Tailoring treatment strategies in prostate cancer: expert insight

Prof Eleni Efstathiou - MD Anderson Cancer Center, Houston, USA
Dr Maria De Santis - University of Warwick, Coventry, UK
Dr Nina Tunariu - Royal Marsden Hospital, London, UK

EE: Good morning from Vienna. Rainy it is but it’s always Vienna and I’m here with Maria De Santis and Nina Tunariu. Two of us are medical oncologists; I’m a medical oncologist, Eleni Efstathiou from MD Anderson, and Maria De Santis just moved from the very cosmopolitan Vienna to the UK and the University of Warwick and she’s going to tell us more about this exciting new area, probably at the centre, the headquarters, of STAMPEDE, would I be able to say that? Of course Nina Tunariu is also from the UK from the Royal Marsden where all the big drug development in prostate cancer has occurred and I think she’s been working side by side with Johann de Bono doing all that. We’re going to have an interesting conversation, I believe, and I hope you will enjoy it too, on what’s really happening to put together the puzzles on prostate cancer therapeutics today. So I actually wanted your opinion, where do you think… we’re actually getting convoluted, it’s pretty obvious. We’ve moved a lot of drugs from the very advanced prostatic carcinoma to a little earlier in CRPC and now we’ve come to hormone sensitive disease with docetaxel just showing some real data of advancement in this field. Now we’ve come to the controversy – should we use docetaxel in just high volume disease for prostatic carcinoma that’s metastatic hormone sensitive or should we just apply it to all patients? I would like both your opinions and, Nina, you’ll be able to also help us from your radiology standpoint. Maria?

MDS: Well, Eleni, I think it is a very timely and a really important question. As you said, docetaxel was the first chemotherapy that showed a survival benefit in prostate cancer but in the metastatic and castration resistant setting in patients that were in pain and had actually a lot of disease. Then the new oral drugs came around and docetaxel chemotherapy moved a little bit to a later stage in the castration resistant setting. Now we have new and very exciting data, the first data was presented more than a year ago at the ASCO meeting and now being fully published, showing that docetaxel in those patients who just started hormone treatment in the hormone naïve or hormone sensitive setting, that docetaxel in that setting provided a survival benefit that was huge. It was a survival benefit, it is a survival benefit that we have never seen before in the treatment of prostate cancer, like 12-15 and in some groups 22 months. So this is a huge benefit and I think it is important to discuss in detail and to think of which patient group and if, as you said, only one patient group or if all patients in the setting of hormone sensitive disease with metastasis should receive it or if we at least have the obligation to inform the patient about the data and to offer that therapy. The data itself, and we have three studies to be more precise, we have one study, an older study, a French study that was actually a negative study but two other studies, the more recent publications, the CHAARTED trial, the US study, and the STAMPEDE trial, a UK-based study were the bigger ones and they showed that survival benefit in the whole population, in the intention to treat population. This is, I think, a very robust indicator that actually all patients and all the subgroups benefitted. There was one subgroup, to be honest, in the STAMPEDE trial, a subgroup of patients that was not metastatic, so M0 subgroup, that was also included but in this subgroup actually the overall survival benefit did not reach statistical significance. We have to take this into account. However, also in that high risk M0 subgroup the failure free survival showed a significant difference in favour of adding docetaxel to the hormone treatment.

So now, at this meeting, a meta-analysis, so putting together all the data of these three trials, reinforced the impression that actually all patients, at least with metastatic disease would derive a benefit from adding chemotherapy to the start of hormonal therapy. The discussion has been ongoing if maybe only the subgroup of the high volume disease patients should be treated with chemotherapy. This came because in the CHAARTED trial, in the American trial, the patients were divided into low and high volume disease and in the subgroup analysis the high volume disease patients derived a bigger benefit than the low volume disease patients. But still the whole trial was positive and the intention to treat population was positive. So I think from the methodological point of view, from the trial point of view, we have every reason to offer chemotherapy to all patients and not only to high volume disease patients.

