flipper44,
Those are the scenarios I contemplated as well. With trial integrity in mind I thought your second option of continued randomization might be the easy solution especially if OS benefit is mostly seen at crossover. If patient outcomes between arms were already significantly different, though at PFS eventing and the benefit was being maintained, I thought they might then enroll only treatment arm patients, your first option, so as to knowingly do no harm. There would be very obvious need for FDA to have more data to put this last option into practice because it would essentially change the trial whereas continuing in a randomized fashion would not create this issue. To me it seems as though they may have waited for the DMC PFS analysis at 248 events to announce the lifting of the hold so as to hide their intentions and better maintain blinding but from the hold to the lift it seems like the Germans at least only wanted treatment arm patients without NWBO's direct knowledge. This might explain why some posters suggested, with alleged supporting evidence, that more patients were enrolled in the US than originally called for in place of the full German quota. I wish we had further validation of this claim. I would also like to suggest an analysis of OS based on the last 31 possibly being all treatment arm. Best wishes.
Market Data 
Markets 
