This is Jennifer Williams and we are happy you could join us today for our conference call. Joining me are Mitch Gold, Jon Schiffman, David Urdal and Mark Froelich. (disclaimer stuff)…turn over to Dr. Gold:

Gold: Thanks Jennifer…First let me remind you of the design of this study. The IMPACT study is a randomized, double blind, placebo controlled phase 3 study that enrolled 512 men with metastatic, androgen independent prostate cancer.

The primary endpoint of the trial is overall survival. The statistical plan was designed using the integrated results of two previous Phase 3 studies, 9901 and 9902a. The interim analysis was performed by an independent data monitoring committee, or IDMC. The IDMC reported no safety concerns and recommended that the study continue to its final analysis. While we remain blinded to the data, the committee reported to Dendreon a 20% reduction in the risk of death in the Provenge arm relative to placebo, a hazard ratio of .80. Alternatively the hazard ratio can be expressed as 1.25, representing a 25% increase in the risk of death in the placebo arm relative to the Provenge arm. At the final analysis, which is anticipated in the middle of 2009, if the study demonstrates an approximately 22% reduction in the risk of death, based on 304 events, the company would expect the study to meet its primary endpoint of overall survival.

While we would have liked to have received results that would have allowed us to amend our BLA at this time, as we said before, the final analysis is designed for a higher probability of success. Here are a few key considerations as we look forward to the final analysis from IMPACT. At the time of the data cutoff for the interim analysis, which occurred in May of this year, the median follow up time in the IMPACT study was approximately 24 months. As we reported today, we observed a 20% reduction in the risk of death in the Provenge arm with a hazard ratio of .80. Now if we look at the integrated analyses of results from 9901 and 9902a at a similar follow up time of 24 months, we observed a 22% reduction in the risk of death, a hazard ratio of .78. Therefore, these results are consistent with those from our previous completed Phase 3 studies of Provenge in this patient population at a similar follow up time.

As you may recall, the final analysis from our integrated studies showed an approximately 33% reduction in the risk of death, a hazard ratio of .67, supporting the hypothesis that there is a delayed treatment effect of Provenge, which tends to get larger over time. Given that the results reported today are consistent with those from our previous integrated analyses, at a similar follow up time, we are looking forward to the final analysis of the IMPACT study next year.

Prostate cancer is, without question, a difficult disease to treat, with only one product in the past 20 years that has been approved by the FDA that has shown an improvement in overall survival. Before we conclude, I would like to thank the hundreds of men who have participated in the IMPACT study. In the 10+ years that Provenge has been under study for the treatment of advanced prostate cancer, more than 800 men have participated in the clinical studies. These men have demonstrated courage in their participation in our clinical trials, and have contributed to what we hope will be an advancement in medical care for prostate cancer patients. At this time I’ll turn it over to Q and A:

Mark Monane, Needham and Co.: Thank you for reviewing the results with us…the hazard ratio that you reported, is that an adjusted log rank analysis? I remember you talked about adjusting for certain factors and maybe you could go over those factors and answer that question first.

Gold: I’ll let Mark take that one.

Froelich: The analysis was an adjusted Cox model in which we adjust for the prognostic factors PSA and LDH

Monane: OK, that was helpful. And then in the follow up, I guess the question is, in the final analysis will the patients still alive will continue to contribute data points in the trial, patients that are dead will not, but the question is, in the trial recruitment, the patients that were recruited later in this trial study, can you tell me whether these patients were similar to those recruited earlier in the study? Is there any reason to believe those patients would be different?

Gold: What we’ve said in the past, and what we know is that we’ve conducted a Halabi analysis on the total patient population from the IMPACT study and compared to that from 9901 and 9902a and we see that the patient population is very similar to that what we saw in 9901 as well as the integrated analyses. So the Halabi model shows that the baseline characteristics are consistent with what we’ve seen in our other studies.

Monane: When you say with the 22% reduction in risk you’re talking about a Hazard Ratio of .78, is that how we should interpret it?

Gold: That is correct.

Monane: …and at that point when you say meet the primary endpoint outcome you’re saying that that p value would be less than .05 with a Hazard Ratio not eclipsing 1?

