Dr Chris Parker - The Royal Marsden, London, UK
Prof Charles Ryan - UCSF School of Medicine, San Francisco, USA

CP: Hello, I’m Chris Parker from The Royal Marsden Hospital in London in the UK and I’m here in Chicago at ASCO 2017. With me I’m very pleased to welcome Charles Ryan, Professor of Cancer Medicine at the University of California in San Francisco. We’re here to discuss the prostate cancer data. Obviously the most exciting data is STAMPEDE and LATITUDE but we’re going to talk about what you might have missed. So, Charles, what else is exciting in prostate cancer this year?

CR: Well, there are a lot of exciting things going on. We’re seeing the development of new molecular diagnostic tests that will allow us to choose the right therapy for the right patient at the right time. We’re seeing some data emerging with PARP inhibitors in prostate cancer, I should say continuing to emerge with PARP inhibitors in prostate cancer. We’re also getting some refinement on what’s the appropriate length of androgen deprivation therapy for men with high risk localised disease which is the culmination of a long study which started decades ago really. So it’s gratifying to see those come to fruition.

CP: Thank you, yes, we’ll come to that. So tell us now about the data that you’re presenting at this meeting.

CR: I’ve been involved for many years now looking at the role of serum androgens, a very simple test looking at ultrasensitive assays for serum androgens. We looked at testosterone and DHEA and androstenedione. Several years ago we did an analysis where we showed that in CRPC patients with lower androgen levels by these ultrasensitive assays had an adverse survival; we showed that in the chemotherapy naïve patient population in patients receiving ketoconazole. We then went on to show it in patients receiving abiraterone or even placebo on the abiraterone trials and we showed that independent of therapy patients with low androgen levels had an adverse survival. So then what we’ve done more recently is we took ketoconazole and abiraterone out of the equation and we did an analysis of androgen levels as prognostic indicators in patients receiving chemotherapy. What we report here is that the data hold, which is that low serum androgens, in particular androstenedione for some reason, appears to be associated with a poor survival. I think that this is likely not because androgens are protective of survival necessarily in the castrate state but rather that the evolution of a cancer in an ultra-low testosterone milieu is associated with a higher, more aggressive form of the disease.

So by demonstrating this in a phase III trial, which was the CALGB-90401, this was a study of docetaxel versus docetaxel plus bevacizumab. We basically demonstrated that it was a prognostic of survival and therefore we think potentially moving forward this will be one of the stratification variables for clinical trials. That requires a little bit more work; I’m working with Susan Halabi from Duke who has developed prognostic indices in prostate cancer, as you know. So we’re going to see if that is something that we should be routinely testing or routinely stratifying patients by on trials.

CP: Were androgen levels prognostic of response as well as survival

CR: That’s an interesting question and in the abiraterone treated patients it does appear to be associated with response. With respect to docetaxel we haven’t fully done that analysis to know for sure yet. We first did all the survival data and then we’re going to go back and do that later. So that will be a future poster.

CP: Do you think that we should be measuring serum androgen levels routinely in our patients in the clinic?

CR: Of course if we’re treating a patient with mCRPC we would need to confirm that they have castrate levels for a variety of reasons. So that’s a simple answer of yes we should be doing it. I don’t think we’re at the point where we should be doing the liquid chromatography mass spectroscopy assays to look at the panel of androgens. I need to do further work to decide how that fits in to a treatment decision algorithm. We think it’s going to be potentially additive to the prognostic nomograms and of course if you want to make accurate prognostic predictions you need all the components of a prognostic nomogram. So Dr Halabi and myself are working on that very concept right now.

CP: Am I right in thinking there’s no suggestion it’s a predictive biomarker?

CR: I don’t think we should call it a predictive biomarker at this point for non-AR targeted therapies. With respect to AR targeted therapies, patients at the higher range of androgen levels did have a higher PSA response rate. But I don’t think you can make treatment decisions around that.

