COMBI-AD: Results from a Phase III trial of adjuvant dabrafenib and trametinib for resected Stage III BRAF-mutated melanoma

DR WEBER: The good news was, the data were presented at ESMO, actually, in Madrid. And for the first time, the mature data from 2 large randomized trials. One was the CheckMate 238 trial that I presented, and the other was the COMBI-AD trial, which was presented by Axel Hauschild.

Why don’t we start with the targeted therapy first? That was a large, randomized trial of over 800 patients who were randomly allocated to receive dabrafenib and trametinib, which are the BRAF and MEK drugs that are approved over the last couple of years for metastatic disease versus placebo in a trial that was done internationally. But what distinguished that trial was that they had Stage IIIA, IIIB and IIIC patients. They did not have Stage IV resected patients. It was only Stage III. And that’s a pretty high-risk population on average. You figure it’s at least a 50% risk of relapse at 5 years.

The good news there is that there was a major difference in both relapse-free and overall survival in that trial, of course favoring dabrafenib and trametinib versus placebo. And if you look at the p-value for the difference, it was probably the smallest p-value ever seen. It was listed on the slide that Axel Hauschild presented. It was 0.00000000000001. So there were actually 13 zeros in front of the decimal point. So it was incredibly clearly positive in terms of relapse-free survival. And if you look at the overall survival, there was enough follow-up — and the follow-up was about 3 years in all patients — to be able to calculate a survival benefit, which was clearly present.

DR LOVE: What was the hazard rate?

DR WEBER: The hazard ratio was about 0.5. I think it was 0.48. And if you break it down by substage, by almost any other indicator amongst the subgroups on the forest plots, the hazard ratio has remained highly favorable and all clustered around 0.5. They were between 0.48 and 0.55.

DR LOVE: Now, that’s for survival or progression-free survival?

DR WEBER: For relapse-free survival, whether it was Stage IIIA, Stage IIIB, Stage IIIC, clear advantage for dabrafenib and trametinib versus placebo. And the survival advantage also held, although because there was less follow up for survival than for relapse-free survival, the p-values were not as favorable. But nonetheless, clear advantage for dabrafenib and trametinib, without question, versus placebo in the adjuvant mode in Stage III resected melanoma.

DR LOVE: Now, for how long was the therapy given?

DR WEBER: It was given for 1 year, and then everyone stopped. And, of course, it’s — these are 2 oral drugs: dabrafenib twice a day, trametinib once a day. And the surprising thing, though, was when you look at the side effects — generally we feel these are fairly well-tolerated drugs — but something like 26% of patients had to stop because of a drug-related or a treatment-related toxicity. A little surprising. And the toxicities were the ones you would expect that we’ve seen in metastatic disease. It was fevers, nausea, diarrhea, liver function abnormalities, fatigue. But 26% is a fair rate. And again, a bit surprising. And one wonders whether the toxicity is looked at differently in the adjuvant mode where, frankly, many patients will be cured just sitting there compared to the metastatic mode, where patients obviously who don’t get treated will do poorly.

DR LOVE: Also in terms of people coming off therapy, I would think maybe a significant number got a few months of therapy or even 6 months of therapy. So maybe it wasn’t like nothing.

DR WEBER: No, I agree. The people who stopped tended to get between 3 and 6 months of therapy, and you wonder whether that was enough. Now, when you compare those data, which had relapse-free and overall survival curves that separated and pretty much stayed separated, there was no “banana-shaped” curve like we saw on that BRIM 8 trial. Which, again, were data that were also presented at ESMO. In fact, there was a Presidential symposium where there were 3 adjuvant trails. COMBI-AD I just described, which was dabrafenib/trametinib versus placebo. BRIM 8 was vemurafenib alone — that’s the BRAF drug — versus placebo. And there, you didn’t see as favorable a result because it was an unusual trial. It segregated the Stage IIIC patients, and there were only a modest number of them. There were a couple of hundred patients who got randomized. And if that had been positive, then the entire trial would be reported out.

It turned out, in the Stage IIIC patients, for vemurafenib alone, it was not a positive trial. The relapse-free survival curves came together, and the p-value was not favorable. And the hazard ratio was quite poor.

So, unfortunately, that was not a positive trial. And on the other hand, who would think that using a BRAF inhibitor alone, which is almost never done in metastatic disease, would be appropriate as adjuvant therapy. So that, unfortunately, was a pretty low-impact situation. And those data have not yet been published.

COMBI-AD was published in the New England Journal simultaneously with the presentation at ESMO in Madrid.

CheckMate 238: Efficacy and safety of adjuvant nivolumab versus ipilimumab in resected Stage III/IV melanoma

DR WEBER: And then that brings us to the other trial, which is actually the one that I presented, and that was also simultaneously published in the New England Journal, where a somewhat different adjuvant trial, there it was nivolumab, the PD-1 blocking antibody, versus an active control of ipilimumab, the CTLA-4 blocking antibody.

As we all know, the CTLA-4 antibody ipilimumab is approved by the FDA for resected high-risk Stage III melanoma in the United States. It’s not approved outside the US. But it is an approved drug. And it is used somewhat as adjuvant therapy in the United States for resected, high-risk melanoma. And that was based on a trial where there have been published survival and relapse-free survival data favoring ipilimumab versus placebo. Those are all data published in the New England Journal back in 2016, where we saw that there was clearly a survival advantage to getting ipilimumab versus getting a placebo.

The current trial, which was called CheckMate 238, which was the one presented just a month ago at ESMO in Madrid, was nivolumab versus ipilimumab. And in that trial, there was clearly an advantage in every way, shape or form for nivolumab versus the active control of ipilimumab. The hazard ratio was actually quite favorable. It was 0.66 or 0.65. The p-value for relapse-free survival was 0.0001. Not 13 zeros, but still a very good p-value and a good hazard ratio, which reflects a 35% decrease in the risk of relapsing over time for nivolumab compared to ipilimumab.

DR LOVE: And again, for how long was the therapy?

DR WEBER: Again, therapy lasted 1 year. And there are a couple of caveats. This was a trial that was stopped at its first interim analysis, which was preplanned. And here you only had 18 months of follow-up in all patients, compared to COMBI-AD, where you had 3 years-plus of follow-up. So for the CheckMate 238, nivo versus ipi trial, you really didn’t have any mature survival data. In fact, very, very few patients died, which is good. So you had no comparison of survival. But you did have relapse-free survival and you had another parameter, which is very important, which a lot of folks think is a close surrogate for overall survival, and that’s distant metastasis-free survival. And in that study, clearly nivolumab was favored. And again, the p-valuewas 0.02 and the hazard ratio was around 0.7.

So here you had, again, by any parameter of efficacy, advantage for nivolumab versus ipilimumab. And if you look at the toxicity — again, you decide what parameter you want to measure — treatment-related, high-grade toxicity was clearly favoring nivolumab versus ipilimumab. It was something like 10% versus 42%. Proportion to patients who stopped because of high-grade toxicity, we’re looking at 4% versus 28% or 30%. So clear advantage in terms of side effects for the nivolumab.

And again, if you do the cross-study comparison, which is not necessarily valid, but if you do a “back of a napkin” calculation, if you had only 8% of all patients stop because of toxicity in nivo, compare that to 26% in the COMBI-AD arm, one might argue nivolumab is a pretty well-tolerated regimen as adjuvant treatment for a year.

DR LOVE: So I only have about 218 questions after hearing all that, so let me get started. Incredible. I guess a lot’s going on.

Okay. So first of all, before we start talking about, like, what this all means, what did we know, prior to this, about adjuvant treatment with PD-1 inhibitors?

DR WEBER: There have been or has been a single pilot study, which, interestingly, I was the PI of, where we treated patients, a small number — 33 patients — with nivolumab alone in a pilot adjuvant trial. And to do this, we chose the worst actors. We chose Stage IIIC and IV resected melanoma patients. Of those 33 patients that we treated, we felt it was fairly well tolerated — only something like 3 of them had to go off because of side effects. And of those 33, we have now 5 years-plus of follow-up. There have only been 6 deaths and 10 relapses or — excuse me — 11 relapses over time in that cohort of patients. And that relapse rate has plateaued. I don’t think there’s been a patient that’s relapsed in over a year on that trial, suggesting that there’s a plateau in the relapse curve.

So again, two thirds of the patients continuously, without relapse, over 5 years. You figure most of those patients are probably going to be cured of melanoma. And in that trial, 31 to 33 had resected Stage IV disease. So that’s a pretty favorable scenario.

DR LOVE: Wow! Resected Stage IV?

DR WEBER: These are almost all resected Stage IVs. We had 10 patients with resected brain mets, Neil.

DR LOVE: Wow!

DR WEBER: And of those 10, 9 are currently alive and free of disease and have never relapsed.

DR LOVE: But on the other hand, no prior randomized data of anti-PD-1 adjuvant?

DR WEBER: There had been no adjuvant experience with PD-1 antibodies — in any way, shape or form — prior to that study.

DR LOVE: So what an amazing study. But again, before getting back to that, again, adjuvant, randomized targeted therapy trials prior to these ones you just mentioned?

DR WEBER: There had been none. Not only had there been no randomized, adjuvant targeted trials, there had been no pilot trials of adjuvant targeted therapy. All adjuvant experience was in the metastatic mode.

Potential benefit of targeted therapy in the adjuvant versus the metastatic setting

DR LOVE: So anyhow, first of all, congratulations. Wow! This is incredible that these data came in. Fantastic. But now, let’s talk about what it means.

Let’s talk about the targeted therapy first. What fraction of these patients got anti-BRAF therapy on relapse in the control arms?

DR WEBER: That is a very interesting question. So we don’t have all the full data on this, although the majority of patients who relapsed, of course, would then have gotten further therapy. The majority of them were not re-resected, rendered free of disease. I think 10% of them who relapsed could again be rendered free of disease. And that was evenly balanced, I think, between the arms. And in those groups, actually, the majority, or a significant minority, got immunotherapy, not targeted therapy. Although, obviously the people who had been in the targeted therapy arm of COMBI-AD who relapsed, generally the investigators were going to give them immunotherapy to stay would then break the code because it impacts the treatment.

In the placebo arm, I think a significant group also got immunotherapy. They did not necessarily get targeted therapy.

So the interesting question which you’re getting at is, could you simply wait and treat until later to see how those patients do? You can only do that if you guarantee that there’s an absolute crossover in the placebo arm. Since that could not be guaranteed and was not part of the trial, you’re not, Neil, going to be able to answer the question of targeted therapy now or later. Amongst my colleagues, most feel that once you metastasize, you should be giving immunotherapy front line.

DR LOVE: Sure. But let me ask you another question. I think you mentioned it was international. Most of the patients, were they in environments where they could get further combination targeted therapy, theoretically? Or because there are countries where you can’t access it.

DR WEBER: It’s an interesting point. That trial started something like 4 years ago. So in not all countries in the EMEA — Canada and Australia — was dabrafenib/trametinib approved. So you’re right. There was no guarantee that they’d be able to get it, which probably, in part, accounts for the fact that in the placebo arm, only a small number actually got targeted therapy.

DR LOVE: Did you know that in breast cancer, I don’t know how many years ago, there was the Scottish trial that randomized between adjuvant tamoxifen and tamoxifen on relapse?

DR WEBER: I didn’t know that one.

DR LOVE: And guess what? They saw the exact same risk reduction, adjuvant.

DR WEBER: That’s interesting. There is a trial that’s ongoing in Stage III resected melanoma in Europe — it’s an EORTC trial where patients are randomly allocated to get the PD-1 antibody pembrolizumab for a year or placebo, who have resected Stage III disease. And they’re guaranteed, if they were in the arm that got the placebo, they could cross over on relapse. So that is Alex Eggermont’s trial. And that will, we hope, answer the question of whether you should get it now or whether you should get it later. I would be willing to venture a bit that you’re better off getting it earlier as adjuvant therapy than later as metastatic therapy.

DR LOVE: Yes. I mean, the tamoxifen trial showed there was a survival benefit to getting it as adjuvant therapy. So but rarely that you see that. But look. This kind of hazard rate, I think nobody’s going to quibble with the fact that it’s a great risk-benefit move, so to speak. And certainly it sounds like going through a year of nivolumab is going to be lot better than going through adjutant ipi. So do you think that it’d be possible to actually access that right now, outside a trial setting?

DR WEBER: Nivolumab, in most places in the US, is available through the patient access program. In other words, if you apply to insurance and you have resected Stage III or IV melanoma and you apply for nivolumab and you get turned down, you’ll get it for free through the Patient Assistance Program. Which I think is a very wise move.

DR LOVE: Yes, that’s awesome.

DR WEBER: There will be no expanded access trial as there was for ipilimumab previously. And the reason why is, I think that the FDA today moves so quickly that given the quality of those data, I would venture a guess that both dab/trem and nivolumab will be approved early in 2018. So there’s no reason to do an expanded-access trial. They just make it available.

DR LOVE: Right. Yes, that’s great.

Choosing between dabrafenib/trametinib and nivolumab as adjuvant therapy for BRAF-mutant melanoma

DR LOVE: So you’re the first melanoma investigator I’ve talked to about all this. I’m just kind of trying to process it. I can’t wait to hear what everybody else says. But it sounds like it’s going to be kind of an interesting question about what you do about BRAF patients. Are you going to give them both?

DR WEBER: That’s a very interesting question.

DR LOVE: Do you think you can give them both?

DR WEBER: I don’t think that most insurance companies would pay for both as adjuvant therapy. The question is, does someone need to do a trial of triple therapy versus double therapy? That’s a different scenario. That may be an upcoming trial.

DR LOVE: But meanwhile, what, you have to choose?

DR WEBER: I would say if you’re a BRAF-mutated patient, you’re going to have a couple of good choices. Remember, in the dab/trem trial, it was Stage IIIA, IIIB, IIIC. And you have to have more than a millimeter of tumor in your node if you were IIIA, meaning a slightly high-risk group. With the nivo/ipi trial, there were no IIIAs — it was only IIIB, IIIC and IV. So again, different groups, and to decide which therapy looked more promising, you’d have to do what I described at ESMO as a not very kosher comparison — you’d have to break out the IIIBs and IIICs from the COMBI-AD trial, break out the IIIB, IIIC from the 238 trial and try to compare the nivo versus the dab/trem group.

At the end of the day, I would predict that they’ll look very similar. And I think either treatment for a BRAF-mutated patient would be quite favorable.

DR LOVE: So, I mean, I would assume that those staging things are kind of just artificial. I mean, you really think it wouldn’t work, and — you make the decision you want to give adjuvant therapy, are you really going to look in terms of what subset it is?

DR WEBER: That’s a good point, Neil. But if you look at the ipilimumab trial, which was the EORTC — I think it was 18071 trial, whose updated survival data were published in the New England Journal. The IIIA patients didn’t get all that much benefit. It was all in the IIIBs and the IIICs. So you wonder, does immunotherapy have a differential effect, depending on your burden of disease? I’m hard pressed to believe that’s true with nivolumab.

DR LOVE: How about just random chance in a trial?

DR WEBER: Hard to believe with several hundred patients who had Stage IIIA disease in that trial it was random chance.

I would predict that the FDA will be asked to approve nivolumab as adjuvant therapy for Stage III and IV and they will not differentiate. And they will think, just like you did, that IIIA, IIIB, IIIC, it’s just a continuum, why would we differentiate?

DR LOVE: And how about the anti-BRAF?

DR WEBER: That’ll get approved for all Stage III, I would assume. That’s where the registration would be applied for, although I don’t know. So I think you’ll have, for any Stage III patient who is BRAF mutated, you’ll have some very good options. I would say it’s going to really depend on what part of the country you’re in and whether you’re in an academic or an office setting, whether you get dab/rem or whether you get nivolumab. I think either one sounds like a great option. With longer follow-up, if the survival curves for nivolumab stay flat but the survival curves for dab/trem come down, that may influence people, but we won’t know that for years. And I suspect that you’ll do well either way with either treatment.

Obviously if you’re BRAF wild type, you’re going to nivolumab. Nobody’s going to give ipi up front as adjuvant therapy. However, a lot of patients who get nivolumab adjuvant therapy will relapse with locoregional or isolated oligomet — or unimetastatic disease, a single met, and get re-resected. And then what do you do? That’s where maybe if you’re wild type you’ll get ipi, or if you’re mutated you’ll get dab/trem.

DR LOVE: You mean you convert from a nonmutated to mutated?

DR WEBER: No, no, no, no. If you are mutated patient and got nivo adjuvant therapy and relapsed locoregionally, Neil, then you’ll probably get dab/trem. And it’ll go the other way. If you got dab/trem in the beginning and then you relapsed and then you were rendered free of disease, you’d get nivo adjuvant therapy.

Therapeutic options for patients who experience disease progression on adjuvant treatment

DR WEBER: The big question, though, is, what do you do when you relapse with metastatic disease? This is what everybody wants to know. So…

DR LOVE: You mean after adjuvant therapy?

DR WEBER: Yes. What if you relapse with unresectable disease after adjuvant therapy, and it’s only been a few months and you had nivo? What do you do then?

DR LOVE: You tell me.

DR WEBER: What if it was 3 years ago? So what I think will happen is, I think first, we have to biopsy everybody, everybody who relapses after adjuvant therapy. And you have to start characterizing the tumors. And we have to get better information on predictive markers pretreatment and then markers at the time of relapse for what’s going on. And we need to understand the mechanism of resistance, say, to nivo or dab/trem in the relapsing tumors. That’s number 1.

Number 2. In the real world, if you relapsed less than 6 months after one of those treatments, you’re not going to get treated with the same drugs again. People will switch to something else. If you relapsed more than 6 months from prior nivolumab, I think you’re fair game to get pembro plus epacadostat or nivo alone or pembro plus something else or ipi/nivo as your next therapy.

If you were BRAF mutated, it wouldn’t be unreasonable, say, to give dab plus trem or dab/trem plus an immunotherapy added together.

Risk of disease relapse with adjuvant immunotherapy versus targeted therapy for patients with node-positive disease

DR LOVE: So I want to ask you about epacadostat in a second. But I still want to go back to this BRAF adjuvant situation. And first of all, I’m going to say this is unprecedented in the history of oncology, the current situation that you just described. You have 2 very effective adjuvant therapies. And you theoretically could give either one to the same patient. I don't ever recall that happening before. But anyhow, it’s very, very provocative. I have no idea how people are going to sort this thing out. It’s very, kind of, fresh. But let me just ask you a question. Can you describe an adjuvant situation of a high risk of relapse? And then describe it in terms of number of nodes, high risk of relapse and how you think that risk would change if you gave, for example, adjuvant nivo. So can you describe a case of, say, have a 50% chance of relapse and how you think it would be modified by giving nivo?

DR WEBER: Take any patient who has a Stage IIIC resected situation, where you have multiple positive lymph nodes with macroscopic disease where you can palpate it, or a Stage IIID disease, which is in the new staging category, where you might have matted lymph nodes, multiple palpable nodes, extracapsular extension with a thick primary that was ulcerated, you know those patients with no adjuvant treatment, you can virtually guarantee they’re going to relapse.

DR LOVE: So you think 90%?

DR WEBER: Oh, yes. I mean, let’s say someone had — for example, I saw a lady yesterday who had 25 of 27 positive lymph nodes with actually not even that thick a primary. I mean, that’s a disastrous situation.

DR LOVE: BRAF-positive?

DR WEBER: She was BRAF-negative.

DR LOVE: Okay.

DR WEBER: So she’s in her sixties, 25 of 27 positive lymph nodes.

DR LOVE: Okay. So what’s her 5-year risk of relapse?

DR WEBER: The risk of relapse in her is virtually 100%.

DR LOVE: Okay. And how do you think that’d be modified by giving adjuvant nivo?

DR WEBER: I think that based on the data we know, you’re probably going to drop it by a significant margin. It may well be as much as a third, or 35%, which is explained by the hazard ratio of 0.65.

DR LOVE: So she would go from a roughly almost 100% chance of relapse to 65% chance?

DR WEBER: Correct. Which is a significant increment. You’re not going to guarantee curing, but that’s —

DR LOVE: Absolutely. But so my question is, if she were BRAF-positive, and at the end of that year of nivo or even during the nivo you gave her the dab/trem regimen, do you think it would come down further?

DR WEBER: I would say no, nobody knows the answer. So we’re doing hand waving. I’m literally waving my hands.

I would say that in a patient like that it is entirely possible if you gave dab/trem plus nivo or dab/trem plus pembro you could have a further decrement in relapse or an increment in relapse-free survival. And that is a trial that is begging to be done if the current metastatic trial of, like, vem/cobi/atezo or dab/trem/pembro — there are 2 trials — shows that giving triple therapy is better than just giving either the immunotherapy or the targeted therapy alone. Then you’re going to see a complicated expensive adjuvant trial where you look at triple therapy versus either dab/trem or PD-1 antibody alone or PD-L1 antibody alone. That’s a big trial.

DR LOVE: Just out of curiosity, a highly educated patient comes to you and says, “I’ll pay for it. Will you give it to me?”

DR WEBER: You mean, let’s say dab/trem plus nivo or dab/trem plus pembro?

DR LOVE: Yes.

DR WEBER: If they had a high enough risk and I could get it, I guess I would try it. One might argue, “I have no data to justify that regimen” because it has not been shown to be more advantageous in the metastatic mode than just giving targeted or immunotherapy alone. And you could also argue, “I’m spending an awful lot of society’s money on something that’s unproven.”

I guess my answer is, I’d be tempted. I might well do it. I’m not sure exactly — it would depend on the individual scenario.

DR LOVE: I think it’s a lot different. I mean, you described a situation where you think there’s 100% chance of relapse. That’s really metastatic disease.

DR WEBER: Absolutely.

DR LOVE: I think if you had a situation where you had maybe a 50% chance of relapse and you were bringing it down to 30% or whatever, then it might be a little tricky. And you don’t know, you could be losing benefit. You never know. There might be some kind of weird interaction. I mean, anyhow. I just think —

DR WEBER: That’s, Neil, why I would tailor it to the individual patient. My patient with 25 of 27 positive nodes, you’re absolutely right. That’s effectively Stage IV disease. You know and I know that she’s got disease lurking everywhere. And it’s just a matter of time before she relapses if she doesn’t get treated.

She’s going to start treatment tomorrow, as a matter of fact. But would I consider a triple regimen for someone like that? Yes, I would.

Emerging data with checkpoint inhibitors added to BRAF/MEK inhibitor combinations for metastatic melanoma

DR LOVE: That’s really interesting about what you’re saying in metastatic disease, combining checkpoint inhibitors and anti-BRAF therapy. And again, you get into this issue of, suppose you get it later. But what is the design of those studies? Any thoughts about what you think that you’re going to see? I guess you would expect at least to see higher response rates initially, huh?

DR WEBER: There have been a couple of trials that are small pilot Phase I/II trials that have been done. There’s a vemurafenib/cobimetinib/atezolizumab trial. Atezo is a PD-L1 antibody.

DR LOVE: Approved in bladder. And lung.

DR WEBER: Yes, it’s approved in bladder and other cancers, but it’s not approved in melanoma. And then there’s a KEYNOTE 022 trial that Toni Ribas presented, and I think that’s dab/trem/pembrolizumab. And we’re looking at 80% plus response rates and waterfall plots where virtually everything goes down below the origin.

On the other hand, you could argue, with the proper selection with 20 or 30 patients, you could see the same thing just with dab/trem or vem/cobi. So it’s hard to know.

DR LOVE: Any safety signal issues?

DR WEBER: Actually, not.

DR LOVE: I heard some bad things about toxicity, but not here with BRAF?

DR WEBER: No. I mean, what you look at is, you’re probably seeing additive or even less than additive toxicity when you add the PD-1 antibody to the BRAF/MEK combination. So they appear to be well tolerated. They’re both involved, I gather, in large randomized trials. And we’re not going to know probably for another year whether it’s better to give what, as you configure, is going to be a pretty darned expensive regimen of 3 drugs then either 2 drugs or, say, PD-1 antibody alone. It’s hard to know.

Long-term survival rates for patients with metastatic melanoma after treatment with targeted agents or immunotherapy

DR LOVE: What a fascinating situation. I wonder what could be done to help patients understand this. But anyhow, getting back to this adjuvant situation for BRAF. One of the things that probably people think about, and I want to ask you about this anyhow because we’ve talked about this before, is what we know about long-term remission, cures, that you see with anti-BRAF therapy and with checkpoint inhibitors.

In the past, there’s been this feeling that you’re much more likely — it’s not necessarily likely, but it’s more likely with immunotherapy. And then I started to hear from people, including you, that sometimes you see really prolonged responses, maybe even cures with targeted therapy. And I think that could influence how people see it in the adjuvant setting, although yes, it’s different. But where are we today in terms of understanding, are we curing any of these patients?

DR WEBER: If you look at the long-term data, where you’re looking at 3-, 4-, 5-year survival from the original dab/trem trial, and that trial originally was published almost 5 years ago, so now you have 5-year data. You probably have a plateau of survival in the 20% to 30% range, which, again, you don’t have good comparator for, say, ipi/nivo, because you don’t have 5-year survival from ipi/nivo. But the 3-year survival from ipi/nivo would suggest that you’re going to be higher. You’re going to be closer to 50% out at about 3 years. And it’s actually above 50% for ipi/nivo at 3 years.

So I would have to believe, at the end of the day, if you look at ipi/nivo 5-year survival, it’s going to be better than getting up-front targeted therapy.

Do you see some patients who are on that plateau of survival? Absolutely, with dab/trem. Is it going to be as high as pembro or nivo alone, or ipi/nivo? That’s not clear. And I would say ipi/nivo is probably going to end up plateauing at a higher level and you’ll have more patients that are cured.

I’ll give you, again, an example. The last patient that I saw yesterday was a lady who traveled from another state to see me just for the opinion of whether she should stop treatment with nivolumab, having been on ipi/nivo a year and a half — a year and three quarters ago. Had some significant side effects with hypophysitis. Got her fourth cycle of ipi/nivo and then went on maintenance. And she’s been on maintenance for a year and a half. And her question was, “I’m told I should continue nivolumab forever. Can I stop?” And I said, “Your PET-CT and CT scan show that you have no evidence of disease and you’ve been that way for months. I personally would stop. I think you have a very good shot at being cured.” Because again, if you look at the patients in complete remission who either got ipi/nivo or pembro alone, those patients, 90% of them over time are going to stay in remission. And we don’t have a lot of follow-up, but there’ll be a manuscript coming out from Caroline Robert soon in the Journal of Clinical Oncology. She’s already presented these data at ASCO and ESMO suggesting that, with a modest amount of follow-up, once you stop treatment, at least, a year later 90-plus-percent of the CRs stay in remission.

So I told this lady to stop. I don’t think you need to get treated forever. At least with immunotherapy. I think you can quit after a year, maybe a year and a half.

DR LOVE: I want to ask you about ipi/nivo in a second.

Checkpoint inhibitor-associated immune-related adverse events

DR LOVE: You say there was hypophysitis?

DR WEBER: Hypophysitis.

DR LOVE: Right. Yes. So when you say the word “hypophysitis,” of course I’ve always got to ask you, because I’m so fascinated by that. I’m just kind of curious, how did she present clinically?

DR WEBER: Oh, boy. She got pretty sick. She was incredibly fatigued. Had a sodium of 128. I believe got admitted to the hospital. Was dehydrated. Just felt like hell. And there was a little delay in recognizing what this was.

DR LOVE: Did she have a headache?

DR WEBER: No. Actually, she didn’t really have a headache.

DR LOVE: Because you’ve told me before some of these people have headaches, right?

DR WEBER: Every once in a while, patients will have the worst headache of their life. And your assumption is, oh my god, they’ve developed a giant brain met and they just hemorrhaged. And you get an MRI scan of the brain, and it shows no evidence of disease, but the pituitary is slightly enlarged and inhomogeneous. And then you realize, a-ha, it’s hypophysitis.

DR LOVE: So how long did it take, then, to figure it out? And did she have it on imaging?

DR WEBER: She did not have it on imaging. And I don’t think she had an MRI scan of the brain, because she had no headache symptoms. She had CTs, of course, of her chest, abdomen and pelvis. It actually took several weeks to figure this out.

DR LOVE: Wow!

DR WEBER: Because the treating oncologist in the community wasn’t that experienced. It may have been his or her first time using the 2 drugs together.

So in an experienced center like where I work or any one of many academic centers, whenever you see a sodium of 118, your first thought is, oh my god, it’s essentially an Addisonian situation where you’re either hypoadrenalism or hypopituitarism. And we will check ACTH cortisol, T4, TSH, et cetera, just at a drop of a hat. So we make that diagnosis very quickly. Obviously, without experience, it’s not your first intuitive thought.

