Oliver Sartor: Thank you. Thank you. Pleasure to be here. One of the pleasures this conference is almost every time you quote a paper you have the opportunity to have someone in the audience who wrote the paper, perhaps upon, so this will not be an exception. Radium-223 in bone health lessons concerned. I have been active with Bayer in terms of consulting and active in clinical trials. I hope that doesn't influence me and of course there are a variety of other potential conflicts that are listed here.

I'd like to start off by reviewing some data which I think is very important and that's the first use of an alpha emitter and that's called Radium-223 in survival of metastatic prostate cancer. I think this really was the landmark paper because it showed the radiopharmaceuticals in the treatment of bone. One particular organ, in fact, could prolong survival and it turned out that this is a beautiful graph, but of course, there's a little bit of caveats and I'll be covering that here in a moment.

One of the caveats that was raised relatively early on, was the fact that in the lower bone metastasis burden patients less than six, plenty metastasis, the confidence between intervals very clearly overlapped one, even though it was remarkably underpowered in this subset and every other subset, there was a very apparent trend toward benefit including those with the prior-docetaxel use and bisphosphonate use, yes or no, which I think were important to know. When we looked at the symptomatic skeletal events and just for clarification, the majority of these were in fact radiation to bone. There were segregated out pathologic fractures and those pathologic fractures are listed here, down in the bottom right where we have pathologic fractures being 4% and the radium arm and 5% for placebo. So really not a difference in terms of pathologic fracture. Very importantly there were no images done in this particular trial and the lack of the imaging really meant that the ascertainment of fractures that were non-pathogenic were not very clearly annotated and there is actually no report over the fracture risk that occurs in radium because of this lack of the imaging and lack of reporting.

That is a little bit problematic but we didn't think it was too much of a problem at the time because we improved the SSC rate and we improve the fracture rate. And then when we did a pre-specified analysis and you know I do want to be careful about the analysis as subgroups. This isn't how they discussed a little bit earlier but this was pre-specified and we did not run an interaction term because the question wasn't so much how do these two compare? But whether or not you might have an effect on SSC rates in the presence or absence of bisphosphonates, you could see on the right-hand side, the hazard ratio is 0.77 the P-value did not quite reach the magic P equals 0.05 or less, was 0.07 and you can argue perhaps it was a bit underpowered, but clearly you could see in an analogous underpowered subset that with the use of bisphosphonates that the radium group had a pretty dramatic effect, has ratio being 0.49 here and lots of zeros before the 0.048 on the P-value.

So I think one might conclude that the utilization of bisphosphonates in combination with radium led to a substantial reduction in the SSC rate in this trial. So what did we learn? We learned that radium was safe, we learned it was associated with better survival in symptomatic skeletal event rates. Concomitant therapies were fine. That external beam radiation or bisphosphonates given commonly, we're just fine. We could use any old hormone like bicalutamide and dexamethasone, DES, etc. and that was fine too. But there were a whole lot of questions we did not answer with ALSYMPCA cause there was no abiraterone, no enzalutamide and no imaging, very importantly. So clearly we had much more to learn and that's some of the things that we will now be discussing.

Now I'm going to give a little bit of homage to Ken Pienta who is not here but I've always been enamored with his concepts about the multi-targeted therapy in bone, which he and I have discussed over many years and it turns out that within the tumor cell in the microenvironment there are lots of ways that we can intervene therapeutically. Radium of course will have a microenvironmental effect. It actually doesn't even bond the tumor it bonds to the two, the hydroxyapatite in the inorganic matrix, but perhaps using other therapies in combination with radium would augment the effects. And currently that was a very reasonable hypothesis and there were two large trials designed to be able to look at this hypothesis in a little more detail. One of them called the ERA 223 and Matthew Smith sitting over here in the front row happens to be the PI on that particular trial. So if I screw up Matthew, please correct me, midstream if you like. And then of course we have PEACE III and there's somebody called Silke Gillessen and Bertrand Tombal who managed to head that particular trial. So Silke and Bertrand, please correct me there, right there in the front row if I screw up in any way.

Now Fred Saad also, in the front row over here had published some data from an international, early access, open label trial and in this case non-randomized utilization of abiraterone or enzalutamide actually seemed to be associated when given concomitantly with radium with better survival. So what could be better? We have a hypothesis, we have preliminary data, everything seems to be going in the right direction. And then thank you Bertrand for this particular slide from ASCO. And it turns out that there was a little bit of caution that perhaps we should have paid attention to because all of these agents, whether or not they be apalutamide or enzalutamide are all going to be associated with an increased risk of fractures and abiraterone as well. So we perhaps had a little bit of caution there, but not too much. So off we go with the ERA 223, and we were very avidly accruing to the trial. We had a lot of enthusiasm, what could be better than giving radium and abiraterone together as a potential life-prolonging combination of therapies, which certainly had a good rationale. And then all of a sudden there was this call and the call was, "Oh my goodness gracious, the trial was being stopped early by the IDMC".

