Combination Chemo + IO Therapy in Advanced Non–Small Cell Lung Cancer

Transcript:

Balazs Halmos, MD, MS: A combination of chemotherapy and I/O [immuno-oncology] is a favorite regimen for many specialists. Kristen, explain the data that support the use of chemotherapy–I/O.

Kristen A. Marrone, MD: KEYNOTE-189 and KEYNOTE-407 looked at chemotherapy plus pembrolizumab. There was an update to the survival for the KEYNOTE-407 regimen in squamous cell carcinoma of the lung. The median overall survival was 17.2 months, and the 5-year overall survival rate was 18%. That was an exciting update. The other thing is that when they looked at patients who had completed all 35 cycles, the response rate was 91%. The 3-year overall survival rate after completion of that—5 years after they started the study—was 70%. That’s information to share with patients in the clinic in terms of what we’re looking for and what individual responses mean.

The EMPOWER-Lung 3 study with cemiplimab plus chemotherapy in both squamous and non-squamous was recently published. That showed median overall survival at 22 months vs 13 months. That study was a little late to the party in terms of approval and regulatory evaluation, so that’s not something I have a lot of experience with. But we can talk about how you’ve thought about that regimen and how it will change your practice.

Balazs Halmos, MD, MS: Martin and Josh, do you have any data to add? What are your impressions? How do you use them?

Martin Dietrich, MD, PhD: Those are 2 separate patient populations. We’ve seen some encouraging data in the squamous subset for EMPOWER-Lung 3 that’s equivalent to KEYNOTE-407. The question is why we would change practice. CheckMate 9LA focuses on a very different patient population. Because it’s missing a chemotherapy-nivolumab arm, it has to be contextualized in some transfers with CheckMate 227. Otherwise, it’s difficult to interpret. The main data set is PD-L1 negatives against PD-L1 positives. They look identical in terms of the long-term outcomes. That’s something that we didn’t expect and hadn’t seen in a similar way before. Those are all legitimate regimens in those respective spaces, but it has to be tailored to the clinical and molecular situation for treatment.

Balazs Halmos, MD, MS: Do you find it appealing that CheckMate 9LA used only 2 cycles of chemotherapy?

Martin Dietrich, MD, PhD: This is 1 of those questions like with 3 cycles of neoadjuvant chemotherapy: how much is enough? How much do you need? The typical time to response for immunotherapy is 2 to 3 months. It may be nice to have an additional option to extend chemotherapy depending on how symptomatic patients are.

Later on, we’ll discuss the POSEIDON study. They’ve done a trial design feature that’s a combination immunotherapy outside the context of immunosuppressive chemotherapy. I think of chemotherapy and immunotherapy as 2 antagonistic mechanisms supplied at the same time. Moving away from chemotherapy may unleash some helpful immunotherapy options. This would be a very good discussion to stir up big controversy: 2 vs 4 cycles of chemotherapy. Nobody is going to want to answer that prospectively.

Balazs Halmos, MD, MS: I spent the first 10 years of my career discussing that issue. That was enough.

Joshua Reuss, MD: In KEYNOTE-407, the squamous PD-L1 negative benefit is less clear. That’s where CheckMate 227 and CheckMate 9LA are very appealing, with PD-L1–low and PD-L1–negative squamous cells in particular.

Regarding the 2 vs 4 cycles of chemotherapy, you can make an argument in this format that clinical practice is a different ballgame. If we give 2 cycles of chemotherapy now, we can always add chemotherapy down the road. But if that patient is progressing, you’re thinking, “What study can I consider? What are some other things I can offer this patient?” You might add chemotherapy back in as a stopgap. I’ve done that, but it’s not a long-term solution.

We need to consider patient-specific factors in making these decisions. How functional is someone? We can talk about their NGS [next-generation sequencing], subgroups, and what we’d do in an ideal scenario, but patient status is going to drive functional status and these decisions as well. We can talk about the benefits of adding CTLA4. I’ve given CheckMate 227 and CheckMate 9LA a few times, but as my patients progressively develop grade 3 colitis and hepatitis, that’s less appetizing. To quote 1 of my colleagues from a recent congress: “Is the juice worth the squeeze?” That’s a question we have to ask ourselves about the litany of regimens that are available.

Balazs Halmos, MD, MS: Is there anything that you do differently for a patient who is frail and older when you come to these big decisions?

Kristen A. Marrone, MD: Numbers don’t always tell the full story. Knowing how functional people are is important, and so is looking at age. I’m more lenient with a few of my older patients. For them, I’d call 50% high. Sometimes I’ll consider I/O alone for those patients, especially because they often have comorbid conditions, such as renal insufficiency, so chemotherapy is going to be harder for them. I take age into account.

It’s about the individual risk stratification of when you talk to them and what they think. What have they heard about chemotherapy? Are they willing to try it? That factors into these decisions. The most important thing has been burden of disease and putting that together with their age and talking through how much they need chemotherapy. Especially our older patients, we have only 1 chance up front to make sure we’re picking the right thing for them. Our older patients are the ones who fall off more quickly if we don’t get it right at the first shot, so that’s my impression.

Martin Dietrich, MD, PhD: I’ll echo that. In my practice in Florida, we see challenges with chemotherapy for patients aged 65 to 75. Chemotherapy isn’t necessarily helpful. For the 1%-to-49% patients, there is minimal improvement over backbone with single-agent immunotherapy. Chemotherapy does have a role in the non-squamous population. We’ve made progress in terms of tolerance, but for the long-term prognostication of a patient, chemotherapy has very little impact mechanistically and all the clinical data follow suit.

Transcript edited for clarity.