Gold: …um, would meet the pre-specified level of fiscal analysis plan for the final analysis…yes.

Monane: OK. Then we can assume that today the date that you reported today did not meet that endpoint…is that a fair assumption?

Gold: Correct. It did not meet the pre-specified criteria to allow us to amend our BLA. That being said what we can take away from the data that we received on the interim analysis is that it is very consistent with what we saw on the integrated analyses of 9901 and 9902a and that both studies have shown an increased benefit over time. I guess another caveat to that is that not only in the integrated analysis but in 9901 and 9902a we saw the treatment effect increase over time.

Monane: Have any patients been lost to follow up…do you have data points on all of the patients going forward?

Gold: Right now its difficult for us to say but there’s been very few patients lost to follow up…it’s not 100% as it was in 9901 but its very close.

Monane: I’ll get back in the queue…thanks for the added information.

Joel Sendek, Lazard Capital Markets: Can you tell us what the P value was at the interim analysis.

Gold: As we said in the past, we keep the details of the fiscal analysis plan and we look at that as proprietary to the company, but what I can tell you is that the final number of events is 304 events, and we’ve preserved a substantial amount of alpha for the final analysis.

Sendek: That was my second question…which is…what is that?

Gold: Let me just emphasize, as Mark just mentioned, the company did not receive a p value from the IDMC for the interim analysis.

Sendek: Can you tell us what the alpha spend is at this point so we’ll know what you’ll need at the final analysis?

Gold: What I can say is that we’ve preserved a substantial amount of the alpha for the final.

Sendek: In the answer to Mark’s question you said that you’d need a .05, so I’d presume that number would be less than .05, right?

Gold: Mark said .05…we never said .05. I think the best way to look at it Joel is to look at the stats, so if we say we need a 22% reduction in risk of death in the final analysis then we would expect to achieve the statistical hurdle.

Sendek: Presumably what you are saying this morning is that you just missed it. If the interim were .78 as opposed to .80 you would have stopped the trial?

Gold: All I can tell you is that we did not meet the pre-specified criteria for stopping and amending our BLA but I see going forward we need a 22% reduction. We’re at 20% today and, historically we’ve seen an increase in the treatment effect over time in our clinical studies so we’re looking forward to the analysis that we’re going to receive next year.

Sendek: Just one more…so if it’s in line with what you thought but you missed the interim, I’m wondering, did you make a mistake when you came up with your interim statistical analysis plan, you know, because there’s a disconnect there in my mind and I’m wondering if you could explain.

Froelich: The study was powered on the final results of the integrated results of 01 and 02a, which was a risk reduction of 33% so the interim result has less power than the final analysis by design so we still feel reasonably comfortable with our projections for the final analysis

Sendek: Thank you.

David Miller, Biotech Stock Research: The interim would have had to have a much lower hazard ratio than .22 to be successful, correct?

Gold: Much is a descriptive term…

Miller: Would have to have a lower than…

Gold: It would have had to have had a lower hazard ratio than we observed today for us to meet the pre-specified criteria for stopping…

Miller: Right, but you said you have to have a 22% risk of death or a .78 Hazard Ratio, that will be sufficient to get success at the final analysis, correct:

Gold: That is correct. Two things to take into consideration. One is the treatment effect, the other is the increased number of events.

Miller: But if you had shown that .78 Hazard Ratio today because of p value spend and things like that that have to do with the interim it would have not have been successful.

Gold: We have not talked about that specifically, i.e. what would we have needed at the interim to be successful but I think we wanted to give, going forward, that if we have a 20% reduction of risk today, and based on how much alpha we’ve preserved for the final we need to meet a 22% reduction of the risk of death.

Miller: I’m going to move on to something else then. Can any of the patients on the control arm roll over now or are they all past that point?

Gold: The patients that are in the control arm can still roll over if they want and the roll over rate to the frozen version of Provenge has been consistent with what we’ve seen in the previous Phase 3 studies.

