Professor Noel Clarke - Salford Royal NHS Foundation Trust, Salford, UK
Dr Eric Small - University of California, San Francisco, USA
NC: Welcome to San Francisco. This is ecancer, my name is Noel Clarke, I'm a consultant urological surgeon from Manchester in the UK and Professor of Urological Oncology at the Christie Hospital in Manchester. I'm joined by Eric Small, a medical oncologist and a friend of long standing, we've known each other for many years. We're here at the GU ASCO conference in San Francisco and one of the highlights of the conference has been the presentation of two large scale multi-centre randomised trials looking at novel anti-androgenic treatments in men with non-metastatic castrate resistant prostate cancer who have got rising PSA. These are the PROSPER and SPARTAN trials. Eric, do you want to just talk to give us a brief thumbnail sketch of what that trial has involved?
ES: Sure. Probably it's fair to say this is a thumbnail sketch of both trials because they were remarkably similar in design. As you know, Noel, this group of patients is one for whom there is no standard of care. These are patients that we know will develop metastatic disease, they're castrate resistant and they're considered high risk as identified by a very short PSA doubling time. These patients are assumed to have micro-metastatic disease and, indeed, untreated as in historic studies we've seen a significant number of patients develop metastases. So the principal goal for both of these studies was metastasis prevention. Because second generation antiandrogens or androgen synthesis or signalling inhibitors have been shown to have activity in patients with metastatic castration resistant prostate cancer, it was a rather simple decision to make to test patients with non-metastatic disease who are at risk for progressing. The design of both studies was to randomise patients at high risk of metastasis to receive either the active drug, apalutamide or enzalutamide, versus placebo. In terms of the SPARTAN trial we randomised, actually it was true in both studies, 2-to-1 ratio to receive one of these agents in conjunction with ongoing androgen deprivation therapy. The primary endpoint of these studies was metastasis free survival, essentially looking to see if metastasis could be prevented, and then there were a number of secondary endpoints on the SPARTAN trial - we were very interested in endpoints approximate to overall survival so symptom management, what we've termed progression free survival 2, progression after the next treatment.
NC: It was nice to see the matching of the populations, I thought that it was striking that they were large scale, 1,200 and 1,400 patients roughly speaking, and that you'd both dichotomised your PSA in pretty much the same way, which was less than six months and 6-10 months, based on the information which came from the old denosumab study about the development of metastatic disease, that that inflection point at about the ten month PSA rise point in M0 castrate resistance. But looking at the characteristics in both of the studies, many of the patients were good performance status patients - about 80% were performance status 0, about 20% performance status 1. So whilst that's representative of a population with prostate cancer, prostate cancer patients, one of the questions that has been raised is the rationale for treating these just on the basis of a rising PSA. I guess we might have some differences across the Atlantic as to how we would approach a patient like that.
ES: Yes, I think the issues… you raise two very important points. The first is not all climbing PSA patients are the same, even if they're all castrate resistant. We know from, as you say, the old denosumab data who are the patients that are destined to die from their disease and to develop metastases. So it's important to note that these both studies selected patients who are at very high risk of developing those events. I think the data from both studies is fairly incontrovertible - there's no question that using apalutamide or enzalutamide delayed the time to metastasis. The data is fairly compelling, it's a very strong hazard ratio, p-value of 0.0001 in both studies, really impressive impact in terms of what we saw. In the SPARTAN study we saw a two year delay in time to metastasis for people who got apalutamide versus not. So I don't think there's going to be any discussion or argument about are metastases prevented, there's no question that they are. I'll come back to the performance status in a second.
