Read the Full Video Transcript

Karim Fizazi: Now, we have about 10 to 15 minutes for Q&A. Please feel free to come to the mic. I see already a question, please.

Anthony Joshua: It's Anthony Joshua, Sydney, Australia. Question for Bertrand. So you have been talking to patients about the De-ESCALaTE study. Janssen are jumping up and down to say they've also been talking to patients, and they're very happy to be randomized to an arm that gives them a 50% rate of gynecomastia. Isn't the best study the one that the patients get to choose what they get to de-escalate with?

So we're still possibly in the wrong mindset completely. What we should be saying to the patients is, your PSA is 0.01. We think it's appropriate to de-escalate. Which of these five strategies do you want to de-escalate with, if any at all? Because a large proportion of patients will say, no, thank you. So I just put it to you and perhaps also to industry that we've got this patient input completely backwards, and we should be asking the patients what they want to de-escalate to.

Bertrand Tombal: So when you speak to patients, they can only answer to questions you ask them. So if you ask them would you stop the ADT, they're going to say yes or no. If you say I'm going to stop everything, they're going to say yes or no. The whole idea about the discrete choice experiment project is to put in place a system where we can assess this question with the patient and ask them what is valuable for them. And in the work, it's done by a young PhD. All these questions will be addressed.

So trying to say, OK, what's the toxicity of ADT? What's the gain of stopping ADT? What's the toxicity of another pathway with ADT and alone? And so we can better understand exactly what we will say, which one stop. Then there is something interesting is that if you ask them to choose and you've not prepared well, what could be the theoretical consequences of the choice they may choose by their own experience? If they experience toxicity, they're going to stop if they worry about overall survival. So I think it's an extremely-- it's an extremely complex field.

There are four strategies, as you said. And I mean, we thought about that initially. We thought about doing the multiple arms trial. But then it's becoming extremely complicated to randomize because, basically, you don't-- you introduce a patient selection, and you would expect a patient with a more favorable response and no toxicity to keep on the drug.

So that's why we understand it's much more complex. We don't understand fully the complexity, and that's why we chose, OK. Industry is doing an approach. We're doing an approach. And in the end, we're going to try to understand better what is there for the patient. But it's extremely complex. But I agree, in the ideal world, we should ask them what is the consequence of your choice. We don't know because we don't have the data.

Nicholas James: Can I comment as well? So STAMPEDE2, we thought about randomizing patients to stop or not, and particularly in the low-volume arm with SABR. And in the end, we decided it wasn't feasible to power it up and stuff. But what we have written into the protocol, particularly on the basis of the abi M0 data, where if you look at the survival, these are clearly low-volume M1s had we done PET scans and stuff on them. And they only had two years of therapy. Is the option to stop at two years? And that option will be the patient's. It's not as much as ours.

And so I think that is an alternate way of testing. I'm sure it's open to criticism, but I think it seems reasonable because I think in the oligomets arm, we're effectively testing the question, can we cure patients with low volume metastatic disease? And I think we probably can cure some of them. We really don't want them on ADT for the rest of their lives, which are going to be very long even if they're not cured.

Karim Fizazi: Maybe a follow up question from online. Dana, I will direct this one to you. So what about re-scanning patients at six months because you talked about PSMA-PET and irradiating remaining lesions in case of a detectable PSA? Is that something you're doing in your practice or not, or would you do it?

Dana Rathkopf: So I find it challenging to re-image-- I mean, for the population that I was talking about, the metastatic hormone sensitive prostate cancer patients, they're starting treatment. And usually, they're responding to treatment. So their PSA is undetectable. And my understanding of next generation imaging, particularly PSMA-PET imaging, is that it's dynamic study. So if the patient is responding and the PSA is undetectable, I don't know what the value is going to be for subsequent imaging.

We heard there's some question about whether AR inhibition can actually upregulate PSMA. So maybe we'll see worse, or I think Michael talked about flare. So it's complicated, I think, re-imaging with next generation imaging. It's something that people are looking at, but I tend not to re-image with PSMA-PET when a patient has an undetectable PSA.

Karim Fizazi: All right. Niraj and then Dan, please short questions, short answers.

Neeraj Agarwal: Yeah. Bertrand, excellent talk on de-escalation. Just a very quick question on use of GnRH agonist versus antagonist when we are pursuing de-escalation approaches. I still remember some patients from SWOG-9346 trial when we were checking testosterone level on GnRH agonist like leuprolide. Testosterone sometimes won't recover for months to years. So just a thought whether we should ideally be using GnRH antagonist for de-escalation trials.

