Dr Eleni Efstathiou – MD Anderson Cancer Center, Houston, USA
Prof Boris Hadaschik – Essen University Hospital, Essen, Germany
Prof Johann De Bono – The Royal Marsden Hospital, London, UK
Prof Nick James – University of Birmingham, Birmingham, UK

EE: Hello, I’m Eleni Efstathiou and in this new COVID era I’m actually stuck in Houston. So if you see me floating around it’s because a storm is ongoing and you might hear some guest-starring cats or peacocks coming through so don’t pay attention. More importantly, I’m joined at this ecancer video by colleagues that really don’t need any introduction. The star of this ESMO has been Dr De Bono, Johann De Bono from the UK, Boris Hadaschik who is also from Europe, from Germany, a urologist, and, of course, Nick James who updated us this year with data from the STAMPEDE trial that I think are quite important for refining the therapy development in prostate cancer. So we’re gathering through Zoom today to actually discuss this data and get more of an insight and intimate, let’s call them, thoughts on the outcomes.
We were discussing before starting how we were going to go about it so we were thinking of culminating towards the more exciting and trendy new data that came through. So let’s start with what we already know and I will turn first to Nick to give us a little bit of an overview of the refinement of the data of STAMPEDE and what is being offered to us through these analyses and then get everyone’s thoughts. Nick?

NJ: Thank you very much, a pleasure to be here today. The STAMPEDE data we presented this time was just the metastatic patients in the abiraterone arm. So we’ve separated out the M0 patients for a separate later pooled analysis with the abiraterone/enzalutamide data, probably next year, in the hope that we can potentially nail a survival advantage in the M0 subgroup with the two arms together. The M1 patients, the thing that is particularly important is what happens in the low risk group because the LATITUDE trial clearly was confined to the high risk group which is only around about 40-45% of the total on our own assessments of the STAMPEDE population. Therefore for regulators and so on, purchasers, it’s important to know whether the benefit is the same in both subgroups. There’s no particular reason why you’d expect the biology to be different but that was the thing we wanted to look at.

That then sits alongside the data with apalutamide and enzalutamide which are obviously a different class of drug but show similar magnitude benefits in the same setting. The other thing I’d like to highlight, which we presented at GU ASCO but it sits alongside the data Boris is going to talk about, is we did comparative quality of life out to two years for the STAMPEDE abiraterone and docetaxel patients in comparison to ADT. There are some striking similarities between what we showed at GU ASCO and what Boris showed with the apalutamide data from SPARTAN.

So basically we’ve got choices now, we’ve got a lot of agents that prolong survival, they have different characteristics. So one of the ways to choose is between the different side effects and quality of life and so on profiles of these and also to ensure that they have the same benefit in different subgroups.

EE: Thank you for that, Nick, and I wanted a little bit Johann’s comments because it’s actually quite interesting – you brought up a specific point and we’re in a transition of times. First of all, we need to reach the level where our physicians, our community practitioners, are actually considering using two rather than just old-school ADT for sure. But we’re still in this era, and this is not judging in any way, but it was the era where we would go for a one fits all approach. We’re getting used to this but now with the data that we’re seeing in CRPC we’re going to have to transition to a more precise approach to life that’s not just about the safety, as you pointed out, and the specific characteristics of the patient but is also about the numerator in the therapeutic index which is the efficacy part. So, Johann, how difficult is this transition going to be, given where we stand right now?

JDB: I think there are many challenges. Clearly we have to do the trials and show significant benefit for patients. We will then have to educate pathology labs and ensure good preservation of tissue and nucleic acids and tissue. Then we have to educate the urologists and oncologists regarding how to manage these complex diseases, looking not only at inter-patient heterogeneity and patient selection for treatments but also the intra-patient disease emergence or sub-clone emergence.
When we first developed abiraterone we reported many years ago that the PI3K pathway was associated with a shorter time to progression in abiraterone, a lower response rate. So we’ve known for a long time that this is a particularly bad subgroup of disease with PTEN loss or PI3K pathway activation. What the ipatasertib phase III has shown is that if we block AKT signalling with an oral TKI tyrosine kinase inhibitor in this disease with the AR blockade, with abiraterone, we can improve rPFS.

