You missed the design of the trial -- it was a NAIVE GBM trial and the crossover was the rescue medication for the placebo arm. Only when the trial was virtually completed and the PFS primary endpoint had FAILED, did they attempt to post hoc dredesgin the trial to add an R/R/ GBM arm using the placebo arm.
It was designed this way, it was post hoc redesigned this way. ECA arms are required to be done BEFORE the trial begins to avoid the inherent bias built into ECA designs (per FDA Guidlance document). NWBO's outside consultants scrambled to assemble ECA dataset without having access to the patient evel data. They actually just created the data of the the graphs in the trial publications. The screeners had variantions -- some permitted biopsy only surgery vs. NWBO's trial requiring complete resection. There are 7 publications that detail the deficiencies in the trial and the bias that makes any conclusions from the trial suspect.
NWBO had done the Interim analysis that gave them a clear view that the primary endpoint was not likely to show a positive outcome. WHen all patients had similar results, it is pretty obvious. So they make the decision to flip the trial into the totally new protocol. It was desperation to try to survive, but they did it. The ECA post hoc already has a significant test product bias (per FDA Gudiance), and now they have to deal with regulators that have plenty of experience watching biotechs trying to spin trials to create positive outcomes. Don't FDA is going to buy it.
Market Data 
Markets 
AI
