Prof Kurt Miller - Benjamin Franklin Medical Centre, Berlin, Germany
Prof Silke Gillessen - Kantonsspital St Gallen, St Gallen, Switzerland
Prof Heather Payne - University College Hospital, London, UK
Dr Eric Small - University of California, San Francisco, USA

KM: Welcome to our ecancer session here from the Genitourinary Symposium ASCO in San Francisco. We have agreed upon we start from the left to the right. Silke Gillessen from St Gallen, Switzerland, Heather Payne from London, UK, and Eric Small from San Francisco, he's the local hero. My name is Kurt Miller, I'm a urologist from Berlin, Germany. Just to start with that, Eric, you're not only the local hero, you're also today's hero because you're closest to the data we're going to discuss on non-metastatic castration-resistant prostate cancer. You had the privilege to present the SPARTAN data today, can you give us just a very short introduction on that and then we're going to discuss PROSPER and see what that means for our daily practice.

ES: Sure. So, as you know, the background for this study and the clinical imperative is that patients with non-metastatic prostate cancer, frankly whether it's hormone sensitive or not, but in castration-resistant prostate cancer we know that many of those patients will develop metastases. Work from Matt Smith has previously shown us from a denosumab study who those patients are so we can predict pretty well who is going to develop metastases or die from prostate cancer and it's defined as now by patients with a very short PSA doubling time. So we selected those patients and the principal goal of the study was to delay metastases. The metric that is used is metastasis free survival and the study is very straightforward - patients who had non-metastatic castration-resistant prostate cancer, PSA doubling time of ten months or less, had negative conventional imaging. For some of our colleagues in Germany that may be a strange concept.

KM: We'll come to that, we'll come to that later on.

ES: To choose PSMA-PET but by conventional imaging negative bone scan, negative CAT scans of chest, abdomen and pelvis. Those patients were randomly assigned to receive either apalutamide or placebo and the primary endpoint of the study was metastasis free survival - can we prevent or delay the time to metastasis? This study, as the PROSPER study, was profoundly positive. It met its endpoint easily, the hazard ratio was quite high, there's a two year prolongation in time to metastasis if you take apalutamide. So in terms of the primary endpoint, I'm sure we can talk about some of the secondary endpoints as well, this study really was able to establish for the first time, because we don't have any standard of care in these patients, that we can delay the time to metastases or death.

KM: Silke, I was surprised by the magnitude of the advantage, to be honest, how about you? Did you expect it be so positive in the primary endpoint? We had a couple of negative studies with Eric mentioned denosumab, a lot of things going on. So it took me by surprise, how about you?

SG: Yes, I have to admit the denosumab study was actually positive but it was just barely positive. So that magnitude of effect was really surprising to me as well except that, obviously, these drugs are very, very effective in the metastatic setting and denosumab is not, for progression free survival. So it was to believe… we all believed, I guess, that it could be very positive but that positive was really impressive, I have to say.

KM: By and large when you look at side effects not much new signal here, also not from apalutamide as I see it. The only thing that was a little bit surprising, Heather we discussed it before, was that in the PROSPER trial the patients in the enzalutamide arm once they were off  study there was a 15% death rate which is hard to explain actually. Is there any interpretation here? Is that any meaningful or what does that mean?

HP: I think it was incredibly surprising, especially because we didn't see that signal in the apalutamide, in the SPARTAN study. I think it's something that's obviously going to be discussed as time goes on and we need to be a bit closer to the data to know the reason for that. The reasons given during the session were that the men were on enzalutamide for considerably longer in that group and there was a slight increased risk of hypertension and cardiovascular toxicities but it wasn't huge. So I don't really know, I don't think there are any explanations for that at the moment but I'm sure it's something that may become more apparent with more time and investigation of the data.

ES: I can tell you that for the SPARTAN study, I was just looking it up, the percentage of patients in the apalutamide arm that died without evidence of metastasis, so from some other cause, was 1.1%, so it was quite low and in the placebo arm it was 0.7%, so pretty similar. I also didn't understand that data and I'm sure we'll be getting more information over time.

