News Focus
News Focus
icon url

flipper44

12/27/17 11:52 PM

#150934 RE: ae kusterer #150924

Maybe, but the new guidances seem to be focused on early detection and/or significant site number detection of efficacy -- particularly early OS.

Let's assume NWBO started using a more powerful manufacturing system in Germany. Let's say the sites in Germany started showing remarkable efficacy. So remarkable that it is highly significant against the earlier cohorts in the trial.

There is something pretty brilliant here. You would not have to unblind the trial to determine that there was statistically significant efficacy against the earlier cohorts. This might be advantageous, particularly if the therapy helps the crossovers as well.

AVII and Ex will tell you that you have to achieve the correct endpoint, but the site specific language seems to leave discretionary room for interpretation on a case by case basis. Remember also we just saw Avastin get approved for rGBM even though its primary failed. Instead they used another endpoint.

What would the FDA do if the crossover proved completely confounding in Germany after the imagined manufacturing upgrade, but the German manufacturing improvement proved a quantum leap in therapeutic efficacy over historical and the prior trial cohorts ?

Do they act dumb, coldly follow the rules and doom near term progress for society, or do they work with what they are given? Can they think on their feet, or do more people die for years on end in such a scenario?

(Let's take this a step further. Suppose no efficacy interims were done. Suppose they stop countdown at t minus one second and say, hmm....maybe we should not do that so we can determine whether we need to ask for a new endpoint later? Maybe the endpoint should simply be measured against historical and prior trial cohort OS.)

If we are stupid, we (society) will continue to die early.

Remember, this is only for AA not FA -- under this scenario.
icon url

Doc logic

12/28/17 12:16 AM

#150940 RE: ae kusterer #150924

ae kusterer,

Yes, I think approval based on blinded data would be precedent setting but what if something very unusual was happening in this trial. For example, what if Kat's Cure experience of complete remission was not unique but being multiplied in a significant number of cases? Would a DMC or governing body potentially intervene and to what degree if they did? We know the Germans are very willing to intervene if placebo is considered inferior to treatment. Did they intervene first because of seeing something like this but then FDA withdrew the screening hold later once potential changes made in Germany passed the equivalency test and any potential reservations by FDA about changes made to L in process manufacturing were taken care of? Coincidently that might have happened right about the time that blinded PFS data would have been available and under review. Did NIH get involved in this whole process due to their own contamination problems that led to manufacturing changes of their cellular products which led to their patent involvement in the planned Phase 2 combo trials? I don't know for sure but I can't rule this out either. Best wishes.