EE: Well, hello, and welcome to the ESMO 2022 edition of ecancer. We are here in Paris, enjoying pretty lovely weather apart from a great convention. Now, I am Eleni Efstathiou, I am a Greek who resides mainly in Houston, Texas, as some of you know, but the gentlemen who are here with me do not need a lot of introduction, although I will give them the opportunity.

GA: My name is Gerhardt Attard, I’m a medical oncologist and clinician scientist in University College London, UK.

BH: My name is Boris Hadaschik, I’m a urologist now working in Essen in Germany. It’s a pleasure being here.

EE: So we do have diversity – urology and medical oncology. I don’t know how you felt about this ESMO meeting but I felt like you’re watching the vol. 1 of a movie, like we were getting data but we’re not 100% there. So a good example, it reminded me of Dune, the movie that was just presented this earlier year and it gave me that feeling – what’s to come next? With that in mind, we decided before having this filming to actually build on what we heard. We’re going to hear it right from the source when it comes to STAMPEDE, there was a lot of interesting data. But I want to hear from you are we there yet, to move beyond just clinical characteristics? So give us a little bit of what was presented at this ESMO.

GA: I haven’t watched Dune, shame on me, I’ll find time. But the science journey is more like a series. We go through these cycles, don’t we, where you have one ESMO with three positive phase IIIs and then there’s a lull and then there’s another peak. Actually the lulls are as important as the peaks because that’s what builds us up to the design of the next generation of trials and then ultimately those positive results. You can really see that and, in fact, not just for prostate cancer, for quite a few cancer types, there’s a bit of lull for phase III trials but there was some really exciting data in terms of cancer initiation that was presented by Charlie Swanton in the Presidential Symposium on Saturday. 

So if we’re focussing in on prostate cancer and you invited me to speak about STAMPEDE, and thank you, that is a major focus of my work. What we presented at this ESMO is two parts. The first was the closure of a chapter, a really thrilling ride with some really high peaks, and that was testing whether adding or combining enzalutamide with abiraterone is better than abiraterone alone or enzalutamide alone. We’ve had so many positive phase III trials for novel hormonal agents combined with ADT in the past 4-5 years and I really think this is closure. We indirectly compared the metastatic patients who had abiraterone and enzalutamide with abiraterone alone and we see no evidence of a difference. Importantly here, if we were going to give the combination which is more toxic and will cost more, we really need big differences, evidence of big differences in treatment effects. So an indirect comparison between two trials is acceptable and there’s no difference. So the conclusion here is the combination is really not an approach to be pursued. We showed this for localised disease last year; many, including you and I, working on the ACIS trial have shown this for mCRPC. So I think we can draw a line on that, if you both agree?

EE: I think so.

BH: Certainly, absolutely. I agree, but on the other hand now you sound quite negative and the positive data that you presented is that longer term follow-up. We learned from your presentation that patients presenting with M1 HSPC, seven years later still nearly half of the population is alive. So this really we made a lot of progress in the last 16 years since CHAARTED. But now the question is we have so many different options to discuss with our patients, and certainly depending on country you are practising in, but you also looked at transcriptomic analysis to look for…

EE: Well, this is the exciting part.

BH: … and maybe find predictive markers which we lack most of the time. We know a little bit of prediction for very late stage therapy with PSMA for radioligand therapy where we can predict response. We have learned this ESMO that the amount of non-regional lymph node metastases is a little bit predictive against docetaxel monotherapy, which hardly any of us is giving at this point in time any more. So, Gerhardt, if you could tell us about the transcriptomic analysis that was also presented in the orals from your group?

EE: And how practically applicable, if it ever comes to that, will it be? Since you used a test that is clinically available.

GA: We’re all aligned on the backbone for metastatic patients is ADT plus NHA. The two pressing clinical questions which really emerged from ESMO a year ago is, a, how many localised patients… or where do you draw the line for localised patients who are given ADT and NHA? At which point do we start to cause more harm than benefit, question one. That was one of the Presidentials last year. Then, the second is who should get triplet therapy, so chemotherapy plus ADT plus NHA? So, starting with the second one, I don’t think we’re going to have a trial that randomises patients to ADT, NHA plus/minus docetaxel; it’s not ongoing, designing that is going to be challenging, it would be great if we find a way. 