Now, there are many things left to do with this set of data but what overall these data do indicate is that combined AR/AKT blockade could become a way forward in the future as we move forward. We do have to further refine the biomarker for this trial but I think that there’s a lot of potential here.

I should say that this class of drugs does have significant side effects. Another key issue will be managing those side effects, particularly diarrhoea and skin rash to a lesser extent and hyperglycaemia. Some drugs are more toxic than others; some of these drugs can cause significant morbidity although I think ipatasertib, the drug in this trial, is one of the best tolerated AKT inhibitors that I’ve certainly worked with. So I’m really quite hopeful that this combination will become a way forward for our patients in the future.

EE: And I’m going to come back to this great data that you presented recently but my main concern, and I’ll take it to Boris who, let’s say, is sitting in the middle and is from a different country. Boris, Nick and the whole STAMPEDE team has mainly brought us into where we need to be in hormone naïve prostate cancer but still if you look at the data from the US we’re not using combinatorial strategies as we should. The use of chemotherapy, Nick, you might have seen the numbers, is less than 5% in this setting and it’s really embarrassing. But, again, we should not be criticizing, there are a lot of obstacles in community practices. Here comes Johann who is continually throwing at us wonderful data that’s putting us fast forward into the future but we know from other solid tumours the uptake is slower and it might be even be slower in prostate cancer. So, Boris, how do you see this transition? We’re not even catching up with what Nick has taught us until now.

BH: It’s a concern that I absolutely share and I think continued medical education and formats like this have to help convince people that earlier treatment and combined intensified treatment really is prolonging life with a good quality of life in the case of, now, Nick’s great data. So I’m a little bit disappointed that we have evidence that is out there that shows poor uptake. My personal feeling is that at least in Germany we are a little bit more open to intensified first line treatment. That might be related to cost in different countries but here we have the opportunity to give both docetaxel and abiraterone and apalutamide, although abiraterone is only approved for the LATITUDE population. So the data now is really important to broaden that field, especially when abiraterone goes generic, I think that is important for the healthcare sector.
I wanted to point out one more thing that Nick did not mention but I believe is one of the most important datasets because if you compare the data just presented now to the docetaxel data, the docetaxel curves after initial therapy came back together after a longer follow-up. Now with abiraterone we have six year follow-up data and the curves stay apart. I think this is so convincing that early intensified treatment is really helpful in that abiraterone, for example, is a great start. So you have to be extremely ignorant not to offer this to the patients. So I don’t know and I sure hope that it’s an issue of generating this kind of real-world evidence data because at ASCO we had data showing that one AR targeted agent after the other is as good as chemotherapy after an AR targeted agent. From the academic perspective we would disagree and choose to change targets but I don’t know myself what to make of the data. I hope that there are flaws in the generation of this real-world evidence because I don’t get it.

EE: So Boris, thank you for the comment on docetaxel versus androgen signalling inhibition because I want to round it back to Nick and get his iteration. Because last ESMO, not this year, last ESMO we were physically present with Johann and myself were sitting next to him when I heard Karim make a comment about the changes with regards to the outcomes, as he perceived them. Nick, what’s your take? There were people who wouldn’t question the use of chemotherapy for hormone naïve disease and I am concerned with that too.

NJ: Yes, so I think there are two separate things really. Just going back to what Boris was saying, one of the problems I suspect urologists have with giving chemotherapy is that it’s outside their comfort zone and they have to send patients somewhere else. Not necessarily in Germany but elsewhere. So having an oral agent that the urologists are familiar with performing just as well is important. At the point where we presented our data last year I thought that docetaxel and abiraterone probably gave similar survival benefit so with longer follow-up from those two arms that’s probably true for the first few years. I think the curves clearly do come together with docetaxel and clearly don’t come together with abiraterone. So it does look as you get further out that the benefits diverge.

