Jeffrey S. Weber, MD, PhD: One of the big questions that I face when I see a new patient is, what testing should I do before advising the patient as to what therapy they should have? Michael, when you see a new patient, what biomarkers do you test either by immunohistochemical staining or, more importantly, genetic testing? What kind of genetic profiling do you do on a new patient’s tumor?

Michael A. Postow, MD: When I see a new patient with metastatic melanoma—I’ll include stage 3 in this discussion and stage 4, and focus this part of the conversation on stage 4, as I think later in our program, we’ll talk about stage 3—the most important test, in my opinion, is a molecular test for the BRAF mutation. There are a number of BRAF mutations that one can find. The most common is a BRAF V600E mutation, and this can be tested by a number of different platforms, including an immunohistochemical test. However, genetic sequencing is usually the standard approach for testing to see if a BRAF mutation is present. We know that if a patient has a BRAF mutation, they’ll be eligible for BRAF and MEK combination targeted therapy in addition to immune therapy. That’s a big discussion: if we find a BRAF mutation in a patient, do we start with targeted therapy or do we start with immune therapy?

There’s a lot to discuss, and there are pros and cons to each of these different types of approaches. But, in addition to a BRAF mutation, we’ll often test for other mutations, some of which can be treated with drugs as well. We’ll test for KIT mutations and other mutations, like NRAS mutations, in patients. We have some data that have shown some efficacy for strategies to target those kinds of mutations as well. Hopefully, over time, we’ll try to find exactly what’s driving each individual patient with melanoma’s tumor so that we know what kind of approach to tweak just to try to treat that patient individually. I think the most important test among all of these is the BRAF mutation. I think that’s a standard approach for patients with stage 4 melanoma and, over time, with stage 3 melanoma as well.

Jeffrey S. Weber, MD, PhD: So, we’re going to do genetic testing in our patients. Some institutions went from a 15-gene profile to a 30 to a 50 to a 350 and then to a 500-gene profile. But it sounds like mostly you are going to use that single BRAF test. Reinhard, you know a patient with a BRAF mutation has a V600E mutation. What goes through your mind, and what do you discuss with the patient to help make the decision? Do they do BRAF/MEK therapy up front, or do they do immunotherapy up front? That’s always the big decision.

Reinhard G. Dummer, MD: This is a very common and important question, and we have to be very precise about that. I have to say we don’t have data that really would prove one approach is better than the other, so there is no comparison around. There is some place for personal decisions, and for me, there are a number of medical reasons that suggest the patient should go for immunotherapy first. For example, if this is a younger patient, they will tolerate toxicities from ipilimumab well, and I want to go for a really long-term outcome. But there are also practical reasons. Will the patient be compliant, for example? Can they come to the center for the infusion therapies? This is difficult decision making, but over the last few years, I have to say that more and more patients go for immune therapy first because we know that targeted therapies are very powerful tools to rescue patients who progress. For the brain metastases, a good example is, let’s say, that last year I would have said brain metastases were a criterion for me to put the patient on targeted therapy and this has changed. Nowadays, for a patient with multiple small brain metastases, the first choice will be a combination of ipilimumab/nivolumab.

Jeffrey S. Weber, MD, PhD: Although we do have some good data that Mike Davis presented at ASCO. I believe he had an abstract where he looked at dabrafenib/trametinib in patients with brain metastases. As Axel pointed out, I think it sounded pretty good.

Caroline Robert, MD, PhD: The response rate was quite good, but I think we were all a little bit disappointed by the PFS in the brain. Because we have PFS for cutaneous metastatic melanoma, I mean outside of the brain, of 1 year and here we had 6 months. I think we were all a little bit disappointed by that.

Reinhard G. Dummer, MD: It was half. The PFS in the brain was half of the PFS outside of the brain.

Jeffrey S. Weber, MD, PhD: Interestingly, in the CheckMate-204 trial, I thought that the PFS in the brain was also about 6 or 7 months, so it’s a little difficult.

Caroline Robert, MD, PhD: Yes, but for PFS with the immunotherapy, I think we needed to wait a little bit longer.

Axel Hauschild, MD, PhD: In these patients, I’m more interested in looking at the duration of the individual response, particularly in the brain, than looking at PFS because we are coming back to the issue of stabilized diseases and all of this. But allow me to make just 1 point on the testing. We have a German guideline, and the German guideline has said, since 2012, that we are testing from stage 3B onwards, which means we have already implemented testing, 5 years ago, in stage 3—not in stage 4. The reason for this is very simple. We wanted to have results in the files, like how for triple-negative breast cancer, we wanted to have the molecular results in there. We’re testing for BRAF, and only if BRAF is negative, which means wild-type, are we testing for NRAS. Only the patients with acral lentiginous melanomas and mucosal melanomas are tested for c-KIT and only when they are metastasized. I think it is very important that you are testing early because you want to have it in your files. When the patient is sitting next to you and progressed to stage 4 disease, you don’t want to waste time for an additional testing.

But I wanted to make 1 comment to Reinhard. Reinhard was not feeling so well with the question. I don’t feel well, because we have no clear-cut criteria pro-tyrosine kinase inhibitors or immunotherapy. The reason for this is that both analyses—for immunotherapy, single-agent PD-1 ipilimumab plus nivolumab, and for tyrosine kinase inhibitors, namely dabrafenib/trametinib and vemurafenib/cobimetinib—are showing the best-of-the-best patients are always consistently the patients with low tumor load, with low LDH, and so on. So, what we were saying in earlier times, that the best candidates for immunotherapy are those with low tumor load and the best patients for targeted therapies are those with high tumor load, is simply not correct.

And the only reason to believe that tyrosine kinase inhibitors should, in every case, be given first-line is that maybe the patients are symptomatic and need a speedy response because they have a different situation—symptomatic brain metastases and all of this. Because it has been shown that these work within a week or 2.

Jeffrey S. Weber, MD, PhD: Although, the paradox is that the patients who will probably do worse with either therapy are the ones who always get the targeted therapy, right?

Axel Hauschild, MD, PhD: Yes, I agree.

Caroline Robert, MD, PhD: But the difference is that when you initiate a targeted combination, you don’t think that you are going to be able to stop it, whereas when you initiate an immunotherapy, of course, it’s not the majority of the patients, but you have a hope that you might be able to stop it. We begin to have data on discontinuation after confirmed response or after 2 years and maybe even shorter. And what better do you want than to be in response and not taking any treatment? So, I think this is a reason to favor immunotherapy in the first-line if you can. This is what we do.

Transcript Edited for Clarity