Jeffrey S. Weber, MD, PhD: Jason, since you’re on that topic, do the hazard ratios—meaning the reduction, the risk of recurrence—impress you enough that you think that we should be using either dabrafenib/trametinib or pembrolizumab or nivolumab as adjuvant therapy for stage 3 disease? There’s a pretty big distribution of risk in an early 3a patient—a 1.2-mm nonulcerated primary, with 1 mm or less of disease in 1 node. There are other stage 3a patients with a much bigger risk of relapse. Who are you going to treat, and what’s your minimum requirement for a hazard ratio to convince you to embark on adjuvant therapy?

Jason J. Luke, MD, FACP: This is a very interesting question that comes at a time when a lot of things are shifting in our field. I am quite impressed by the hazard ratios for both targeted therapy and immunotherapy. My answer, broadly, is that I lean toward treatment. The question becomes, in my mind, are there stage 3 patients who we shouldn’t be treating with adjuvant therapy? You have to put this in the context of the recent changes to our surgical practices, where we’re not doing completion lymph node dissections in the majority of patients. We end up not really knowing whether or not they were actually stage 3b or 3c. In my practice, everybody is stage 3a. But obviously, they are not.

So, how do you differentiate that? In my practice, I’m fortunate to have a multidisciplinary team. We actually review the pathology and look at what’s going on in the sentinel lymph node that was positive. Sometimes features within the tumor can inform us. But broadly speaking, my default is to treat most patients who are stage 3a and haven’t had a completion lymph node dissection. You have to have a conversation with the patient, to be able to tell them that we may or may not have very specific data. Given the broad environment, that would be the recommendation I would make to most of them.

Omid Hamid, MD: I agree with you. For a time, I didn’t quite know where to go with my stage 3a’s. But the KEYNOTE-054 data clearly look at the stage 3a patients and show a significant hazard ratio for them. That makes me feel more confident in speaking to the 3a patients and offering them adjuvant therapy.

Robert Andtbacka, MD, CM: From a surgical perspective, I agree with both of you. If we don’t do a completion lymph node dissection, if you look at the studies that we’ve done—the MSLT-II study and the DeCOG study—we know that for about 25% of patients, maybe even more, if they had thicker tumors and more disease in the sentinel lymph node, they will have additional lymph nodes with melanoma in them. I think that we know that there’s a patient population that clearly has disease that has not been taken out. That strengthens the reason for us to treat these patients.

An additional thing we need to remember is that all of these studies were done according to the AJCC Seventh Edition Melanoma Staging System. That has changed. We now have the eighth edition. The way that I look at this, the old stage 3a is the new stage 3a and 3b. That sort of splits that patient population, as well. We now have the new stage 3a patients with the thinner tumors—less than 2 mm. There is sentinel lymph node positivity with a small amount of disease in it. The question becomes, do those patients benefit from this treatment?

Jeffrey S. Weber, MD, PhD: And, is it appropriate to give someone with a 10% or a 15% risk of relapse and a minimal risk of death from disease over 5 years adjuvant therapy?

Robert Andtbacka, MD, CM: From that perspective, the common patient that we often see is a patient who has just a few cells in that sentinel lymph node, that has disease in it. In general, we tend to observe those patients. But ultimately, I think it requires a discussion with the patient. You need to talk about the risks and the benefits here. It is also a discussion to have with the team, as well. It’s even a cost discussion. Is it worth it now? We don’t know about treating now versus treating later. The KEYNOTE-054 data was a crossover study. Patients who were on the placebo arm, when they recurred, were able to get pembrolizumab. We don’t have that data yet. I think that’s going to be important for us to see, in deciphering the true difference between treating now versus later.

Omid Hamid, MD: And we will get some of that data when the CheckMate-915 adjuvant trial matures and is presented. That was based on no need for complete nodal dissection and just sentinel node dissection. It randomized a single-agent option versus combination regimen. So, some answers are coming.

Transcript Edited for Clarity