RS validation

[surinzp]

Hello all,

I was asked to validate a related substances HPLC method by % area . But there are no listed impurities for this product. In this case what parameters do we have to perform? Please advise.

Thanks

[prepcolumn]

here is the most confusing validation type. First of all, the validation parameters depends on the sample whether it is drug substance and drug product.

The parameters you have to pay special attention are;

Linearity: You should prove that your method is linear from LOQ (or Reporting thereshold) to 120 % of test concentration
Accuracy: if it is drug product you may spike standard to placebo at LOQ, 80%, 100%, 120% of impurity limit concentration. if it is drug substance no way for accuracy (it is assumed accurate through linearity and precision).

Rest of the parameters are generally same..

[surinzp]

Hi

Thanks for your reply. for this product we only have specification limit for unknown impurities as we do not have any known impurities for this product. Do we do linearity and accuracy just for the main analyte in the drug product?

[prepcolumn]

Exactly, you will not have any other option except analyte ..

As i said on the previous post, if your sample is a drug substance (API) you cannot perform accuracy..

[HPLCaddict]

You CAN perform accuracy even for the API only. Of course it's not possible to do this the usual way used for the finished product, e.g. determining recovery in the matrix.
But there are other ways, e.g. comparison of results with another well-established (and validated) procedure (e.g. from a pharmacopoeia) or analysis of a well-defined sample, e.g. a CRS substance. If this is not possible, then of course the only way is to suppose accuracy is given when linearity, selectivity and precision are OK.
By the way, when no known impurities are available, forced degradation studies are a must.

[surinzp]

Thanks very much guys,

I have another basic question. How do we do linearity and accuracy for a dissolution method?

[gtma]

A follow-up question to the initial question....during the validation, the slope of the API at the assay levels is 20% bias high compared to the slope of the API at the impurities levels (surrogate for impurities). Should I be including a correction factor if I want to use area percent approach for impurities? Usually, a relative response factor of 0.8 to 1.2 assume an RRF of 1.0. But this is not RRF cuz it's the same analyte being assessed at different conc. levels. Should I use the same concept as for RRF and don't include a correction factor? I really don't want to use external std at the impurities level to calc. impurities. Not correcting lead to underestimation of the impurities by 20%.....+/- implications as it related to ICH ID, qualification and ability to pass an already established product specification. What would explain a 20% bias high at the assay levels? Not likely matrix effect.
Note: linearity plots at the assay and impurities levels (correlation coefficient, y-intercept,etc) are very good. The response at the 100% assay level is approx. 0.6Au.