Replies to post #985 on Dendreon Corp fka DNDNQ

[walk]

>so, in effect, you are saying that the pooled data's log rank p-value of .011 with an HR of 1.5 (which suggests that patients who received placebo have a 50% greater chance of dying than those who received Provenge) is not significant/insufficient and requires more data?<

Thats a great sentence, which in a nutshell summarizes where dndn stands. I agree with your implication that the pooled data is sufficient for accelerated approval. With this caveat.

Trial 9901 has really great survival numbers and statistics, no doubt. The only problem is that survival was not an endpoint for the trial, and was so small even dndn and E Small the lead investigator both considered it a non- registrational trial. Now it is the pivotol trial, because trial 9902b was modified to a survival primary endpoint and enlarged to 500 patients. Trial 9902a was supposed to be confirmation or supportive. By itself 9902a was not confirmatory but did offer a trend towards survival. The pooled data between 9901 and 9902a is stat sig, but it seems more that trial 9901 rescued trial 9902a statitically rather than 9902a confirming the survival advantage found in 9901. By lumping the 2 trials togeather the non-stat sig trial 9902a becomes stat sig in the pooled database, and the already stat sig trial 9901 only diminishes slightly in statistics.