Duration of Immunotherapy in Melanoma

Transcript:Keith Flaherty, MD: Jason, Rene brought up this point about duration of therapy. And before we move to combination therapy, which I think is a big-ticket item to discuss, an anti-PD-1 monotherapy—treated patient who’s feeling well, looking well, really nothing to push you in terms of stopping therapy for toxicity concerns, has a nice ongoing response to treatment. The trials looked at one year; then the trials looked at two years, never in a randomized context; then indefinite therapy. That’s when the FDA approvals happened, basically licensing indefinite duration of treatment. How do you interpret that? And how do you guide patients in your practice who are in this common scenario of nice ongoing response, no toxicity?

Jason Luke, MD, FACP: I think, in the current situation, this is a patient-by-patient decision-making process. And that’s where it goes to the quality of the response and to the patient that you’re treating. There are patients who started out with low-volume disease that has regressed and there’s very little left. If a patient has gotten to a stable place for more than three to six months, perhaps we should follow-up with a PET scan to look for any hypermetabolic activity or even a biopsy to look and see if there’s any residual tumor there. Say to yourself, ‘If it’s both negative on PET and there’s no residual tumor in the biopsy, do we need to keep going with this?’ And we don’t know that answer, and I think that that’s where our comfort level is going to come with each patient, as well.

So, some patients are going to say, ‘Wait a minute, you were giving me this drug, it was working. You’re going to stop it? No way’. And I think it’s a negotiation that you have with them. But I think an evidence-based way in clinical practice to do it was, well, what can we measure? Well, is there any hypermetabolic activity in the tumor? Is there any residual active tumor in a biopsy? If those two answers are “No, I think you can have that conversation,” now how long do you go with that? I was talking with Mike Atkins at Georgetown. He said six months—we do it on everybody—and if it’s there, stop. I think it’s more nuanced than that. I think there are definitely patients who have minor regression who I would not stop no matter what. I would just not feel comfortable. But if the patient actually had a response and that was stable for months, I think that’s where you might start to look.

Keith Flaherty, MD: So, Rene, this idea of complete response or functional complete response as Jason is putting it, I think that helps for a set of patients. One of the issues that informs this, I think, is mostly anecdotal data, not really systematically reported yet, of a re-treatment effect. In other words, patients who have been on PD-1 antibody therapy stopped at some point because of protocol—a forced stop at some time point—who then progressed, but was re-treated and responded. I certainly have seen it. I’m interested in your experience in terms of that phenomenon. Does it influence your thinking in terms of stopping, with the idea that you could go back if a patient progressed at a later time?

Rene Gonzalez, MD: Yeah, it does. Going back to the original ipilimumab study, we did that—retreated if they got a good response. The duration of the treatment was limited by the toxicity. There’s only so much ipilimumab patients can take. Yes, they do respond. My sense is that the subsequent responses were less than the first one. Although I have some patients that are out many, many years now on ipilimumab retreatment. I have less experience doing that with an anti-PD-1 because the studies we’ve done have been basically unending. An occasional patient here or there I’ve taken off for a variety of reasons, mainly not related to the tumor or to the toxicity. And I’ve had one relapse and a couple not relapse.

Jeffrey Weber, MD, PhD: Let’s get data-driven. In the IST that I did with nivolumab plus vaccine, which morphed into more of a phase II nivolumab alone and ipilimumab failure trial, 126 patients have been treated so far; it’s an ongoing trial. At least 10 or 12 patients are still on treatment, for a total duration of two-and-a-half years. Seventeen patients have gotten to two-and-a-half years and stopped because they were stable, PR/CR, and got to the end of the two-and-a-half year treatment. Two of them had toxicity before, usually before a year, and then got to two-and-a-half years without progression. All of them, since that two-and-a-half year time point, have remained in remission; nobody has progressed. Seventeen is not a trivial number, and that number will probably get bigger because there are at least 10 or 12 patients still on treatment that have not reached two-and-a-half years but are either stable in PR or CR.

So, to me, that would say you don’t need to get treated forever. And this, by the way, was a trial where they got treated with nivolumab every other week, with a vaccine for 24 weeks, and then they went on maintenance for every 12 weeks for two years. So, again, we’re not talking a huge number of doses. We’re talking 20 total doses, which is less than a year’s worth of nivolumab, as it’s given every other week. I think that the least I would go would be a year.

I tell patients, if you have achieved stable disease, partial or complete response by a year, I would go up to that maximal response, plus either 12 or 24 weeks and stop, to a maximum of two years. I generally wouldn’t treat someone for less than a year, so I don’t agree with Mike. I see his point, but I don’t agree. And I wouldn’t go beyond two years. And it’s a tough one, because put yourself in the patient’s shoes: you’ve been getting this drug either every two weeks or every three weeks—in the case of pembrolizumab, for two years—and then somebody comes in the room and says, ‘Well, listen, I think you look great, it’s okay, you can stop.’ It’s kind of their lifeline, and a lot of patients don’t want to stop.

In Steve Hodi’s and my CheckMate-064 trial, in the beginning, it was going to be treat to progression, which is the labeled indication for nivolumab. Then we decided, okay, we’re going to cut it at two years, and then patients got concerned about it. And they came to us and said, ‘We don’t want to stop,’ so we then amended the trial back to allow those who wanted to to continue beyond two years. And I’d say a little more than a third of the patients will want to continue.

The other cohort will just stop. In that trial, as I remember, I don’t think anybody’s relapsed yet of those who have stopped, although it’s much earlier in follow-up. And to boot, you have Evan Lipson’s article in Clinical Cancer Research and some other anecdotal cases where you can re-treat the progressors and get a response. Now, Jason said an interesting thing. He doesn’t know if he’s had any pembrolizumab or nivolumab patients who have progressed after a response. Didn’t you say that a little while ago?

Jason Luke, MD, FACP: That had a big response.

Keith Flaherty, MD: Right.

Jeffrey Weber, MD, PhD: Someone who has a partial response or a CR in cutaneous melanoma, I’m in the same situation, which is great for patients. I’ve had mucosal patients have a PR and progress at week 36, week 48 in that trial I just described. So not every responder gets to two-and-a-half years. But a couple of those were mucosal patients. And the data that have been assembled show that the mucosal patients will do very nicely with PD-1 blockade, but they don’t do quite as well with not quite as high a response rate, not as long a response or duration of response as the traditional cutaneous melanoma patients. But there’s still benefit to PD-1 blockade.

Transcript Edited for Clarity