Dendreon Corp fka DNDNQ

[walldiver]

OK, I grok what you mean on that first part. As for the second part of your post...

<<Agreed that while weight and LDH were stat sig in 9901 they were nowhere near stat sig in 9902a. But FWIW Location was also stat sig in 9901, but not in 9902a (although it was close). Strictly speaking they went from 5 stat sig covariates in 9901 to two in 9902a.>>

...actually, an argument could be made that p<0.10 could be considered stat sig in a Cox analysis (especially in a trial with only 98 pts), and disease localization was p=0.07. I could have sworn that they have designated PSA, bone mets, and disease localization as the three covariates to be used in 9902B's Cox.