Is a Real Time Stability Study required to determine Expiration Date?

[Mike Blanc]

Hi,

Currently my product (class IIa) shelf-life is 3 years and we had recently completed a 5 years acceleration stability study with favourable results.

Do I need real time stability study or acceleration stability study is good enough for the extension of shelf-life?

 

[Ronen E]

Re: Is Real time stability study required to determine Expiration Date ?

Hi Mike and welcome to the Cove

Normally you could update your shelf life claims based on the accelerated aging study (provided that the science and statistics are sound), however you'd typically be expected to follow up with a real-time aging study starting before you actually place units with the extended shelf life claims on the market. You should then test a statistically significant sample from this population, in pre-determined intervals (say, every 3 or 6 months) from the end of the previously established shelf-life (3 years). This way, your real-time testing will always be one step ahead of the devices on the market, and should anything go wrong you will be able to take timely action. For this reason I think your real-time aging should be at least 3 months ahead of actually placing units on the market (enough time for testing and acting, should it be necessary).

In the longer term, real-time aging is normally expected for substantiating the validity of the accelerated aging model.

Cheers,
Ronen.

 

[Bev D]

Ronen is correct. While your accelerated test conditions are probably theoretically correct, each product will have different actual degradation stresses and different actual stress conditions. almost all regulatory reviewers and statisticians know this and will insist on empirical evidence that your accelerated stress model is predictive of real life.

 

[Ajit Basrur]

I agree with Ronen and Bev.

Refer this FDA document that states Real Time study is required (see below) - http://www.fda.gov/downloads/MedicalDevices/.../UCM081366.pdf SHELF LIFE OF MEDICAL DEVICES - April 1991

Geoffrey S. Clark
Microbiologist
Division of Small Manufacturers Assistance
Office of Training and Assistance Center for Devices and Radiological Health
Food and Drug Administration

Accelerated Aging Parameters, including information that validates the accelerated system. The results need to be supported by real time testing of shelf life samples to confirm the tentative shelf life data collected from the accelerated tests.

.

 

[MIREGMGR]

I understand that the OP's question pertained solely to EU requirements, but FDA expectations were referenced above. My experience and understanding of FDA's approach to this question is partly different from that of the esteemed prior answerers.

The prior answers are correct in that real time aging is mandatory to verify the results of accelerated aging. The difference is in regard to marketing the device prior to completion of at least some of that real time aging.

I'm at home without access to my annotated files, but see this device-type-specific guidance as an example. Refer to the last paragraph of Section E. Note the important caveat that this paragraph applies only to devices with established materials and designs.

Given this caveat, FDA accepts 510(k)s...at least in my experience...that declare device-sterility and device-functionality shelf lives, and include labeling stating the shorter of those two lives, based only on validated accelerated aging. There is no requirement to have completed any degree of real time aging prior to marketing. FDA however does expect the 510(k) to include a statement that real time aging is underway and will be followed by testing at appropriate time-increments (for instance, typically annually for a five-year-expiration-date sterile disposable) to verify the shelf life declaration, with the results of satisfactory completion of that verification testing placed in the device's records for possible later FDA review.

Regarding the requirement for validation of the accelerated aging process used, my understanding is that this validation must occur before the accelerated aging process that results in determination of the shelf life to be declared in the 510(k). This importantly is why the guidance applies only to established materials and device types. The real time aging that is conducted to verify the accelerated aging is not this validation.

 

[MIREGMGR]

Regarding the OP's actual question, this EK-Med document in German (DEAD LINK REMOVED) additionally may be informative. As I understand it, it clearly requires real time aging to verify the results of accelerated aging.

I note that the EK-Med authors reference the same Geoff Clark (FDA) summary document that was mentioned above.

 

[CBAL08]

Is there any guidance available for Devices such as Stand alone software and Electronic Medical Devices ( Class I ) to have shelf life?

 

[Statistical Steven]

Hi,

Currently my product (class IIa) shelf-life is 3 years and we had recently completed a 5 years acceleration stability study with favourable results.

Do I need real time stability study or acceleration stability study is good enough for the extension of shelf-life?

Just be careful. If you can extend to 5 years based on accelerated testing, and start to market and label with 5 year claims, your real time stability performed concurrently can show less than 5 years of shelf life and you will need to take a field action. I advise people to increase by no more than 12 months passed the last known real time timepoint (consistent with ICH Q1). As you accumulate real time data, you can have future lots with longer shelf life. Just my

 

[MIREGMGR]

Just be careful. If you can extend to 5 years based on accelerated testing, and start to market and label with 5 year claims, your real time stability performed concurrently can show less than 5 years of shelf life and you will need to take a field action. I advise people to increase by no more than 12 months passed the last known real time timepoint (consistent with ICH Q1). As you accumulate real time data, you can have future lots with longer shelf life. Just my

Which points up that it's extremely important for the validation of your accelerated aging process to be real, not just a paper engineering exercise.

If you are not sure from a rock solid validation that your accelerated aging process will provide the same answers as your real time aging, all the way out to the longest stability period you envision ever wanting to use, I would go even further than Steven and suggest that you shouldn't use accelerated aging to determine product stability at all.

 

[Gish]

Given this caveat, FDA accepts 510(k)s...at least in my experience...that declare device-sterility and device-functionality shelf lives, and include labeling stating the shorter of those two lives, based only on validated accelerated aging. There is no requirement to have completed any degree of real time aging prior to marketing. FDA however does expect the 510(k) to include a statement that real time aging is underway and will be followed by testing at appropriate time-increments...to verify the shelf life declaration, with the results of satisfactory completion of that verification testing placed in the device's records for possible later FDA review.

I agree with the statement above and also with the requirement for validation of the accelerated aging process.