Biosectinvestor, I have no reason to believe that the DSMB messed up the randomization but that the randomization process was on target. However, perhaps by the summer of 2015 it became clear that the survival of the Treatment group was superior to that of the Control group.
It was therefore decided that it was no longer ethical to enroll additional Control patients but from the candidates already screened, they decided to enroll enough patients into the Treatment group so as to reach the exact number of patients (232) that they had planned to enroll into that group.
Result: they ended up with exactly the planned number of Treatment patients but with only 99 instead of 116 Control patients.
I may of course be wrong but that is the best explanation I have.
Market Data 
Markets 
AI