Bertrand Tombal: So one of the requirements to make that study feasible is that it's run as a pragmatic trial. So basically, this is randomized real world evidence. So meaning people-- and we're going to capture the data. So the main biomarker we will capture will indeed be PSMA-- not PSMA, you know that's a lapse-- PSA, and testosterone. And that's why also we had a lot of discussion-- how long should we keep these patients before we stop? That's why we put the maximum duration of 12 months.
Because beyond that, we may have a lot of patients never recovering their testosterone. And then our study will be similar to the Dutch trial where they only stopped the AR pathway inhibitor. And that's why in the trial we will exclude the patient Nick was mentioning, where we believe we have an intention to cure that testosterone. Maybe one day we're going to use oral antagonist to everybody, and we're going to reduce the variation around testosterone recovery. But indeed, we will measure that because that's one of the difficulty when we-- and the correlation with quality of life will be interested as well.

Neeraj Agarwal:  Thank you.

Karim Fizazi: Dan, please.

Dan: Yeah. So quick question on the quality of life measures. You brought this up, the limitations, in terms of interpreting these specific to hormonal therapies. And it's absolutely true. The biggest concern our patients have is the fatigue, muscle loss, the deconditioning. How important is it for us to characterize and collect data on level of activity in our patients in order to put that PR, longitudinal prom into context?

Because as these testosterones recover, patients don't feel better from testosterone alone. They have to recondition. And it's only the patients that are active that are going to feel any difference. But we don't collect-- how important is it in these studies that we collect some level of activity to be able to really focus on the patients that can recover?

Bertrand Tombal: In the new item list, they will be adding very basic question about level activity. But as you understand, it is very, very unprecise. And you remember when we did the quality of life analysis of the Relugolix, we see a totally different effect between patients who had a high baseline quality of life and those who had a low baseline quality of life. So this will be collected. But once again, in a pragmatic trial, you cannot ask like a physical test. But I fully agree with you. We need to collect this data.

Karim Fizazi: Yeah

Gerhardt Attard:  I have a follow-up question for Bertrand on the De-ESCALaTE trial. It's clearly the prickliest bush in the garden. The primary outcome measure. I think I saw that it is time of treatment. Is it time of treatment? Do you measure that from testosterone recovery?

Bertrand Tombal: So once again, pragmatic trial, we have to go for pragmatic endpoint. The first and most important endpoint which is, is it feasible? So the statisticians have calculated that if you have more than 30% of the patients that are backed to hormone therapy within one year of stopping the treatment, arm A will be in any situation equal to arm A. So that's the first one.

The second one, we wanted to choose a very pragmatic endpoint that was time to starting the next overall prolonging agent. So assuming it's a large trial, you're going to have validation. Then because of the regulation in the EU, you need to have a landmark analysis for overall survival at a very short interval, 3, 5, 7 years, to be sure you're not heading into a wall in terms of killing patients. So that's why it's a regulatory endpoint imposed by the EU system. It's to be sure we don't kill patients. So we need to provide a landmark analysis at 3 years.

Karim Fizazi: Maybe Let me take a few questions from online to you, Gert. In hormone sensitive aggressive variants, do you recommend or use sometimes early use of cabazitaxel plus carbo in contrast to classic triple therapy?

Gerhardt Attard:  I mean, we don't have the data for that. I think we don't have enough data to be set and that all aggressive variants are going to do badly. So there's a group of patients there who may have long responses. I feel quite uncomfortable treating them with carboplatin, extra treatments in that way.

Karim Fizazi: Several questions about the practical timing about stopping ADT in patients with mCSPC, and I'm sure we will provide 10 different answers. But maybe, Chris, what do you do in your practice? Do you sometimes stop ADT in patients without major side effects? And if so, when do you consider that?

Christopher Sweeney: My number is 2 years based on if it's a patient who's had some very good response to therapy, micro metastatic disease of sorts. We know the control with 2 years is better than 6 months in the adjuvant setting. So patients who are at 2 years with a completely undetectable PSA, scan negative, I say to them, would you like to consider stopping. I don't demand it. I see where they stand about it, saying there are some risks. But if there is a concern and your PSA goes up and your testosterone has gone up, we'll restart the therapy.

And I've had a number of patients who've been off therapy, and we've all got these anecdotes now after SBRT for low-volume disease, oligometastatic, who are off therapy for many years, and how can we really collect that data. And it sounds like the STAMPEDE2 will help us get some prospective data, but we need thousands of patients to be able to see how many we actually do cure in that manner. My number is 2 years. I could be wrong.

Karim Fizazi: Gert, a question to you about the potential use of cell-free DNA for monitoring during the break if we stop ADT or systemic treatment. Do you think it has a future as compared to simply monitoring PSA or other things?

Gerhardt Attard: Intriguing. I'd love it to. Half my lab works on ctDNA. But I think beating PSA for a patient who has stopped ADT, who has an undetectable PSA, so just a slight rise, it's going to be hard to do that, and especially given the cost of PSA versus ctDNA analysis.

Karim Fizazi: And maybe a last question, which was to myself. Someone was saying, I'm using triplet therapy as a sequential treatment. Should I change? I think the answer is yes. The trials we have used combination triplet therapy, so we should follow evidence, please, versus personal recommendation.