 Let’s turn to refractory disease. We’ve gone through resectable disease, adjuvant, neoadjuvant, frontline, and then comes the tough nut to crack. This is always the biggest challenge. What do you do with a patient who’s been through your frontline therapy, or a frontline and second-line therapy? Now you’re in third-line therapy, or if they fail adjuvant therapy, and then they get, say, combination immunotherapy [IO], then they fail, then what do you do? Adil, let’s come back to you. How do you approach patients who progress on IO if they’re 

 wild type, or if they progress on IO, then they get 

, if they’re mutated, and now you’re into third-line therapy, then what do you do? What can you tell us about resistance and about how you choose what to do next?

 This is a challenging population, and it’s a hard nut. I have to say that the last couple of years haven’t been super encouraging. From what I can tell looking at the PD-1 refractory tumors, most of those patients don’t have a lot of T cells left. That probably is why we’re seeing disappointing results with additional checkpoint blockade. Once you’ve wrung out PD-1, you’re not going to get much more from other checkpoints, simply because there are not a lot of T cells left in those tumors.

There might be exceptions to that rule. Maybe 10% or 15% of the time, you could restart treatment if you haven’t had it for a while. Your question really goes to the heavy-duty PD-1 refractory field. One encouraging signal we are seeing is getting a new bunch of T cells to come into your tumor. I personally like interleukin-12, but I think you are seeing some encouraging data from these TLR [toll-like receptor] agonists—there was the CheckMate trial that was reported at SITC [the Society for Immunotherapy of Cancer meeting] last year. There are also some intriguing responses, where you’re getting T cells coming into your tumor. You’re turning some lukewarm tumors to a little bit less lukewarm tumors. Then once you add PD-1 to that, you can get responses to it.

With the targeted therapy field, 5 or 6 years ago, there were papers showing that 

 mutation might be responsible. I don’t think they panned out enough in the data that today we can say we really understand what’s happening with BRAF resistance. It’s a challenging field, and you’re probably going to rely on things like the…study where they’re looking at adoptive T cells where, outside of the immune suppressive microenvironment, maybe you can get some fresh T cells to ramp up and then put them back in. You might get some response, although I think the jury is still out on how long term those responses are and whether those are really immune responses, or whether it’s the preparative regimen that’s doing some of the heavy lifting for you in that setting.

Videos

Novel Approaches for Relapsed/Refractory Melanoma

 Let’s head back to Jason. The ipilimumab-nivolumab track record in brain metastases really did change my practice. Those are pretty recent data that have come out over the last year or 2. What can you tell us about your approach in treating melanoma patients who have brain metastases in frontline therapy? How do you decide ipilimumab-nivolumab, pembrolizumab, or nivolumab alone versus 

 This has been an evolving area that still remains the toughest spot in melanoma to deal with. What has really come forward is that there’s just no question anymore that the standard of care for patients with brain metastases that are asymptomatic or low-volume where you don’t have to intervene immediately is ipilimumab-nivolumab. I would even go farther to say nivolumab plus ipilimumab-3, although maybe you guys want to argue with that.

The reason for that is 2 trials. One was out of Australia and led by Georgina Long, the ABC-X trial, as well as the CheckMate204 study that Hussein Tawbi led. If you look at patients treated on these trials, this is a selection of patients with brain metastases. This isn’t the run-of-the-mill patient. The people who present with small-volume or asymptomatic lesions, the landmark 1-year survival is 80%. You think about dating back 5 to 10 years, the median survival for these patients was 6 months, and now we have 80% of them alive at a year. They’re durable. It’s response in the body. It’s response in the brain. It’s really quite impressive.

In my practice, that is enough that I don’t even radiate these patients. I give them ipilimumab-nivolumab, and then we watch and monitor. If we have to, we radiate later. These data really do change the game.

I mentioned that that isn’t every patient. In fact, it’s not even a majority by any means of patients with metastatic melanoma. Many patients present, they need steroids, they’ve got shift, and they’re sick. In those patients, it’s a barrier to think you’re going to give ipilimumab-nivolumab there because if they’re on steroids, it will blunt the effect, and/or they’re already too sick to think you’re going to stress them.

It’s really all hands on deck. That’s where 

mutational analysis is super important. Certainly, BRAF inhibitors are active and 

 combination in CNS [central nervous system] metastases. Mike Davies presented the COMBI-MB study some years ago showing that you get responses as you would expect in the brain. The caveat is that they are not as durable as they would be otherwise in the body. Whereas the median PFS [progression-free survival] for 

 in the body is on the order of about 11 months, in the brain it’s about 6 months. For patients who present with symptomatic brain metastases, I try to think about if there a way we can get them to ipilimumab-nivolumab, because the data for ipilimumab-nivolumab are so impressive.

It’s worth noting as a follow-on that Hussein Tawbi updated last year’s ASCO [American Society of Clinical Oncology Annual Meeting] with the symptomatic population, where they used ipilimumab and nivolumab and got not nearly the same kind of results. They had a response rate of about 20%, perhaps. Even those 20% were outliers who had weird stories and weren’t representative of the usual patient you would see if a patient is symptomatic with these high doses of steroids or has a shift or something like that.

The quick is that ipilimumab-nivolumab, if you can do it, is for sure the right answer. If not, use everything you can get, with the 1 caveat being not whole-brain radiation. At ASCO last year, there was a nice study looking at adjuvant whole-brain radiation after resection. The results were equivocal. There was no difference between the whole brain and not using whole brain.

If you looked at the numerical data, whole-brain was actually worse than doing nothing. All of us have treated patients and given whole-brain, and whole-brain is awful for patients, and in melanoma, it does not add anything. Please don’t use whole-brain radiation. We need to SRS [stereotactic radiosurgery] some lesions, we can use the 

, but try and get to ipilimumab-nivolumab. Do not use whole-brain radiation. Call everybody you know to try to get as much advice as you can because these are hard patients to manage.

 Su, will you forgo stereotactic radiosurgery in an asymptomatic patient who presents front line?

 That’s an excellent question. We usually always discuss these patients at our multidisciplinary melanoma tumor board. We have radiation oncology there, obviously medical oncology, surgical oncology, etc. If they’re asymptomatic and they are up to 3 cm or so, we usually wait. Three is probably pushing it in terms of sizes, but we try to forgo radiation therapy and give them some time on ipilimumab-nivolumab. We follow them very closely with respect to symptoms and radiographic imaging. We’ve had situations where their lesions have responded beautifully, and they haven’t required radiation. With SRS, although it’s fairly well tolerated, there can be adverse effects. We’ve seen radiation necrosis. It’s certainly not a benign modality.

 One last interesting thing is, in the trilogy or the IMspire150 trial, there were some data about the time to CNS metastases. This was presented prominently at AACR [American Association for Cancer Research] 1 month or so ago. Ryan, what can you tell us about that? You know that study pretty well.

 The study randomized patients to receive a triplet of vemurafenib, cobimetinib, a BRAF-MEK inhibitor, for 28 days, and then after 28 days, to fold in atezolizumab therapy. The comparator was vemurafenib, cobimetinib, and atezolizumab placebo. The data for the primary end point was progression-free survival. There was a progression-free survival advantage to giving triplet therapy versus doublet therapy.

