(focus only on the IMPACT data segment)

Gold: (9:43) …the study (IMPACT) was designed to be powered at 88% with 304 death events. As I said in my JPMorgan presentation in January, we have reached that number of events, and we anticipate the final analysis of these data by the end of April.

So what do we know about the IMPACT study and how does it compare to our other studies? Well, there’s a tool we can use in prostate cancer called a Halabi model. You can plug in the demographic factors that the patient has, and Halabi will give you a predicted survival for the patient. All this is is a way for us to predict…are these patient populations similar based on their demographic factors? One of the concerns for IMPACT has been, “is the patient population from IMPACT similar to what we saw in 9901?”

So what we see here at the top in Study 1 is 9901 (referencing Halibi slide) the median survival predicted by Halabi is 20.1 months. Next is Study 2, which was a smaller study and you can see it was a significantly sicker population, the predicted survival by Halabi there was 18.9 months. The next one is study 1 and 2 together, and Study 3 is the IMPACT study, and this is the final data for Halabi and the median survival is 20.7 months. This is very much in line with the 9901 study, so from a demographic perspective, these studies are similar in their patient populations. I’m going to share with you blinded survival curves from the IMPACT, and I caution you not to read too much into this, but what we see here is encouraging.

When we talk about the demographics we are talking about these patient populations behaving similarly when we look at the actual Kaplan-Meier curves. The red curve is the combined placebo and treatment arms from 9901. The green curve are the combined blinded data, treatment and placebo arms from the IMPACT study and what we see here is the 9902b or IMPACT study essential overlaps the 9901 study Phase III study. The blue curve is 9902a represents a study where were know patients are sicker, and we see that that curve drops substantially below the other two curves.

And finally looking at 36 month KM curves the estimates of overall survival again these are combined treatment and placebo estimates. I’ve already shown you the estimates broken out for 9901 when you separate treatment arm and placebo are, but combined…I just want to make sure we’re comparing apples to apples here…9901, 26% of patients were alive at 36 months, 9902A 28%, 9902B very much in line with what we saw in 9901 with a 26% estimate from KM for 36 month survival. So the take home message is that these patient populations appear to be behaving very similarly.

Then we move on to an interim analysis that we announced in October. The company had as part of its SPA a planned interim analysis of overall survival that was conducted by an IDMC and when they conducted the interim analysis at the 2 year median follow up time timepoint, we saw a 20% reduction in risk of death from the IMPACT study with a HR of .80. So how does that compare to the survival we saw at a similar follow up time from 9901 and 9902a when they were integrated together? It’s very similar. For those two studies at the 2 year follow up point we saw a 22% reduction in the rate of death. So we’re seeing a similar patient population based on demographics and we’re seeing those curves behave similarly based on an integrated analysis of treatment arms and placebo arms, and we’re seeing an interim analysis of overall survival that’s consistent at the two year time point.

Now for us to hit to our final analysis, which we’re going to announce in April of this year, we need to go from the 20% that we already have to 22%. So how achievable is that for us? How hard is it for us to get from 20% to 22%? Well let’s go back to our data--a potential proxy for what occurred on the previous two studies. When we look at the 9901 and 9902a studies integrated together, 225 patients, we saw that during that two year median follow up time there was a 22% reduction in the risk of death, and by the end of that study that 22% increased to 33%, so there’s a 50% increase of the…risk reduction of death. So in IMPACT we need to go from 20% to 22%, or a 10% improvement.