InvestorsHub Logo
icon url

flipper44

05/19/14 6:00 PM

#10677 RE: gnawkz #10674

Let me give you just one reasoned (but slightly dramatized) hypothetical to prove why OS should only be used against the historical OS group if PFS is the primary endpoint.


Let's say HISTORICAL standard of care is globally:

6.9 months PFS and 15 months OS. Again, Standard of Care.

Now let's say DCVAX-L = 18 months PFS, and OS = 24 months in the treatment group.
BUT ....Crossover = 6.9 months PFS, but OS = 25 months.

Under the foolish OLD FDA (not the new and improved one) they would say, "sorry, no significant difference in OS between the cross-over treatment group and the treatment group, therefore no approval."

Patients would understandably scream at the FDA, "you idiots!!! The cross-over treatment group and the treatment group OS trounced the historical standard of care OS by 9-10 months!!!"

The FDA can only reach such an absurd result if they primarily compare the treatment group with the cross-over treatment group.








icon url

Doktornolittle

05/19/14 8:02 PM

#10689 RE: gnawkz #10674

gnawz: I have a very strong opinion on whether crossover should prevent early termination for efficacy. Unfortunately, so far, someone that I know has far more familiarity with the workings of the FDA disagrees with me on the point. Scarier, this guy doesn't just know more than me, he seems to know more than just anybody I have heard post. He is on another message board.

I think it is completely obvious that if the crossover group does not fair as well as the experimental group that there needs to be termination for efficacy if the numbers are statistically significant. But this other poster is totally familiar with how the FDA thinks. All I could conclude was that either the FDA doesn't think very well sometimes, or this other poster does not know as much as he pretends to know.

The crossover group might only get 6 months less DCVax-L treatment than the experimental group, but it is early on. And if that proves to be a significant problem for the crossover patients then of course that is a reason to stop the trial.