I think the appropriate question is does that matter? Does it make a difference in terms of the final outcome for these patients and how they do? There's a number of different ways of answering that. First of all, on the SPARTAN study it's really important, we're very delighted to see that every endpoint, secondary endpoint, was met in addition to metastasis free survival and most importantly was symptomatic progression free survival. So with that we were able to delay the time to bone pain and the time to the need for bisphosphonates and so on and so forth, skeletal related events. So that delay, for me, speaks to an important clinical benefit even if you don't believe that delaying metastasis in and of itself. I think if you were to ask, again it depends on what side of the Atlantic you're on, perhaps, I think if you were to ask one of my patients how they would feel about delaying the time until when they had metastases delaying by two years is a pretty good thing to do. The other component of both of these studies is that both showed a trend, but not a significant trend, towards improvement in overall survival. I wouldn't take that to mean that there is or is not going to be, there's only been 24% of all events in the SPARTAN trial before we can even begin to look at survival. That said, with 24% of events in place, hazard ratio is 0.7, p-value was 0.07, it appears that at least the trend is in the right direction. The only other thing I will say, and we can't predict what the ultimate survival will be, of course, but the only other thing that I would say is that on the SPARTAN trial we went out of our way to try to ensure that patients were treated with standard of care at the time of developing metastases per label, per labelled indication. So the study actually provided abiraterone to those patients who had reached the primary endpoint and whose physicians wanted to provide it to them. It was not mandatory and so not all patients got it but 80% of patients on the placebo arm received approved therapy, i.e. life prolonging therapy, and of those the vast majority, no surprise, were androgen signalling inhibitors, either abiraterone or enzalutamide. So what we have, therefore, is a study where the vast majority of the placebo patients have already been treated, have already received appropriate therapy and so this is going to be, once we look at the survival data, a true study of early versus delayed therapy as opposed to what we wanted to avoid which was early versus never. The survival data might look better if some patients had gotten apalutamide and the control group had got nothing. So we're very glad about that piece of it but, you're right, we have to wait for that data to mature.
NC: Yes, that's a fascinating issue, I think. So the question of early versus late and one of the points which was brought up by Philip Kantoff, who was the discussant in the poster session or the podium session to discuss the posters, was what drives those decisions to treat. So a question which remains unresolved is in those patients who develop metastases, and they clearly did that sooner in the placebo arm, did it make any difference because they then got treatment? It didn't obviously make a difference to survival.
ES: Not yet.
NC: Well not yet and I accept that things are immature and patients will need to be followed longer. But, of course, remember that in both trials the age of presentation of the patient was in and around 73, 74.
ES: That's correct.
NC: The average life expectancy in the UK of a 74 year old is ten years, roughly, ten to twelve years, and it's probably the case that in an individual with advanced prostate cancer developing from a certain point their survival is five, six, maybe seven years and it will be longer in some, according to comorbidity. But coming to the point, it is that the early versus late question, the quality of life information which presented in your trial, it wasn't presented in the enzalutamide trial, I'm told that it is to be presented and has been collected, but it didn't change. Now, there are two things in that - one is that the quality of life in the placebo patients didn't go down appreciably but equally the quality of life in the treated patients didn't go up either. Now, if patients are having a treatment then we like to think that it's going to improve their quality of life.
ES: I would counter that. I think Phil Kantoff raised a very important point which is how sensitive are these instruments to pick up these kinds of differences. What we were able to see, in fact, was this benefit that we're seeing in terms of metastasis free survival, progression free survival, symptom progression free survival, symptomatic progression free survival, was not done at the cost of side effects. So, in fact, we were quite delighted to see that patients started with a very high quality of life, and you started off by saying this, these are good performance status patients and, in general, patients with non-metastatic disease don't have symptoms, in general, from their disease. They may have some symptoms from androgen deprivation therapy but, by and large, this is a relatively healthy, relatively asymptomatic group of patients. So I think that any treatment that is going to be used in this setting had best be non-toxic if it's going to have any utility whatsoever. In other words, we could delay the time to metastases but if it's done at the expense of substantial quality of live decrement then it's absolutely a non-starter. What we saw with two different instruments was that patients in fact did not have a decrement in their self-reported quality of life. So it's hard to improve on starting well, so the absence of seeing improvement doesn't worry us much, doesn't worry me as much. I am relieved that there wasn't a decrement. There was a question somewhere along the line about, oh well, maybe we should be thinking about combining in this group of patients Taxotere together with apalutamide or enzalutamide, taking non-cross-resistant agents. I think there you really start worrying about quality of life impact. At least with these agents, I really can't speak to the enzalutamide study because it wasn't reported, but at least in the SPARTAN study there was no decrement and I think that's very positive.
NC: Yes. There were some side effects, though, I think it's fair to say. There was an increase in cardiovascular side effects; there was a small increase in the enzalutamide, and I suspect in yours as well, in mental function; both had a small but significant number of epileptic fits in the treated group. So there is a downside to it but one of the things which was raised and subsequently was raised in a presentation afterwards in relation to cost-benefit from the STAMPEDE trial and that is if these patients are going to be treated for quite a long time with relatively expensive drugs what's the cost benefit? Can the healthcare systems afford it?
