Bioequivalence and Bioavailability Forum

Hi Dr_Dan, Hi all,

I was thinking something along the lines of what Helmut wrote, and now I am just thinking loud:

Let's say the CRO has stored the samples in a freezer. Ok they are now XYZ months old and their content of the API may due to stability issues and a ton of other factors no longer be exactly what they were when the samples were analysed for the study report, but they do have some content and they are still real samples.
So let's say we take then out and re-analyse them. And then we re-analyse them again for ISR purposes. We do not try to make any inference re. the differences between the first analysis and the first re-analysis - we only deal with the difference between first re-analysis and second re-analysis. Would that be acceptable?
Perhaps not but it might be the best we can do?!

After all, ISR gives us info about reproducibility (week-to-week or month-to-month) in real samples rather than in spiked samples as those that are used in validation. An argument for rejecting this approach could be to claim that ISR on post-study samples may not reflect ISR when done at the study time, but on the other hand we need to bear in mind that there is sometimes also a considerable time between validation and study production (yes, I know reval comes into play but still there is sometimes a real time gap), so this could be tried.

Now punish me!