Patient Selection and Checkpoint Inhibition in RCC

Transcript:Robert A. Figlin, MD: Let’s just talk about that. So, you’re in a patient’s room and you’re talking to them after having progressed on frontline therapy, whatever the frontline therapy was. You have data that incorporates overall survival with progression-free survival. I think it’s important for the practicing clinician to understand how you’re going to use that data, and if you’re going to use that data, to incorporate that into decision making. Can you comment about that?

Daniel J. George, MD: Absolutely, because remember the way I translate progression-free survival is cancer control. Patients aren’t going to understand when you say progression-free survival. But when you say controlling your cancer growth, they get that. So, this is where context is everything, and, again, I’ll go back to what Elizabeth said because I think this is a really important point. What we’re talking about is their next therapy. But if I’ve got a patient who has a tumor that is “on fire,” I’m not sure they want to hear that. But if that’s the image that their tumor is just taking off, that’s a patient where I need cancer control. That would point me toward something like cabozantinib because I can get an overall survival benefit and I can get cancer control.

The problem is, when I look at the CheckMate-025 data and I see the poor risk patients getting the biggest survival benefit associated with that, that makes me think, ‘Gosh, that’s who I should go with for that.’ That’s the challenge. But I want to get back to what Elizabeth said earlier, and that is that poor risk doesn’t necessarily mean somebody that is in poor health. Poor risk means that that they have poor risk factors. And let’s be very specific here. There are three Motzer criteria of poor risk factors here that we’re talking about—elevated calcium, performance status, and anemia. So, if I’m looking at those criteria, and I think it’s somebody who doesn’t necessarily need that cancer control issue, they’re not suffering, they’re not as symptomatic, but maybe they’ve got a little bit of decreased performance status but I wouldn’t call it “on fire”—then to me that makes sense to say, “Okay, I want to try nivolumab in that setting.” And maybe I’m going to have a short leash for switching if they get symptomatic or more symptomatic. But maybe I’m also going to see some benefit there even if I don’t shrink the cancer, because even the stable disease patients could benefit.

Whereas, on the other side, I’m saying, “Look, if I’m really worried about the next 2 months in terms of complications, bone pain, and bone fractures, and skeletal events and things like this, that’s where I need cancer control. Cabozantinib makes sense to me.” I hope none of us are ever in this situation, but it’s going to come up. We have cancer patients that fit that, maybe not the myasthenia or Guillain-Barré. I haven’t seen that one, but I have seen this one. And I think you’ve got to be prepared to be open-minded enough.

Robert A. Figlin, MD: David, I want to talk a little science for a couple seconds. So, at 2016 ASCO there will be data talking about patients with mismatch repair deficiencies benefitting from immunotherapies. We know that, as Dan just pointed out, there are poor risk—categorized patients that seem to benefit from immunotherapies. Is there some science behind those observations?

David F. McDermott, MD: There may be. It’s actually one of the more interesting parts of the PD-1 story, which is for most therapies. Up until now, our best patients tended to do best with the treatment for chemotherapy, for example. But in these examples where we’re seeing some of the best impact of the immune checkpoint blockade, PD-1, PD-L1 are often in our worst-behaving tumors. And there may be a connection there between the genetics of the tumor and the response, which you alluded to with the mismatch repair story. Tumors that have more mutations—for example, tumors that are associated with smoking like bladder cancer, lung cancer, and tumors that are associated with other environmental exposures, like melanoma—things that on the scale of mutations, tend to have the group the most, are where we’re seeing some of our best effects. To me, that’s a very exciting finding. Where we had some of the least options for our patients is where we’re seeing dramatic effects.

And that connection between mutations and response may be that when a tumor has more mutations, it’s able to generate these so-called neoantigens, or new antigens, that are coming from the cancer that are then presented to the immune system, and the immune system recognizes some of them. It has more of a chance to recognize something as foreign and then will send T-cells to surround this antigen, surround this tumor expressing this antigen. And, in most cases, the tumor will defend itself with PD-L1 on the surface or in the microenvironment around the tumor. So, that tumor is going to grow and is going to spread, and the patient is in trouble, the prognosis is dismal. But, if you now have something that can block the interaction between the defense—PD-L1 and the T-cells that are sitting there at the tumor recognizing it but unable to kill it—you’re allowing those tumors to kill very specifically. And also that might explain why there’s less toxicity with PD-1/PD-L1 and some of these other strategies because the effect is more at the tumor than systemically the way you see it with interleukin-2 or CTLA-4 blockade. So, we don’t know yet for sure that that’s the story because we don’t have a lot of prospective randomized data. But certainly, a lot of the retrospective data point in that direction that with more mutations, more neoantigens, there is more of a chance of immune recognition where we can unleash the brakes and kill that tumor.

Robert A. Figlin, MD: Tom, some thoughts?

Thomas Hutson, DO, PharmD: I would like to do just a counterpoint to Nizar. The data sets that we have with the three newly approved agents—and we all agree that they’re going to provide great benefit to our patients—are immature across the board. So, we will find that there may be nuances between with the immunotherapies. There may be a population that seems to be more sensitive to that approach. I think, at the end of the day, using any of the three used in second-line is not doing your patient injustice, and I’ve been involved with all three of the drugs in development. I’ve seen all three of them work. And it just reminds us that it was a phase II trial with lenvatinib/everolimus, but our government and the regulatory authorities felt that the efficacy data were valid. It was a randomized trial that also had independent radiologic review. It was a higher-level phase II trial, and it takes me back to the days when Bob and I led the development of pazopanib in the United States, and we had similar issues when we would discuss it. No one wanted to try pazopanib. It was phase II data, and now people at this table will say pazopanib is the best agent we have so far based on the data that we have.

So, I think we have to be open-minded. And I don’t mean that, Nizar, you weren’t saying that you weren’t open-minded, and I hear that it was a phase II trial. But once you get your hands wrapped around the agents, I think they all have a role to play and we’ll figure out the nuances as time goes on. But, as we said, 6 months is all we’ve had with some of these agents on the market.

Transcript Edited for Clarity