Here’s what I’ve found.
FDA Stability Guidance 1998 (Withdrawn)
2500 e. Container/Closure
2501 Changes in the quality of the product may occur due to the interactions between the formulated
2502 biotechnological/biological product and container/closure. Where the lack of interactions cannot be
2503 excluded in liquid products (other than sealed ampules), stability studies should include samples
2504 maintained in the inverted or horizontal position (i.e., in contact with the closure) as well as in the
2505 upright position, to determine the effects of the closure on product quality.
The orientation of the product during storage, i.e. upright versus inverted,
640 may need to be included in a protocol where contact of the product with
641 the closure system may be expected to affect the stability of the products
642 contained (e.g. liquids and semisolids), or where there has been a change
643 in the container closure system.
Couldn’t find any mention of inverted samples
Conclusion: Looks like both a regulation and an expectation. On the basis of the above, any regulator could ask for the inverted data. Some manufactures may, with justification, obtain Lyo/Package contact data in the development phase and use that to skip the inverted testing, but it would be best to check with the regulators before acting on that assumption.
Note that many companies who perform the inverted testing typically do not execute protocols that entirely mirror the upright samples schedule. Often, the inverted sample testing is added to the standard upright sample protocols on a periodic basis. (For example, if the standard protocol was for 0, 3, 6, 9, 12 months and annually thereafter, the inverted samples may also be tested only at 6, 12 and annually) You would need to prepare a justification based on development data and if your studies were intended for registration, speaking with the regulatory authority prior to acting would be wise.
Best regards, John O’Neill