DR LOVE: So wow, it’s a real challenge to be an oncologist. It’s a shame that they didn’t pick that up earlier. Although, we just did a multitumor thing last week, and we probably went through 20 newly approved drugs. I mean, completely new things coming into practice. It’s, like, a lot to keep up. But wow, what a shame. Anyhow, at least they finally figured it out.

DR WEBER: Yes, it was finally figured out. She went on replacement. And, actually, the oncologist was pretty gutsy. He gave her her fourth go-around of ipi/nivo while she was on thyroid replacement and corticosteroids. And got away with it. She did fine. And then she’s been on nivo maintenance and feels okay, except for a little bit of fatigue. And she made this interesting observation. She said, “When I went on nivolumab, my blood pressure went up.”

And I said, “Hmm. Interesting. I’ve never heard of that. And I said, tell me what other medicines you take.” And it turns out she takes fludrocortisone, which is probably why her blood pressure went up a little bit. Because that’s the mineralocorticoid piece of the steroids that sometimes when you get hypophysitis you lack. Most of the time, we put people on hydrocortisone and that’s that. We don’t usually have to put them on hydrocortisone and fludrocortisone because there’s enough mineralocorticoid activity of the hydrocortisone to accommodate for what they’ve lost by having the low pituitary output of the ACTH. But in her case, she needed it. And I think that probably pushed her blood pressure up a bit. So I advised her to play around with her scheduling and maybe take it every other, every third day. Not every day.

DR LOVE: Wow! Amazing.

DR WEBER: An interesting case.

Choice of nivolumab and ipilimumab versus either therapy alone for newly diagnosed BRAF wild-type metastatic melanoma

DR LOVE: Where do things stand right now in terms of first-line therapy of metastatic disease? And I’ll start out with BRAF wild-type situation and the PD-1 versus combo ipi/nivo.

DR WEBER: Again, there’s always a controversy. And Jedd Wolchok recently came out with the first author New England Journal article. Interestingly, it wasn’t presented at ESMO, but it just came out in the New England Journal at the same time as all the other adjuvant stuff. So once again, these last few months, it’s the New England Journal of melanoma, which is good. So there’s clearly at 3 years both a progression-free survival and an overall survival advantage. It’s at least 5 or 6 percentage points. It looks as if the survival at 3 years in ipi/nivo is beginning to plateau. It’s above 50%. So the median survival is going to be well beyond 3 years with ipi/nivo. I predict it’s beyond 40 months. And that beats the heck out of anything that we have seen with any other melanoma therapy in a large, well-conducted study.

So when you talk to patients and they say, “Doc, what treatment will give me the longest survival and the greatest chance of being alive at 3 years, 4 years, 5 years?” it’s got to be ipi/nivo. And I can’t argue with that. On the other hand, I can’t argue with a 55% to 60% rate of Grade III/IV toxicity. And trust me, I’ve seen enough patients on the combo to know that the toxicity is pretty impressive. But in a young person, someone who can take it, someone who wants to take it, I think that’s the way to go. And what we’re doing now is playing around with the dose and scheduling to see if you can drop the toxicity maybe to the 30% range to make it more palatable, maintain the benefit and now adding new drugs to the flip dose, ipi/nivo.

So we have, this month, starting an HDAC inhibitor plus ipi plus nivo trial where we flip the ipi/nivo doses. It’ll take a bit of experience to see whether that looks as good as ipi at 3, nivo at 1 as the induction, which is where you run into all the side effects.

Role of PD-L1 expression as a predictive marker of response to immune checkpoint inhibitors

DR LOVE: Getting back, though, to the decision between ipi/nivo and single-agent nivo or PD-1. Do you look at all or think about PD-1 levels or any other, anything else in terms of predictors?

DR WEBER: Interestingly, in that trial in the New England Journal, virtually all the benefit of ipi/nivo versus nivo alone came in the PD-L1-negative population. The PD-L1-positive population, however you cut it, there didn’t appear to be much of a difference in the progression-free and overall survival between the combo and nivo alone. So after seeing the data, I said, “Boy, I’m going to have to start spending more time getting PD-L1 staining on the pretreatment tumors.” Interestingly, not all of my colleagues agree with that thought. I would say more of them disagree. They’ll just say, “Nah, I’m not going to use PD-L1 staining. I don’t think it gives enough information.”

I’ve always been the guy who was cynical about PD-L1 staining as a predictive marker. If you actually do a formal evaluation on a receiver operating curve of the performance of PD-L1 staining, the area under the curve is, like, 0.65, which basically is utterly mediocre. If I tried to sell you a predictive test today with that area under the curve on the ROC curve, you would laugh at me. However, in this scenario it may be useful.

DR LOVE: So bottom line is, for practical purposes, are you getting the assay or not?

DR WEBER: I am getting the assay if I know I need to make the decision between ipi/nivo and nivo or pembro alone. Or usually it’s not pembro alone, it’s usually a trial of pembro or nivo plus some other drug. In that scenario, I’ll get the staining done and I factor that into the equation.

DR LOVE: And if you kind of put aside whether or not you can look at the PD-1 level, globally, in your own kind of gut, interpreting clinical research, et cetera, in your mind, how much is it adding to add the ipi? So, for example, what do you think the relative — I’m not asking you to quote from the trial, but just what you think. How much more do you think it lowers the risk, the relative risk, by adding in the ipi?

DR WEBER: I think it gives you about a 5% absolute benefit in survival. And again, if you’re 85, you may not want that 5%—

DR LOVE: How about hazard rate, though?

DR WEBER: — the toxicity.

The hazard ratio? Again, it’s difficult to look at the hazard ratio because that trial, the 067 CheckMate trial was powered to compare either ipi/nivo or nivo versus ipi as a control. It was never truly powered with appropriate assumptions of having a narrow difference to see a real hazard ratio between them. But you can do anything you want statistically. I think the hazard ratio was in the mid-0.8 range.

DR LOVE: So yes. That’s why I said, “I’m not asking you to interpret the trial, or any trial, but just in your gut.” So having syphoned through all these data, do you think that — okay, so you have a response rate, a benefit just by using PD-1. Do you think that adding ipi adds another 50% to that? Or much less, or much more?

DR WEBER: Oh, no. It has much less. The absolute benefit is not more than 5%, in my view.

DR LOVE: Because that was my take.

DR WEBER: The hazard ratio is probably in the mid-0.8 range. So figure it’s about 0.82, 0.85. That’s a 15% relative change, a 5% absolute change. Some patients may not want to deal with that. So this is where you need to get the patient involved in the decision-making. This is not such an easy decision.

DR LOVE: This also happens a lot in oncology, though, this balancing things off, et cetera.

Biologic rationale for the addition of HDAC inhibitors to immunotherapy

DR LOVE: Let me just ask you a couple of other follow-up questions, again to all these things we’ve been talking about. First of all, in terms of adding things in to immunotherapy and checkpoint inhibitors, first of all, you mentioned HDAC inhibitors. I’m not sure I’ve heard a whole lot about HDAC inhibitors in melanoma. It kind of doesn’t seem to catch on anywhere that much. But what’s the thinking there?

DR WEBER: HDAC inhibitors are actually drugs that are approved in some of the hematologic malignancies.

DR LOVE: Right.

DR WEBER: So HDAC inhibitors are epigenetic agents, which, as you can tell by the name, they impact on the acetylation of histones which are an integral part of the DNA, and they do a lot of gene expression control. It’s just part of our normal cell metabolic regulation apparatus. So it turns out that —

DR LOVE: Romidepsin and T-cell lymphoma, for example.

DR WEBER: Yes. So yes, exactly. So it’s approved in narrow indications in the hem malignancies, because there it has a direct antitumor effect. In a histology like lung cancer or melanoma, HDAC inhibitors alone probably don't do very much. What they do accomplish, though, is, they have immunologic boosting effects. And that’s only been found over the last couple of years. So they would be perfectly matched to add to a drug like ipi and nivo, because they downmodulate T regulatory cells, downmodulate myeloid derived suppressor cells, which are the bad guys. And that may help unleash more of an immune response when you add either nivo or pembro or ipi plus nivo. And they’re already in combo trials.

Activity and tolerability of the IDO inhibitor epacadostat in combination with anti-PD-1 antibodies in melanoma

DR LOVE: Yes, I think I’ve heard that from Paul Richardson about myeloma, now that I’m thinking about it. What about the other agent, and I’ve been hearing more about this. It sounds kind of promising — or this kind of strategy, epacadostat?

DR WEBER: Epacadostat is an IDO inhibitor. And IDO is a substance made by tumors and by antigen-presenting cells. That’s yet another immune-suppressive influence. It’s not like a checkpoint. It’s not a molecule on the T cells. It’s a soluble material or material that’s at the membrane of the cell that acts to suppress immunity. It sucks up certain amino acids, like tryptophan. And if the T cells don’t have any tryptophan, they are not happy. If you have tryptophan or other essential amino acids, the T cells work better. So it’s an indirect way of boosting immunity.

And if you use an IDO inhibitor alone in melanoma, it doesn’t do anything. It has almost no antitumor activity. But it looks very promising if you add it to PD-1 blockade because in a mouse, it works great to boost the effects of the PD-1 blockade. And in the initial trials in people, of which there have been 2 or 3, there’s been epacadostat plus nivo, epacadostat/pembro.

You get a very high response rate in the 50-plus-percent range. And again, you expect maybe 40% front-line response rate with nivo or pembro alone. But most importantly, the progression-free survival looks like it’s about a year. And PFS is your endpoint, because with PD-1 antibody alone in most trials, progression-free survival is like 6, 7 months. If you go up to 12 months, that’s probably a real change. And if you do enough patients in a Phase II study, figure, I don’t know, 50, 60, 70, 80 patients, and you see a PFS across the board of 12 months, I would say that’s very encouraging. And I’d certainly bet that in a randomized trial, pembro/epacadostat would be better by virtue of better PFS than pembro alone.

And we’ll find that out early in 2018. I think there’s this randomized trial ongoing. And it should have enough time to mature, to give us those data.

DR LOVE: This is an oral agent?

DR WEBER: Yes. I mean, it’s easy to take.

DR LOVE: No tolerability issues?

DR WEBER: Minimal toxicity changes versus pembro alone or nivo alone. I say it looks like a very promising idea. And now we’re looking at ipi/nivo/epacadostat. So we’re doing the double combo. Now we’re looking at the triple combination. That’ll be very interesting. I like that idea.

Efficacy and safety profiles of the BRAF/MEK inhibitor combinations dabrafenib/trametinib, vemurafenib/cobimetinib and encorafenib/binimetinib for BRAF-mutant melanoma

DR LOVE: Getting back to targeted therapy, we were talking about the dab/trem combination. Of course, you have vemurafenib/cobi. And there’s another one, encorafenib/binimetinib —

DR WEBER: Yes. No. It’s not a drug that I’ve spent much time working with. But technically, the data looked just as good with all 3 BRAF/MEK combinations. There’s no reason why, in terms of efficacy and toxicity, it wouldn’t get approved. I assume that’s going to happen.

DR LOVE: That’s what I was going to ask you, do you think that these 3 combinations, in your mind, is there any way to separate out either the efficacy or toxicity ratios?

DR WEBER: At the end of the day, I assume the efficacy will be identical. I think the only difference will be the toxicity profiles. So if you start with dab/trem and somebody has horrendous fatigue and fevers, you can switch them to vem/cobi and then you’ll be dealing with a different set of issues. It might be dermatologic or maybe nausea. And if that didn’t work, you could switch to a third potential combination. But again, it’s going to be based on the toxicity profile. I’m hard pressed to believe that any of those combinations will be any different in terms of efficacy. Median survival is all going to be about 25 months. And to me, that’s the key indicator. And the plateau is going to be pretty far down below 30% at 5 years for all of them.

DR LOVE: That’s interesting. And we’re finding now in non-small cell lung cancer for the BRAF patients, they’re using dab/trem first line. Non-small cell lung cancer.

DR WEBER: I would predict it would work very well. Whether you will see the same duration of response and the plateau with non-small cell compared to melanoma, not clear. I don’t think BRAF is as important a driver in non-small cell lung cancer as it is in melanoma.

DR LOVE: Yes, no. I think you’re right. The waterfall plot looks good. But I don’t think they see that kind of duration. And then, of course, colon is a complete mess because that doesn’t work at all. They’re giving, like, chemo, vemu and EGFR antibody. They’re seeing some responses. But it’s interesting how in the different tumors, BRAF just kind of plays out a little bit differently. Any idea why?

DR WEBER: BRAF plays a very important role in driving melanoma. The MAP kinase pathway is key. In colon cancer, for example, EGFR is a much more important part of the pathway. So it allows a bypass pathway to arise much more readily. In those cells, compared to, say, in melanoma, where resistance to BRAF/MEK inhibition is almost all through MAP kinase pathway reinvigoration, so to speak, it’s not exactly that way in colon cancer. EGFR plays a more important role.

Association of the diversity and composition of the gut microbiome with response to anti-PD-1 blockade in patients with metastatic melanoma

DR LOVE: So I want to get to your cases in a minute, but a couple other things I just want to throw out to you. So here’s one. Just kind of curious. Any thoughts about, so oral from ASCO: Association of the diversity and composition of the gut microbiome with responses and survival in metastatic melanoma patients on anti-PD-1 therapy.

DR WEBER: This, I believe, was Jen Wargo’s work.

DR LOVE: Right.

DR WEBER: And these are very important data in Science Magazine. Outstanding work. And the Science article basically recaps what she presented at ASCO. And there will be a companion article from Laurence Zitvogel, who’s from the Gustave Roussy in Paris. And both of them were similar analysis, although there were some major differences in the outcomes and in the conclusions. But the bottom line is, there are certain bacterial species associated with resistance and with efficacy of PD-1 blockade. There’s no question about it. And bizarrely enough, they appear to be mostly anaerobic clostridiales and bacteroidales are the species that are associated with either doing well or not doing well.

And why anaerobes? Obviously, anaerobes have the ability to metabolize things that probably get into the bloodstream that can be immunosuppressive or immuno-boosting. And 5 years ago, no one would have thought this was the case. But now we’re looking at hundreds of patients’ worth of data, and it’s obviously real. And in this Zitvogel data, I think they came up with a specific bacterial species. It was a group of species with the Wargo article. That was the one that was discussed at ASCO.

So these are important data. And guess what’s going to happen? This will lead to a trial. What’s the trial? You’ll take the bacteria that were associated with the good outcome with PD-1 blockade, and you’ll attempt to isolate some small number of bacteria. You’ll lyophilize them, put them in a pill. And they’ll just be administered as a pill after maybe giving antibiotics to clear out a lot of gut flora to make space microbiologically. Then you’ll try to reconstitute the gut flora in patients who then get PD-1 blockade and compare that to PD-1 blockade where you take an irrelevant bacteria or a placebo of some sort.

DR LOVE: Can you talk a little bit more about what’s going on? Is it, like, a pharmacologic issue? That it’s in some way interfering with the checkpoint inhibitor molecule pharmacologically? Or is biologic? What’s happening?

DR WEBER: No. I think it’s indirect, and I think that the bacterial species cause secretion into the bloodstream of immunosuppressive influences that boost the activity of things like T regulatory cells. And I think both of those articles and Jen Wargo’s presentation at ASCO imply that there were clear effects on the immune system of the presence of some bacterial species versus others.

The problem I have is, to reconstitute human microbiome is not so easy. For example, if I gave you big-time antibiotics intravenously for a week, you’d wipe out 99.9% of your bacteria. But if stopped, a couple of weeks later you would just reconstitute and you’d be back to where you were when you started.

So changing the microbiome in humans is not so easy. The model systems are mice. And in a germ-free mouse, it’s a little different than if it’s you or me. But people will now put a lot of effort into understanding how to reconstitute the microbiome. And the full weight and force of the biotech and pharmaceutical industries will be applied to this. And I predict we can figure this one out in the next 5 to 10 years.

DR LOVE: Yes. Actually, Allan Venook of GI fame or interest brought this up to me recently, and he was also talking about the difference between the bacteria on the right side of — they had this big thing, right versus left colon cancer and the bacteria are different there. When they actually do this work, where do they get the samples from?

DR WEBER: I mean, the work in humans is mostly — they grab it through a colonoscopy. Most of the work we do is from stool samples, though. I mean, for example, when Jen Wargo did her study, she collected poop, pretreatment/on treatment blood, tumor and tried to connect them as did Laurence Zitvogel.

DR LOVE: Really? I guess that sounds strange.

DR WEBER: So you wouldn’t believe — so it’s actually hilarious, Neil. Every patient we see who goes into immunotherapy, we hand them a little box and say, “Would you mind donating a sample of poop for our study?” And they look at you like you’re insane. And they said, “Now, I know you find this hard to believe, but it’s amazing how what’s in your poop may impact on how your body handles this treatment.” And I say, “Look at this way, it’s a renewable resource. There’s plenty of it. It shouldn’t be a problem. It’s not painful to donate it. It’s a little messy, but here’s how you do it.”

DR LOVE: Wow.

DR WEBER: We give them the little kit. And then it gets frozen away, and you analyze the DNA of the bacteria.

DR LOVE: Wow, that’s — I mean, I like it because it’s different. Really fascinating. I’ve heard little things about this from people interested in immunotherapy in different — I can’t even remember whether it was lung or hem or whatever, but this idea of the connection between the microbiome and checkpoint inhibitors. Has that only been looked at in melanoma, or with other cancers?

DR WEBER: No, people are looking at in other cancers. Bladder, lung cancer.

DR LOVE: Huh.

DR WEBER: There’s no question it’s probably real. Five or 6 years ago, I would have thought it was a bunch of crap, no pun intended. But today it is unequivocal, if you look at work by Tom Gaiefsky, by Laurence Zitvogel, by Jen Wargo. I mean, those are the 3 major investigators in the melanoma-related field. There’s no question it’s real. And, actually, Zitvogel had an interesting abstract that was presented by one of her kidney colleagues, but it suggested that people who got antibiotics while they were getting PD-1 blockade didn’t do as well as those who got no antibiotics.

DR LOVE: Wow!

DR WEBER: But, I mean, she had data from a couple of hundred patients. It looked real. So it’s a little scary.

DR LOVE: Huh.

DR WEBER: It suggests the type of antibiotics you take may impact on how well you do.

DR LOVE: Wow.

DR WEBER: So this is an up-and-coming field.

DR LOVE: Hmm. What about — I mean, pretty naïve about this stuff, but, like, when you take — I don’t know, maybe patients take it, too. There are these things you buy in the, whatever, pharmacy. Some of them are refrigerated. What do you call them, macrobiotics?

DR WEBER: Yes.

DR LOVE: What does that do to the biome?

DR WEBER: I have spoken to experts in the field, and remember, I’m not a microbiologist. I am not an infectious disease specialist. But if you speak to people at my institution like Marty Blaser, who is a world-class expert on the human microbiome and an infectious disease doc, who’s written lay books on the subject, Marty Blaser would tell you that taking probiotics will have minimal impact on your microbiome because you have this homeostatic reset and you’ll take the pills and then a couple of weeks later you’ll just be back where you started. So he doesn’t think that that would work unless somehow you blasted the body with big-time antibiotics and reset the playing field for the microbiome and then gave the bacteria. So this is going to be the subject of some serious research over the next 5 years.

DR LOVE: So here’s another one I want to ask you about. I know nothing about it, but I’ve always got my eye open for antibody-drug conjugates. And I see there was presented at ASCO, glembatumumab vedotin. Anything there?

Ongoing trials of immunotherapy combinations in patients with melanoma refractory to immune checkpoint inhibitors

DR WEBER: Yes. There were some interesting data with that drug alone. And that drug with CD27. Some interesting responses have been seen in patients with brain mets, potentially patients with ocular melanoma. We have seen anecdotal responses, small numbers but real responses in patients who failed other therapy. So there’s something interesting going on there, and that is worth pursuing. I don’t know how far that’s going to go. But that’s pretty much all that you can say. It looks quite interesting.

Other interesting stuff from the ASCO meeting that were also presented at ESMO in Madrid were the LAG-3 plus nivolumab trial that was presented at both meetings by Paolo Ascierto. And there, I mean, that’s one of the more interesting combinations where you had some responses in patients who were either PD-1 relapsed or PD-1 refractory.

So again, there are many trials in either PD-1-refractory patients who never responded to PD-1 or those who were stable for a while or responded and then relapsed. And this looks like one of the more promising ones. So this will certainly be pursued.

There were responses mostly in those who were stable to PD-1 and then progressed and then got PD-1 plus LAG-3 or nivo plus LAG-3. It looks very interesting. There are a couple of LAG-3 antibodies out there. So you’re going to hear a lot more about that.

DR LOVE: And I remember hearing about LAG-3. And is it TIM-3? Are there other checkpoints?

DR WEBER: So, Neil, there are at least 8 or 9, maybe 10 different checkpoints on T cells, all of which are receptors for various ligands. And the so-called exhausted T cells tend to PD-1-positive, LAG-3-positive, TIM-3-positive. So if you are resistant to PD-1 blockade, it may be because you have too high a level of the other breaks, the other checkpoints, and if you begin to combine the blocking of those checkpoints, you may get higher response rates. So that will be strongly pursued by a number of groups and a number of companies.

Case: A 72-year-old man with basal cell carcinoma (BCC) achieves a very good partial response to the hedgehog inhibitor vismodegib

DR LOVE: So let’s try to get to a couple of cases. We haven’t talked about basal cell for a while. So, want to tell me about your 72-year-old man?

DR WEBER: Yes. I mean, I occasionally will treat patients with basal cell. So this was a guy treated some time ago who was treated with the hedgehog inhibitor vismodegib. And the problem is, we see a lot of patients who have facial basal cells that often are ignored. And they end up needing to have disfiguring plastic surgery. They get radiation. It returns. And then you have to operate in a radiated field. So the nice thing about that and the newer hedgehog inhibitors is, you can often do neoadjuvant therapy with the hedgehog inhibitors, shrink the tumor and have a much better therapy.

So I think that this was the classic case. But this was a guy who was — he was an alcoholic. He was not taking care of himself. He let this thing get very large. He had a very good partial shrinkage of his tumor but had the typical side effects: the asthenia, the poor appetite, fatigue, muscle aches. He elected to go off. And then he was able to have a decent surgery.

So I think this a typical story and, both with the hedgehog inhibitors in basal cell, this will be a common phenomenon. As you also have heard, there’s a PD-1 inhibitor that has activity in squamous cancer. Actually, several of the PD-1 antibodies do, with a 50% response rate in locally advanced or metastatic squamous cancer of the skin. So you’ll see the same thing, someone with a big, ugly squamous cancer that recurs on the face will get treated with PD-1 blockade, and then as a neoadjuvant treatment and then you’ll operate on them. And I think it’ll be better.

It’s an expensive way to deal with tumors, many of which are ignored. But I think that the patients will clearly benefit and you’ll have a much easier surgery. You won’t have these unbelievable surgeries that the plastic surgeons show you at your conferences, where they have these giant flaps and have to remove large portions of someone’s face. I think it’ll make the surgery much more palatable.

DR LOVE: That’s really cool, though, when you think about it, neoadjuvant checkpoint inhibitor like debulking people. That sounds really cool.

Efficacy and tolerability of the hedgehog inhibitors vismodegib and sonidegib

DR LOVE: Getting back to basal cell. We heard about vismodegib for a long time. What about sonidegib?

DR WEBER: I mean, they’re both excellent drugs. Both of them are being used currently, and I think both of them will have relevance in the neoadjuvant mode, which is, for me, what I find most interesting.

DR LOVE: And what about tolerability? Because you mentioned the problems that this patient had. First of all, is there any difference between these 2 hedgehog inhibitors in terms of tolerability profiles?

DR WEBER: It appears that the more recent drugs have slightly better tolerability profiles. There are a lot of problems with fatigue and asthenia with vismodegib. It seems like they’re a little better tolerated. No question.

DR LOVE: What about this — you mentioned lack of appetite. But what I actually heard about was so-called dysgeusia, where food doesn’t taste good. Is that what they tell you?

DR WEBER: No. When the food doesn’t taste good, they lose their desire to eat. So they just don’t have an appetite. It just tastes terrible. It has this horrible metallic taste, which can happen, by the way, with interferon. Sometimes it can happen with checkpoint inhibitors. But it’s a very common phenomenon with hedgehog inhibitors.

DR LOVE: With checkpoint inhibitors. That’s interesting.

DR WEBER: Yes. It’s a rare but not unheard-of complication, having this metallic taste. But it’s just so much more prevalent and common with hedgehog inhibitors.

Monitoring and management of immune-related adverse events associated with immune checkpoint inhibitors

DR LOVE: I know you have long interest and a lot of publications in terms of the issue of side effects, immune-related side effects of checkpoint inhibitors. I thought I saw a paper that you were involved with some kind of system to monitor for autoimmune side effects? Does that ring a bell?

DR WEBER: There are now coming up position papers from ESMO and from ASCO.

DR LOVE: Online tool.

DR WEBER: Yes, online tools for how to handle the side effects.

DR LOVE: ASCO 2017.

DR WEBER: Yes. So it’s clear, and this contains — it’s more management tools than surveillance tools — but there are a lot of things in there that make it clear that what the experts do who have been doing this for the last 10 or 15 years should be done in the community, meaning you should, at frequent intervals, you should be talking to patients who are on ipi/nivo more than you might have. Have your staff talk to them to find about side effects. Check their endocrine parameters. You should be inquiring about their GI tract, level of energy and all these things, much more so than — I don’t know. Let’s say you give them nivolumab alone or pembro alone and you say, okay, see you in 2 or 3 weeks. And you kind of forget about them. But with ipi/nivo, and sometimes with PD-1 blockade alone, you just need to be more on top of the patients. So it’s more work for the staff, the nurses, the nurse practitioners.

DR LOVE: Different mindset too, in terms of whole pathophysiology. And we’ve been conditioned to think about chemotherapy and that profile. And here, we’re looking at something totally different.

DR WEBER: Yes. The example, Neil, is that patient who came to me who had been treated by a community doc who probably wasn’t quite as experienced. I don’t think they — the institution where they were treated didn’t pick up on the hypophysitis quickly. I mean, the first thing that someone here at NYU would think about would be, “Oh my goodness, do they have an endocrinopathy?” That’s always the first thing you think about with fatigue, low sodium, that kind of thing.

DR LOVE: Right. I was thinking about that case also.

Case: A 37-year-old man with Stage IIIC resected melanoma discontinues adjuvant nivolumab after 9 months due to a stress fracture of the left tibial plateau

DR LOVE: Let me ask you about a couple of other cases. We were talking about adjuvant therapy. Maybe you can talk a little bit about your 37-year-old man.

DR WEBER: So yes, it was a 37-year-old guy with Stage IIIC disease. He was on nivolumab. And then he had pretty long-term therapy. He was out at 9 months, doing great. He was actually on a clinical trial. And then he developed these myalgias. And they slowly got worse. And he was a pretty tough guy. He was a relatively young guy, in good physical shape. And then he said that he was playing soccer with his kids one day on a, like, a Friday. And then the next day he woke up, he had pain in his left lower knee. He had it elevated. Then I saw him the next Monday and was a little concerned that slowly but surely, the muscle, the myalgias had gotten worse. His CPK was slightly elevated. I decided he probably had a little bit of myositis, but it wasn’t disabling. He got a brief steroid taper.

And then his pain just got worse and worse in the left knee. And I got just got him an MRI of the knee and sent him to one of our orthopedists. And he had this weird stress fracture of his tibial plateau. And I spoke to the orthopedist who said, “We don’t see this in young people. This would be almost unheard of in a 37-year-old.” So I had to make the executive decision, “Am I going to continue him, or am I not?” Do I figure this was an unlucky guy who was kicking a ball with this kids and had some weird fracture unrelated to treatment? Or was this all treatment related? I elected to stop him and take him off treatment. And then more than a year later, the guy’s — and this is a while ago — more than a year later, the guy’s fine. So I hope I made the right decision. Although remember, with adjuvant therapy, with melanoma you’re never really out of the woods. But if he gets 5 years without a relapse, the vast majority of relapses would have occurred by 5 years.

So the toxicity issues get even more complex in the adjuvant mode.

DR LOVE: Absolutely.

DR WEBER: Because this guy, he has a 70% chance of relapse, but he has a 30% chance that he’s cured just sitting there. So you’ve got to be careful how you judge the risk and the benefit.

DR LOVE: Absolutely. But, I mean, on the other hand, how much treatment did he actually get?

DR WEBER: He got a fair bit of treatment.

DR LOVE: So, I mean, maybe that’s all he needed.

DR WEBER: He got 9 months’ worth of therapy.

DR LOVE: Nine months. And who knows. We’ll never see that trial, 9 months versus a year. But I mean pathophysiologically, what do you think was going on? You said you don’t see it in young people. Do you see it in the elderly?

