Episode 102: Adjuvant therapy in melanoma – how it will help inform the renal community

Dr. James Larkin describes learning from adjuvant immune checkpoint inhibitors in melanoma.

 

Tom:
I’m joined today by my great friend, James Larkin. I’m with Brian as well, of course, which is less exciting. Because Brian and I have been doing this together for a while, but James, you haven’t been with us before. Thanks so much for joining us today. What we’d like you to do, we always get people to introduce themselves because we sometimes struggle with that. And then what I was keen to do was just talk a little bit about this upcoming issue around the adjuvant pembrolizumab, renal cancer space, and actually what we can learn potentially from the melanoma data that’s come out previously?

James:
Okay. Well, first of all, thanks for the invite. Introduction, I’m the medical oncologist at the Royal Marsden in London and I spend half my time treating melanoma and then the other half treating kidney cancer. And then I guess, straight onto adjuvant pembrolizumab and kidney cancer, we’ll, we have a press release at the moment at the time of discussion. So we don’t know what the hazard ratio is. And that’s obviously a critical question. I guess in melanoma, we’ve had data now for three or four years for adjuvant treatment. So, that’s basically stage three melanoma-involved lymphnodes.

James:
And I guess the first thing to say is there’s a massive health warning here, talking about melanoma and kidney cancer at the same time. And kind of a maybe it’ll, maybe not, person to that is the fact that targeted treatment in melanoma has a clear benefit in the adjuvant setting. So, that’s [inaudible 00:02:06] targeted treatment. With what looked like comparable hazard ratios with checkpoint inhibitor therapies. So we don’t know, but I guess we’re not necessarily anticipating the kind of comparable efficacy shall we say in kidney cancer, given the experience of the last few years with adjuvant targeted treatment and kidney cancer. So that’s kind of one thing to say upfront, I think. Do you want me to keep talking?

Tom:
James-

James:
You’re going to ask me some questions? Go ahead.

Tom:
Yeah, I am James. So the first question I wanted to ask there was what is the absolute benefit of adjuvant pembrolizumab or nivolumab? Is it a progression-free survival benefit in melanoma, or is it just a progression free or is it progression and overall survival?

James:
Yes, this is the crux of it, I think. But if you look at the hazard ratios for relapse-free survival, it’s around about 0.5, maybe slightly north of that. Actually for targeted therapy and for checkpoint inhibited therapy. And we don’t yet have evidence of an overall survival benefit in melanoma with about four years of follow-up, something like that. But there is a slight caveat to that in melanoma. And I’ll just explain that, which is the sort of background in melanoma is that the first sort of modern effective adjuvant treatment was actually ipimonotherapy at a dose of 10 milligrams per kilogram.

James:
So that’s a higher dose than we’re used to obviously. And that trial, which reported a few years ago now showed a benefit for relapse re-survival hazard ratio about 0.7, that kind of thing. And a benefit for OS. This is in comparison with the placebo. But the catch there was that trial was done at a time when there was no effective treatment for metastatic disease. And obviously that’s changed now in melanoma. And of course it’s different in kidney cancer. So, to some extent, some of the data we have at the moment we’re slightly extrapolating from that. But to date, we still don’t have evidence of OS benefit for say [crosstalk 00:04:03].

Tom:
How long did the adjuvant ipi OS benefit take to materialize?

James:
Well, what I’m quoting there is a sort of five year follow up. So, that’s exactly it. You need to wait a long time.

Tom:
James, what is the overall survival benefit for ipi versus placebo?

James:
Yeah, it’s around about the same. So we’re talking about 0.7. So then absolute difference at five years, you’re talking about sort of nine, 10%, that kind of ballpark.

Tom:
Okay.

James:
And I think the other thing you have to say here is that ipi 10 is a pretty toxic treatment. The FDA approved it, the EMA didn’t. And I think this comes back to one of the critical issues for adjuvant treatment. And actually, while I think of it. A big question we have in melanoma is that stage three melanoma is quite a broad church. You’ve got some people at one end, very high risk of relapse with stage four. You’ve got other people at the other end and they’ve had sentinel node biopsy. They’ve got microscopic nodal involvement. A lot of those people will be cured already.

Tom:
So, what is that range of risk?

