Prof Martin Bögemann – University of Münster, Münster, Germany
Prof Nicholas James – Queen Elizabeth Hospital, Birmingham, UK
MB: Hello everybody, we are here with ecancer and we’re here at the ESMO annual meeting in Madrid, ESMO annual meeting 2017. This meeting there has been lots of new data and especially there has been new data on the STAMPEDE trial. I’m here sitting with Nick James from the UK and he is leading the STAMPEDE study and he and his team presented some new data and we’re happy to talk about this a little bit.
NJ: Yes, we’ve presented three pieces of data at the meeting, one yesterday and the other two tomorrow. There are a number of things that arose out of the data we presented at ASCO in the summer. So we showed that up front use of abiraterone improved survival really quite substantially, about 30% odd. Now one of the things that intrigued us when we looked at the data, which is the thing I’m going to present in the morning, is that the effects appear to be bigger in the patients with non-metastatic disease. The STAMPEDE trial includes about 50% of the patients in this particular comparison who have either high risk node negative disease, so T3/4 Gleason 8/9/10 or a PSA above 40, or pelvic node positive but no distant metastases. So these patients in the main got radiotherapy as part of their package of care and for those that had radiotherapy treatment was truncated at two years, so it was a fixed duration of treatment instead of indefinite. What we saw was that when we did some further analyses, so we looked at overall survival is pretty immature, we had very few deaths fortunately, we’ve only got around four years’ follow-up. But for metastasis free survival, failure free survival excluding PSA and failure free survival including PSA we saw very big effects in favour of adding abiraterone to the package of care and these appear to be biggest in the ones who also got radiotherapy. So if we look at the subgroup for failure free survival including PSA who got androgen deprivation therapy, radiotherapy and abiraterone, adding abiraterone gave you almost a 90% improvement in failure free survival at three to four years out i.e. beyond the end of the abiraterone and the androgen deprivation therapy. If you look through the other endpoints it’s very consistent, so for metastasis free survival in the node positive subset, that difference, the hazard ratio is a little bit softer, it’s 0.26 which is a 75% reduction in relapse at three to four years.
Now, this is clearly a huge effect and we’ve just separately published, we’re part of a consortium called ICECaP which Chris Sweeney leads looking for intermediate endpoints for high risk non-metastatic disease. So we’ve contributed the STAMPEDE data to this, as have a lot of other research groups, and one of the things we’ve just published in JCO is that metastasis free survival in non-metastatic disease is a very strong and consistent surrogate of overall survival. So there is a move to have the regulators accept this as an intermediate regulatory endpoint which we think they probably should. So on that basis it would seem that we’d have pretty solid data to say that abiraterone should be part of the package of care at the very least for the node positive patients and obviously we’ll wait to see for the node negative high risk patients. But the effect on failure, PSA, looks to be pretty good in those as well.
So that was the first bit of information that we’re presenting, so it’s a more in-depth analysis of the work we already presented. The other two pieces of information we’re presenting, one was in the oral session yesterday presented by Matt Sydes who is the lead statistician for STAMPEDE and one of the architects of the whole trial and a fantastic, very talented statistician. The other piece of information will be presented tomorrow by Claire Vale who is one of the lead statisticians in the MRC Trials Unit’s meta-analysis group and we’ve collaborated with them a lot. The question that arises is we’ve got two trials now, or two sets of trials, that show that survival can be improved either with abiraterone or with docetaxel. Obviously one of the things that’s of interest is how do the effects stack up one against the other – is there a better treatment, is there a clear winner between the two. Within STAMPEDE the tail end of the docetaxel arm overlapped by about a year and a half with the abiraterone arm. So we have around 600 men randomised between abiraterone and docetaxel. Matt presented the data on the head-to-head comparison and essentially what we saw was that for overall survival there doesn’t appear to be a difference. But what we did see also is there is really quite a striking difference in favour of abiraterone for failure free survival based on PSA. That got progressively diluted down as you went to harder endpoints like skeletal related events and metastasis free survival, particularly skeletal related events didn’t appear to be different between the two. So what that tells you, and this was, for me in the clinic where we’re not allowed to use abiraterone at the moment while it goes through all the regulatory processes, is that I can tell my patients that we’re not compromising their survival by not being able to give them abiraterone but they will have a different passage through their disease if we give them docetaxel to if we give them abiraterone. With abiraterone the patients are staying on the first treatment much longer which implies that they’re going to have a shorter castrate refractory prostate cancer period at some point in the future. So this is important information for us to be able to communicate with our patients in the clinic.