So for the patients who survive longest they’re probably going to do better with an oral agent. For the patients with the bad survival it probably doesn’t matter, those curves are probably running together. Then you’ve got to make decisions on other criteria. So cost is one of them and in the UK at the moment, driven by cost, we can only give docetaxel. But the other is things like quality of life and our quality of life data suggests that your quality of life goes up on abiraterone and down on docetaxel for as long as two years from the start. Boris’ apalutamide data looks strikingly similar and the LATITUDE data looks strikingly similar. Quality of life, for me, is a really good reason to make a treatment choice and the treatment choice is in favour of office urology practice as well. So particularly having prominent urologists talk about this is a good way to get it into practice everywhere really. The biggest gain is to be had by applying what we already know to everybody.

EE: Great, thank you for that. I want to bring up, as you all know and some people watching us, now abiraterone is generic in the US and buying a month’s worth of abiraterone is $200. So it’s become the cheapest of all as long, of course, as we’re monitoring our patient. Since we’re in the topic of androgen signalling inhibitors, we saw eventually all the data coming together from all three studies in the non-metastatic, conventionally speaking, as, Boris, you know better than anyone, CRPC. The bigger question for some of us remains even though we saw primary endpoint, secondary endpoints, at the end of the day this long-term treatment with this agent, what’s the impact on quality of life? I know you’re one of the major Principal Investigators in the SPARTAN study and have been involved in the review of the other studies, what are your thoughts?

BH: Thank you Eleni. Here at ESMO Stéphane Oudard presented the updated quality of life data of the SPARTAN trial and, to me, it was very reassuring that we show now with more than double the follow-up, instead of 20 months in the beginning at the first quality of life reports, now we are 52 months of follow-up and we confirm that despite adding another drug, here apalutamide but it’s the same for enzalutamide and darolutamide, in particular that you maintain good quality of life in these elderly gentlemen who start their therapy in the nmCRPC space with a very, very high quality of life, so there’s not a lot of room to improve. When we presented the first quality of life data it showed that our median is higher than the US general population so these are fit gentlemen and they maintain their quality of life although they take one more drug with anti-tumour effect and they take this drug for a median, in SPARTAN, 32 months. So this is reassuring again that early intensified treatment makes sense, you can prolong life. You certainly generate some costs but quality of life is maintained and in the placebo arm there was some deterioration of quality of life starting at one year after inclusion in the study. So we see that the androgen receptor targeted agents have good quality of life and they show that we can prolong overall survival. But certainly if we go further out we have to individualise the therapy more and start with the other drugs. With that we can go over to the real highlights of ESMO.

EE: I wanted to also get usually the UK side of things with regards to the non-metastatic CRPC treatment. I’m going to go back to Nick before we go into the PROfound study and that final survival analysis. Nick, what are your thoughts in the treatment of non-metastatic CRPC and any qualifiers with regard to these outcomes? Because these were men on trials and it’s a little different in reality.

NJ: In my practice these patients have almost entirely disappeared because I’m not at all keen on putting men on ADT for a rising PSA. Obviously nmCRPC is an iatrogenic disease – you create it by putting people on ADT. You have to be mindful of the fact that ADT harms your quality of life, as shown. It drops in the trial, in the control arm and it drops because you gain fat, you lose muscle, your bone density goes, all of these things. You get fatigue, you lose your sex drive, all of these things are bad. We are, without a doubt, over-treating an awful lot of people. What I’m seeing in practice, and my own practice, is that we are tending to track these people and give PET directed, metastasis directed, radiotherapy where we can and a minority will relapse with worse looking disease and have to go straight on to long-term ADT. But the majority don’t need to go on long-term ADT so these patients are largely disappearing in most UK practices, I think. And I think they probably should disappear, I think we’re over-treating them.

The other thing I am concerned about is that there was a PSA doubling time cut-off to go into the non-metastatic CRPC trials but I wonder if in practice how tightly that will be adhered to and people with very slowly creeping up PSAs will find themselves on these agents as well at probably huge cost to the healthcare system and no benefit to the patient.

