Transcript:Robert A. Figlin, MD: David, let’s move on and talk about the complications and challenges associated with sequencing. We’ve just talked extensively about treating the frontline patient, newly diagnosed. We’ve had a very rigorous discussion about side effects, efficacy, and choice of agents. So, some of the challenges for our community doctors are, when is it time to change, and how do you decide when a second-line agent is really appropriate? And then, what’s the conversation going to start to sound like when choosing a second-line agent? Please address that as well as integrating the imaging in our decision making about what is truly progression versus a radiographic change that may or may not justify a changing of an agent.

David F. McDermott, MD: I think there are certain situations where that decision is relatively straightforward. So, if a patient is on a TKI in the front-line setting and they get a scan, and there are new lesions, for example, that person needs another therapy, in my mind. If the patient has symptoms of progression in some ways regardless of the scan—if they feel worse, they have new pain, new cough, something like that—that patient needs another therapy. I think the places where we all struggle are in the patients that are feeling well, they’re not tolerating the dosing schedule of whatever they were on because you’ve figured it out with your personalized medicine à la Dr. George, but their scans are worse. You know, how much worse is enough?

If they don’t have new lesions, but they have growth, we have a bias—certainly in our center because we have trials for those patients—to maybe pull the trigger on a second-line therapy sooner than in the community where you’re trying to extend the benefit of any one drug. But we roughly try to use RECIST criteria. But we’re not strict in adhering to that, even in our practice. But we like to move on when we see significant growth, I’d say 20%, 30% growth of the lesions that exist.

And then when you ask the question of what to choose next. Our group obviously has somewhat of a relative bias towards immunotherapy for patients both in the frontline—as Dan mentioned with IL-2—but also in the second-line with PD-1 blockade with nivolumab. I think the data from the CheckMate-025 phase III trial is about as clean a trial as we’ve ever seen in the second-line in kidney cancer. You see clear benefits and overall survival over everolimus, you get improvements not only in safety, decreased toxicity relative to everolimus, but also improvements in quality of life, which likely reflect both things—less toxicity and clinical benefit.

So, from our point of view, that trial confirms our bias to switch people in that direction. But it would be nice in some ways to compare, which we don’t have, checkpoint blockade with a TKI in second-line. We don’t have that data. For some patients that might be the right approach, and it’s hard to argue that it’s not right for some patients. We just can’t tell who should do which sequence. But I think in our practice, we use that second TKI for after they’ve had a chance to receive nivolumab.

Robert A. Figlin, MD: David, I’m going to put you on the spot because you’ve got a lot of experience in this area. Suppose tomorrow we had an immuno-oncology agent in the frontline setting. You’ve just articulated very nicely the use of WHO and RECIST criteria for PD—and there’s a whole group of efforts, including some that you’re involved in which are called immune-related-response criteria, or IRRC. How do you envision using that to make a decision about when to choose a second-line therapy? And how do you do it in patients that are on clinical trials in the frontline setting and then need second-line treatment, and how are those patients analyzed?

David F. McDermott, MD: That’s an advanced placement kind of question right there. Yes, I think it’s going to be hard. Assuming we have frontline approvals for some of these combinations, we’ll have to wait for it. That’s an assumption. I think most of us in practice outside of a trial will make that switch, whether it’s frontline failure of PD-1-based therapy or even second-line failure. Do patients develop new symptoms? In those patients, I’ve never seen a case of pseudoprogression where the patient felt worse, personally. So, in the best cases of pseudoprogression I’ve seen, I’ve seen new lesions but the patient comes in feeling great.

Robert A. Figlin, MD: In the context of this conversation and you have some presentations at ASCO this year, define for us what pseudoprogression is, how you characterize it, and what you do with that information, even in the context of a clinical trial.

David F. McDermott, MD: Pseudoprogression, or an unconventional response, has been classified as developing new lesions, for example, in the presence of decline of the original lesion. So, the stuff you started with is smaller but you developed new things, or in the setting of an overall decline after a progression. It is when you see growth in lesions that then subsequently leads to a shrinkage. That situation is more difficult to manage in my experience because most of the patients who actually grow initially are going to be true progressives. Those are going to be patients who are going to need something else.

Whenever I see growth on a scan, I don’t assume unconventional response, I don’t assume pseudoprogression. I assume it’s going to be real progression and we’re already starting the discussion of what are we going to do next. But then we try to give the benefit of the doubt, if a patient is not deteriorating asymptomatically and, say, let’s come back in 6 weeks or 8 weeks with another scan. If that scan looks worse, we’re going to do X, if not, we’re going to feel good because there’s some benefit to continuing treatment beyond progression.

I think what makes it hard is some of our best responses on some of those early trials—some of the people we remember the most—are the people who originally grew and then had a dramatic shrinkage or developed new lesions and then have a shrinkage. But most patients with growth on scans are really going to be true progressives. We have to keep in mind that with nivolumab, the best response for more than one-third of patients was progressive disease. So, we have to be aware that both things are possible and make sure we don’t prevent patients from going on to another therapy because they deteriorate while you’re waiting for that next scan.

Daniel J. George, MD: I think there are a couple of key points that you bring up here. And one is that you don’t necessarily have to see evidence of a disease of response immediately. Some of these patients are coming off a VEGF inhibitor. There could be a rebound effect. Just removing that VEGF inhibition—having that angiogenic growth factor come back—there could be some growth of existing lesions, and maybe even a new lesion that was there but not really apparent until it got reperfused. So, I’m always careful about that. But when we see multiple new lesions, new organ sites involved, symptoms, or impending issues or complications, that’s when we get really nervous about this. I think the other key point is that you don’t necessarily need a PR to get a clinical benefit from that agent.

David F. McDermott, MD: Right, absolutely.

Daniel J. George, MD: And what you’ve seen, what we’ve seen in the data with stable disease, overall survival in patients who have had stable disease or progressive disease as their best response really speaks to the fact that this may be working in an unconventional timing away that’s not going to necessarily cause that traditional tumor regression because it’s not direct treatment, indirect treatment.

Elizabeth R. Plimack, MD, MS: Right.

David F. McDermott, MD: Right. And that’s what we’re seeing we think in those responses; meaning the growing new lesions are, we think, inflammation surrounding lesions that weren’t recognized by the immune system that now can be recognized because the T cells can actually get there and kill the tumor. This was first noted in the CTLA4 experience with ipilimumab, but we saw patients with new brain metastases in melanoma. And we said this has to be disease so let’s take these patients to the operating room, only to find that when those lesions were removed, they were dead melanoma. So, in some cases what you’re seeing as worsening on the scans is really the effect of the drug, a T-cell response to the tumor.

Transcript Edited for Clarity