In due course the question will become obsolete because the real question is should you give both. We can’t answer that from STAMPEDE but other trials like PEACE-1 and the ENZAMET trial…
NJ: And ARASENS, yes exactly, will potentially answer that. So my hunch is that for docetaxel fit patients we’ll end up giving them both docetaxel plus one of the new androgen receptor targeting drugs.
MB: In other parts of Europe, next to the UK, are not considering Brexit.
NJ: We’re out on a little limb all on our own in so many things.
MB: Absolutely. In Germany, where I come from, we will probably get hold of a broader label of abiraterone early next year, maybe even in January. So what would you recommend if you would have free access to the compound for your patients? Would you rather go for abiraterone or would you rather go for chemotherapy in this setting?
NJ: I think in the States as well pretty much they have free availability because the drug is off patent and therefore will presumably come down in price. In a sense it’s the same situation you have with CRPC - you relapse the CRPC, you’ve not had either drug before you have to pick drugs. So I think if you sit down with a patient and say, ‘Would you like to have a pill that may cause your blood pressure to go up a bit and we’ll have to do monitoring but it’s not very tough, to be honest, or do you want to have chemotherapy?’ if you put the choice like that everyone will say, ‘Well, why on earth would I take chemotherapy?’ If you put it slightly differently and say, ‘Right, do we give you chemo now and then abiraterone later or abiraterone now and then chemo later?’ in CRPC you find a lot of the time patients say, ‘Actually, I’ll go with the chemo first while I’m well and I’ll save the less toxic treatment for later when I’m less well.’ So to a certain extent the answer depends on how you frame the question. I do think probably most people, looking at the data, looking at a very long period trouble free, really, are going to go for abiraterone up front, it would be surprising if they didn’t. There are some people who might say, ‘Well, let’s just get the hit out of the way and I’m not shackled to monthly blood tests for the next three, four, five years,’ but I think they’ll be a minority.
MB: As a urologist whenever I talk to a person who has a high risk prostate cancer by the D’Amico criteria, that’s your N0 M0 subpopulation in STAMPEDE, we would always tell them, ‘OK, you need radiotherapy,’ sometimes you operate on them but mostly you would give those guys radiotherapy and you would always tell them, ‘And you need ADT to boost the effect.’ So do you see that in the future we will do that with abiraterone?
NJ: I think we probably should. There’s quite a lot of certainty now around the role of combining radiotherapy with ADT, so the PR07 trial and the SPCG-7 trial both tested the proposition of comparing hormone therapy with and without radiotherapy. They both showed you got a 50% reduction in the risk of death adding a decent period of hormone therapy, years. The questions has been asked the other way around – radiotherapy plus or minus hormone therapy, Michel Bolla's trial for example, and that gives the same answer – you need both. And generally in oncology combinations are usually better than single therapies so it’s consistent with the principles of oncology.
Now, there’s no randomised data combining radiotherapy plus or minus hormones for the node positive subgroup but one of the things we published from STAMPEDE last year was we had roughly half the patients in the node positive group in the control arm had had radiotherapy and half hadn’t. What we saw, consistent with the node negative data that I’ve mentioned already, was that you roughly halve your risk of relapse at three years if you gave radiotherapy and hormone therapy to node positive patients. These are not pathologically node positive post-surgery, these are radiologically node positive at diagnosis with no pathological staging. So we’re assuming that the radiological assessment is telling us the truth. If you look at the prognosis of these patients they’re clearly doing really quite badly if they have hormone therapy, only half of them have relapsed by three years. So the very striking thing was that, say, half of the node positives have relapsed by three years, 75%, roughly, are disease free if you add radiotherapy and that’s non-randomised data but it looks pretty compelling. If you then add abiraterone as well the relapse rate drops to round about 10% which is really very striking. And it’s a time-limited thing, it’s two years of treatment. So clearly we need to do a bit more follow-up on this and harden the data up as we get more events because it’s only four years median follow-up but it’s a very easy to manage treatment and it’s a very big effect. So I think there’s a very strong case for it and it probably is cost-effective as well because relapses are so expensive.
MB: Yes, you have pain when you relapse; if you get metastases you will have debility and you have hospital stays. It would maybe even reduce cost.