EE: Nick, you might be more disciplined in the UK but I can tell you, I just saw this week, a man who, as you called it, iatrogenic, you had to use a Greek word, and not just been exposed to that but had received back-to-back in four months, and a physician, abiraterone, enzalutamide, apalutamide. Wildly exploding disease. The reality is that you guys in the UK might be more disciplined but the rest of us in the world still will have a lot of those - Yeah, go ahead Boris.

NJ: I was going to say, paradoxically, although we often have worse access to drugs than you do in the States, we often have much better access to imaging. So we get good access to what you might call fancy imaging and that allows you to make more informed decisions.

EE: You have better access to education, I would say, as well. Sorry, I’d just add, it’s my belief. Go ahead, Boris.

NJ: Sorry, I cut across Boris. Sorry.

BH: No, no, I completely agree but I think the education has to be at the time of hormone sensitive biochemical recurrence. If these men reach castration resistance and a short PSA doubling time, systemic treatment is the way to go. But I fully agree that at time of hormone sensitive biochemical recurrence you have to do good imaging, do a PET then and discuss with the patient whether you do some kind of targeted therapy to not start ADT or whether it’s a case where you start ADT early and then you are in an indication where you could consider the hormone sensitive disease state, adding apalutamide based on the TITAN data. So this nmCRPC niche should go away, I fully agree, because if we have better imaging early we do understand more the disease. Maybe even we have to sequence these men then very early and then decide which way to go.

EE: For the time being, though, we should just remember to not put them on autopilot and monitor them while we’re treating them. With that I’d like to actually go to the exciting data. Johann, back to you and the PROfound data. It wasn’t too exciting to me because we saw the data already presented last year and published and it was a no-brainer we were seeing the survival coming as a benefit even in that interim analysis and even though there was all this crossover. To me it signifies the importance of earlier exposure to agents such as a PARP inhibitor for these specific patients. But walk us a little bit through it and your thoughts for the imminent future.

JDB: I think the first thing to say is that when we started this journey I guess fifteen years ago with PARP inhibitors, we from the outset saw anti-tumour activity in breast, ovarian and prostate cancer. I still, today in 2020, regret it has taken so long to get where we are in prostate cancer because as early as 2007 we were seeing responses in men with BRCA2 germline with single agent olaparib. Nonetheless, through the TOPARP-B trial we have proven that actually olaparib has single agent activity.

But I do want to point out there seems to be a major, how shall I put it, negative movement about treating patients with DNA repair defects other than BRCA. I worry that this is coming from people that have not really treated the patients. Certainly for the BRCA population of patients these drugs are highly active with very durable responses. In fact, the longest time I’ve given olaparib to a patient is over eleven years with a BRCA ovarian cancer. So in these BRCA populations these drugs can work very well. But I do remain convinced in a subset of ATM loss prostate cancer, from my own experience, these drugs can have impressive single agent activity. But I don’t think we’ve figured out the ATM biomarker well enough to date and that definitely needs more work.

So the trial was a randomised phase III study called PROfound which I designed with our steering committee. Two to one randomisation with two cohorts, one was the main primary endpoint cohort with BRCA and ATM alterations, the second was the cohort B, that was the exploratory evaluation of a number of other DNA repair genes, the most important of which is probably PALB2, to a lesser extent RAD51, both uncommon but both important synthesisers to olaparib. The key thing is that the trial allowed crossover. This was because after much discussion we felt it was sensible and ethical to pursue a trial that was open label with crossover at central radiologically reviewed progression. So the primary endpoint was rPFS in cohort A based on central radiological review, OS was a secondary endpoint.

Because of that crossover I always had concerns – would we see an overall survival benefit? I am pleased to say that my concerns were unfounded in that in our New England paper that we’ve just published this weekend concurrently with ESMO you will see Maha Hussain et al, I’m the last author, a significant overall survival benefit that met the predefined, the pre-specified, alpha. The two-sided p-value was 0.02 with a hazard ratio of 0.69, the curves clearly separate for cohort A which is the BRCA and ATM.

