 Locally advanced disease has a spectrum of possibility. That could include disease that is not resectable but is locally advanced. For example, patients with lots of intrasametastases are nonresectable, and they have their own separate treatment algorithm. For patients who are resectable, we can get them to a status that we call “no evidence of disease” and treat them adjuvantly. Clearly, there are 2 major pathways for treatment.

Around 45% of cutaneous melanomas contain the 

 mutation, and there are at least 3 targeted combination therapies available to give them adjuvantly. Alternatively, there is immunotherapy. Immunotherapy is mainly anti–PD-1–based immunotherapy. Although there is anti-CTLA4, which is approved based on data from Europe. Because the head-to-head analysis of nivolumab vs ipilimumab in the CheckMate 238 trial continues to have better relapse and resurvival rates than the nivolumab group, we assume that the same would be so for pembrolizumab, so most people are using anti–PD-1 agents.

In addition, the problem with anti-CTLA4 is that the approval is to use 10 mg/kg, but even at 3 mg/kg, it’s pretty toxic. There was a trial called ECOG-E1609, which compared 3 mg/kg with 10mg/kg. The 3 mg/kg appeared to be less toxic, as expected. It was still toxic, but it did have the same efficacy. Regardless, we rarely use it.

 positive, which is what we call it if you have mutation, then you have 2 options: anti–PD-1 therapy with a BRAF/MEK inhibitor or the final option, which is observation.

I’m a surgeon, but I am also involved in quarterbacking the patients’ care, including running many of the clinical trials that may affect patients with locally advanced melanoma. Let’s take it from the beginning with stage. Let’s say that someone comes in and is stage IIIA, IIIB, or for some patients stage IIIC disease. The reason we know they are stage III is because they had a sentinel lymph node biopsy that was positive. By definition, we would call that micrometastatic disease.

In general, once we know that information, most patients go on to adjuvant therapy. There is some controversy with stage IIIA disease even though there is FDA approval for it. It’s usually pretty straightforward. I will see a patient and give them the results. Many times, I have already told them that the whole point of getting a sentinel lymph node biopsy is to help predict the next steps of therapy, so the patients are not usually surprised by the next discussion.

Before we can do that, we need to open up the possibilities, so we are going to get 

 testing. We’ll usually do hot spot testing because, unfortunately, if you are stage III, you may fail and become stage IV. Getting hot spot analysis, which includes things like 

, and other potential targetable druggable targets, is useful because it is usually bundled in the cost.

Depending on whether they are a patient with 

 mutation, we would offer both options. It depends on experience because there is no head-to-head study. You can extrapolate in terms of response and relapse/resurvival rates based on trials, but we don’t like to analyze across trials. In some centers, they highly prefer BRAF/MEK inhibitors if you are a candidate for that, the reason being that it’s a pill. Pills are simpler. You can hold a dose; it’s quite easy. Many people feel that, from a lifestyle perspective, that’s way better.

Anecdotally, we at the Cleveland Clinic have seen some patients who, when they failed BFAF/MEK inhibitors, their failure display and their relapse seem to be worse. We haven’t proven this with scientific rigor. However, almost all patients have adverse events: not level 3 or 4 adverse events, but a lot of them get level 2 and 3 adverse events. It’s not necessarily a dose-limiting adverse event, but it’s annoying to patients.

That could be fever, or it could be photophobia. There are all kinds of different adverse effects that are off-target effects. Although there are many out there who feel that a BRAF/MEK inhibitor is good for the lifestyle of many of the working patients, I would argue that many of them have the fatigue and unhappiness that I used to see with interferon for the same patient population.

If we flip that over to anti–PD-1–based immunotherapy as adjuvant therapy, 9 of 10 patients have limited adverse effects in our experience, which is pretty good. The 1 in 10 patients who have significant adverse effects are much worse than those that we see in the pill BRAF/MEK inhibitor form. Of course, they have to come in every 3, 4, or 6 weeks, depending on the therapy. We don’t really do the 2-week dosage anymore, but there are various options. Every 6 weeks is obviously less cumbersome. Some people want to see the patients more often because of the potential adverse effects, and they want to get ahead of that. To me, if you look at the data that were presented for stage IV melanoma, especially at ASCO [American Society of Clinical Oncology Annual Meeting] a couple of years ago, it seems that when anti–PD-1 therapy cures, it has the more durable cure of 2 cancer treatment types. Moreover, there have been separate data to show in stage IV with pembrolizumab that lesser tumor burden equates with better anti–PD-1 efficacy. What is better from a burden standpoint than having no evidence of disease?

I have spoken to my colleagues who are in well-known institutions who may disagree with us, but in our practice, we tend to favor anti–PD-1 immunotherapy. Many of our patients seem to be OK with that even if they are working. You have to tailor your therapy to the patient.

Let’s jump to patients who walk in the door with palpable macroscopic disease. It depends on how fast growing the disease is and how much the quality of life is being affected by it. For example, somebody walks in the door with a 15-cm mass, which is quite large, in the axilla. They may be suffering. They may have other adverse effects, though that may require them to delay their surgery, so even though it’s not FDA approved to give it this way, we sometimes give those patients a neoadjuvant dose or 2. Many have read the neoadjuvant papers. They are small series, but the data perhaps show you have some time before you have to take the patient to the operating room. The fact is that giving someone 1 or 2 doses could give you predictive, prognostic information based on the response you get. In the age of COVID-19 [coronavirus disease 2019], where we are worried about getting patients in and when hospitals are getting shut down and reopened, having the flexibility of not having to worry about the exact time to treat is a big reason to consider neoadjuvant therapy.

Most of the data are in anti–PD-1 based immunotherapy. Some trials have shown that if you give monotherapy, patients grow to the point that you couldn’t operate on them. If you give ipilimumab-nivolumab, people have died with it, although it may be much more powerful. The bottom line is that we do it on occasion, but in general, the plan would be to take the patient to surgery and give them adjuvant therapy. They then end up with basically the same options as the patient with the micrometastatic disease. Therapy in the adjuvant setting is usually for a year with close imaging follow-up.

