Extension of a stability study for an API

[fall4p]

Dear all,

Firstly I would like to thank you for this fantastic forum.

I was wondering if any of you has any experience with extension of a stability study (for an API).
The norm for ICH is to test up to 5years, which means that protocols for these studies foresee enough samples to be stored (under suitable conditions) for 5 years.
However, if for some reason a manufacturer decides to extend this study after 5 years there would be of course no samples left to perform testing.
My question is: Is there a way to incorporate this description in a protocol, so that enough sample can be stored to be tested after 5 years and avoid deviating from the protocol?

I am looking forward to your answers.

 

[Statistical Steven]

Dear all,

Firstly I would like to thank you for this fantastic forum.

I was wondering if any of you has any experience with extension of a stability study (for an API).
The norm for ICH is to test up to 5years, which means that protocols for these studies foresee enough samples to be stored (under suitable conditions) for 5 years.
However, if for some reason a manufacturer decides to extend this study after 5 years there would be of course no samples left to perform testing.
My question is: Is there a way to incorporate this description in a protocol, so that enough sample can be stored to be tested after 5 years and avoid deviating from the protocol?

I am looking forward to your answers.

You can store material for 10 years if you choose. The ICH Q1 guidance is just a suggestion. The question is why would you want to extend your stability passed 5 years? You can always write a protocol that does yearly or bi-annual pulls from year 5-10 for example.

 

[fall4p]

I think that would be the best thing to do.
Although it might sound strange this is more of a resources/logistic brainstorming question.
For example for the stability of product A, there are X samples prepared for up to Z years but while the study is on-going a decision is taken to extend it to more than Z years. It is sensible to ask why would you extend the study - but let's leave that to the side for now.
So where do I find the extra samples? ..
The only reasonable answer I can think of is to foresee extra material in the protocol "just in case"..
Is this just enough or are there other ways? Does anyone ever had any similar issue?