NJ: Good afternoon. My name is Nick James, it’s my pleasure to welcome you to this ecancer session live from the ESMO meeting in Barcelona. I’m joined here by Nicholas Mottet, who chairs the EAU Guidelines Committee and a very prominent urologist, to discuss the data presented in the various prostate cancer sessions here.
NM: Quite. As you know, regarding guidelines there has been a major issue with the androgen sensitive metastatic disease where the endless discussion is high volume/low volume, high risk/low risk. In the guidelines we said from the very beginning that it’s purely subgroup analysis, we had to be very careful. I heard that it was a little bit changed at ESMO.
NJ: Yes, obviously this has been a very live controversy and we’ve always been very pleased, actually, that the EAU took the same line as the STAMPEDE group which was that we didn’t see a volume effect in the data that we had. But the specific thing that we hadn’t done was split our patients by the criteria that they used in CHAARTED – high volume/low volume driven by number of metastases and visceral metastases. So this is a retrospective analysis, of course, but we centrally pulled out around 90% of the scans, maybe 80% of the scans, centrally reviewed them and classified the men into high or low risk. Basically about 45% are low volume, 55% are high volume using the CHAARTED criteria. The other thing that I would highlight here is that in CHAARTED a very high proportion of the low volume patients were relapsed after prior therapy, and the same in GETUG-15 which was a negative trial but with a trend towards a positive result that was previously reported. Whereas, in contrast, in STAMPEDE around 95% of the men in both the high and the low volume groups are de novo so that’s an important difference. So what the CHAARTED group had previously reported and kept on stressing is that there didn’t appear to be a survival benefit in the low volume patients, only in the high volume patients and we have always taken the same view that you just put which is this was an underpowered subgroup analysis and it really didn’t fit; there wasn’t any biological reason why, apart from anything else, that we could see.
So we were very keen to do this subgroup analysis and we coupled it with a long-term follow-up. So the follow-up is more mature. In terms of the follow-up for the whole cohort the original result holds true. So docetaxel improves survival, improves failure free survival, improves all the other endpoints. So, unsurprisingly, what we reported is correct or is consistent. Then if we split it by high and low volume what we see is that there is a survival advantage in both low and high volume patients. Actually the absolute magnitude of benefit is bigger in low volume. So it goes from about a 15% absolute gain at five years in low volume, a 10% absolute gain from about 25% to 35% at five years for the high volume patients. So depriving low volume patients of chemotherapy you’re depriving them of a survival advantage at five years of 15% which we think is very important. The thing that we drew out was around the differences not between high and low volume, it’s that low volume relapsing patients are probably different to all of the other groups. Your fellow countryman, Karim Fizazi, discussing it essentially agreed with that view.
NM: That’s such a surprise. The first thing, it’s very reassuring that we were correct. That’s an issue that when you interpret trials you have to be very careful about subgroup analysis when it was not pre-planned. We try to insist on that. That’s why for EAU guidelines, for example, we accept the subgroup analysis for radiotherapy to the primary in M1 disease because the trial was powered for such a subgroup analysis but it was not the case for all the others. I remember several talks by specialist Susanna Labi saying, ‘Be very careful. Subgroup analysis is the best way to lie.’ Not on purpose but to find something just by chance. So we have to be very, very careful about subgroup analysis, it’s very reassuring.
I completely buy the fact that relapsed and newly diagnosed are different but in fact that’s common sense. If you relapse you are followed very currently. If you relapse one year after local treatment, come on, you were already metastatic before; the difference and probably are the same. But it’s against interpretation of data.
NJ: It is. The other thing about the relapsing patients, because it relates to the non-metastatic CRPC trials, is that the trend in the UK in particular is that we are PET scanning these patients after surgery or radiotherapy. You pick up relapses early so it completely alters the landscape. Rightly or wrongly, we’re not putting patients on ADT for a rising PSA only, we’re at the very least having PET detected, if not actually detected, metastases, if you like, on CT and bone scan. I’m sensing your scepticism here. Which is why we have such a low proportion of relapsed patients in our metastatic STAMPEDE population because we’re managing these patients rather differently. Obviously that’s a separate issue around the role of PET in detecting relapse versus CT bone scan, whatever.
NM: But then if you use that kind of imaging modality for relapse what you’re doing is a major lead time bias compared to bone and CT scan. So is the lead time bias associated with a prolonged survival? Not just lead time but the prolonged survival, yes or no? Nobody knows.
NJ: The thing that the early detection allows us to do… One of the things I’m increasingly concerned about is that when we set up STAMPEDE the average survival with metastatic prostate cancer was three years. If you look at the low volume metastatic patients that we’ve just reported results on, if they get docetaxel we’re looking at a five year survival of 80%. So these men are looking at very long durations of ADT so we need to look for ways to not put people on ADT for very long periods of time. One of the concerns I have with the non-metastatic CRPC trials is these men have already been on ADT for a while and then you’re intensifying it before they’ve even got any metastases. I think we’re going to be stacking up a lot of bone problems, heart problems, dementia, goodness knows what, because we know all these things are associated with ADT and we’re shoving the duration of exposure up. The thing about doing the PET scan is that we can give targeted radiotherapy and no long-term ADT to these patients. So it feels like the right thing to do and I know you’re going to tell me I have no evidence to support it.
NM: No, I agree with that. It might be the right thing to do. Again, I’m biased because of guidelines but we have to be very careful. You can do it; it’s not because we can do it that we have to do it. We can irradiate one or two single spots, for sure – the toxicity is absolutely acceptable, no question. The question is so what? Does it improve anything? Nobody knows. At the same time is there very strong and clear evidence that these patients need immediate ADT? Not at all. But the guidelines – salvage ADT at relapse – should be reserved for those with the shortest doubling time, the highest risk. That’s far away from being the majority, it’s mainly the minority. So it’s an overtreatment that leads to a non-metastatic CRPC.