ES: Yes, that's an excellent question. I can't tell you at all what the cost will be of apalutamide so it's hard to do that analysis quite yet. Nick James' approach to that, which is a way of looking at it, is that the longer you keep people in the non-metastatic setting, yes, you do have some of these adverse events to deal with but the longer you keep someone in the non-metastatic setting and keep them out of the metastatic setting that there's potential benefit from that and gain from that. And it has to be evaluated, it has to be measured. I do want to say one thing about toxicity, having used this drug. At the end of the day 8-10% of patients dropped out for side effects so it was not substantial and, really, if you look at the individual side effects it was, in my estimation, very well tolerated therapy. The epileptic fits, the seizures, were seen in 0.2% of patients, not zero, and there's no question that that's a class effect. We've seen this and I think it's true for both drugs. I don't think it's a common event but certainly in the case of the SPARTAN study, both cases were explainable - one patient had a pre-existing condition and one patient had a traumatic fall - but nevertheless it's a real event and something that we absolutely have to be aware of and tell our patients about, no question.
The other thing that I want to point out which was discussed a little bit in the context of the enzalutamide study was this… and I think there will be more data coming out, that the preliminary data that Dr Hussain presented suggested that 15% of patients, a stunningly high number of patients, on the enzalutamide arm died without metastatic disease. A second issue really, in terms of the tolerability of these agents and potential significant side effects, is what is the proportion of patients that are dying from events that are not prostate cancer related? I can't speak to the enzalutamide data, except for what was presented by Dr Hussain, on the SPARTAN study it was 1.1% of patients. So that's encouraging that we're not having patients die of things other than prostate cancer, that would be very distressing I think.
NC: Yes, and I think that in the presentation that Dr Hussain gave on the enzalutamide data which was picked up by Dr Kantoff I think there was a misconception because this figure of 15% came out. Of course, the audience, of which I was one, all sat up and thought, 'Well this doesn't sound very good.' But actually when going down to see the data and speaking to the people who ran the trial, in fact this was patients who had not developed metastases on the treatment arm who had died from other causes. And the view was that they had naturally had an extended period of life and died from something - cardiac event or whatever - and that is explainable in terms of the natural life expectancy of individuals. As we mentioned earlier, the average age is 74 so some of them would have been older and one has a natural lifespan and that explains that reason. So I suspect that that was a misconception in terms of the presentation.
ES: Yes, there's no question that one needs to… and we'll see it once they publish their data, that one needs to see the data corrected for duration of exposure. So, as you say, the patients who are on the treatment arm on the active agent live longer or are exposed to drug longer and so are subject to many more of the typical side effects that you'd be looking for. I think that's going to be really important data to look at.
NC: So a big question, I guess, is in the US is this population of patients going to be receiving treatment much earlier? In other words are we going to be adopting enzalutamide or apalutamide in the M0 castrate resistant space? I guess the question I have unresolved for myself is what will happen in Europe and other parts of the world.
ES: Well, that's the £85 million question. I don't know the answer to that. Some of it obviously depends on the regulatory authorities - is this going to be approved by the FDA? We know that the companies have submitted the data for consideration, I have no way of knowing what will happen. But making the assumption that it is approved, the question then becomes is this something that is likely to be adopted readily. As you, I don't know, I think that in this country there's a lot that is driven by patient request and expectation. As Maha Hussain pointed out, it requires an individualised discussion between physician and patient to understand potential risks and potential benefits. I suspect at the end of the day, and obviously once survival data come out that changes things altogether because if earlier treatment prolongs survival then that will be a less difficult decision to make; that's many years down the line. So based on this, I suspect that delaying metastasis for two years, delaying symptomatic progression, will probably carry the day. I suspect that for patients that have particularly aggressive disease, whose PSA is climbing rather rapidly, I suspect it's going to be adopted but it really remains to be seen. The other unspoken thing, and I know you would agree with this, is we have the luxury in the setting of a study like this or in an academic setting to calculate a PSA doubling time, many clinicians don't and will be extrapolating or flying by the seat of their pants saying, 'Oh well, the PSA last month was 0.2 and this month it's 0.5, ergo….' And that would be a mistake because we know that one or two PSA values don't determine, by any means, what the trajectory will be. It is very clear in my mind that patients with a very slow PSA doubling time do just fine being watched.
NC: Yes. I think that if I might conclude by congratulating you and Maha and the teams from the two trials because we have clearly got active agents, we've got data from two extremely well conducted trials. It gives us a problem of how to sort out who to give it to and when and this will be resolved in the fullness of time. But, for sure, it's been fascinating to pick up the data and we will be going through it for some time yet, I'm pretty sure of that.
ES: I suspect we will be, yes.
NC: So that's all we have to say from San Francisco. Thank you for listening, it's been a great pleasure. Thank you.