It didn’t allow patients with active brain metastases, but there was a subset analysis looking at the time to develop brain metastases in patients who received doublet versus triplet therapy. It seemed to be a numerical difference, meaning the patients who got doublet therapy had a shorter time and were more likely to develop brain metastases than those who got the triplet therapy, though it didn’t meet statistical significance. It was a subgroup analysis. It wasn’t a ton of patients. I can’t really say that’s the way to go in terms of treating patients in frontline with brain metastases that are asymptomatic and overtreated. That’s a different experiment, so to speak.

There will be those types of trials. Already, there are a couple of trials, including a trial in SWOG that’s looking at triplet therapy.…That will be an important trial to understand whether targeted 

 and combining that with PD-1 is as effective in patients with brain metastases than just nivolumab.

Management of Brain Metastases in Melanoma

 Let’s turn back to Ryan. Jason had alluded to a couple of scenarios where alternate doses and schedules of immunotherapy were used. At this ASCO [American Society of Clinical Oncology annual meeting], there was at least 1 very interesting oral presentation by Mike Postow. What can you tell us about the different doses and schedules? Tell us a little more about how you treat your patients in real life, off protocol.

 In real life, off protocol, I tend to treat patients as my colleagues do and often give single-agent PD-1 or combined immune checkpoint inhibition. The reason I don’t give combined immune checkpoint inhibition to everybody is the high toxicity rate. That’s obvious. It’s probably also why the difference in overall survival between 52% and 44% at 5 years isn’t 70% and 44%. Because the toxicity is high, a lot of patients need corticosteroids. It’s not a high percentage, but some patients certainly die from the toxicity more with combined.

There’s been a push in the field to look at alternative ways of delivering combined immune checkpoint inhibition. One way of doing it is to lower the dose of ipilimumab. Jason alluded to the CheckMate 511 study, which was comparing alternate dosing of 3 mg and 1 mg ipilimumab-nivolumab versus 1 mg and 3 mg ipilimumab-nivolumab times 4 doses and then maintenance nivolumab.

What’s clear is the toxicity is less with the flipped dose. What’s also interesting is it’s probably a little early. There was only about 18 months of follow-up to see a real difference in overall survival with that. When you think about when the overall survival advantage was seen with ipilimumab and nivolumab versus nivolumab in CheckMate 067, it was actually 2 and 3 years when you started to see those curves separate, not 1 year, not even 18 months. We can feel good about the fact that it was less toxic, and the relapse-free survival looks about the same. I don’t think we can comment much on overall survival with alternative dosing of ipilimumab until you have longer-term data.

At this ASCO, Michael Postow and his colleagues at Memorial Sloan Kettering Cancer Center presented data on a trial where they thought they could get away with 2 doses of ipilimumab and nivolumab for patients and then do a scan. If the scan looked good, meaning there wasn’t any growth, then that would be it, and they could switch to maintenance nivolumab. If patients were having some additional growth of their disease, then they would continue to receive the final 2 doses of ipilimumab and nivolumab. This is the standard dose of 3 mg/kg of ipilimumab and 1 mg/kg of nivolumab.

What was interesting about the data is essentially 2 doses was about what you’d expect with 4 doses, meaning the majority of patients had no progression at 2 doses. In fact, the 12-week response rate was slightly less than 50%. But the total response rate, when we looked out further, was 57%, which is in line with what you’d see with when you intend to give 4 doses of ipilimumab and nivolumab.

Of those patients who may have had some evidence of tumor progression after 2 doses, none had a response. That was a minority of patients. That was about one-third of the patients who fell into that category. It suggests that continuing with ipilimumab-nivolumab when you’re not seeing benefit may not actually be the right thing to do. Although quite frankly, it doesn’t seem it was the wrong thing to do, because the toxicity rate was virtually the same you would expect if you were intending to give 4 doses.

What it ultimately shows us is that response is an early phenomenon with these drugs. So is toxicity, particularly in combination. Thinking about how we can optimize therapy, just thinking about a lesser number of doses, would be fine. It’s hard to say we should change our standard of care based on that. We certainly should feel OK if all we can do is give 2 doses because the patient is not tolerating it.

This is 1 of those types of questions that can be done only at an institution, and then ultimately maybe in a cooperative group to try to change the amount of dose that we give to be less than what’s standard. It certainly is something that may be considered in the future to do a larger trial, just comparing the 2 doses.

 The amazing thing is that 19 of 60 patients had more than 4% growth. Why they chose 4%, I don’t know. That’s a good question. That means any evidence of growth, as opposed to stability or regression, but none of them responded. Those were all the patients who are going to do badly. I was absolutely struck by that. The median survival is not reached. They haven’t even reached median PFS [progression-free survival], as I remember.

 It is pretty interesting, although I assume they met median PFS, but actually, that’s surprising if they had 20 months of follow-up, and they haven’t seen the median PFS.

 It’s 60 patients, so it’s hard to lean on even the PFS and the OS [overall survival], but certainly the response data are fascinating, as is the toxicity.

 Eighty percent of patients who were alive at either 18 or 24 months. That’s pretty impressive.

 Those investigators may be onto something. We’ll see. Maybe less is more.

 There also were the correlative data. They mentioned that they looked at T-cell activation and cell proliferation, and the parameters didn’t change after 2 doses. The increases they saw after 1 seemed like that was sufficient. That that data also support biologically what’s happening.

 They chose the 2 classic pharmacodynamic sine qua nons of getting PD-1 blockade alone or ipilimumab alone: Do the T cells proliferate, so Ki-67, and do they upregulate ICOS? I was blown away by the fact that 2 weeks later, it happened. Then you can keep going for as long as you want. Nothing changed. That’s a pretty early indication. Very interesting data.

I/O Dosing Schedules in Metastatic Melanoma

 Let’s talk about the data. Let’s turn to Su. There have been some relatively long-term follow-up studies of at least ipilimumab plus nivolumab-pembrolizumab, eventually nivolumab alone, and 

. The longest would be the dabrafenib-trametinib story. What can you tell us? Do you feel there are enough data to drive you in metastatic first-line disease in a 

-mutated patient to want to choose 1 or the other? Sorry to put you on the spot, but those are the data.

 Absolutely. That’s an excellent question. First, I have to point out that the options we have in front of us today have really blown the field wide open for melanoma, right? The response rate we see with anti–PD-1 alone is somewhere around 40%. The response rate we see with combination anti–PD-1 and CTLA4, is probably somewhere between 55% and 60%. There’s certainly a survival benefit, which I’m happy to get into. Then the response rates with targeted therapy are probably somewhere over 70%, potentially with less of a durable response.

It’s important to note that we have so many choices in front of us in terms of anti–PD-1 alone, combination, targeted therapy for those who have a 

 mutation. I do use a lot of I/O [immuno-oncology], even in the first-line setting. That is certainly a little a bias. The COMBI-d and COMBI-v trials have now 5 years of data. The 5-year overall survival is 34%, which is not insignificant by any means. Compared with that, for example, the CheckMate 067 data, where over 900 patients were randomized 1:1:1 to ipilimumab plus nivolumab, nivolumab alone, and ipilimumab alone with matched placebos. The response rate was almost 60%. I believe it was 58%, albeit the grade 3/4 adverse-event rate was also almost 60%. The overall survival at 5 years was 52%, meaning 1 of 2 people are still alive at 5 years.

The data are really exciting. I had alluded to the fact that I do use a lot of I/O in the first-line setting for the 

metastatic population. Part of it is because I have seen in my practice that if I use targeted therapy in the first-line setting—and once they become resistant, I barely have time to get them onto an I/O regimen, because their disease seems to just be accelerating—the performance status declines. There are just a lot of factors that prohibit me from then trying a second-line therapy. Now I know it’s doable. I’ve certainly been able to do it in some patients. That is a worry that I always have.

