Leora Horn, MD, MSc, FRCPC: Right now, first-line therapy in greater than 1% patients—I think that’s going to lead us to the next set of trials that we’re going to discuss. A lot of data has come out in the past 6 months. Maybe we should break it down by squamous and nonsquamous histology. When we think about patients who do not have driver mutations—so we’re taking out the EGFR, the ALK, the ROS1, because we know that those patients should get targeted therapies—there have been 3 major trials in the last 6 months in nonsquamous non—small cell lung cancer. We had the KEYNOTE-189 trial that looked at carboplatin or cisplatin/pemetrexed with or without pembrolizumab. And then we had the 2 IMpower trials. We know about the trial that recently came out at the World Lung Conference looking at a platinum with or without atezolizumab. And then we had the trial that looked at carboplatin, paclitaxel, bevacizumab with or without atezolizumab in patients with nonsquamous non–small cell lung cancer—slightly different patient populations but similar enough. How are these 3 trials impacting what you’re thinking about in those patients who are less than 50%-positive but also in those patients with a high PD-L1 [programmed cell death ligand 1] score with nonsquamous non–small cell lung cancer?

Thomas E. Stinchcombe, MD: For the people who are high PD-L1, greater than 50%, I still preferentially use single-agent pembrolizumab. To me, these trials apply more to the under 50% group. If we look at those 3 trials, the IMpower150 and KEYNOTE-189 trials have overall survival benefit. With IMpower132, we have progression-free survival benefit, not overall survival benefit. I think KEYNOTE-189 showed a survival benefit in the intent-to-treat, all PD-L1s, and I think that’s a commonly used therapy in my practice. The carboplatin/paclitaxel/bevacizumab/atezolizumab regimen did show a survival benefit. Like you said, it’s very hard to compare patient populations in the 2 trials because bevacizumab tends to select for a patient population with a better prognosis.

With that 4-drug combination, I do have some hesitancy related to the taxane and the bevacizumab toxicities when there’s an alternative. For the carboplatin, pemetrexed, and atezolizumab regimen, I’m going to wait for the overall survival data to make sure that it shows a survival benefit. Then I think it’s going to be compared with KEYNOTE-189, in terms of what the hazard ratios are and other aspects. So, I’d say most of my patients get the carboplatin/pemetrexed/pembrolizumab combination.

Leora Horn, MD, MSc, FRCPC: Eddie, sometimes you hear people talk about these patients who are rapidly progressing. They need a quick response to therapy, especially those patients whose PD-L1 score is greater than 50%. Is there a group of patients with high PD-L1 expression for whom you are thinking about using the combination rather than single-agent therapy?

Edward B. Garon, MD: It’s interesting. Before I had both options, I really was not very excited about the idea that there would be a group of patients for whom I would want to add chemotherapy. I thought that may be about 10% or less. In my practice, I found it to be a little bit higher than that of people for whom I really do worry as to what happens to them over the next several weeks. As you know, there are patients who somehow subjectively feel better almost immediately from this. I have a hard time explaining that, from a biologic perspective. But, oftentimes, even in patients who respond, it’s going to take some time to see that response. There are patients for whom I worry about a delay in that response or the consequence of a lack of efficacy. There are certainly patients with a PD-L1 expression that is greater than 50% who have rapid progression despite a single-agent PD-1 inhibitor. So, there is certainly a group of patients who seem less clinically stable for which I have found a role for using both.

Also, I know that there’s been a lot of enthusiasm lately about cross-trial comparisons. People say, “Well, if you compare the KEYNOTE-024 data with the KEYNOTE-189 data, here’s what I would do.” The reality is, the control arms do quite differently between those 2 studies. And, in fact, if you wanted to use the KEYNOTE-042 data as your comparator group, with KEYNOTE-042, overall, even the 50% group didn’t do as well. And so, it’s not as clear. In general, my practice is to look in these greater than 50% patients and give them single-agent pembrolizumab. But, I am finding it creeping into my practice. There is a group of patients for whom I give the chemotherapy as well.

Leora Horn, MD, MSc, FRCPC: And what do we do for a patient when they’ve got toxicity? What do you stop? You’ve got a patient. They’re 7 months into therapy. They’ve had a response. They’re coming in and they’re saying, “I’m tired.” How do you decide? Do you stop the pemetrexed? Do you stop the pembrolizumab? Do you stop both? What drives your decisions?

Thomas E. Stinchcombe, MD: I initially try and stop the pemetrexed because I think the pembrolizumab is probably more active in this situation, going back to historical controls. By stopping the pemetrexed, I may alleviate the fatigue and any anemia. I guess the one exception is if I think there’s an immune-related toxicity. In that case, you are sort of forced to address the pembrolizumab.

Edward B. Garon, MD: The one additional thing that I’ll say about that is, I don’t know that we are going to be the ones making that choice. I think that the patient is often going to be the one making the choice. The promise of immunotherapy is so strong to the patients. I think we would have a hard time, in most situations, even if we thought it might be medically reasonable, to stop the pembrolizumab and continue with the pemetrexed. I think patients are going to have a strong desire to stick with the immunotherapy approach.

Leora Horn, MD, MSc, FRCPC: That’s a good point. I have patients who are hitting 2 years on their immune checkpoint inhibitors and do not want to stop. In these studies, at 2 years, patients, even on the combination therapies, were supposed to stop the therapy. I think that you’re right.

There’s 1 subset analysis in the IMpower study that people talk about a lot. What do we do with the EGFR- and ALK-positive patients who technically have not responded well to checkpoint inhibitor therapy. In IMpower150, there was some suggestion that those patients may have benefitted from the taxane/atezolizumab/bevacizumab combination. And so, are you using this in patients who progressed on their TKIs [tyrosine kinase inhibitors] and have exhausted their TKIs [tyrosine kinase inhibitors]? How are you interpreting this data?

Edward B. Garon, MD: From my perspective, I think it is intriguing data. There was a progression-free survival benefit. The overall survival looked impressive, although the confidence interval includes 1. Overall, I think that this is a somewhat appealing approach. We also know that this is a population of patients who, frankly, have not done well with single-agent immunotherapy. And so, I’m not that eager to hold out to give them single-agent immunotherapy. There is this suggestion of a benefit. Also, I’m intrigued by data from KEYNOTE-189, which excluded EGFR and ALK patients. Nonsmokers did appear to derive a very clear benefit from the addition of a PD-1 inhibitor. That’s a bit tempered by the CheckMate-227 data. There, they indicated that really was for patients who had higher mutational burden, which would be anticipated to be former smokers and patients without EGFR- and ALK-mutated disease. But, in my own practice, it is something that I have found patients to be very interested in. I have been willing to do it—either this regimen or another chemoimmunotherapy regimen that is available.

Leora Horn, MD, MSc, FRCPC: You see a lot of those patients. Have you been using it for your patients who are a little more like those in Tennessee, or where there are not as many EGFR- and ALK-positive patients in your part of the world?

Thomas E. Stinchcombe, MD: I generally offer them this combination as an option. Then, I offer carboplatin and pemetrexed without immunotherapy. When you talk to the patients, some of them prefer the carboplatin/pemetrexed profile. I think both are reasonable options at this point. If I do use immunotherapy, it would be in combination with the other 3 drugs. That’s really the subset for whom that’s proven to work, to date.

Leora Horn, MD, MSc, FRCPC: I guess we’ll have to wait a while for the data, but there is a trial that’s looking specifically at patients who progressed on their EGFR inhibitor. They are looking at carboplatin and pemetrexed with or without pembrolizumab. So that may help us answer one of those questions.

Transcript Edited for Clarity