CP: So while we’re talking about AR targeted therapies we saw some interesting randomised data comparing abiraterone and enzalutamide from Kim Chi. Can you tell us about that?

CR: So this is a very interesting study that we’ve all wanted to see for a long time and it’s not yet fully told. What they did in Vancouver and other sites is randomised patients to abiraterone or enzalutamide, metastatic castration resistant prostate cancer, and he reported the initial analysis of the first line. The trial will eventually report what happens when patients on either therapy progress to receive the other therapy so that will be interesting in terms of our ongoing conversations about whether sequencing is of any value. My own personal bias is that it is probably not and my bias is I think being confirmed by these and other data. But Kim Chi’s data show that the response proportions are fairly similar between the two therapies; the progression free survival was virtually the same between the two therapies. But what was also important is they did a cell free and circulating tumour DNA analysis. This is something to really keep your eyes on over time. What we see is that we are now able to detect circulating tumour DNA in prostate cancer patients. This can essentially, for lack of a better term, be sequenced and we can identify androgen receptor aberrations, BRCA2 aberrations and other mutations like p53 that are common in this disease.  So what he showed was that, number one, the higher proportion of circulating cell free DNA, that is circulating tumour DNA, as that number goes up the prognosis gets worse. So really what that means is this is an indicator of tumour bulk most likely. So that’s point number one and that’s something that we’ll learn more about. Point number two is, for example, they were able to identify BRCA2 mutations in the circulating tumour DNA. This of course may obviate the need for a biopsy which would be good but he also showed that patients who had circulating BRCA2 mutations had an adverse prognosis. That’s consistent with a number of other studies that have shown that the presence of a DNA repair defect in prostate cancer is associated with an adverse prognosis.

Now, if I might, it’s really interesting because Dr Hussain from Northwestern presented a really fascinating study of veliparib, a PARP inhibitor, in prostate cancer - abiraterone plus veliparib versus abiraterone alone. What she showed is that the patients who had BRCA2 mutations in that study actually did quite well on the abiraterone and almost looked like a positive predictive factor. So those two trials’ results were in contradistinction to one another, however one was biopsy based, one focussed on BRCA, one was the circulating tumour DNA based and in an adverse prognostic group. In other words it could be that the patients who have higher levels of the circulating tumour DNA simply have a worse prognosis. But it’s good to see conflicting data in a way because it generates questions.

CP: Yes, although to try and generate a simple message it’s fair to say that the weight of evidence is that men with DRD do worse rather than better.

CR: Yes, exactly. We are seeing the weight of evidence is telling us that this is a poor prognostic feature, that these are common events, relatively speaking, 25% or so of the CRPC population, and we’re getting more and more data on the utility of the PARP inhibitors in this space. That’s something we’re all eager to hear more about.

CP: And at least in our practice when we detect DNA repair defects we are actually changing our treatment. So if we have trials of PARP inhibitors, well that’s well and good, but if not we’re using carboplatin as a single agent and, anecdotally at least, are seeing some very good responses. Do you support that?

CR: Very much, our practice is very similar. I said in a talk that I gave this morning; I said that, although that is true, knowing the DNA repair defect presence earlier is a good thing, it allows you to think about PARP inhibitors, to think about carboplatin, it does not necessarily mean that you need to go to those right away. So Dr Hussain’s data supports the fact that we can still use the abiraterone and similar therapies in that group of patients and that’s actually reassuring because it gives us some space to think about how we plan the next therapy. That’s good, I hope that data are validated in other studies. Just on the final point is that it’s probably not a coincidence because there are mechanistic reasons why the androgen receptor may in fact conduct DNA repair or contribute to DNA repair. So targeting the androgen receptor in cancer cells may in fact be a form of synthetic lethality by essentially inhibiting or downregulating that DNA repair capability of the androgen receptor. So it’s very interesting clinical science.

CP: Yes. Another interesting question that is being tackled at this ASCO relates not to CRPC but to treatment of locally advanced disease with radical radiotherapy. What are your thoughts on that?