DR WEBER: He said this is something you would see in a patient in their seventies or eighties.

DR LOVE: So what do you think was going on pathophysiologically?

DR WEBER: I think that he had some inflammatory component, either in the joint or in the soft tissues with the myositis. And I think that led to weakening of his underlying bone structure and led to the stress fracture.

I think that this was somehow autoinflammatory. And off treatment, he slowly is getting better. His fracture is healing. Or he got better. I mean, he was off for a while. But it took, like, 6 months for him to feel that he could go out and play soccer again with his kids. And I said, “No soccer for at least 6 months.” And now he’s more physically active and he’s fine.

DR LOVE: Wow! Really interesting. It also brings up a generic question, you’re kind of the dean of autoimmune toxicity with melanoma in checkpoint inhibitors in general. Do you think that in people where you see autoimmune toxicity that in some or many of these cases what’s going on is sort of the activation of subliminal autoimmune diseases these people have that aren’t clinically manifest?

DR WEBER: I would agree. And there’s a colleague of mine, Iman Osman, at my own institution, who’s been exploring siRomics. That is, she looks at preexisting autoantibodies. And she finds certain types of autoantibodies appear to be more elevated in those who have autoimmunity than those who don’t. And she even has a mouse model with a human FC receptor knocked into the mouse, where she can give the serum or she can purify antibodies from the patients, give them to the mouse and reproduce some of the side effects. It’s very interesting. And we’re going to try to work on understanding with her how can you profile patients pretreatment and predict who’s going to have side effects.

Role of chimeric antigen receptor T-cell therapy in melanoma

DR LOVE: So let me ask you about your last, really interesting case. But another thing just popped into my head I’ll just throw out there, CAR-T in melanoma.

DR WEBER: CAR-T in melanoma has not really panned out. In fact, CAR-T cells in solid tumors has not been that promising. As you know, there have been 2 CAR-T treatments approved by the FDA at a phenomenal cost. One is pediatric ALL. The other is adult hematologic malignancies. And solid tumor CARs have not been that promising. One of the more promising ones so far is the folate receptor in ovarian cancer, and that’s the subject of an ongoing trial.

DR LOVE: CAR-T?

DR WEBER: But yes, with CAR-T cells.

DR LOVE: Because I heard about — I think it’s mirvetuximab as an antibody-drug conjugate. But this is CAR-T you’re talking about.

DR WEBER: This is CAR-T cells against the folate receptor, which is highly overexpressed in ovarian cancer.

DR LOVE: Right. Right. Yes, they had this antibody conjugate.

DR WEBER: Remember, Neil, with CAR-T, you need a cell surface molecule that’s recognized by the antibody piece of the CAR-T cell to transduce a signal into that T cell. So again, there are not that many absolute tumor-specific cell-surface molecules easily recognized by antibody. Those that are often are present in normal tissues. And you get all this off-target toxicity. So the folate receptor looks interesting. It’s something that I think is being pursued. But I wouldn’t hold your breath in melanoma.

Case: A 45-year-old man with recurrent BRAF wild-type Stage IV melanoma and scleroderma achieves a complete response to pembrolizumab after disease progression on multiple therapies

DR LOVE: So let’s finish out. I’m just curious about your 45-year-old man. This sounds really fascinating, with the scleroderma.

DR WEBER: Oh, boy. That was a tough one. He came to me with scleroderma and melanoma, because his rheumatologist felt this was paraneoplastic. And his rheumatologist was a very experienced woman who’s been doing this for 30 years. And she has good reputation in the community. And I figured yes, she was probably right.

DR LOVE: So this patient had multiply relapsed melanoma and this scleroderma syndrome?

DR WEBER: So he had multiple melanomas, relapsed, rendered free of disease and finally relapsed in his chest wall with multiple sub-q nodules, multiple lymph nodes. Obviously could not be rendered free of disease surgically. He comes to me, and only relatively recently, over 6 to 8 months, had had the scleroderma. And I spoke to the rheumatologist and I said, “Give him the least immunosuppressive influence you can,” and she gave him hydroxychloroquine, which has modest immune-suppressive activity but can help with disease like scleroderma.

DR LOVE: How was the scleroderma manifest?

DR WEBER: Oh, his skin on his arms, his hands and his shins was like leather. And his feet — it was painful to walk.

DR LOVE: Have you seen this before with melanoma?

DR WEBER: I have never seen a case of scleroderma like that.

DR LOVE: Wow!

DR WEBER: It is not an association with melanoma. That I can say for sure. Hydroxychloroquine can control the disease. It’s a pretty modest immune suppressant. I gave him the pembro, and he slowly had his scleroderma get worse and worse and worse. And the skin is so fibrotic on the knuckles, the skin breaks, so he has to put Band-Aids^® over his knuckles. It spreads to the esophagus. It doesn’t have motility that’s as good as it used to be. So it’s hard to swallow large amounts of food. You have to really cut your food up.

DR LOVE: And just to mention, I see he’s BRAF wild type.

DR WEBER: Yes, and he’s BRAF wild type. It would have been so easy to give him BRAF/MEK, obviously. I had hoped he was BRAF mutated, but he was wild type, of course.

So with some reluctance, I gave him pembro every 3 weeks. And he got out to 6 months, and I think his tumors completely regressed. The current scan shows no obvious evidence of disease. No evidence on exam. But his scleroderma slowly got worse. His legs are swollen. Lower extremities are swollen. He has to wear different types of shoes. He’s a schoolteacher and still goes to work every day to detach, but it’s hard to stay on his feet because his feet hurt. I said, “Look, this is really affecting your quality of life. You’ve no obvious disease. Let’s quit and give you 12 weeks off, a drug holiday. Let’s put you on moderate-dose steroids. Keep up the hydroxychloroquine.” And his doc added mycophenolic acid. I’m going to see if we can make is quality of life better off pembrolizumab. But boy, what am I going to do when he relapses?

DR LOVE: First of all, it sounds like this is not a paraneoplastic syndrome.

DR WEBER: No, I don’t think it is.

DR LOVE: So he just probably has scleroderma.

DR WEBER: I think he has scleroderma. If it was paraneoplastic, it would have gotten better with tumor regression.

Use of immune checkpoint inhibitors in patients with preexisting autoimmune diseases

DR LOVE: I mean, I was going to actually ask you what was new in terms of giving checkpoint inhibitors to people with autoimmune disease. But specifically, what do we know about scleroderma? I’m not sure I’ve seen anything on that.

DR WEBER: First of all, it’s not that common a disease. Second of all, the only experience that’s published is anecdotal, plus a couple of small series, one of which was published a year or so ago by Alex Menezes, who’s from Australia. Good, young guy. And Menezes probably treated around 50-some-odd patients who had various types of autoimmune diseases. And, actually, of the 56, only 3 or 4 of them flared so badly he had to stop the PD-1 inhibitor or ipilimumab. He tried both.

Ipi was worse. Caused more problems. Pembro or nivo was better, which makes sense. They’re more directed.

And the bottom line is, you can actually, amazingly, get away without killing somebody who has an autoimmune disease by using PD-1 inhibitors. I sure as heck would not give ipi/nivo to someone like that.

DR LOVE: But what about scleroderma specifically? Does that flare when you give a checkpoint inhibitor?

DR WEBER: In this guy, I think he slowly got worse.

DR LOVE: But what about in this other series? Did they have any scleroderma patients?

DR WEBER: I don’t think he had scleroderma patients who flared. He did have scleroderma patients, and he thought that they were quiescent.

DR LOVE: So we actually had a case come up in lung cancer where the patient had a long history of ALS, and they wanted to know, could you give a checkpoint inhibitor? Because I had heard previously that people were very, very nervous about giving it to MS. So what about those 2 entities?

DR WEBER: The only time I’ve ever treated someone with MS with immunotherapy was high-dose IL-2. And, amazingly, I got away with it. But they did not have severe MS. Someone who has progressive, not relapsing or remitting but progressive severe MS, I think you could kill them by giving them serious immunotherapy.

DR LOVE: What about ALS?

DR WEBER: Boy, that’s even tougher. It depends on how sick they are.

DR LOVE: But, I mean, is that immunologic?

DR WEBER: We assume it probably is.

DR LOVE: Is that right? Is the thinking ALS is immunologic?

DR WEBER: It may be. And you would have to really push me to give someone with ALS immunotherapy. But if they did not have severe ALS, meaning they weren’t at the point of being paralyzed, having breathing issues, I would gingerly — and they had no other options and they were going to die of melanoma, I would try pembrolizumab or nivolumab. I wouldn’t give the combo.

DR LOVE: That’s kind of the thing that I’ve heard a lot about in a patient — we’re not talking about adjuvant situation, but in a patient with metastatic incurable disease, regardless, that it’s a consideration. But I do hear that people, whether it’s Crohn’s disease or whatever, that people are okay if they’re not getting biologic therapy or aggressive therapy or they’re stable, but they start getting very uncomfortable if they require more extensive therapy. Is that kind of the way you approach it?

DR WEBER: Yes, I would agree, Neil. And if I see a patient with active autoimmune disease, I cross my fingers and toes and hope that they’re BRAF mutated so I don’t have to think about giving them immunotherapy up front.

I just had someone with active colitis. I had him come off steroids and go on vatelizumab, which is the alpha4beta7 integrin antibody, which can control colitis without really being grossly immunosuppressive. And then I put him on BRAF/MEK. Thank God he was mutated. But if he fails BRAF/MEK, I’m going to have to take the plunge and give him pembro or nivo. That’s going to be tough.

Effects of novel therapies on the long-term outcomes of patients with metastatic melanoma

DR FLAHERTY: This paper, and one that preceded it a couple of years before, are basically tracking all of the definitive trials — targeted therapy trials and immune checkpoint trials — year by year as we get more published data from each of those studies, trying to understand kind of short, intermediate and then now what in melanoma we’re referring to as long-term benefit. So it’s kind of the 3- to 5-year time frame, let’s say.

So as I think you know, Neil, the kind of preconception dating back from even the cytokine era, but certainly ipilimumab, the first immune checkpoint, was that basically, there might not be very many patients who benefit, but that those who do do it in a very durable way.

So as the PD-1 antibodies came forward with a significantly higher response rate, so up-front benefit, 40% response rate in the front-line setting versus 10% with ipilimumab, the hope was that intermediate and then particularly long-term benefit would kind of track and that we’d see a tail of the curve that was ever higher. And so that’s been the importance of following long-term data on the immunotherapy side.

On the targeted therapy side, of course in other cancer types where targeted therapy has been introduced years longer than in melanoma, we know that there are both patients who can develop resistance in the early course of treatment and then some who can develop it over longer time frames. But in melanoma, specifically in the setting of BRAF inhibitor therapy and, admittedly, a kind of genetically complex tumor compared to, like, CML, for example, or gastrointestinal stromal tumor, which are genetically less complex, what does our long-term benefit profile look like? This was something that really people didn’t have a lot of confidence about a few years ago, with the tail of the curve — if there even was a tail of the curve — what it might look like.

So that’s been the undertaking of these kind of cross-study meta-analyses, to understand what we already knew, which is that the targeted therapies have a very reliable early impact, and do they maintain that in anybody? And if so, who are those patients? And then on the immunotherapy side, do PD-1 antibodies follow the ipilimumab precedent in terms of maintaining their benefit over longer and longer follow-up time?

So that was the objective. The obvious caveat before jumping to conclusions is, the cross-trial comparisons are hazardous. And I don’t need to tell you that. But just to warn people as they digest this type of data, whether it’s in this paper or they just look at the independent data set, you can try to adjust for patient characteristics like serum LDH and other measures of disease burden and aggressiveness, but the trial populations will never be the same.

And so having said that, what we’ve seen in the early data now followed out for several years, but still more follow-up to be done, is that sure enough, the targeted therapies show their overall survival benefit looks like more protection, if you will, against melanoma fatality in the 1- to 2-year time frame. And then you keep tracking over more time, and the immunotherapies, the curves cross, if you will, out around 2, 3 years of follow-up, where immunotherapies are now holding on to the benefit they were able to produce in a larger proportion of patients. And so now it’s where it’s getting really interesting. Because with more follow-up out through 3, 4 years, we’re starting to see this slight separation — and people are wondering, obviously, are we going to see with 5-, 7-year follow-up, very clear fraction of patients with immunotherapy who are maintaining disease control, survival. But we haven’t gotten there yet. It’s an interesting kind of maturation of the data so far, but not yet a completed story.

DR LOVE: I’m just kind of curious in terms of long-term survivors whether it’s immunotherapy or targeted therapy, how often is this in the presence of actual disease that you can see? Somebody has a PR or stable disease, but they are alive 5 years later. And how much of this is CR?

DR FLAHERTY: Yes. Great question, and that’s at the heart, not of this paper but of the data that’s emerged around this paper — absolutely clear difference in complete responders versus partial responders. So as you keep going out 3, 4 years, and for where we have 5-year data, those patients who are still alive and well then, many, many of them had complete responses early in the course of therapy. And you’re absolutely right that if you’ve got credible disease burden remaining, and by credible I mean not little kind of shreds where there were tumors that separate someone from a complete response, but FDG-avid tumor nodules that remain, those patients are absolutely the ones who are at risk for losing control over time. And that’s equally true on targeted therapy and immunotherapy.

Survival of patients with metastatic melanoma who receive immunotherapy compared to targeted therapy

DR LOVE: So you, obviously, are going through all the data. You have tremendous clinical experience. Let me ask you kind of your gestalt right now that you think about when you see a patient who, let’s say, is presenting for initial treatment of metastatic disease — typical melanoma patient in terms of site of mets, et cetera — if you think in your own mind — BRAF-positive, you think in your own mind the likelihood that this patient is going to be alive and well at 5 years, for targeted therapy, so BRAF/MEK combination. Immunotherapy. I’m curious about your thoughts in terms of single-agent PD-1 versus combination. And what global numbers in your mind do you carry for these 3 strategies in terms of risk of being alive and well in 5 years?

DR FLAHERTY: Yes. I mean, I’ll tell you, the take-home number is emerging at about 30% to 40% likelihood of 5-year survival. That’s based on small amounts of data, and we have behind that, even out to 4 years now, Phase III cohorts. So if you look at, kind of, Phase II cohorts, we’ve got maturity of data that would suggest that we may even be hitting close to that 40% number. But, of course, smaller trials will always have more patient selection — you may be enriching for better patients, if you will. So we need to see those Phase III cohorts come out. But it’s not going to be less than 30%.

DR LOVE: But when you say that number, is that for one of these modalities individually, or using them all together?

DR FLAHERTY: Okay. Great — fair point. I’m starting with just single modality. And in our practice, for most patients — and as you were posing it for BRAF mutations — it’s BRAF/MEK versus PD-1 monotherapy as the most commonly pursued choices — we’ll come back to PD-1 CTLA-4 combination therapy. But anyway, no, you’re right. The data we’ve got really doesn’t have patients — those clinical trial participants having access to all the therapies. And so I often tell patients, based just on our center data, that we think we’re in that 40-plus-percent range if we leverage the entire toolbox.

DR LOVE: But, now, when you say 30% to 40%, are you saying both for immunotherapy and targeted therapy, that kind of a number?

DR FLAHERTY: That’s right.

DR LOVE: Really? That’s amazing, actually, if you think about it.

DR FLAHERTY: Yes. I mean, this is — again, remember, we’re talking about overall survival. And you look, with either the immune checkpoints or the targeted therapies, and you’ll see progression-free survival duck down to, let’s say, 25% out to 4 years’ follow-up. But there are still patients alive and well probably because they’ve crossed over and gotten access to effective salvage therapy.

DR LOVE: So, but you’re saying 30% to 40% long-term survival just with targeted therapy?

DR FLAHERTY: That’s right. You have to remember, I mean, and this is in that paper to a degree as well — that it all depends on the patient’s baseline characteristics. True for targeted therapy, true for immunotherapy. So when those trials were conducted, these were attractive options, and patients with all measure of disease burden — low disease burden, intermediate, high — were going on those studies.

So it’s the patients with low disease burden, not surprisingly, who are the ones who are much more likely to have complete responses and much more likely to be alive and well 3, 4 years and counting. If you only offer these therapies to patients with high disease burden, then all of those numbers start to decay in terms of likelihood of long-term survivorship.

So the data is very clear on that point. And those trials, the targeted and immune trials, accrued a healthy fraction of patients who had low disease burden at baseline. So if you include those patients in the mix, that’s what drives the 3-, 4- or 5-year benefit rate.

DR LOVE: I’ve got to say that I have had this impression, and I don’t know whether others have, that the likelihood of long-term disease control was a lot greater with immunotherapy than targeted therapy. Do you think that a lot of people think that way?

DR FLAHERTY: Yes, absolutely. And it’s all in advance of the data. In other words, you absolutely, you’d see talks given with hypothetical Kaplan-Meier curves drawn that made exactly that point. You’ll still see those representations in advance of us having the evidence. And then year by year, as the evidence has been produced, it continues to confound that. You see this tail of the curve emerging on both, where you see some decay on the PD-1 antibody side that people were hoping and thinking wouldn’t happen, and you see this persistence on the targeted therapy side that people were thinking didn’t occur.

Selection of targeted agents versus immunotherapy in the front-line setting for patients with BRAF mutation-positive melanoma

DR LOVE: So that leads to the, seems like, age-old question, but I guess it really hadn’t been around that long, at least it hadn’t been this interesting for that long, of targeted therapy versus immunotherapy first line. So I’m kind of — in the BRAF-positive patients. And I am kind of curious where you stand today on that, both in the patient who you don’t need a quick response in as well as the patient where you really want to see them respond.

DR FLAHERTY: Right. So it’s the first category that’s the real nail biter right now. I say, “nail biter” because you’re right, you don’t need a response in the moment, but it’s just that the data would tell you, you’ve got 2 good options. And I would tell you that we’re still trying to unpack the molecular features of those who derive long-term benefit from immunotherapy versus targeted therapy. And there’s some hope there, but nothing definitive, that we may be able to subset — in other words, we may be able to define those, in the BRAF-mutant population, those tumors that have low mutation burden, who, basically, are the ones who are disproportionately in the running for long-term benefit from targeted therapy, and those patients — again, still BRAF mutant — who have high mutation burden, who are the ones who are deriving long-term benefit from immunotherapy.

So we’ve got to continue to develop that data. It’s not going to require big randomized trials, necessarily, but just big cohort analyses where we kind of go deeper than just BRAF itself to understand who are the patients who get long-term benefit from either.

Effect of PD-L1 expression on response to immune checkpoint inhibitors

DR LOVE: What about PD-L1 levels?

DR FLAHERTY: Yes. So in melanoma, we don’t test for it. It’s not part of the FDA approvals, and there is no standard test. But there’s the availability of commercial assays, admittedly. As in every tumor type that you’ve deliberated on where PD-1 antibodies are approved, yes, there’s an association between higher response rate with PD-L1-positive, and the higher the expression level, the higher the response rate. But the issue in the field has been that melanoma, perhaps more than other tumors, has an appreciable response rate, even in the PD-L1-negatives. And that’s what’s precluded people from embracing the testing. But I think you’re getting at the point that what about a BRAF-mutant patient?

So you know about the BRAF status. Would it change your thinking if you knew that they had a, let’s say, a 50% likelihood of response versus a 10% to 15% likely response based on PD-L1 testing? And that data exists to discriminate, exactly as you’re suggesting. The problem is, and this is the point I always make to patients, that data, whether it’s with pembrolizumab or nivolumab, comes from clinical trials that used a specific antibody and specific methods for calling positive, negative and establishing a cut-point. We don’t have those tests available in practice. I mean not just at the hospital level. I mean even at a commercial centralized lab. You can’t order the antibody and use the cut-point and the method that was used in the trial population to apply to clinical practice. You’re using a third-party antibody, and you’re getting a report back that doesn’t even tell you what they’re scoring in terms of percent positive and what their methodology is.

DR LOVE: I want to get back to this long-term thing, but just to divert out a little bit since we’ve talked about PD-L1 levels. The one situation that, I mean, for several years, seems like you would want to be able to use it, although I don’t really hear people talking about it, is trying to decide about whether you’re going to use combination immunotherapy or single agent.

DR FLAHERTY: Yes.

Comparison of the efficacy and safety of combination therapy versus monotherapy with immune checkpoint inhibitors

DR LOVE: It kind of looks like the people with high PD-L1 levels don’t really need the second drug. But is that actually the way it works in practice?

DR FLAHERTY: Yes. No, that’s — I mean, so that’s the data. And again, this is this problem of not routinely using PD-L1 testing in practice confounds your ability to leverage that data. So if you’re a practitioner who follows the script that I was just kind of laying out there and saying, “Okay, look. We don’t have a credible assay that relates to the data produced in these trials, then I’m kind of stuck,” which is, frankly, my view, at least from the most evidence-based perspective.

So just revisit the data real quick, which is, we have about a 15% to 20%, let’s say 18% based on the published data, response rate difference in favor of PD-1/CTLA-4 combination, nivolumab/ipilimumab specifically in melanoma. So you do get a bump, 40% to 58%, in response rate. You get a progression-free survival difference that’s about 10% to 12%, depending on what time point you’re looking at, as you track the population over time. There’s a PFS difference.

What we only learned about in 2017, first at AARC and then at ESMO, with definitive overall survival data now published, is that there’s a much more modest impact on overall survival, 5% delta absolute at 2 years, 6% at 3 years. So that gives you kind of a numerical picture of what that looks like. So that’s what we’re wrestling with in terms of the data.

Now, the apologies that are made are that that trial was not designed as an overall survival primary endpoint to make that comparison. But look, in clinical practice, this is what we’re trying to decide between is, combination versus monotherapy.

And so most people look at that overall data and say, if PD-L1 testing is not part of my routine practice, then I’ve got to have a patient in front of me who’s willing to put up with the risk of significant toxicity for the combination for that type of delta in terms of overall survival. And if this is a low disease burden patient that we’re talking about, which was kind of scenario number 1 that we’re considering, we now know that those patients can be taken through a sequence of PD-1 monotherapy and salvage, with still a 10% response rate with ipilimumab second line. We didn’t know that 2 years ago, that ipilimumab retained its efficacy as a salvage therapy. And I admit the 10% response rate doesn’t sound like much, but remember, that’s the profile of the drug that got it approved in the first place in 2011.

So a low disease burden patient, in my view, with confidence, can be counseled about a sequential approach, PD-1/CTLA-4. If they have a BRAF mutation, of course you’d favor BRAF/MEK combination therapy before you ever went to a 10% response rate jog like ipilimumab. But this is where clinical practice wrestles right now with kind of how to think about sequencing targeted immuno. And again, this ipi/nivo combination data is so heavily weighing on this decision with overall survival looking — if it’s real, if it’s even statistically real it’s a pretty slight difference, combination versus monotherapy, but not a slight difference in terms of toxicity.

DR LOVE: Just to be clear, though, in terms of the sequential thing. When you give the ipi second line, are you also adding in a PD-1 inhibitor, or just —

DR FLAHERTY: No. So that’s — no, not routinely. That was a clinical trial concept that got kicked around a couple of years ago, the idea that maybe you need to keep the PD-1 around to be able to leverage the mechanism. We have only Phase II level evidence that’s now been published though, that indicates that actually you don’t do better than that 10%, 15% response rate with ipi/nivo combo as a salvage therapy. And with ipi itself having 10% response rate now documented in a couple of large cohorts, I’d say that basically you’re still tempting toxicity with that approach. And without more promising evidence than what I’ve just summarized, ipi as a monotherapy is the salvage.

Duration of immunotherapy and targeted therapies to achieve optimal clinical benefit

DR LOVE: And another related issue in these long-term survivors is duration of treatment. How do you approach that, both in terms of immunotherapy as well as targeted therapy?

DR FLAHERTY: Right. So let me give you the quick bullets. So with targeted therapy, it’s indefinite. In other words, you just keep going, even in complete responders. And I’d say that at least until we know more. But what we are seeing, and even at Congress, as we’ve seen some data presented regarding complete responding patients who discontinued, and there are definitely are examples of patients who only after discontinuing had reemergence of disease. We don’t know what the percentage of that is, but I mean — truly. But the fact that it can happen reinforces the idea that these patients probably need to stay on therapy long term.

On the PD-1 side, PD-1 monotherapy, when it was FDA approved was an indefinite duration of treatment in melanoma. No defined duration. But the clinical trials that established the benefit of those drugs most commonly used a 2-year duration and then stopped, regardless of what was going on in the scans and so on. And so in the centers that conducted those trials, we’ve got comfort with that approach. And many people are starting to dial that back and even think about stopping earlier, all depending on the patient’s response. If it’s a complete or near-complete response within, let’s say, the first 6 or 12 months of treatment, then think about truncating even short of 2 years. There’s no consensus on this point. This is really a hot topic right now in the melanoma field, because it drives a lot of drug cost and patient inconvenience to keep going into this second year. But going beyond that, I would say, we never had evidence that going longer periods of time was important.

Correlation between tumor mutational burden and response to immune checkpoint inhibitors

DR LOVE: So I’m just going to take a moment and ask you more macro questions that go across tumor lines. As you mentioned, we were in your city, Boston, doing a multitumor thing, and we had your colleague Lecia Sequist there talking about lung cancer. And we were also talking about BRAF-positive disease in colorectal cancer. And those things are kind of buzzing around in my mind as I talk to you. And one of the things that’s kind of interesting, and Lecia’s really been talking about this a lot for the last year, the thing that they see in lung cancer, which is that people with targetable mutations don’t do well on immunotherapy, and yet that’s really not the case. Right? The patient with BRAF is just as likely in melanoma to respond to a checkpoint inhibitor as BRAF-negative. Is that right?

DR FLAHERTY: Absolutely right. Yep.

DR LOVE: I’m just kind of curious what your thoughts are about this, particularly in lung cancer, where they’re really delaying checkpoint therapy.

DR FLAHERTY: Yes.

DR LOVE: And yet in melanoma, it’s just a totally different ballgame.

DR FLAHERTY: Yes.

DR LOVE: What do you think is going on biologically?

DR FLAHERTY: Yes. I think the evidence is there, actually, to explain it. So remember I was commenting before about tumor mutation burden and this kind of genetic complexity concept? We’ve got a bunch of retrospective data that tells you that the tumors that are driving high PD-1 response rates are the ones that have the highest mutation burden. Now, ALK translocated non-small cell lung cancer and ROS translocated non-small cell lung cancer are probably the extreme examples of really quite genetically simple tumors. EGFR mutant I would maybe kind of put adjacent to those. But similarly, these are not high mutation burden tumors. Remember, these are mostly nonsmokers that have those genotypes in non-small cell lung cancer. So hold those for a moment. Come over to melanoma.

Almost all of melanoma is at a higher mutation burden level at baseline than those subsets of non-small cell lung cancer. It maps over to smokers, non-small cell lung cancer, in terms of mutation burden. So we have our UV-associated melanomas in smoking non-small cell, bladder, these kind of high DNA damage-related cancers. That’s where the through line seems to be in terms of where there’s consistent and pretty high response rate to PD-1 antibodies. But if you go fishing down into the low mutation burden tumors, that’s where it starts to weaken.

So I was suggesting before, in response to one of your questions about how do we break this tie between for BRAF-mutant patients targeted therapy versus immune therapy, we know from the published data that there’s a whole spectrum in that population of some younger patients, particularly with quite low mutation burden, and then others, particularly the older folks who have quite a high mutation burden in association with their BRAF mutation. And so I suspect that in melanoma we’re actually going to start to be able to discriminate, maybe not as starkly as in non-small cell cancer where they’ve got EGFR, ALK, ROS and they kind of triage those patients up front to targeted therapy and then reconsider immunotherapy later, in melanoma I think we’re going to have a group for whom we’re going to have even more confidence assigning them to BRAF/MEK therapy if they’re low mutation burden and then PD-1-based therapy if they’re high. But we’ve got more work to do on this topic.

There’s prospective trials that are being mounted now across tumor types, in fact, to really try to set a mutation burden cut-point on a couple of hundred, few hundred gene NGS test, the type of test that you can get now from a handful of commercial laboratories as well as academic labs.

DR LOVE: That’s interesting. Any explanation — I don’t know how well it’s documented in the literature, but I’ve heard a couple of investigators — I know Alice Shaw at your place has done a lot of work on ALK — talk about patients who have very high PD-1 levels with ALK but yet don’t respond to checkpoint inhibitors. What do you think is going on there?