We talked about disease that has intrasametastasis; it may be that they are beyond resection. In our practice, and we have published this, we will use combinations of intratumoral T-VEC [talimogene laherparepvec], which is an FDA approved anti–PD-1 therapy. We published our first 10 patients’ real-world data at least 2 years ago, I believe, and the data were impressive: 7 of 10 patients with unresectable stage III or stage IV disease had complete responses, including 2 patients who had distal, uninjected tumors. With that kind of data, we were very impressed.

It’s clear, from our own experience, that for patients who are in that situation, who have not already failed anti–PD-1 therapy, I would tier it. You have micrometastatic disease, you have macroscopic metastatic disease, and you have unresectable, local-regional disease.

 The approach to multidisciplinary management of melanoma has become increasingly complex over the years. Here at Johns Hopkins Medicine, we’re fortunate to have practitioners from several disciplines: dermatology, pathology, surgery, radiation oncology, medical oncology, etc. We get together every week and discuss all the new patients we’ve seen. We get good perspective from all the different disciplines. It’s important to know what the disease looked like when it began.

The dermatologists are often the first to see the tumor in situ

 It’s also important to understand what the proper surgical approach is. Certainly, the COVID-19 era has complicated some our surgical management possibilities over the last several months. As we learn more about the way the coronavirus is transmitted, we’ve become more facile and better able to treat patients as we usually do.

It’s also important to understand what the possibilities are in terms of medical management for both clinical trials and standard of care with adjuvant therapies, for example. As we discuss these patients 1 at a time in our multidisciplinary meetings, everybody chimes in to volunteer information about the latest data and how we might best serve patients from all different disciplines.

The Optimal Approach to Treating Advanced Melanoma

 The MSLT-II data have been practice changing for multidisciplinary teams, surgery in particular. It used to be that patients with a positive sentinel lymph node in the setting of melanoma would almost invariably go on to a completion lymph node dissection, or at least completion lymph node dissection was the standard of care. The results from the MSLT-II study, of course, demonstrated that there was not an overall survival benefit to doing that. The practice with patients now is that those with a positive sentinel node are oftentimes done with surgery unless there’s a compelling reason to continue with a lymphadenectomy. Those patients, after a positive sentinel node has been found, are oftentimes transferred over to the care of a medical oncologist for the consideration of adjuvant medical therapy.

 MSLT-II and the ECOG study are in an area of strong interest to me, not just because I am a surgeon but because I do new therapy clinical trials. I’m also interested in the studies’ ramifications. To go back a bit, part of what happened was timing. MSLT-II reported out within around 12 months of KEYNOTE-054, CheckMate 238, and COMBI-AD, which were all the adjuvant trials. 

Here is the problem: MSLT-II shows that completion lymphadenectomy after a positive sentinel node—we’re talking about microscopic metastatic disease by definition—did not improve survival. It did improve local-regional relapse-free survival, and that might have ramifications. In that trial, nobody had access to the modern immunotherapies, not even the modern anti-CTLA4 therapies.

Fast-forward to when all those trials were read out, within 12 months. They all required, for those patients with the same scenario, completion lymphadenectomy, so neither trial intersected completely. MSLT-II didn’t have the benefit of the adjuvant therapies, but it said, “Don’t do completion lymphadenectomy.” The 3 major adjuvant trials required completion lymphadenectomy, and most people in their minds have just decided that they will give up completion lymphadenectomy.

If you look at MSLT-II carefully, you will see that there was 1 subset of patients who were head and neck melanoma patients. It is probably because they were understudies, but they were almost better off if they had a neck dissection, not to mention the fact that, when patients recur in the neck, it might be a much worse situation than if they recur elsewhere. They can injure cranial nerves and create long-term permanent functional disabilities as well as cosmetic, aesthetic, and psychosocial problems related to deformity.

The question then is this: For head and neck patients, especially because all the adjuvant data assume they had a neck dissection, we will offer some a neck dissection. We will certainly look at how much tumor burden was in the sentinel node and will make that decision. There is a trial that is percolating now in a cooperative group that may eventually get approved; we will see. We are trying to answer some of these questions, but that is pending as we speak.

For the most part, because we have so much good salvage therapy, we can still operate on the patients later, including anybody below the clavicles with a positive sentinel node regardless of how much disease was present. In fact, in MLST-II, the more disease you had in the lymph node, the less likely lymphadenectomy had any difference, and that’s likely because the cat was already out of the bag: It’s probably really stage IV disease.

We won’t do completion lymphadenectomies unless the patient was absolutely adamant about it. Above the clavicles, I would say maybe one-third of my patients will go on to get a completion lymphadenectomy because of MSLT-II and because of the ramifications of going back and salvaging the situation, if it is even salvageable, with neck dissections, parotidectomy, and so on. You are now dealing with much larger disease.

Impact of MSLT-II Data on Completion Lymph Node Dissection

 When do we think about neoadjuvant treatment? Neoadjuvant treatment is important in the era of COVID-19 [coronavirus disease 2019] because we may not be able to get a patient in anyway, so the fact that there is a concept of neoadjuvant treatment is important. I was part of the NCCN [National Comprehensive Cancer Network] ad hoc guidelines for what to do with melanoma during COVID-19, and we were talking about delaying sentinel node biopsy, getting the tumor out, and palliating to deal with the cell node biopsy later. 

The thought of being able to palliate or at least treat a resectable but macroscopic lesion with neoadjuvant therapy is highly palatable and more attractive. Giving a patient 1 dose gives you a lot of opportunities: No. 1, it’s unlikely, based on the data, that the patient with 1 dose will become unresectable, and No. 2, if the patient has an obvious meaningful reduction in tumor size, you can probably feel good that giving them the same therapy postoperatively is the right thing to do. 

We are not quite there, but if someone has no change in their tumor status with at least 1 dose of anti–PD-1 therapy, there might be a thought to treating them with ipilimumab-nivolumab postoperatively or switching them over even if they are BRAF positive. We’re not doing that, but in theory, that’s the kind of information that this may drive. A lot of these are still ongoing questions. In terms of the impact of neoadjuvant therapy, you also have to look at how big the tumor is. 

In theory, some of these tumors are difficult to get out. I wouldn’t say it’s that often that I have to, for example, resect the axillary vein, but once in a blue moon, when I have to, I also sometimes have to dissect out the various brachial plexus branches. Some of those patients can wake up with poor motor function even if it’s temporary.

