Immunotherapy Leads the Way in Lung Cancer, but Creates Complex Treatment Decisions

Joshua Reuss, MD

Whether to recommend single-agent immunotherapy, dual immunotherapy, or combination chemotherapy has become the focus of frontline treatment discussions in advanced non–small cell lung cancer (NSCLC), explained Joshua Reuss, MD, who added that PD-L1 status, the longevity of the data, and durability of benefit are among the primary elements of treatment selection. 

“The easiest thing to look at is PD-L1 status. It’s an imperfect marker, but I believe we will have better markers for response in the future. Especially for PD-L1–high patients, you could feel comfortable with single-agent immunotherapy,” said Reuss. “When you start to get into an intermediate PD-L1 status, that’s where you could really have some individuality in terms of how you treat particular patients and take into account tumor histology, patient preference, along with other factors.”

In an interview with OncLive^® during the 2020 Institutional Perspectives in Cancer webinar on lung cancer, Reuss, a thoracic medical oncologist at MedStar Georgetown University Hospital, and an assistant professor in the Department of Medicine at Georgetown University Medical Center, discussed the lung cancer armamentarium, the role of immunotherapy combinations, and the significance of biomarkers. 

OncLive^®: Could you shed light on some of the landmark trials that we’ve seen with single-agent immunotherapy in lung cancer?

Reuss: There are several differentiating features between the studies that have resulted in the approvals of [single-agent] immunotherapy in advanced NSCLC. KEYNOTE-024 evaluated a PD-L1–high population, and it’s also the study that we have the most long-term data from. Although it might be tough to differentiate atezolizumab [Tecentriq] and pembrolizumab [Keytruda] other than the fact that one [inhibits] PD-1 and the other [inhibits] PD-L1, we really don’t have enough data to determine whether one therapy is better than the other.

[Given that] KEYNOTE-024 has the most robust long-term follow-up data, we can say that this is a regimen that has withstood the test of time, including crossover, and has durable benefit.

KEYNOTE-042 and IMpower110 [enrolled somewhat] different populations. In the case of KEYNOTE-042, tumor [proportion scores (TPS)] were 1% or greater. IMpower110 [enrolled patients with any] level of expression of PD-L1. There was a hierarchical design in IMpower110, so we know that there was a statistically and clinically meaningful benefit in the PD-L1–high population that did not appear to transcend into the lower PD-L1 subgroups. 

In KEYNOTE-042, although the overall study was positive in the TPS population of 1% or greater, the [benefit] did appear to be driven by the PD-L1–high population.

If you’re used to one therapy over the other, be it pembrolizumab or atezolizumab, I believe you could go with what your preference is. Pembrolizumab monotherapy is approved [for use] every 6 weeks and atezolizumab is approved every 4 weeks. I don’t know if we are going to see [a battle between] who can expand [the administration of the drug] to a longer interval. 

Not only does pembrolizumab have the longest [administration interval] that I know of, but it also has the most robust data. [Those are the things I consider when] selecting therapy for patients with PD-L1 expression of 50% or greater.

Could you highlight the role of chemoimmunotherapy, along with how you approach treatment selection? 

Ultimately, there’s a lot of provider preference regarding chemoimmunotherapy. For patients who have PD-L1–high expression, [single-agent] immunotherapy is an excellent choice. There are select circumstances where you might want to add in chemotherapy, such as if patients have high tumor volume, significant symptoms, or threat of organ compromise, and you’re worried [about responsiveness]. Immunotherapy takes longer than chemotherapy [to elicit a response], and we can’t always wait for that. That’s when I would typically add on chemotherapy to someone who’s PD-L1 high.

For those that have lower levels of PD-L1, there’s a combination of factors [to consider]. You have to look at the toxicity profile of the therapy, patient-specific factors, as well as the duration and longevity of the data. The KEYNOTE studies have more long-term data. For nonsquamous histology, we see a durable benefit to immunotherapy plus chemotherapy that really seems to transcend PD-L1 levels. This may not be the case for patients with squamous cell lung cancer that express PD-L1. 

