Comparative Effectiveness of Chemoimmunotherapy vs Immunotherapy for Non–Small Cell Lung Cancer

PracticeUpdate: How are decisions regarding use of chemoimmunotherapy versus immunotherapy made in the first-line setting for patients with advanced non–small cell lung cancer?

Dr. Goldberg: We've seen several studies over the last few years that have tried to inform us on how best to use immunotherapy as first-line treatment for non–small cell lung cancer.

Single-Agent Immunotherapy

We've seen the single-agent immunotherapy trials, including trials that include pembrolizumab, atezolizumab, and then more recently, cemiplimab. It seems from the single-agent trials, that really the majority of the benefit of single-agent immunotherapy is in the patients who have PD-L1 expression of at least 50%, and that's compared to chemotherapy.

I think that that really has become our standard as far these high PD-L1 expressing tumors, is to typically use a single-agent immunotherapy. There are cases when we might use combination, but depending on clinical features, if there's a high tumor burden or really progressing disease, we might want to add chemotherapy. But I think, typically, the high PD-L1 expressors of greater than 50%, we typically will think about using single agents.

It gets more complicated in the middle of the road PD-L1 expression, the 1% to 49% expressing tumors. There we have several different options that are reasonable to consider and have data to support them. A trial of pembrolizumab versus chemotherapy that included patients with PD-L1 at least 1% had a positive result and an improvement in overall survival for pembro versus chemotherapy and led to approval. Technically, single-agent pembro is an option for patients who have PD-L1 of 1% to 49%.

Combination Immunotherapy

There's also several other combination regimens where the combination seems to have an improvement in overall survival compared to chemotherapy alone. That includes chemotherapy plus pembrolizumab, chemo plus atezo, chemo plus atezo and bev, and then also ipi-nivo, as well as ipi-nivo plus chemo.

All of those, independently, have an improvement in survival compared to chemotherapy alone. How to make the decision of which treatment to use is complicated. It would be great one day when we have head-to-head trials comparing all of these different regimens. In the meantime, we're using clinical factors and patient wishes, and some patients wish to avoid chemotherapy. The use of ipi and nivo can be considered for the lower PD-L1 expressing patients.

Typically, I think, we're still using combinations in the moderate PD-L1 expressors, the 1% to 49%. In the low PD-L1, the less than 1% or 0%, the expression there is really where we're thinking about combinations and not single-agent drugs really at all. That really hasn't been shown to be effective, or hasn't really been studied as first-line treatment.

In the less than 1% group, again, there are several options that now we have. We can think about, again, chemotherapy plus immunotherapy, and that can be with single-agent immunotherapy along with the chemo, or it can be ipi-nivo along with chemotherapy. Again, all of these regimens compared to chemotherapy seem to have benefit, and using patient-specific characteristics and wishes is really, I think, the best way to think about making these decisions at this point, while we don't have prospective data comparing these regimens.

PracticeUpdate: Your group took a deeper dive into the data regarding comparative effectiveness of these two approaches. How did you conduct your analysis?

Dr. Goldberg: There's been several attempts to try to compare these different regimens amongst each other because, again, all of the data we have from prospective trials compares the newer regimen, the immunotherapy-containing regimen, to chemotherapy alone. How do the regimens compare to each other?

We looked at this by doing a network meta analysis, and we looked at all of the various first-line trials of immunotherapy alone, or immunotherapy with chemotherapy, and tried to see which of the regimens seemed to have benefit. What we found was that the regimens all appeared basically similar to each other in terms of overall survival. There were some differences in terms of progression-free survival, but what we found was that overall survival appeared similar. When you're looking at different regimens, depending on the PD-L1 expression, in the PD-L1 less than 1%, it's looking at combination immunotherapy with ipi and nivo compared to chemo and immunotherapy together.

In the PD-L1 positive patients, it's looking at single-agent immunotherapy compared to chemoimmunotherapy. Again, we found really no difference in overall survival with the methods that we used with these different regimens, basically lending support to using your clinical factors and patient preferences to guide decision making.

PracticeUpdate: A study at ASCO looked at the difference in outcomes between these two approaches in the first-line treatment of advanced NSCLC. How did this compare to your group's meta analysis?

Dr. Goldberg: At ASCO, we saw another study that aimed to do similar things that we did in our network meta analysis. They aimed to look at a difference in outcomes between single-agent immunotherapy and combination chemotherapy with immunotherapy in the first-line treatment of advanced non–small cell lung cancer. They specifically were looking at the PD-L1 scores between 1% and 49%, this low, positive PD-L1 expression, where we have a lot of different treatment options and it's not entirely clear which one is best.

They pulled data from several different studies, and what they found was that there did seem to be an improvement in overall survival in the chemotherapy plus immunotherapy arms of these trials compared to immunotherapy alone. Again, this isn't a prospective trial that compared these two different types of regimens, but it's using the various different trials and then looking across the trials at IO alone versus chemo-IO.

Again, they found something different than what we found in our study. The way that we did our study must've had different methods than theirs, again, because we found different results, or it's possible that there were different trials included. Regardless, what they found was that for the PD-L1 1% to 49% expressing patients, there was an improvement in overall survival with chemo-IO versus IO alone. The medians there were 21.4 months with chemo-IO and 14.5 months with IO alone. A reasonable difference there: the hazard ratio was 0.68.

I think this is intriguing. I think many of us were doing this clinically, anyway. We were treating patients with chemo-IO who had these moderate PD-L1 expressing tumors of 1% to 49%, even though the study with pembro showed benefit in the 1% or greater patients, I think we were typically using chemo-IO in these patients anyway.

It really does show that seems to be the preferred way to go, when feasible, to give chemotherapy plus immunotherapy. There are prospective studies that are ongoing to compare these two types of regimens, IO alone and chemo-IO. I think those studies will, of course, be the definitive answer in the next few years, but until then, we have this study that shows that chemo-IO probably has an improvement in outcomes compared to IO alone and should be considered when a patient's not on a clinical trial.