$EYPT

Sounds about right. Lower bound for non-inferiority is -4.5 letters. Looking at typical wAMD Ph3 trials below, ~300 patients per arm (KOD recent trial similar n), a 1-1.5 letter difference already gets the lower bound CI to ~4.5+. So if $EYPT shows up in a ph2 with 2 letters or greater difference then logic would dictate they'd have to run massive ph3s to hit non-inferiority. Not to mention the typical degradation of effect from ph2 to ph3. Not sure how they are characterizing -1.5 to -2.9 letters as a "great" outcome given this fact pattern which argues the exact opposite