So to try and address that question we’re all collaborating to generate molecular data across all our large phase III trials and we’re doing that in STAMPEDE. To allow this cross-trial comparison we’re all using the same clinical test and we’re lucky that we have partnered with really great collaborators, Decipher and now Averacite. They’re analysing our samples using this standardised test and at this ESMO we presented the data on the first 781 patients who were randomised ADT plus/minus abiraterone. 

There are a few really encouraging results there. First, a very high success rate, much higher than we see with genomic analysis or formalin fixed tissue. Secondly, from one test we derive large amounts of information on signatures and profiles, on biological pathways relevant to prostate cancer. Thirdly, we have signatures that are highly prognostic. We tested both the prognostic and the predictive question and none of the signatures have an interaction with treatment effect. I was initially disappointed, I said, ‘Okay.’ Then as I thought about this over the past 4-5 months, we’re adding abiraterone to ADT, they’re both targeting the androgen receptor. This is not ADT versus no ADT as has been studied previously, or docetaxel which works in a different way. So maybe that’s to be expected. But we have signatures that are highly prognostic. 

The question about prognosis leads me to what I referred to earlier, the second burning question in localised disease. Further to your point, Eleni, the signatures are going to have a role in better selecting patients with localised disease. Because the data we’ve presented from STAMPEDE were very high risk. These were far to the right, we’ve proven the efficacy, but how can we better make that decision. Signatures that improve our prognostication above clinical variables better than NCCN risk criteria etc. are really going to help us with making that decision. 

The third thing, and then I’ll hand over to Boris, is treatment de-intensification. That’s the third burning question – stopping treatment – and there the signatures may help. If you need to make a decision at eight months you then can use more up-to-date information.

EE: You see, that’s what I said, Dune number one – you’re just giving me previews for what’s to come. So I think our audience is getting it, that’s what I meant from the beginning. So actually we’re all very excited and what you said, mainly for me and Boris mentioned it, we have so many of our patients live longer and longer but it is undoubted that we have that 20% of de novo metastatic patients or are high risk localised, very aggressive, who do not survive beyond three years. You showed it first, CHAARTED even before that, even with docetaxel. So that’s where we’ll need your insights, probably focussing on those patients that some of us would like to say we’re going to throw the kitchen sink but we still don’t do anything with that. 

GA: That’s interesting. So, yes, you’re right. In fact, we have some molecular information to tell us something about those patients. They have a very high cell cycling signature score, this came out of the transcriptomes, loss of PTEN, loss of p53.

EE: As expected, so what do you do? How do you intensify? I love your comments about de-intensification and the rest. It’s been something I have been also preaching, as you know, for years. That’s why we have several discussions during this meeting. It’s still early days but in the interests of time do you want to make a comment, both of you, on the PEACE-1 data regarding PSA kinetics whilst on treatment?

BH: Thank you. I think that’s an important question and, Gerhardt, you have been challenged in the discussion a little bit because the Decipher test is not available in Europe at this point in time. It’s a great test because it academically provides much information so I’m really looking forward to more of your analyses. 

EE: May I stop you for a minute? May I say that Ki67 data you presented last year might have been easier to go for? Do it in the lab – easy.

BH: So immunohistochemistry would maybe be easier or, very easy and cheap, we know that PSA at certain time points is also prognostic and can help us identify patients that are doing poorly. This year there have been data from the PEACE-1 trial being presented that an eight-month PSA can show us a good separation of the curves. So if the PSA is undetectable at eight months then these patients are doing extremely well and we know that already from TITAN and ARASENS. So it’s not completely new but it may help us to design trials relying on this cheap biomarker…

EE: As a surrogate.

BH: As a surrogate to discuss to intensify treatment in the poor responders. Probably I’m with you that we need to change the mechanism of action and add something new for this patient population that are doing poorly. It was also the poor PSA responders, both in the double and the triple therapy arms, they had a median survival of three years. So chemotherapy was in both arms but the addition of abiraterone did not add too much in this poor risk population. So maybe we have to really find out and focus on these to make them survive longer. So it provides opportunities for the future.

EE: Just a small comment here, PSA is very good for monitoring but it does not replace imaging, just for those watching us. Because sometimes we tend to forget that in the hormone naïve setting and I’ve had some unpleasant surprises. In the interests of time I want to move over to the big ambiguity in discussion these days. Who would have thought that mCRPC would be sexy and exciting again? I thought we were done with it. So now during this ESMO we had the sequel of the PROpel trial and actually would you like, Gerhardt, to give us a little bit of an overview of the PROpel? Did you want to make a comment before we go there?

