Assessing an HPLC method for determination of impurities

[Imad]

Dear All,
In assessing an HPLC method for determination of drugs impurities, is it applicable to depend on these impurities relative retetion times, given by the method author instead of using reference standards and to how extent does variation is acceptable

[KC]

I cannot imagine confirmation of identity being performed based on the retention time specified by method author. Even concurrent analysis of a certified reference standard of the compound that matches perfectly does not suffice for some clients if the analysis is relatively non-specific (GC/FID or HPLC/UV). I have used both of these types of analyses in conjuction with one another (when I can detect by both) and it is usually sufficient to confirm identity but if you can go with MS detection, then you have retention time matching as well as mass confirmation. Nobody needs more than that!

[Imad]

Dear All,
In assessing an HPLC method for determination of drugs impurities, is it applicable to depend on these impurities relative retetion times, given by the method author instead of using reference standards and to how extent does variation is acceptable

Thanks KC for ur response, If i made my quiry more specific, I am evaluating method of analysis used for checking impurities of drugs, in some methods authors specify relative retention times of impurities/active for detection and quantification of any detected impurities ie authors use reference standards to identify the their retention times and then depend on their relative retention times for further investigation, hope I didn't complicate the matter

[tom jupille]

Depends a bit on how many peaks you have and how well they are resolved. If few peaks and wide spacing, then specified RRT may be OK. If it were my problem, I would run standards in any case. "Better to be safe . . ."

[ntruong]

Hi Imad,
I have a similar situation: one HPLC method with about ten impurities. All of these impurities were identified, qualified and validated. RRT and RRF of these impurities with respect to the main compound were determined during method validation. After that, we have a very simple assay method where RRT and RRF are used for any detected related compounds. FDA investigators have no issue with this since it is impractical to prepare all standards for every analysis.
Regards,

[Imad]

Hi Imad,
I have a similar situation: one HPLC method with about ten impurities. All of these impurities were identified, qualified and validated. RRT and RRF of these impurities with respect to the main compound were determined during method validation. After that, we have a very simple assay method where RRT and RRF are used for any detected related compounds. FDA investigators have no issue with this since it is impractical to prepare all standards for every analysis.
Regards,

Thanks ntruong,
What u said is what Iam looking for, but still what do u mean by FDA investigators have no issue, do u mean it is acceptable for them. for me working in a regulatory body we accept such method, but in reality when using it and due to variation experienced like columns, instrument etc, we don't get the same RRT and at the same time we know it is impracticable for companies using such method to supply us with these impurities standards. Ur comment will be highly appreciated.

[ntruong]

Hi Imad,
Using RRT (relative retention time) and RRF (relative response factor) to identify and quantify "impurities" is acceptable by the Agency as long as your LC method is validated. I mean you have to identify and qualify known impurities (if you have to synthesize them or if you are lucky you could buy the commercial ones available on the market ) and then validate your method (the RRT won't change that much if you know the concentration of each impurity with respect to the main compound and prepare accordingly). You can still apply the RRT rule even there are some variations in column and instrument (same makers) because the elution order of your main compound and impurities should be the same (if your method is isocratic, you won't see much change, but you might run into problem if it is gradient). Remember, you will have to perform amended validation for every new addition (>0.1%) to the impurity family. You don't need to prepare impurity standard solutions every time
provided there is no change in the validated method.

[Imad]

Hi Imad,
Using RRT (relative retention time) and RRF (relative response factor) to identify and quantify "impurities" is acceptable by the Agency as long as your LC method is validated. I mean you have to identify and qualify known impurities (if you have to synthesize them or if you are lucky you could buy the commercial ones available on the market ) and then validate your method (the RRT won't change that much if you know the concentration of each impurity with respect to the main compound and prepare accordingly). You can still apply the RRT rule even there are some variations in column and instrument (same makers) because the elution order of your main compound and impurities should be the same (if your method is isocratic, you won't see much change, but you might run into problem if it is gradient). Remember, you will have to perform amended validation for every new addition (>0.1%) to the impurity family. You don't need to prepare impurity standard solutions every time
provided there is no change in the validated method.

Thanks ntruong for ur valubale comment, actually I agree with all what u said and I had discussion with my colleagues in work about this issue and I told them i will highlight the matter in the forum to look for a support to convenice them, thanks again.