In my daily practice I must admit I offer and I discuss chemotherapy with all patients, telling them that I think chemotherapy will be one part of their treatment anyway at one point. We would expect them to be in need of chemotherapy and I tell them that the overall survival benefit we derive is largest when we use it right away. The chemotherapy right away at the start of treatment, together with hormone treatment, is six cycles and then we stop. I think this is also an advantage for the patients, they have five months of therapy and then we can have them off treatment, just stay on hormonal treatment. I think this is quite an attractive offer for those patients to start with that, knowing that in five months the whole chemotherapy treatment is over. And we do not have to give it together with corticosteroids because in the CHAARTED trial, for example, no prednisone was used. So this would also be maybe an attractive difference to the use of docetaxel chemotherapy at a later stage.

So, my answer in summary would be that in my opinion chemotherapy should be offered to all patients and not only high volume patients. So to the whole spectrum of metastatic patients at the start of their hormonal therapy.

EE: So that’s fair. Even though we have to admit, at least from the US perspective, the number of patients with de novo metastatic disease is very limited. The question remains, of course, because the CHAARTED trial gives kind of a diverging opinion there and we’re going to see how the paper looks when you do the meta-analysis when it comes to recurrent patients. But I think that Chris Sweeney wanted to introduce a scientific rationale and make it more clinically apt by trying to focus on those high volume patients. That’s why I wanted, Nina, your opinion because we use the old school approach, bone scan, look outside the actual skeleton for metastases, it’s very rough, very crude. Where would you put it at this point, if it was right now you designing a trial to try to identify through your novel methods the really more aggressive, high volume disease? What would your comments be here?

NT: I think that the volume of disease, there is no doubt it’s prognostic and we have data on whole body MRI showing this, on bone scan index showing this, on PET choline showing this. So every single imaging method says, yes, you have a high volume of disease you are going to live less. However, I think that the time perhaps has come for some subtleties. So none of us talk about what is the difference between having a sclerotic pattern of disease and lytic pattern of disease. It is my experience that what makes and worries me when I report a scan is when I see the disease breaking through the bone. That is what gives patients cord compression, that’s what gives patients fractures. Then there is the other emerging issue, but not in M0, but we seem to pay, perhaps, less attention than we should to local disease. Because we use CT and bone scan we miss these nuances because on CT you will never pick up early local disease recurrence. So if I would have a trial and if it is really a clinically important question is if you want to subdivide this and just, how shall I say, fine tune this rather than giving to everyone, just look a little bit more at the patterns of disease. Because I think it’s a little bit too… not finesse enough to just say high volume and low volume, we need to say what kind of low volume. Because if you have low volume but it’s lytic and it’s going to give you cord compression then you are definitely in a much worse category than someone who has, perhaps, fifty metastases but they are all sclerotic and progress very slowly.

EE: I couldn’t agree more and that brings me to the question I have. Yesterday we had a very nice meeting with Nick James and Bertrand and we were all discussing all these nuances. As Nick put it, and I know you work with him very closely now, they could not find any classifier that would help them identify with some finesse, within what they had as a group, which patient would be the utmost candidate for chemotherapy versus not in the hormone sensitive setting. That brings us now exactly to the question in hand for castrate resistant disease. There is completely different… you brought up, we’ve never seen in the hormone sensitive setting such survival differences but we don’t have a lot of trials there, to be honest with you. The only other trial that has been reported recently was Maha Hussain’s non-inferiority trial of continuous castration versus intermittent. So now we come to the castrate resistant phase where we have all these drugs that are now in one space – abiraterone, enzalutamide, docetaxel, cabazitaxel is coming up there because we’re going to be using docetaxel up front in some patients, and there’s a question there. Sipuleucel-T in the United States and radium-223 where there are limitations in some countries where you cannot use it if you used chemotherapy, and that was the case for Belgium for instance, you have to use it before. So we are limited by some very objective bureaucratic issues but where they are asked to pick the ideal sequence of treatments for these patients, where do you see us going with that? How will we practice medicine in five years?