Miller: The first number of patients into the trial had a lower Gleason score by trial enrollment and you have said that in our previous trials, the previous trials, that the Hazard Ratio improved, can you talk a little bit about how you would respond to the question or to a charge that this data kind of shows the healthier patients, the patients later on the trial, by definition, by Gleason score, were sicker, therefore its unlikely that they would see the same improvement as the trial goes forward.

Gold: What I would emphasize, first off, is that the patient population as a whole in IMPACT is based on Halabi and is very similar to what we saw in 9901. And 2nd what I would point you to…I know you’re familiar with this poster, this poster that Dr. Aaron Small presented at the ASCO Prostate Cancer meeting a couple years back, we showed that biggest percentage differences in survival were in patients with the highest with the highest Gleason scores. So the take home message is that the survival benefit that we’re seeing is independent of Gleason score.

Miller: My last question is…given that there is this clear tail effect, and that the data looks better going forward, the longer out you go, are you worried or do you have any second thoughts about moving that final analysis forward?

Froelich: The final analysis is pre-specified under the protocol to happen at 304 events so we can’t change that at this time but as I noted before we comfortable that the trial was
adequately powered for an overall reduction of 31% which is consistent with what we saw in our prior trials and integrated analysis.

Gold: And David what I’d like to emphasize is, I don’t know if you picked it up in my prepared comments, the data cutoff date for the interim actually occurred in May of this year…

Miller: Yeah, I did pick that up.

Gold: So…

Miller: You guys made an amendment that essentially accelerate that final analysis from sometime in 2010 to I assume it’s still second half of 2009…are we still 2nd half of next year for the final?

Gold: We said middle of next year…

Miller: OK, middle. So you accelerated that from 2010. Given that there is this tail effect are you having 2nd thoughts about doing that?

Gold: No, because I think what we see is that a 20% reduction of risk of death today with a median follow up of 24 months is very similar to what we saw in the previous Phase 3 studies and if that occurred in May and the final is going to occur sometime in the middle then I think we’re pretty comfortable with the…if the treatment effect repeats itself and it tends to increase over time in effect as it did in 9901 and 9902a and the integrated analysis then I think we’re comfortable with that.

Miller: I guess I do have a final question. I just want to make sure that I understood the answer to Mark’s first question…this is the adjusted Cox hazard ratio?
Gold: Correct.

Miller: OK. Thanks. I’ll jump back in queue.

Gold: David? The last thing is that when we gave you the .78 number at 24 months for the integrated analysis that was also on an adjusted Cox basis as well.

Miller: OK, perfect, thank you.

William Ho, BOA: Could you discuss the recent failures of some of the other companies in the area of immunotherapy and what perhaps makes you different and increases you confidence in success next year.

Gold: Without getting into specifics, in general, prostate cancer has been a very difficult disease to treat and we’ve several studies out of late that have shown a negative treatment effect on overall survival. We’re very encouraged by the results that we’ve seen today in the interim analysis showing a 20% reduction in the risk of death. I can’t speculate as to why those studies have failed and we’ve seen the data that we’ve seen today, but I think there are some inherent differences that I should highlight. First, we use a recombinant protein as our antigen delivery cassette, and that recombinant protein technology I do think is a key piece of why we are able to generate a substantial immune response against the antigen delivery cassette. Second, we use an ex-vivo process to load up the dendritic cells and as a result we take it out of the immunosuppressive environment that exists in vivo, within the cancer patient, and are able to activate the dendritic cells and we give a very large number of active antigen presenting cells that comes back to the patient. So I think there’s probably a whole host of factors that may be contributing. I don’t think there is anything that we can conclude right now but I think we’re encourage by the results that we’ve seen today, particularly in light of the results that we’ve seen come out over the last year or so.

Ren Benjamin, Rodman: Can you let me know if a futility analysis was part of this interim analysis, and if so, what would have been required to stop the trial because of futility?

Gold: As we’ve said in the past there was no futility analysis as part of this interim analysis.

Benjamin: The hazard ratios that were reported…was in a different manner than it was presented in the past…was there a reason why there was a switch here?

