There has not been an accepted explanation for the results of the Provenge trials - IMPACT being the main one. But the notion has been that "the data is the data", and the rationale, the mechanism of action, will become evident as time goes on. Part of the puzzle, the puzzling part, is that Provenge did not prolong progression free survival, did not delay progression, but did extend overall survival compared to placebo, the placebo arm. And people who were on the placebo arm, if they progressed and decided to cross-over, did better than those who did not cross-over. The cross=over consisted in the reinfusion of white cells previously extracted, in a product cleverly called Frovenge - frozen Provenge.
These treatments did not have an effect on PSA.
[PSA is created in the prostate cells when the Androgen Receptor, a transcription factor, moves into the nucleus and binds near the gene for PSA and expresses that gene - ie creates the mRNA that is edited and exits the nucleus, enters the ribosome and where the PSA protein is constructed very rapidly out of material delivered by tRNA. Made by magic essentially.]
That is to say that Provenge does not seem to have an effect on the activity of the Androgen Receptor as reflected by PSA.
Rather than being a medicine in the Androgen Receptor line of treatments, its approach is to try to get the body to see PAP as a foreign molecule, to see PAP as an antigen, and so to turn against this chemical.
So while prostate cancer can be attacked by using its prostate features or its cancer features, Provenge, like ADT, seems to focus on the prostate features, but not on the AR; not on the hormone mechanism that ADT uses.
Another clue, as to how it works, is that Provenge seems to work better in older men than younger men, according to one analysis of the IMPACT data released. It seems to not work at all, in fact, in younger men (<65), as opposed to the matching group in the placebo arm, is the claim.