MDS: That’s the crystal ball question actually. I think the problem here is that, as you said, we have a lot of drugs in hand, a lot of approved drugs for the metastatic castration resistant setting, which we don’t have for the sensitive setting. Here the big question is when to start treatment in the course of disease and how to sequence those treatments. So far we don’t even have a predictive biomarker that helps us to choose the first treatment, not to speak about the second or third line treatment. Maybe one biomarker will come soon but still we do not have it now and here. The second question is… so here we do not have guidance at all for the first pick. In daily clinical practice we discuss with the patient and we look at the pattern of metastases and we look at the pattern of disease progression, mostly, at least to define the start of treatment. Because a patient with a slow, very slow, rising PSA might not be in need of chemotherapy right away but we would like to start with a hormonal therapy like abiraterone or enzalutamide, a patient that maybe has a few bone metastases and a rising PSA. So this would be one way to go. It would not be wrong to start with chemotherapy right away but most patients would choose to have an oral therapy most probably. The start of treatment is tricky, is a tricky question, because in former times we only had chemotherapy in hand and we tended to start when the patient became more symptomatic and the volume of disease was increasing, just to save what we have. Now we do not need so much to save because we have several treatments in hand. We tend to start earlier and I think it makes sense to start earlier before the patient becomes symptomatic, before pain comes up, before the first event like spinal cord compression comes up. So we actually want to avoid it, we want to secure the patient a disease free interval and a long time until problems come up and a long time until the quality of life and the health does deteriorate. So this is the background of starting early with preferably one drug that has little toxicity and not so many side effects.

EE: So I’m going to push that towards Nina. You said early, I start early but then when we had this recent St Gallen meeting everyone voted against using the new technologies to identify metastatic disease in the castrate resistant setting, early with whole body MRI and the like. When I went after that to the European Urology meeting it was all about implementing these new methodologies. In the United States I’m getting a lot of pushback, and I think we agree that you’ll tell me about that problem, why it can’t get done as much whole body MRI. We don’t know if it makes sense to implement treatment when you identify the metastatic lesion with a whole body MRI, some of our colleagues say. From my standpoint it’s still metastatic disease and if you tell us how you believe it can shape the future as a methodology.

NT: I think for me the surprise comes that prostate cancer is, apart from the breast, the only cancer in which we don’t use imaging to guide us. I think it’s a paradigm which is old and we need to change it and it comes from ten years ago when everyone said patient prostate cancer is old, the only treatment I have to give is some hormones and sometimes docetaxel. So where is the point to do imaging? I don’t need an accurate tool because even if I know nothing will change. I think we need to accept that in the 21st century perhaps we should use slightly more precise imaging methods. Even more importantly, and I was so happy that Howard Scher in his prostate cancer working group 3 at ASCO mentioned this, we should stop treating the patients until or beyond unequivocal progression i.e. until they break a leg. Change the treatment when they stop benefitting, and I think that’s an absolutely crucial point and it’s going to change the way we will do in the future. Because how do you decide when someone has stopped benefitting from follow-up, from just watch and wait, from hormones, from abiraterone. You need to have a little bit more than your clinical gut feeling which means you will need to have imaging which can pick up early cord compression, can pick up early local disease recurrence because we are now seeing this new pattern in which we cured the metastatic disease but we left behind a huge local tumour recurrence and we had nothing to treat it with. So what I would say is I appreciate it is expensive and I appreciate it is hard because all of our data comes based on routine scans, CT and bone scan. But we need to accept that everything we have published and done until now we had no imaging response criteria and the patients with prostate cancer have bone disease and at least half of them have only bone disease which means that the imaging, the way it is, cannot help.