Now, let me be very clear, the trial is not powered to look at this gene by gene and if you start looking at the ATM control arm it is a very small group of patients. So those individuals that are trying to parse this out to gene by gene, let’s be very clear, there is definitely single agent anti-tumour activity with both olaparib and talazoparib, two trials that have run and led in this disease, single agent activity that can be durable in some patients now over a year with radiological and CTC and PSA responses. But obviously PROfound arm A is a limited patient population. With PROfound arm B I cannot really say that in this arm we have any evidence for the overall population of benefit but I can assure you that there are subsets in there, PALB2, FANCA, probably as well RAD51, where we do see benefit. The CDK12 arm is still somewhat, it’s fair to say, needing more work. We published in Cell, with Arul Chinnaiyan leading that work, that CDK12 alterations sensitised to immunotherapy, work is ongoing in that space. But I think the key thing here is that the main activity for this drug is in the BRCA and some other genes population.

This trial pre-specified a sensitivity analysis that the EMA have in the past recommended using well-described methods to evaluate the survival benefit if crossover had not happened, so to adjust for crossover. I’ll be honest, I have read the papers, I’ve looked at the analysis, I don’t fully understand the statistics except that it is a recensoring of the data. But you will see that the analysis indicates that there may be an underestimation of the survival benefit because of that very high level of crossover of two-thirds of the patients overall crossed over from the control arm to olaparib which was, in my mind, the right thing to do when we designed this trial.

I think it’s important to note that even in arm B we have separation of the curves, there is no evidence of harm to the patients but there is some evidence of anti-tumour activity for the arm B patients. But the FDA approved this agent for all of the DNA repair genes bar one, the EMA have been much more thoughtful, I think much more conservative. In my opinion we probably should be best served by being somewhere in the middle. But clearly more research work is needed to further evaluate these subgroups of patients.

But I’m proud to say that we now have molecular stratification for prostate cancer with mismatch repair defects, CDK12, BRCA, ATM even, PALB2 and maybe even now PTEN with the AKT data. So the dream we all had to catch up with our breast cancer colleagues is rapidly coming to fruition.

EE: I like that you’re always so optimistic but I think we have ways to go and I’ll go back to Nick with that. We have a lot of ways to go because, as I said, you’re moving us fast forward and potentially these patients who were treated on PROfound are what would be the metastatic CRPC L1, not treated with anything else apart from what they were exposed to in the hormone naïve setting, let’s say abiraterone or enzalutamide and apalutamide. Based on the data that’s coming from 97, as you pointed out, from Rosie Eeles and others, we could see that men who harboured such mutations, especially germline back then,  when we would see who had the most grim prognosis. So Nick, I don’t know if still to this day the UK has updated the guidelines to actually look for germline events, even germline, basic. What’s the story and where are we going with that?

NJ: So patients with a strong family history are able to access genetic services in the UK and that has been well established for a long time. In terms of whether we should be doing it up front, at the moment none of these mutations allow you to help you select an up-front treatment, essentially you’re selecting on quality of life or funding or whatever grounds, not on any molecular test that says you should have abiraterone or docetaxel or whatever. We’ve got analyses underway within STAMPEDE which hopefully will give us some sort of insight into which subgroups might benefit more from one or the other drug.

EE: That’s what I wanted to hear, yes.

NJ: So we’ve got many thousands of blocks pulled in and multiple parallel analyses looking at DNA, RNA and protein level data for these. So I hope we will be able to provide answers to allow us to select patients up front because, coming back to Johann’s point about breast cancer, in breast cancer tumours are classified as whatever – ER positive, ER negative, triple negative and all the rest of it – up front, they’re not classified on relapse. At the moment we have good data on how to potentially break out some subsets on relapse, like the BRCA positive ones would benefit from PARP inhibitors, probably from platinums as well. None of that helps us up front yet. Doubtless it will in due course but we’re decades behind the breast cancer doctors in applying molecular classifiers to the initial therapy of men with prostate cancer.