NJ: We’ve just been looking at the health economic data from the docetaxel part of STAMPEDE and this is not published data so I have to be not too leaky about what I say about it. It won’t surprise you to know that the cost of delaying relapse, the benefits financially of delaying relapse, wipe out the costs of the docetaxel even without a survival benefit. Because, as you say, things like even just radiotherapy is not cheap and once you get nastier complications, outflow obstruction needing nephrostomy tubes or fractures, spinal cord compression, these all add up to a) a lot of unpleasantness for the patient, but b) a lot of expense for whoever is paying for it. So the same sorts of gains will apply to abiraterone and even though it’s a more expensive drug the gains, the effect is so big, it’s certainly clinically effective but I think it may well stack up against the rather strict financial criteria we have in the UK. There’s a separate issue around the licence, of course, as well but we’re talking to the regulators about this.
MB: Next to the financial toxicity that is involved in this theme, if we look at the data which were presented yesterday by Kim Chi concerning the life quality assessments within LATITUDE, which is another trial addressing M1 patients, he clearly showed that life quality was improved and maintained using abiraterone compared to ADT alone. So if we extrapolate this on the irradiated patients, the M0 population, we might have another positive fact for the patients.
NJ: Yes, that data was quite surprising, in one sense, because it was the patients on the control arm and the abiraterone arm up to the point of relapse. So to see an improvement in quality of life in patients who are responding to treatment by adding abiraterone was really quite surprising. You might, on the face of it, have expected if they’re responding to hormone therapy and haven’t relapsed yet they’ll have the same quality of life. But it actually goes to show, slightly tangentially, that patients that we think are asymptomatic probably aren’t really completely asymptomatic and you can improve their quality of life if you control their disease better which I think is one of the interesting conclusions you come to from Kim’s paper. Again, you’re right, it’s going to be key to driving the health economics. If you’ve got a quantity of life gain and a quality of life gain that starts to look like a very good treatment.
MB: Yes, quality of life is something worth so it’s worth investing into it as well.
NJ: Kind of simplistically, the average length of therapy in the CRPC setting with abiraterone is about 15 months whereas our average length of therapy in the metastatic patients is roughly double that in STAMPEDE. So, give or take, you give 15 months of abiraterone and you get 3-6 months survival gain; you give another 15 months of abiraterone by giving it at the beginning and we’re projecting you’re going to get around about a 3, 3½ year survival gain. So your extra 15 months buys you a lot of extra survival. So probably it’s going to stack up reasonably well in terms of costs per quality adjusted life year.
MB: The third trial, or the third part of the STAMPEDE output from this conference will be?
NJ: From the meta-analysis. STAMPEDE isn’t really a single trial, it’s a collection of trials. We’re up to ten different treatments we’ve assessed in the trial now, we’ve just passed the 9,000 patient mark which obviously we’re very pleased about. If you treat them all as separate trials we can compare the abiraterone STAMPEDE with the docetaxel STAMPEDE and we can also then build in a meta-analysis where we look at docetaxel STAMPEDE, docetaxel CHAARTED, abiraterone STAMPEDE and LATITUDE and we can compare all these things head-to-head. And GETUG-15, the French trial, is in there as well. Now the important thing about this is that in the head-to-head directly randomised patients the survival, if anything, shades in favour of docetaxel but it’s not really significant. In the network meta-analysis, which is the methodology that Claire Vale is using, all of the parameters, including survival, come up in favour of abiraterone. So we’re presenting apparently discrepant data, in other words, one’s showing one thing, the other’s showing an effect in the other direction for survival. The explanation for this is that the docetaxel trials are mostly carried out in the 2005-2010 era, or 2013, in terms of recruitment; the abiraterone trials all started 2011 or afterwards by which time abiraterone, enzalutamide, cabazitaxel, radium were all pretty freely available in all of the populations being studied. So if you were a man going into STAMPEDE in 2005 the only salvage therapy you had available was docetaxel whereas a man going into STAMPEDE in late 2013, early 2014, when the abi arm closed had five different treatments available. So what we see is a time gradient in terms of prognosis, the later docetaxel patients do much better than the earlier ones. So if you compare the doce trials and the abi trials they’re time shifted so you expect the abi trials to do better. So we think the two different bits of data are actually consistent although on the face of it you’d think actually they clash.
MB: We can talk about some other things that were presented yesterday. What’s your take on the PSMA therapy that we saw yesterday by Michael Hofman from Melbourne?
