I guess as a lay person, the way I understood it is this:
1/3 of the group was to receive a placebo- but many ‘crossed over’ at varying points and actually received DCVAX. This makes sense, of course, from a humane standpoint.
However, when the scientists and statisticians go to compare the control against the treated group, is there any way to have a control arm at all? It seems like this would become a very difficult calculation- proving that DCVAX did indeed extend the lives of patients- whether it was given right away, one month in, three months or longer- with such a small number (my understanding is that 331 is a relatively small sample for a stage 3 trial) it would not be easy to pinpoint the efficacy of the treatment.
I completely understand then, the need for the length of this trial and the OS to be calculated to the very last patient- it seems like this would be a pathway to approval.
Would they, in that case, simply compare length of survival for the patients to a ‘typical’ group of GBM patients, not necessarily from the trial? Of that could be done, it seems like the DCVAX patients have MUCH longer life expectancies- which is a fantastic thing.