COMBI-AD: Dual-Targeted Therapy for Stage 3 Melanoma
Panelists: Jeffrey S. Weber, MD, PhD, NYU Langone Health; Reinhard G. Dummer, MD, University Hospital of Zurich; Axel Hauschild, MD, PhD, University Hospital Schleswig-Holstein; Michael A. Postow, MD, Memorial Sloan Kettering Cancer Center; Caroline Robert, MD, PhD, Gustave-Roussy
Jeffrey S. Weber, MD, PhD: Tell us about the data from COMBI-AD.
Axel Hauschild, MD, PhD: I have presented at ESMO a clinical trial on the adjuvant use of dabrafenib and trametinib, the BRAF and the MEK inhibitors, which were compared to a double placebo. It was a double-blinded clinical trial, and it showed an amazing hazard ratio of 0.47. So, it has a clear-cut benefit with an even more amazing P value. It was 13 zeros, and I need to say this is convincing, but everybody was asking for overall survival and distant metastasis-free survival. There was a hazard ratio in the same range for overall survival, 0.57. These are the best results I have ever seen in an adjuvant trial in melanoma, and it shows that for the BRAF-mutated patients, we have a good alternative to either conventional interferon; pegylated interferon, which was never approved in Europe; or high-dose ipilimumab, which is not approved at all in Europe. So, we had only low-dose or high-dose conventional interferon available here. For us, it’s something completely new.
Jeffrey S. Weber, MD, PhD: When you compare dabrafenib/trametinib, BRAF/MEK, to ipilimumab, it sounds like you have a superior hazard ratio by far. Obviously, the P value is superior, although there you have longer follow-up and survival. And then, with the CheckMate 238 study, which I presented at ESMO, it’s a bit like apples and oranges. There was a trial with an active control arm. It was comparing nivolumab with ipilimumab, but the hazard ratio also looked pretty good; it was 0.65. And there, you have a trial that was stopped at an early interim analysis, so you only have RFS. You don’t have survival data, and it will be confounded by the fact that there’s an inherent, or de facto, crossover. But nonetheless, it sounds like we now have multiple options for terrific adjuvant therapy, which I guess your point is that it’s going to reduce the surgeon’s role in the management of stage 3 disease.
Axel Hauschild, MD, PhD: I learned from you and others, for decades in medical oncology, that you should never compare one clinical trial to another because the eligibility curves here are varied. One clinical trial was conducted in stage 3 patients, namely the COMBI-AD trial, whereas the other one was refraining from stage 3A and taking some patients, a small proportion of patients, with stage 4 disease in the adjuvant setting. So, it’s really difficult to compare across all clinical trials, although in the future it will be done. There will be a couple of minor survival curves superimposed to speculate which is better. But I like your clinical trial result, the 0.65, because ipilimumab alone with 10 mg/kg has shown a survival improvement already over placebo. So, you can speculate that if you add one hazard ratio to the other, you come up with more or less a hazard ratio in the same range. But this is, again, pure speculation.
Jeffrey S. Weber, MD, PhD: I think we would call that a back-of-the-napkin calculation. The nice thing about the nivolumab versus ipilimumab trial is that it wasn’t just a hazard ratio that was highly favorable. It wasn’t that it had a P value of .0001 that was highly favorable. The toxicity was considerably less. Something close to less than 9% of patients had serious toxicity, grade 3/4, or had to stop, versus something like 30% or 40% who either had to stop or had serious toxicity. Obviously, it was very favorable. But what can you tell us about the toxicity of COMBI-AD? Any words?
Axel Hauschild, MD, PhD: There was typical toxicity observed for stage 4 melanoma patients, with no big surprises. It was used for a year in the adjuvant setting. But the treatment discontinuation rate was 24%, so it appears to me that it will be a little higher for reasons we don’t understand at this point in time. Why did so many discontinue? I believe it’s the situation of an adjuvant treatment, where there was not such big pressure as for stage 4 melanoma treatment in patients. Because if you look at the individual grade 3/4 toxicities, it’s not higher than in every clinical trial, particularly COMBI-D and COMBI-V for stage 4 disease.
Jeffrey S. Weber, MD, PhD: Is anyone going to use adjuvant ipilimumab anymore? It’s not even approved in the European Union. Michael, it’s approved in the United States. Will anyone use it anymore after these data?
