Of course you left out the confounding and the "highly successful" was another posters comment, certainly not mine.
This trial is a mess and has so many biases that I highly doubt FDA will buy it. Cherry picking data, like adding the small recurrent GBM arm is epitome of data dredging -- wasn't in the original trial and was ginned up after the data was unblinded when they needed some positive news. It is only 64 patients, no prior DCVax-L rGBM trial and not a planned protocol -- only data dredged.