KM: Eric, you talked about the other endpoints and Phil questioned the endpoints a little bit in terms of what is the real benefit for the patient. If you look at the trend, also overall survival shows some separation of the curve. I'm sure there is a plan to have further evaluation of the data, seeing if overall survival gets significantly positive.

ES: Yes. The thing to remember, again for SPARTAN, is that there are only 24% of events have happened with regards to survival so it's very, very unstable and immature data. Nevertheless there's a hazard ratio of 0.7 and p-value is 0.07 and survival… So I was actually very pleasantly surprised when I saw that because it's certainly trending in the right direction. We'll have to see, it's just way early. One of the unique things about SPARTAN that we were able to do was by providing second line therapy to patients once they developed metastases, label, by the way, on label, it wasn't off label use of abiraterone, the majority of patients have already gotten second line therapy. So I don't think that we are going to be seeing this late effect of survival going away because suddenly all the control patients get treated.

KM: I need to address what is called your PFS2 which is counterintuitive to me because, if I recall your tables correctly, even in the apalutamide arm the treatment number one was abiraterone in second line, is that true?

ES: That's the most common treatment.

KM: Yes, so I would expect abiraterone is not as effective following apalutamide than following ADT and the contra is true because PFS2 was longer in the apalutamide arm than in the ADT arm. How does that…? Is there any explanation for that?

ES: It was certainly one of the most intriguing pieces of information. So PFS2 was measured, was the time from randomisation through the primary event, through metastasis, through the secondary therapy, through progression through the secondary therapy, and that was measured as PFS2. So you're absolutely right that it takes into account time following the primary endpoint, it's obviously approximate to survival. The curves separate and it's statistically significant, the hazard ratio is 50% in favour of apalutamide. So the statement that I made was, I'm not going to get the mechanism yet, the statement that I made was it's fascinating that treatment used in the non-metastatic CRPC setting has an impact on therapy that is administered per label in the metastatic setting as measured by PFS2. Now it's not PSA response proportion, we don't have that data, we are obviously very interested in looking at it, but what it does point out is that these drugs are not mutually exclusive, that you still gain benefit by the late addition of abiraterone. It's very perplexing.

KM: Which is totally counterintuitive because abiraterone following enzalutamide has an 8% PSA response rate, so it's quite surprising.

HP: We're getting two really good endpoints from one study, aren't we, with this, the whole sort of sequencing.

ES: But you're absolutely right, we need to look at the PSA response proportion because we don't know what it is and all we're really looking at is the aggregate time. So it may be that the benefit that you get up front more than outweighs any downside at the back end.

SG: I sat close to Matthew Smith and I asked him the same question because I think I thought the same like you, that the PFS2 is just from beginning the second therapy, because in some other studies that's how it is defined. But what it is defined is really from the start of randomisation so it's treatment one and two.

KM: Yes, correct.

SG: So this is a bit different from when you just think about PFS2 from start of the second treatment, but it is not longer than… if I understood Matthew correctly.

ES: We don't have that data. That's correct.

SG: It's taking away a little bit of the second one but it's still longer, the two in addition.

KM: Silke, Eric mentioned PSMA PET and obviously I learned from your colleague, I think Dr Feng was his name.

ES: Felix.

KM: Yes, and he said, 'Well, we're using it every day here.' So he talked about oligometastatic disease, actually. Silke, are you using it in daily practice? Specifically the typical patient for M0 per se is rising PSA then he gets ADT for rising PSA and at some point you do some imaging. Do you use PSMA PET in these situations?

SG: The thing is actually it's a bit different because the Swiss are a bit reluctant to start ADT early. So we are actually generally not starting in the non-metastatic setting so we don't have that many M0 CRPC patients, just because we don't start the ADT off so early. But coming back to the novel imaging things what's happening now is that not in the CRPC setting but in the castration naïve setting we are starting to do a lot of the novel imaging. We found a lot of oligometastatic disease that is treated all over the place, as we all know, even if we don't have any good data for it. So we don't have that many non-metastatic patients anymore because with the novel imaging obviously you find much earlier a lot of metastases. Just a question would be does it make a difference because you had conventional imaging and some people nowadays do just PSMA PET so they make in conventional imaging M0 patients to metastatic patients and then you would start with one of the novel endocrine agents anyway. You probably do the same in Germany, right?