James:
Well, it’s from a risk of relapse of 10 or 15% at one end. And then at the other end, you’d be up to 50, 60, 70%. So, when you’re sitting down with the patients and you’re talking about a year of adjuvant treatment with potential side effects, it’s quite a nuanced discussion I suppose, as well.

Brian:
So I mean, that would cover like T1 renal cancer through node positive, pretty much that range of risk.

James:
Yeah. And I mean, I think that’s-

Brian:
That’s everybody.

James:
Going to be one of the really interesting questions about the hazard ratio in subgroups in keynote 564. And I mean, for the record in melanoma, it looks like the hazard ratio for relapse re-survival is the same across the stage groupings, which is fair enough. But on the other hand, the absolute risks in these patients with 3A melanoma are actually quite small.

Brian:
Sure.

James:
So, and again, this is quite a complex discussion to have with patients. If you’ve got someone with metastatic disease, the downside of the treatment not working is kind of very different from adjuvant when they might be cured already obviously.

Tom:
James, you might have spotted, but Brian and I are not melanoma experts. So just for my benefit here.

Tom:
So, ipi has a survival benefit advantage of 0.7 in five year survival data. And are you quoting data of ipi nivo versus ipi or nivo versus ipi with the DFS?

James:
Yeah, it a great question, Tom. Sorry for not being clearer, but there’s a trial looking at nivo, year of nivo versus ipi [inaudible 00:06:39] 238, and the four year data were presented and published six months ago at ESMO, something like that. And then, so the hazard ratio for benefit for nivolumab versus ipi is 0.7 on top of that. So a further benefit, if that makes sense.

Tom:
And that’s a DFS benefit, not an OS benefit?

James:
Correct.

Tom:
Great.

James:
And then, just for completeness, so hopefully it’s clear, there’s a pembro trial, which was done versus placebo and melanoma and the benefit there is the hazard ratios of about 0.5. So if you do the maths on those, which obviously one isn’t allowed to do, it’s sort of fairly consistent.

Brian:
Yeah.

James:
In terms of benefit. And again, actually, oh, no, sorry. I was just going to say for the record as well, this whole thing, nivo versus pembro, are they different? In the melanoma world where we’ve used a lot of both, the view is in terms of efficacy and safety, they’re pretty similar. And a lot of it comes down to sort of frequency of administration, that kind of thing. Sorry, I interrupted someone.

Tom:
Boy you’re allowed to interrupt as much as you like, it’s actively encouraged. So James, the next question I’d like to ask then is, so it seems to me that it’s reasonable to say that both pembro and nivolumab have a disease free survival advantage over four cycles of ipi and the hazard ratio per survival of ipi versus placebo is about 0.7.

Tom:
So it’s reasonable to give nivolumab and pembrolizumab to unselected patients in this group in melanoma and the risk ranges from about 10% relapse to about a 60% relapse. Is that a fair summary?

James:
I think that is a fair summary. I think the questions start coming though with checkpoint inhibitors in this lower risk group, because the reality is that 90% of people I think will have a year of nivo or pembro without much trouble. Bit of itch, bit of fatigue, that kind of thing. But I think we all know that sometimes even single agent therapy can cause major side effects and clearly some of them are irreversible. And the irreversible ones, endocrine, I guess we’re, sort of reasonably comfortable with you can take thyroxin and that kind of thing. But I quote a risk for adjuvant treatment with nivo or pembro of about sort of one in 200 or one in 300 of developing insulin-dependent diabetes, which is lifelong. And that’s real you can look it up, you can read the trials, that kind of thing. But when you say that to patients, that opens people’s eyes, I think quite significantly, I mean, having diabetes for the rest of your life is not trivial. And if you may well have been cured already for melanoma, that sort of makes people think in practice.

Tom:
So [crosstalk 00:09:25] you go, Brian.

Brian:
I was going to say, how many patients after that discussion will decline adjuvant therapy? And maybe it depends on their risk of course, but.

James:
Yeah, I mean my experience, I mean, again, and I work at a tertiary center, so we’ve got a selected group of people that we see, but most people who are coming in the first place want treatment, I think is one thing to say. Everyone’s very scared about their cancer coming back. So, at the end of the day, most people, if you have the discussion, will say well, I’d like the treatment that reduces that risk. So, not many. To answer your questions, Brian. And the other thing just to chuck in here is the business about targeted therapy and melanoma. So actually if you’ve got BRAF mutant melanoma, you can actually say, well, look, we’ve got immunotherapy or we’ve got targeted therapy.