DR FLAHERTY: Yes, I know. So I think to this issue about measuring tumor inflammation is the other side of the coin, right? So I’m talking about mutation burden as kind of like adjacent to the kind of testing of the genetic alterations themselves. But you’re absolutely right, that the flipside of this is, how much immune engagement is there at baseline in any of these tumors? And how much does that tell us about the people who are going to respond or not? And you’re absolutely right, that there are definitely patients who have baseline tumor inflammations — they’ve had T cells in there, they’re secreting interferon, that’s driving PD-L1 expression on the tumor cells, and yet they are not 1 drug away, if you will, of single-agent PD-1/PD-L1 antibody away from responding. They’ve got other layers of immunosuppression. Different than the folks, particularly those with high mutation burden, who also have baseline tumor inflammation, PD-L1 expressions on the tumor, who are 1 drug away.

But remember, even in the tumors that we call baseline, highly inflamed and high mutation burden, with single-agent PD-1 antibodies, we’re still not cracking above 50% response rate, right? So MSI-high colon, when you were talking about colorectal cancer and the BRAF-mutant subpopulation of colorectal cancer, those are mostly MSI-high cases, the BRAF-mutant, right-sided colon cancer cases. So those are in the running for response to PD-1 antibodies. But even in that population, it’s still a 50% response rate.

So it’s telling you that even in the tumors that are pre-wired with all of these mutations, what we think of as mutated neoantigens, lots of baseline T-cell recognition in the tumor, that still there must be other mechanisms that are protecting those tumors and basically mediating de novo resistance to PD-1 antibodies.

So, of course, this is the hottest area in all of oncology drug development, is trying to come up with strategies to be able to elevate that response rate in those baseline T-cell inflamed tumors. And there are people who obviously have been putting a lot of time and effort into next immune therapies to try to crack that nut. But I think this discussion points to the fact that maybe there’s actually targeted immuno combos that are going to be impactful in terms of trying to be able to potentiate a response in these tumors.

Activity and tolerability of BRAF/MEK inhibitors in combination with anti-PD-1/PD-L1 antibodies

DR FLAHERTY: In the melanoma case, that’s the most popular clinical trial category right now, are BRAF/MEK plus PD-1/PD-L1 antibodies. There’s 3 Phase III trials now ongoing. And you could say, “Look, they are all active drugs, why not put them together?” But I think the story’s actually a bit more biologically and mechanistically based, that these approaches might actually really potentiate each other. And we’ve been doing Phase I/II trials of this sort for the past several years, and the results are obviously preliminary but certainly look promising.

DR LOVE: Yes. I saw that when I was looking through your papers, and I was going to ask you about that. Again, it’s a different disease, different drugs, but I know they ran into major problems when they combined EGFR TKIs with checkpoint inhibitors. Why did they see that? And I guess you’re not seeing it.

DR FLAHERTY: Yes. Pneumonitis was the thing that came up there. And I think that may be somewhat of a different situation in terms of combination talks. Now, I should remind you, though, that in melanoma, interestingly, with the very same PD-1 antibodies and the same dose, we see very little pneumonitis. I mean — and very rarely do we see clinical significant pneumonitis. And yet in lung cancer patients, clearly more.

So is this a disease state thing? Is it a smoking history thing? When you have the exact same drugs producing a different toxicity profile like that, you wonder what accounts for it. But anyway —

DR LOVE: You also have different host immune system, because they’ve — old people versus younger people.

DR FLAHERTY: Precisely. So there’s a number of factors to consider. But suffice it to say, we don’t see that problem when we try to combine targeted immune therapies in melanoma. But we’re using BRAF and MEK inhibitors. And maybe there the issues are just going to be fundamentally different.

So we’ve had our eye on liver tox as being the main issue, because both classes of therapies can cause transaminase elevations. And those were the things that really kind of showed up in the early trials, if there was anything to really catch one’s eye in terms of combination toxicity. But it hasn’t been a showstopper. And like I said, multiple combination regimens are now moving into Phase III.

Association between the gut microbiome and response to anti-PD-1 antibody-based therapy in metastatic melanoma

DR LOVE: So I had the pleasure of interviewing your colleague Jeff Weber, and there are a few things that he said to me that are still spinning around in my brain that I want to run by you. A couple of them were very practical clinically, but one that I’ve been really starting to hear about and I think is really fascinating, again, just sort of globally, always interested in picking your brain, is this so-called microbiome. And he was telling me about it, and I know there was some stuff presented at ASCO, and particularly as it relates to checkpoint inhibitors. Can you talk about kind of what your view is of this and where you think it’s heading?

DR FLAHERTY: Yes. I mean, obviously microbiome has been considered known as potentially having a lot to do with how cancers form in the first place, potentially responsiveness to various kinds of therapy. And I’m interested in all that. But I would tell you that the immunotherapy setting, in my mind, has got to be the area in which microbiome would have the most influence.

DR LOVE: It seems like it.

DR FLAHERTY: Because, I mean, put it this way, the branch of the immune system that we are modulating with these drugs is right at the borderline, if you will, of the gut, where microbial infiltration is an all-day, everyday reality, right? So we’re defending ourselves against pathogens in the gut as just part of our everyday business. And the immunosuppressive mechanisms that exist to protect us from an overly exuberant immune attack there are, again, just part of normal biology.

So when we come in with, in this case, the immune checkpoints, and that we see colitis as the most common life-threatening consequence. I think you shouldn’t be shocked to see that given that, again, this same branch of the immune system is so active at the mucosal boundary. So — but your point about microbiome, if you alter the microbiome, so now you’re altering that kind of host defense composition or its state of kind of alarm and attack, I think the likelihood that that has ramifications in terms of what the immune system can then do or not do in terms of mounting an antitumor effect to me is very logical. Now, having said that, the data that’s being generated in these early studies is basically just trying to decode the entire universe of the microbiome in the gut, which is just inordinate numbers of organisms that exist there. To me, this feels 1 step away from whole genome sequencing in the sense that you probably have to look at large populations to be able to try to understand whether there’s a subgroup that has a different microbiome makeup that would really powerfully influence likelihood of response or long-term benefit from these therapies.

But I think it’s really — it is a very logical thing to me that microbiome, how it relates to immune defenses at the gut boundary and then the systemic effect of these drugs are very logical steps. But we’re going to have to nail down beyond phenomenology, real mechanism that would actually support changing it, right? I mean actually thinking about giving either supplements to alter the microbiome, potentially even antibiotics to mold it, if you will, in a desired way. And we’re a long way away from that. We don’t even yet know what the microbiome phenotype is that we would be trying to create. But that’s what the early work will aim to demonstrate.

DR LOVE: That’s really fascinating. I never thought about this idea of the immune system in the gut dealing with the bacteria.

DR FLAHERTY: And liver right behind that. I mean, this is, again, the biliary tree deals with bacteria infiltration all day every day. It’s the second most active site of host defense, if you will, against the commensal organisms that we live with. And again, that’s the second most important site in terms of autoimmune toxicity when we give a PD-1 antibody — is the liver. And again, these mechanisms are part of tamping down these branches of the immune system. And perhaps then not surprising that all of these dots potentially connect.

So understanding gut microbiomes is going to be critical. There’s others that are starting to interrogate skin microbiome, try to understand if there’s some relevance there, at least in terms of having relevance to cutaneous toxicity. Which, of course, is even more common than gut and liver, usually not life threatening, obviously, but still a very common feature of these checkpoint antibodies.

DR LOVE: Hmm, that’s interesting, the viewed toxicity within the idea of this normal physiology that’s going on.

DR FLAHERTY: Yes.

Effect of disease burden and type of response on outcomes of patients with metastatic melanoma

DR LOVE: Anyhow, getting back to practical decisions. So I take it that if you see a patient with BRAF-positive metastatic disease in the first line, and they say to you, “What’s the strategy that’s going to mostly likely have me alive and well in 5 years?” you’re going to say, kind of, it’s a coin flip?

DR FLAHERTY: Right. It is. And I tell them further, this is the low disease burden patient, and I tell them further, I’m going to watch you like a hawk through the first few months of therapy. And I’m going to do scans at least every 2 months. Because how you respond — we’re going to back to complete responders doing well and less than complete response not doing well — is very much going to influence my thinking about how long we stick with treatment choice number 1. Basically, I don’t want to wait around and get burned by the acquisition of resistance over time in a patient who’s having a really suboptimal response to either approach. This is, I think, a key point about the long-term follow-up data, is that the progressors are coming from those who had a middle of the road partial response but where there’s still definitely disease burden remaining.

The high disease burden patient you asked about a while ago, and we didn’t talk much about them, but unfortunately, these are the patients who just don’t get long-term benefit with any reliability to either targeted therapy or immunotherapy. So patients who walk in the door with symptoms, and then you measure the serum LDH and it’s elevated or particularly markedly elevated, like more than 2 times upper limit of normal, if you look at the data from these long-term follow-up analyses, they are just not in the running for a more than 2-year survival outcome. So if they have a BRAF mutation, sure, we think about BRAF inhibitors for their ability to induce disease control, symptom benefit and so on. And it’s not that immune therapies can never have an early effect in those patients. They can, but long-term benefit is just, unfortunately, exceedingly unlikely. And that’s true even with PD-1/CTLA-4 combination, which I’ll tell you is, if there’s a population in whom that approach is favored, it’s those high disease burden patients.

So if you have patient in front of you who doesn’t have a BRAF mutation and you’re purely weighing PD-1 monotherapy versus PD-1/CTLA-4 combo, if they have high disease burden, that is oftentimes the driver of telling them, look, there’s more toxicity associated with this regimen, but we’re not going to have the benefit of being able to get you through 2 lines of therapy. If we’re going to go for it, we’re going to go for it up front with both drugs.

DR LOVE: That’s really interesting. Roughly, globally, in clinical practice — I don’t know, maybe there have been surveys done — but what fraction of patients in first-line metastatic disease fit into this high tumor burden —

DR FLAHERTY: Yes. So yes. So there are survey data. I think they do vary between the referral centers who always see kind of a selected set versus community practice. In our practice, I would say, no fewer than 30% of patients meet the definition I’m describing of high disease burden at baseline. You can look at some survey data and talk to community docs, and they’ll tell you it’s probably closer to 50% of patients who walk in the door by the same rough measures that classifies high burden.

So it’s a nontrivial thing. And remember, in melanoma, before we had these drugs 8 years ago, the average survival time was 6 to 8 months, even in clinical trial populations, right? So this is an overall aggressive disease. And the patients walk in the door with symptoms and elevations of LDH. To me, it’s amazing that we actually can get a benefit of even 1 or 2 years’ duration with the available treatments. But these are not the long-term surviving patients.

Potential implications of results of the COMBI-AD and CheckMate 238 trials for adjuvant decision-making for patients with resected Stage III/IV melanoma

DR LOVE: So when I sat down to do the interview with Jeff, I kind of looked at what’s been going on recently, and I looked at it and it looked like ESMO was, like, huge, huge, huge, in terms of adjuvant therapy. I see these two trials, one was Jeff’s trial of nivo and the other was the targeted trial with the dabrafenib/trametinib. And I was like, “Whoa, I guess things are different now.” Is that the case?

DR FLAHERTY: It was different the next day. I mean, we’ve had all this discussion about the metastatic setting. And now, the revolution is happening in the adjuvant setting. And it’s, of course, going to have ramifications for everything we just talked about in the metastatic setting. Because now we know that’s — back up a step. We already knew that ipilimumab improved relapse-free survival and overall survival in a way that we had never seen historically, with interferon or any other approach. That was 3 year ago data.

But now we see nivolumab clearly beating ipilimumab. I mean, just hands down on efficacy and on safety. So right. So basically just — that’s it. Now for the BRAF wild-type patient. Very easy. You say nivolumab is the new standard. And literally, I would say homogenous opinion, coming right out of ESMO, that the next patient you saw in clinic who had the same Stage III or even resected Stage IV burden of disease, so that match the clinical trial criteria, you’d confidently tell them that’s their situation.

On the BRAF side, the first result we see is BRAF/MEK versus placebo. There was in parallel, by the way, of vemurafenib, a BRAF inhibitor monotherapy versus placebo trial, which captured less news because it didn’t have the same impact and, in fact, missed its primary endpoint, whereas BRAF/MEK versus placebo, big impact in terms of improvement in relapse-free and overall survival even in this adjuvant setting.

So the complicated math that we’re doing in the field now for a BRAF-mutant patient who walks in the door with Stage III or even resected Stage IV disease is, okay, we had ipi versus placebo. Now, nivolumab versus ipi. And you can kind of look at this step-wise improvements there and it gets you to, roughly, and just — I’m apologizing for the cross-trial comparisons and all this stuff — it gets you to a roughly 50% protection. Fifty percent in relapse-free survival. I mean, again, we’re talking about melanoma, where we were anguishing about interferon with its 10%, 15% relative reduction in risk. So 50% relapse-free survival benefit. Just more than 50% was shown in the dabrafenib/trametinib versus placebo study.

So I tell patients, look, we can’t be so precise. These data feel roughly of equal impact. So if you have a BRAF-mutant patient, my take-home from ESMO and every day since then is, you’ve got basically equal efficacy data for 1 year of nivolumab or 1 year of dabrafenib/trametinib in the adjuvant setting for these patients with resected Stage III/IV disease. Ipilimumab now basically has been displaced.

And so what’s going to be fascinating, as we now shift to doing this in clinical practice routinely is those patients who do relapse, what do we do for them? Because they’ve now seen these drugs in the adjuvant setting. And as you can imagine, we have essentially no data whatsoever regarding what strategy you should prefer. Do you always switch mechanistically in a BRAF-mutant patient depending on what your choice was in the adjuvant setting? I mean, that’s tempting to say yes. But we really are going to be entering a new territory where we’re going to have to rediscover what benefits are maintained for these therapies if a patient does relapse after adjuvant.

DR LOVE: Yes, I’m dying to put together a patient decision aid on this one.

DR FLAHERTY: Yes. Yes. Totally. I mean, but it’s amazing data, I’ll tell you.

DR LOVE: It is.

DR FLAHERTY: Just stare at the outcome data, relapse-free survival for — we have that for all the data sets. And we had this overall survival finding from the dabrafenib/trametinib trial. I mean, it really is better than expected on both sides, the immune checkpoint side and the BRAF/MEK side. And who would have thought that would happen?

Comparison of the mechanisms of action, activity and safety profiles of encorafenib/binimetinib, dabrafenib/trametinib and vemurafenib/cobimetinib for BRAF-mutant melanoma

DR LOVE: So a third targeted combination, encorafenib and binimetinib. That’s something that you were involved with, the so-called COLUMBUS trial. Any way to differentiate out the 3 doublets right now?

DR FLAHERTY: Yes. I mean, you can imagine I’m cautious on the cross-trial efficacy comparison. I mean, for what it’s worth, the encorafenib/binimetinib combo does produce the best progression-free survival result we’ve seen in the field, with a median of just under 15 months, and that’s compared to 11, 12 months for the other 2 approved regimens. That feels like a potential difference, but the response rate wasn’t different. It was, in fact, identical across the Phase III cohorts for each of the BRAF/MEK combos. So are we really looking at a difference, in this case, of duration of disease control and overall benefit?

DR LOVE: In terms of efficacy, any way to differentiate their mechanisms? Is this kind of like an aromatase inhibitor kind of switch thing, or more complicated?

DR FLAHERTY: Yes. I think it’s a little more complicated. I mean, we think it’s all about the kind of biochemistry of these drugs. The MEK inhibitors look really similar. So, actually, I don’t think there’s really a case to be made there. On the BRAF side, though, encorafenib was actually developed in the shadow of vemurafenib and dabrafenib. And what I mean by that is, there were actually preclinical studies done and now published head to head showing differences in potency and in overall specificity as well, for BRAF and kind of unique target engagement and pathway inhibition dynamics. I mean, I’ll be brief about it. But suffice it to say that this a drug that gets cleared from the plasma very quickly, and yet it has a very long on-target time. And that’s different than vemurafenib and dabrafenib. The head-to-head data preclinically make it look like it’s a better BRAF inhibitor.

Interestingly, when we tested it in Phase I, and we presented this data and now published as well, that BRAF inhibitor, encorafenib, actually looks like it has a worse tolerability profile than the other drugs. I mean, I could clarify that in great detail. But let me just skip right to the fact that when we gave a MEK inhibitor with that BRAF inhibitor, we saw an even bigger drop in toxicity. In other words, improvement in toxicity by adding a MEK inhibitor to that BRAF inhibitor than we had seen in years past with vemurafenib or dabrafenib.

So because BRAF/MEK is the relevant unit of therapy, combination therapy, established already definitely a few years ago, it’s of historical interest that that’s true, that encorafenib, if anything, is a tougher agent.

In combination, and this is where I’m coming to my kind of tiebreaking comment, what’s very, very clear is, encorafenib/binimetinib is a better tolerated BRAF/MEK combination than the other BRAF/MEK doublets. It basically drops the 2 major quality-of-life impacting toxicities. So the fevers and the complicated fevers with chills and lightheadedness and things that can happen with dabrafenib/trametinib do not occur with this combination. And then the photosensitivity that vemurafenib produces, which can be a month-by-month ongoing issue and really a quality-of-life limiting issue for patients also just disappears with this combination.

So some of the other BRAF/MEK class effect toxicities still exist. But if you look at the safety profile, this combination, it really does look different than these other regimens. So that that would be, as a practitioner, the reason to be interested in yet another BRAF/MEK combination.

DR LOVE: Yes, I remember you telling me about that a couple of years ago. And I’m curious. In terms of selection, I’m going to guess that if you’re going to use targeted therapy in the adjuvant situation, you’re only going to use dabrafenib/trametinib because that’s where the data is?

DR FLAHERTY: Yes. I think that’s right. I mean, obviously post-ESMO, it’s been, and before FDA approvals, we’re obviously petitioning on behalf of patients with payers to get coverage of this approach. And I think we’re not going to have much of a strong leg to stand on if we’re lobbying for something beyond dabrafenib/trametinib in the adjuvant setting. So I think you’re right on that point. It’s in the metastatic setting, obviously, where we’ll have the full menu of options and can choose, based on preferences, including tolerability.

Selection of first-line therapy for patients with BRAF-mutant metastatic melanoma

DR LOVE: That’s what I was going to ask you. Right now, first-line metastatic disease, what’s your usual nonprotocol combination?

DR FLAHERTY: So we’ve basically told patients and within our own group, we’ve said, “Look, pick your poison, dabrafenib/trametinib versus vemurafenib/cobimetinib in terms of tolerability. The efficacy data is almost superimposable. So just go, based on toxicity preference.” And that’s not a very strong statement in terms of picking between the two.

And we don’t have any reason to believe that there’s any ability to give patients salvage targeted therapy if they’ve failed one of these other regimens. So it really is pick your regimen and go with it. And immunotherapy, obviously, is the rest of the discussion.

Ongoing trials of MEK inhibitors with or without anti-PD-1/PD-L1 antibodies in NRAS-mutated melanoma

DR LOVE: So I know at one point there was a lot of interest in patients with NRAS mutations. And I think this combination was being looked at. And where did that end up?

DR FLAHERTY: Yes. So MEK inhibitor monotherapy was put through all the way through a Phase III trial, and NRAS-mutant patients, and produced a 15% response rate. Produced an improvement in progression-free survival. It was statistically significant, but it was just really modest. Right? We’re talking 1 and a half months’ improvement in progression-free survival in a disease where we’re now used to seeing 6- or 12-month improvements, even in medians. And so basically, while the trial was positive, it was data produced in a population who mostly hadn’t seen immune checkpoint therapy and that, of course, had become the standard along the way. So there was a lot of reluctance to embrace MEK inhibitor monotherapy as clearly a new treatment to bring into the armamentarium. But where there is almost as much enthusiasm around that approach is in the targeted immuno-oncology combo strategy. So in other words, I mentioned already BRAF/MEK/PD-1 or BRAF/MEK/PD-L1 triplets in Phase III, quietly to the side of them are a couple of Phase III trials with MEK inhibitor plus PD-1 or PD-L1 in the BRAF wild types, half of whom have NRAS mutations and half of whom have other drivers. But the point being that where MEK inhibitors may find their first real inroad in melanoma would be in combination with PD-1 and in the BRAF wild types.

DR LOVE: And again, thinking back to our GI session that we had, we had Johanna Bendell, and I know she at one point had looked at using a MEK inhibitor with a PD-1 for MSI-stable colon cancer. Is that kind of the same strategy?

DR FLAHERTY: Yes. In fact, that same trial included a melanoma cohort, again, BRAF wild types. And in the MSS colon data was the most clearly — there was a clear signal, if you will, because MEK inhibitors have so little activity there, and PD-1 antibodies also and the fact that the combination as producing a nearly 20% response rate with what looked like some really durable responses in that group, that was taken as being kind of clear validation of the concept.

In melanoma, as I already said, we already know that PD-1 antibodies have activity. We know that MEK inhibitors have some activity, even in the BRAF wild types. So the signal that was generated in the melanoma population looked actually just as good. It was 50% response rate of MEK/PD-1 combo. But is that clearly different than PD-1 monotherapy? We need a bigger data set, obviously. Because these drugs are active, you can imagine, again, these ongoing randomized trials, they are just accruing like gangbusters because patients who are enrolling on them are getting access to active therapy on whichever arm of the study they’re assigned.

Rationale for the investigation of immune checkpoint inhibitors in combination with MEK inhibitors

DR LOVE: I know I’ve heard Johanna explain what she thinks is the biology. But what’s your take on why a MEK inhibitor would improve response to a PD-1 agent?

DR FLAHERTY: Yes. So the data that’s been generated in patients, actually, not just in models, and including in our group where we’ve done some of this work, it’s pretty clear on the tumor cell side that you see MHC or antigen expression go up as a consequence of MEK inhibitors. You see immune cells in the microenvironment change. So you see that macrophages appear to be kind repolarized back to an M1 phenotype from an M2 phenotype. And you see more T-cell entry into the tumor as a consequence of just MEK inhibitor monotherapy.

So what’s the chicken and the egg of that? In other words, how much of this is a tumor effect that’s then having manifestations in terms of immune recognition? How much of it is purely immunologic? We don’t know. But all of those phenomena that I just described have been confirmed, not just in model system but in patients.

So to me, that’s enough reason to keep investigating. But on the back of these trials, let’s say we have positive Phase III trials for progression-free survival with combo versus monotherapy, are we really going to know who’s getting the benefit and why? And that’s going to take a bit more work.

Mechanism of action and activity of the cancer stemness inhibitor napabucasin

DR LOVE: So I want to get to your cases in a second, but a couple of other quick questions. One, I noticed that you were involved, I know you do Phase I work as well, but with the drug napabucasin, the stem cell inhibitor. And, of course, we’ve talked about that a lot in the GI cancers. Just kind of curious on your quick take about where that’s heading.

DR FLAHERTY: Yes. So that concept, I think, is alive and well, that there’s a population of cells that survive our best available therapies. And these cells have stem cell-like features. Are they truly stem cells? To be debated and discussed. But, basically, they’re less differentiated than the cells that are more responsive to therapy. And everything I’m saying so far appears to emerge with targeted therapy and immunotherapy as an equal theme.

So there’s this surviving population of cells, and these are not rare, one in a million cells. These are pretty common. And we need therapies to target them. We need to be able to try to kill that residual population. And that’s where this agent came forward as one of the very early efforts to try to poison that cell population. Having said that, there are definitely other approaches emerging now that seem to be ultimately trying to solve the same problem, to really be toxic to this dedifferentiated subpopulation or try to push them, kind of in an epigenetic way, back towards the better-differentiated type, phenotype, and then, therefore, sensitize them to these available drugs.

And so that’s been the broad proposition. As you said, this agent’s been most developed in GI cancers because multiple Phase III trials have been mounted in GI tumors. But this is emerging as a common theme across multiple therapeutic contexts. Of course, it was described in hematologic malignancies years before, that the chemo surviving, or even in the case of some hem malignancies, targeted therapy surviving subpopulations are these stem cell-like populations. But we see it in melanoma and non-small cell lung cancer, breast cancer. It’s an emerging and repeated theme.

DR LOVE: Do you think that this particular agent is going to end up in practice?

DR FLAHERTY: I think it’s too early to tell. I mean, it’s got a perfectly acceptable toxicity profile, so it can be broadly combined and is going to need to be given in combination. But it’s because of the biology that’s being pursued and the design of these trials where you’re going to need kind of full mature Phase III populations followed out to completion before you’re going to really make a call. We’ve been working with this drug specifically in Phase I for, oh, it’s been at least 4 years, maybe 5 years now, and really difficult to glean signals from Phase I or Phase I/II studies as to whether it’s moving the needle. So this is one where I think we’re going to have to kind of keep holding our breath and see what Phase III trials look like.

Choice of vismodegib versus sonidegib for advanced BCC

DR LOVE: So one more question, which is, we now have 2 hedgehog inhibitors approved in basal cell cancer. I’m curious what your thoughts and experiences are in terms of using these agents, when you use them and how you choose between the two.

DR FLAHERTY: Yes. So vismodegib being the first and then sonidegib coming right behind it. They look pretty comparable by response rate in the early going. Tolerability, these drugs have a real issue in terms of fatigue, altered sense of taste and then usually what we think is a consequence of that, weight loss, making it difficult to treat patients for a long period of time. And yet the flipside is that most of the addressable population that we’re talking about with truly advanced basal cell carcinoma are kind of locally advanced, borderline resectable patients. They’ve already been through a surgery or two and radiation as well, typically. And so we oftentimes think about these drugs as being, let’s call it induction therapy of a sort. I mean, like, we get patients into a response, recognizing we’re going to have to stop after a few or several months. And so usually the idea with talking to patients about this prospectively is, look, give us a few months, 2 to 3 months to kind of get you moving in the right direction. See if we can keep you on the therapy beyond that time frame. But we’re not talking about indefinite 1-, 2-year durations of therapy. These drugs, just unfortunately, are not tolerable in the vast majority of patients after that time frame.

So between the two, though, sonidegib, based on the emerging data that supported its approval and now in clinical practice, it does seem better tolerated. There was this randomized Phase II study that was done of kind of lower dose and higher dose and they looked equally effective, but the lower dose, which was the approved dose, clearly better tolerated. And so while it does have the same class effect toxicity that I already summarized for vismodegib, it really does have a little bit less in terms of those kind of constitutional type toxicities, if you were to look across the trials and now in clinical practice.

So in any case, suffice to say that these are therapies that I think in this population are best considered as kind of a course of therapy, plan on taking a break, you’ll be able to get kind of buy-in from the patient if you up-front communicate that you’re not going to be threatening them with indefinite therapy, but get them into a response, plan on taking a break and then watch them closely. You may have to re-treat, but like with other targeted therapies and oncogene-defined populations, it seems pretty clear that you can re-treat. In other words, if you stopped for reasons other than emergence of resistance, the likelihood of being able to reinduce a response seems to be quite high.

So that’s been our practice, even in patients who could never entertain another surgery. In patients who could be rendered surgical candidates with a response, obviously that’s a real sweet spot for these drugs. And that’s really where their uptake has taken hold, is in locally advanced patients where they would have to undergo a disfiguring surgery. And if you can get them into a significant response where it makes the surgery much less morbid, that’s a real value proposition.

Dose modifications and treatment holidays to mitigate the side effects associated with hedgehog inhibitors

DR LOVE: It must be difficult to do trials because there are not that many patients. So maybe a lot of this comes down to individual experiments with patients. Do you ever attempt, like, intermittent therapy or different schedules of therapy?

DR FLAHERTY: Yes. So scheduling is everything. It’s pretty clear that dose reductions don’t do a whole lot to mitigate toxicity. So it really is a matter of treatment breaks. And as you’re alluding to, do you just plan those up front. In other words, do you get a patient through a month or two of therapy, see if you’re lucky and there’s someone who really tolerates treatment exceptionally well? And if not, which is most patients, do you then just build in breaks? Three weeks on, 1 week off is a common strategy that people have tried. But some people need more time off drug than that to really have mitigation of these toxicities. So if you need 2 weeks off to be able to get some relief, then how long do you stay on?

This has been done really just ad hoc in practice. There haven’t been follow-on trials to really tease out these schedule issues. And really, as you said, it’s because of the rarity of the population.

New trials going forward are going to be tough, because sure, you can demonstrate activity in patients who have truly unresectable metastatic basal cell, but that’s really rare. The much larger number of patients that exist are locally advanced patients. And if they’re going to get surgery after systemic therapy, man, that makes it hard to design trials with definitive endpoints when you’ve got people differentially pursuing surgery or not. It’s kind of like the old A&L issue of patients being bridged to transplant and that making it difficult to judge outcomes.

Case: A 72-year-old woman who presents with a rapidly enlarging subcutaneous nodule on her right arm is diagnosed with Merkel cell carcinoma

DR LOVE: So I want to ask you about a couple of your cases, briefly. And I was interested by the fact that we saw an approval specifically in Merkel cell cancer, avelumab.

DR FLAHERTY: Yes, exactly.

