DNDN: you forget that the TTP for cancer patients in the 9901 was 0.052 (I believe for no progression for the Provenge arm), while a completely different endpoint "survival" was statistically significant (<0.05). What are the odds that you would get these two statistics together in this small trial by chance? That would probably be close to the 10 "reds" on the roulette wheel in a row territory imho. Maybe Walldiver or David M. can offer a probability on this.
Putting Provenge out there isn't going to do a diservice to cancer patients. They can always indulge in the Taxotere treatment soon after Provenge treatment too. Or just not take Provenge based on their doctor's opinion. Pick your poison. Well in the case of Provenge, a reasonably pain-free sword. In a few years this will be resolved with more certainty, but you get the picture I hope.