The impact to validation would require a “change management board” decision. This should be done with a variety of people at the company. Based on the changes, assessment, small scale/lab batches, etc. you might feel that the API’s should operate the same. If during document review, chemical analsysis, purity review, manufacturer audit, etc. you might determine that there is more risk that things will change. This would require full revalidation.
Regardless, I need to refer you to stability requirements (not my forte). I believe stability guidelines are clear what changes require stability lots to be tested and how many stability lots are tested. For example flavorants might not require stability, while preservatives do require more stability testing. For API changes (what we have here) stability testing will definitely need to be done (how many? I can’t help you on this).
In my personal oppinion, stabily testing will test much more rigorously than validation testing. Typically validation testing happens at T=0 (right after a lot is made). Stability testing can happen at time points (T=9 months for example). So, I think stability lots have much more information than validation lots.
One possibility is to do one “confimration run” and do increased sampling at T=0. In the validation report you can reference the T=0 results, but also reference the eventual stability results.
I don’t think the new API lot can be released for commercial release until stability is complete - so the validation report can wait (until stability is ready) and not mess up commercial release.
Please plan ahead because obviously, this can take many months to process before commercial release happens - so it doesn’t happen overnight.