NorthWest Biotherapeutics Inc (NWBO)

[MolBiol]

Thanks, I did read it and I couldn't find any language that suggested to me that the threshold for being statistically significant is flexible for the surrogate data used to seek AA.

Pyr wrote

"Reliable" is loosely interpreted with AA. If it was wholly "reliable" (statistically significant, I think you mean--at least it's what I meant), they wouldn't require a post approval (post AA), placebo controlled, confirmatory trial. Follow longusa's link here:

The post-marketing approval is solely based on the belief that the trial will meet its clinical benefit endpoint. That belief is supported by the surrogate data. Therefore, the confirmatory trial is to ensure that the endpoint is indeed reached. If the confirmatory trial fails to support the AA the approval is revoked.

I don't see how boundaries can be set for seeking AA that are more liberal than boundaries set for halting. I believe they would have to be the same. If the data were good enough at 66 events for AA then it would be good enough for a halt since the data are one in the same (surrogate=primary endpoint). I guess then you could go for accelerated approval after a halt.

In my opinion, if seeking AA using surrogate (PFS) data were the strategy NWBO was following then they would have had OS as a primary and PFS as a secondary surrogate endpoint. Once they had strong PFS data they would seek AA while OS data accumulates post-marketing. Since PFS, without any OS benefit, is probably good enough to get approval the way the trial is setup is the smart/safe way. Again, I don't see how you use PFS data at 66 events to get AA while the remainder of the data accumulates on the same endpoint, PFS.