 I wonder, do your colleagues agree with that? Jason, do you agree that if someone fails frontline targeted therapy, then the immunotherapy becomes a difficult scenario?

 My bias is to say yes. I also think that patients who fail any therapy in the frontline are also harder to manage. I look for the high upside, and because of that, those ipilimumab-nivolumab data look quite impressive. It’s worth noting that we certainly have not figured out the optimal regimen in terms of immunotherapy in the front line, or even combinations with targeted therapy. All this is evolving.

In the last couple of years, we’ve had some discrepancy around the most appropriate dose of ipilimumab become an issue in the field, with CheckMate 511 showing potentially no difference in the benefit between nivolumab 1 mg/kg–ipilimumab 3 mg/kg and nivolumab 3 mg/kg–ipilimumab 1 mg/kg, but that’s in contrast to previous randomized studies. There are data for ipilimumab plus pembrolizumab with a lower dose of ipilimumab. All this is to say that we have a lot more to do. When I think about which treatment to use in the front line, I am somewhat biased toward immunotherapy as well, going into long-term overall survival with PD-1 around the low 40s and doublet I/O in the low 50s. I like those data. That is usually what drives my perspective.

I’ll also note that you have the BRAFinhibition in the back pocket, so if you don’t get the kind of long-term sustained benefit you want, you can get disease stability, which allows you to regroup and think about future options. All those are ways I think about it. I will note, just for the record, that the best patient with 

-mutant disease to treat with a BRAF inhibitor, though, is the patient with low-volume disease who is almost always going to be the person in the front line. On some level, we are putting ourselves behind the book when it comes to using targeted therapy if we do it that way. Probably everyone here is going to, generally speaking, agree with that sentiment.

Novel Therapies for Metastatic Melanoma

 Let’s move on from adjuvant to neoadjuvant. Let’s talk about systemic approaches. Back to you, Adil. It’s the same question that we think about in the adjuvant mode. How do you make decisions in choosing therapy, 

 versus immunotherapy, based on if it’s a 

 wild-type patient? What are the biomarkers that help you to make the choice?

 That’s a great question, and I loved the previous discussion that Jason Luke and Sunandana Chandra were having about adverse events. In terms of neoadjuvant, I lean toward immunotherapy. That has to do with some biases from trials in terms of thinking that immunotherapy will have more lasting benefit. I have to say, parenthetically, that we understand a lot more about immunotherapy response, and we understand the mechanism of immunotherapy response a lot better than we understand 

-mutated patients who will have a lasting response versus who won’t have a lasting response. There’s been some interesting biomarker work done in 

-mutated patients too. Today I would say that none of us can really predict, given the 

-mutated patient, who’s going to have a lasting benefit versus who’s going to have some kind of bypass pathway.

I feel like in the neoadjuvant space, unless somebody has a history of previous adverse events—colitis, multiple sclerosis, or some major issue—I lean toward using immunotherapy. Then the choice between using single-agent or double-agent immunotherapy comes down to what I think is likely to work, how bulky the tumor is, and how urgent the treatment seems to be. Most of our neoadjuvant patients today are not organ sparing. As Jason was mentioning earlier, you’re not really converting non-surgically resectable to resectable. You’re just basically trying to see what works better. Some of the ASCO [American Society of Clinical Oncology] trials are interesting in the sense that they are suggesting 2 cycles or a shorter duration of treatment is enough. Alex Huang’s trial also suggested that 1 single dose might be enough.

I treated this 1 patient—and I mention this because it just happened. I gave her 6 months of treatment with single-agent nivolumab. She had a complete response, pathologic complete response [CR]. Then we had a long discussion. I was like, “Yeah.” She was like, “It’s really important for me to have maximum survival.” I was like, “OK, we’ve given 6 months. Why don’t we give you 6 months more adjuvantly afterward?” In the 6 months, her sugar started rising, and then she developed type 1 diabetes. Now I’m just cringing inside, and I’m thinking maybe I should have stopped with the path CR and not done that extra few months of treatment. Maybe she wouldn’t have type 1 diabetes. I feel like 1 lesson for me from this year’s ASCO Annual Meeting is that maybe shorter duration is OK. I hope I don’t overlearn that lesson and have the opposite problem. Maybe the 1 advantage of neoadjuvant therapy is that we can get by with shorter, less intense treatment.

 Do you feel the same way about metastatic therapy? Meaning, do you tend to lean to immunotherapy first, rather than targeted therapy?

 I do. I just want to check my biases here. I feel like the data are maybe not fully supporting my position. I feel like immunotherapy offers longer-term success.

Systemic Therapy for Metastatic Melanoma

 Su, how do you talk to patients about the toxicity issues? Jason just started to get into that. Maybe you could continue. In other words, in front of you is the stage IIIA patient with some minimal amount of disease in the lymph nodes, and they don’t have ulceration in primary. I just saw such a patient yesterday as a new patient. It’s a tough discussion. How do you handle explaining the relative toxicities to them?

 First of all, I do go over the data somewhat with the patient. For example, if they have less than 1-mm focus of tumor in their lymph node, then based on the trials, they probably don’t even need therapy, right? That’s something worth considering if they truly are eligible for adjuvant therapy based on the trials we have.

It’s really a balance between the 2 paradigms of therapy, 1 being targeted, and 1 being immunotherapy. By balance, with targeted therapy, you can certainly see adverse effects. They can be pretty life-impactful in the sense that you could have pyrexia, for example, with a particular combination of BRAF and MEK inhibition. You can see transaminitis. There are certainly adverse effects that can be impactful for the patient, impactful for whether we can continue treatment. As opposed to immunotherapy, where there are certainly also adverse effects, these adverse effects may actually, in some cases, be permanent. They may lead to endocrinopathies requiring lifelong supplementation. They can sometimes be life-threatening adverse effects, albeit very rarely.

I just have a frank discussion with our patient population in terms of this is what we can expect. They both have rigorous data. We have overall survival benefit now that has been shown specifically with the combination approach. Together, we come to a decision. The nice thing is that they’re both very valuable and efficacious options. It’s just a matter of going through the nitty-gritty and trying to figure out what kind of toxicities are acceptable to the patient and their family, as well as us.

 Jeff, can I follow on that really quick, just because I’ve thought about this toxicity question? It’s important to note, and none of us has alluded to it yet, that escalation of cancer therapy has been associated with improved benefit in general. We’ll talk about the metastatic—the data where ipilimumab and nivolumab together really do appear to have the best long-term overall survival advantage. In the adjuvant setting, we already have a study now, the CheckMate 915 study, that actually does not seem to suggest that we have an added benefit by adding double I/O [immuno-oncology] therapy.

The reason I bring that up, to dovetail into what Su just said, was that the toxicity associated with ipilimumab and nivolumab is not insignificant in the metastatic setting, let alone in the adjuvant setting. That’s a really important consideration when you think about choosing adjuvant therapy. The toxicities that we would tolerate, potentially in the metastatic setting, are not the same toxicities that we would tolerate in the adjuvant setting. Being very aggressive up front to educate patients—to let us know so we can interrupt therapy, we can intervene very quickly, whether it’s I/O or targeted therapy—is truly an important part of the care of adjuvant patients.