Brian:
Right.

James:
And some people might say, well, actually if the efficacy looks similar, which we think it is in the absence of head-to-head trials, I’ll go for the targeted therapy, please. Because if I get side effects, they’ll probably go away as when I stop the treatment. So you that’s a discussion that we have in melanoma that obviously we’re not going to have in kidney cancer.

Brian:
So do you, for BRAF mutant melanoma in the adjuvant setting, do you actively encourage targeted versus immunotherapy or do you let the patient decide?

James:
I mean, I’d like to hope that we’d sort of talk about all the options actually put them all on the table.

Brian:
Mm-hmm (affirmative).

James:
There might be people who’ve got contraindications for one reason or another to one of the treatments or that stuff. But I think if other things being equal, we would often lower risk disease, 3A melanoma, that kind of thing, point people towards targeted therapy rather than immunotherapy for some of the reasons we talked about.

Brian:
And do you for, one more question Tom.

Tom:
Okay.

Brian:
Do you for patients who they’re not BRAF mutant, so you’re considering adjuvant immunotherapy, do you actively discourage some patients like a low risk patient? Who’s 10% risk?

James:
Yeah. It’s all about competing risk, I think.

Brian:
Yeah.

James:
And comorbidities. If someone had autoimmune disease or something like that.

Brian:
Sure.

James:
Talking about that would make you think. And the other side thing, and this isn’t ageist, I’ll say that now is that if you’ve got someone who’s relatively elderly and the life expectancy might not be that great, you have to actually ask yourself the question. Well, here’s a year of treatment that you kind of might not need. Why not just sit tight monitor carefully. And then if the patient does develop stage 4 disease, then you could treat them later. Which of course is one of the big questions here. Sooner versus later, now or later, which we don’t really know the answer to.

Brian:
So it sounds to me like, you there, isn’t a tipping point where you say, I wouldn’t be recommending below, you know, below that risk. It sounds like you you’re happy. Is there other subsets within the forest plot, which you at and say, I wouldn’t be treating this population or that population based on other parameters or do you just accept the ITT trial and you offer everyone therapy?

James:
Yeah. The forest plot’s pretty consistent, I think to answer that question and if there’s some sort of hard cutoff we have in our head, it’s a one millimeter burden in the sentinel node, which is what was done in the trials. In other words, more than one millimeter evidence, less than one millimeter not really evidence. So if you’re going to give it to someone who’s less than one millimeter, you need to really be able to justify that pretty well. And I think the other stuff does come down to competing risk and co-morbidities. If you’re going to give someone a year of this treatment and sometimes again people might not want to come up to hospital every four weeks or every six weeks or whatever for a year to have a treatment they quote unquote may not need. So, it’s a lot of different factors actually.

Tom:
James, in melanoma, you have a survival advantage and therefore you’re giving these patient’s therapy. Ipi has a survival advantage and it sounds like nivo and pembro have substituted ipi. If you didn’t have that survival advantage, would you be as positive about treating these patients in the knowledge that there is underlying risk of long term adverse events?

James:
I think obviously is more nuanced is all I would say. When you come back to it and you’ve got positive trials, presumably regulatory approval and remembering that RFS or slash DFS is the regulatory standard in this area. And you’re having a conversation with someone about stopping them developing metastatic disease. There is quite a lot of pressure there, shall we say, but I think inevitably in the absence of OS data, it’s more nuanced than the kind of extrapolation that I’ve just talked to you about in melanoma. So I’ve sat on the fence slightly there.

Tom:
I’m not sure what the answer was, but my question is does, has the attitude changed over time? So you’ve had these treatments of around three or four years, right? Have people’s attitudes changed over time. Maybe you see that patient as bad diabetes or has some bad outcome or?

James:
Yeah. I mean, I think as these treatments kind of roll out and people get more experience, what tends to happen. I think with single agent anti PD, one is you’ll treat a few the first few patients things will be well. And then I think what focuses people’s mind is when they get a bad side effect actually.