NJ: Absolutely and we lack data on this.
NM: Yes, exactly.
NJ: The other, for me, interesting data which we were discussing was the data presented in the prostate session on Friday around adjuvant versus salvage radiotherapy. So there was data from three trials presented: the RADICALS trial Chris Parker presented, and this is around 1,200 men randomised to adjuvant or salvage radiotherapy with high risk features, so positive margins, positive nodes, things like that. The results were very reassuring. Firstly, it didn’t matter what you did, very few of these men had died, that’s the first thing. So it shows we don’t need to rush to do anything because the ten year metastatic rate is probably less than 5% in these. But the second thing was that in the RADICALS trial only around a third of men had had salvage treatment whereas obviously nearly 100% had had adjuvant. There was nothing to suggest a difference in the subsequent relapse rate. Then, linked to that, and part of a very interesting programme running out of the MRC Trials Unit in London, it was a meta-analysis of three trials: RADICALS, RAVES, which is a Trans Tasman Oncology TROG trial presented at ASTRO, and a French trial, GETUG-15 which hadn’t actually even reported results yet but had shared the data with the MRC Trials Unit. So it was an individual patient data meta-analysis, not pooled, so done in real time. This is a very interesting trend because all three trials showed the same thing, that adjuvant was unnecessary, salvage was just as good. Slightly different ways of doing it but basically there was nothing to suggest it was bad and, of course, the toxicity is less because you’ve only treated a third of the patients.
NM: That’s very reassuring and, again, it won’t change the guidelines but it will reinforce the proof we have and the level of evidence and the strength. What we wrote, 2019, adjuvant delays relapse. That’s evidence-based, all the three trials showed the same thing. None of the trials except one showed a survival benefit. None of the trials had a pre-defined salvage treatment. We know that in many of them patients were treated with a PSA as high as 2, 3, 4, 5, where it’s almost completely useless, all the evidence shows the same thing. So it’s very reassuring to see a randomised trial, and even better to see three randomised trials grouped together, confirming that adjuvant versus early salvage are equal in terms of efficacy, less toxic and you only salvage those who are in need for salvage and you don’t need to treat everyone. So that’s very reassuring.
NJ: Absolutely. The positive adjuvant trial is the oldest one, a lot of it was pre-PSA and you had to have a palpable nodule to get salvage and stuff like that. So they were very late salvage whereas the subsequent ones like the Bolla trial that was presented, the EORTC one, as you say there was a PSA benefit but no subsequent downstream benefit. So absolutely I would agree with that.
NM: That’s very reassuring for guidelines again. When you analyse very correctly what is available suddenly the message is a little bit different than when people say, ‘Oh, we have this strong evidence that it works.’ For PSA if it’s all relevant.
NJ: A secondary message from that is that PSA probably isn’t relevant here. Because if you look at, for example, the RT01 trial, which is the UK radiotherapy dose escalation trial but every other dose escalation trial has been the same, they all showed a benefit in PSA relapse free survival. No trial to date has shown anything on metastasis, overall survival. So you trade the dose up front for a bit of hormone therapy later but it doesn’t appear to have any survival penalty or any relevance whether you PSA relapse.
NM: What might be of major interest, I don’t know if it’s planned or not planned for the UK trial, adjuvant versus salvage, is to stratify by risk at relapse. Because we strongly suggested that many of the relapse patients don’t need anything because there’s such low PSA, increase of such a long doubling time, that okay.
NJ: Absolutely. I think the very low metastasis rate and the very good survival…
NM: There’s a very strong signal for that.
NJ: There’s certainly scope for saying exactly what you just said – can we back off even further? The other striking thing was that 1,200 men in the trial and it’s under 100 relapses each side of the thing with quite mature follow-up. So it shows that even patients with high risk features mostly are not relapsing, it’s only a third of them had relapsed.
NM: That’s absolutely interesting because we had a very large cohort of patients relapsing. If I remember it was [..] from 12,000 men. The specific survival at 15 years was something around 85%. Okay, relapse is bad news but you won’t die the next morning. It’s very nice to see randomised trials reinforce what is supposed to be there.
NJ: Yes, and it also reinforces the fact that the prognosis is getting better and better and we don’t need to urgently and radically over-treat people in a lot of cases. I think there’s a more general trend; I think we’re giving them too much.
NM: Exactly, we give too much local treatment for low risk disease; we give too much treatment for relapse. Maybe we give not too much treatment for the highest risk provided we know them. We don’t give too much treatment where they have metastasis.
NJ: The other thing we just need to say just before we wrap up, because we haven’t discussed it yet, is the DNA damage repair story. We haven’t seen yet the data from the olaparib in DNA damage repair randomised trial that’s going to be presented tomorrow at the Presidential Plenary, but there was a whole load of stuff in the prostate discussion session today around patterns of DNA damage repair abnormality. So it looks quite likely that we’re going to see a PARP inhibitor heading into the clinic with a licence off the back of the olaparib data tomorrow, I suspect. We haven’t seen the data yet but I think that’s something we have to anticipate.
NM: And there was a press release on that.
NJ: Yes and there’s a press release, so publication by press release.
NM: New England Journal?
NJ: You publish it in The New York Times and your shareholders.
NM: Exactly, before New England.
NJ: So doubtless this is something we’re going to be discussing in a lot of detail at a future meeting.
NM: Exactly.
NJ: So with that, that’s probably where we wrap up. So I’d like to thank you for listening and look forward to speaking to you again at a subsequent meeting.