 It’s going to be even more important. I was about to turn to you, Jason. I know you’re the PI [principal investigator] of the KEYNOTE-716 trial. It’s going to be even more important because now we’re treating stage IIB and IIC patients. If that’s a positive study, which I assume it will be, you’d better bet that there will be a lot of stage IIA patients who get treated in the community. Your point about having a different threshold for stopping and a different threshold for choosing based on toxicity will become even more important because if those studies, the KEYNOTE-716 and the BMS [Bristol-Myers Squibb Co] CheckMate76K trials, I’m assuming they’ll be positive. Assuming they are, that opens a lot of patients to adjuvant therapy. It’s like, 20,000 patients in addition a year.

 The number of patients with stage II melanoma who would be amenable if the approval comes through would be approximately the same as the stage III population. We would essentially be doubling the number of adjuvant patients. But again, as we talked about in the beginning, the risk of those patients for relapse is actually as high as for the ones we already treat. It makes sense to potentially extend that to those patients, but yes, it will be many more. Being cognizant of that toxicity is important.

Adjuvant Therapy for Melanoma: Safety and Efficacy

 Let’s take another turn, and let’s go back to Jason. Jason, let’s talk about treating our patients with the higher risk—the stage IIIB, the IIIC, and the IIID by the AJCC [American Joint Committee on Cancer]. Those are really the bad actors, because frankly those with stage IIIa, by the current AJCC ratings, do pretty well. They have only a 20% risk of death at 10 years. It’s those with stage IIIB, IIIC, and IIID who do badly. How do you decide I/O [immuno-oncology] versus 

-adjuvant therapy? What’s the deal? Even add IIIA if you want.

 It’s an important thing to think through. You mentioned earlier, do you test for 

 in neoadjuvant? It becomes the same question. Do you test for 

 in adjuvant? When you think about trying to make these decisions, you have to have that data to make such a decision. It’s very common that we get the patients referred to us through the usual referral channels, and they haven’t had the test. That actually often limits our ability to make that discrimination between therapies, and sometimes you need to delay that conversation.

I would emphasize that for anybody viewing, because that’s a really important need, that test. That should almost be reflexively routed by the pathology department at this point. If you have that data, then you can set up this question about which therapy you want to choose. It comes down to patients’ perspectives as well as doctors’ perspectives. Ryan Sullivan, MD, was just alluding to some people who use 

, some people use I/O. The data, as you just mentioned, look pretty similar going out. There are some caveats that I would just note. One of them, around the 

-mutated patients, is that 1 slide from KEYNOTE-054 was interesting to look at the disproportionate improvement in the hazard for relapse in the 

 population, specifically relative to the total. Pembrolizumab in the adjuvant setting worked better in 

-mutant patients than it did in the total population.

I don’t want to exaggerate that more than anything else, except to say that it definitely works in 

-mutant patients as well. You can see that in some other I/O trials also. In terms of which stages to treat, for all of us here, there would be no question that we would try to get a patient who’s stage IIIC or IIID melanoma to take adjuvant therapy. In stage IIIB patients, most of us probably would say that. The stage IIIA population, as you alluded to, that’s the tough one. Their absolute benefit from adjuvant therapy, while it could be improved by 50%, ends up being single-digit improvement and absolute improvement.

When you think about the toxicity spectrum of these drugs, that’s a consideration, because there are long-term irreversible endocrinopathies that develop on I/O therapies—the thyroid, diabetes, etc—and I have difficulty with that population, stage IIIA especially, for using immunotherapy in the adjuvant setting. In those patients, I tend to defer, which I think you were alluding to in general. If they want adjuvant therapy, that’s where I would definitely tend to go for BRAF inhibition.

The data from COMBI-AD really highly support that. The data look really good in the stage IIIA setting as well. In that consideration point, you have to put it in front of the patient and think about what’s meaningful to them: coming for an infusion every 4 to 6 weeks or taking the targeted therapies each day?

I’d also summarize by noting that these—I won’t call it extreme benefits that we’ve seen with these hazards for improvement in the stage III setting—are likely going to be extended into other disease settings as well over the relatively near term. Due to idiosyncrasies in our staging system, all of you are aware that stage IIB and IIC melanoma is, in fact, higher-risk than stage IIIA. That really goes to the biology of the disease we treat, which is that the primary depth of the melanoma is actually more important than the involvement of single lymph nodes.

Because of that, we have this oddity where we have the on-label ability to treat that stage IIIa population that I just said I wouldn’t usually treat with immunotherapy, yet we do not have on-label usage of the drugs for a higher-risk population of stage II with deep primaries. Two trials are being investigated in that space, the KEYNOTE-716 study with pembrolizumab and the CheckMate76K study with nivolumab. The former study, KEYNOTE-716, is likely to read out much earlier. We’re really hopeful that we’ll be able to extend this benefit of adjuvant therapy into earlier stages of melanoma.

When you marry that to our previous discussion around neoadjuvant, we’re starting to get to a place where medical management of melanoma over the next 5 to 10 years may come forward as the primary modality for managing patients. That’s really interesting to think about. It would have been heretical to say that 10 years ago, but these data really do support that, as the field has brought all these data forward, moving into different disease settings. It’s really exciting.

Adjuvant Therapy for Melanoma: Ongoing Studies of Interest

 Speaking of adjuvant, let’s go back to Ryan. I think at ASCO [the American Society of Clinical Oncology annual meeting], there were 2 presentations, both of which were updates. They were not new trials. One was COMBI-AD, and that was excellent. The other was the readout of Lex [Alexander] Eggermont, [MD, PhD,] for KEYNOTE-054. What can you tell us about updated data? Does that change your mind about using one versus the other at all?

 I think the way to set the context is for so long, we had interferon, and there were people who would use interferon. There were people who wouldn’t use interferon. They would draw sabers at conferences and fight each other about this. Then thankfully, we had more effective therapy in the metastatic setting that was able to be brought forward into the adjuvant setting.

You know full well, Jeff, that the CheckMate 238 trial, which is nivolumab versus ipilimumab, but even before that, there was the EORTC study comparing ipilimumab versus placebo. That was a positive trial with regard to relapse-free survival [RFS] and overall survival, really resetting the template, if you will, for adjuvant therapy. Interestingly, what was presented last year, ipilimumab versus interferon, is not a slam dunk for ipilimumab. We would have anticipated that ipilimumab would have been way better than interferon. Interestingly, there were 2 doses of ipilimumab looked at, 3 mg/kg, 10 mg/kg. If you had to choose one of those from a toxicity standpoint, you would choose 3 mg/kg because it’s much less toxic. Remarkably, against interferon, the difference with 3 mg/kg versus interferon was better than the difference with 10 mg/kg, suggesting that the toxicity may actually be influencing the ultimate benefit because of presumably the immunosuppression utilized to counteract that toxicity.

Then bringing forward the more contemporary, even though that was probably the most recently first presented adjuvant trial, the CheckMate 238, that is nivolumab versus ipilimumab, showing clear improvement in relapse-free survival. I believe you presented the 3-year data last year showing about a 60% relapse-free survival for nivolumab.

This year, Lex presented the KEYNOTE-054 3-year update, again showing about a 60%, I think the number was 58%, 3-year relapse-free survival for pembrolizumab, comparing that to placebo, which was in the mid-40% range. I’m not sure those data help us to distinguish between those 2 drugs, but it’s comforting that the relapse-free survival is pretty solidly above 50% at 3 years.

