solvent having different group of API.

Now you check the solubility of all APi in single solvent group, Mark least solubility API.

Secondary select longest chain of equipment use for production of individual API in above selected group and consider longest equipment chain area for MACO calculation.

Third calculate ADE or PDE values for each same solvent group API &. Calculate MACO considering lowest ADE or PDE value of group. This would be total Maco for that group.

Now devide MACO in for all individual equipment using longest equipment chain considering equipment service area.

Here MACo limit for individual equipment is considered as worst case for this solvent group. Same way repeat for other’ solvent group API s.

Only the method uses for cleaning should be same for all individual API of same group.

I hope it would be helpful for worst case determination.


Dear Nil…never is too late! thank you very much…
we keep in touch.
Kind regards,