Contribution of the product matrix

[KM_USA]

My supervisor and I have been discussing this validation question for cGMP test procedures for over the counter pharmaceuticals in the USA. These are fragranced products, and there can sometimes be tiny peaks throughout a chromatogram, visible only when such chromotagrams are greatly zoomed using the data station software.

I’ve got a question about trace contribution of the product matrix: in this case a tiny peak that is not resolved from the peak of interest. The analyte of interest (the API) is significantly larger than any such response from a product placebo, typically more than 100 times larger or greater.

So the question is a general one: for selectivity, how small of contribution must the product matrix be and still be considered negligible for a cGMP test procedure and its validation? In the modern world of software and zooming, very little is “zero” these days. Thanks, and any official references or guidelines would be great.

[rgamage]

You can define the limit in your SOP. eg. any interfering peak less than 0.5% of API peak is insignificant.

[KM_USA]

@rgamage

You can define the limit in your SOP. eg. any interfering peak less than 0.5% of API peak is insignificant.

Thanks. Question: did you write 0.5% of the API because you feel that is reaonable, or because that is a level generally accepted, or accepted by regulation?

[rgamage]

I haven’t seen a regulatory criteria on this. Generally guideline requires any interfering peaks to be separated. Closest I can recall is allowed placebo or back ground contribution in dissolution which is 2%. This shows there can be some acceptability as long as it will not affect quality and safety. If your peak appears only sometimes and visible only when zoomed, you may want to adopt a condition like I mentioned. If it is consistently appearing, better way is to substract it from main peak and validate your action. The value 0.5% is reasobale as cromatographic data often carries a marigin of error greater than that.

[KM_USA]

Rgamage - thank you very much for your input, and I have shared this with my QA representative who says it sounds reasonable.

The QA representative would like to know if other validation experts on this site agree or disagree, so please post follow-up. I’d be grateful, thanks.

[KM_USA]

@rgamage

I haven’t seen a regulatory criteria on this. Generally guideline requires any interfering peaks to be separated. Closest I can recall is allowed placebo or back ground contribution in dissolution which is 2%. This shows there can be some acceptability as long as it will not affect quality and safety. If your peak appears only sometimes and visible only when zoomed, you may want to adopt a condition like I mentioned. If it is consistently appearing, better way is to substract it from main peak and validate your action. The value 0.5% is reasonable as chromatographic data often carries a margin of error greater than that.

[quote=KM-USA]Rgamage - thank you very much for your input, and I have shared this with my QA representative who says it sounds reasonable.

The QA representative would like to know if other validation experts on this site agree or disagree, so please post follow-up. I’d be grateful, thanks.[/quote]

So, 2.5 months later and no additional responses (but lots of views): can anyone post if their companies do something similar, and/or if 0.5% seems reasonable? Thanks.