Pauline Funchain, MD, and Caroline Robert, MD, PhD, on Melanoma: New Data on Encorafenib, Binimetinib, Ipilimumab, and Nivolumab

Pauline: Caroline, so many of us admire your work. Thank you so much for coming here. So we'd love to hear a little bit more about EBIN. Can you tell us about the question you wanted to answer and the design of the study? Caroline: So the question was is it a benefit for the patient to begin a treatment with targeted therapy? Of course I'm talking about patient with BRAF-V600E or K melanoma with advanced disease for whom we have several possibilities of treatment, but is it a good option to begin with anti-BRAF, anti-MEK, and then before waiting for a progression, initiate immunotherapy? Because that's something that we're very often tempted to do because we all know that when we have a progression under BRAF-MEK therapy, it's very difficult to stop treatment because although the patient progress under treatment, we know that if we stop, they will progress even more rapidly and sometimes we don't even have time to give immunotherapy and to see if it will work. So that was the big question. And in fact, this question had also been addressed by the second big trial by Paolo Ascierto and his colleagues because he had a three-arms treatment. One, the first one was initiating with anti-BRAF-MEK and switching at progression to immunotherapy combined ipi-nivo. The other arm was ipi-nivo and then switching at progression with ... And he had this, what he called the sandwich arm, which is really the question we address. When he did a short treatment, two-month treatment with a targeted therapy and then ipi-nivo. Unfortunately this trial was not powered enough. There was only about 60 patients in each arm, so we could not conclude. So it's quite a good thing that EBIN could address this question with the power statistically that was sufficient to say if it's yes or no, if it's different. And the main objective was on PFS. Pauline: So the design of the study, what were the arms and how many patients? Caroline: So two arms, 271 patients randomized, one-to-one, to receive either three months of encorafenib plus binimetinib, standard dose, and then one week of drug vacation, only one week. And then ipilimumab plus nivolumab. The dose of ipilimumab was one milligram per kilogram, the low dose, which at the time the trial was initiated, considered totally acceptable. And the second arm, the control arm, is the same, ipilimumab plus nivolumab regimen, low dose. And of course after the four infusions of ipilimumab plus nivolumab, we continue with nivolumab with a standard dose, fixed dose, 480. So we really wanted to stratify on parameters that we thought were important because we know that for some patients, we really want to initiate with targeted agents. So we stratified on what? On LDH activity in the blood, which our patient ... with a poor prognosis, usually, we stratified on a stage and we also stratified, sort of obliged, by centers. And so we have now ... we could produce the first results and the main objective, which was PFS and the follow-up, the median follow-up, was 21 months. Pauline: So what did you find in these cohorts? Caroline: So finally, we found that overall, if you take into account the total population of patients, the intent-to-treat population, there is no difference in PFS. I mean, the two curves ... and it's interesting to look at the curves dynamically because you see the first three months after the induction, the patient who received targeted therapy, they are almost all of them, 99%, without progression. Whereas we see a lot of patients progressing with ipi-nivo. And then if you continue to follow the patient, the second evaluation, the two curves begin to get together and then they remain together. So we have a hazard ratio that is very close to one. We have a large confidence interval that encompasses one, so there is no difference. But I think ... then I will talk about the subgroups, but even though if we have no difference, because we know from the DREAMseq trial, Mike Atkins, that it's better to initiate with an immunotherapy. If we can put in the same curves the patient who initiate the treatment with anti-BRAF, anti-MEK, with the same curves as the immunotherapy patients, it's not bad. But then let's look at the patient that were pre-specified, the population that were pre-specified high LDH, really high LDH, like twice the upper limit of the normal. If you look at the hazard ratio of this subpopulation, it's close to 0.5. So it touches one, but still it's very in favor of initiating with anti-BRAF, anti-MEK. So it's like you rescue this patient at the beginning and then you can switch to ipi-nivo. The other subpopulation that seemed to really benefit from this induction period of three months were the patient with liver metastasis. It was not pre-specified, unfortunately, but it's very clear that the hazard ratio is really on the left of the one limit. So it's interesting. So it gives me and us, I think, the community, more confidence to initiate this trial with targeted adjunct and then not to wait for the progression and to try to have a benefit of immunotherapy afterwards. Pauline: So we found that in accruing to DREAMseq, it was really difficult to accrue patients with high tumor burden who were rapidly progressing. Did you have that type of accrual question when you saw a patient in clinic and were accruing to EBIN that there were certain patients you felt were growing so quickly it would be difficult to put them on the trial? Caroline: Yes. We had actually two patients that were randomized in the immunotherapy arm, the control arm, and they could not enter the trial because we decided that they should receive a targeted therapy. Pauline: So that did happen sometimes. Caroline: It can happen. We know that it happens. I think we all know that if we can, we give immunotherapy first line, but if we can't, maybe now we have some robust data to say, "Let's switch before the progression." Pauline: So when you're looking at these data and seeing no difference in PFS, do you see then EBIN as an alternate option, complementary option with two different types of patients, high risk patients and less high risk patients? Or do you see these data as potentially saying that that approach, that the EBIN approach may supersede in BRAF mutant patients, ipi-nivo first? Caroline: No, I think it's really to be reserved for specific population of patients for whom you feel obliged to begin with a targeted adjunct. Honestly, I feel that. Pauline: Yeah, totally makes sense. And when you read these data, what are those subgroups that you feel are really the EBIN favorable patients? Caroline: Patient progress rapidly. So high LDH, lot of liver meds because liver meds, I mean specific area, the liver, we know it's sort of immunosuppressive area. We don't like that. Also, patients who have a location of the metastasis that are threatening their lives rapidly. Of course, in this trial we did not enroll patients with central nervous system, non-controlled lesions. But maybe also these patients, we can try to switch them also when we begin with targeted adjuncts. So I think it's really, I would say all the patients for whom we feel no time for them ... I mean, they don't have time to see if immunotherapy is going to work. We need to give them something that we control their disease right now, and now let's try to switch them earlier. Pauline: Yeah, it is really great to have that data supporting the targeted therapy approach first and feel reassured that there's an equivalent PFS. So it was fantastic to see that data. Thank you so much for your time, Caroline. Caroline: Thank you very much.