DR LOVE: And I was like, oh, is that really a completely different disease? So I asked you to pull a case so you could educate me on Merkel cell. Let’s hear about this 72-year-old lady.

DR FLAHERTY: Yes. So this is a patient who presented like many do, with kind of a subcutaneous lesion on her extremity. And it was kind of watched initially clinically, because it just looked like it could be a lymph node or could be anything but didn’t have a real superficial finding to it in terms of at the skin surface. But it really progressed quickly, as Merkel cells typically do, which ultimately prompted biopsy and excision.

So these lesions, we really don’t know how to stage them well. But they do confer really high risk of microscopic metastatic disease and, obviously, ultimately relapse even following definitive surgery. There’s obviously surgical excision plus/minus on the sentinel lymph node approach. In this population, it’s kind of the usual. There have been debates and discussions in the field whether you should radiate the area around the primary tumor. But, of course, that doesn’t help the patients who have microscopic metastatic disease to lung and elsewhere, which are the most common long-term sequelae, which was exactly her situation.

So adjuvant therapy hasn’t really convincingly existed for this population in years past, and particularly in an older patient giving platinum-based chemotherapy, which has been considered a standard of sorts in the metastatic setting and kind of debated in terms of its relevance in the adjuvant setting, for an older patient where you have some concerns about offering a regimen like that that doesn’t have clear compelling efficacy rationale, observation is kind of the most common approach.

But now, with the emergence of these data, I would say it is particularly critical to be keeping an eye on these patients because (A) they have a very high risk of recurrence, and because we have a therapy now that’s clearly impactful.

Pathophysiology of Merkel cell carcinoma and rationale for the use of anti-PD-1/PD-L1 antibodies

DR FLAHERTY: So just a couple of reminders about Merkel cell carcinoma. There’s kind of 2 varieties of it, and the clinical studies have reidentified this with the PD-1 antibody approach. There are some that are actually very high mutation burden tumors that kind of look like squamous cell cancers in terms of thatfeature, but you’ve got this viral-associated form of Merkel cell carcinoma that’s fundamentally different in terms of its pathophysiology. And, obviously, as patients were treated and tested for the presence of this virus, presence or absence basically, and having that correlated with outcome, it turns out that actually patients can respond to a PD-1 antibody whether they have a virus-associated cancer or not. And it makes sense when you consider the high mutation burden variety is probably capable of responding simply because of that fact.

So in any case, these are tumor types — the virus-associated/nonvirus-associated one that we think are truly immunogenic, and, therefore, it lines up with melanoma data. It lines up with cutaneous squamous cell carcinoma data actually presented at ASCO that shows quite a promising response rate with PD-1 antibodies as well.

So I think it’s perfectly logical that we’re seeing the efficacy signal that we’re seeing with PD-1 antibodies. In this case, actually avelumab was the first to get approved, but we already have Phase II level data being produced with the other PD-1 antibodies showing the consistency of that effect.

So the real issue that this case brings up that I wanted to mention is that we’ve got an approval that specifically indicates that it’s for use after chemotherapy in patients with advanced disease. And yet we really have such little confidence in the chemotherapy regimens, platinum-based chemotherapy regimens that have been used in this population over the years. And so I’ll tell you that in clinical practice, the temptation is to really move front line with these agents. And particularly in an older patient like this to not be potentially losing really valuable time, where the patient has no response, this aggressive tumor type, Merkel cell, will progress so quickly that you can have patients go from asymptomatic to symptomatic in a matter of weeks. And given the efficacy data that we see with these PD-1 antibodies, the real temptation is to really move to first-line treatment. But what’s it going to take for that to become a guideline and even to alter the potential indication, we’ll see.

DR LOVE: What do you see in terms of response rate, particularly durable response?

DR FLAHERTY: Yes. The response rates are about 50%, which is really about as high as it gets in solid tumors with PD-1 antibodies. Having said that, the data that we have right now, which is just about a years’ worth, or maturity in terms of follow-up of the FDA-approved population or indication, it looks like the durability is not as great in Merkel cell carcinoma as it is melanoma. So if you just track the progression-free survival curve, you’ll see this kind of turn in the curve at about 50% of the population, even a little higher than that, actually, for patients who have stable disease, but you see a decay. Almost half of those patients who initially had disease control or response will lose it over the subsequent year. So telling you that there’s probably a decent degree of heterogeneity and complexity of these tumors that’s driving resistance even more so than we see in melanoma.

Incidence and clinical presentation of Merkel cell carcinoma

DR LOVE: Can you kind of do a brief Wikipedia type of summary of Merkel cell in terms of number of cases, cell of origin and clinical presentation?

DR FLAHERTY: Let me start with cell of origin. That continues to be an absolute mystery. We still don’t know what the cell of origin is. And there’s all sorts of work that’s been done to suggest that it might be sweat glands. It might be hair follicle. It’s a deep structure, and typically in the dermis is where these things arise, which is why, in a case like this, one that I presented of this woman, she had a truly subcutaneous nodule. And again, it supports this kind of dermal origin of these cells.

In terms of cases, we’re not sure that we really have adequate capture of all the cases. Maybe this is one that now that we have more effective therapy, we’ll get a more complete and accurate assessment. These are small blue-cell tumors, though they look like small cell tumors in other parts of the body. But when they arise in the skin primarily, then they’re rightly called Merkel cell carcinoma. We think there’s a couple of thousand cases a year.

DR LOVE: Really?

DR FLAHERTY: But not more than that.

DR LOVE: That’s a lot.

DR FLAHERTY: So it’s a population that has, I think, fallen between the cracks, even at major centers. Many of them are head and neck. About 50-percent-plus of them present in a head and neck location and will often be triaged to head and neck surgeons for initial surgical management. But the rest of them happen elsewhere, and most common in the extremities. So they’ll be managed up front by a more general surgical oncologist. And then who sees them for medical oncology, it totally depends on kind of referral patterns and otherwise. But if you look at the registry data, it’s really hard to know whether we’re even fully capturing this population of patients. Many of them might be getting called small cell cancer NOS and not accurately classified.

DR LOVE: It’s interesting. If you were to see a patient, for example if you had seen this patient, or a patient like her earlier on, after the primary tumor removal and lymph node dissection, who would be willing to pay for the drug, whatever. Would you use avelumab in the adjuvant setting?

DR FLAHERTY: Yes. I think it’s perfectly reasonable. There are a couple of Phase II trials, adjuvant trials being mounted now to at least try to track these patients in that setting to understand, even without a proper control arm initially, what the effect might be. In the absence of that data, what I used to say before ESMO was, we really don’t know about the adjuvant benefit of these PD-1 antibodies in any setting. Now in melanoma, we know. And so to then cross over to a rare tumor type like Merkel cell, where who knows how long it will take for us to get definitive data, I think it’s perfectly reasonable. The safety profile, in my view, of PD-1 monotherapy in the adjuvant setting, to say that given the activity of these drugs in the metastatic setting in this population, that is, I think, a reasonable discussion to have with the patient. Recognizing that we don’t know if the benefit is maintained or possibly worse, possibly better, to deploy it in the adjuvant setting. But for a patient who understands those concerns and would be willing to cover the cost if their payer won’t, I don’t see a reason to withhold.

Neoadjuvant therapy for melanoma

DR LOVE: Incidentally, I don’t know if you consider postchemoradiation therapy in lung cancer adjuvant, but kind of seems adjuvant to me.

DR FLAHERTY: Yes.

DR LOVE: Standard of care. But another thing I was going to ask you about, not having adjuvant data, is neoadjuvant. We saw this paper at ASCO from the Memorial Group looking at neoadjuvant checkpoint inhibitors in lung cancer. It was pretty interesting. They saw a lot of responses. What do we know in general about neoadjuvant therapy in melanoma?

DR FLAHERTY: Yes. So there’s 2 strains of that. There’s the immune checkpoint approach and BRAF/MEK neoadjuvant. And, actually, it’s MD Anderson group that’s been producing the most data early on with either approach. Really fascinating. I mean, because we get complete pathologic complete responses — 50% of patients on BRAF/MEK and pushing about 40% on PD-1, neoadjuvant.

Now, the thing about this data that you have to keep in mind is that we have a very small fraction of the melanoma population who are ever in that clinical window where you kind of classify them as being neoadjuvant candidates. Remember, the vast majority of our patients present with a small primary tumor that’s easily excised.

So to have, let’s say, bulky lymph nodes that are resectable and where they’re candidates for these types of trials or this clinical approach, that’s a pretty niche population. But they could be a huge opportunity for us to do studies in that population, building on this early evidence of a pathologic complete response, right? Because if we have new agents that we can throw into the mix and show in small numbers of patients, like in the breast cancer context with neoadjuvant studies, that we can bump up the complete response rate. That might be a very effective way of triaging new approaches before scaling up massive randomized Phase II and Phase III trials.

DR LOVE: That was always the hope in breast cancer, but I guess it didn’t quite exactly work out.

DR FLAHERTY: No, right.

DR LOVE: It makes a lot of sense, but sometimes things that make sense don’t always work.

Case: A 21-year-old woman with node-positive, BRAF V600E mutation-positive Stage III melanoma receives dabrafenib/trametinib after disease progression on talimogene laherparepvec and anti-PD-1/anti-CTLA-4 therapy

DR LOVE: So one final thing I want to ask you about is, I noticed, looking through these cases, that you had a patient who got T-VEC. And I want to kind of get an update of where that is. This is this 21-year-old. What happened there?

DR FLAHERTY: Yes. So this is a patient who had this recurrence of disease that was in transit, as we call it, these dermal metastases, kind of in lymph node, actually. It was the first thing that was injected in her case, but then she had the in transits. And T-VEC, that’s right in the sweet spot of when T-VEC was tested prospectively in Phase II and then Phase III trials.

DR LOVE: Maybe you can just backtrack and explain what T-VEC is.

DR FLAHERTY: Oh, I’m sorry. Yes, of course.

It’s a herpes virus that’s basically replication incompetent. So when you inject it, it can’t disseminate widely as an active herpes virus. But it will survive in the tumor cell long enough to be able to induce expression of GM-CSF. So it’s genetically engineered, the herpes virus, to express or to deliver the gene that expresses GM-CSF, which is a cytokine that’s been of interest in melanoma for a long time. And if you can express high local concentrations of that stuff, it would appear that you could then induce a local immune response. So that was the design of T-VEC, was to basically use the herpes virus as a carrier to get the GM-CSF gene in there, induce GM-CSF expression and then invite in dendritic cells and inflammatory responses. Clearly, it does that. And as you probably recall from the Phase II and Phase III data, one of the issues was that you could get a response in injected lesions pretty reliably. But the issue is, do you have an effect beyond that? Because a patient like this who had a very worrisome initial presentation, even more worrisome recurrence, Stage III recurrence initially, these patients are inordinately likely to develop subsequent metastatic disease.

So if you’re thinking about T-VEC as an approach, you want a systemic effect. I mean, that’s really — you know that you’re going to be facing recurrent disease in 80%, 90% of these patients, systemically.

So the point being that in this case, it was used very much on label, kind of right in the way that T-VEC was investigated in trials. And when you looked at the subset data from the Phase III trial of T-VEC, it was these patients, these Stage IIIB, IIIC or even in the M1A category, who seemed to even have an overall survival improvement, suggesting a systemic effect.

Now the fact is, all of that data was very interesting at the time. But with the immune checkpoints coming, and of course BRAF/MEK therapy coming as well, is much more active systemic therapies. One wonders, like, what’s the place for T-VEC anymore? Right? Because even in this exact patient who presents with this Stage III recurrence, wouldn’t you just use that as a window to now initiate these much more systemically impactful therapies, that you know, high likelihood would give local response as well?

So turning all this and putting to the scenario now, where you have a patient who presents with metastatic disease of some kind and they have an injectable lesion, is there are role for giving T-VEC, let’s say, for a few weeks before you start your “definitive” systemic therapy? Or in a patient who’s been getting systemic therapy but has residual disease that’s injectable, do you use that lesion to deliver T-VEC to be able to try to augment the systemic response and the immune response? These are the questions that the field’s wrestling with now, post the approval of T-VEC. But remember, when it was developed it was tested by itself as kind of a stand-alone therapy. That’s just not how it’s going get incorporated if it is maintained as a treatment approach in the field.

DR LOVE: The thing that I wondered about is, I heard about these issues in terms of how you actually do the injection and, like, the infection precautions and stuff. It sounded like it was kind of a hassle.

DR FLAHERTY: Yes.

DR LOVE: Is it?

DR FLAHERTY: It is, unfortunately. It needs to be treated essentially like a live virus. So as I said, it’s replication incompetent, but that doesn’t take it below the radar of how a practice or an institution handles it. So it has to be in a special pharmacy. It has to be refrigerated in special conditions. You can’t keep the stuff around forever, right, so you have to kind of refresh your stock pretty frequently. So there is all of those issues. And then in terms of the administration part, basically you need to have a nonpregnant practitioner. That’s kind of rule number 1. But otherwise, it’s pretty straightforward.

The volume that you basically mix up of the stuff, you’ll inject it across the lesions that you are dealing with. If you have 1 large lesion, you try to put it all in there. If you’ve got a few small lesions, you try to distribute it across. And that’s not perfectly codified, right? In the clinical trial that established the benefit of the approach, different patients were treated differently in terms of exactly how you used the material, if you will.

So yes, there’s some art involved in terms of how to incorporate into practice.

DR LOVE: So final question. I see that this woman actually had a BRAF-positive disease and that you, I guess, were thinking about or maybe you already have initiated systemic therapy. I see too, PD-L1 IHC-negative.

DR FLAHERTY: It’s a real tricky issue about your — okay, she has a BRAF mutation and she’s PD-L1-negative, does that make it then obvious that you give BRAF/MEK, because she would have a 10%, 15% likelihood of response to PD-1 antibody and she has a 70% response likelihood to BRAF/MEK. Some practitioners would say yes, that’s exactly how and why I would use a PD-L1 test. My objections I already made before.

So, in fact, in talking with her, when she was referred to me with all of that data available, I told her, “Look, I can’t confidently incorporate that result in decision-making. What I can tell you is, we’ve got these 2 approaches. They’re both certainly on the table. Because she’s so young, PD-1/CTLA-4, as a combination approach, even with its not magnificent benefit at the level of overall survival, it’s what I kind of refer to as the young, highly motivated. These are the patients who we used to talk about high-dose IL-2 with, those kinds of patients. When you have a conversation about PD-1 monotherapy versus PD-1/CTLA-4, you find more traction, if you will, for throwing the kitchen sink there with the combination.

So it’s kind of that combination versus BRAF/MEK combination was the relevant discussion with her. She ended up choosing PD-1/CTLA-4 combination as the initial approach, with BRAF/MEK then as the back-up.

DR LOVE: Yes. That’s kind of what I predicted when I saw this case. How did she do?

DR FLAHERTY: Yes. So she didn’t respond to PD-1/CTLA-4 after, basically only 2 induction cycles and severe toxicity that precluded getting more. So she kind of basically hit the stopping point on induction therapy after 2 rounds. And at least at that point didn’t have evidence of a response. So she switched to BRAF/MEK right at that point.

DR LOVE: Which combination?

DR FLAHERTY: Dabrafenib/trametinib.

DR LOVE: What kind of toxicity did she have?

DR FLAHERTY: So she had colitis and what seemed like gastritis. I mean, true treatment-related gastritis as well. It was symptomatic, led to a scope, and she really did have quite an inflamed gastric mucosa. That’s not a very common site of checkpoint-associated autoimmune toxicity, whereas, of course, the colitis we would comfortably attribute. But because of the timing, they really were at the same exact time frame. We ascribed all of that to being checkpoint antibody related. But she needed steroids. So, basically, someone who hits that level of toxicity is, at least for the time being, done with immune checkpoint therapy in our practice.

Incidence and spectrum of immune-related adverse events associated with immune checkpoint inhibitors

DR DAUD: If you look at immune-related adverse events, there are somewhere between 20% and 60% of all people treated with PD-1 inhibitors, depending on whether you’re talking single agent, combination, combinations with ipi or combinations with IDO inhibitors. So the percentage of serious adverse event varies somewhat depending on the precise PD-1 drug or PD-L1 drug that you’re talking about.

And I think the most life threatening, the most worrisome are things like colitis, liver function test abnormalities or hepatitis, pancreatitis. And then some of the endocrine abnormalities — things like Type 1 diabetes, they’re not particularly common, but they are pretty life altering. I’ve had patients who’ve had adjuvant PD-1 and — I have a 35-year-old who had adjuvant pembrolizumab and has developed Type 1 diabetes. And we don’t think that is going to change over time. We think he’s stuck with checking his sugar and giving himself insulin probably forever. And things like myocarditis and some of the neurological side effects that are not as common but very life altering.

And so what is the common theme behind these adverse events? I think one of the common themes seems to be activated T cells infiltrating tissues that probably they shouldn’t be infiltrating, like the colon, like the liver, like the neurologic system, like the heart or some of these endocrine organs.

In some cases, we think the problem is aberrant or unusual expression of adhesion molecules in these target organs like the colon or liver. And that is getting activated T cells attracted to them. And once the T cells are there, just by random chance, just stochastically, some T cells will react to whatever tissue they appear to be trapped in. And then once you have that damage created, then it kind of sets off this vicious cycle where damage and more attractive adhesion molecules.

So currently what people do is use steroids, immune-suppressive drugs like mycophenolate and stuff like that. But if you do that, you’re also killing off some of your tumor-reactive T cells. So there’s a delicate balancing act there.

But what we think is that there could be more specific agents that might be able to prevent trafficking of these T cells into the colon or into the liver and still preserve the T cells, per se, and maybe they can still traffic to tumor sites and still tumor. That’s the vision, that you might be able to do that without destroying the T cells themselves.

DR LOVE: And where are we or where are you in terms of research looking at alternative strategies?

DR DAUD: To me, one of the really interesting pieces of research has actually been the already-done research that people in ulcerative colitis and Crohn’s disease have done. And basically in ulcerative colitis and Crohn’s disease, what people have figured out is that there are these integrin antibodies that can prevent integrins, which are the sticky substrate on T cells, and basically can prevent integrins from binding to adhesion molecules.

And if you do that, you can get some effective treatment for things like ulcerative colitis and Crohn’s disease and multiple sclerosis and some types of autoimmune liver disease. So I guess what we’ve done, our preliminary data indicates that a similar thing might be going on in immune checkpoint-induced adverse events. And you might be able to prevent it by just preventing these T cells from kind of getting waylaid in your colon or in your liver rather than going to the tumor.

Perspective on the utility of immune checkpoint inhibitors in patients with preexisting autoimmune diseases

DR LOVE: That leads into the issue of the use of checkpoint inhibitors in patients with prior autoimmune conditions. Where are we today in understanding that scenario? And how do you approach that situation clinically?

DR DAUD: Yes, I think it remains a very difficult issue, especially when you’re talking double — and now people are talking about triple immune checkpoint inhibitor combinations where you might have a PD-1 agent. You might have a T-cell-activating agent. And then you might have an IDO inhibitor or something else, a triple combination of some type.

In these cases, once you have a runaway activated T cell, it’s very hard to stop that without also stopping a lot of the tumor-reactive T cells. And so what people do is use steroids plus mycophenolate for liver toxicity or steroids plus TNF blockers for colitis and for bowel problems.

And then there’s a further class of problems like multiple sclerosis, myasthenia gravis, myocarditis, where essentially there isn’t much that can be done, except to maybe cope with the symptoms of the damage. You could try and reduce afterload on the heart. And there isn’t much you can do to prevent the myocarditis or the damage. But once you’ve recognized the damage, maybe you can put, like, a pacemaker in or you could — I’d say managing the symptoms more than preventing that toxicity from happening.

So I think for things like diabetes and myocarditis, I don’t think we know today what the problem or a good approach to prevent that problem. But I think for things like colitis and hepatitis, which — and thyroiditis, which — these are common side effects. And I think in many cases, these are the things that bring treatment to a halt. Especially the colitis brings treatment to a halt. I think in these cases, I feel like the blocking of adhesion receptors/adhesion molecules might be, like, a productive path to prevent them from happening.

DR LOVE: Any thought about whether or not some of these patients that you see autoimmune toxicity that don’t have a, quote, clinical history of autoimmune disease maybe do have an autoimmune disease that’s subclinical that’s being brought out?

DR DAUD: Yes. I think that’s a great point. If you look at patients, a lot of times you’ll see that when you get a PET scan on someone, you’ll see increased sugar uptake in their thyroid glands even before they start. Of course, you see thyroiditis after they start. But it could be that you have some subclinical, but I think it could even be that all of us, or many of us, have T-cell infiltration in our thyroid or in our hearts. But because of PD-L1 blockade, it kind of keeps those T cells from doing much damage. Once you get the inhibitor, now those T cells are running amok.

Case: A 35-year-old man with Stage III melanoma and a history of Guillain-Barré syndrome develops diabetes after receiving adjuvant pembrolizumab

DR LOVE: You mentioned the case you had of the patient with diabetes. Can you compare what we think the pathophysiology is of the Type 1 diabetes to start with and how that compares to what happens when you give somebody a checkpoint inhibitor and they get it?

DR DAUD: So the patient I’m talking about is a 35-year-old. And he, in fact, had this very interesting history where I think he was actually — he had a history Guillain Barré syndrome that happened after an apparent E coli infection, which I think he picked up like a traveler’s diarrhea in one of his trips somewhere. And then he had that Guillain Barré syndrome. It was very transient. It was just a polyneuropathy, then just regressed. Otherwise healthy, active, very fit 35-year-old.

So he got adjuvant pembrolizumab. He had a Stage III melanoma. He had a couple of different sites of disease, but Stage IIIA, microscopic disease in both sites. And I think about month 6, he just came in for a routine visit and his sugar levels were — I want to say 250. And when I talked to him, he was like, “Yes, I had a smoothie. I was, like, working out in the gym. I was feeling kind of dehydrated. I had a smoothie.” I was like, “Yes. Okay. That makes sense. Maybe it’s just one of those things.”

And then the next time he came in, again he had a really high sugar level, again, it wasn’t a fasting sugar — and I wasn’t as concerned. But we did send him to the endocrinologist. And lo and behold, he did more testing and confirmed that it was diabetes. And I’ve noticed the same pattern, that you could go months and months of treatment, and then abruptly you have, like, a spike. And he didn’t have any pancreatitis. He didn’t have lipase elevations, which you’d think you might have lipase elevations if you have T cells in your pancreas. But his was just kind of, like, out of the blue.

DR LOVE: Interesting. So he got what, adjuvant pembro on a trial?

DR DAUD: Adjuvant pembro. It was actually off trial. So it was not on trial. It was just adjuvant pembro, because he had this previous — and you could question whether this was a good idea. I mean, I certainly question that now. But I think because of his history of Guillain Barré syndrome, I was actually worried about giving him interferon, because that would have been a standard drug. I was worried about giving ipi, because I thought ipi was much more likely to cause autoimmune side effects.

He didn’t have a history of colitis, but he did have some irritable bowel-type symptoms. And so I thought pembro was the safer of the two. In fact, I made the decision to give him pembro and he actually paid out of pocket for it. He was one of those people who could.

DR LOVE: I guess in retrospect you gave him the right treatment anyhow, right?

DR DAUD: With this auto, maybe ipi wouldn’t have caused this diabetes. So I was worried at that time. I was like, “Wow! The guy has this history of neurologic toxicity.” Ipi, we think, can cause all kinds of autoimmune toxicity.

Viewpoint on the use of immune checkpoint inhibitors or targeted therapy in the adjuvant setting

DR LOVE: No. I was just thinking of the adjuvant nivo trial that was just presented at ESMO. And now — but on the other hand, actually if you stop and think about it, I mean, what were you projecting at the time as his risk for recurrence?

DR DAUD: Back in those days, I thought his recurrence risk was about 30%. And I have to say that I would have been shocked if the ipi versus nivo adjuvant trial had shown anything other than what it showed, only because there’s been so much data in the last few years.

DR LOVE: I guess I — just thinking, though, about it moving forward — so now, I guess, everybody’s going to use adjuvant nivo. But, I mean, would you use adjuvant nivo on a patient like this with a history of Guillain Barré, knowing they might be cured of their melanoma anyhow?

DR DAUD: That’s a great point. I was actually at a debate with Mario Sznol. And he was arguing the other point. I was arguing that adjuvant therapy makes sense. And he was arguing that adjuvant therapy does not make sense in some of these low-risk patients. And I think he convinced me, partly because of my experience over the last few years, because now I think essentially you have 2 shots at patients.

You have the adjuvant shot, and then you have the chance to possibly take a certain percentage of melanoma patients, perhaps it’s 20%, Stage IV patients. And you could move them into a complete response category. And then those are going to be lasting CRs. Perhaps it’s 30%, if you look at ipi/nivo, or maybe even more.

So the fact that you have 2 shots should really be tackling these Stage IIIA patients at an early stage and perhaps leaving them, a small percentage, with lifelong toxicity. I wonder if we should now reserve adjuvant therapy only for really high-risk, IIIB, IIICs and Stage IV patients.

DR LOVE: That’s really a good point. I didn’t even think about it. Because in other parts of solid tumor oncology, breast cancer, non-small cell lung cancer in patients with targetable mutations, people talk about you really ought to compare adjuvant therapy to treatment on relapse. And I guess really the melanoma trials don’t do that, right?

DR DAUD: They don’t do that.

DR LOVE: Although I guess that’s what happens anyhow.

DR DAUD: That’s what happens. I think one of the things that brought this point home to me is that if you look at the COMBI-AD trial, which is the BRAF/MEK inhibitor adjuvant study, if you look at the relapse-free survival curves, they are really impressive. They’re pretty impressive. But if you look at the overall survival curve, the hazard ratio today, it’s not statistically powered to show a difference, or it’s not currently showing a difference even though it has a lot of zeros in it. But it isn’t as impressive. And many of these patients were treated in Europe or Australia, not in the United States.

Now today, if you were to do a trial like that, I would — I mean, I’m guessing the survival benefit for some of these lower-risk adjuvant patients might not be very considerable, which is another way of saying, should you really be treating everybody at that early point, giving them a year’s worth of treatment? Or should you be treating them at a later point when a small percentage of them — like, say, 30% of them — relapse, and then treat them at that point? And then possibly you have just as good a chance of rendering them disease free.

Hepatic and dermatologic side effects associated with immunotherapy

DR LOVE: In terms of treating people with prior autoimmune problems, is your global take that in general you cannot uncommonly see reactivation or exacerbation of this across the board? It’s worse if you give anti-CTLA-4? Is that kind of globally the way it’s shaking out right now?

DR DAUD: Yes, I think that’s the way it’s globally shaking out. I think a lot of times what my experience has been is that the autoimmune toxicity that you worry and lose sleep about, a lot of times that is not the autoimmune toxicity that actually ends up happening, just like with this case. I’ve had this experience that you’re thinking, “Oh, yes. That guy had colitis. That’s the thing that it’s going to aggravate.”

But lo and behold, it’s hepatitis. It’s serious hepatitis that seems to be the problem. And again, I’ve been reading the autoimmune disease literature, and apparently what happens with autoimmune hepatitis is that you have aberrant expression of this adhesion molecule called MADCAM1 in the liver. And MADCAM1 typically is only present in the colon. Once it’s in the liver, then T cells, which should have been sticking around in the colon, now decide to go to the liver. And once they’re in the liver, they just stochastically start reacting against the liver.

DR LOVE: That's interesting. What about skin toxicity? First of all, clinically, how often do you see it? What’s the typical presentation? And again, anything in terms of the pathophysiology that we know about?

DR DAUD: It is very common. And people dismiss it as being trivial, but the problem is that if you’ve been on treatment for 2 years or a year and a half, and if you’re constantly itching yourself at night — and a lot of these skin toxicities do seem to happen at night or after you shower — it could just be really life altering.

I’m just thinking about this one gentleman who’s on a clinical trial, has had a complete response, brain mets. He was on the brain mets study with ipi/nivo. Complete response. So he’s grateful. He was BRAF mutant. We didn’t give him gamma knife, just — he had multiple, like, I want to say over 8 brain metastases, many of them very sizable. So has had a complete response, but now, about a year and 8 months later, his skin toxicity is really getting to him.

He had some kind of hay fever dermatitis or something like that even before coming on the treatment. But now, even despite the fact that he’s using steroid all over his body, with Saran wrap and stuff, he’s just itching himself. Just wakes up every night and he told me, he’s like, “Can I just stop this treatment? What do you think are the chances that this response will stick?”

And I just said, “I think it’ll stick, although I’m not sure. I’d like to do a few more months of treatment to get a couple more negative PET scans.” But he’s leaning towards stopping treatment early.

DR LOVE: What is he receiving?

DR DAUD: He’s getting ipi/nivo.