Gold: There is a lot of debate in scientific literature about what is the best way to report these hazard ratios. The way that we reported it today is one that is becoming an increasing trend in the lit. and we felt it was important to go to that trend as we look forward to publishing this data and having it consistent with what the community is seeing.

Benjamin: Based on the rate that you’re seeing right now is there any chance that your final analysis comes in even earlier than what you’re predicting now?

Gold: As we said in our prepared comments we’re comfortable with projecting out to the middle of 2009.

Aaron Reems, Wachovia: Could you provide us with the confidence interval for the 24 month hazard ratio that you provided us at .78.

Gold: I don’t have than right now Aaron.

Reems: OK. With data cutoff coming in May can you walk us through what needed to occur between data cutoff and now, and was it expected to take 5 months to gather that data for the review or were there other things that were affecting things for the Data Safety and Monitoring Board getting together to review that data?

Froelich: Just to clarify, we had to clear the entire data base, not just the survival but all the safety and other info needed to file a BLA should the interim have been positive. That’s the amount of time, with more than 80 clinical sites, to bring all the data in, clean it, and lock the data base.

Monane: In the previous data set we saw some long survivors…a certain percentage of patients were alive 3 years later…can you tell us anything about that with the new data set?

Gold: No. We’re blinded to the data Mark so we have no additional data on, for example, the shape of the curves or long term survivors, etc.

Monane: How about the standard deviation of the trial? Obviously that’s going to be an important statistical analysis. Can you describe that or tell us if that was consistent with what you saw in the previous 9901 analysis?

Gold: Mark, everything that we received from the IDMC is included in the press release today.

Monane: The company published some information about the role of a booster shot and the utility a trial you did in some Provenge patients. Can you talk about that a little more and did any patients get an opportunity to get a booster? I know it would probably be a frozen booster. Was that opportunity given?

Gold: The boosting data that we’ve shown is in the P-11 trial and that is in an earlier stage prostate cancer population. No patients in 9901, 9902a or IMPACT have been eligible for a booster. But we’re happy to talk about the concept of boosting if you’d like.

Monane: So no patients in this trial got a booster, correct:

Gold: Correct.

Monane: Did the company have any expectations of this analysis? Are you happy with this analysis? Disappointed? What have you really learned from this analysis in terms of Provenge development going forward and thinking about it as part of the company?

Gold: We’re encouraged by the data that we received by the IDMC. Consistency and reproducibility are one of the hallmarks of biological activity of any product, and the ability to impact survival, particularly in a disease like late stage prostate cancer, is potentially very meaningful. So the fact that we’re seeing data today—20% reduction in the risk of death—that’s very consistent with what we’ve seen in our previous Phase 3 trials and very encouraging to the company. As Greg and I have said when we meet with the investment community we always position the interim analysis as an opportunity to accelerate the regulatory pathway but that the final analysis, by design, has a higher probability of success. The fact that we shot a 20% reduction in risk today and that target we are shooting for in the middle of next year is a 22% reduction and that in our previous Phase 3 studies we’ve seen this increasing treatment effect over time, I think we’re looking forward to the results of the IMPACT study next year.

Miller: Do you assume that there would be a faster clean of the data next time or would it be back to the five months that it was this time?

Gold: I don’t think we can speculate on that. Mark’s team did a tremendous amount of work to get ready for this interim analysis and he had to do a tremendous amount of cleaning of the data to get it ready for the IDMC so I think it’s difficult to speculate on how long it will take…we’ll do it as rapidly as possible.

Miller: You wouldn’t have to repeat this on the patients that have died already?

Froelich: That’s correct.

Miller: Are you going to meet with the FDA on these data?

Froelich: This data has already been discussed with the FDA.

Miller: Anything you can characterize from their comments on it?

Gold: There really isn’t anything meaningful to discuss other than that we’re encouraged by the data and we didn’t meet the pre-specified criteria to be able to render an amendment to the BLA and we look forward to the final data.

We look forward to the final results of the IMPACT study in the middle of next year. In the meantime, we’ll continue to update you on our latest Phase 2 trials of Provenge and the other products in our pipeline that we’re excited about. Thanks for participating in the call and have a great day.