>>I believe the main question about filing on the 9901 + 9902A data is whether or not the statisticians will emphasize the missed primary endpoint of progression.<<
This is clearly the key issue. FWIW, my view is set forth in #msg-5915049.
As a simplifying rule of thumb, I think TTP (and its endpoint cousins, PFS and DFS) is highly relevant in non-metastatic cancer, but is much relevant in metastatic cancer. In metastatic cancer where survival itself has been calculated, TTP’s importance ought to be de minimus.