James:
Practically speaking, I think the other kind of important point, it comes back to targeted therapy and this whole balance between BRAF mutant inhibitors and checkpoint inhibitors. I think in the early days of the targeted therapy adjuvant trials, a lot of us thought that you’d get benefit for while you were on treatment, that year of treatment. And then the curves would come together again for targeted therapy and that didn’t happen. So we’ve got 5 year data now for targeted therapy so that we can be confident that there’s potentially durable benefit there. And I think seeing that pushed people more towards BRAF inhibitors in a situation where there’s a choice. So, I think inevitably though, to try and answer the question, there’s a lot of enthusiasm to start off with. People get a bit more experienced and, yeah, again, it becomes nuanced.

Brian:
Did any of your colleagues come to you James and said, I’ve given this a few times. I can see there is a benefit, but you know what? I think the risks outweigh the benefits. Are there any of the melanoma community who just say, this is too risky for our patients?

James:
I think there’s a blanket statement, you can’t really say that, particularly if you’re dealing with really high risk. I think that the lower risk ones 3A, let’s say, it’s the kind of thing people talk about at meetings, but I mean, does anyone really say you shouldn’t have treatment? No. I think people are just more circumspect actually. At least people I’ve come across anyway.

Tom:
And 3A is the 10 to 20% chance of relapse?

James:
Yeah. That’s that kind of ballpark. Microscopic involvement as Sentinel notes. It’s a big group of patients because a lot of patients with melanoma will only have that, they’ll have their primary [inaudible 00:16:20] positive Sentinel node. So in the clinic, actually, it’s quite a lot of patients there.

Tom:
In years gone by, we’d be sitting in Chicago in a week and a half time listening to Tony Chiari. Well, Brian might not be, Brian might be somewhere else, but the rest of us would certainly be in the room, listening to Tony present his data. We in the renal cancer community are not really used to these presentations. And what I’d like you to do James now is just highlight some of the pitfalls of data assessment in the adjuvant setting. And some of the nuanced points. Maybe central review versus non-central review, randomization, risk population, event rate, maturity of data. What should the discerning oncologists be looking for when they’re making a decision about whether they feel this is practice changing?

James:
Yeah, well I think yet at that hazard ratio number one, but we know that’s positive already. And so let’s assume that there’s a solid hazard ratio. I mean, the next thing I would always say is what’s the inclusion criteria for the trial? How representative is that of the patients in the clinic and especially subgroups, we talked a little bit about that already, but again, kidney cancer, receptive kidney cancer can be quite broad. So are we seeing the same level of hazard ratio in all the different subgroups? Scan review. That’s obviously important. That’s something to think about. Because we all know that sometimes we’ll see chest CTS and that things are equivocal. I think, to say that you need to biopsy everything to demonstrate metastatic disease is probably a bridge too far, but I think, that that’s an important variable.

Tom:
Do you think central radiology review is important?

James:
Well, it can’t hurt, would be my answer to that. Whether or not it’s mandated by regulators or others, I think is a separate question, but-

Brian:
I assume there was central review in this study. Right.

Tom:
But does it need to be your primary endpoint.

James:
Based on the central review?

Tom:
Yes. Do you think investigator-assessed review is adequate or does it need to be a central review?

James:
I think central review is better. I don’t think that’s a controversial statement, but then on the other hand. I mean, there’s quite a big history of this question, investigator versus central. But I don’t have a really strong view on it, Tom, but only to say that I think, central views bound to be preferred if possible, but I’m not sure at the end of the day. Go on, for it.

Brian:
I think it’s probably mostly in the baseline scan where central review’s going to find a lot of people who have little stuff, which may or may not be [inaudible 00:18:57], but that’ll end up getting excluded from a trial. I think that’s probably where it makes more of a difference.

James:
Yeah.

Brian:
But again, it’s controversial and I mean the inclusion for the pembro study, I mean, we know that that’s public information. I don’t know it off hand, but I assume it was T3 node positive and maybe high grade T2 or something. That’s been typical. I don’t know if Tom you know?

