Chromatography Forum

Hi,

we frequently use the raw material pharmacopeal method to analyse our API in our finished product.
This way we avoid to repeat the whole selectivity part of the validation (only selectivity against excipiens).

But now we have a method for a neutral component which uses water and THF as mobile phase.
But in our excipients, we have some charged molecules, so these peaks elute as broad peaks, shifting more than 3 minutes, ....

When we add 0.1% phosphoric acid to the water, all our problems are solved, and no shift of retention times has occured for our API, the RT is exactly the same for all peaks in the system suitability.
But can we conclude that we are using a method which doesn't need to be revalidated for selectivity, as these are all neutral components, not affected by the addition of a pH modifier?

Best regards

Ace

I just happened to have the EP Allowed Adjustments sitting on my desk! The USP is similar.

IF this is a compendial method, the changes allowed are given below and I suggest you do an equivalency. IF IT IS NOT compendial (e.g. in house), I encourage you to contact your QA department to make a risk assessment if an equivalency is needed.

Allowed conditions:
For isocratic: pH can be adjusted +/- 0.2 or +/- 1 when non-ionisable substances are to be examined.

For Gradients, no pH adjustment is allowed.

I'd suppose that adding 0.1% phosphoric acid to plain water will change the pH more than just 0.2 units. Moreover, EP's compendial procedures as well as the allowed changes are supposed to be used for the pure APIs only. You're using it for the finished product, right?. So you'll have to validate it anyway. With all the work involved going through accuracy, linearity & co. do you really want to take the risk of omitting selectivity just to save a few days? Even if it might make sense from the analytics point of view, nice people like auditors are often not very prone to argumenations like that.
During development stages, I'd just use the method as you do. But when it's getting serious, I'd do complete validation including selectivity. Call it regulatory paranoia...

I'd suppose that adding 0.1% phosphoric acid to plain water will change the pH more than just 0.2 units. Moreover, EP's compendial procedures as well as the allowed changes are supposed to be used for the pure APIs only. You're using it for the finished product, right?. So you'll have to validate it anyway. With all the work involved going through accuracy, linearity & co. do you really want to take the risk of omitting selectivity just to save a few days? Even if it might make sense from the analytics point of view, nice people like auditors are often not very prone to argumenations like that.
During development stages, I'd just use the method as you do. But when it's getting serious, I'd do complete validation including selectivity. Call it regulatory paranoia...

Agree. Since you'd need to validate anyway as it's now an API in a finished product, even if you made zero changes to the compendial procedure, you'd still need to validate. So make your mobile phase adjustment as necessary to achieve required robustness and selectivity.

Agree. Since you'd need to validate anyway as it's now an API in a finished product, even if you made zero changes to the compendial procedure, you'd still need to validate. So make your mobile phase adjustment as necessary to achieve required robustness and selectivity.

Thanks for all your responses.
To further comment on the need to validate anyway:
We already got away by doing only precision and accuracy on our finished product.
Accuracy was done by doing side by side analysis of raw materials and raw materials in the sample matrix.

Maybe we were just lucky, but we saved some time!

Ace

Definitely lucky you haven't been audited

The EP method is validated to be used for the API, any changes other those allowed by the EP (column length, diameter, particle size, adjustment of the mobile phase...) imply a new method validation. On the other hand if you want to use the method for the finished product a complete validation must be carried out. As Lusby said You are lucky that you haven't been audited.