The COMBI-AD study, which is now for something completely different, BRAF-MEK inhibitor therapy versus double placebo, so dabrafenib/trametinib versus double placebo. I think the most remarkable part about these data is they continue to show, really for the first time in an adjuvant study with a molecularly targeted therapy, continued improvement in relapse-free survival of the treatment arm versus the placebo arm. Most of those studies look like this, where there’s an initial benefit and then it comes back together….

Suggesting that you haven’t actually cured patients, you’ve just delayed when their disease is going to come back. The data with COMBI-AD, 5 years down the road now, suggest that they’re probably curing patients with BRAF-targeted therapy, or if you’re not, you’re delaying those relapses by 5 or 10 years. Interestingly, the 3-year data look almost identical to the nivolumab and pembrolizumab data, 59% relapse-free survival, the 5-year data, 52%. There’s a little bit of decay in the relapse-free survival curve. It’s not a complete plateau, and we’ll have to see in the future whether we see complete plateaus with nivolumab and pembrolizumab in those respective studies. I think the bottom line is if you were somebody who would think about targeted therapy for patients with stage III melanoma, you’re still going to offer that. If you were somebody who wanted to go up-front PD-1 inhibitor therapy for patients with a 

 mutation, you’re still going to offer that. The data from KEYNOTE-054 and COMBI-AD reinforce what you’re previously thinking of doing. Then if patients are 

 wild type, the data from KEYNOTE-054 and most recently last year, CheckMate 238, again support the use of nivolumab or pembrolizumab in patients with high-risk stage III disease.

 Yes, certainly it looks like the KEYNOTE-054 and the CheckMate 238 curves are very close at 3 years. I was absolutely fascinated by the fact that COMBI-AD looks like it’s coming to a plateau. The urban legend was that the targeted therapies would have an early high RFS, and then it would dip down below immunotherapy, but that hasn’t really happened. Actually, they have the longest follow-up of any of the current adjuvant trials. I was actually quite impressed that the patients on COMBI-AD who got treated with dabrafenib/trametinib did very well in the long term.

Adjuvant Data Explored for Locally Advanced Melanoma

 Jason, there are other neoadjuvant trials out there. As I said, there’s been a flood of interest in this. Can you tell us about any of the other trials? I think there’s been one with pembrolizumab alone.

 There are some very interesting results. Going to that question about the pace of the responses, when you start a BRAF inhibitor, you get rapid responses. We know that. The study that you are alluding to by Alex Huang, [MD,] et al, from University of Pennsylvania, it’s very interesting that they gave a single dose of pembrolizumab prior to surgery. They got roughly similar results to giving 3 or 4 doses of nivolumab. A single dose of PD-1 can generate a substantial treatment response. I think that’s really interesting as we think about kinetics of response and a long-term benefit that we can see in some of these patients. As we start thinking about the feasibility of combination regimens, especially, we don’t necessarily need to think about long durations of treatment prior to neoadjuvant approaches. I think that’s exciting.

I also want to note quickly, there are some other intratumoral injection studies that are also starting to come forward. We’ve seen some data for TVEC [talimogene laherparepvec] that look interesting. I’m not sure I would do that, but there are combinations of PD-1 antibodies with other sorts of innate molecules that we’re going to talk about a little bit later that do look quite promising, and again, I think do have the potential to come forward as a strategy in the future.

 OK, I like the TVEC idea. But again, that’s a tough one. What about BRAF-MEK? One would think with the excellent response rate that BRAF-MEK would be the perfect neoadjuvant strategy. What can you tell us about the data? I think there was a presentation probably last year. I don’t think it was updated this year. I believe it was summarized in Alex Menzies, [MD’s,] review that he presented last year.

 Stephanie Blankenstein, [MD,] reported some data. It was a small trial. It was only about 25 patients or so. These patients had either stage III resectable or they could have had stage IV disease, but they had to have somewhere up to 3 lesions that were resectable. The patients underwent 8 weeks of dabrafenib plus trametinib. Out of the 25 patients, 18 patients went through the neoadjuvant regimen, and 17 of them underwent resection. Of the 17 patients, 16 patients had complete R0 resection, which I think is remarkable.

Of course, we don’t have a lot of long-term outcomes yet with these trials. It certainly poses a question of therapy options that we didn’t have before. Then again, it leads to more questions of, for a 

-mutated patient, for example, should we think of neoadjuvant immunotherapy versus targeted therapy? I think those questions are still unanswered.

 I would say in Menzies’ article, though, it looked as if the immunotherapy neoadjuvant patients had a better relapse-free survival rate than the BRAF-MEK patients, but again not enough long-term data. Adil, weren’t there more long-term data in that [Elisa] Rozeman, [MD,] presentation at the American Society of Clinical Oncology [annual meeting]? This was the late update on OpACIN-neo, the very latest data. What can you tell us about that?

 I think it’s very interesting. The one editorial, general comment I’d make is that I’m having a hard time disentangling partial responses [PRs], near-complete pathologic CRs [complete responses], pathologic complete CRs, pathologic partial responses, and comparing them…. I feel like all 5 of us could probably look at a scan and say, “OK, this is a CR or a PR.” I have a hard time understanding if in Holland, or in Australia, or at UCSF [University of California, San Francisco], pathologists are going to look at it and come to the same conclusion. I think in the Rozeman article, what you can see is that if you had a pathologic CR, you’re probably going to have a great longer-term outcome. But the one thing to remember with these trials is that there are very few patients in each arm, and I think Su was talking a few minutes ago about the OpACIN trial. It’s interesting that the AEs [adverse events] are seen in the 20% range for different ipilimumab-nivolumab combinations. I simply think that when you’re having 20, 30 patients in each arm, you’re probably not capturing the full range of AEs here as you are, say, in CheckMate 067, where every time patients come in, you’re checking all kinds of laboratory tests, you’re following them for 5 years afterward. With hormone adrenal insufficiency and stuff like that, are you really capturing them in the first few weeks? I apologize for this inconclusive, long-winded answer, which doesn’t help us necessarily.

I think that the pathologic CR issue and path response issue is, it can be gamed in many different ways, as opposed to RECIST [Response Evaluation Criteria in Solid Tumors], which you and I and everybody else can agree on means something. Actually, in one of the trials, I want to say it was the Blankenstein trial, it was interesting that they had a comparison of pathologic CRs and radiologic CRs. It was about 20% lower for the radiologic CRs compared to the pathologic CRs, illustrating that maybe we should minus a certain percentage from the pathologic CRs.

 That seems to be a common scenario, both in my experience in squamous, basal melanoma, and cutaneous cancers. When you treat them up front with immunotherapy, sometimes you’ll get a chance to resect them fully. I think that the radiologic response lags way behind the pathologic response. The nice thing for me about the Rozeman presentation was that even, I think now with 2 and a half, 3 years, of follow-up, if you had a pathologic CR, you stayed in remission something like 88% or 90% of the time. That’s pretty impressive.

Now, you might argue that getting a pathologic CR simply picks out the people who, with adjuvant therapy, would have done well anyway, and the ones who are going to relapse are the ones who didn’t have pathologic CR or a pathologic high-grade PR. But again, only time is going to tell. There is a BMS [Bristol-Myers Squibb] trial, and there’s an ECOG [ECOG-ACRIN Cancer Research Group] trial. In those randomized trials, we’re going to find out, I hope, about the utility of neoadjuvant versus adjuvant therapy.

