Treatment of Advanced NSCLC With Dual Immune Checkpoint Inhibitors: 4-Year Update of CheckMate 227

John V. Heymach, MD, PhD: Sandip, 1 thing that’s different from the drugs we’re talking about is the combination of PD-1 and CTLA4. We’ve got the CheckMate 227 study. We saw the update. Do you want to take us through that? There’s also a group that has brain metastases. Maybe you can talk about that.

Sandip P. Patel, MD: One of the key concepts of most of these frontline immunotherapy studies is that we’re talking about a duration of therapy for 2 years. We’re talking about 4-year survival, 5-year survival. That’s 2 and 3 years off their last dose of scheduled immunotherapy, showing that these immunologics, they’re prodrugs. The actual drug is the patient’s immune system and we’re getting functional durable responses. We saw some really nice data that we discussed earlier, about one-third of patients with PD-L1 high are making it to 5 years.

If we look at the CheckMate 227 study, in terms of the 4-year overall survival rate, for less than 1%—PD-L1 negative, the opposite end of the spectrum—it’s 1 of 4. These are remarkable results compared with traditional chemotherapy, where you may expect 5%, 8%, maybe 10% of patients to get out to that 4-year mark. Thus, we’re really moving the needle across the PD-L1 spectrum. Part of this is that PD-L1 IHC [immunohistochemistry] isn’t a great biomarker, but in the land of the blind, the 1-eyed biomarker is king, so TMB [tumor mutational burden] is 1 of the biomarkers that didn’t pan out from the study as 1 of the prespecified end points, tumor mutational burden. This is a great regimen.

For those doctors who treat multiple types of cancer, the ipilimumab dosing is different. It’s 1 mg/kg every 6 weeks continuous. This is very different from what you give for induction in melanoma or renal cell carcinoma. The toxicity profile is different. It’s a lower rate of colitis for these patients compared with the higher-dose induction regimen. I just wanted to make that distinction. But after 2 years of therapy, a quarter of patients—even if they’re PD-L1 negative—are approved for this regimen in the United States based on its complicated analysis for PD-L1–positive patients, greater than 1%. In that cohort, if you’re greater than 1%, about 1 of 3 made it. It’s remarkable, especially in a nonsquamous cohort, for all these regimens to have such dramatic activity.

One asterisk is that Merck conducted a study adding ipilimumab 1 mg/kg every 6 weeks to pembrolizumab. That was the KEYNOTE-598 study, where the PD-L1 was greater than 50%. In that study, there was no difference. It reads like a statistical Rubik’s cube—high complexities on which patient populations benefit from particular combinations. I know we’ll be talking about some of the chemotherapy–I/O [immuno-oncology] combinations later, but 1 of the areas in which ipilimumab may play a role is in brain metastasis. This is a burgeoning body of literature also in the melanoma space as well, in which patients with CNS [central nervous system] metastasis—in a cross-trial comparison, of course—tend to have more robust intracranial responses with the addition of a CTLA4 agent in combination with PD-1 relative to historical data with PD-1 alone. This is a subgroup of patients in my own clinic, along with PD-L1 negative and squamous cell. I’ll consider a CTLA4 plus PD-1, more commonly with chemotherapy, which we’ll discuss later, because of some of this provocative data.

John V. Heymach, MD, PhD: Thanks. That was great—all the points, you covered. It’s interesting that in the PD-L1 greater than 50%, it doesn’t look like ipilimumab or CTLA4 adds benefit, but it’s clearly beneficial. Also, you pointed out that the label says greater than 1%, but the less than 1% had the same hazard ratio for survival. This just had to do with the design of the study as well. Interesting cohorts. One other interesting thing about this regimen is that not everybody responds. But if you do respond, the duration of response is surprisingly long. There are those early progressors that happen as well. Do you use the regimen? If so, what’s the approach? What sort of patients are you likely to use this in?

Sandip P. Patel, MD: It’s a great question, and it’s great to have all these options. In my clinic, I’m utilizing KEYNOTE-189 trial data, so carboplatin-pemetrexed-pembrolizumab for most of the nonsquamous. With PD-L1 less than 1%, patients with squamous histology, and patients with CNS [central nervous system] metastasis—I know we’ll be talking about this soon—we use CheckMate-9LA, which is 2 cycles of chemotherapy plus CheckMate 227. The reason for that is related to PD-L1 negative, the approval, and also those 2 cycles of chemotherapy avoid those early progressors. You can eat your cake and have it too, in terms of helping those patients. Because it’s hard to predict, even for PD-L1 greater than 50%, which patients will respond or have weak responses to I/O vs have that early progression. We want that ability to predict. Even in the PD-L1 greater than 50% space...hypothesis is chemotherapy-pembrolizumab. Then I have to reject that by low-volume disease, lymph node only, asymptomatic, where there’s enough of a runway. But for patients with bulky disease, PD-L1 greater than 50%, I’ll often do chemotherapy-pembrolizumab because I can’t predict who’s going to be the I/O monotherapy responder.

John V. Heymach, MD, PhD: We’re having the same experience at The University of Texas MD Anderson Cancer Center. We have a study opening soon taking patients who progress on ipilimumab-nivolumab and adding the chemotherapy after progression, to see if you can rescue those the way with the CheckMate-9LA you start with adding chemotherapy. We’ll look forward to seeing that study move forward.

Transcript Edited for Clarity