Further, the OS endpoint change, they flat out said it was based on data from the trial. The high crossover count was data form the trial.
Of course the overall crossover rate was not blinded because the patients had to request cross-over, and almost all did. They knew there were pseudoprogressors who had been incorrectly determined to have progressed. Some crossed over from placebo to dc-vax and some crossed over from dc-vax to placebo. IIRC even some early vaccinated patients had 'hot' dc-vax doses remaining because different numbers of doses were obtained depending on the amount of tumor tissue that was extracted on the original surgery, and they were given those doses until they ran out.
LL said years before unblinding 'everybody is living longer.' They knew who died but until the data was unblinded they never knew of those who continued to survive whether and when they had received dcvax or the placebo.
WRT the PFS endpoint being invalidated due to pseudo-progression, they knew very well from the progression of symptoms that they had a problem because the radiology wasn't correctly distinguishing who actually progressed vs who were pseudo-progressors.
Spin as much as you want. We have clear valid data that dc-vax works as well as the clinicians say it does. I will be very surprised if the RAs will insist on throwing to satisfy complaints coming from those with a financial interest in delaying dc-vax coming onto the market.