Neoadjuvant Data Explored for Locally Advanced Melanoma

 Speaking of practice changing, let’s get back to the neoadjuvant sphere, where there could be some things that will change practice. Do you utilize an assessment, like a pathologic complete response, to determine how you will subsequently treat an adjuvant patient? Let’s say they get neoadjuvant off-protocol or on-protocol ipilimumab-nivolumab, then they have a pCR [pathologic complete response]. What does that tell you? How does that impact your treatment?

 That’s an excellent question. One of the things we’re trying to answer in our field of melanoma is if pathologic complete responses or near-complete responses will actually translate to survival benefit, just like in our breast cancer population? There’s been some unique retrospective and prospective studies that are certainly intriguing and I think require further study in this patient population. For example, Dr Alex Menzies last year discussed a retrospective pooled analysis. I believe it was 6 neoadjuvant trials—well, it was a combination of neoadjuvant and adjuvant. It was a very heterogeneous patient population. The patients were treated with targeted therapy as well as immunotherapy. Even in the immunotherapy cohort, some were treated with anti–PD-1 alone, some with anti–PD-1 and CTLA4 alone. Looking at these in a pooled analysis is challenging to draw conclusions from; however, I think we learned some important things from the retrospective data. I believe the pathologic response rate was about 41%, in terms of complete responses. We learned that with targeted therapies, you can get rapid responses; however, with the combination of ipilimumab-nivolumab, the response rates in terms of pathologic complete responses seem to be higher. The data were retrospective, so we need to keep that in mind.

More recently, this year, Dr [Christian] Blank discussed the OpACIN-neo trial. In that trial, it was around 114 patients. They were studied with 3 dosing schemas of anti-CTLA4 and anti–PD-1, settling on ipilimumab 1 mg/kg plus nivolumab 3 mg/kg for 2 cycles, and that showed a pretty favorable pathologic response rate of about 77%. Interestingly, the grade 3/4 AEs [adverse events] were pretty tolerable, at only about 20%.

Some big questions are about the number of cycles needed. We don’t know, but it certainly is quite intriguing. What I found the most interesting about the trial is that the investigators looked at whether therapeutic lymph node dissections could be omitted for some patients. Patients with a complete pathologic response or a near-pathologic response actually were omitted from a further therapeutic lymph node dissection. I believe 70 of 99 patients achieved a pathologic response in what was noted as their indexed lymph node. Once there was a pathologic response in the indexed lymph node that was significant, those patients did not have to go further—

 Right. That was the PRADO trial. That was Christian’s oral presentation. I thought that was pretty interesting.

 Correct. I think the surgery-related adverse effects were lower in that group. There’s a lot to be further understood about which patients benefit from neoadjuvant trials and which patients actually could potentially skip surgery.

pCR and Neoadjuvant Therapy for Melanoma

 Along the same lines, let me turn to Adil, to go a little further with the issue of molecular testing. How important is 

 testing that it helps you guide adjuvant or neoadjuvant therapy? Does it play a role? It sounds as if, for Jason, it doesn’t play a role for neoadjuvant. Will it play role for you when you try to decide about adjuvant therapy for patients?

 It could potentially be an important tool for patients where you’re maybe worried about autoimmune adverse effects, or if you have an extremely bulky tumor—I just saw this patient last week, the tumor is so large that I have a hard time, and he’s already having nerve compression symptoms in his arm and isn’t grip strength. That may be 1, and actually, I did order 

 testing on him. I’m going to see him tomorrow, so I’m considering the possibility of starting him on targeted therapy, but I rarely do that. For me, the 

 testing is something to do in case your immunotherapy doesn’t work and you’re looking for a plan B. For us, we have this home-grown T-cell assay that we use, and we’ve published on it before. Basically, it measures exhausted T cells in the tumor microenvironment, and it’s run by Mike Rosenblum at UCSF Helen Diller Family Comprehensive Cancer Center. That’s the test. Although I’m a big fan of 

 testing, if you’re trying to make decisions on an individual patient, it’s going to be hard to really slot a melanoma patient based on 

 Su, do you ever look at any other biomarkers? Do you ever assess tumor mutational burden or anything else on a routine basis, or is it all just a research tool?

 For our resectable patients, our priority at my institution is to try to get them on trial. We do have some adjuvant as well as some neoadjuvant trials that we try to prioritize for these patients. In terms of routine testing, besides what’s been mentioned, looking at tumor mutation burden, of course, is a certainly interesting research question. We often have that information in front of us as we’ve ordered next-generation sequencing, etc, but it doesn’t really inform my decision-making at that time.

I do think that, as more is learned about these neoadjuvant as well as adjuvant approaches, we’re probably going to have to fine-tune them a little more to be able to identify the patients who would benefit more compared with others, and that’s when these parameters will be probably more looked at and maybe acted upon a real-time basis that would actually be practice changing for our patients.

Molecular Testing in Locally Advanced Melanoma

 Let’s turn to Jason. There’s been a lot of attention given to neoadjuvant approaches with immunotherapy in not just melanoma but also other cancers. Whom do you consider for neoadjuvant immunotherapy, and how often do you see these patients?

 That’s a good question, and I appreciate everything that everybody else has mentioned so far. I think there’s neoadjuvant, and there’s neoadjuvant. What I mean by that is, in the old days, we had patients who had disease who were probably not really resectable. We had a hope that by giving some intervention initially, we can shrink it enough where we can do surgery. Those patients are not that common, but they do happen, and all of us have seen them in clinical practice. That actually contrasts somewhat with the whole concept of neoadjuvant as we’ve been discussing in the field over the last couple of years. We’re just trying to identify patients who actually are resectable, but where we could intervene in a therapeutic approach, either for research consideration or potential therapeutic implications prior to the surgery. Those are 2 different things.

The extent to which these are common patients is suspect. At our center, because we have had an interest in this for quite an extended period of time, we see a fair number of these. We have had institutional investigator-initiated trials where we’ve put on 20 to 30 patients per year. But again, recognize that there’s a long-standing practice of all patients with cutaneous malignancies being referred to UPMC Hillman Cancer Center as the hub in western Pennsylvania, owing to the work of John Kirkwood for decades now.

In clinical practice in general, this isn’t a particularly common presentation. I will note, however, that the results initially—and others will to speak to them—are quite impressive for intervening. That being said, it’s still a research question, and we really guide all the patients toward research protocols. We would necessarily take a patient through neoadjuvant therapy if they were resectable, if they weren’t going to go on a protocol. That’s an important differentiator. For community practitioners, they should be cognizant that there actually is a cooperative group trial now that’s a SWOG clinical trial, trying to look at the impact of neoadjuvant therapy relative to adjuvant. I strongly recommend people consider that if they actually do see some of these patients.

 When you see these folks, do you test them for BRAF or PD-L1? Does it matter to you?

. For BRAF, in the setting where they have palpable disease, our expectation is that their risk is going to be quite high for recurrence, and it seems very reasonable to get that data up front. That data don’t always inform what we do. Most of the trials we’ve been pursuing in the neoadjuvant space are immunotherapy trials. We sometimes get the PD-L1, but that’s more in the context of research, not to guide standard practice.