Well, the first thought that went through my head is, well, maybe it's so positive that they decided to terminate the trial early. That was true for ALSYMPCA by the way, the interim analysis, which had absolutely no anticipation of being positive, that trial was stopped early. This trial was stopped early but it wasn't good news. Now, in fairness to Matthew and the ERA 223 crowd, I'm going to point out Matthew's publication with a little more mature analysis. That doesn't show why the IDMC stopped the trial, but this is survival and a little bit longer follow up, but this is why the IDMC stopped the trial. This was the Kaplan-Meier for fractures. Oh my goodness gracious. I think you could probably drive a truck through those curves. You know that certainly is an important looking difference and then you start to look and you say, "Oh my goodness, look at the fracture rate in the combination trial and, my goodness gracious, all sorts of things are going wrong here". And then you looked at the survival at the time that the IDMC stopped the trial hazard ratio 1.347 and obviously this is not a preplanned analysis this was done by the IDMC, but you'd turn out with a P-value of 0.02 in this unplanned analysis is going the wrong way. My goodness gracious. This is really dreadful. And when the IDMC reviewed, this is actually part of the slides that were obtained, you know, deaths 34.7%, 27.4 hazard ratio is wrong, median survivals, not looking good, fractures 26% versus 8.1. By the way, if you would like to access this data, you can access it through the site in the EMA where they did the safety review in pharmacovigilance. And furthermore, fractures were actually correlated with debt and Fred Saad would say, "I told you so", but it turned out that most of the deaths were actually due to progression, were not associated with the fractures. Nevertheless, you did have a fracture associated with a higher risk of deaths, with the relative risk of about 3.27 in this trial. 

Interestingly, almost half of the fractures were in patients with the less than six metastases. And you know, that was a little bit of a surprise to me. When I first saw that and I said, "my goodness gracious, you know, how do we interpret that?", and this wasn't what I was expecting at all and Matthew's paper and I quoted directly so I can't get this wrong, this exactly what was said 'Among 76 patients with one or more independently assessed fractures in the Radium-223 group, 60 or 79% occurred at a skeletal site with no bone metastasis'. Now where I might differ a little bit, these are often referred to as osteoporotic fractures and I might take a little bit of an exception with the use of osteoporotic in this particular context. I'll just say that they were non-pathologic and there's, you know a mixture of things that can happen that can be traumatic fractures, these fragility fractures, it can be osteoporotic fractures, plenty of things could occur. But the important thing to me was that most of the time this was occurring in the context of a non-metastatic bone site and that I thought was very important.

Now when we began to look at the use of bone health agents, and this is something that of course is very important, we begin to see that several components of the SSE composite were actually reduced with use for bone health agents. By the way, is predominantly the zoledronic acid or the denosumab. It really wasn't much else, so those two agents. And here you can see, and these are not statistically analyzed because obviously, we got lots of subsets, unplanned analysis here, but EBRT was 16.1% and the radium with the bone health agents since 26.4 otherwise, and you can see that if you gave the bone health agents that it was really quite analogous to the non-radium use where the, there was obviously a difference in the, in the radiation with longer-term, the symptomatic bone fractures were pretty dramatic.Look at that line sort of in the middle of the red box. We have 15% without bone health agents and 1.3% with. So that's a pretty dramatic reduction in the risk of symptomatic bone fractures if you're giving these bone health agents. That was pretty convincing to me in orthopedic surgery, which is going to be surgery to bone is going to be nine versus one which is 3.7 versus 1% and no real differences found in cord compression, but I really wouldn't expected that.

The ERA-223 insights to me are as follows, abiraterone and radium started concomitantly and these asymptomatic patients dramatically increases fractures, and possibly deaths in men with this castrate bone metastatic CRPC. So that's obviously not something we want to do and it's not recommended in clinical practice. It shouldn't be recommended in clinical trials. I think this is a pretty definitive study that there really did not appear to be any evidence of benefit, and the fractures occurred across all subsets, particularly those with low volume metastatic disease, but 50% of the fractures were in this low-risk disease and most of the fractures were not at sites' metastases and the bone health agent,s zoledronic acid and denosumab, substantially reduce the risk of fracture.

Okay, now we're going to move over to Bertrand and Silke's trial that was presented, PEACE III, and this is very preliminary results but very important preliminary results consequently it was an oral at ASCO. And the trial, the PEACE III, was very similar to ERA 223. This is going to be the bone predominant metastatic CRPC patients who are asymptomatic, modally symptomatic and this case, the randomization was enzalutamide plus or minus the radium, one to one randomization and moving forward, and it so turned out that the IDMC decided to do an early intervention on this particular trial and using the ERA 223 data and a little bit of early data. And I don't want to go into confidential matters.... Turns out that I'm on the IDMC for this particular trial, but I'll simply say that the IDMC intervened, and with the intervention which was recommendation that the bone health agents be started before radium and be given consistently, there was a dramatic effect.

And here you can see it, and thank you, Bertrand, for this particular slide, in which you see that without exposure to bone health agents that you have a dramatic fracture risk with enzalutamide versus radium even though it's a very, very small number of patients. But you can see in 12 months it was 37.4% which is pretty remarkable. And if you actually gave the bone health agents, that was zero. So that's a pretty substantial difference. And even though this is preliminary and really an unplanned analysis, I think it's a lesson learned. So thank you, Silke and Bertrand, for helping to inform us about this important issue.

I think we can ask if novel hormone agents are used, we can first of all ask should they be used or should we be adding radium? We don't really know, but I'll just sort of conjecture. We know at this point they should not be started in a 302-Cougars sort of setting or ERA 223 sort of setting. But we don't know exactly when the right time might be. And there are lots of different options and perhaps these should be explored or sure have not.

So hopefully I'm well-timed because that little bell meant I had one minute to go, so I'll finish right on time. So the conclusion, I'd say that combinations of radium and abiraterone or enzalutamide, when you used concomitantly, dramatically increase the fracture risk in chemotherapy, naïve, asymptomatic metastatic CRPC. And then most of the fractures in ERA 223 occurred in sites without metastases, and then bone health agents can dramatically reduce the fracture risk. And I'll just simply say, when you're using radium, I think you ought to strongly consider the use of concomitant bone health agents. So with that, I'm all done. And again, thank you for the opportunity to be here today.