"In addition, the company has received the final audited results from the D9901 trial, which confirm the previously reported preliminary results. Comparison of the Provenge treated group to placebo group using the Kaplan-Meier method revealed a clinical benefit, as measured by a delay in time to disease progression, in the Provenge treated group (p = 0.061) that closely approached, but did not achieve, the pre-specified endpoint of the study (p = 0.05). The p value previously reported in the preliminary analysis was (p = 0.085). "
"Placebos were reinfused with their own blood untouched per trial protocol."
I'd have to check but it isn't relevant to my point. In any case, they could have taken a small amount, Provenged it, checked the cd54 result, and indepedently re-infuse the untreated bulk of the sample.
As for the p-value, they found an error in the original analysis of TTP in 9901, which was first reported as 0.063 (or something like that). By the time of submission, they had found that the CRO (apparently) had miscounted. The corrected (and submitted) p-value for 9901 TTP is 0.052. Which is really quite nice, if you think about it.
News 
Market Data 
Discover 