Identifying Patients for Neoadjuvant Melanoma Therapy

® titled “Advances in the Treatment of Malignant Melanoma.” I’m Dr Jeffrey Weber. I’m from the Perlmutter Cancer Center at NYU Langone Health in New York, New York. Joining me today in this virtual discussion are my colleagues Dr Sunandana Chandra, who’s from the Robert H. Lurie Comprehensive Cancer Center at Northwestern University in Chicago, Illinois; Dr Adil Daud, who’s from the UCSF Helen Diller Family Comprehensive Cancer Center in San Francisco, California; Dr Jason Luke, who’s from the UPMC Hillman Cancer Center in Pittsburgh, Pennsylvania; and Dr Ryan Sullivan, who’s from Massachusetts General Hospital Cancer Center in Boston, Massachusetts. Welcome.

Today we’re going to review the latest data on melanoma released from the 2020 ASCO [American Society of Clinical Oncology] Annual Meeting. Let’s get started on our first topic.

The first topic talks about perioperative approaches to melanoma. A question that we medical oncologists deal with is: What’s the optimal approach to the multidisciplinary management of locally advanced melanoma? Ryan, tell me, how do folks at your institution handle the multidisciplinary management of the disease?

 Thanks, Jeff, for the introduction. It’s great to be here with such wonderful colleagues. One of the key parts of managing patients with early-stage high-risk melanoma is the multidisciplinary approach. I think there are 3 critical team members in terms of those who are patient facing: the dermatologist, the surgical oncologist, and the medical oncologist. Then there are a number of other critical people as well, including the dermatopathologists, radiologists, and of course all the support teams that help our patients.

What we tend to do at Mass General is a 2-step approach. The first step is the dermatology, cervical oncology clinic to meet patients who are maybe diagnosed with melanoma. Then we have a review at a multidisciplinary tumor board, and then typically we have a surgical oncology, medical oncology appointment that’s jointly visited. It’s a little spin on the traditional multidisciplinary clinic, where all the patient-facing doctors are seeing patients, but that’s typically how we do it.

Then we discuss the great majority of our cases, certainly any case where there’s a question of what’s the best management, and move forward. I’ll also mention that in the time of this COVID-19 [coronavirus disease 2019] crisis, we’ve found that these have been immensely helpful, because when we’re maybe making a slight change to what would have been the standard of care in the setting where we have open ORs [odds ratios], if there are limitations to that, it’s allowed us to apply novel approaches that have been certainly supported in literature to move forward with something that may not be the standard of care by NCCN [National Comprehensive Cancer Network] Guidelines.

 In your discussions with your surgical colleagues, how have those MSLT-II data—which frankly are now a couple of years old—they changed the extent and the role of the completion lymphadenectomy at your institution?

 For a few years before that, we were often discussing—for a patient who had a positive sentinel lymph node—whether the value of completing the lymph node dissection was appropriate for individual patients. A few years before the MSLT-II data came out, there were data from a randomized trial in Germany, which randomized patients with a positive sentinel lymph node to either completion lymph node dissection or monitoring with serial ultrasounds and exams, which found that there was no significant overall survival difference or distant metastasis presurvival among patients randomized to observation versus surgical complete lymph node dissection. What was important about that, though, is that there was a lot of trouble with that study in terms of powering. From a statistical standpoint, it did prove the principle that there’s probably some patients who we don’t necessarily need to do the completion lymph node dissection.

A year or 2 later, when the data came out almost overnight, it was data that was just directly published in the

 that had never been presented before. I remember the week before in tumor board, we were talking about doing completion of lymph node dissections, and the week after, we were talking about never doing another one. There are some scenarios where a completion lymph node dissection may be considered. Certainly, if there’s a palpable node that wasn’t part of the MSLT-II, those patients deserve a complete lymph node dissection, but in patients with positive sentinel nodes, it is the rare exception when we, in concert with our surgeons, recommend a completion lymph node—

 Let’s ask some of our other colleagues. I assume they have the same philosophy. How do you folks monitor the lymph node basins when you haven’t done a completion lymphadenectomy? Adil, what is done at UCSF?

 We use ultrasounds to monitor the lymph node basin. As Ryan mentioned, when that DeCOG-SLT study came out—and that was, I want to say, ASCO 2015—we essentially stopped doing completion lymph node dissections. I’ve always thought it’s a staging, but sentinel lymph nodes are a staging tool, not a disease-control tool. I’ve never really thought that any of these studies would show a benefit, unless you’d have to think that you’re doing disease control from completion lymph node dissection. There’s been really no evidence from the previous WHO [World Health Organization] trials a long time ago, which maybe many of you don’t remember, but it was 20 years ago or so. To me, the whole sentinel lymph node biopsy, completion dissection, is really more of a diagnostic issue. We basically monitor people’s lymph node basins every 6 months with ultrasounds if you are not doing scans on those patients.

Optimal Management of Locally Advanced Melanoma

Jason J. Luke, MD, FACP, associate professor of medicine, Division of Hematology/Oncology, and director of the Cancer Immunotherapeutics Center within the Immunology and Immunotherapy Program, University of Pittsburgh Medical Center Hillman Cancer Center, discusses predictive biomarkers in metastatic melanoma.

 mutations and PD-L1 status have helped guide therapy decisions in patients with metastatic melanoma, explains Luke.

Beyond these biomarkers, clinical factors can also inform prognosis, says Luke. For example, patients with low disease burden prior to treatment tend to respond best to therapy.

Tumor mutation burden and interferon levels associated with gene expression are also predictive biomarkers of response, explains Luke.

Investigational biomarkers including LAG3, myeloid cell gene expression, and adenosine signaling may further aid in stratifying which patients are likely to respond to treatment, concludes Luke.

Jason J. Luke, MD, FACP, discusses predictive biomarkers in metastatic melanoma.

Dr. Luke on Predictive Biomarkers in Metastatic Melanoma

Jason J. Luke, MD, FACP, associate professor of medicine, Division of Hematology/Oncology, and director of the Cancer Immunotherapeutics Center within the Immunology and Immunotherapy Program, University of Pittsburgh Medical Center Hillman Cancer Center, discusses current and emerging treatment approaches in metastatic melanoma.

Five-year analyses with treatment regimens such as dabrafenib (Tafinlar) and trametinib (Mekinist), pembrolizumab (Keytruda), and nivolumab (Opdivo) and ipilimumab (Yervoy) are showing survival rates ranging from 40% to 55%, says Luke. 

These data have further solidified the use of these agents in the metastatic setting and have provided a rationale for which to evaluate additional combinations that may possess the potential to augment responses. For example, in an ongoing phase III study, investigators are evaluating the addition of the oncolytic virus, talimogene laherparepvec (T-VEC; Imlygic), to pembrolizumab. Additionally, several studies are evaluating PD-1-based combination therapies, concludes Luke.

Jason J. Luke, MD, FACP, discusses current and emerging treatment paradigms in metastatic melanoma.

Dr. Luke on Current and Emerging Treatment Approaches in Metastatic Melanoma

At the 2015 annual meeting for ASCO we witnessed an extraordinary explosion into the general oncology consciousness of the Programmed Death-1: Programmed Death Ligand-1 (PD-1:PD-L1) axis and drugs that target that interaction. These drugs have changed the field of melanoma over the past few years, but at ASCO the presentations highlighted advances in non-small cell lung cancer (NSCLC), urothelial and renal cancers, mismatch repair (MMR) colon cancer and hepatocellular cancer, head and neck cancers as well as triple negative breast cancer. In fact, clinical trial results were presented that will immediately change the standard of care in NSCLC with the anti-PD-1 antibody nivolumab

 and possibly in melanoma with the combination of nivolumab with the anti-cytotoxic T-lymphocyte-associated protein 4 (CTLA4) antibody ipilimumab.

