Steve, as you talk about recurrence and second-line treatment options, we saw an interesting study at ESMO [the European Society for Medical Oncology 2020 meeting] called the PIONeeR trial, which looked at later lines used, but was a biomarker-focused trial. Tell us about that study.

I like this study. It is important work, and this is not a therapeutic study. This was presented by our colleague, Fabrice Barlesi, MD, PhD, and is looking at patients who are receiving immunotherapy with either nivolumab, pembrolizumab, or atezolizumab in a second-line setting, or in combination with chemotherapy in the frontline setting. This was not an intervention study. Patients had a repeat biopsy and blood samples, trying to understand the science of immune responses and if we can understand better what’s happening. When we look at these drugs, they’re superior to chemotherapy. We see long-term survival and amazing results. People are startled to understand how little we understand about exactly how these drugs work and why they work in some and not others.

This is looking beyond PD-L1, and they saw some interesting things there. While PD-L1 did serve as one predictive marker of who’s benefitting, they saw an important role for tumor infiltration by cytotoxic lymphocytes, building on work done by our colleagues in melanoma, where if you have more immune cells within the tumor, those patients seemed to derive those antitumor immune-mediated responses, and high infiltration of regulatory T cells, coupled with weak PD-1, seemed to associate with worse outcomes. It’s not just activation through the PD-1/PD-L1 pathway. Those activated immune cells need to get to the tumor, infiltration into the tumor, and this type of biomarker-specific work is critical to increasing the sophistication of delivery for immunotherapy.

Yes, interesting, and it will be nice to know if you have a tumor that’s not responding, what can be done to increase the infiltration of T cells and to provide the right kind of microenvironment for those patients to respond better to therapy. This also brings up the importance of asking the question about the immune microenvironment and how that reflects the patient’s likely response to immunotherapy. We’re seeing some information here from various trials.

One study that I thought was informative is the subset analysis looking at the genomic characteristics of the tumor and how that relates to outcome. This was a subset analysis from IMpower150. This talked about the role of 

 mutations. These are mutations we see in lung adenocarcinoma, and 

 we see in squamous as well. It’s traditionally been thought that these mutations would be associated with poor outcomes with immune checkpoint inhibition, based on prior experience.

At the ESMO meeting, the IMpower150 investigators looked at the tumor samples in a subset of patients who had the 

mutations, and what was interesting is these mutations definitely have a prognostic impact, but the chemotherapy plus bevacizumab plus atezolizumab regimen was still better for 

 mutations compared to what they saw in degree of benefit in the wild-type population. There’s also something that we’ve seen with the pembrolizumab-based studies looking at the impact of 

. We’re still learning more about various genomic events that might either sensitize or portend a poorer outcome with immune checkpoint inhibition.

I couldn’t help but talk about TMB [tumor mutational burden] when we discuss biomarkers, and I want to turn this question to you, Christine. What do you think about TMB as a biomarker in this situation?

: There have been a lot of challenges with using TMB as a biomarker, as I’m sure all of my colleagues on this panel have also appreciated. One of the biggest issues that we’ve had is standardization of TMB across platform and between institutions. It’s hard to know how to apply a nonstandardized biomarker to some of this information. We’d all been hoping that TMB was going to pan out as far as a complement to PD-L1, giving guidance and putting patients into buckets of high PD-L1, low TMB, what does that mean, how can we choose our therapy? But unfortunately, it doesn’t appear that we’ve had quite as much efficacy using TMB as a biomarker as with PD-L1.

: Any other thoughts on the TMB issue? Steve?

Part of the reason why we keep coming back to TMB and why we won’t let these negative studies put the issue to rest, is it just makes sense. It’s a rational biomarker. When we have more somatic mutations, you expect more neoantigens, and you’re just increasing the odds that you have a T-cell receptor that’s going to interact. It’s such a rational biomarker, that it seems like it should work despite evidence to the contrary, and we need to continue to refine that. But as of today, it doesn’t influence my treatment practice.

Videos

Understanding Immune Response in Non–Small Cell Lung Cancer

For us, having heard this, who would use the chemotherapy plus IPI/NIVO [ipilimumab/nivolumab] approach, and if you choose that approach, is 2 cycles of chemotherapy, in your impression, sufficient?

 Suresh, before I go to the chemotherapy/nivolumab/ ipilimumab, the nivolumab/ ipilimumab combination has proven itself in multiple other tumor types prior to the non–small cell lung indication. We’ve seen that in melanoma, we’ve seen that in GI [gastrointestinal] and renal cell carcinoma, and here it is the same story in non–small cell lung cancer. If you believe the science behind this combination improving the function, as well as the quantity of the T cells, then you would expect to see better results with this combination. What is interesting about this combination, it tends to overcome histology, it tends to overcome PD-L1 suppression, since we’re seeing the same results regardless of histology as well as regardless of PD-L1 status.

Now back to your question on the combination of I/O [immunotherapy]-I/O plus a platinum doublet, the CheckMate 9LA protocol, I was one of the primary investigators on this study. It’s a very nice, well-designed, phase 3 clinical trial looking at the standard of care chemotherapy platinum doublet vs just 2 cycles. This is a novel approach. We’ve never done just 2 rounds of chemotherapy, we tend to do 4 to 6 cycles, that’s the standard in our clinic. But it’s a novel approach with only 2 doses of chemotherapy, trying to get some of the synergy from chemotherapy, plus Opdivo [nivolumab] and Yervoy [ipilimumab], and then patients will go into Opdivo/Yervoy maintenance.

