Long-Term Results for Immunotherapy in Melanoma

Transcript:Jeffrey S. Weber, MD, PhD: Let’s go on to what happens in most melanoma patients in community practice, which is that they get treated with a single-agent PD-1 inhibitor. I think Caroline Robert had some very nice data, Georgina, looking at the long-term outcomes in some of the pembrolizumab patients. Can you tell us, what was the take-home message?

Georgina Long, MD, PhD: She presented results from KEYNOTE-006, which is the phase III randomized trial of 2 different dosing schedules of pembrolizumab versus ipilimumab. And we know there was a clear benefit in terms of response rate, progression-free survival, and overall survival, compared with ipilimumab. But what was most interesting about what she presented were data on patients in this trial—people only got the pembrolizumab for 2 years. So, there were data about what happens once you stop the pembrolizumab.

There were 555 patients who had received pembrolizumab, and only between 100 and 150 actually got to the 2-year mark and got the full schedule of pembrolizumab. It was a reasonably short follow-up, 9 months. We saw that over 90% of patients were still in response, but there were some relapses. Now we’re just looking at “OK, you’ve had a response or you’ve been on pembrolizumab with clinical benefit for a long period of time. Now you’re off it.” The majority of patients at this short follow-up were still doing well and remained responsive. What happens when you stop and rechallenge with anti—PD-1—will patients respond? There was a great poster at ASCO, again with small numbers, which suggested that you may respond, but we still don’t have those data. That was one of the major take-home messages.

The other one was the acquired resistance, meaning once you have a response—and this is taking the whole study, not just the patients who got the 2 years. So, everybody who had a response to pembrolizumab, what happened to their response over time? We did see, interestingly, at 2 years, only approximately 70% were still in response, which means 30% of responders are developing an acquired resistance. So, that’s another issue we need to consider.

Jeffrey S. Weber, MD, PhD: I was struck by the initial data that we saw last year, again presented by Caroline Robert: 97% of those who had a CR stayed at remission for a year. Now it’s 91% of even PRs in stable patients.

Jason J. Luke, MD: That’s the one that really stood out to me, because that’s not the way I had previously thought about this. I often think about patients who have stable disease who are getting out a ways; I’m starting to be on the lookout, as I have concern that they’re going to progress. And yet that data set argues that that’s not from the randomized data…

Georgina Long, MD, PhD: I don’t know, I was concerned because of the ones who have progressed. They were mainly from that group. Jeffrey S. Weber, MD, PhD: In the stable group.

Jason J. Luke, MD: Sure, but we’re talking about more than 2 years later.

Georgina Long, MD, PhD: Well, no, because they were the ones who got 2 years, and we’ve only got a median follow-up of 9 months. So, that’s why I’m cautious. I think it’s early yet, and I’m worried about that. I’m still worried about the stable disease group.

Robert H.I. Andtbacka, MD, CM: I agree with Jason. I think it was striking that you had the stable disease patients, the majority of whom were 9 months out and had not progressed at that point in time. I think it shows several points. First of all, it was a very small number of patients. I think it’s really difficult for us to truly make a statement on what “stable disease” means there. But the second thing it also shows—and from the surgical side we see it all the time—is that you may have a tumor that is there on radiographs, but when we surgically resect or biopsy it, there’s no viable tumor left.

So, this is fibrosis, this is inflammation, but there’s no viable tumor left. And to me, that was the take-home message from this. For many of these stable disease patients, maybe we need to reconsider: If you stay stable for a long time, is there actually any viable tumor left? Do we need to continue this anti—PD-1 therapy for the full 2 years? We don’t have the answer to this yet, but I think that this is the next step, where we need to really consider more biopsies and try to understand how long we really need to continue treatment in these patients with stable disease.

Jeffrey S. Weber, MD, PhD: Yes, but the nice thing is that all of us have had patients who had a 3-cm lung mass and some other small volume disease and got treated with a single-agent PD-1, and they come back, never regress, never grow—they just sit there. Then it’s biopsy proven, so there’s a tumor, and a year later, 2 years later, or 3 years later, they’re still coming back. Those are the patients in that category, and I wish I could understand what was going on. Georgina?

