Accelerated Stability test for biologics/drugs

Question

Hello everyone, today I received a question from my customer about estimating the expiration date of a drug using data that includes accelerated conditions such as temperature, not just based on the ICH guideline stability test. She mentioned that I should estimate it using the Arrhenius equation in the Fit life by X model. She provided me with a Discovery Summit materials 2019 EU, but current link seems that the demo video and PPT materials have all been deleted. Could anyone provide any insights or resources related to this?

Any datasets with analysis datatable will be appreciated.

https://community.jmp.com/t5/Discovery-Summit-Europe-2019/Taking-Accelerated-Stability-Assessment-to-the-Next-Level-in-JMP/ta-p/156431

They urgently requested a response by today, and I currently lack understanding of this matter. Assistance from an SE who has conducted the same analysis in the past would be greatly appreciated. (raw datasets, analysis workflow, interpret final results for expiration date)

They are currently storing three batches of pharmaceuticals under accelerated conditions (40 degrees) and three batches under long-term preservation (25 degrees). they aim to secure stability data using these conditions. they want to be advised  how to analyze that can handle both temperature conditions simultaneously?

Accelarated Stability Assesment JMP summit 2019.pdf

lkimhui · Accepted Answer

Hi HyeWon,

Recently I encountered a similar request as well.

I hope these links will be helpful for your consumption ;)

https://community.jmp.com/t5/Abstracts/Reduce-Cost-and-Avoid-Nonconformities-Doing-Smart-Shelf-Life/ev-p/708166

https://community.jmp.com/t5/Mastering-JMP/Diving-Into-Accelerated-Life-Testing-for-Product-Reliability/ta-p/669956

DonMcCormack · Accepted Answer

Thanks @Ben_BarrIngh! The Destructive Degradation platform will likely be the easiest way to fit this type of model. As pointed out, it can be done in Nonlinear, which is more flexible but harder to set up. It can also be done in the old Degradation platform using the Destructive Degradation tab. This is not the same analysis as Destructive Degradation and differs slightly, but only offers an advantage for custom models (like Nonlinear). Fit Life by X is not appropriate because it is meant for hard failures (i.e., when something stops working completely) not soft failures (where failure means something is unusable because a measured characteristics cross a threshold, upper or lower).

I am putting together materials on degradation analysis to live in Learn JMP and will let the SE Community know when they're available. @HyeWonJung, your post is very timely. It underscores the need for an easy way to perform this type of analysis. While the methods mentioned above work, JMP users who having been exposed to this type of modeling will find it very challenging and possibly overwhelming. The Project Management teams is aware of this and were working on making this functionality more accessible. Maybe @mia_stephens can lend a little more insight.

DonMcCormack · Accepted Answer

Thanks @Ben_BarrIngh! When it is posted, I'd love to get feedback from you and the others in this thread. One of the challenges with this materials is that it is broader than reliability modeling. It needs to be present it in a way that is accessible to a wider audience, not only what is said in the written materials and videos, but where it lives in the Community.