Guard columns and method validation

[miro2009]

Hello chromatographers and validation experts,
I would like know what would be the impact of adding a guard column to an already validated method, I assume a partial validation would be required, but which items exactly would be needed to repeated?

Also, in case of validation of a method for quantification of a lyophilized product, how would the analysis repeatability (injection of 6 different preparations) tests be performed:reconstitution of 6 different vials and injection? wouldn't this include an additional variable of content variation from one vial to another (even if content uniformity is acceptable)

Finally, is it acceptable to specify that my Accuracy range is 70-130% (as its a method used for content uniformity), and for the Range test, to specifiy the same (70-130%)? I'm having accuracy problems with the 50% so I can't go for the 50-150% range.

I would appreciate your feedback and comments.

[cody84]

Since you would have to add a guard with the same phase as the column, I do not think anything would be needed, the system suit would be enough. You are not increasing the column lengths more that the range specified in USP for instance.

[Consumer Products Guy]

Cody - I agree with you. I say the USP/cGMP/FDA ranges account for stuff like that without validation being required.

That said, our QA people would say that there's been a change, and would need to validate that change.

[cody84]

Same hear CPG, afterall, who cares what R&D says, just let them validate all your methods for you and have no idea what's going on....

[miro2009]

Thanks all for your opinions and I totally agree!
Although our "QA" here only go by "documented" opinions, is this described anywhere in official literature, guidelines?

@Consumer Products Guy

That said, our QA people would say that there's been a change, and would need to validate that change.

My question would be what exactly to validate? a complete reptition of the whole validation?

[Peter Skelton]

If you're running an impurities type method with lots of peaks, run the same samples with and without the column to demonstrate that you're getting the same results and not having any detrimental effect on resolution.

If you're running an assay method then repeat the accuracy portion of the validation with the guard column.

That should cover it, essentially you need to show your QC people that by adding the guard column you aren't changing the results that the method gives. You shouldn't need to re-validate to do this. Alternatively, show them the USP section on permissible method paramater changes that do not require re-validation

[lynzjm]

If anything it might affect your retention times slightly and give you slightly broader peaks but it'll be fairly minimal effects to be honest. If it was me I'd document it in a "deviation" to the method and leave it at that, stick in a before and after chromatogram on the deviation/change control and that'll keep QA/QC happy. I certainly wouldnt re-validate the whole method - as said above its the same column material you're only making it a bit longer really to help protect your column!

lynz x

[miro2009]

Thanks again everyone!
Anybody also have a reply about my question about validation of methods for lyoplilized products in the very beginning of my thread?

how would the analysis repeatability (injection of 6 different preparations) tests be performed:reconstitution of 6 different vials and injection? wouldn't this include an additional variable of content variation from one vial to another (even if content uniformity is acceptable)

Finally, is it acceptable to specify that my Accuracy range is 70-130% (as its a method used for content uniformity), and for the Range test, to specifiy the same (70-130%)? I'm having accuracy problems with the 50% so I can't go for the 50-150% range.