EE: And I’m going to go to Boris with that because what you mentioned, the family history, we’re way beyond that. We’ve known this data for four years – it has to be any patient with metastatic and I would argue locally advanced disease. These are the men who are enriched in those first two years where you lose patients in your STAMPEDE trial, if we take into consideration a potential non-responsiveness to taxane based on the docetaxel data that I’m looking back to. So, Boris, what’s the status in Germany? Are you moving forward towards that level, at least? I’m not even talking about somatic NGS, I’m talking about basic germline.

BH: Basic germline, it’s also in Germany hard to get an appointment with a geneticist, so what we’ve established in our current practice now, given the data that Johann presented, is that at the time of CRPC… so men get intensified therapy at the start and then at time of CRPC we try to get molecular pathology on board and then generate the data so that for the next lines of therapy after chemo we can choose and we can already contact the payers to approve treatment with olaparib, for example. So this is for now our way to go. With the patients with a strong family history there are some initiatives in Germany right now to try to initiate kind of the Israeli prostate cancer screening trial with MRI and PSA. So, when we identify those men early on this will lead to more genetic testing and then maybe therapy stratification. But, to be honest, in general practice I will be happy if we can get molecular pathology at the time of CRPC but before that I don’t see it right now.

EE: So it’s not happening but based on the data that was presented last year in the PROfound trial we could as early as diagnosis assess the status and use that moving forward, right Johann? We don’t have to go for the very exquisite approach of doing biopsy at the time of CRPC. If we can, like in your institutions or mine, sure. But we could go for that archival biopsy right now.

JDB: Yes, and I think we need to work with our pathologists to preserve these diagnostic biopsies better, to get better nucleic acid preservation. I think once we do that, and this is not going to be costly or difficult, we will get good genomic data with much less in the way of screen fails and quality control fails. But, no, I agree, we can get these from diagnosis although it is possible and, indeed, likely that the diagnostic biopsy will not truly match the CRPC genomic data. So there is merit to our further expanding the study of circulating and CRPC samples because the diagnostic sample may not really give us all the data.

EE: Thank you for that and in the interests of time I wanted your thoughts on two more things, Johann and, of course, wrap up opinions from Nick and Boris on that. Before we go to the IPATential trial again, just to speak to that, to the outcomes and what you expect in the future, I wanted to comment on this mini oral retrospective analysis suggesting, with all the flaws of a retrospective analysis, that you could potentially treat DRD positive patients with cabazitaxel when we have data in the past from others suggesting that taxanes or docetaxel are the right way to go. Because based on the comment you made earlier, I would beg to differ a little bit and say any patient with a DRD positive tumour right now should have the opportunity to be exposed to PARP inhibitor and specifically olaparib since we don’t have any other proven efficacy treatment in mCRPC. I was very worried about that mini oral and I wanted your thoughts.

JDB: I’ll be honest, I don’t like the terms DRD and HRD. I think that cancer is so complex that we just have to focus on detail. I think we have to look at each gene separately, although I do take where you’re coming from. Certainly for many, but not all, of these DNA repair defects, which is what DRD means, PARP or platinum makes a lot of sense. And don’t forget carboplatinum is a good drug, it’s well tolerated and if the patient cannot afford PARP or PARP is not approved for, say, ATM in Europe which is quite likely then I’m giving them carboplatin and that can work very well.

EE: Okay, I’m going to stop you there because you’re preaching to the choir, I use it all the time. But, Nick, is there data to support it? Prospective data?