So what I would like in an ideal world is that the patient comes to you, the patients is asymptomatic, you look at the PSA doubling time and you look at all everything else and then you do an accurate test like a whole body. The advantage of whole body will be that it will give you both nodal and bone disease but by all means in your hospital if they expect it is PET choline or  fluoride PET, combine them and try to get a feel of how fast that patient progresses. If you have colorectal cancer, if the patient comes with a 5mm lung lesion you don’t start and treat them but if the lung metastasis becomes 10mm in four weeks then you treat them. So we forget about all of this and I think in fairness the USA doesn’t help because you are correct. I think that if in the USA the people would be more inclined to do whole body MRI rather than only expensive nuclear medicine tracers the technique would have been much more successful. The problem mostly, in fairness, comes from the radiologists themselves because one is in the UK and in Europe the radiologists do not have time - a whole body MRI takes an hour. In the USA the whole body MRI is incredibly expensive because the principle in the USA is that each radiologist records a body part. In Europe it doesn’t matter, I report urogenital imaging, in the USA you report chest, you report abdomen, you report spine so therefore the cost becomes incredible. But I think if you would ask a patient what tests would you like? Come to hospital, have a CT, have a bone scan for which you spend four hours because you need to have the injection, and then at the end I’m going to tell you you have bone disease only, I have no idea if you are responding or progressing; my gut feeling says that you are OK, let’s continue. Or I give you the possibility to do a test in which it’s telling me are you progressing or responding. And by all means radiologists are at fault here because we don’t have enough trials. But I think if you are going to have an answer we do need to use slightly more accurate tools and I think we need to move prostate cancer in the 21st century i.e. is my patient benefitting, not only have you progressed? No? Let's continue.

EE: You brought up some very important issues that had to do exactly with how much we’ve changed the disease in ten years by not even realising what you just said. Because we just saw two trials a couple of years ago, it was the 302 trial on abiraterone in chemo-naïve patients and the PREVAIL trial with enzalutamide showing a median survival of three years, essentially, which is very different from 2004 when the TAX-327 trial was conducted, a different, probably, era with an equally good drug, you would say, docetaxel. Then again 75% of patients being asymptomatic in that trial and it was only 18 months of survival. I think, and Maria I’m going to ask you a question on that, beyond drug development is also that multidisciplinary approach that should now incorporate beyond urology, radiation therapy and medical oncology, radiology, pathology and, let’s hope, molecular pathology. So where do you think we’re going? Are we going to see informative radiology and predictive markers that will help us choose treatments very soon? I’m coming back, I want to know – do you see it in a year, in two or three years? We’re getting some diagnostic companion studies happening.

MDS: Most probably we are going to see a predictive marker at least for one little spectrum of hormonal therapy. There is one marker in development and most probably others to come. I’m also quite confident that radiologists like Nina will be successful in implementing more imaging into our prospective trials and those trials then will help us to develop the imaging being part of decision making which makes a lot of sense and we have heard all the arguments now. So I’m quite confident that it is coming up, I’m not so sure if it is within the next two to three years but maybe in five years we won’t rely on PSA and bone scan anymore and wait until we see more metastasis coming up and confirming those metastases before we feel obliged to change treatment.

EE: What do you think from your perspective?
NT: I think that the other thing which imaging can help and a lot of centres are doing now, and we were talking with Bertrand Tombal several times about this, is the so-called discordant or differential response we see. I think that’s a specific group for prostate cancer and I call them oligo-progressors because you have someone who has a successful response to abiraterone, the patient feels amazing, or enzalutamide, and then suddenly the PSA goes up. So what do you do? Suddenly the PSA goes even higher and even higher. Now if you do a CT or a bone scan the CT will not help you in bone, so we have established that. The bone scan may because if it shows two new lesions then you have the chances of taking the patient off the trial but if you would do a slightly more advanced imaging you may find out that that PSA comes from one or two lesions only. It is not uncommon and it is a shame because we could, if that tends to be true, we could continue to treat the patient with the hormonal therapy and just treat with radiotherapy those particular fossae, one. Secondly, we can biopsy them and let’s find out why it’s different and we use imaging to guide the biopsy which means you go exactly where you want. Perhaps you do a PSMA or an FDHD and try to figure out what happens with those lesions which are escaping your blockade.

EE: So I just had a real case example, exactly that – a very young man, 52, responded very well by the initial imaging and PSA criteria on abiraterone. He went on starting the PSA to go up, I did exactly what you’re saying and confirmed that some lymph nodes in the pelvis were the ones growing. Then I did exactly what you said, by adding on radiotherapy and it worked like a miracle. It was very localised progressive disease. It can give you answers and it does not have to be a highly specialised institution. I think you just gave a very practical answer for the future to us. I would like to thank you for being part of this discussion even from your office and I hope that in the future we will share more moments like this. Thank you very much.