Once you have the new med you use the courts to prevent the off-label use. See eg lucentis or avastin for Wet AMD.
"Re-patent" would mean something different like Shire patenting Vyvanse by adding an amino acid cap to amphetamines, changing it enough that it becomes a new drug with a new patent. That would require clinical trials and all the expense that goes with them.
[1] https://www.fda.gov/drugs/cder-small-business-industry-assis...
[2] https://www.fda.gov/drugs/development-approval-process-drugs...
[3] https://www.fda.gov/drugs/cder-small-business-industry-assis...
They wouldn't need to modify it for exclusivity via Abbreviated New Drug Application - I just don't see the FDA accepting any evidence short of a phase III clinical trial that HCQ helps with Alzheimers to fulfill the new use requirements. At best they'd be able to run a proper clinical trial for some cocktail of generic drugs and use ANDAs to prevent competitors from using those generics in their own cocktail for a few years.
It's also a highly international forum and divisive topics play very differently in different places. Readers often mistake a comment for something weird and radical coming from a person nearby, when in fact it's an unremarkable comment coming from a person far away. It's unfortunate that this leads to so many brutal online arguments. In person, people instinctively modulate their communication with respect to such variables; online they don't, because the information isn't available and we can't make it available.