This is the idea of being able to shrink the tumor to begin with, especially when you are dealing with other cancers; you are hoping to get true negative margins because it’s the surgery with maybe a little radiation and chemotherapy that will salvage, save, or cure that patient. It is not so in melanoma. In melanoma, if we can salvage and save major blood vessels or nerves even though we might leave a bit of tumor behind, we know that the surgery is helping the immunotherapy, not the immunotherapy helping the surgery. 

What about PD-L1 and BRAF testing? Absolutely. With the BRAF testing, you need to know, No. 1, do you have that as an option? You need to be able to offer your patient that option if it’s available. An exception might be, for example, if it’s a foot melanoma or an acral lentiginous melanoma because the chances of the patient being BRAF positive are extremely small. The question is then about what type of BRAF testing you can do. If the patient is rapidly progressing, and you want to do BRAF-based neoadjuvant therapy, you can do immunochemistry, which is quick. These are usually relegated more for the patients with fast-growing, stage IV disease, but we’ll usually do a PCR [polymerase chain reaction] hot spot type of analysis. 

For PD-L1, the honest truth is that we don’t use it. It’s still a research tool. There are probably some data that, if you’re PD-L1 low, then adding ipilimumab is beneficial, certainly in a stage IV setting. If you look at CheckMate 067, it went back and forth. At the current point, we’re not using it routinely, but it is an important research tool. If you want to go back retrospectively and see what your data are in the real world, it may be that, if you live in a certain part of the country with a certain socioeconomic class of patients, certain insurance coverage, and a certain amount of compliance, you might see differences that you wouldn’t see in a clinical trial. For example, we [at the Cleveland Clinic] published the data looking mainly at the national cancer database. Those are real-world patient data. That’s not highly regulated ECOG status 0 and 1 clinical trials, and there were some differences. For example, there was a meta-analysis of multiple trials—it was more lung cancer than melanoma—but it showed that women did poorer than men with a checkpoint-based immunotherapy. We didn’t see that, for example. The real-world data, especially regionally or locally might differ, and having those PD-L1 data might be useful for your practice. It’s worth having, especially if we are even considering adding ipilimumab to a patient with stage IV disease.

The next rendition in all this in terms of trials is a SWOG trial comparing pembrolizumab with interferon and ipilimumab. I’m not sure that’s going to change practice much, but there may be some interesting information there; it’s certainly confirmatory. The other trial, which has not been published, but you can google it, is CheckMate 915, which is the BMS [Bristol-Myers Squib] trial of ipilimumab-nivolumab vs nivolumab. It confirmed the power of nivolumab, but it showed that adding ipilimumab to ipilimumab-nivolumab didn’t make a difference. What’s interesting is that it’s the first trial that I know about that is a major phase 3 trial that did not require completion lymphadenectomy. 

Nobody knows how many patients did or did not have completion lymphadenectomy; it depends. The trial spanned from when we did it to the time we didn’t do completion lymphadenectomy. We didn’t participate in it, but this has been brought up. It wasn’t powered to ask the question, so even if it was fifty-fifty with half who did and half who didn’t have completion lymphadenectomy—I doubt it’s that ratio—it wasn’t powered to ask that question.

It is interesting to note that this trial, which didn’t show ipilimumab-nivolumab benefit over nivolumab, may have had a lot of patients who didn’t have completion lymphadenectomy. Nobody knows if that affected the trial’s outcome because we haven’t seen the data, and they may not have been able to have the power to answer that. This raises the question about MSLT-II and whether there is a role for completion lymphadenectomy, not because of completion of lymphadenectomy by itself but because the role of surgery here is to help the immune checkpoint inhibitors, not the other way around.

Neoadjuvant Melanoma Therapy: Patient Selection

Molecular testing of melanoma is tremendously important in the stage III setting and across the board as well. I’m an advocate for early next-generation sequencing of melanoma tumors in the high-risk setting, and by high-risk, I mean stage IIB and above, in general.

Those results could inform the possibilities for eligibility for clinical trials in the neoadjuvant setting. It can inform eligibility for standard of care and trial participation in the adjuvant setting. It’s also something that’s good to know about a melanoma patient early in the course of therapy. You never know when a disease might come back, and it sometimes happens rapidly, so knowing the targeted options for a particular patient early on in the course is often very helpful.

As you know, there are 6 medications that are FDA approved for patients with BRAF-mutant melanoma. The hot spot mutations that we hear most about are V600E and V600K. Having said that, there are advantages to knowing what the genetic defects are in melanomas beyond just BRAF and beyond BRAF V600E and V600K. There are mutations that are sensitive to BRAF and MEK inhibition at the 600th locus, for example, and V600R is fairly well reported to be sensitive to BRAF/MEK inhibition.

Additionally, knowing about the NRAS mutation status of a patient’s tumor is important. The NEMO trial demonstrated efficacy in some patients using MEK inhibition. The same is true with c-KIT, which is oftentimes mutant in mucosal melanomas. There, we have the option of using drugs like imatinib as part of the NCCN [National Comprehensive Cancer Network] Guidelines.

I advocate for early knowledge of what the next-generation sequencing results are for patients with high-risk melanoma.

We do not typically test for PD-L1 expression, nor do we test for TMB [tumor mutational burden] in patients with melanoma. Those metrics have been looked at in terms of biomarkers for various clinical settings, such as immunotherapy and targeted therapy. It is sometimes helpful to know those elements for consideration when you’re thinking about clinical trials and eligibility. But in terms of standard of care, we don’t typically send for PD-L1 expression testing or for TMB analysis.

Importance of Molecular Testing in Stage III Melanoma

 CheckMate 238 is an important study. It is a multicenter, double-blind, randomized phase 3 trial. It’s in the adjuvant setting for patients with resected stage IIIB, IIIC, or IV melanoma by the 

, 7th edition. This trial required that patients undergo complete regional lymphadenectomy prior to enrollment. This was important for several reasons. This trial is several years old, but there is an active comparator arm, which is the ipilimumab. We know that ipilimumab has a benefit in the adjuvant setting from prior studies.