DR LOVE: And you said he’s BRAF-positive. What about switching him over to there?

DR DAUD: Yes. But the thing is, he’s had a complete response. He had brain mets, so he avoided radiation treatment. He’s had, like, a fabulous — so he has had a complete response. It’s just a question of when to stop him. And I think I might decide to stop him earlier just because this skin toxicity is just getting the better of him.

Management of brain metastases in patients with melanoma

DR LOVE: Can you comment a little bit about your approach to brain mets both in BRAF-mutant and wild-type disease?

DR DAUD: I think that’s been a huge change, like, I’d say in 2017. Prior to 2017, I think most people, if you polled them, most melanoma specialists would lean towards surgery or Gamma Knife^® for a percentage of patients with brain mets. But now our practice has changed.

I think for many patients, we just start them on just a systemic treatment, whether it’s BRAF and MEK inhibitor or whether it’s ipi/nivo. I don’t use PD-1 single agent for patients with brain mets, generally. I think the response rate — Harriet Kluger had a very nice study in Lancet, where she showed that there was — I think it’s about 20% response. But to me, ipi/nivo, that’s a good case for using ipi/nivo.

And generally, we reserve Gamma Knife for the lesions that are not responding or not shrinking, or residual tumors or, in some cases, surgery for tumors that bleed, which can happen. You can have general shrinkage, but you can have bleeding tumors or, in some cases, if you use BRAF/MEK and you have similar — like, a discrepant response.

Or, sometimes when you have a nadir, I’ll just zap the ones that are left behind at the nadir, because I think that BRAF inhibitor — my thinking still is that most patients will recur on a BRAF inhibitor who have brain mets. You might have bought them time, but you’re probably not getting a long-term response in a patient with brain mets specifically.

DR LOVE: Interesting. And globally, do the responses in the brain kind of mirror what you see systemically? In other words, you see more responses with BRAF therapy?

DR DAUD: I think this ASCO, there were a couple of back-to-back presentations. And it actually seems like the response rates are pretty comparable. I didn’t see that there was a greater — I think it’s about 40% to 50% for BRAF/MEK. And I think it’s about 50% with ipi/nivo. So I think in the brain, there doesn’t seem to be a discrepant response by brain-specific measurement criteria, so the response for BRAF inhibitors, BRAF/MEK inhibitors, is a little bit lower in the brain than it is systemically. And at least in that one — the Hussein Tawbi presentation, it seemed like the responses were pretty similar. And so I think all things being equal, I guess I somewhat prefer starting with ipi/nivo.

Approach to single-agent versus combination treatment with immune checkpoint inhibitors as first-line therapy for metastatic melanoma

DR LOVE: And speaking of ipi/nivo, I’m curious where you stand today in terms of using that combination in first-line metastatic disease as opposed to just a single anti-PD-1 agent and whether or not you incorporate PD-L1 level assays into your decision.

DR DAUD: Yes. Thanks for bringing that up. Yes, I do incorporate PD-L1 assays. Now, I’m somewhat of an outlier, because I think a lot of people don’t seem to think that the PD-L1 assays are meaningful. But my view is, just like we were talking about a few minutes ago, I think that the PD-L1 assay is technically reproducible. There have been studies done comparing different antibodies. It works fine.

I think the problem is, you’re not converting all of the PD-L1-positives or all of the hot tumors to responses with PD-1. And the problem there, I think, is that you have inhibitory cells or bystander cells. Your PD-1 somehow stimulates those guys. That’s my view today of the tumor microenvironment. So you’re both stimulating the antitumor cells. You’re also stimulating some of the inhibitory cells. And who gets there first determines whether you clinically have a response or not.

So — but I think in this specific issue, single agent versus combination, I think a T-cell assay, whether you use our homegrown assay here at UCSF, use PD-L1, use gene expression profiling that Toni Ribas has talked about, any of those assays will give you some idea whether you’re talking PD-1 responsive or you’re talking needs more checkpoint stimulation. And I use that assay to inform. So the bottom line is, I think I impart 50-50 with ipi/nivo versus PD-1 monotherapy.

Effect of the gut microflora on response to immunotherapy

DR LOVE: What about other predictors? I know you’ve been very interested in that general topic. Where are we in terms of getting better predictors?

DR DAUD: I think we are making headway. There’s the clinical factors like the liver metastases. There’s the LDH, performance status, just a clinical mucosal, uveal, acral lentiginous versus the standard garden-variety skin melanoma. And, actually, there’s a couple of really cool papers showing that microbial flora has an influence on response.

But I think that the issue with the microbial flora question is that — one paper was from Europe by Laurence Zitvogel and one was Jen Wargo and colleagues at MD Anderson. And the problem still seems that we know that a diverse flora is good for you. And we know that if you take antibiotics, you’re disrupting your gut flora.
But within the spectrum of normal floras, it’s hard to know if 1 particular bacteria or 1 particular organism makes a difference. In some of the papers, it seems like rumenal coxi is more important. In some papers, it seems like ackermannii — if I’m pronouncing that correct — is the right flora. So I don't know if the particular kind of flora is what these studies are pointing out to a diverse flora that’s not antibiotic, modified or extinguished. And so I think in terms of clinical applicability — I mean, I tell patients to try to avoid using antibiotics. Use probiotics. But beyond that, specifically, what can you say to a person who walks into your clinic? I’m not sure today what we can do.

DR LOVE: What’s the pathophysiology of why the different flora change the way the drugs work, or is it a pharmacologic thing or biologic thing?

DR DAUD: It isn’t super clear, but one theory is that some of these bacteria, it’s the filaments in them, probably our filamentous structures are stimulating TLR9 in your gut. And that, in some way, acts on your liver and, in some way, that changes your immune rheostat and gives your T cells a little extra activation. That’s the simplistic way I look at it. In some way, it’s changing your whole body immune rheostat.

There was actually a really interesting paper from this guy, Kevan Harold at Yale, who’s done some really cool work with autoimmunity. And, in fact, he showed that even immune suppressants don’t tend to work as well when you’ve had antibiotics that have wiped out your gut flora. So it’s not just immune checkpoint, but you do seem to disrupt immune agents in general if you change the temperature of your immune system.

DR LOVE: The microbiome. I love that one.

Emerging data with immunotherapy combinations in patients with melanoma

DR LOVE: So you mentioned combination immune checkpoint therapy. Of course, we have anti-CTLA-4 and anti-PD-1/PD-L1, but what about some of the other agents? You mentioned IDO inhibitors. I know there’s one that’s actually got a name now, epacadostat. What are those? And in general, what is the thinking now in terms of looking at other checkpoints — there’s TIM-3 and LAG-3. Any thoughts about those?

DR DAUD: One of the interesting things with epacadostat is that PD-L1 is actually a pretty good marker for IDO inhibition. And IDO1 expression is a pretty good marker for PD-1 inhibition. And I think it kind of points to a common etiology. And I think what you’re doing is you’re taking and, say, converting 50% of the PD-L1-positive tumors into responders with IDO plus PD-1. You’re converting maybe 60% to 70% of the hot tumors into responders.

And I think something similar is happening with the LAG-3 and TIM-3 inhibitors. One of the things that was fascinating to me at ASCO was the drug that’s known as relatlimab, if I’m pronouncing that correctly. I think Paulo Ascierto presented that data at ASCO. Basically it’s a LAG-3 antibody with PD-1.

The response rate was modest. It was, like, in the, I think, 14% range, 11% to 14% range. So not an earth-shattering response rate. But this was in the PD-1-refractory population. And they have this marker, LAG-3 expression. And the LAG-3-positive group is where most of the responders were. So I feel like that’s a huge step forward, because now we have a marker. You could probably hone down on that PD-1-nonresponsive population and get to the marker-positive.

Biology of Merkel cell carcinoma and implications for treatment with anti-PD-1/PD-L1 antibodies

DR LOVE: Maybe you can talk a little bit about what Merkel cell is and what we know about it in terms of immunotherapy.

DR DAUD: I feel it’s really interesting, because it shows just the tremendous benefit of immunotherapy. So Merkel cell is basically a neuroendocrine cell that’s present in your skin. And it’s a receptor. It’s a sensory receptor. And so for reasons that we don’t fully understand today, this Merkel cell polyomavirus, which is present, which is very prevalent in the United States — maybe 40% or 50% of the population has traces of Merkel cell polyomavirus — can randomly integrate into your Merkel cells in the skin.

And then depending on where exactly it integrates in your DNA, it can turn on cell proliferation and produce large T and small T antigen that’s part of the Merkel cell polyomavirus. And then the large T antigen predominantly, we think, drives continuing cell proliferation. And so that’s about 80% of all Merkel cells.

Twenty percent of Merkel cells it seems like it’s just UV damage that causes Merkel cells. And one of the amazing things is the high response rate with PD-1. It’s, like, 40% with avelumab and, I think, 60% with pembrolizumab, although the trials were done in different populations. I think avelumab was second line. Pembrolizumab was first line.

And the real response rate might be probably 30% to 40% once these trials are done in a comparable population. Because most of the Merkel cell population is elderly, our previous chemotherapy for them has been etoposide and platinum, which is super toxic. I think being able to offer people, whether it’s pembrolizumab or nivolumab or avelumab, either one of them, that’s a huge advance. And many of these responses, just like in melanoma, are long lasting. You’re talking about possible complete responses in some proportion of cases, years’ long duration. So I think that’s an amazing thing.

DR LOVE: What locations do you see these lesions?

DR DAUD: There’s 2 types of Merkel cells, predominantly. One of them is in the sun-exposed skin. And a lot of them are in much older folks. And then there’s a second class of Merkel cell that’s in the somewhat younger people, like fifties and sixties. And that seems to be on the buttocks and — not 100% sure why these different populations have different locations, but one of the interesting things that stand out is that when you look at PD-L1 expression or when you look at Merkel cell polyomavirus versus negative patients in terms of their response to PD-1 drugs, it doesn’t seem to make any difference.

So the virus-positives/virus-negatives, about the same response rate. PD-1, also — again, many of these trials are small, so we have to give them a — you’re not talking hundreds of patients. So it could be that there could be some small difference in PD-1 expression. But it doesn’t seem to make any difference whether you’re virus-positive or not or PD-L1-positive. It seems like the response rate is pretty equivalent in all these populations.

DR LOVE: What’s the typical clinical course? They present with a localized lesion that is excised surgically?

DR DAUD: Yes. It’s like a red — it almost looks like a boil on your skin, but it’s not painful. It’s solid. It’s not a painful — it’s not a white boil. It’s a flesh-colored boil. Oftentimes, it’s on your face or it’s on your hands for most people. And it just grows rapidly and, oftentimes, presents with either just that primary lesion or with a regional lymph node. And the previous treatment paradigm has been wide local excision, sentinel lymph node biopsy, adjuvant radiation.

Many of these lesions are very sensitive to radiation. And I guess the one difference between melanoma and Merkel cell is, melanoma is generally radio resistant. But Merkel cell, even in advanced cases, can be very radio sensitive. And we’ve had a couple of patients where they’ve had responses to PD-1, but then there’s a residual tumor left behind somewhere. You radiate that tumor and, lo and behold, you have this abscopal kind of response, which I haven’t — at least I haven’t seen a lot of abscopal responses myself in melanoma. But in Merkel cell, I have seen that with PD-1 that you radiate one tumor site, and then everything just responds.

DR LOVE: That's interesting. How often is local disease control a problem?

DR DAUD: Merkel cell can be a problem with local disease. For reasons that are not totally clear, Merkel cell has a habit of metastasizing to the pancreas.

DR LOVE: Wow!

DR DAUD: We don’t know what the biology there is, but it often presents as a very large pancreatic mass. And near there, there’s a celiac artery. There’s a possibility of having intra-abdominal DVTs. And so local and regional control can be a big problem in Merkel cell. And they’re rapidly — so a lot of times you’ll have surgery, and then when the patient is recovering, again, you’ll see the Merkel cell lesions pop up.

DR LOVE: Globally, what’s the overall cure rate with Merkel — just with local therapy with surgery?

DR DAUD: I think it’s high. It’s about 50%. Fifty percent is cured.

DR LOVE: Anything other than the pancreatic mets that’s characteristic in terms of metastatic disease? I’ve heard it said it’s usually rapidly progressive. Is that your experience?

DR DAUD: It’s rapidly progressive. It has a different biology than melanoma. It doesn’t tend to metastasize as much to the brain, from my personal experience, as melanoma does. And it’s generally an older population from what I can see.

Benefits and risks of adjuvant nivolumab versus dabrafenib/trametinib for BRAF-mutated melanoma

DR LOVE: We were talking about adjuvant therapy before. And now, of course, we have an instant dilemma in BRAF-positive patients. You have the positive dabrafenib/trametinib trial. You have the positive nivo trial. It kind of looks like similar risks and benefits. How are you now, since ESMO, discussing this with patients?

DR DAUD: I had a 31-year-old who had a Stage III melanoma. And I want to say he had a left arm primary/axillary mass resected. And I had this question just now. His dad is a pathologist, so he’s pretty knowledgeable. And basically I made the decision — and he was BRAF mutant, also. I made the decision based on the fact that he was PD-L1-negative. And so I just thought that his chance of responding to nivo was maybe a little bit lower.

And — because he was BRAF mutant, I thought — and in his case, he’s 31. So I didn’t think that waiting and watching was a great option in his case. And I thought there was some chance of long-term cure. So that’s how I think. That’s one way to make that.

I think what I’m hearing from people is that in most cases, they’ll use immunotherapy. And that’s perhaps reasonable. But I think there are cases where immunotherapy won’t work. And I’ve seen that. You had patients recur while they are on adjuvant nivo or pembro. So I think that just points to the fact that if you treat patients who just don’t have a lot of T cells, you’re probably going to fail.

DR LOVE: But in general, is it your take that both of these strategies have similar benefits, or do you kind of feel like immunotherapy might, in the long run, be a better strategy?

DR DAUD: I think immunotherapy might be a better strategy in the long run. And the reason I say that is that again, without knowing a whole lot about survival, which I think it would be nice to see some survival benefit from the nivo versus ipi trial, I think we know that ipi has a survival advantage, but does nivo have a survival advantage versus ipi? So we don’t know that. And you’re talking 18 months, which is not a lot of time. But my bias is to think that immunotherapy would be better, because I see the banana-shaped curves, relapse-free survival curves, with the COMBI-AD trial.

And then I see the negative vemurafenib trial, which Carl Lewis reported at ESMO 2017. It was vemurafenib versus placebo, vem for a year versus placebo. And I think the MEK inhibitor does add, but you have to ask yourself, “Is the MEK inhibitor really going to change the natural history of the disease, or is it just — how much of a survival benefit is there going to be?”

DR LOVE: So this man you’re going to start on dabrafenib and trametinib?

DR DAUD: I think that if you could convince yourself that immunotherapy won’t work because it’s a PD-L1-negative tumor, I think that’s who I’d reserve for targeted therapy.

DR LOVE: Any thoughts about the challenges of getting a patient through a year of targeted therapy in the adjuvant setting?

DR DAUD: Yes, we put a fair number of patients on adjuvant dabrafenib/trametinib a few years ago. And it is a problem. And again, I think the fevers and chills — just my personal feeling is that you do have them more often in adjuvant patients, just like you have any side effect. I think it’s more noticeable when you don’t have metastatic cancer. Again, I think — just the earlier point we were talking about is that it would be nice to have a briefer regimen rather than putting everybody though this extensive costly treatment when there’s a potential that you could save some of those patients at the back end if they relapse.

Case: A 47-year-old man with previously treated BRAF mutation-positive metastatic melanoma receives vemurafenib/cobimetinib

DR LOVE: So to maybe get a little bit more into the challenges of using BRAF therapy, maybe you can talk about your 47-year-old patient. I don't know if you recall this. Back lesion, positive sentinel node, 2008. Recurred, started on ipi/nivo. Does that ring a bell?

DR DAUD: That rings a bell. Yes.

DR LOVE: Okay. So you want to present that patient?

DR DAUD: Sure. So this is a 47-year-old who had a back lesion and then had recurrence and had very extensive disease, including multiple sites of disease. And started on ipi/nivo, had rapidly progressed on ipi/nivo. And now the question is — and luckily was BRAF mutant. So we had that option, treated him with a BRAF and MEK inhibitor combination, but just had really bad fevers and chills, just couldn’t tolerate it.

I tried all kinds of maneuvers. I tried intermittent treatment. I tried giving him low-dose prednisone. I tried, like, doing 21 days on and 7 days off and tried taking him off the MEK inhibitor for some time and just various maneuvers that we do. And it didn’t work. Switched him over to vem/cobi, and he tolerated that much better. Now, I’ve done the reverse, too. But in this case, this was a gentleman who was happy using sunscreen and staying indoors for the major part of the day.

And for him, the fevers and chills really were just disabling. And he continues to have really good disease control. And given how extensive his disease was before — his LDH was, I want to say, 3,000 or something to start off with. So in my opinion, this is a real success. He’s now about a year and a half into treatment. And from what I can tell, tumors are generally stable. There’s been some slow progression. And I recently added a PD-1 to his regimen, hoping to get some control.

DR LOVE: You added it to the targeted therapy, or you stopped targeted therapy?

DR DAUD: Added it to the targeted therapy, but I think that’s not something that there’s any data support for. But I feel that if you stop targeted therapy in patients with very extensive, you can get such a tremendous blow-back that before immune therapy has a chance to work, you could probably be blown out of the water. And this is a patient who’s failed ipi/nivo, so I have reason to think that he doesn’t have a highly T cell-infiltrated tumor to begin with. So I don't know how likely it is that this’ll work. I’m just waiting to put him on a clinical trial that. That’s my plan B.

Indirect comparison of the activity and tolerability profiles of BRAF/MEK inhibitor combinations (dabrafenib/trametinib, vemurafenib/cobimetinib, encorafenib/binimetinib)

DR LOVE: So you wrote a paper in the Journal of Hematologic Oncology, “Indirect Comparison of Dabrafenib/Trametinib Versus Vemu/Cobi.” So maybe you can add in the third doublet, encorafenib and binimetinib, and kind of tell me in your own mind the way you see those 3 doublets and how you choose between them.

DR DAUD: Yes, that’s a great question. I think they’re relatively equivalent, I would say, in terms of response rate. Where I see the differences are just in the side-effect profiles. And from what I can tell with vem/cobi, you don’t have the fevers and chills as much. But you do have more skin toxicity. With dab/tram, you have — in general it’s tolerable, except for the fevers and chills, which in the first 6 months, anyway, they are a pretty serious issue for most patients, especially young, healthy patients who — minimal disease. They seem to have really high levels of dabrafenib. And they seem to have the high fevers and chills.

In terms of the 3 doublets, encorafenib and binimetinib, I think from my reading the literature, avoids the side effects of either one of those and has, at least numerically, a better PFS compared to either one.

What I read from the literature is that PPE seems to be the major issue with encorafenib/binimetinib.

The palmar-plantar — the red palm, so that’s a relatively trivial side effect compared to the fevers and chills and compared to, like, the extensive photosensitivity that you have with vemurafenib.

At SMR, there was one of these debate things. And I was on one side of a panel. And I think my take on the BRAF inhibitor field is that even today — this is about 5 years after they were introduced into clinical practice. I think we don’t understand today what the problem is with BRAF inhibitor resistance. Is it pharmacokinetic? Is it pharmacodynamic? Is it the development of resistant mutants, or is it something totally different, like just some kind of epigenetic regulation and the cells just hunker down into, like, this hibernation state and have EMT and just dedifferentiate and just outwait?

So it isn’t totally clear, because if you look at the last 5 years of publications on this, there’s been so many targets pointed out. Either it’s MEK or it’s NRAS. Either it’s IGF1 and not — they don’t seem to be reproduced in subsequent studies. So to me, it seems like in a certain percentage of cases where you might have resistance or you might have problems with one of the drugs, you might be able to salvage them by simply changing the BRAF and MEK inhibitor.

And again, these drugs haven’t been tested in that setting where if you are progressing off, let’s say, dab/tram, you can salvage them with vem/cobi. But I certainly have seen that people can have different responses to either one. You could treat patients with dab/tram, and you could have a certain degree of response. You change them over to a different inhibitor, and you could see further shrinkage of tumor. So I think having an extra choice is great.

Clinical outcomes for patients with metastatic uveal melanoma treated with anti-PD-1/PD-L1 antibodies

DR LOVE: So I want to ask you about uveal melanoma. I know you had a paper in Cancer. Can you talk about what we know about that and particularly about the use of checkpoint inhibitors?

DR DAUD: They seem to be refractory or resistant to checkpoint inhibitors in general, especially to single-agent PD-1. I think with the ipi/nivo combination, I think there was a retrospective review that showed that there might be about 15% response to ipi/nivo combination. But our paper looked at just single-agent PD-1. And I think we had only, like — I want to say 2 or 3 patients responding out of 60. And if you look carefully at those patients who were responding, they were atypical. They were patients who had prior chemotherapy or, in some cases, patients who didn’t have liver metastases. That’s typically what happens with uveal melanoma. They’re associated with liver metastases. But these were patients who atypically had lung metastases or some other metastases.

And it could be that the immune-suppressive microenvironment of the liver, combined with the fact that these tumors have low mutation burdens to start off with, you’re just stacking one immune suppression mechanism on top of another. And you end up with a very immune not-sensitive tumor.

DR LOVE: Do uveal melanomas have BRAF?

DR DAUD: No. They almost never have BRAF mutations. They rarely have NRAS mutations. The kinds of mutations they have are in these G proteins called GNA11 and GNAQ. And they activate the MAP kinase pathway, but not through RAF or RAS. But this is higher up on the cell membrane. These are receptors, G proteins that are apparently present in your uvea to a high degree.

Targeting NRAS mutations with CDK4/6 inhibitors in patients with melanoma

DR LOVE: Any other mutations or targetable mutations that look promising in melanoma? There was a lot of interest in NRAS. I see you have a paper looking at ribociclib. That’s kind of interesting, a CDK inhibitor we use in breast cancer with binimetinib. What do we know right now about NRAS disease?

DR DAUD: So NRAS disease is very interesting. Because of our problems targeting RAS in general, KRAS, NRAS, HRAS, to date nobody’s figured out the challenge of how do you get an inhibitor into a protein that’s so smooth? It just doesn’t have a lot of grooves and crevices in there. So these inhibitors don’t fit in.

So if you’re stuck trying to inhibit downstream of NRAS — so, which basically means MEK or PI3 kinase. And MEK and PI3 kinase cannot be inhibited together. So you have to focus on one of the two. And if you focus on MEK, you still have a lot of signaling downstream. And so people have used CDK4/6, which actually looks pretty effective. That combination looks pretty effective.

But I think we’ll have to look at bigger trials to see whether there’s a certain percentage of people who have long-term disease-free survival, like you have with immunotherapy or in general. Are you just inhibiting the disease for 4 to 6 months? Which, if that’s the case, I think it would be not as interesting.

Case: A 58-year-old man with metastatic BCC experiences a good response to vismodegib but discontinues treatment due to dysgeusia

DR LOVE: So let’s move on and talk a little bit about basal cell. And you have a 58-year-old patient I wanted to hear about.

DR DAUD: Yes. I remember this patient pretty well. He’s a guy who had a basal cell. And, like happens with a lot of basal cell patients, I think, may have waited too long to see the doctor. But he had axillary lymph node metastases and lung lesions.

DR LOVE: Where was the primary?

DR DAUD: Primary was on his shoulder.

DR LOVE: Was he in a social situation that was related to delay? Is that what you typically see?

DR DAUD: In his case, he’s actually, like, has great health insurance and is a government employee. I don’t believe it’s due to social factors. But I think it could have been anxiety about seeking medical care. It could be one of those type of things.

DR LOVE: How big was the primary when he sought —

DR DAUD: The primary basal cell had been resected when I saw him, but from reading the notes, it was about maybe half a centimeter in size.

DR LOVE: That’s all?

DR DAUD: Yes. So not the biggest basal cell in the world.

DR LOVE: And he had metastatic disease?

DR DAUD: Yes. What happened in his case is that he had a local-regional recurrence before it became metastatic. He had this axillary lymph node. And I think if that had been resected earlier, that might have — this is all hindsight, but when it was resected, he did have extranodal extension. And perhaps it was too late by that time. So I started him on vismodegib, yes. I used vismodegib and, he had a response both in his lung nodules and in his axillary metastases, but after a prolonged period of time, the ageusia that can happen from these hedgehog inhibitors can be a problem. In his case the muscle cramping wasn’t as big of a deal. That’s the other major side effect of these agents. His electrolytes were normal, but the ageusia really became a problem.

And after several months of trying to talk him into intermittent treatment — there’s some data with vismodegib and hedgehog inhibitors that you can get some good mileage by intermittent treatment — I tried giving him drug holidays, but ultimately it just proved too hard to do. And basically he stopped, decided to stop taking vismodegib.

And I still follow him and I still see him. And his tumor is growing larger. And hopefully I’ll be able to convince him to do something. I think he was looking for compassionate-use PD-1. And he was looking for immunotherapy trials, but to my knowledge he’s still not treated and he’s been about 6 months off treatment now.

DR LOVE: What do we know about checkpoint inhibitors in basal cell?

DR DAUD: There was a paper in squamous cell that showed some activity in squamous cell, and it was a PD-1 antibody. I want to say it was presented at ASCO.

DR LOVE: Right. But that was squamous.

DR DAUD: Squamous. Basal, what I’ve heard anecdotally through my colleagues is that people have noted responses with basal cell. But I don't know if an actual piece of data — that there’s a certain response rate. If I had to guess, I’d say there might be a response rate, because these are UV-induced tumors. They have a lot of mutations. If you follow that line of thinking, I think it’s possible that you could have some response rate.

DR LOVE: There is now another approved hedgehog inhibitor, sonidegib. What about using that in a patient like this?

DR DAUD: I mentioned sonidegib to him. And for whatever reason, when we tried to switch him over to sonidegib, the insurance that he was on denied that. And nowadays, insurance and appeals is becoming, like, a standard part of oncology. And I wasn’t able to convince the reviewers to allow that. And I think the problem was that I couldn’t show that he had actually progressed from vismodegib. And I think that seemed to be what the insurance company wanted, that evidence that the vismodegib had failed.

DR LOVE: Hmm. What did he actually describe to you in the ageusia?

DR DAUD: He almost made it sound like it was a painful sensation, not just that he wasn’t able to taste stuff but that he had this constant metallic almost like a painful sensation when he ate stuff, especially food that was a little spicy, which he previously was used to and enjoyed eating. Like, it almost seemed like it hurt him when he ate it. So that’s what he described that as.

Now, I’ve had a few patients with basal cell. It doesn’t seem to be a problem for everybody. I’ve had a patient who’s been now for 3 and a half years on. And we only see him every 6 months, because that’s the cycle that he’s established. And for him it’s the cramping that’s more of an issue, but it’s very tolerable.

Incidence and management of KIT-mutated melanoma

DR LOVE: So I want to move back to melanoma and ask you about KIT-mutated disease. I see you had a paper looking at nilotinib. But can you just review the whole issue of KIT-mutated melanoma?

DR DAUD: So KIT-mutated melanoma probably affects about maybe 2% to 4% of all melanomas. And most of these are mucosal or acral lentiginous melanomas. So most of them are not your standard garden-variety skin melanomas. There’s a small percentage on the head and neck in extremely sun-damaged skin. But for the most part in the United States, it’ll be acral or mucosal.

And there’s many different types of KIT mutations and there’s many different KIT inhibitors, starting from imatinib. And there is a response rate of about 10% to 20% with KIT inhibitors. And I can’t tell for sure if one KIT inhibitor is any better than any other KIT inhibitor. What the problem is with the KIT-mutant melanoma right now is that there’s a distinction between actual mutations in KIT, amplification and immunohistochemistry overexpression.

And it seems like the KIT inhibitors really only work in the mutated population or in the mutated-amplified population, not just amplification or not in just overexpression. So that’s a relatively small proportion of patients. And then to me, the other issue with the KIT inhibitor treatment currently is that the response rates are what they are. But the duration of response is not as — you’re talking immunotherapy duration of response. You’re talking a few months, 4 to 6 months.

So I think with all that being taken into consideration, I think most people start KIT-mutant melanomas with PD-1 or immune checkpoint combinations and then move on to other treatment. And I think that’s even true for these fusion melanomas. There’s a class of melanoma that involves gene fusions, either BRAF fusions, ROS, NTRK, ALK fusions. You can use the — and we have a clinical trial here at UCSF where we are looking at NTRK, ROS or MET inhibitors for this or the different-type BRAF inhibitors like regorafenib for fusions.