When we’re trying to decide between pembrolizumab plus chemotherapy or atezolizumab plus chemotherapy or atezolizumab plus bevacizumab [Avastin] and chemotherapy, the hardest regimen there is the one from the IMpower150 study, which [consists of] atezolizumab, bevacizumab, and chemotherapy. Whenever you add another agent to a regimen, there are additional toxicities.

There are some patients where I might consider this [regimen] up front, such as if patients have significant tumor burden in the liver, for example. That was one of the subgroups that were evaluated in the subgroup analyses. I do tend to favor the combination of pembrolizumab and chemotherapy for the nonsquamous non–small cell population that has PD-L1 positivity. We could even make a case for [that regimen in] the PD-L1–negative and the squamous PD-L1–positive populations as well.

Could you spotlight the role of dual immunotherapy combinations?

Studies that are evaluating immunotherapy combinations, along with the combination of immunotherapy and platinum-doublet therapy, are the CheckMate 227 and CheckMate 9LA trials. The regimens evaluated [in these studies] were recently approved. Now that a lot of physicians have experience with single-agent immunotherapy and chemoimmunotherapy combinations, there’s some hesitancy to explore other regimens. This is due to the prior history of significant immune-related toxicities with nivolumab [Opdivo]/ipilimumab [Yervoy] regimens. [We saw] a lot of significant toxicities in earlier studies across tumor types with high-dose ipilimumab at 3 mg/kg as opposed to a lower dose of 1 mg/kg, which was dosed every 6 weeks. [The latter dosing schedule was used in] CheckMate 227 and CheckMate 9LA.

One of the big take home messages for CheckMate 227 is that the long-term outcomes of this study appear to be robust, irrespective of PD-L1 status or tumor histology. We’re seeing a median overall survival [OS] that does not appear to be significantly different in the PD-L1–negative and PD-L1–positive populations, with a 3-year survival [rate] around 33%. In addition, the median duration of response [DOR] in the PD-L1–positive population is around 2 years, and a there is a very similar 18-month median DOR in the PD-L1–negative population. 

One of the things that I would consider [with regard to treatment selection] is tumor histology. Based on the updated analysis of KEYNOTE-407, we can see that in the PD-L1–negative population, the benefit of pembrolizumab appears to be less robust and less clear, even after accounting for crossover. In that study, the median DOR was 8.8 months. So, when you look at a median DOR of 8.8 months vs 18 months or close to 2 years, there is a real difference there. That’s a population where we would strongly consider an immunotherapy combination in the right patient. Specifically, a robust patient who understands that immune-related toxicities are very different from chemotherapy toxicities. A severe immune-related adverse effect could be lethal and could lead to prolonged steroids and the inability to receive subsequent therapy.

[An immunotherapy combination] could also be beneficial in patients [with resectable disease] who might have progressed rapidly on adjuvant chemotherapy, or soon after durvalumab [Imfinzi] monotherapy in a patient with locally advanced disease treated with definitive chemoradiation. Those are populations where the regimen studied in CheckMate 227 would be reasonable to consider. I’m still not sure where the CheckMate 9LA regimen will fit into my practice; we need longer-term follow-up data for this population.

If you need the cytoreductive power of chemotherapy, is there really a benefit to giving nivolumab plus ipilimumab and chemotherapy vs pembrolizumab plus chemotherapy or atezolizumab plus chemotherapy? We may never a study [that compares those regimens], but [that’s something I think about] when [determining which] regimen [to recommend]. One advantage of the combination of nivolumab and ipilimumab [is that when] patients progress, a platinum-doublet should be used, which, up until this past decade, was the standard for patients with advanced NSCLC. 

If you administer the CheckMate 9LA regimen, we [could] consider platinum-doublet chemotherapy, depending on when patients progress, but it’s a little less clear [that we should]. Those are some reasons that I’m not quite sure yet, where the CheckMate 9LA regimen will be incorporated into my practice. It’s a little clearer for CheckMate 227 in all cases, but we just need more data and better biomarkers [for that regimen].