Additionally, this was 1 of the trials that included patients with stage IV disease; not all the adjuvant therapy trials have. This also included mucosal melanoma, which is important because not all the adjuvant therapy trials have included mucosal patients. Lastly, this trial included patients with in-transit metastases that did not have nodal involvement, so that oftentimes makes patients stage III but without lymph node involvement. That’s an important patient population to think about.

In terms of relapse-free survival and distant metastasis-free survival, nivolumab was statistically better than ipilimumab on both of those counts. The median recurrence-free survival in the nivolumab group was about twice as long as it was in the ipilimumab group. It was about 52 months vs 24 months. This appeared to be maintained across the subgroups that were analyzed. The patients with BRAF mutations vs patients with BRAF wild type 2 were particular patient populations of interest, and both of those groups demonstrated benefit in CheckMate 238.

The same was true of PD-L1 expression, so patients with or without PD-L1 expression benefited. Disease stage was also an important consideration: patients with stage III and IV resected melanomas seemed to benefit.

Overall survival was looked at as a part of the study. Though there was no significant difference in overall survival between the 2 treatment groups, it is important to note that patients were allowed to go on to get another therapy after progressing on either of the arms. Oftentimes, patients who received ipilimumab in the adjuvant setting and then progressed would go on to get anti–PD-1 therapy. You can see how that would confound the overall survival calculation.

It is worth noting that fewer patients needed subsequent systemic therapy who were in the nivolumab group than patients who were in the ipilimumab group. Although the overall survival analysis was a bit confounded by patients who went on to receive alternative therapies, the data about RFS [recurrence-free survival] and distant metastasis-free survival are enough to make this an encouraging study. Nivolumab is now 1 of the standard-of-care options that we commonly use for patients with resected melanoma in the adjuvant setting.

Another trial that looked at anti-PD-1 treatment in the adjuvant setting was KEYNOTE-054. This was a phase 3, double-blind trial that looked at pembrolizumab, which is another anti–PD-1 agent, vs placebo in patients with resected high-risk stage III melanoma. The patient population here was a bit different from the KEYNOTE-054 trial. This study included patients with stage III disease but not patients with stage IV disease. In this trial, they again required complete resection of cutaneous melanoma that was metastatic to lymph nodes. This trial did not include in-transit metastases patients. 

The patients with stage IIIA disease had to have a lymph node metastasis that was greater than 1 mm of disease. This randomized about 1000 patients to either a year of pembrolizumab or a year of placebo.

There are a couple of interesting characteristics of this study. First, the placebo arm is tremendously helpful here because when these data mature, that’s going to allow us to better understand whether using adjuvant anti–PD-1 is helpful compared with waiting until the disease recurs, and it will not recur in all these patients. Some of them are surgically cured because of their lymph node dissections and the attendant surgery. It’s going to help us understand whether we need to use anti–PD-1 treatment in the adjuvant setting or if we can wait until the disease comes back and then administer the same medication. That’s tremendously important.

Without that data, even without understanding that, the latest analysis showed a 3-year median follow-up looking at sustained improvement in relapse-free survival consistent across all the subgroups similar to the KEYNOTE-054 trial.

If you take all these data looking at anti–PD-1 in the adjuvant setting together, then it’s clear that nivolumab or pembrolizumab is a relatively safe standard-of-care option for patients with resected melanoma.

Similarly, for patients with BRAF-mutant melanoma, the COMBI-AD study was a randomized, double-blind, placebo-controlled phase 3 trial that compared dabrafenib plus trametinib: BRAF inhibitor plus a MEK inhibitor compared with 2 matched placebos in patients who had undergone resection of stage IIIA, IIIB, or IIIC cutaneous melanoma.

Each of the patients received either a year of therapy or a year of placebo, and relapse-free survival was better in patients who received dabrafenib plus trametinib vs the patients who received placebo. The recently published data continued to demonstrate a benefit at 3 years and again at 4 years, so there’s approximately a 20% absolute difference between the treatment arm and the placebo arm. This effect persisted regardless of baseline disease stage, metastatic load, or tumor ulceration status. This is a terrific option for some patients with BRAF-mutant melanoma who have undergone surgical resection of disease.

The US Intergroup E1609 trial was a phase 3 randomized study looking at adjuvant ipilimumab either at 3 mg/kg or 10 mg/kg vs high-dose interferon alpha. Interferon alpha was the adjuvant therapy that was used before the immune checkpoint blockage era. This was an important study with a stepwise approach: It first compared interferon with the standard dose of ipilimumab, which is a 3 mg/kg. This included patients with AJCC7 resected stage IIIB, IIIC, or IV M1a or M1b stage 4 melanoma. It looked at overall and relapse-free survival. I say that this trial was stepwise because it first evaluated ipilimumab at 3 mg/kg vs high-dose interferon, and it then went on to compare ipilimumab at the higher dose of 10 mg/kg vs high-dose interferon.

The bottom line for the study was that adjuvant therapy with the lower 3 mg/kg dose ipilimumab benefited patients when compared with high-dose interferon. Overall survival was better, and relapse-free survival was better.

This study then went on to look at about 1000 patients in the ipilimumab 10 mg/kg vs high-dose interferon study. Here, although there were trends in favor of ipilimumab 10 mg/kg, it did not achieve statistical significance, in part because the toxicity of ipilimumab at 10 mg/kg is quite high. This was an important study because the standard-of-care adjuvant therapy prior to the immune checkpoint blockade era was high-dose interferon. This answers the question about whether survival is improved when using immune checkpoint blocking therapy: In the case of ipilimumab, 

Adjuvant Clinical Trial Data Impact on Treatment Selection

 In the world of stage III melanoma or resectable stage IV melanoma, the only molecular testing that we routinely do is BRAF with additives, because in most hospitals, there is usually a base hot spot analysis that they’ll throw in. Tumor mutational burden [TMB] is clearly associated with benefit. The issue is that melanoma usually has a high tumor mutational burden. It is sometimes like asking this question: “How yellow is the sun?” Almost all your melanomas have high tumor mutational burden: There’s high and there’s higher, but there’s not that much low in melanoma.	