It’s the typical V600E BRAF mutation that you see in melanoma, the kinds of inhibitors that work for that, like vemurafenib or dabrafenib, do not work in these fusion melanomas, because you have intact BRAF. You don’t have that mutation. So you need an inhibitor that can work in that situation. And we are looking to see what the response rate is. And we are screening a large number of patients to try to find these fusion melanomas.

DR LOVE: And at what point do you recommend KIT testing in patients with melanoma?

DR DAUD: I think for patients who have mucosal melanoma or who have acral lentiginous melanoma, I think it makes sense to do KIT mutation testing. In patients with normal skin melanoma, I think I would only do that if you had exhausted immune checkpoint inhibitor therapy and BRAF and you’ve already tested for BRAF and it’s wild type. Because it’s likely that you’ll find BRAF or one of the more common mutations in those patients.

DR LOVE: Anything that we haven’t talked about today that you want to comment on, any new research going on, any clinical issues that you want to comment on?

DR DAUD: What I’m excited by is there’s this data showing that there’s these tertiary lymphoid structures that are present in colon cancer and breast cancer, in a lot of cancers. And what’s in the tertiary lymphoid structures is follicular helper T cells, B cell as well as T cells. And one of the interesting directions that I feel is worth exploring is trying to reproduce these tertiary lymphoid structures in your melanoma or in your treating cancer. And we think that if you have specific chemokines, you might be able to reproduce those TLS, tertiary lymphoid structures. And maybe that could turn some proportion of immune checkpoint-negative melanomas to positive ones.

Results of the Phase I/II ECHO-202/KEYNOTE-037 trial of the IDO inhibitor epacadostat in combination with pembrolizumab in advanced melanoma

DR LUKE: The field of immune checkpoint blockade for melanoma and for all cancers is rapidly advancing. And the backbone of therapy in that sort of setting is really PD-1 antibodies. But the next generation is coming forward, and it’s not totally clear on which targets would be most ideal to combine with PD-1.

And at the University of Chicago, and a number of other investigators have really been interested in the interaction between the tumor infiltrating lymphocyte and the tumor in the tumor microenvironment. And so we’re aware that when T cells get activated in the tumor microenvironment, they elaborate the cytokine interferon gamma. And that’s a particularly important cytokine for driving the productive antitumor immune response.

In response to interferon gamma, tumors can upregulate PD-L1, which is why we think PD-1-blocking antibodies have their activity. It turns out that other pathways are also upregulated in the context of an interferon gamma-associated response. And particularly, one that’s important is indoleamine dioxygenase. And so multiple inhibitors of indoleamine dioxygenase have come forward. And the most advanced of them is epacadostat. And so there’s a rationale then, in the context of tumor infiltrating lymphocytes, elaborating interferon gamma and upregulating PD-L1 and IDO together. It then makes sense to combine pembrolizumab, a PD-1 antibody, with epacadostat and IDO inhibitors.

And that is really the basis for the Phase I/II clinical trial that was presented at ESMO. It’s been presented at multiple meetings prior to this, where we’ve observed that, whereas with a monotherapy PD-1 antibody we’d expect a response rate on the order of 40%, here with this combination, albeit in a single-arm, nonrandomized controlled study, we see a response rate of 60%, on the order of 60%.

Similarly, whereas we see the PFS with monotherapy PD-1 antibody at about 6 months, with the combination we’re seeing things out more closer to a year. So again, these are small data sets, but they really suggest that there could be a powerful combination regimen with PD-1 plus IDO inhibition.

Mechanism of action and safety profile of epacadostat alone or in combination with an immune checkpoint inhibitor

DR LOVE: So I’m trying to kind of remember what it means when something ends with “stat.” And so, like, it’s a small molecule, right?

DR LUKE: Correct.

DR LOVE: So what kind of small molecule? And is there any toxicity?

DR LUKE: So epacadostat, or most of the IDO inhibitors as a class, inhibit the enzyme IDO1. So it is a small molecule inhibitor of the active site in the enzyme that converts tryptophan, the amino acid, into kynurenine. And that process, it turns out — we’re learning more and more — appears to be very specific to the tumor microenvironment or an interferon gamma response. So you would be concerned that if you are adding checkpoints you might get more toxicity, but it turns out that in the clinical trials to date, both for the combination of pembro and epic but also for other IDO inhibitors, there really is no suggestion of increased toxicity relative to what you would see with a PD-1 antibody alone.

And we think that’s because it’s this interferon gamma-associated T-cell activation that’s really only present in the tumor microenvironment, so that by inhibiting this enzyme, it’s not really of relevance anywhere else in the body and, therefore, you’re not generating more systemic toxicity.

DR LOVE: Just thinking about it, I hadn’t thought about this before. But from your point of view, let’s maybe just talk about, say, an anti-PD-1 antibody to start with and then maybe get into these other things. Do you think that essentially all the toxicity is autoimmune? Or do you think there’s, like, some more traditional type of toxicity that you see with these agents?

DR LUKE: When thinking about toxicity in the context of checkpoint blockade, it is important to recognize that there could be 2 categories. There’s just straight-up drug toxicity like we’ve historically thought of for chemotherapy, but really, the major toxicities we worry about are these immune-related adverse events.

DR LOVE: So, I mean, for example, I hear about these skin changes. Do you think they’re a drug type of effect, or autoimmune?

DR LUKE: So dermatitis or cutaneous eruptions in the context of checkpoint blockade have actually been shown to mostly be immune related. So you can do biopsies of the skin and see infiltration of lymphocytes. And that’s commonly thought of as the most likely source. Now, whether or not there’s also an aspect of just straight-up drug toxicity, I think that’s a little bit of an open question. But broadly speaking for PD-1 antibodies, that’s not considered to be the major mechanism.

I will note, however, that, as we bring other combination partners forward, like, say, a small-molecule IDO inhibitor, we will need to be cognizant of everything else that we learned about cancer drug development over the past several decades, that these drugs, in and of themselves, could have direct toxicities.

So the one toxicity that was observed to perhaps be slightly increased with the combination of pembro plus epacadostat was dermatitis. And there it’s kind of an open question as to whether or not that’s truly an immune-related issue or a drug-specific issue. Along those same lines, there’s recent data of nivolumab plus the IDO inhibitor known as BMS-986205. And in that combination, they saw a slight increase in hepatic toxicity, again, only slight, but again raises this question of are there drug-specific properties or is this an immune-related adverse event?

I think some of these questions are really only going to be teased out in the Phase III trials that are starting to be launched now, where you’re going to have randomized populations who either get PD-1 monotherapy or the combo and to really try to look at both efficacy and toxicity there. But it is a wide-open question as we get combination regimens. Pulling apart the different toxicity profiles will be difficult between the different agents.

DR LOVE: Yes. And thinking about it also in terms of, for example, epacadostat, it sounds like you’re saying that you don’t see an obvious increase in autoimmune problems. Is that the case?

DR LUKE: Yes, that’s correct.

DR LOVE: And then the other — so the related question is, why do you see so many problems with anti-CTLA-4 agents like ipilimumab? What’s the biology of why you see worse toxicity?

DR LUKE: Right. So that’s a really good point. And so the difference is that blockade of CTLA-4, dogmatically speaking, has been described as being a peripheral immune effect, meaning at the point of T-cell activation, CTLA-4 acts to dampen initial T-cell activation. So you get a much more global activation of the adaptive immune response in the context of CTLA-4 blockade. That’s in contrast to the interferon gamma-associated tumor microenvironmental changes. So if a T cell upregulates gamma in the tumor microenvironment and you get PD-L1 and IDO upregulated, that effect, we believe, is really only happening in the context of the tumor microenvironment and, therefore, we have less of an off-target — in quotes — effect of immune peripheral immune activation.

Emerging role of LAG-3 and TIM-3 inhibition in immune checkpoint blockade strategies

DR LOVE: It’s funny. Sometimes I ask people questions and I just assume somebody out there smarter than me is going to understand what you’re talking about. But I do get some of it.

Anyhow, one other thing about other immune targets, again, just hearing people talk, I’ve heard things about other checkpoints like so-called TIM-3 and LAG-3. What are they? And what are some of the other immune strategies that particularly integrate with checkpoint inhibitors that you think are most promising?

DR LUKE: Absolutely. So as we really drill down and focus on what’s happening in the tumor microenvironment at the point of attack between a T cell and the tumor, we recognize that it’s not all just PD-L1. But it does all stem off of this biology of the T cell elaborating cytokines and the tumor doing things in response. So one of those is IDO, as we talked about, but there are other things that happen to that T cell as it sits in the context of the tumor. And just like in autoimmunity, where you have an antigen that the immune system can’t clear, your T cells, if they’re in the tumor microenvironment will become progressively dysfunctional.

And that can be observed in the context of which receptors are shown on the surface of the cell. And so one that’s become especially of interest is one known as LAG-3. That molecule is upregulated both in parallel but also in sequence with PD-1. And it’s another marker of T-cell energy or, quote, exhaustions, close quote.

And so the idea then is, could we block LAG-3 and potentially reverse that exhausted state and get those T cells to be functional in the tumor microenvironment again? And data has been presented now at multiple meetings for the anti-LAG-3 antibody relatlimab. And there are Phase III clinical trials of nivolumab plus relatlimab coming forward for refractory melanoma and, I think, in other disease settings as well.

LAG-3 is not the only T-cell exhaustion checkpoint, however. Another that’s of interest is TIM-3. And there are a number of other receptors that are also of interest, like OX40, CD137 and a host of other ones that are also targets for drug development in terms of trying to relieve this exhausted state or to try to push T cells to be more functional.

ADVISE: A planned Phase I adaptive study to match patients with solid tumors to various immunotherapy combinations based on biomarker assessment

DR LOVE: How close are we to maybe moving into a situation where we’re going to know which of these to use, the old personalized oncology concept? Do you think that’s coming to immunotherapy?

DR LUKE: So I absolutely believe that this concept of personalized immuno-oncology is coming. And I think it’s coming sooner than people think about. Currently, we’re treating populations of patients and retrospectively doing immunohistochemistries. But I can tell you that I’m going to be the principal investigator for a trial that we’re calling ADVISE, or the adaptive trial after nivolumab resistance, to look at specifically prospectively profiling tumors by immunohistochemistry to look for which receptors are upregulated and then put patients into buckets based on what we find on their immunohistochemistries.

DR LOVE: I think there’s some stuff, like, in lung cancer that did it that way.

DR LUKE: So the trial you’re referencing was at the dawn of targeted therapy. There was a study known as the BATTLE trial that was done at MD Anderson.

DR LOVE: And Roy Herbst — yes.

DR LUKE: Right. So it’s a very similar sort of concept, but beyond just that, we’re excited about this because we think not only could we profile by immunohistochemistry, but could we actually even drill down and then do, in parallel, gene expression profiling to look at RNA-based genomics? Because maybe in the future, we could do genomics instead of immunohistochemistries, because that would really open up the world to be a much more robust biomarker.

DR LOVE: What’s the name of the study again?

DR LUKE: The ADVISE.

Diagnostic comparison of CT scans and colonoscopy for immune-related colitis in patients with ipilimumab-treated advanced melanoma

DR LOVE: Okay. Another paper I want to ask you about, this little informal journal club, is a paper you were involved with, “Diagnostic Comparison of CT Scans and Colonoscopy for Immune-Related Colitis in Ipilimumab-Treated Advanced Melanoma Patients.” So I’m always curious about colitis in general, not just so much with ipi or CTLA-4, but you hear about it with checkpoint inhibitors sometimes.

I mean, we’re hearing about bladder cancer. Everybody’s using checkpoint inhibitors now. And it doesn’t seem that common without the anti-CTLA-4, but you hear about it. And I’m very interested in what you looked at here to further understand really how these should be managed.

DR LUKE: Absolutely. So management of immune-related adverse events is probably the most difficult aspect of immunotherapy for practicing physicians who don’t have a background in this. And I would say that it’s really the opportunity for us to put our internal medicine differential diagnosis hats back on.

But in this investigation, what we looked at was patients who presented with diarrhea who then were admitted to the hospital and the comparison between those that had colonoscopies to get histology compared to the CT scan findings when they came in. So on the clinical trials of the various checkpoint inhibitors, it was very much recommended that if people got sick that we were very aggressive in trying to determine the etiology of what happened. And, therefore, on the ipilimumab clinical trials almost everyone got a colonoscopy.

As these checkpoint antibodies have exploded and moved into clinical trials all over the place, it’s become less frequent that people do the colonoscopy and they’re more likely to treat empirically. But an open question then was, like, how much of this are we really missing? And I think that this was a useful investigation, because what was shown was that in the context of patients at the Dana-Farber treated with ipilimumab, that those that presented and had a CT scan that showed stranding or evidence of a colitis process by radiomics, almost 100% of the time, it was above 95% of the time, had colitis when they had a scope.

So to the point that then in general practice, if there’s a high pretest probability that the drug is causing this problem, this diarrhea syndrome, then it’s probably unnecessary to do a colonoscopy in all of these patients. The key caveat that I would make there, however, though, is, there are some investigations that you can do without admitting a patient that really should be done.

So many of our patients are exposed to antibiotics or they’re in the hospital or around a healthcare setting quite often. So it’s really essential that practicing physicians check for things like C diff and they ask patients just general history elements like, “Has anybody else in your family been sick?” and so on and so forth, because it does happen that we get confused and that occasionally we do these colonoscopies and we do find the patient actually has C. diff or something else. And you don’t want to miss that, because if you just start someone on steroids, it could potentially do 2 problematic things.

One is, steroids are by no means benign. They cause a lot of side effects in and of themselves. And secondarily, if you confuse the situation and you tell the person that they’re refractory to immunotherapy, depending on what the disease setting is, you may have eliminated a very significant treatment option for that person. So I tend to be rather dogmatic about it. I’ll check C. diff and some of these things.

Despite this paper, I still have a lower threshold for really doing a colonoscopy, especially if I start outpatient steroids and patients don’t get better very quickly. But generally speaking, if your pretest probability is high that this drug is causing this problem and you do some relatively minimal investigations in the outpatient setting, you can feel confident to give steroids in this setting.

DR LOVE: What is it that they see on colonoscopy that’s useful?

DR LUKE: So you’ll see, generally speaking — and it depends on — because the spectrum of effect for colitis can be quite broad. If you catch it early, you’ll really just see erythematous pink mucosa that just looks inflamed. If it goes on for more than a week or two, however, you can start to see patches of — holes in the colon, quite honestly. So colitis that really is histologically obvious but even clinically is obvious that the colon just looks really beat up, just like inflammatory bowel disease.

And if you do cold forceps punch, you can see penetration into the epithelial layers of various different types of immune cells, both CD4 and CD8 T cells, as well as macrophages. And it can really be a range of what you can see, anything from microscopic colitis to full-on inflammatory bowel disease-like histology.

DR LOVE: But you can learn something from the histology?

DR LUKE: Yes, because sometimes, actually, when we do a colonoscopy in these patients, we actually don’t see evidence of friable mucosa that is just super obvious from a clinical perspective. And so I tell my colleagues in gastroenterology that even if they go in there and they’re not convinced just by visual inspection, I ask them to please do random biopsies.

Because it’s commonly the case, actually, that we will pick up in the histology, the presence of this colitis syndrome, which, if we didn’t know it was going on, we might not have done that. And we would still be stuck with a patient who has diarrhea, but we don’t know how to treat them.

DR LOVE: Maybe everybody knows this, but I don’t recall anybody mentioning the idea that CT scan can suggest or make the diagnosis of immune colitis or colitis more generically.

DR LUKE: Yes. So colitis more generally, you can definitely pick up stranding or inflammation in bowel segments on a CT scan. And so then again, it goes into your general internal medicine practice, in that if you have a high pretest probability for colitis and you get a CT scan that shows inflammation, you can put 2 and 2 together there. Again, it requires, though — there are other reasons that people could have colitis, so you don’t want to miss something else that would be treated in a different fashion.

Underlying pathobiology leading to immune-related adverse events in patients receiving immunotherapy

DR LOVE: And in general, do you view colitis that comes as a secondary effect of, for example, anti-PD-1 therapy alone as qualitatively the same as what you see with anti-CTLA-4, just more of it with CTLA-4 or the combination?

DR LUKE: Yes. So that’s the right way to think about it. These are class effects. And it’s just they’re incidents that’s different. So with CTLA-4 blockade, it’s obviously greater than with PD-1 blockade. And with the combo together, it’s even higher. But yes, generally speaking, they are the same.

Now, the intensity of them will also vary, so that with ipilimumab alone, the intensity of the colitis is commonly greater than with PD-1 blockade. And certainly with the combination when patients get colitis, we’re going to be more likely to be aggressive with anti-TNF agents and superhigh doses of steroids very quickly, because this is the kind of thing that can really spin and become problematic.

DR LOVE: Do you think that some of these patients, or maybe a lot of them, have underlying autoimmune bowel disease that’s being reactivated that’s subclinical, is being made clinical? Or do you think these are totally normal bowels that become diseased?

DR LUKE: So it’s a very good question about what is the underlying pathophysiology that leads to immune-related adverse events. And that’s actually a major area of research in the field. I actually agree with what you suggested surrounding the idea that some patients have — you might think about it as a higher immunological rheostat in terms of if they’re just primed a little, they’re ready to go into an autoimmune state.

So there are T cells that are antigen specific for something in the bowel that are normally controlled by your natural mechanisms to suppress autoimmunity, but by just relieving these checkpoints, that’s enough to get them into play, such that you start to get clinical symptoms. Now, how frequent that is just in the general population is unknown. We would call those people normal, but then in the context of this intervention that we made against cancer, they become abnormal. So I think it’s a little bit hard, because I don’t know that we necessarily have the right categories for this, just in the way we usually think about clinical medicine.

Activity of immune checkpoint inhibitors and targeted therapies in patients with advanced melanoma and brain metastases

DR LOVE: So one other question related to some recent papers, but just a topic I want to throw out at you is the issue of brain mets with melanoma. And I see that there were a couple of papers at ASCO looking at checkpoint inhibitors in people with brain mets. But also I’m curious what we know about responsiveness of brain mets to PD-1 antibodies alone and also targeted therapy and how you integrate systemic therapy into their management.

DR LUKE: So absolutely. Brain metastases in melanoma is really the last frontier, where we haven’t made the kind of progress that we wish that we had, although we’re starting to get there. It’s been known for a long time that BRAF inhibitors as monotherapies, even when it was just vemurafenib, had activity in brain mets, though it was attenuated relative to what we saw systemically. Similarly, we know that PD-1 antibodies do have a response rate in the brain, though again, it’s not kind of what we would want.

So at ASCO 2017, there were a couple of different presentations. One was on dabrafenib/trametinib, specifically in the population of patients with brain metastases. And what we saw there was that the response rate looked quite impressive. So systemically, we looked for a 60% to 80% response rate. And that’s similar to what was seen in the brain metastases population.

Now a difference, however, was that whereas in cutaneous disease in the Phase III trials the progression-free survival was on the order of about 11 to 12 months, in the brain metastases population that progression-free survival was more about 6 months. So while it’s still a very useful therapy, it definitely was attenuated in terms of long-term benefit.

Put that in contrast, however, with the ipilimumab plus nivolumab immunotherapy combination, which showed, again, a similar response rate in brain to what was seen systemically on the order of 40% to 50%, the difference being with that combination, those responses were more durable. And so very similar to what we see with systemic cutaneous disease, those patients with brain metastases who do respond to ipi/nivo seem to go into long-term response.

And now, the follow-up on these patients is not as long as we would wish, but we’re getting over a year. And it’s very likely to be the case that much like we see with other immunological phenomenon, when we get true responses, those patients go into durable disease control.

Now, the caveat is, ipi/nivo for brain metastases is toxic, meaning that patients do get pretty sick from it. I’d say that in my practice, however, this is one of these situations where we may not have multiple shots on goal. And so I tend to use PD-1 monotherapy for most patients for systemic disease who kind of are in a reasonable place clinically. But for brain mets, this is a situation where you pull everything out and you just go heavy hitter right away, because you may not get another chance.

DR LOVE: Do you use PD-L1 assays in making the decision?

DR LUKE: So I, generally speaking, don’t use a PD-L1 assay in melanoma. When I say “generally speaking,” I don’t ever do it. And the rationale there is that that actually doesn’t inform my decision-making. And so when I think about for melanoma monotherapy PD-1 versus a combo PD-1/CTLA-4, what I’m really asking myself is, what are the adverse prognostic features of this patient, and/or would this patient have the potential to get these agents in sequence?

So there are some data, although they’re not as robust as we wished, for PD-1 followed by ipilimumab. And those data suggest in small series that the impact is approximately what you get for a combo, both in terms of response rate after both drugs together as well as toxicity. So then from my clinical perspective, I’d rather then expose someone to a monotherapy PD-1, because if they respond, then I don’t have to give them ipi.

But in those patients who have adverse prognostic features like brain metastases, like bone metastases, like rising LDH and poor performance status, there I want to come in as quickly as possible, because if we don’t get something going, then you’re never going to get the chance to give 2 drugs in sequence over a 6-month period. You’ve got to get something happening immediately.

DR LOVE: What about in the younger patient, highly motivated but otherwise wouldn’t fit the criteria of being a concern, no brain mets, et cetera?

DR LUKE: Yes. So I think that that’s a patient-level conversation to have with them. I personally don’t think that just because someone’s young that means we necessarily need to expose them to more toxicity. There are certainly patients who are of that opinion, that, “Just give me everything you’ve got at once.” But if you explain to them that “I believe that PD-1 followed by CTLA-4 is probably roughly equivalent to the combo up front,” then again, the rationale is, why expose yourself to toxicity if you’re going to get the opportunity to get that benefit over time?

And I think that this is probably where it’s important to emphasize that some of these next-generation combination immunotherapies really look the most exciting. So a PD-1 plus IDO inhibitor may end up being roughly equivalent to ipi/nivo, but not with the side-effect profile that we’ve been talking about.

DR LOVE: Yes. That would be great.

Role of radiation therapy in the treatment algorithm for patients with melanoma and brain metastases

DR LOVE: One final question about brain mets: How does radiation therapy, whether it’s stereotactic or whole brain, fit into your treatment approach?

DR LUKE: So this is an area of controversy. So the data for ipi/nivo actually looked quite good. And, unfortunately, our data for radiation in melanoma brain metastases is not particularly good. And so there as a community, we are all in agreement that whole brain radiation really should not be used for advanced melanoma until the situation where there’s really just no other option and, rather, stereotactic approaches should be preferred.

But that being said, even stereotactic approaches can lead to radionecrosis over time, especially in the context of checkpoint blockade, as there are patients now getting out 2, 3, 5 years. There, the incidence of radionecrosis is going up. So in the, quote, old days, just a few years ago, we didn’t have this problem, because unfortunately patients didn’t live long enough to get it. So many of us now believe that we should really be trying to use systemic therapy with combination immunotherapy for patients if they have low-volume brain metastases and there’s no urgent need to radiate, that we really ought to consider systemic therapy first.

And this is a whole area of investigation. And there are clinical trials ongoing to try to look at combo, at sequence and so on and so forth. And I think that this is an open area where we’ll just need to learn more about it over time.

DR LOVE: What’s your vision of the, quote, radiosensitivity of melanoma, let’s say, compared to other solid tumors, breast, lung cancer or maybe hematologic cancers, lymphomas, et cetera?

DR LUKE: Yes. So radiosensitivity as a concept is not something that we generally employ in melanoma. When we think about consolidative radiotherapy, we can say carte blanche that that does not work in melanoma. So it’s never the case, let’s say, you didn’t get a negative margin on a surgery, you would never ever radiate that with the idea that you would clear that margin.

So going to your point, we do do consolidative radiation in breast cancer, in head and neck cancers and some other cancers. And that’s been shown to have an impact on survival that’s useful over time. That is not the case in melanoma.

That being said, melanoma, like other tumors, is amenable to high fraction radiation. So if you do an SBRT to a specific lesion, you can get treatment effect. So it’s not that radiation doesn’t work at all, but rather that for all patients, it’s not something that we generally think of.

DR LOVE: But, I mean, is it true that the melanoma cell itself is more resistant, is harder to kill, so to speak, than a breast cancer, or not necessarily?

DR LUKE: Via DNA double-strand breaks that are induced by radiation, yes, a melanoma cell is more difficult to kill than an adenocarcinoma-like breast cancer. But again, I think there’s more biology under there. I think there are probably ways we could structure our radiation with immunotherapy approaches to optimize that. But generally speaking, we don’t do it.

Risk factors, incidence and mortality rates of melanoma and nonmelanoma skin cancers

DR LOVE: What do we know about the risk factors and the incidence of melanoma and nonmelanoma skin cancers?

DR LUKE: So when we think about skin cancer as oncologists, we’re often thinking about melanoma, because that’s the cancer we often think about treating with systemic therapy. But it’s important to point out that melanoma is actually much less common than other types of skin cancer, such as basal cell carcinomas and squamous cell carcinomas.

So whereas melanoma is increasing in incidence, it’s still only on the order of about 80,000 cases annually in 2016/2017. And that really pales in comparison to squamous carcinomas, which are now approaching 300,000 to 400,000 cases annually and basal cell carcinomas, which are now over a million cases annually. So in terms of the public health problem of skin cancer, the squamous and basal cell carcinomas are really a much more pervasive problem than melanoma.

DR LOVE: And it’s interesting, because, I mean, that’s not an inconsequential number of deaths from basal cell and squamous cell, almost 1,000 deaths. I’m kind of surprised at that.

DR LUKE: Yep.

DR LOVE: People die of basal cell a lot?

DR LUKE: Not a lot, but —

DR LOVE: I mean, a thousand cases, that’s more than, like, a lot of T-cell lymphomas. Just kind of interesting. Okay.

DR LUKE: Yes. And when we think about the risk factors associated with the development of skin cancer, everyone understands that sun exposure is the major risk factor. But that being said, it is slightly different in that with melanoma, we commonly think of high-intensity sunburns, especially when people are younger, leading to long-term damage that can become melanoma, whereas with other forms of skin cancer, it’s really the chronic sun exposure over the course of one’s life that leads to these lesions.

So in melanoma, we commonly see a younger patient population who can develop advanced disease, whereas, with basal and squam, it’s very much the case that the median age of diagnosis is now in the late sixties, kind of, and seventies range, where people have had chronic sun damage over the course of their entire life. And most commonly, these lesions actually end up being small lesions that can be removed in a dermatologist’s office. But people who have 1 squamous or basal cell lesion very commonly have more than 1, because it’s really a phenotype of a person who’s spent a lot of time in the sun and who has chronic sun damage over all of their sun-exposed skin.

DR LOVE: In terms of squamous cell, how many of these cases are related to transplant or immune suppression?

DR LUKE: So really all skin cancers can be associated with immunosuppression. And in terms of the overall incidence or burden of squamous or basal, or even melanoma in the context of transplants, it’s quite low, actually, relative to the global burden of the disease in the population. But it’s certainly the case that physicians should be aware that if a patient had a transplant or has been immunosuppressed, knowing that that could be contributing to the development or the aggressiveness of their skin-based lesion.

ONTRAC: Results of a Phase III trial of nicotinamide for nonmelanoma skin cancer chemoprevention

DR LOVE: Can you discuss the ONTRAC study of nicotinamide for prevention of nonmelanoma skin cancers?

DR LUKE: So there has been major progress in the area of public health around prevention in the context of a nonmelanoma skin cancer. That was presented a few years ago at ASCO and published in the New England Journal. It was this ONTRAC study where, from Australia, they randomized patients to either get a placebo or to receive the vitamin nicotinamide at 500 mg twice a day. And this is really a very, very useful study.

And when we think about using vitamins, they often get thrown around that there’s no dose consistency. They’re not followed properly. But in this study, it was quite a rigorous prospective study. And they really showed a dramatic reduction in the incidence of basal and squamous skin cancers — again, not melanoma, basal and squamous cancers — with the use of nicotinamide, 500 mg BID, again showing that the relative reduction was on the order of about 20%. And this was effective both for basals and squams. The hazard was actually 0.77.

DR LOVE: So it’s 23% reduction.

DR LUKE: Yes, 23% reduction. But when you think about the overall disease burden, we were thinking about from the epidemiological perspective and relative to other kinds of cancer, that’s a dramatic decrease in the overall disease burden in the population. And I think the most important part of this, perhaps, is that nicotinamide is really not toxic at all. It’s a vitamin supplement you can take. It’s not going to cause any problems of any significant threshold.

Now, in this study, it was important to point out that patients had to have at least 2 nonmelanoma skin cancers to be eligible for this study. But again, for the population of patients who have chronic sun damage, that’s a fairly common incidence. And so if you had more than 1 or 2 squamous or basal cell cancers, taking nicotinamide is a very reasonable consideration and is now compendium-listed for major cancer societies.

DR LOVE: That’s really fascinating. Any evidence or reason to think it would be helpful for people with an established cancer or even metastatic disease? What’s the mechanism, incidentally?

