Note the deliberate wording—there is no assertion here that these 470 patients constitute all of the available data in the Proellex safety database.
I would suggest you are reading too much into this. The wording is almost certainly a result of the fact that they had to draw the line somewhere in on-going trials. Does a patient who has been on drug 3 weeks with no liver problmes qualify as clean? They could clarify exactly where they drew the line - but in all seriousness I don't think it really matters inre the stats.
PS As calibration I would point out that RPRX does not seem to cherry pick data. They draw a reasonable line - but don't describe unless asked. E.g. the endo mITT was a reasonable line and predictable (I predicted almost precisely where they would draw the line) - or at least as reasonable as any post hoc analysis.