Michael A. Postow, MD: I think adjuvant ipilimumab, despite its approval, will be replaced. If you’re going to give adjuvant immune therapy, you’ll choose adjuvant nivolumab. If you’re going to consider targeted therapy, it would be the BRAF and MEK inhibitor combination, dabrafenib and trametinib, for stage 3 patients. I think the question will really be if you have a BRAF-mutant patient in the stage 3 setting, do you give adjuvant nivolumab or do you give adjuvant dabrafenib and trametinib? Of course, we have no randomized data. This is the same problem we have in the metastatic setting. Now we’re just shifting our conversation from earlier to stage 3 resected melanoma.
Jeffrey S. Weber, MD, PhD: Although, there may be a way out that is not statistically supported. If you break out the patients with stage 3B/3C BRAF-mutated cancer—which, again, is slicing and dicing from CheckMate-238—who got nivolumab and look at their RFS, you’d have to then take the stage 3B/3C patients out of COMBI-AD and look only at their RFS, because we only have RFS data from CheckMate 238, and do the comparison. At the end of the day, it will be an interesting argument to see how well those patients did.
Axel Hauschild, MD, PhD: The COMBI-AD trial was stratified for 2 factors. One factor was patients with V600E versus patients with V600K. The other one was stage 3A versus 3B versus 3C. And, if you look at the subgroup analysis, you will see such consistent data for every subgroup factor, and I think it was 15 different categories that have been evaluated. So, there is no doubt it works as well in stage 3A, 3B, and 3C.
Michael A. Postow, MD: One of the other issues I was impressed with was really not only the COMBI-AD trial, but also the adjuvant study of vemurafenib versus placebo. The duration of targeted inhibition was limited in these studies. These patients weren’t on targeted therapy for years and years and years, yet the benefit in terms of reducing their likelihood of disease recurrence did seem to be maintained beyond the period of time that these patients were on the active treatment. There may be something immunologic happening, or maybe something more permanent happening, when patients are on somewhat of a limited course of targeted therapy as adjuvant treatment. We don’t really completely understand what’s going on there just quite yet.
Caroline Robert, MD, PhD: We will hopefully have very soon, by the next AACR annual meeting, the results of the EORTC trial sponsored by MSD with pembrolizumab versus placebo. Then, it will be more comparable to COMBI-AD, because we will have a placebo arm, so that will be important.
Jeffrey S. Weber, MD, PhD: Were the eligibility criteria similar?
Caroline Robert, MD, PhD: Yes.
Jeffrey S. Weber, MD, PhD: So, 3A, 3B, and 3C. That will come out. You would have thought it would have come out already, at least with RFS, but it will sometime in 2018.
Caroline Robert, MD, PhD: Yes.
Jeffrey S. Weber, MD, PhD: Now, at this ESMO, the data for CheckMate versus COMBI-AD will generate a whole series of arguments. Then the arguments will get even more complicated when we have the EORTC data. And then, it will get even worse if you start thinking about patients going on ipilimumab/nivolumab versus nivolumab alone. Again, each trial may change the parameters. We had a few years where we used ipilimumab. Now we’ll have a time where we use BRAF/MEK or nivolumab and who knows what it will be.
Axel Hauschild, MD, PhD: I found it a pity that a new trial, the CheckMate trial that uses ipilimumab plus nivolumab, is aiming for a 12-month period. That would be an ideal situation to test just a 3-month regimen, which could be very attractive to certain patients if you explain to them that they don’t need to be treated for a whole year. The aim is for 3 months, and even if they have sequalae from toxicities, it could be a very attractive scheme replacing others. You have the option of 1 year, or even longer treatment, compared to just 3 months of treatment.
Caroline Robert, MD, PhD: Yes, but imagine it doesn’t work, then you don’t know if it’s because it was not given for long enough. It’s difficult.
Jeffrey S. Weber, MD, PhD: That’s a tough sell, and I’ll say one thing: I’ve done piloting of ipilimumab/nivolumab adjuvant therapy, and ipilimumab at 3 mg/kg and nivolumab at 1 mg/kg was just not tolerable in the adjuvant setting where you have patients with stage 3B/3C disease, for example, who might have a 50% chance of relapse. But that means they have 50% chance of being cured just sitting there. That’s a bit of a tough sell, which is why we flipped the doses, and now we’re actually giving nivolumab maintenance every month, which makes it even more palatable. And then, with the CheckMate-915 trial, I think they stretched the ipilimumab out to 6 weeks. So, I think that will make it easier for patients.