Exactly. Post hoc and data mining means looking at failed, unblinded data and then mining the other data for results.
The FDA specifically provides room for flexible trials and updates to SAP’s BEFORE unblinding, and considers those clearly as NOT post hoc and it’s not post hoc in fact.
In this case, driven by the scientific world’s change of a definition that MUST apply to this disease going forward for all approvals by the FDA and similar agencies, it would seem requisite and entirely appropriate for the FDA itself to inquire with regard to this redefinition and the company would need to have that data. That it helps potentially to show efficacy is completely irrelevant to the inquiry but not to ultimate approval. The FDA needs to ask that question and the company needs to be prepared to answer it. If it helps illustrate efficacy, then the FDA would necessarily take note. This could annoy some who think otherwise, but their logic simply can’t withstand appropriate, rational scrutiny in the context of rational drug regulation.