TMB is probably not going to make a difference, necessarily. That is probably also why PD-L1 is not that effective either, because many of our tumors are PD-L1 positive. The idea of a cold or desert tumor where PD-L1 is probably low and certain aspects of immunology like Th1 signaling and tumor-infiltrating lymphocytes are low occurs, and we may see it more in stage IV disease, but whether we see it depends on which part of the tumor you are looking at. It’s less of an issue in the early stage III diseases. Regardless, no study has formed the basis of ordering it as standard of care, so it does not impact our standard of care.

To answer the practical question for stage IIIB, IIIC, and IIID melanoma, there are a couple of things. First, there is a question about stage IIIA melanoma. The FDA approved stage IIIA, and the KEYNOTE-054 trial of pembrolizumab vs placebo included stage IIIA, as did the COMBI-AD trial. You could give stage IIIA, and the questions are why you would and why you wouldn’t. For why you would, you would also want to look at the amount of tumor in a lymph node. To be stage III, you have to have a thin tumor with a minimal number of lymph nodes involved. However, if a lymph node has significant disease, you could still be a patient in stage IIIA, and you might consider doing adjuvant therapy. It’s important to point that out. I don’t want to negate that completely.

There may be some molecular prognostic testing that people might do to help them make the decision 1 way or the other. For stage IIIB, IIIC, and IIID melanoma, these run the gamut. To be a patient with stage IIID melanoma, if you look at the AJCC [American Joint Committee on Cancer] data, a rare number of patients who were stage IIID because they had to have a of number of positive nodes and trans-metastatic disease. Many of those patients may be unresectable, or if you are resecting them, then you are not getting resection with the confidence of negative margins or full-on resection that you would get otherwise, which is something the radiation people would call R1.

An R0 means there is nothing left in my path. An R2 means there is grossly visible disease left behind, and R1 is microscopic disease. In stage IIID melanoma, you may not be able to resect, or you might end up with R1 disease. That needs to be considered. The prognosis for stage IIID melanoma is similar to stage IV, so for that group, you might consider ipilimumab-nivolumab because they are such a high-risk group.

For stage IIIB and stage IIIC melanoma, there’s a bit of heterogeneity there. You can have macroscopic disease and be there, and you can have microscopic disease and be there. In general, for resected disease for stage IIIB through IIID melanoma, you are going to get your molecular testing. If you have a BRAF mutation, then you have 2 options: a BRAF/MEK inhibitor, and within that you have 3 options—some of it is about comfort, and some of it can be hospital negotiations with the companies. Among those 3 options, there are some people who like it so much that, if 1 option gives an adverse effect that the patient can’t stand, they might go to a different combination before going to immunotherapy. 

On the other side is immunotherapy. We are talking about anti–PD-1 immunotherapy: nivolumab or pembrolizumab. Why would you choose 1 agent over the other? Everybody has their own empirical experiences with them. At 1 point, it was nivolumab every 2 weeks and pembrolizumab every 3 weeks. If you want to see your patient more often, then you get it every 2 weeks. If you want to see them less often, then it’s every 3 weeks. Then nivolumab became every 4 weeks. Now with pembrolizumab you can do every 6 weeks.

You have to tailor it to your patient. You have to look at what your patient wants and what their knowledge base is. Some people come in having seen commercials about anti–PD-1 therapy for lung cancer and melanoma, and they are going to tell you, “That’s what I want.” Some patients are going to say, “No I don’t want the IV [intravenous] therapy. I want a pill.” Another question that still exists in the field is this: If you have both options, and it’s dealer’s choice, which do you give? If 1 fails, you are going to give the other when it fails, so maybe you should give the BRAF/MEK inhibitor first. Or should we give anti–PD-1 therapy first?

Like all things in melanoma, we usually ask those questions in the stage IV setting, and we then bring it to stage III melanoma. EA6134, which is about answering that question, is about 90% accrued, but it is still going to be years before we know the answer. It was a tough trial to recruit too, so I’m not sure there will be an appetite to even answer the question in stage III melanoma, but the results of that in the stage IV setting may drive what we do first. For 1 or the other, the bottom line is that you have to tailor to your patients: Do they want a pill, or do they want an IV? In my mind, I look at the data that pembrolizumab in the stage IV setting did better with less disease, and you are more likely to get a durable cure stage IV melanoma with anti–PD-1 therapy. There is no better time to have less disease than in the adjuvant setting. 

I tend to lean toward anti–PD-1 therapy, and I’ll throw in the fact that, although the adverse-effect peaks are lower with BRAF/MEK inhibitor, the chances of having some annoying adverse effects are much more common with the BRAF/MEK inhibitors. Even though you get lifestyle improvement with a pill, you might get lifestyle reduction with some of those chronic annoying problems. This harks back to the days of interferon, which weren’t that long ago, where patients had no other choice: they did it, and it was an annoying thing to be on because they were always feeling sick all the time. You have to tailor it, and you’ll sometimes do 1 option, and they can’t handle it, so you do the other option because they couldn’t handle the first therapy.

CheckMate 915 is an important study. It looked at adjuvant ipilimumab plus nivolumab. This is mostly important because we know that, in some patients, combination therapy of ipilimumab plus nivolumab is highly effective in the stage IV setting. The question that CheckMate 915 asked was about whether you could extrapolate and perhaps say that ipilimumab plus nivolumab would be beneficial in the adjuvant setting.

This was a randomized phase 3 study that evaluated ipilimumab plus nivolumab vs nivolumab alone for patients who had undergone a complete surgical resection as stage IIIB, IIIC, IIID, or IV melanoma according to AJCC Cancer Staging Manual, 8th edition. To some people’s surprise, adding the 2 drugs, ipilimumab plus nivolumab, did not result in a statistically significant improvement in recurrence-free survival in the intent-to-treat population compared with nivolumab by itself.

For some patients in the stage IV setting, ipilimumab plus nivolumab is a good way to go. It does not appear that, in the adjuvant setting, there is a benefit in using dual therapy, at least in CheckMate 915.

