Late-Stage Trials: Immunotherapy Combinations for NSCLC

Transcript:Mark A. Socinski, MD: We’ve established that the standard of care in the second-line setting has changed. It’s now 1 of the 3 immunotherapy agents. We’ve established that in the first-line setting, there’s a small population that does better with immunotherapy. We’ve had some tantalizing data about combining immunotherapy with chemotherapy, and the next wave of data sets that we’re going to see over the next year or so really fall into 2 categories. The first I’d like to talk about are the Phase III chemotherapy plus or minus immunotherapy trials. I think David alluded to the KEYNOTE trial that’s really the follow-up of KEYNOTE-021G. We have a number of trials with atezolizumab. In fact, all of the agents, or many of the agents, are in phase III testing.

David R. Spigel, MD: Yes, it has been a busy year.

Mark A. Socinski, MD: Yes, and the question is, how’s this story going to end?

David R. Spigel, MD: I don’t know. I think in the next 6 months, we will all be wrestling with results coming out, as you mentioned, with first-line atezolizumab and bevacizumab and platinum-based regimens. We’ll be wrestling with nivolumab and ipilimumab data.

Mark A. Socinski, MD: That was the second camp, the immunotherapy combinations.

David R. Spigel, MD: Yes, as well as durvalumab and tremelimumab coming out—all in this late fall time frame. What I spend too much time thinking about is, if all those hit, what would you use? And that’s difficult, because you want to see the data, how good they were, and what the toxicities are that you pay for. But it could be an environment like that a year from now, where we have multiple options in the first-line setting.

Mark A. Socinski, MD: You can only imagine how confusing it is to the community oncologist who doesn’t focus on lung cancer. It’s crazy.

David R. Spigel, MD: Yes, and a little scary, too. Not only must we understand how to deal with these drugs, but you also know these immunotherapy—immunotherapy combinations can be a little bit hard to handle sometimes. If you have all these things out there, I do worry a little bit about a lot of different toxicities that doctors aren’t used to with just single-agent checkpoints.

Mark A. Socinski, MD: We’ve been very comfortable with platinum doublets for a long time. We’ve been very comfortable with adding targeted therapies to platinum doublets in terms of managing them. Like we say, most of the time it increases toxicity, but it’s not prohibitive toxicity and we can manage it. You start to get into the I-O combinations and this is forging new frontiers, so to speak. What are your predictions about the first-line setting? I think the first data set we’ll see is from the MYSTIC trial, which is single-agent nivolumab versus durvalumab/tremelimumab versus standard platinum-based chemotherapy. That’ll be the first to read out. What are your thoughts? Did anyone have any early predictions on that trial?

David R. Spigel, MD: The other thing I’ll just say is, all these trials are designed to hit on every level. They have several primary endpoints.

Mark A. Socinski, MD: They don’t want to lose.

David R. Spigel, MD: They’re stratified for PD-L1-high and negative patients, so you don’t know what scenario will be positive or not. I’ll let you comment on it though.

Mark A. Socinski, MD: Ross?

Ross Camidge, MD, PhD: Well, I think durvalumab versus chemotherapy is ringing bells whilst we had the positive KEYNOTE-024 trial. We had the CheckMate-026 trial, which was a negative study. That’s why I’m asking the question, what are the differentiating factors? Is it just the level of enrichment for PD-L1? But tell me what you think about CheckMate. There were some issues other than the PD-L1 level, then.

Mark A. Socinski, MD: Yes, there were issues. There are a lot of lessons in trial design in thinking about CheckMate-026. We only used 2 stratification factors. There was an imbalance in known prognostic factors—sex, presence of liver metastases—that favored the chemotherapy arm. So, that arm performed better than average. If you look at the overall survival and Kaplan-Meier curves, both arms had overall median survivals of about 14 months, just shy of that on both sides. And those survival curves looked completely superimposable.

So, usually in lung cancer 2 drugs beat a single drug. Here we have a situation where you can argue in a relatively unselected population, meaning greater than 1%, that that’s the majority of patients—we have a single agent that looks very similar to a platinum-based doublet. But, at the end of the day it was a negative trial. I look at this like I look at the EGFR TKI story, initially using it in unselected populations where we knew that the majority of patients were wild-type and not getting substantial benefit, but there were these patients in there who might be driving whatever benefit you would have seen. But then when we isolate those patients, akin to KEYNOTE-024 with the high expressors, you see the real value of the drug in that population. In CheckMate-026, we just didn’t enrich enough, plus you add a few items of bad luck like the imbalance in female sex, which is a known prognostic fact.

David R. Spigel, MD: But you know, Ross, I don’t know what to guess. Would the MYSTIC trial be positive or not, or the NEPTUNE trial be positive or not?

Mark A. Socinski, MD: Right. The CheckMate-227 trial is a similar design with nivolumab/ipilimumab.

David R. Spigel, MD: Yes. The differences between the CheckMate-026 and KEYNOTE-024 studies have not made me confident to guess which of these trials would be positive.

Mark A. Socinski, MD: Right, right.

Jared Weiss, MD: I won’t guess which are positive, but I will guess how the bars will be set by the oncologic community. Just as there may be a culture against drugs such as bevacizumab, I think there is a very strong cultural tidal wave in favor of immunotherapy, in particular for immunotherapy without cytotoxics. And so, I think compared to come of the other conversations the bar is going to be lowered a little bit and people are going to be looking in the broader community for how to look at these combination immunotherapy trials positively.

David R. Spigel, MD: Although, you’ll have probably 6 months, maybe a year, worth of experience of carboplatin/pemetrexed/pembrolizumab and a lot of comfort there. And then, you start talking about CTLA-4 combinations. Doctors might still be most comfortable with a carboplatin/pemetrexed backbone.

Mark A. Socinski, MD: Yes. So, I guess the question is, how does the panel feel? If you had to place a bet today, are you betting on adding immunotherapy to chemotherapy because that’s our comfort zone? Or, are you betting on the immunotherapy combination strategy with the 2 regimens that we have out there?

Ross Camidge, MD, PhD: I’m betting on the path of least resistance, which is that people will start to use pembrolizumab in combination with chemotherapy and not worry about PD-L1 testing. I’m not saying that’s scientifically valid, but I think that’s what most people will do.

Mark A. Socinski, MD: Others?

David R. Spigel, MD: I think with the nivolumab/ipilimumab and durvalumab/tremelimumab trials, something has got to hit. There are too many endpoints that are powered to show some difference and every study so far has confirmed, even in melanoma, that the response rates are higher. The toxicity is also higher. Whether or not that translates into a PFS benefit and an OS benefit—you may not even need an OS benefit, just a PFS benefit—is unclear. But I’d bet on the immunotherapy combinations.

Mark G. Kris, MD: I’d bet on them too, and I think a lot of it depends on goals of care here. I think there’s such a push away from chemotherapy and it’s a win if you can do something without chemotherapy. And I see that trend continuing, I really do. I think that’s what’s going to happen. My money would be on the immunotherapy combinations, and people are going to believe the data that support their belief. That’s what people are going to do.

Ross Camidge, MD, PhD: It’s called echo chain bearing.

Transcript Edited for Clarity