BH: I think I should give the overview and then you comment, Gerhardt.

EE: You want to give the overview. Okay, you’re both going to give comments.

BH: Yes, so we are waiting for it to get some more granular information because the first presentation at ASCO GU of the PROpel data, they were really marketing the all-comer population.

EE: Remind to our audience what it is about because some people…

BH: Yes, yes. So PROpel is a first line mCRPC study and it has tested whether the addition of olaparib to abiraterone is better than abiraterone alone. But this patient population is a little bit outdated because most of the patients had ADT only to start with and roughly 20% had some docetaxel before. But there was no NHA before. So first-line mCRPC and the primary endpoint is radiographic progression free survival, looking both blinded and investigator. We had the data showing a hazard ratio for the all-comer population of 0.66 earlier this year and now the update confirms this response rate, a hazard ratio of 0.67 in the all-comer population. 

But we all remember the MAGNITUDE data showing no effect in men without HR defects. So here the exciting part of the PROpel presentation was that we have now got more information on the subgroups. The subgroups of patients with BRCA1/2 mutation really, in my opinion, drove the results because there we have a fantastic hazard ratio of 0.2. So really now the big question is do we need all that expensive molecular testing or don’t we need it? Is it really a drug combination for all comers or should we more focus on the patients with mutations, whether that might be only BRCA1/2 or a bigger panel where the hazard ratio was 0.5? So the overall results are still positive for radiographic progression free survival. The overall survival data, also at this congress, is immature and the hazard ratio there is 0.88.

EE: It’s early.

BH: It’s early but I don’t know whether we will see positivity in the longer run. Reading the question is about subgroups and it’s not approved in Europe yet, this combination. So we also have to wait for the EMA discussions and what they put in the label. But it would be nice from a urology perspective to not have to test, to just prescribe it. But I personally would, at this point in time, more focus on the really mutated patients. But I’m very interested in your…

EE: So full disclaimer, you’re not an actual investigator on either trial, MAGNITUDE or…?

BH: No, no.

EE: So we both are on MAGNITUDE, so that’s our full disclaimer. You alluded to your comment right there, you put something ‘but’ there, you kind of felt it. But, for the audience, the interesting part now, of course, is that the US has already priority FDA review for the combination. So we’re in the throes of approval, essentially. I don’t know about the EMA but you may have some insight. With what you mentioned, Boris, what are your thoughts?

GA: Two comments. As Boris said, this is a population who received ADT alone. If they’ve had ADT and NHA then we’re in the PROfound indication. We already know, you test for BRCA and associated genes, if you find an alteration you give olaparib alone. There’s a licence. I think we have some clarity there, the data from PROfound was very strong. Ideally I would like to see no patient who is going to progress who hasn’t had NHA but we know in the real world that is going to happen for a multitude of reasons. So that population would probably remain. We will probably debate on what proportion of CRPC and we’re going to see global and intranational differences but that population is going to remain.

So, the second comment – at last we’ve seen some granularity. We’ve spent the past year arguing on splitting or lumping. You’re obviously going to test if you have access to a test; more detail is always better. If you don’t have access to a test are you going to try it? But if you can test you are going to test.  What PROpel showed really clearly, really excitingly, as Boris said, if there’s a BRCA alteration the hazard ratio is 0.23.

EE: That was a crazy number, it’s not reached on the BRCA2 yet.

GA: Yes, so we need to be careful because there are confounders there because the analysis is done post hoc with multiple biases in how tissue can be collected, how plasma is collected. Because plasma drives some of the analysis, you’re enriching for patients that have higher tumour fractions, for example. In fact, we see that the median survival in the control arm, so those with ADT abiraterone, was 8.4 months which is shorter than anything we’ve seen. So there are confounders there but nonetheless really exciting. 

The second point, is there efficacy in the HRR wildtype and that’s what we should focus on. Now, there are two hypotheses here. The first is that there are contaminants of genomically HRR altered that would have been detected in a prospective test but have been missed using the retrospective analysis. I agree with Fred that it’s hard to fully explain that hazard ratio of 0.75 just with contaminants because how high a proportion of the patients can that be?

EE: They already detected 28% so how much more could have been?

GA: Yes, so the second hypothesis is that there are either other genomic alterations that sensitise to PARP that we’re not testing for. A couple were suggested by one of the questions yesterday. Or, as was the original hypothesis in the trial, AR inhibition is inducing BRCAness. But we should draw a line under this, that the efficacy in the HRR wildtype is not as high as in the BRCA.