Georgina Long, MD, PhD: I was just going to give the same anecdote. In fact, I’ve got quite a few patients on KEYNOTE-006, and the ones on pembrolizumab have now stopped. There’s 1 in particular who has very pigmented, raised, in-transit disease throughout his whole leg. We had a biopsy before he started—melanoma. We had a biopsy at week 1—starting to regress. And we’ve done multiple biopsies now that he passed 2 years, and it’s just macrophages. But it looks exactly the same as when he started the trial.

Robert H.I. Andtbacka, MD, CM: We see this on the legs—it’s melanophages, and there’s no viable tumor. So, I think that for us within the melanoma community, this also challenges us now to really understand who these patients are who have visible disease, yet there’s no biological disease there. How do we pick these patients out? Because I think we do have to remember that with continuing many of these therapies, such as the anti—PD-1 therapy—if we continue them beyond 1 year, up to 2 years—there are more chronic side effects that we see, including arthralgia and so on, that can become debilitating for some patients.

Jason J. Luke, MD: So, I have a provocative question then. Does anyone on the panel feel like there is a justification for stopping patients at any point? And this is something we also talk with each other about all the time. I think we all feel good when we see CRs. Then you ask, “Sure, maybe could we talk about it, but what about PRs and how much of a PR?” and so on and so forth. In my practice, after about 6 months to a year of treating a patient, if there’s a quality response—which I would say is 50% or more—I start to entertain that and talk with them: “What does it mean to you to come to the clinic, and what would it mean to you to stop treatment?” But I think some people are on the end of “Look, never stop, because we don’t know what happens.” Some patients want to keep going. Some don’t want to come at all. So, I think it’s a really wide-open question.

Jeffrey S. Weber, MD, PhD: I feel the same way you do. Does everyone else have the same view that at some point you have to stop?

Robert H.I. Andtbacka, MD, CM: I think that this is also one of the challenges that we face: With an anti—PD-1 therapy, we go for 2 years, yet with anti¬–CTLA-4 therapy, it’s usually 4 doses. So there, even when you have the 4 doses and you stop, we know there is a good group of patients who will have stable disease. Like you said, Jeff, they don’t grow, they don’t get smaller, but they just sit there for a long time. But with this therapy and the way it’s set up, we go for the full 2 years. But I agree, I don’t think we necessarily need to do that.

Michael A. Davies, MD, PhD: I think what’s really going to be interestingly important is that most of us have been having those conversations when we have patients achieve a CR very quickly—how long do we feel like we need to treat them? And, again, the provocative part is with these results: Are there other patients who haven’t achieved a CR where you would feel comfortable stopping them earlier on? And I think that’s a moving target, but it certainly would be something that would be interesting to think about over the next few years.

Jeffrey S. Weber, MD, PhD: So, it’s duration. But the other question is, how much drug? Whom do you decide to treat with the combination, and [whom] do you decide to treat with the single agent? Jason, how do you look at this? When the patient comes into clinic, how do you have that discussion?

Jason J. Luke, MD: Well, to make this binary, we’re going to just postulate this is a BRAF wild-type patient.

Jeffrey S. Weber, MD, PhD: That’s correct.

Jason J. Luke, MD: Because if you add in the mutation, you get a whole other question about if you should lead in with the targeted inhibitors. But, as I was mentioning previously, I really look at basically the same clinical factors that have been published by this group about clinical factors that impact targeted therapy. Because I think that those are broad factors that impact the outcomes of the patients with melanoma.

So, if I see a patient who has a rapidly rising LDH level that’s out of the normal range, many sites of disease, and brain metastases—which I now really strongly feel validated about, and we’ll talk about later—those are the patients where it’s just a no-brainer. Unless there’s a clear reason where I think that there’s no way they can tolerate it, that’s what we’re doing. Once it becomes less than that, it’s much more nuanced then, and I talk to the patient about toxicity and so on and so forth. But there’s a certain set where, for sure, that’s what we’re doing. Then it becomes much grayer if it’s much less than that.

Transcript Edited for Clarity