Tom:
Yeah. I mean it’s a high risk population clearly. And it was, as you would imagine, it’s a clear cell population of patients with high risk disease. It includes a subset of patients who have had a metastasectomy,

Tom:
So an n1 population, subgroup of patients, obviously a small subgroup and their risk may be the same. It may be different, but essentially it’s a high risk clear cell study as you would. And there’s nothing super unusual about it. James, when you first saw the melanoma data, the ipi data with a survival advantage, it probably caused a big stir in the community. We are not going to see that because we have a progression-free survival advantage. What sort of data would you see as being in your mind, practice-changing, assuming the forest plot analysis as is the case in melanoma. And we don’t know the data obviously, but assuming that’s been the same throughout and what sort of figures are you looking for to say, okay, that’s really meaningful to me. I want to use this on my patients as soon as I can?

James:
Yeah. Well, if we’re talking about hazard ratio, I’m assuming that it’s going to be 0.7 or better, or thereabouts. And I think it’s pretty difficult to ignore that actually assuming there isn’t any, anything unexpected in the data. I mean, we always want more mature follow-up, but of course we have to make decisions in clinic in the early days. And just to come back to something you said, actually at the stage four NED group receptive stage four, that’s actually a pretty important question.

James:
And for what it’s worth in melanoma, there were a few patients like that in the trials for hazard ratio stayed the same as for stage three. So, that’ll be something else that’ll be interesting to look at. And actually while I’m talking, the other thing is this is clear cell, right?

Tom:
Mm-hmm (affirmative).

James:
And I think it would be remiss not to mention non-clear cell because it’s 25% of people in the clinic. We don’t have data or we won’t have data if it’s clear cell only, but all the patients with non-clear cell and we’re asking the same questions. So there’s a gap there, I think that we’re going to probably need to deal with somehow.

Brian:
So two questions, let’s say the hazard ratios is 0.6, 0.7. Obviously no survival. Yeah, the tolerability is as expected. How will you apply it in practice? Will you say, okay anybody who is eligible for the trial, whatever that may be? Or will you focus it towards high risk, given the melanoma experience and the toxicity and whatnot?

James:
I think I’d have the discussion with everybody who fits the bill as it were in terms of the trial. And it’s going to be a well tolerated treatment. We know that already. I don’t think there’s going to be any surprises there at all.

Brian:
Yeah.

James:
So, the question sort of is, when would you not give it? I think.

Brian:
Yeah.

James:
I mean, the hazard ratio is solid. And it might come back to some of the stuff we talked about for melanoma. Comorbidities, age risk, etcetera, etcetera, etcetera,

Brian:
Yeah. And how about non-clear cell? Would you give it to a non-clear cell patient?

James:
Well, that’s a great question, isn’t it? And assuming you’ve got approval and reimbursement.

Brian:
Yeah. Let’s assume that.

James:
Yeah. It’s data. It’s going to be data free, isn’t it?

Brian:
Yeah.

James:
So then you’re going to end up having to extrapolate from the metastatic experience.

Brian:
Right. Where we know it’s active, but less active at least than say papillary et cetera, or less active [inaudible 00:22:50] chromophobia so you’re doing a lot of extrapolation.

James:
Yeah. It is a lot of extrapolation. No doubt about that at all, but I think again, if you can access it and you’ve got someone with high risk papillary, would you not treat them? I actually, I think if they understood all of that stuff. I’d be prepared to do that in principle. I don’t have a massive problem there. Shall we say.

Tom:
James, we’ve got a last question for you. Duration of follow-up on these trials. What happened in melanoma? We’ve seen in metastatic clear cell kidney cancer, we’ve seen hazard ratios for the [inaudible 00:23:24] TK immune therapies for survival drift from 0.5 to 0.6. Still a massive difference, fantastic results. In melanoma did you also see the hazard ratios? Were the early hazard ratios better, and will we get the best results now and what sort of follow-up for DFS do you think is relevant? Because some people could argue if the follow up is really short, all you’re actually doing is just treating early metastatic disease?

James:
Yeah. All great questions. So the hazard ratios, melanoma stay pretty constant with follow up, moving around a little bit, but not too much. I think the business about early metastatic disease. I mean, we know that’s an issue in kidney cancer with some patients they’ve got metastatic disease. We’re not just picking it up. So, I think we’re obliged to have decent duration of follow-up. And I mean, unless there’s some reason not to, I mean, I think we should be seeing 5 year data. There’s all sorts of interesting questions. One of which, is it better to treat in the adjuvant setting or not? Are we curing some patients with checkpoints and [inaudible 00:24:37] of metastatic disease, etcetera, etcetera. And I don’t see how you can get any traction on those questions without having mature follow-up in the trials, actually.