CR: I’ll say the top line data and I’d like to get your thoughts because, as a clinical oncologist in the UK, you do a lot of radiation. The basic story here is that over the course of twenty or maybe thirty years now the field has been progressively demonstrating that the addition of hormonal therapy to radiation therapy for high risk localised disease is beneficial. As you know, the study started with lifelong androgen deprivation therapy, three years and then on down, as risk decreases the duration of androgen deprivation therapy decreases. But there have been relatively few studies that have looked at one duration versus another; there have been a lot of studies that have looked at radiation with hormones versus none. So this study was 36 months of ADT versus 18 months of ADT demonstrating that there was really no difference in the overall outcome of patients who received 18 months versus 36. As we treat patients with ADT we know that there’s a pretty significant quality of life benefit to getting them off the treatment early. So this was again reassuring news. The data has a few holes in it and I’d be eager to hear your thoughts but overall a reassuring piece of data.

CP: Yes, I’m not sure it’s quite as simple as it was presented. It was presented as saying 18 months is the new standard because it’s just as good as three years. I’m not convinced about that; it seems to me that until now three years is the standard and this trial was asking a non-inferiority question – is 18 months as good? While it failed to show a difference between 18 months and 36 months it was a relatively small trial, about 630 patients. So it’s still consistent with the possibility that there’s a 30% detriment for 18 months rather than 36.

CR: Yes, I agree with your point and we do need to be mindful of the statistical design. I was struck by the conversation that took place after that presentation about how the statisticians review this data and say it failed to meet the statistical standards set forth. But as a clinical person and understanding the quality of life benefits, this point was raised in the discussion, this will change practice for a lot of individuals I think. It brings in the tension that we have as clinicians between statistical rigour and empathy for our patients and their life. My conversations with colleagues after that presentation was about the art of medicine and whether that’s good enough for us to shorten the duration of androgen deprivation therapy.

CP: Yes, it’s a very good point actually because the data, while it may be limited, will certainly be helpful when it comes to discussing with patients what duration they want. The other point which occurs to me is that these are the same patients that were also studied in the STAMPEDE abiraterone comparison. What we can say is that the benefit of adding an extra 18 months of ADT is much, much less than the benefit of adding abiraterone.

CR: That’s a very interesting perspective. So now if we were to go back to our clinical life and see these patients in light of the STAMPEDE data they may be treated with two years of ADT with abiraterone, which is what the STAMPEDE study did, versus 36 months of standard ADT. So very interesting and, again, it’s decreasing the duration but increasing the intensity. We have trials that are ongoing that are looking at that concept of triple hormonal blockade etc. in relapsed patients, for example, and that’s a concept that is as yet to be fully explored, this idea of shorter duration, increased intensity.

CP: Have you seen anything else this ASCO that you think is going to change practice in the clinic?

CR: This is a practice-changing ASCO from the standpoint of the STAMPEDE and the LATITUDE data which will be discussed by you and by others in a separate context here. My overall sense right now is that this would be a resurgence of enthusiasm around hormonal therapy and molecular diagnostics. There have been a couple of things that we were enthusiastic about a couple of years ago that have gone away, perhaps. I don’t think many people are talking about AR splice variants like we used to, so that’s interesting. We are now seeing so many quality centres present quality data on the role of sequencing, both from a prognostic standpoint and a treatment decision standpoint that I feel like this is becoming a standard. We should be considering doing tumour sequencing in all patients. I realise that a vast majority, or not a vast but a majority, of patients who undergo a sequencing test may not have an actionable finding in terms of the mutation leads to a drug choice or a therapy choice but the sheer number of mutations is important. Whether a patient has an unmutated prostate cancer, that’s actually an important piece of data now. So I’m eager to see that story evolve.

CP: Yes, I’m sure you’re right, it’s definitely coming isn’t it? We can’t be dinosaurs; we’re going to be treating different prostate cancers with different treatments in the years to come. So thank you very much.

CR: My pleasure.