PROCESS VALIDATION CEPHALOSPORIN - Matrix

Dear all,
we are a cephalosporin unit and during the issuing of Process Validation Protocols we came up with the following inquiry:

Consider a Cephalosporin dry powder product A, which is being filled in a glass vial 1g.
Lets say that the dry API is being held to our facility by the manufacturer in three types of containers 5kg, 10Kg and 20Kg.

Lets also say that these are the batch sizes also deposited in the regulatory authorities for the specific finished product.

Should the Process Validation be conducted on three batches per batch size (of 5Kg, 10Kg, 20Kg) or should there be a matrix set up?

Could we for instance arrange so that the first three batches made are of the highest batch size (20Kg) and the Process Validation is only conducted on these three batches? Could we therefore write a rational in our protocol that the filling line is the same for all three batch sizes and that the worst case is that of the 20Kg due to the extra time that the filling will take place?

What should be our approach (if you agree with the above rational) if for some reason the API manufacturer does not have available the 20Kg container and send us a 10Kg container? Should in this case perform a PV only for one batch and wait until 20Kg containers are delivered so that to conduct the PV in three consecutive batches?

Thank you in advance.

Best regards,

[quote=Gigastorm]Dear all,
we are a cephalosporin unit and during the issuing of Process Validation Protocols we came up with the following inquiry:

Consider a Cephalosporin dry powder product A, which is being filled in a glass vial 1g.
Lets say that the dry API is being held to our facility by the manufacturer in three types of containers 5kg, 10Kg and 20Kg.

Lets also say that these are the batch sizes also deposited in the regulatory authorities for the specific finished product.

Should the Process Validation be conducted on three batches per batch size (of 5Kg, 10Kg, 20Kg) or should there be a matrix set up?

Could we for instance arrange so that the first three batches made are of the highest batch size (20Kg) and the Process Validation is only conducted on these three batches? Could we therefore write a rational in our protocol that the filling line is the same for all three batch sizes and that the worst case is that of the 20Kg due to the extra time that the filling will take place?

What should be our approach (if you agree with the above rational) if for some reason the API manufacturer does not have available the 20Kg container and send us a 10Kg container? Should in this case perform a PV only for one batch and wait until 20Kg containers are delivered so that to conduct the PV in three consecutive batches?

Thank you in advance.

Best regards,[/quote]

IS THERE ANY FEEDBACK ON THE ABOVE QUESTION?

BEST REGARDS,

Since nobody has responded, I’ll give it a go.

First (Question 2), the use matrices for Process Validation is acceptable. However, it’s usually used in conjunction with variation in the output, not the input. That is to say, same process with outputs of 1X, 2X and 3X.

Second (Question 1), the approach is more along the lines of reducing the overall number of batches from the theoretical maximum, which in this case is 9, as opposed to only doing three batches.

Third (Question3), a scientifically based rationale would need to be provided. I’d typically expect to see this in the Validation Strategy document. However, it depends on how you construct your documentation set. If you don’t have a strategy document, then the protocol would be the next best place. The thing to remember here is that the approach is scientifically justified and instantiated prior to the execution of the study.

Now, for my 10 cents (used to be 2, adjusted for inflation and the falling value of the dollar). I’m struggling to identify the components of variation within the process you defined. It appears to me that your process is re-packaging bulk. The only variance that you’ve identified is the size of the input. This would impact to absolute number of vials and labels used, the amount of processing time, and the size of the input container. What else is there? The process and the output (1 gm vials) remain the same, as per the information you provided. Again, the key is that your approach be scientifically justified.

Dear Gigastorm,

I assume that the input batch size of your batch is same irrespective of type of containers (5kg, 10kg or 20kg). In such situation, if all the containers are of 5kg, then total number of containers will be 4x and 2x respectively for 20kg and 10kg containers.

Now, you are handling 4 times more containers than that of 20kg containers. It means you are providing very high probability of microbial and particulate contamination. As far as microbiological contamination is concerned, for Media-Fill 5kg scenario is worst case and the Media-Fill can be validated with such justification using 5kg containers.

Otherwise, I don’t see any problem with respect to process Validation. You can take think of taking single run of each size container and thus making three runs with proper justification.

Regards,

Jasbindersingh
(kamal.consultants@hotmail.com)