Brian:
So, my last question, and Tom, I’d like to say, I think we’re asking better questions for a disease we don’t know anything about. So we may want to focus to melanoma.

Tom:
We could do ovarian cancer next week, Brian.

Brian:
You have long-term follow-up in melanoma. So that was my question, of delaying versus curing, a proportion of patients. Is it the thought in melanoma that you’re actually curing a fraction of patient, call it 10% as opposed to just delaying? Or can you say?

James:
Are we’re talking about metastatic disease, Brian?

Brian:
No, no, no melanoma in the adjuvant setting. So is the thought that, yeah, we’re actually curing that extra 10% of patients that we wouldn’t have?

James:
Yes. Yeah. Well, that’s what we think. If you look at distant metastasis, free survival in the trials and various other things, it’s all pretty consistent basically. So I think, we think we’re curing people. Yes. To answer that.

Tom:
Yeah. Brian question for you. You’ve been around the block a few times. What are you looking for in the dataset when it comes out week’s time? What sort of numbers, what duration of follow-up, what would be practice changing for you? Do need to see an OS curve too, as well. What’s going to float your boat?

Brian:
As we’ve talked about, I think we’re expecting to see DFS hazard ratios in the 0.6, 0.7 range, just based on how the trial was designed, I assume. We’re not going to see overall survival. And I don’t know if they’ll even show those curves, but I wouldn’t expect them to look different. We’ll see the tox profile we expect from single agent pembro. So I’d be interested in those numbers, but in the forest plot and the subsets, certainly receptive M1 disease, and lymph node positive disease, have a very high risk, I could get my head around. As it gets lower risk, I think that’s going to become a little more of an interesting question. And again, I don’t know that we’ll ever see OS given we have, at least four regimens that extend OS in the metastatic setting, if not more. So I think it’ll be as expected and I think it’ll be widely adopted. the question is what happens over time and will the other trials also be positive? And we don’t have the targeted therapy option like they do in melanoma.

Tom:
James, I’ve got a last paradox that I wanted to put to you. In metastatic disease in good risk patients. It looks like immune therapy does not outperform sunitinib as it currently stands. Which is why?

James:
For survival.

Tom:
Yeah. For survival, its very [inaudible 00:27:05] not outperforming sunitinib. However, it now, and clearly in the adjuvant setting, by almost definition, those patients are super good risk or good risk disease. And yet we are seeing, or potentially we are seeing this DFS advantage for immune therapy. So, the biology points us towards immune therapy working in more aggressive disease. But now in this early setting, it looks like we have immune therapy working very well, but we know sunitinib has not worked well previously. How do you pull that together?

James:
I can’t. It’s a great question, but I mean, maybe one way, and I’m speculating is to say that in the adjuvant setting, intrinsically, maybe the biology is determined in individual patients in terms of sensitivity to drugs, rather than it being something to do with the disease progressing and things changing as you go along.

James:
But no, I can’t answer it. And it’s pretty striking actually that if you compare with melanoma that you’ve got this benefit for adjuvant treatment, with targeted therapy, whereas we know targeted therapy has worked quite well for metastatic kidney cancer, but it was a real struggle in adjuvant trials to show any benefit. I can’t answer it. Sorry.

Tom:
Well, James, thank you very much indeed. Brian, last question from you. We’ll call it there.

Brian:
No, let’s call. This was great. I learned a lot. Appreciate setting the stage for the data next week.

Tom:
Yeah, James it was terrific. Thanks so much for your time.

James:
Thank you very much.

Brian:
Take care, gentlemen. See you soon.

Tom:
See you.

Related Episodes

Episode 168: Mechanisms of response to PD1 therapy in renal cancer
Episode 167: Highlights from EIKCS oral session
Episode 163: The Uromigos Debate – Treatment of favorable risk renal cancer
Episode 150: Comparing the quality of life data in renal trials
Episode 149: Treatment-free survival for ipilimumab and nivolumab
Episode 147: Toni Choueiri Renal Cancer Christmas Special
Episode 117: PD1/VEGF adverse event management in renal cancer
Episode 140: Ipilimumab in renal cancer
Episode 135: The STAR Study
Episode 138: Renal Cancer ESMO Guidelines Update