A question that sometimes comes up is about whether there’s a role for ipilimumab in the adjuvant setting. Given the studies that we’ve talked about and given the benefits of using anti–PD-1 or BRAF/MEK therapy in the adjuvant setting, there’s a limited role for ipilimumab in the adjuvant setting. In some patients who have received anti–PD-1 therapy in the adjuvant setting and then progressed, they would like to do something additional now that they have been rendered free of disease after surgery. In those cases, if BRAF/MEK therapy is not an option and they’ve already progressed to anti–PD-1 treatment in the adjuvant setting, then ipilimumab can be considered there. The same is true in a patient with BRAF-mutant melanoma who has progressed through anti–PD-1 treatment and BRAF/MEK inhibition adjuvantly. In the right patient who’s enthusiastic about pursuing a third adjuvant treatment, ipilimumab can be considered in that setting. We do not consider ipilimumab as first-line adjuvant therapy for anybody.

Adjuvant Therapy in Stage IIIB/IIIC/IIID Melanoma

When I see a patient who is potentially a candidate for adjuvant therapy, I try to lay out the landscape of what the possible options look like. We talk about the benefits of using adjuvant therapy as a treatment strategy. We talk about the potential benefits of using treatment at an early stage where perhaps we’ll delay the return of disease or prevent the return of disease. We talk about the uncertainty around using therapy in the adjuvant setting vs using therapy if and when the disease comes back because it is still a bit uncertain. 

We talk about what the potential risks are for each of the approaches. We talk about the potential toxicities for immune checkpoint blocking therapy, specifically anti–PD-1 treatment. We talk about some of the short-term issues, and we talk about some of the longer-term issues. We also discuss the potential risks of BRAF/MEK therapy. We talk there about pyrexia, rash, and some of the possible impacts on cardiac function. We also talk about the general reversibility of most of the toxicities that you see with the targeted agents. It’s a pretty wide-ranging discussion that we have with patients who are considering adjuvant treatment for melanoma.

In some patients, the decision is quite clear. For example, the decision is clear in patients who have an immunological comorbidity where you’re concerned about exacerbating some underlying condition like severe rheumatoid arthritis. In patients at risk of exacerbating their arthritis or exacerbating some other autoimmune disorder, if targeted therapy is an option there with BRAF/MEK inhibition, then that often becomes the treatment of choice. With patients with a solid organ transplant, for example, BRAF/MEK therapy is an option that’s absolutely at the top of the list there.

There is also a lot about patient preference that comes into play. Some patients would rather come in on say an every 4-week or 6-week basis to get their therapy and not have to worry about taking a medicine on a daily basis at home. There is a lot that goes into these discussions. At the end of the day, it’s a case-by-case consideration in terms of which therapy we recommend for patients.

There is a particular patient population: patients with resected stage IIIA melanoma, for whom we dig down into the risk-benefit analysis. And the reason there is because, in many of the patients with stage IIIA disease, they will be surgically cured after resection of their disease. For some of them, the risk of a serious toxicity with these therapies often outweighs the risk that the disease will come back. There, we have a careful discussion about not just the adjuvant options, but also about observation alone. Some patients, after hearing what the potential risks and benefits of therapy are, especially the patients with stage IIIA, choose to monitor the disease carefully. They know that they’ve got therapy options at the ready if the disease were to come back.

The management of toxicities of the targeted agents is distinct from the management of the immune-related toxicities. For example, BRAF/MEK inhibitors are associated with pyrexia. Pyrexia can be mild, but pyrexia can also be severe and be associated with hypotension, chills, rigors, dehydration, and sometimes renal failure.

Fortunately, for most of the patients getting targeted therapies, the serious toxicities are quite rare, they are in the single digits. Having said that, it’s important to talk about these potential toxicities with patients and discuss what might be the response if these toxicities were to occur. There are well-established guidelines for when to hold the targeted agents and how to restart them, sometimes at the same dose, sometimes at lower dose. There is also guidance about when to permanently discontinue these agents. For example, this is the case in patients with a severe cutaneous reaction, such as a Stevens-Johnson–type syndrome, which can be life-threatening. In patients who are on targeted agents, we often have to discontinue that therapy permanently. 

We have a frank discussion with the patients about exactly what the toxicities could be, and we ourselves make sure that we are armed with guidance about how to manage toxicities of targeted agents vs immunotherapy appropriately.

An important consideration of immunotherapy vs targeted agents is that, in general, the targeted agent toxicities will reverse once the drug is stopped. That’s not an absolute truth, but for the majority of cases, that is the case. The same cannot be said of the toxicities that sometimes occur with immune checkpoint blocking therapies. For example, in patients who experience hypophysitis, which is an immune attack on the pituitary gland, that is often a lifelong toxicity that needs to be managed with replacement hydrocortisone. That needs to be discussed with a patient: you can sometimes get toxicities that will never go away.

In a young woman who is hoping to have a family at some point, it’s important to understand that the endocrine toxicities of some of the immune agents can be lifelong, and these therapies are worth discussing in detail to make sure that patients understand what the long-term consequences of some of the rarer toxicities might be. 

Balancing Safety and Efficacy in Adjuvant Therapy 

The National Comprehensive Cancer Network [NCCN], shortly after the COVID-19 [coronavirus disease 2019] pandemic arrived, put out some recommendations that gave some guidance about how to care for patients in the adjuvant setting with melanoma.

These recommendations were developed at a time when we knew a bit less about how the virus was transmitted, and it was a bit less clear about how we could keep patients and practitioners safe.

The guidance was important at that time, and it remains important. However, we now have a bit of a better bead on how to keep patients safe even when we’re performing surgeries and treating patients in the adjuvant setting.

One of the guidance points was that therapy may be initiated up to 12 weeks from the time of definitive surgical resection of melanoma. That’s standard; many of the trials include outpatients who were about 12 weeks out. Another recommendation was that adjuvant therapy should be deferred during the COVID-19 pandemic for patients with less than a 50% chance of disease relapse. Back in May or June of 2020 when these guidelines were being developed, that seemed reasonable. Now that we know a bit more about how the virus is transmitted and how we can keep patients safe, if you talked to a patient who was perhaps at slightly lower risk than that, they would still be in favor of initiating adjuvant therapy. The threshold is perhaps a bit higher than 50%. In any case, we are still thinking carefully about risk-benefit in the setting of COVID-19.

Another recommendation included choosing regimens that are the least taxing on the health system and the patient. That remains important for sure.