Could you expand on the significance of biomarkers in lung cancer? What challenges remain? 

When it comes to other biomarkers, we’re still not clear where they will fall into this decision-making tree, whether that be [tumor mutation burden (TMB)] or other aspects of [a patient’s] molecular profile. Now that we’re getting next-generation sequencing and detailed molecular analyses on all our patients, what are some other mutations or alterations that may affect this decision? That still has to be determined.

Another important factor is treatment schedule. So, if a patient can receive immunotherapy every 6 weeks as opposed to a chemotherapy regimen every 3 weeks or an immunotherapy combination every 2 weeks, that is huge in terms of quality of life. Patients are first and foremost people, with their own lives and their own interests. We must respect that. Unfortunately, the status of the pandemic still looms large. Multiple studies that are evaluating patients on chemotherapy believe patients will have worse outcomes with the coronavirus disease 2019 [COVID-19] pandemic. There are studies looking into the effect [of COVID-19] on patients receiving immunotherapy as well. 

How has the COVID-19 pandemic affected treatment selection? 

The data change so much. Early on, we felt that chemotherapy had a significant affect on mortality rates from the COVID-19 virus. Although, subsequent studies make that a little less clear. Similarly, with immunotherapy, there has been mixed retrospective studies.

Overall, given the neutropenia, lymphopenia, and the count depleting nature of chemotherapy, in a situation where you could go either way with adding chemotherapy to a regimen, let’s say pembrolizumab monotherapy, for example, I would lean on the side of [single-agent] immunotherapy. 

It was right around the time when the pandemic hit that pembrolizumab, administered every 6 weeks, was approved. Now, when you’re starting out treatment, you probably don’t want to start off at that frequency, just because you want to assess toxicity and [potential disease progression]. Those are reasons that I would steer towards [immunotherapy] in this era. 

It’s one thing to have a routine visit in the virtual setting but trying to assess a new patient and their functional status in that setting is nearly impossible, along with determining an optimal therapy to give them.

What other emerging therapies are you excited about? 

There’s always a new checkpoint [inhibitor] or a new immunotherapy agent that is exciting. One therapy that we’re excited about in the lung cancer field for patients without driver mutations is a TIGIT therapy called tiragolumab. The agent was evaluated in the CITYSCAPE trial, which randomized patients to atezolizumab with or without tiragolumab in the treatment-naïve setting. The patients were PD-L1 positive.

The results of that phase 2 study were exciting. [Though, the benefit] appeared to be primarily driven by the PD-L1–high population. We saw response rates of 37% vs 21% in the intention-to-treat population. In the PD-L1–high population, the response rates were 66% vs 24%. When looking at [progression-free survival (PFS)], there is a benefit that also appeared to be driven by the PD-L1–high population without significant increased toxicity, which is also important.

It’s exciting and it’ll be interesting to see which patients will derive the most benefit from this therapy. This [regimen] is going to lead to greater responses [than we have seen historically]. At the end of the day, aside from chemoimmunotherapy studies, only around half of patients respond to immunotherapy. 

What is your advice for approaching treatment selection? 

Any big academic center is going to have their own decision trees and their own ideas in terms of how to approach a particular patient. We have a lot of options. There is generally no wrong answer and you’re not going to completely ruin a patient’s long-term survival. The best thing you can do is have a discussion with your patient about the different options, the different toxicity profiles, and, ultimately, have a shared decision-making plan.

What is the take home message of your presentation?

We have many options now, but in reality, it feels as though we’re on a ship in a stormy sea. 

Functional status is still the most important factor in terms of determining a prognosis for a patient and, to a degree, [provides] predictive value in terms of how patients are going to do with a specific therapy. I believe the toxicity profile is important. There are patients that feel strongly that they want a chemotherapy-free regimen, either due to a family experience or other factors. If they are physically fit and understand the difference in the toxicity profile, they should absolutely play a role in the decision-making process