@ELS @mcnado It's neither in this case. Mycobateria in their latent states exhibit phenotypic rather than genotypic drug resistance, which is the problem with eradication since infected people host multiple sub-populations in different growth sates and you have to keep up treatment until you've gotten all of them (it's much, much easier to kill actively growing Mtb than those in their latent state). I did Mtb research as a postdoc that included studying latent vs. active TB, if you'd like references I'd be happy to dig some up later.