One recommendation was to use the extended dosing of nivolumab or pembrolizumab, so that would be nivolumab every 4 weeks or pembrolizumab every 6 weeks. That’s reasonable on many levels. You’re keeping patients out of the health care setting as much as you can, and as we’ve all ramped up our telemedicine capabilities, it’s easy to do a midcycle check-in at the 2-week or 3-week mark and have them get a set of blood tests locally. It is easy to be able to look at them via video and make sure they’re doing OK, so that we’re keeping them out of the health care setting.

Another point was that the BRAF/MEK inhibitors are a good option because they’re administered at home, so it again keeps patients out of the health care setting. For the appropriate patient with resected BRAF V600E or V600K melanoma, dabrafenib and trametinib is the evidence-based option, but the NCCN guidelines provided guidance that alternative BRAF/MEK inhibitors like encorafenib/binimetinib or vemurafenib/cobimetinib may be substituted if the drug supply is limited.

We, ourselves [at Johns Hopkins Medicine] have not run into that issue, but those recommendations are reasonable given the pandemic. Anything that can help with less frequent clinic visits, with less frequent infusions, and with increases in telehealth interval symptom checks are certainly are going to be helpful in keeping patients out of the risky settings of the health care environment.

 The impact of COVID-19 is happening now because it’s starting to impact our center. It’s not yet impacting cancer care, and quite frankly, even when we had the worst part of COVID-19 in April and May of 2020, it was probably the worst because of the unknown factors. We shut things down prophylactically, not because we were in the throes of it like New York.

The biggest thing that we noticed right away was that the screening for skin cancer went away all of a sudden. I was available, the hospital would let me operate, but there were no patients: literally nobody. I was there to work, but there were no patients to see me. If you think about it, unless it’s really obvious with a huge, painful, bleeding mass, if it’s a mole that is itchy and irritating with maybe a bit of blood, then in the throes of COVID-19, it may not prompt you to be seen. 

I’m worried about that, these delays in therapy. Once the patient is diagnosed, the next big issue is that we get a COVID-19 test on everybody. I’ve had a couple of positive patients [at the Cleveland Clinic], but not many, which is shocking given the number of cases of COVID-19 in my state, whose visits have been cancelled because of COVID-19. The next question is this: what do you do with it? What happens if they are persistently positive? Should you stop testing? This is the Cleveland Clinic formulation, and it may be so at many hospitals, once the patient is past a certain number of days of asymptomatology.

It also creates a bit of fear and trepidation in the staff as well. From the standpoint of both adjuvant therapy and regular stage IV immunotherapy, there are patients who sometimes need to be rehabilitated after surgery, or they are just not strong enough. You would normally send them to a skilled nursing facility, but right now, those places feel scary to most people and their families, so a lot of that has had a change in paradigm. Despite the fact that patients may not have ever gotten COVID-19, there is something about the environment that we are living in right now that is somewhat depressing, and it chronically weighs on patients. I feel that I’m seeing more negative outcomes with melanoma, certainly in the stage IV setting, than I normally would have seen last year or the year before. 

This is without scientific rigor; this is just empiric information, but I worry that where there is smoke, there is fire. There is something about COVID-19 in general. I also treat oral cavity cancer, and I had a patient who came in with oral cavity cancer. She was in her 80s, and her husband had just died of COVID-19. The effect that all that stress has on patients with cancer could lead to worse outcomes. This is on top of the idea that you have cancer and have COVID-19. That’s a whole other discussion that I don’t have much experience with, thankfully. Living with COVID-19 has been very tough on our patients.

Here at [Johns] Hopkins [Medicine], this has, in many ways, been the epicenter of COVID-19 data. When you turn on CNN you’re always hearing statistics as reported by Johns Hopkins University, so we, like everybody else, have been very tuned into what the ebbs and flows of the virus are. We’ve followed many of our academic colleagues in masking and face-shielding protocols to keep ourselves and our patients safe. In doing so, the patients have appreciated being able to come into an environment that is relatively protected. With some exceptions, we don’t allow visitors into the hospitals in most cases. We’re minimizing the number of patients who can be in a waiting room at any given time. We make sure patients are socially distanced while they are there.

At my own clinic, we are prohibited from double-booking, which has been nice for us. It’s streamlined the flow of patients in clinic to keep everybody safe. In that way, we’ve done our best to mitigate the risk.

In terms of patients’ enthusiasm for embarking on these various therapies, we’re now seeing that we know a bit more about how to control the spread of the virus so that patients are more willing to come into our cancer center and be engaged in clinical trial participation. That was not happening as much at the beginning of the pandemic because we weren’t as comfortable with what it meant to keep patients out of harm’s way. As we’ve been able to open up again, we’re getting more toward normal.

We still certainly abide by the NCCN recommendations to try to minimize the frequency of clinic visits and the frequency of infusions. We use telehealth when we can and when it’s appropriate. We make sure that we’re checking in on patients for symptom monitoring, and we get outside laboratory tests when we have to so that we’re not bringing patients onto the Hopkins campus. We’re finding creative ways to keep patients safe.

Short-Term COVID-19 NCCN Recommendations

The clinical utility of molecular testing of the tumor at the outset of disease is important. It’s helpful to know what the genetic makeup of the patient’s tumor is right at the get-go. We support doing next-generation sequencing for the melanoma hotspot mutations even outside of BRAF, so NRAS, c-KIT, etc. This is good for knowing what clinical trial eligibility might look like, but it’s also good for understanding what some of the standard of care therapies might be: MEK inhibition for NRAS-mutated melanoma or imatinib therapy for c-KIT, for example.

The question about whether to send for repeat molecular testing at recurrence or progression is a good one. Most of the time, that does not bear much fruit, although there are some exceptions to that. When you’ve sent the initial sample, you’ll sometimes get a report back from the laboratory saying that that there was either insufficient tissue or there was a low yield on the sample. For whatever reason, the DNA couldn’t be amplified, or what have you. There, it is important to send another sample so that you’re sure that you’re getting the best information you can.

Another situation where it’s good to send more than 1 sample is in a patient who has had more than 1 primary melanoma, where there’s some uncertainty about what the metastatic lesion is. There, we will often send circulating tumor DNA [ctDNA], looking at the tumor mutations floating around in the bloodstream as a marker of what’s happening in the total body. There we have found that, with tumor mutation testing of the tissue, we have not seen mutation X. When we’ve sent it in the ctDNA, low and behold, we have found it. In general, we don’t send repeat molecular testing at recurrence or progression; however, there are some clinical scenarios where it is helpful to do so.

