Replies to post #62816 on NorthWest Biotherapeutics Inc (NWBO)

[exwannabe]

Re: crossover and AVI11s post.

Though I agree with what he posts, some here are talking at tangents.

The FDA will accept OS from crossover data. This is easily demonstrated by the example of Provenge, approved in 2010 on a design with OS as the endpoint and crossover (to frozen product of unknown quality) allowed.

BTW, that trial was initially designed with PFS as the endpoint, then changed a couple years into the trial to use OS as the primary.

The earlier Provenge submission also had trials with crossover, but in these the primary was PFS. The PFS endpoint failed, and DNDN was trying to get approved on the OS p value being <.05 Close, but no cigar. That is my basis for saying PFS has to be stat sig (unless NWBO actually does get the primary changed to OS).

In evaluating crossover, the FDA never even looked at if it was causing the trial to look worse. All they looked at was the opposite, could the crossover have made the trial look better (by delaying docetaxel).

Anyway, AVI11's point was that they MUST look at the original randomized arms in evaluating OS, despite crossover. You can not say "Look at those who received DCVAX-l".

In that, he is clearly right.

[Doktornolittle]

I will have to read that a few times, but thanks. As a long, AVII scares the heck out of me, but I am not sure if I am persuaded here. Just because crossover is possible in a PFS trial doesn't mean it should be done, and doesn't mean it is the FDA's business to make that decision.

It would be interesting to see how often the FDA has insisted on crossover when PFS is the primary measure but the drug is not an immunotherapy.

But I will read this a few times. When AVII talks I listen, as much as it hurts sometimes.
----

And of course, I was trying to say Touche' this morning, but my darn auto-spell doesn't speak French.