DR LUKE: I’m not sure the mechanism is completely worked out. I think it is around immune surveillance, but I don’t know that that’s been elucidated to the degree that we’d want to comment on it absolutely. In terms of taking it in the metastatic setting, I think there’s little harm, although to date we don’t have evidence that says that that would augment the treatment effect.

DR LOVE: That’s really interesting.

Pathophysiology of BCC

DR LOVE: You want to talk a little bit about basal cell?

DR LUKE: Yes. So in the nonmelanoma skin cancer world, basal cells and squamous cell carcinomas, again, are less frequent but can be quite devastating. And so for basal cell carcinomas, we think about the basal cells are our cells that sit just below the epidermis, in the dermis. And they’re really pluripotent progenitor cells that can do quite a few things.

They generally tend to have less metastatic potential but can really cause quite problematic things in terms of local aggression. And generally speaking, they can be removed, but upon recurrence or lack of complete excision, they can start to cause severe dermatological problems both from a cosmetic as well as a functional perspective.

And so when we think about the biology of basal cell carcinoma, it’s interesting to look at the syndrome known as basal cell nevus syndrome, which really can help us understand why do these things happen. So basal cell nevus syndrome is an autosomal dominant syndrome that has a prevalence on the order of about 1 in 50 to 250,000 people. And these are folks that develop really hundreds of basal cell carcinomas all over their body.

These lesions do look very similar to sporadic or sun-induced basal cell carcinomas. And on a molecular level, these lesions have been identified to have PTCH mutations as the most common lesion that leads to this syndrome. So germline mutation in PTCH will lead to this recurrent basal cell carcinoma syndrome. And that’s very interesting then, when you look at what’s going on with sporadic or sun-induced basal cell carcinomas, where we see, actually, very similar biology, in that the majority of basal cell carcinomas that are in the context of sun exposure have a similar genetic lesion.

Now, it can happen in different ways, but you can get activating mutations in this pathway, in smoothened, which regulates PTCH, in about 10% or 20% of basal cell carcinomas. But in others, you can have epigenetic regulation and other ways to really impact on this pathway. And that really suggests that this hedgehog pathway, which is the name of this overall smoothened PTCH pathway, really regulates this basal cell carcinoma phenomenon in multiple settings.

DR LOVE: Can I just ask a simple question, which is, why it’s called basal cell nevus syndrome? Are they nevuses?

DR LUKE: So the lesions that develop are nevi like, and these patients don’t tend to die from basal cell nevus syndrome immediately, meaning they will have these lesions all over their body for extended periods of time without developing metastatic disease. And so that’s why it’s referred to as nevus syndrome.

DR LOVE: Interesting.

Role of the hedgehog signaling pathway inhibitors in BCC

DR LOVE: You want to talk a little bit more about the biology here?

DR LUKE: Sure. So hedgehog signaling pathway. Under wild-type conditions, PTCH interacts with smoothened to shut down signaling through this pathway. But when PTCH is not regulating smoothened, you get downstream activation of these different genes, including GLE proteins that eventually lead to a transcription factor to activation of hedgehog target pathway genes.

The drug development in this space is evaluating molecules that impact on the various different signaling nodes in this pathway to interrupt signaling here, such that we might be able to shut down hedgehog signaling and lead to a reduction of hedgehog-driven cancers like basal cell carcinoma.

And multiple hedgehog inhibitors have come forward into clinic now. There are 2 of them. One is vismodegib. The Phase II study of vismodegib. In the metastatic basal cell carcinoma population as well as with the locally advanced basal cell carcinoma population and, as might be expected, the therapy is more efficacious in the locally advanced setting than in the metastatic setting.

But in both, you can see rather dramatic responses at more than 30% tumor reduction in the locally advanced setting in more than 50% of patients and, in the metastatic setting, in more of a 20% to 30% therapy setting. Now, that’s quite important, because other therapies, systemic therapies for advanced or locally advanced disease for basal cell carcinoma really are of limited utility. We don’t have particularly effective systemic therapies. We do use chemotherapy. However, it tends to be quite toxic without necessarily a big impact.

The other drug, a hedgehog inhibitor, that’s available in clinical practice is sonidegib. And again, the sample sizes are smaller in these investigations. However, the data sets look fairly similar, such that in the locally advanced basal cell carcinoma setting, you can see response rates to sonidegib of about 50%, and again then in the metastatic setting it’s lower, again on the order of about 20-ish-percent. But again, both of these agents being quite useful to manage basal cell carcinoma when it cannot be removed by surgery.

Side-effect profiles of hedgehog inhibitors

DR LUKE: Now, the issue with hedgehog inhibitor is that they are not the most easily tolerated drugs to take. And the adverse events associated with sonidegib, these are really class effects for hedgehog inhibitors.

Patients will complain of muscle spasms, hair loss, dysgeusia or abdominal discomfort, nausea, et cetera, so really GI-like symptoms that can develop. And these can really be, unfortunately, quite dose limiting, meaning that patients will have a difficult time maintaining oral hydration and food intake in the context of these drugs. And it’s very much the case that sometimes drug interruption is required or treatment holiday, because patients just have difficulty tolerating that out over an extended period of time.

DR LOVE: Anything particularly typical about the muscle spasms? When do they occur, and is there anything you can do about it?

DR LUKE: I think that managing the muscle spasms is really a patient-per-patient sort of thing. We see people who actually take this drug and don’t get them at all. And we see other people, though, that have them quite dramatic. And so we recommend NSAIDs. We recommend trying to sometimes take just short treatment breaks. And exercise can sometimes help with those as well.

DR LOVE: Any comment about this issue of survival?

DR LUKE: These were not randomized controlled studies that have led to the approval of these agents. So it’s been an open question and debate in the field about the long-term efficacy of hedgehog inhibitors. And multiple case series have been presented and really compared with historical series to try to make the case that there is an impact on long-term survival.

A Kaplan-Meier estimate for overall survival in the vismodegib clinical trials demonstrating a 1-year landmark survival at about 85%. If you take the largest series from the literature, we would expect only approximately 50% to 60% of patients to still be alive at 1 year. So while these are not a randomized and direct comparison, it does appear that there’s a benefit to those patients who are taking hedgehog inhibitors in terms of improving overall survival.

DR LOVE: I don't know how often in your own practice you actually see a patient or you treat a patient with one of these agents. Have you had patients, like, in the last year or so die? What do people die from? Do they get mets?

DR LUKE: So there unfortunately are a couple of different scenarios for patients who have locally advanced basal cell or squamous cancers. One can be that the locally invasive aspect eventually can become metastatic. And that most commonly leads to lung metastases, somewhat like a sarcoma or a head and neck cancer, where eventually just the tumor burden overwhelms the patient. But also very commonly, unfortunately, are the problems with management of the local invasion of the lesion.

So I have a patient that I’m remembering who had a basal cell on his temple and, unfortunately, that lesion was quite aggressive, eventually eroding into his skull. And so management of that area was very difficult. It became a radiotherapy-refractory lesion, and that area would eventually get infected. And we had the person on long-term antibiotics. And eventually you have breakthrough even with the antibiotics and you get meningitis, et cetera, and you can really lead to infectious consequences for these patients.

DR LOVE: I know I’ve heard from a lot of people about the challenges of using these drugs. You mentioned the toxicities and the fact that it’s pretty hard to keep people on them for a really long time, trying to use intermittent treatment, et cetera. But is there also evidence that they actually develop a pharmacologic resistance, or is it just you can’t get the hedgehog inhibitors into them clinically?

DR LUKE: Oh, no. It’s very much the case that resistance will eventually develop. And the median time on treatment for most patients is less than a year. And that’s actually, generally speaking, due to treatment resistance. Often what will happen with these drugs is that patients will have a complaint about how they’re having difficulty with the drugs, and you take a treatment break. But the tumors will come back and, therefore, the patients are more amenable to try to find a way to fight through. Eventually, unfortunately, however, similar to many if not all targeted therapies, we do eventually see treatment resistance. And at that point, then the disease usually takes off and starts to grow much more aggressively.

Case: A 62-year-old man receives vismodegib for locally recurrent, unresectable BCC

DR LOVE: So why don’t we hear about a case from your practice? You have a 62-year-old man with basal cell. What happened there?

DR LUKE: Yes. So this individual was — I mean, it could be just about any of us, slightly overweight with diabetes and really had had a long history of sun exposure. And he developed a lesion on his temple, which — he observed it. And his wife was nagging at him to try to get it removed, eventually did see a dermatologist and had it biopsied. And it came back as a basal cell carcinoma.

And this is a person who had had previous lesions, actinic keratosis and other things taken off of his skin, really didn’t worry about it too much. But it was observed that after the lesion was resected, that it turned out that the margins actually were positive. And so they went back and they did a wider excision of the lesion. And this time, again, the margins were close, but they felt confident, at least, that they had removed the basal cell carcinoma at that point.

DR LOVE: Just out of curiosity — and a lot of times, I’ll have it in my mind when I hear about locally advanced basal cell, a person who’s on the fringe of society, who’s not getting medical care. What about this man?

DR LUKE: Oh, so absolutely not. For this guy, at least. I mean, this was a very well-off gentleman who had had a lot of sun exposure over the course of his life and really just kind of thought, “This just isn’t a big deal. It’s a little bump on my skin,” and it really took quite a bit to get that person to the doctor. And I think this really does raise that as we get older in society these days, with the amount of sun exposure that the population is taking, it’s really going to become essential that people do see their dermatologists and really consider having any new lesion biopsied and removed.

DR LOVE: So what was the next step with this man?

DR LUKE: So for this man, after this lesion was removed it was assumed that things were taken care of and it’s not going to be a big issue. Unfortunately, it was not more than about 6 months before the skin graft that was required to close the previous lesion started to change in color and eventually became nodular. And so at that point, he returned to the dermatologist — or, I think, actually a surgeon at that point. And a rebiopsy of the area was taken and unfortunately showed that there was still residual or locally recurrent basal cell carcinoma that was present.

And so that was quite unfortunate. He did have an attempted resection of that lesion again but unfortunately again had positive margins on the reresection and now had a very significant problem, because now he had a surgical wound that actually would not heal right because there was residual basal cell carcinoma that was present. So the patient was referred to a radiation oncologist and had radiation to that area. Again, unfortunately, while there was some initial control of the disease for a period of time, over a relatively short period of time we started to see a disease outgrowth. Again, now in more of a miliary pattern extending into the skin, going sideways.

And what we saw was that the initial surgical lesion really started to degrade and started to fall apart, such that now there was an open wound that really wouldn’t heal after multiple surgeries and radiation. So a biopsy of that lesion did show residual basal cell carcinoma and, at that time, unfortunately, he was referred to us, being deemed to be unresectable basal cell carcinoma.

So realize that this is a patient that not with metastatic disease, like we would commonly think about in the medical oncology setting, but rather with a local problem of a lesion on his temple, now with basically a hole in his skin for opportunistic infections to take route.

DR LOVE: Just to clarify, at the point he saw you, had he seen a medical oncologist?

DR LUKE: To that point, he had not seen a medical oncologist, because this had been managed with a dermatologist and a surgeon as well as a radiation oncologist, but there hadn’t been thought to have a role for a medical oncologist at that point. So this patient was in the community being managed by his providers there.

DR LOVE: How long ago was this?

DR LUKE: This odyssey for him started less than 2 years ago. He got referred to me more about a year ago.

DR LOVE: Hmm. Because I am kind of curious, since it’s not that common out there, and certainly not common in medical oncology, whether you’re seeing patients — I mean, to me it sounds like it’s maybe a little — before he got stereotactic radiation that maybe he should be considered for a hedgehog inhibitor before that.

DR LUKE: Yes. That’s a very reasonable question. I think that that’s an area that’s unknown and just workflows in the community oncology setting. I don’t know that it’s right or wrong. There aren’t any studies that look at radiation before or after hedgehog inhibitors. But you could actually ask a quality-of-life question that if we had intervened earlier, would he not have quite as much of an issue? But I think that’s hard to say.

DR LOVE: No. I mean, I was just wondering in terms of awareness in dermatologists and surgeons about hedgehog inhibitors. You would think if they had somebody with an unresectable one, they would be aware that these agents are there.

DR LUKE: Yes, but I think that uptake in the dermatologic as well as the dermatologic surgical approach settings, I think they are broadly aware. And I think it often comes down to the patient in front of you. Do you, as the physician, think, “Oh, I can get this. And I can really take care of this problem”? I think often that takes us down that route, sometimes to a place where, perhaps, if we had, a few steps earlier, reconsidered, we might have made a different calculation.

DR LOVE: So what happened next?

DR LUKE: So being deemed unresectable, he came to see me. And the conversation then was really about how are we going to manage this as long as we possibly can? And we did present options to him of a hedgehog inhibitor, which, for all the reasons around hedgehog signaling in basal cell carcinoma, could be very reasonable. We did discuss platinum-based chemotherapy as a potential option as well as a few clinical trials. And there are some things in the immunotherapy world as well as just in the Phase I world that might be of interest.

Vismodegib-associated side effects

DR LUKE: Given the biology — he was a sophisticated guy. And he said, “Yes, I’d like to try the hedgehog inhibitor.” We did educate him around the potential toxicities of the drug, and he did start taking it. And very quickly — I mean, within days — he did see, actually, the area start to improve. And so he had less oozing around the area and just in general it looked better.

Unfortunately, he also very quickly started to notice that taking the drug wasn’t as easy as he had expected it to be.

DR LOVE: He got vismodegib?

DR LUKE: Yes. But he started to develop abdominal discomfort and a sense of nausea. And he wasn’t so much someone who had the muscle aches but rather just kind of just didn’t feel well all the time. And so he struggled through that, and we tried some different things around his diet and trying to get in the appropriate nutrition, but by about 3 months, even though we had dramatic treatment effect, we had really gotten to a place where he was kind of fed up with this, because it was really impacting on his quality of life. And so he said, “What will happen if we stop?” And I said, “It’ll probably grow back, but it’s reasonable to give it a try.”

So we did, but unfortunately it only took a few weeks before we started to see the lesion returning. And so we did then transition to put him back on the hedgehog inhibitor and did various things about dose reduction and some interruptions and continued on. But, unfortunately, eventually he did develop resistance and we saw outgrowth of the tumor with the surgical area again starting to break down. And we realized that he had had progression of disease at that site.

Thereafter, we did try a few other things with standard chemotherapy as well as some off-label sorts of approaches, but none of those really made a real big difference. And eventually, unfortunately, he did develop an opportunistic infection that could not be controlled adequately, and he did pass away.

DR LOVE: Wow! What a terrible story. It seems like he probably got treated with vismodegib maybe before sonidegib was approved. But I am just curious now that both are available, how do you choose between the two?

DR LUKE: I think that when I look at the data sets, it’s not entirely clear to me that one is distinctly better than the other. So this sometimes comes down to formulary questions about what’s available to a patient. We did start using vismodegib earlier, so we have a little bit more experience with it, but it’s not that one, I think, necessarily is better than another, per se. So in my practice, it’s a little bit “path of least resistance.” And the one that we can get for the patient is the one that we use.

Epidemiology and biology of Merkel cell carcinoma

DR LOVE: Let’s talk a little bit about Merkel cell, though, because, I mean, amazingly, you actually have a new agent approved.

DR LUKE: Yes. So Merkel cell carcinoma is another form of skin cancer, but again, much more rare, so the only approximately 2,000 cases per year in the United States. And this is a very interesting disease that has a complex biology that’s both related to infection as well as exposure risk in the environment.

And so we see these lesions most commonly in older adults over the age of 50, but commonly, honestly, more in the late sixties and seventies, people who have had a high degree of UV ionizing radiation exposure from the sun and also patients who have had a history of immunosuppression. And that can be in the context of a transplant. It can be in the context of HIV and several others.

Unfortunately, this is a very aggressive form of cancer, meaning that more than 40% of patients who develop a cutaneous Merkel cell cancer will develop metastatic disease, and often quite rapidly, such that historically we describe the median survival in the context of metastatic disease at less than a year, at about 9.6 months.

Very interestingly, this disease stratifies in between those that are associated with a viral etiology and those that are not. But about 80% of patients who develop a Merkel cell carcinoma, we can find evidence of infection for the Merkel cell polyomavirus. And that is actually a virus which exists at normal circumstances on almost everyone’s skin. It just turns out that most of us don’t get Merkel cell carcinoma, presumably from immune surveillance that will eliminate the virus. But in some folks, unfortunately, it doesn’t and it can develop Merkel cell.

It is important to point out that 20% of Merkel cells do not harbor this virus and, in those patients that have been profiled to date, we see quite extreme mutational loads. And that’s where we see dramatic effect of sun exposure leading to cancer with mutation of common genes, as you’d expect, like p53, RB and all of these canonical cancer-driving genes.

DR LOVE: So just to clarify, though, in terms of the viral-related ones, is the issue there some kind of immune deficiency or something, primarily?

DR LUKE: Yes. So we do believe there’s some element of immunodeficiency that leads to the development of Merkel cell, but it’s not as easy as being able to identify one gene, per se. We note this association in clinical settings with immunosuppressive medications, like steroids, anti-TNF and so on and so forth, in the context of transplant. But certainly there are people who develop Merkel cell who have no history of any direct immunosuppression and who otherwise don’t manifest symptoms of an immunodeficiency syndrome. So we don’t really understand, in those patients, what exactly is going on.

Management of Merkel cell carcinoma

DR LUKE: That being said, this is a tumor that we’ve now learned can be dramatically amenable to immunotherapy. And so because of the viral etiology, these tumors commonly have very high expression of PD-L1 and infiltration of T cells. So historically, our management of Merkel cell carcinoma — and you can see, Merkel cell carcinoma is that we don’t have a direct FDA approval, but rather, just like other small cell cancer, like small cell lung cancer, which is also a neuroendocrine tumor — and I didn’t point that out earlier, but Merkel cell carcinoma is also a neuroendocrine tumor.

We treat it very similar to small cell cancer with platinum plus etoposide chemotherapy. And we’ve done that historically. That has definitely been, before very recently, the standard front-line therapy for metastatic disease. It’s also sometimes used for adjuvant therapy as well, where radiation is probably more common, however. That being said, much like small cell carcinoma of the lung, we see that responses are quite common initially but rarely very durable.

A historical data set for progression-free survival on chemotherapy for Merkel cell carcinoma shows there are rare patients that go out a long period of time, but the median unfortunately is on the order of just over 3 months in terms of progression-free survival on this chemotherapy.

DR LOVE: Just a side question: You mentioned that it’s a neuroendocrine tumor. And I can’t remember what the scan is that you do in people with GI neuroendocrine tumors and the radiolabeled lu-dotatate thing that’s being used. I don't know if you’re familiar with that and the —

DR LUKE: Yes. So — but those scans are more commonly used for low-grade neuroendocrine tumors in the liver or in the pancreas, as opposed to these neuroendocrine tumors, much like small cell, are much, much more aggressive. And so we tend not to use those radiolabeled investigations in Merkel cell carcinoma.

Response to PD-1/PD-L1 blockade in Merkel cell carcinoma

DR LOVE: Right. So how about checkpoint inhibitors?

DR LUKE: So because of the high degree of T-cell infiltration associated with the virus in Merkel cell carcinoma, there’s been dramatic progress with the development of PD-1 and PD-L1 blockade for this disease.

A first presentation of pembrolizumab for Merkel cell carcinoma showed that in the initial series, more than 50% of patients in the front-line setting had RECIST-quality partial responses to treatment. And these can be quite dramatic, meaning within days we see the tumor starting to shrink massively.

And so these are quite effective therapies, pembrolizumab as well as avelumab, the PD-L1 antibody, again, both in the second line after chemotherapy and in the front-line setting. Both have been shown to be highly effective. And I’ll point out that avelumab has actually gained regulatory approval as the standard PD-L1 therapy in this space, although pembrolizumab also has a compendium listing with the NCCN.

The kinetics of treatment response and the efficacy of treatment out over time, again, from the CITN study of pembrolizumab in Merkel cell carcinoma shows the majority of patients had relatively rapid decrease in the sum diameter of their lesions, which was durable out over extended periods of time. The time to response, again, showing here in the dots, again, essentially at the time of first scan, we see that most patients actually are already demonstrating a RECIST-quality response.

And so in this study, the initial reported response rate was approximately 70%. Subsequent studies have described that in the front-line setting more around 50% response rate. In the second-line setting, that also drops a little bit further, but really there’s no question that PD-1- and PD-L1-blocking antibodies have become the default standard of care treatment for Merkel cell carcinoma and really should be considered for patients with reasonable performance status and probably in the front-line setting.

It isn’t to say that we don’t use chemotherapy for these patients anymore. However, a lot of these patients will present with bulky metastatic disease, where we really have to think about doing something aggressively immediately to try to get their symptoms under control. But in my practice, I really try to convert patients over to immunotherapy really as quickly as we can, because I think that the long-term benefit from PD-1- or PD-L1-blocking antibodies really dwarfs any kind of benefit from chemotherapy outside of that initial benefit that you might get from tumor debulking.

Case: A 61-year-old man with metastatic Merkel cell carcinoma

DR LOVE: You had a patient, a 61-year-old man with Merkel cell. What happened there?

DR LUKE: So recently, I had a patient who presented to see me. And he noted that really over a very short period of time, only a few weeks, he had had a bump just above his left elbow, which had gone from being something that he noticed to being the size of, like, a tennis ball. It really had just dramatically and rapidly increased in size. And so that lesion was biopsied and came back showing Merkel cell carcinoma.

We then went forward and did a PET scan because of the size of this lesion and saw axillary lymphadenopathy as well. And so I had a discussion, because the standard of care obviously would be to try to resect these lesions in the local and in the regional lymph node basin. However, the surgeon said, “This is a very aggressive disease. It’s bulky. I’m not totally sure I’m going to be able to get it.” So in this patient, we talked about using immunotherapy, but because of how fast the cancer was growing, we actually chose to use platinum-based chemotherapy.

And we gave the patient 3 cycles, actually, of cisplatin plus etoposide and saw a dramatic shrinkage of the tumor. So the tumor on his elbow nearly resolved. And the palpable lymphadenopathy actually decreased quite a bit as well. So the patient was then able to actually transition to go to the operating room and had wide local excision above the elbow, as well as axillary completion lymph node dissection.

Interestingly, whereas on the elbow there was no residual disease, in the axilla there still actually was a deep-seated lymph node mass. And so given that, we actually went forward with adjuvant radiotherapy both to the arm and to the axilla, which is, importantly, a standard of care in the adjuvant setting to use radiation. So we then continued to follow the patient forward. We didn’t give him more chemotherapy, because chemotherapy has not been associated with an overall survival advantage in the adjuvant setting, but continued to watch.

And, unfortunately, fairly quickly, within 6 months, he had developed abdominal discomfort. And restaging imaging now already showed liver metastases and omental caking. And so again, the question was what to do. And in discussion about the available options, we said, “There’s this benefit to a PD-1/PD-L1, so we should go forward and use avelumab as a front-line therapy.”

So we started avelumab. And I think the patient got 2 doses, but unfortunately his symptoms actually continued to accelerate with worsening abdominal pain and weight loss. And he got hospitalized, and when he was discharged, we again had further conversation. It really wasn’t clear that we’d given him a real shot at immunotherapy. He’d only gotten a few doses. But it was just clinically obvious that we couldn’t go further down that route.

So we again switched back to cisplatin plus etoposide chemotherapy, and he had a marked improvement in his symptoms of abdominal discomfort. He was able to eat. And we saw, actually, after 2 cycles and then again after 4, dramatic reduction in his overall tumor burden. So again, then, at that point, we said, “Okay. Why don’t we give immunotherapy a try again? Now we’re in a lower disease volume setting, symptoms are better.” So we reinitiated avelumab at that point and, actually, the patient now has gone — he’s still in a stable disease situation, but he’s now gone more than 6 months on that therapy and seems to be tolerating that quite well.

So I think that this really emphasizes how clinical practice is not what we do in clinical trials and, really, we manage the patient that’s in front of us. But we can use the concepts that have been taken from randomized Phase III trials and try to incorporate that into the management of our individual patients. So with this patient, my priority was to get him to immunotherapy over time, but it took a couple of shots on goal in order to do that. We had to debulk him and get him into a clinical situation where we thought the immunotherapy would have a reasonable shot. And so far, that has been an effective strategy.

DR LOVE: And this case kind of reinforces what you’re saying about it being an aggressive disease. I’m kind of curious, if you looked at Ki-67 or mitotic rate or whatever, what do you see?

DR LUKE: So similar to small cell lung cancer, you see very high rates of Ki67-positive cancer cells in Merkel cell carcinoma, especially in the metastatic setting. If you get resections of these lesions from lymph nodes or other places, you really just see sheets of these neuroendocrine cells that are looking quite dangerous and very, very aggressive.

DR LOVE: Fascinating.

Activity and tolerability of anti-PD-1/anti-CTLA-4 combination therapy for Merkel cell carcinoma

DR LOVE: Any data or thoughts about using combined therapy with anti-PD-1 and anti-CTLA-4 in Merkel cell?

DR LUKE: So absolutely. There’s an interesting combination checkpoint blockade in Merkel cell, just like there is in every other kind of cancer. And the discussions have been had around using CTLA-4 plus PD-1-based immunotherapy. The caveat there is that the patient population for which you’d be employing this is not the same as what we’ve seen in melanoma, because the majority of patients who have Merkel cell carcinoma are going to be of advanced age and, therefore, their ability to tolerate the side-effect profile associated with that combination therapy, I think, is a little bit more in doubt.

So those sorts of questions are being asked. And clinical trials to investigate those questions are coming forward. I think there we’re going to need a lot more flexibility around the dosing strategy associated with the CTLA-4 antibody, whether it’s to drop it from 3 mg/kg to 1 mg/kg for ipilimumab or perhaps to space out the dosing, not every 3 weeks but rather every 6 weeks. I think we’re going to have to take a broader approach, because, unfortunately, the idea that we’re going to be inducing a Grade 3/4 immune-related adverse event in more than 50% of our octogenarians, I think, is going to be problematic in clinical practice.

Epidemiology, etiology and therapeutic options for squamous cell carcinoma of the skin

DR LOVE: So I want to finish out and ask for your thoughts about 1 other nonmelanoma skin cancer, which is squamous cell cancer and what the epidemiology of that is, particularly as it relates to oncologists and where we are today with therapy, and particularly the issue of checkpoint inhibitors.

DR LUKE: Yes. So squamous cell cancer of the skin is a, again, much more common entity than melanoma, being on the order of 300,000 to 400,000 cases annually in the US. Again, whereas melanoma is associated with high-intensity sunburns, squamous cell cancers are really associated with long-term exposure to ionizing radiation or UV light over the course of one’s life, so that people often develop multiple of these lesions later in life, in their sixties and seventies.

Again, these are, generally speaking, easily managed with local resection, either by Mohs surgery or even sometimes a surgical resection. But they can usually manage locally. Unfortunately, they can develop as locally recurrent lesions after surgery. And a small fraction do develop metastatic disease. And most commonly, that will be to regional lymph nodes and/or to the lung. And management of locally advanced or metastatic squamous cell carcinoma from the skin follows algorithms, actually, that are fairly similar to what is more commonly thought of as oropharynx head and neck cancer.

And so we think about platinum-based chemotherapy. We think about radiation. We think about EGFR inhibitors like cetuximab. But most recently, we’ve really started to focus on the utility of immunotherapy in that space, because it’s been known for quite a while — and it completely makes sense — that the driving etiology of these lesions is the tremendous amount of sun damage. So the number of mutations that we see in squamous cell cancers of the skin are dramatically high. And so following along on the research underpinning the utility of checkpoint blockade in other cancers, there’s a hypothesis then that PD-1-based immunotherapy might be quite effective in squamous cell carcinoma of the skin, given the number of putative mutations or neoantigens that might be present.

DR LOVE: Can you comment on the recent data on cemiplimab for squamous cell carcinoma?

At ASCO was the first presentation of a PD-1 antibody specifically in squamous cancer of the skin. And there, we did see a quite dramatic response rate, on the order of 50%. And, actually, in a subset was 70% of patients responding. And larger data series are coming forward, with one agent now being granted a breakthrough designation for squamous cell carcinoma of the skin.

DR LUKE: It was presented at ASCO.

DR LOVE: But this is just squamous of the skin, or has it been used in other cancers?

DR LUKE: Their trial is in other cancers as well.

DR LOVE: Wow! That is interesting. Fascinating.

We should note that subsequent to recording this discussion, nivolumab was approved for the adjuvant treatment of patients with melanoma with involvement of lymph nodes or in patients with metastatic disease who have undergone complete resection. Nivolumab was previously approved for the treatment of patients with unresectable or metastatic melanoma.