KM: Yes, we do it all the time. Eric has mentioned, what does it mean, non-metastatic? That's a matter of how closely you look, how many cells you find.

ES: That's exactly right.

KM: So obviously the question will be in the future, Heather, what would you do with a patient, you do a PSMA PET, I know you do, and then you find four to five lymph node mets spread across the body, not a typical candidate for stereotactic body radiation. So what do you do, is it still the apalutamide candidate or is it already the abiraterone? Or you give enza which is then approved for both? What's your strategy in that? What would be your strategy?

HP: I think it would depend what's licensed, really, at the time. But PSMA PET scanning has changed practice considerably and it's certainly changed practice, as you say, for oligometastasis and for patients with rising PSA. But with all of these men in SPARTAN and PROSPER they all have microscopic metastatic disease and it's just finding the correct imaging to be able to detect that. I think treating earlier, delaying the burden of metastasis must be a good thing.

ES: Yes, actually there is a subset of our patients, I don't know the number, it's small, that had PSMA PETs and also had conventional imaging and these are some patients at our institution because we are one of the few places in the US that has this, but also there are German patients on our study, there are Australian patients, so patients did get PSMA PETs. I can't imagine that the outcome is any different for them if they're conventional imaging is normal but we are going to look at that subset. Our standard is exactly what we just said - it's no surprise that these men had micrometastatic disease, they were selected, in a sense, to have micrometastatic disease and either you can't detect them with conventional imaging or you can with a PSMA PET, it doesn't really change matters. So in terms of the efficacy of the treatment I don't think it has an impact. I think in terms of the label, that might become an issue, like how do you label? It's worth pointing out that in the US abiraterone and enzalutamide are approved for metastatic castration-resistant prostate cancer defined by conventional imaging, not defined by PSMA PET.

KM: Yes, but the label doesn't say by conventional imaging.

ES: No, it doesn't, but that's what the studies were, right?

KM: Of course.

ES: And it's sort of what we're stuck with, this anachronism.

KM: At the end you're right, all the data point in one direction - use these drugs earlier. It's in the hormone sensitive setting, it's now in the M0 setting and also additionally what Nick James presented in his talk also using docetaxel earlier, also in the M0 setting. That's what the whole trend goes to, would you agree with that?

ES: Absolutely. Again, the PFS2 data supports that, right? You would worry that the one downside of using the drugs early is that you use them and then you're out, you're done. And the PFS2 data suggests that we don't need to worry about that. So the body of evidence, I agree with you, suggests earlier treatment, less disease on board. It shouldn't come as a surprise, this is true in every other malignancy, right? We're just a little slow in prostate cancer to realise that.

KM: Heather, Phil mentioned the side effects, are you worrying about the side effects of these drugs? Like apa, enza, abiraterione, is that something? You give them for a long time at the end of the day, is that something you're worried about?

HP: Not really. I think it's important to measure blood pressure and I think everybody measures blood pressure on abiraterone but perhaps in enzalutamide or with apalutamide it's something that should be put into the follow-up regime, that we need to monitor them for hypertension because we can treat that.

KM: Silke, will it show up in this year's St Gallen conference if it…?

SG: Yes, definitely in 2019.

KM: This is a completely new field.

SG: So we had it in the 2015 and we decided not to put it into 2017 because there was no new data and we knew the data would come out, it was reading out really fast. So we definitely will have it in the 2019 and discuss what people are doing. But I still think, coming back to the M0, the STAMPEDE data for the M0 with abiraterone looked pretty amazing as well, right?

KM: It does definitely, it's the same, yes.

SG: So that will be even earlier then, and then comes the big question what happens if you have given the abi in the M0 setting for two years.

ES: Good question.

SG: And what's happening now when they get perhaps in the M0 CRPC setting but they had two years or something of abi? It will be interesting to see the new biology.

ES: Yes, the mechanisms of resistance will be very compelling, I agree.

KM: So I think we had an interesting discussion and I think, as always, we had some new answers and many new questions. I'm really looking forward for our next year's discussion here and see what daily practice is like. Until then thank you very much. Thank you.