Our approach [at Johns Hopkins Medicine] to treating patients with metastatic melanoma differs a bit by several factors. Certainly, BRAF wild type vs mutant is a part of that. In any patient who is potentially eligible for a clinical trial, that is oftentimes our first consideration, and that’s consistent with the NCCN [National Comprehensive Cancer Network] guidelines that endorse clinical trial participation for appropriate patients.

Sometimes understanding the patient’s BRAF tumor status can speak to eligibility about targeted therapy options. In general, in patients where they have bulky disease, and the disease is often impeding the function of organs, we are enthusiastic about starting with up-front BRAF/MEK inhibitor therapy in a patient with BRAF-mutated melanoma because those therapies often work quickly, and you can sometimes reduce tumor burden within a matter of a couple of weeks.

In patients where the disease burden is not so high, we have a frank discussion about potential risks and benefits of starting with either targeted agents in the case of patients with BRAF-mutated melanoma or immune therapy. That’s a lengthy patient-by-patient discussion about what the short-term toxicities and benefits might be as well as what we know about the long-term toxicities and benefits of either of those options. It’s not an across the board answer for all patients. It certainly differs patient to patient.

In the era of COVID-19 [coronavirus disease 2019], patients are often more enthusiastic about a therapy they can take at home, where they don’t have to come into the health care setting. But as we’ve learned how to keep patients safe in the COVID-19 era, we’re seeing that patients are more and more willing to come into the hospital, undergo infusions, and receive treatment. 

BRAF-Wild Type vs Mutated Unresectable Melanoma  

 In patients with metastatic melanoma, KEYNOTE-006 was an important study. It’s an open-label, multicenter, randomized, controlled phase 3 trial, and it was done around the world. This trial included patients with metastatic melanoma who were naïve to ipilimumab, and they were assigned to receive either pembrolizumab or ipilimumab. It assessed overall survival [OS] among other things, and the 5-year outcomes were recently published, and pembrolizumab continued to show improved overall survival vs ipilimumab in the 5-year follow-up.

There were about 43% of patients estimated to be alive at 5 years vs only 33% of patients who received ipilimumab. This is similar to some of the other outcomes we’ve discussed; this demonstrates that an anti–PD-1, in this case pembrolizumab, is superior in this way to ipilimumab in patients with advanced disease.

CheckMate 067 is one of the most often cited studies. This was a trial that randomized almost a thousand treatment-naïve patients with advanced melanoma to either ipilimumab by itself, nivolumab by itself, or the combination. It was powered to compare the nivolumab-containing arms to the ipilimumab-alone arm. It was not powered to compare both of the nivolumab-containing arms to each other.

The results of CheckMate 067 at the 5-year mark were presented at ESMO [the European Society for Medical Oncology annual meeting] in 2019, and they essentially showed that both of the arms that contained nivolumab provided significant improvements in OS and PFS [progression-free survival] over the ipilimumab by itself in this patient population. This, of course, is important information, but it leaves some of the questions unanswered.

There seems to be a durable benefit across many of the subgroups. What we don’t totally understand is whether a particular patient is in need of dual therapy with ipilimumab plus nivolumab or whether nivolumab by itself will suffice. There, the search goes on for biomarkers to better understand whether using both agents in combination is worth the increased potential toxicity signal vs using nivolumab by itself. The toxicity issue we discussed with using ipilimumab plus nivolumab together vs nivolumab alone has been a topic of a lot of further study. 

One of those further trials is CheckMate 511, which is a phase 3/4 study to look at whether a different dose combination of nivolumab plus ipilimumab can improve the safety profile of the combination. This study evaluated previously untreated patients with unresectable stage III or IV melanoma, and they were randomly assigned to get either the high-dose nivolumab and low-dose ipilimumab, or the reverse. Once the patients were done getting either of those combinations, they then went on to get the nivolumab monotherapy at the standard dose of 40 mg every 4 weeks.

The primary end point of the study was a comparison of the incidence of treatment-related grade 3 to 5 adverse events. There was a secondary end point that included descriptive analyses of the objective response rate, PFS, and OS. But it was not designed to look formally at whether nivolumab 3/ipilimumab 1, which was the experimental arm, was not inferior: it was not a noninferiority study of that vs the standard dosing of nivolumab plus ipilimumab.

What it did reveal is that it met its primary end point and demonstrated a significantly lower incidence of treatment-related grade 3 to 5 adverse events when you use the nivolumab 3/ipilimumab 1, so the higher dose nivolumab and the lower dose ipilimumab vs the standard combination for melanoma, which is the lower dose nivolumab and the higher dose ipilimumab.

Some of the descriptive analyses demonstrated that there were no meaningful differences between the groups for the efficacy end points: PFS and OS. However, we need some more data to investigate some of those points further.

What this has done is that it has supported decision-making in the clinic that, in particular patients, flipping the dose can perhaps mitigate some of the adverse events that we see with the standard dose of ipilimumab plus nivolumab. I don’t know that there’s a hard and fast rule on this just yet, but it does at least introduce the possibility that, for the right patient, flipping the dose of the drugs may be a reasonable way to go.

Another interesting study that was published in the last year or so looked at the 5-year outcomes with dabrafenib plus trametinib in patients with BRAF-mutated melanoma. This was a pooled analysis of 2 different trials. The trials it looked at were the COMBI-d and the COMBI-v, and these were studies that looked at the combination of dabrafenib plus trametinib. One of them compared that combination to dabrafenib alone, and the other trial compared that combination to vemurafenib, which is another BRAF inhibitor. 

In this pooled analysis, what they looked at was several factors, including performance status, age, sex, number of organ sites, and LDH [lactate dehydrogenase] level. In the multivariate analysis, several of those variables were significantly associated with progression-free survival and overall survival when you looked at the dual therapy in comparison to either of the single therapies. 

The bottom line for this analysis was that frontline treatment with dabrafenib plus trametinib can lead to long-term benefits, in this case, in approximately a third of the patients with unresectable or metastatic melanoma with either a BRAF V600E or V600K mutation.

Modern Perspectives in Advanced Melanoma | Insights