Combined Immunotherapy and Chemotherapy for NSCLC

Transcript:Mark A. Socinski, MD: I raised the issue of the KEYNOTE-21G trial. A couple of weeks ago, we had an accelerated approval for the use of pembrolizumab in combination with standard platinum doublet carboplatin/pemetrexed in nonsquamous non—small cell lung cancer. They went as far as to say that it did not matter what your PD-L1 status was for that regimen. Jared, I’d like you to give us your perspective on that, and I’ll ask the others.

Jared Weiss, MD: Well, my bias is that I was trained by the presenter of that trial. So, that was a randomized phase II trial with nonsquamous histology, as you mentioned, and pemetrexed/carboplatin versus pemetrexed/carboplatin/pembrolizumab. What we knew before ASCO this year was that there was a PFS advantage to the triplet and added toxicity. It looked roughly additive. But in the early data that we saw, there was no survival advantage. And I think, to the surprise of some, approval was granted based on the results of those data.

Mark A. Socinski, MD: I just want to ask, was that a general consensus—that people were surprised—with this accelerated approval?

David R. Spigel, MD: I think all of us have been a part of randomized phase II studies of about 60 in each arm where one arm looked better, and that did not lead to an FDA approval.

Mark A. Socinski, MD: Right, well said.

David R. Spigel, MD: The primary endpoint was response rate—55 versus 29, that looks pretty good. PFS looks better. But as Jared said, the OS doesn’t seem to be there. And what’s a little surprising is that the confirmatory phase III is just around the corner.

Mark A. Socinski, MD: Around 6 months.

David R. Spigel, MD: And it’s not like that trial hasn’t been done. It’s enrolled; we’re just waiting on the results. I’m supportive, and I’m hoping the phase III data will confirm it. It’s a mouthful in terms of the regimen, but it could be a step forward for patients.

Mark G. Kris, MD: Please remember, accelerated approval was set up just for this situation. You have some compelling early data and a confirmatory trial close on their heels, so why not approve it, knowing that more data will be coming in?

Mark A. Socinski, MD: Yes. The only issue there is, what if the confirmatory data do not confirm it? And then, what do you do?

Jared Weiss, MD: Then you take it back.

David R. Spigel, MD: We have this issue right now, actually in bladder cancer, with atezolizumab. But yes, hopefully—I agree with Mark. That’s why the system is set up this way. I think for those who are involved in research, the rules have changed a little bit. But for patient care, it could be a great thing.

Mark G. Kris, MD: Plus, they’re drugs that we know. We know that they each have single-agent activity, and it’s likely to be additive. The toxicity is likely to be additive. Then, more data will help us make those choices.

Jared Weiss, MD: And there are a little more data from ASCO this year. We got a peek at the survival curves: a hazard ratio of 0.69, but not reaching significance with those small numbers—P of 0.13. What was interesting was that this happened despite very heavy crossover from the chemotherapy arm to immunotherapy—about 75%.

Ross Camidge, MD, PhD: Yes. There are other criticisms. There are imbalances between the arms. There’s a 2-to-1 imbalance in never-smokers in favor of the experimental arm. So, who knows what it is. But actually, I think Mark brings up a very good point: We were struggling with saying, “Well, why did they approve it when the confirmatory trial is coming around the corner?” Maybe that’s exactly why they did it. If they messed up the approval, it wasn’t going to be for very long.

Mark G. Kris, MD: Again, it’s not a stretch. These drugs work in this disease, and they likely add in somewhat, but how much is to be decided. Jared Weiss, MD: I don’t think these results are adequate to mandate that a practitioner do this.

Mark A. Socinski, MD: That wasn’t where I was going with this. So, how many of you have given this regimen off of a clinical trial in a patient since the accelerated approval?

David R. Spigel, MD: I have a patient coming in next week where it has been ordered. So, we’re poised to start it with 1 patient.

Mark A. Socinski, MD: Ross?

Mark G. Kris, MD: Wait, did you check that person for PD-L1 status first?

David R. Spigel, MD: Yes, the patient’s negative.

Mark G. Kris, MD: OK. I think there’s a consensus that we’d all check first.

David R. Spigel, MD: Yes.

Mark G. Kris, MD: But everybody agreed to that, though: If they were high in PD-L1, we would treat them with a single agent?

David R. Spigel, MD: Yes.

Mark G. Kris, MD: But technically, this trial would let us throw chemotherapy on, as well.

Mark A. Socinski, MD: Well, that’s a question. It says it’s regardless of PD-L1 status. So, the implication is, “If it’s regardless, why do we have to test for it?” And we’ve been advocates of saying you do, because if you have a high expresser, single-agent pembrolizumab is an effective and safe therapy. Right?

Mark G. Kris, MD: Absolutely.

Jared Weiss, MD: Patients value, and I think oncologists value, a noncytotoxic regimen.

Mark G. Kris, MD: Correct.

Ross Camidge, MD, PhD: We should point out that even though they said regardless of PD-L1, the data they showed were that response rate didn’t change by PD-L1 status. They didn’t show anything about whether the PFS or duration of response varied. We haven’t seen that yet.

Mark A. Socinski, MD: And it’s very small numbers, so it’s going to be hard to tell.

Mark G. Kris, MD: But if you had a case where response rate was important, a very symptomatic patient, you can make a case for giving the 2 drugs—the 2 classes, rather.

Transcript Edited for Clarity