NJ: Sorry, I’m not quite sure which question we’re answering here. The thing I would like to comment on is just coming back to something that Johann was saying about the archive samples. So we’ve done some work on this within STAMPEDE – we pulled out samples from our huge archive of varying ages and what we found was that when we started sequencing them that the older samples you lost as many as a third for the reasons Johann said which is that the samples degrade. What we found when we did it prospectively and then to pull the blocks in, because we were scoping out doing a PARP based trial, was that the fail rate once you get fresh samples, still formalin fixed paraffin embedded, the fail rate drops to just a few percent. So that does argue in favour if we’re going to be wanting to look at BRCA or whatever down the line, getting it in the can relatively early rather than waiting until they’re castrate refractory five years later when you may be faced with having to re-biopsy because your tissue has now degraded. So I think there’s a good reason to do the up-front sequencing but it doesn’t help you with your initial up-front choices. But I think it’s worth getting that data in the can.

EE: Sure, and we agree. I was more referring to, and I know Johann’s beliefs and I agree, but honestly do we have data on platinum that we should use right now instead of going for PARP inhibition?

NJ: There’s some data.

EE: Some data.

NJ: Talking about preaching to the choir, my experience has been, I suspect, the same as everybody else’s which is that a subset of patients will respond well to platinum and particularly if you know they’re BRCA positive but because it’s such a well-tolerated drug, particularly in young fit men who have had multiple lines of treatment, even unselected it’s worth a trial of therapy in the absence of the ability to get tested because you will see spectacular responses.

EE: Okay, I’ll just turn to the German who is probably more purist. Tell me, do we have prospective data? I use it and Johann does but for everyone listening to us, that’s all I’m saying.

BH: We do use it as well although here in Germany it’s a little bit more complicated because we can use this PSMA radioligand therapy very freely as well. So there are now plenty of opportunities for these later disease status so we go back and try to squeeze out some molecular data and sit together with our pathologists and then we decide what to give to the patient. But we do give platinum as well but I am not aware of good randomised prospective data. There is a lot of prospective data that it works but I’m not aware of good randomised data. So you have to discuss it with the patient.

EE: So all we’re saying is we need the data. Right?

BH: We expect approval of olaparib this year so then this would certainly be our first choice in these patients.

EE: Okay, we’ve completely run out of time but I want in once sentence, first from Johann, ipatasertib. I expected more of a wow difference and you explained to me earlier a couple of reasons why. Would you like to, in brief, say what you think about the outcomes and how we can improve them in the future?

JDB: What we’ve seen here is a 23% reduction in the risk of progression during the trial for the combination of AR and AKT blockade over AR alone in the PTEN loss population. PTEN loss was defined as 50% of the cells having no PTEN staining by immunohistochemistry. That cut-off was perhaps too loose a cut-off, not stringent enough. Our phase II data looked at essentially an H-score of less than 10, which is minimal staining. The genomic data indicated that there was much more separation of the curves, the hazard ratio then was 0.65. So overall we need…

EE: Sorry to interrupt, that was RNA data right?

JDB: That was DNA NGS for PTEN loss by NGS but only for a portion of the patients.

EE: That’s more stringent, essentially, because you’re missing part of…

JDB: So, bottom line, the combination is superior to single agent. We need more data, we need more follow-up and you will need to see, although I have seen it, how we break-up PTEN loss by differing cut-offs.

EE: So, Nick, do you think that this agent is the next one to be approved for mCRPC PTEN loss?

NJ: I think there’s a little way to go yet, isn’t there? I think it’s highly likely. There are other agents competing for that slot of the next approval – PSMA lutetium is going through registration trials, for example, and will report next year. So I think that will be next. So it’s going to get a more and more crowded space and I think we’re going to be segregating the patients up by PSMA uptake, PTEN loss, BRCA and so on into different treatment tracks which is great. I think you’ll get better choices of therapy rather than just try everything on everyone.

EE: And more precise. Boris, final word?

BH: I fully agree and I think it’s a great opportunity. We have seen so much progress and now that we individualise our patient treatment even more it remains an exciting area. I’m very optimistic.

EE: Well on that note thank you all for giving us your insight and thank you those of you who are watching us from home or your offices. I hope you enjoyed ESMO, looking forward to the next one.

NJ: In person, hopefully.

EE: In Paris too.