>This is an apple-orange comparison with the D9901 data. The logrank p-value of the DN-101 trial was .07, i.e., not statistical significant while the logrank p-value of D9901 was .01. Even after applying a Cox regression (with some unknown model), the p-value in DN-101 only improved to .035. This should be compared with the Cox p-value .002 for D9901.<
So if the follow up to DN-101 eventually brings the log rank p value under 0.05, would you support approval?
> On the other hand, the logrank p-value of D9901 was .01, highly stat sig and, per the FDA sensitivity analysis, the various Cox models developed by both Dendreon and the FDA supported the survival benefit finding.<
How do the FDA's alternate cox models (#2, 3, and 4) that gave p values above 0.05 support the survival benefit finding?
As usual, it is easy to spin things one way or the other.