NJ: That’s generated a lot of discussion and I’ve just done a satellite session on targeted radionuclide therapies with Johann de Bono and Joe O’Sullivan and Aurelius Omlin from St Gallen. There’s a number of these things around, as you know particularly coming from Germany because a lot of this stuff has originated out of Germany. The cute thing about PSMA radio-tagged diagnostics is that you’ve got the possibility of both imaging what you’re treating and then treating it and then imaging it again afterwards because you can hook your PSMA binding molecule to a radioisotope like lutetium which is the one in the study but other ones are being used as well like actinium which is an alpha emitter which is potentially more interesting. You can hook it up to technetium as in the standard bone scan imaging and then you’ve got a technetium PET type image that you can reconstruct where the disease is which allows you to do dosimetry as well. You can image, a bit like conventional radiotherapy, you can work out the uptake ratios and you can calculate a dose but the doses are a bit weird to conventional radiotherapy because conventional radiotherapy you’re giving the dose, bang, in two minutes whereas if you’re giving it as a radioisotope you’re giving the dose over several weeks so you’ve got a big dose rate difference. So although the numbers look familiar, 20Gy, 30Gy, they don’t quite mean the same as 20Gy given as a stereotactic dose, for example. But it does allow you to quantify it and compare the different things.
So what Michael Hofman showed was some really very interesting responses in hugely pre-treated patients, just very heavily pre-treated patients who had really pretty much no options left. So I’m sure this is something we’re going to hear a lot more about. The isotope matters, so lutetium is a gamma emitter so your radiation goes further so you get more bystander lysis damage of things a bit further away potentially. If you use an alpha emitter like actinium then you get a much shorter distribution of the radiation from where the molecule is. The one side effect that would concern me, that does concern me a bit, and apparently it’s more prominent with actinium than with lutetium, is that you get PSMA uptake to quite a high degree in the salivary glands so dry mouth is a side effect. You’d predict it would be reversible and apparently that’s more prominent with the alpha emitters than the beta emitters. You’d think dry mouth, well I’ll put up with dry mouth to make my cancer better, but I don’t know if you used to do this in medical school, one of the games that came up after rugby matches was how many cream crackers, which is a dry biscuit that you eat with cheese, can you eat? And you eat the first one fine then you put the second one in and it’s like chewing sawdust. Without saliva you can’t eat, you can’t taste anything and it’s horrible. So profound xerostomia, and dry eyes as well, is potentially not to be underestimated as a side effect. So if you optimise these treatments, give bigger doses, the danger is that you’ll be doing a lot of damage there as well. So there is some work to be done on how you optimise the one versus the other and whether there’s some way of blocking it, pre-loading or something with a ligand, I don’t know, that allows you to protect it. But they certainly look very attractive; it looks like a whole new class of drugs potentially.
MB: You can do a whole [?? 20:35] thing. In Germany we have access to the lutetium PSMA therapy and what we do, we just do cooling of the salivary glands. It looks funny, you take a band-aid and then you put those liquid cooling things from the fridge to the cheeks and it helps a lot. We only see transient dry mouth, periods like for a couple of days and it almost always vanishes completely.
NJ: Yes, I think it’s a very attractive treatment. Exactly that thing works for hair loss so there’s no reason why it shouldn’t.
MB: In Germany there’s only one site which produces actinium PSMA therapy which is in Heidelberg where the compound was developed. Because it’s an alpha emitter it does a whole lot of damage wherever it works so the guys which I sent to Heidelberg and I saw afterwards they had dry mouth problems which was very prominent. What I heard is that you can’t easily offer actinium to the public domain because you have to sit on an atomic bomb, so to say, to milk it, to get this substance. You can’t produce it on site so it’s hard; even though it would be very interesting to have it working in a metastasis it’s hard to get hold of.
NJ: So lutetium, in that respect, is more user friendly, isn’t it? You’ve got a longer half-life and stuff. There are some radiation protection issues, aren’t there? You have to keep them in hospital because it’s a gamma emitter, beta emitter, so you’ll get gammas so you are radioactive to your surroundings whereas alpha emitters that’s not so… because it’s such a short track length. As with radium you can just inject the stuff and off they go, there’s no real radiation protection issues with that.
MB: What we saw yesterday, the talk from Michael Hofman, is promising concerning… it warrants further studies on large populations randomised to a control arm then we’ll have to see what it is worth after a while.
NJ: I think there are other… I had a session this morning with a company with another similar compound saying what trial do we do. So there’s a number of ways of doing the same thing.
MB: I guess we’ve come to an end and we talked about some very interesting topics today. We hope that we helped a little bit for you guys at home to make decisions in this complicated set of patients. Thank you very much.