 Beyond that however, pembrolizumab (anti-PD-1), atezolizumab (formerly MPDL3280a, anti-PD-L1), and durvalumab (formerly MEDI-4736, anti-PD-L1) with and without tremelimumab (anti-CTLA4) all demonstrated groundbreaking results in various tumors. Each of these agents is on a rapid track to approval by the Food and Drug Administration.

There is tremendous optimism about PD-1:PD-L1 immunotherapy and perhaps we will eventually see a time when this approach is used in nearly all cancers. That being said, it is important to note that the response rate to these drugs as single agents remains in the 20% range in most tumors and that there is significant heterogeneity of response. A great deal of work remains to develop biomarkers of treatment benefit and to learn to apply these to the clinical management of our patients.

Where is the field at currently in the development of biomarkers of activity? Much further than you would likely expect. An obvious biomarker for PD-1:PD-L1 benefit has been staining by immunohistochemistry (IHC) for PD-L1 in the tumor, infiltrating immune cells, and/or stroma/vasculature. There are a number of problems with PD-L1 IHC, however. As you can tell from the previous statement about PD-L1, there is disagreement about the most appropriate cell type to test, what the cut off for positivity should be (1%, 5%, 50%?), when the testing needs to be done (fresh tissue at the time of each line of therapy?) and several other questions. Probably most importantly however, PD-L1 testing is inadequate because approximately 10% of patients of all cancer types who have had responses to PD-1:PD-L1 antibodies have stained negative by IHC. There are various reasons for this but pointedly, if we were to select patients by PD-L1 IHC alone, we would effectively deny at least 10% of the cancer patient population access to a drug that might have otherwise led to durable disease control.

There are other biomarkers that have been recently advanced as well. One of these is the number of mutations in the tumor being treated. Influential studies have suggested that higher mutational load corresponds with response to ipilimumab in melanoma

 Similarly, the approximately 60% response rate in MMR-deficient colon cancer to pembrolizumab

 (0% in MMR-proficient colon cancer) advances this logical hypothesis that a higher number of mutations in a cancer will lead to an increased frequency of neoantigen (non-self molecules that can be seen by the immune system) and therefore a higher likelihood of an anti-tumor immune response. While appealing as a hypothesis, this correlation is not perfect. Melanoma and NSCLC commonly do have high mutational load and high response to PD-1:PD-L1, however, renal cancer also has high response to PD-1:PD-L1 but demonstrates much lower mutational load.

Another approach to biomarker development is to more specifically examine the tumor microenvironment. Elegant translational work has demonstrated that those patients who are most likely to respond to PD-1 treatment are those whose tumors have the presence of both tumor infiltrating lymphocytes and PD-L1 at the invasive tumor margin.

 Further, work over several years has suggested that a productive anti-tumor immune response is driven broadly by the cytokine interferon-γ

 and multiple abstracts at ASCO this year demonstrated that the presence of an interferon-γ gene expression signature correlates with response and progression-free survival.

 Perhaps more importantly however, the lack of the interferon-γ gene signature correlated almost exactly with those patients who had no benefit to PD-1 immunotherapy. Tumors that demonstrate an interferon-γ gene expression signature have been previously described as “T-cell inflamed” tumors and are contrasted “non-T-cell inflamed” tumors which lack the signature.

Understanding this model is important as it sets up rational drug development strategies for the future. For example, in tumors demonstrating a high interferon-γ gene expression profile, combining PD-1 antibodies with other interferon-γ inducible gene targets such as indoleamine 2,3-dioxygenase (IDO) is likely to be highly efficacious. Such an approach is unlikely to be useful in non-T-cell inflamed tumors, however. Alternatively, understanding the molecular mechanisms that drive the lack of an immune response in non-T cell inflamed tumors will be essential to open up the tumor microenvironment and facilitate the activity of PD-1:PD-L1 antibodies. The WNT/β-catenin pathway was the first of these pathways to be discovered where in 48% of non-T-cell inflamed melanoma, tumor-intrinsic molecular signaling through β-catenin leads to immune exclusion and lack of activity of immunotherapy.

 Upon application of a small mole-â€¨cule to block β-catenin signaling however, the activity of anti-CTLA4 and anti-PD-1 antibodies can be restored. For both T-cell inflamed and non-T-cell inflamed tumors, a drug development future of “precision immunotherapy” is becoming apparent in which understanding the patient’s antecedent immune response will be essential to pick the appropriate combination of targeted-/chemo-/radiation therapy to combine with immunotherapy.

To summarize, ASCO this year was the coming out party for anti-PD-1 immunotherapy broadly in cancer. Amazingly, we’ve already learned a great deal about the limitations of this approach and are positioned to move forward very quickly to improve and someday optimize PD-1 approaches. To that effect, further progress will only be made if patients continue to participate in novel early phase clinical trials that push the field forward. Though PD-1 therapy as a standalone has been ground breaking, it still benefits only a small proportion of patients. So here’s to a future in which every patient has the chance for a durable long-term response to immunotherapy!

		Paz-Ares L, Horn L, Borghaei H, et al. Phase III, randomized trial (CheckMate 057) of nivolumab (NIVO) versus docetaxel (DOC) in advanced non-squamous cell (non-SQ) non-small cell lung cancer (NSCLC). Presented: 2015 ASCO Annual Meeting.

		Wolchok JD, Chiarion-Sileni V, Gonzalez R, et al. Efficacy and safety results from a phase III trial of nivolumab (NIVO) alone or combined with ipilimumab (IPI) versus IPI alone in treatment-naive patients (pts) with advanced melanoma (MEL) (CheckMate 067). Presented: 2015 ASCO Annual Meeting. 

		â€¨Snyder A, Makarov V, Merghoub T, et al. Genetic basis for clinical response to CTLA-4 blockade in melanoma. 

		â€¨Rizvi NA, Hellmann MD, Snyder A, et al. Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer. 

		â€¨Dung T. Le, Jennifer N. Uram, Hao Wang, et al. PD-1 blockade in tumors with mismatch repair deficiency. Presented: 2015 ASCO Annual Meeting.

		Tumeh PC, Harview CL, Yearley JH, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. 

		â€¨Diamond MS, Kinder M, Matsushita H, et al. Type I interferon is selectively required by dendritic cells for immune rejection of tumors.

		Ribas A, Robert C, Hodi SF, et al. Association of response to programmed death receptor 1 (PD-1) blockade with pembrolizumab (MK-3475) with an interferon-inflammatory immune gene signature. Presented: 2015 ASCO Annual Meeting. 

		â€¨Seiwert TY, Burtness B, Weiss J, et al. Inflamed-phenotype gene expression signatures to predict benefit from the anti-PD-1 antibody pembrolizumab in PD-L1+ head and neck cancer patients. Presented: 2015 ASCO Annual Meeting. 

		â€¨Spranger S, Gajewski T. Rational combinations of immunotherapeutics that target discrete pathways. 

		â€¨Spranger S, Bao R, Gajewski TF. Melanoma-intrinsic β-catenin signalling prevents anti-tumour immunity. 

. 2015. doi: 10.1038/nature14404. [Epub ahead of print]

Articles

At the 2015 ASCO Annual Meeting, we witnessed an extraordinary explosion into the general oncology consciousness of the Programmed Death-1: Programmed Death Ligand-1 axis and drugs that target that interaction. 

Jason J. Luke, MD, FACP, UPMC Hillman Cancer Center