The people posting on this board follow their stocks closely. If they didn't know about an endpoint, then it certainly wasn't the headliner.
That's a separate question. Your traditional argument has been about the prognostic factors, all of which were prespecified. So was the Cox analysis.
But the Cox was not pre-specified as the primary analysis in terms of efficacy for 9901, 9902a, or the pooled analysis. Anyone claiming it was is either erroneous or doesn't understand the precise meaning of "primary analysis" in terms of a statistical analysis plan. The fact it was prespecified is nice, but it doesn't give Dendreon a free pass to substitute it for the prespecified primary analysis -- which they have never done.
You won't catch me saying this is a pristine statistical submission, and part of the reason it isn't pristine is the reliance on Cox to have 9902a be stat sig for survival. Whether that even matters depends on your view of the definition of "single trial plus adequate supportive data" in the FDA regulations.
Cox isn't required for 9901 or the pooled data, however. Those Cox p-values are simply informative to demonstrate the benefit observed was not due to imbalances between the arms. Every trial ever submitted gets that type of analysis by FDA staffers (and the company). The only difference here is the company released it for us to debate ad nauseum.
Market Data 
Markets 
AI
