Jack West, MD: Let’s turn to what this actually means in practice. What would you say are the distinguishing features now, Charu, that will lead us to favor one treatment over another, both in the nonsquamous and the squamous populations?

Charu Aggarwal, MD, MPH: So, I think, as before, what hasn’t changed is that I still look at histology. I think histology still drives treatment selection. I think getting a sense of what mutation may be driving a particular patient’s cancer is still important. But for those patients who don’t have any driver mutations, I think triplet is very relevant, be it any histology—nonsquamous or nonsquamous—we have data to support triplet therapy for both. However, I will say that even though some people may write off PD-L1 now and ask, “Well, the benefit is for all PD-L1 subgroups, why do you even need PD-L1?”, I will still say that PD-L1 is important because there is that subset of patients with really high PD-L1 expression, about 50%, where if we go back to the KEYNOTE-024 data, can we just use pembrolizumab monotherapy and perhaps spare them the nephritis or the neuropathy that may accompany some of the doublet chemotherapy adage to the pembrolizumab?

And then again, in the same vein, I think for the completely PD-L1 negative, the 0s—you know, data from TMB, as well as CheckMate-227—are there some patients where we can spare chemotherapy? Perhaps just use an I-O combination or just not give them immunotherapy? So, I think we still have questions at both ends of the spectrum. But really, clearly, there are many more treatment options for our patients now.

Jack West, MD: My view is that we’re increasingly in a world of, for at least the patients without a driver mutation, you’re giving pembrolizumab, and the question is, is it with or without [a] histology-appropriate doublet? For the high—PD-L1 group, I would say the data are very strong with either choice. But in a patient who I don’t feel an urgent need to give everything up front and those patients where I would still try to frontload—it would be those becoming increasing symptomatic, losing weight rapidly before you start them, or the ones who have had marked progression between their initial ERs, CT scan, and the PET CT who got to finishing staging 3 weeks later—those are the ones who I’d try to give a combo up front and just give everything you’ve got. Cause even with monotherapy, the response rate’s 45%, not 90%. But for a lot of the others, if you can spare them the toxicities and the other challenges of chemotherapy for a long period of time when it may just be along for the ride, I would rather do that and spare the chemotherapy doublet for later. But for the lows and the negatives, most of those I would be most inclined to give a combination with either the KEYNOTE-189 regimen for the nonsquamous patients or the KEYNOTE-407 regimen and now one of the taxanes for squamous. What are your thoughts on this?

Gregory J. Riely, MD, PhD: I think I really want to emphasize what Charu just said, that we still need PD-L1 testing for all of our patients.

Jack West, MD: Yes.

Gregory J. Riely, MD, PhD: We still need it, and we still need mutation testing. We still need to understand which patients have EGFR mutations, we still need to find those with ALK-positive lung cancer. Because the treatment really depends on knowing these things for all of our patients. So, I think this is something where we can’t let our enthusiasm for combination chemotherapy and immunotherapy, regardless of PD-L1 status, take over. While it’s a huge accomplishment, we still need to pick our patients and we still need to be good doctors and find the right therapy for the right patients.

Jack West, MD: Yes, you make a great point that we wouldn’t want to just say, “You don’t even need to think about it anymore. We will give you a platinum doublet and immunotherapy. You’re never going to be wrong.” Yes, it still is valuable to distinguish who might be most likely or least likely to benefit.

And another thing that was not a featured presentation at ASCO but is worth thinking about is the group from UCLA reporting their results in a very small study of giving first-line pembrolizumab to high—PD-L1 patients with an EGFR mutation instead of first-line EGFR inhibitor. It showed horrible results, that none of these patients with a proven EGFR mutation responded and 2 of 10 died in the first 6 months, which is very unusual with this. It really suggested that these are not the same patients and we shouldn’t take shortcuts and not do the testing for these patients who are going to be much better certified doing targeted therapy first. It doesn’t mean you would never give immunotherapy, but one of the toxicities was pneumonitis developing on erlotinib after discontinuing pembrolizumab. Just highlighting that sequence may matter.

Transcript Edited for Clarity