EE: Obviously.

GA: And there is a toxicity cost. So information is going to help if you can get it. What we should now focus on, and hopefully AstraZeneca will subscribe to this, is let’s really look at those HRR wildtype in detail to try and answer that second question.

EE: I think we all agree, then, that of course the answer is not black and white and there’s no discordance. There is ambiguity in the sense of MAGNITUDE dropping that arm early on whereas PROpel went all the way agnostically. So the answer may be somewhere in the middle, right? 

GA: It always is.

EE: It could be everyone is correct and there is a subset of patients in that wildtype that really might benefit and we need to tease them out. Anyone working on a PARP inhibitor, all pharma and investigators, should come together. 

GA: And I think if you’re in a situation where you cannot get a test then there’s a judgement call there. Are you going to use olaparib with increased toxicity? Are you going to try and get the test? What other treatment options do you have for that patient?

EE: Correct, if you’ve already given chemo, let’s say in the hormone naïve, would you go cabozantinib way, what would you do? So, as I said, previews of what’s going to come. Final couple of comments. We heard a little bit of data on metformin again. We went through the disappointment of the big phase III trial on breast cancer being negative, what’s happening in prostate? Do you want to comment on that?

GA: The preclinical and epidemiological data with metformin is always giving these teasers but it’s so hard to run a randomised trial because there are many confounders. But if anyone can grab this thorn it’s the Swiss Cooperative Group.

EE: Absolutely.

GA: Who really now have specialised in these possibly difficult but, what’s the word, niche, boutique…

EE: Elegant.

GA: Elegant phase II trials. They randomised about 160 patients to enzalutamide plus metformin versus enzalutamide. No evidence of efficacy in this small phase II study. They also showed an intriguing but counterintuitive result. In the subgroup analysis by BMI the patients at a higher BMI appeared to have a better outcome. I think we need to be careful about artefacts there but it does remind us, let’s start thinking about BMI in our studies and especially in mHSPC.

EE: Weight loss is a treatment plan here.

GA: BMI could have an impact on outcomes and the discussant, [?? 22:13] raised the question about PTEN loss and its interaction with metabolic pathways. So we need to start, we’ve talked about molecular data, maybe even start throwing factors like BMI into the mix. 

EE: Absolutely. I want to wrap it up going back to good old-school chemotherapy that sometimes has been put down these days that we’re all excited about molecular data. I will turn to a urologist but, of course, a German urologist, to tell us a little bit your thoughts on cabazitaxel. It’s been pushed around a little bit but we lived through recently listening to the ASCO data suggesting there’s some life in that drug there and it’s very interesting. Even the phase II showed no difference with PSMA-Lutetium in all of that. The problem with cabazitaxel is we’re limited by the fact that there is some toxicity and sometimes concern about giving it to patients who are more vulnerable, forget the frails. So there was a presentation discussing the different dosing, what did you think about it in the sense of using it in practice? You do treat with chemotherapy still?

BH: Yes, so in Germany we do treat with chemotherapy as urologists. I think cabazitaxel is a great drug. We have seen the CARD data a little while ago and also you referenced that TheraP.

EE: And we shouldn’t forget that CARD data because some people did forget it.

BH: Yes. The TheraP data showing equal overall survival in a phase II against Lutetium-PSMA. So it’s definitely a helpful and good drug but many physicians and some patients are afraid of toxicity. So in this trial that Stéphane Oudard presented, the older dosing of 25mg every three weeks was compared to a regimen of 16mg every two weeks. Both arms had the support of DCSF and what they nicely showed is that the two-week schedule had really a marvellously decreased number of neutropenic fevers and toxicity. So it was clearly a positive trial with regards to their primary endpoint of toxicity but we have to be aware that it’s not a non-inferiority trial with regards to oncologic effectiveness. We know, yes, for an elderly population this two-week regimen is probably very well tolerated and should be explored and offered to patients but we cannot guarantee that it’s as potent from the oncologic perspective as the standard regimens of 25mg or 20mg. 

EE: This is the kind of study that makes you feel good inside if you’re doing it at home but you’re not 100% sure you’re doing the right thing. It’s like those of us who use 50mg of docetaxel every two weeks for instance. So, in the interests of time we want to thank you all for joining us and not keep you further. We had a great time at this ESMO, even though it didn’t change the lives of our patients yet from a prostate cancer perspective. But it made it a little better with the information we got. Thank you all for joining us.