The dosing is different. Yes, we’re used to giving Opdivo/Yervoy in our clinic, but we’re not used to giving Yervoy at the lower dose, but I don’t think any of my colleagues here will argue against a lower dose of Yervoy. I think we’re just going to give enough of the drug with limited toxicities or adverse effects. It was only 1 mg/kg every 6 weeks. Then the primary end point is to look at progression-free survival as well as overall survival, and the study did meet its primary end point, and about two-thirds of patients are still alive at 1 year with this quadruplet regimen.

Now, is this an appropriate regimen for everybody? I don’t think so, but definitely it’s a good option. If you don’t want to miss the opportunity of using chemotherapy in the first-line setting, but at the same time you don’t want to give the full force of 4 to 6 cycles, then here is another option for our patients.

Great. As we talk about these new regimens, we don’t want to forget the ones that we’ve already had in the clinic. In fact, we heard about long-term outcomes with pembro [pembrolizumab] in the high PD-L1 expression population. For us, this was the update at ESMO [the European Society for Medical Oncology annual meeting] about the KEYNOTE-024 study. Can you talk to us about those results?

This is a dear study to my heart because it was the first time that we were able to use a checkpoint inhibitor as a single agent for patients with metastatic non–small cell lung cancer across different histologies, but in patients who have strongly positive or a high expression of PD-L1, more than 50%. We all are aware of the initial data with the trial having an overall survival advantage compared to platinum doublets, but here we have the 5-year data. We’re always talking about flattening the curve with COVID-19 [coronavirus disease 2019] and we’re doing the same thing with lung cancer. We are trying to flatten the tail of the curve with lung cancer, we want to see long-term survival. We start to wonder if we are curing some of our metastatic patients with immune checkpoint inhibitors or with immune therapy, and with this study, more than one-third of patients who received pembrolizumab in the first-line setting are still alive. These are wonderful data.

You mentioned, Suresh, the 3-year data also of Opdivo/Yervoy. We’re seeing more long-term survival. You look at numbers, 5-year survival with chemotherapy for those patients, 5 years ago you were seeing about 0%, or definitely less than 5%, and here today we’re talking about one-third of patients are still alive at 5 years. That’s wonderful.

What’s also nice to see with the trial, there were no added toxicities, nothing out of the ordinary. We’ve learned from different trials that immune-mediated reactions, if they are going to happen, they tend to happen early, usually in the first 6 months. They could happen at any time, but are more likely to happen in the first 6 months rather than later, which allows us to continue treatment beyond that period.

Absolutely. It is amazing that we have a third of the patients make it beyond 5 years. There was a discussion whether that represents a cure, and Roy Herbst, MD, PhD, the discussant for that abstract, certainly said those patients are cured. The important message there is you give 2 years of therapy and you stop. If the patient progresses, reintroducing the drug was associated with activity there, so that’s what I have done in my clinic for those patients. Steve and Christine, how do you approach that situation, when they progress on immunotherapy?

In similar fashion to you, I first rechallenge with immunotherapy, single agent, and do a short interval CT scan. I’ll take a peek at 6 weeks to make sure we’re headed in the right direction. But seeing another set of data about rates of efficacy at reintroduction, I’m comfortable. The majority of patients had either a response or at least disease stability, so this also helps me feel more comfortable in coaching patients in that decision at 2 years of whether to stop therapy, because I think that’s a huge point of patient anxiety in their treatment course. They’ve been told they have metastatic lung cancer, then they’re told they have this miraculous immunotherapy option. They get disease control for 2 years, and stopping the drug is such an anxiety-provoking episode for them. Now it’s nice to have some additional data to give them about, “OK, well if the cancer does come back, here’s what we would do, and here’s the likelihood of success in that case.”

: I agree with that. Fortunately for me, it hasn’t happened yet. All of my patients who have gotten to 2 years where I’ve stopped, it simply hasn’t recurred. I like having that assurance, but so far, so good. Ultimately it’s not going to be a one-size-fits-all. I think there probably are predeterminants of who can stop early and who can’t, of who needs specific strategies. All of my patients who have made it to 2 years and we’ve stopped electively, their disease remains in control now off therapy.

Chemotherapy and Immunotherapy in Non–Small Cell Lung Cancer

Another regimen that enters the landscape, I want to hear, Martin, your overview of the CheckMate 9LA trial.

This is a 2-part study with PD-L1 positive, overlapping with the pembrolizumab approval for greater than 1% in the PD-L1 negatives. I want to start with the positive ones. This was a 3-arm study looking at chemotherapy as the comparator arm and then nivolumab, and then nivolumab plus ipilimumab, and there were incremental benefits. Both were better than chemotherapy, but particularly interesting for me in the overlap of the intent-to-treat population, but also in the individual subgroups, a benefit of the combination immunotherapy [I/O] over the nivolumab arm alone, which suggests to me that single-agent PD-1 is not covering exactly the entirety of our needs in our population, and that there is utilization and benefit to our original immunotherapy agent, which is ipilimumab. It’s been around for 9 years and finally is making its way into non–small cell lung cancer. This is interesting. If you will allow for the discussion from earlier that our PD-1 inhibitors are similar and are performing similarly, there is at least an absolute numerical benefit to using the combination if you choose to have a chemotherapy-free regimen for the PD-L1 positives.

What was even more exciting to me is the question of can you utilize immunotherapy alone in a population that has previously only produced negative results? CheckMate-026 was the hallmark trial that defined that we need at least some positivity of PD-L1 for benefit. There’s a bimodal distribution between the PD-L1 highs and the PD-L1 negatives, but ipilimumab serves as an immune sensitizer that further enhances and stimulates toward a PD-1 response. From a relative benefit standpoint, for a PD-L1 negative patient who is interested in immunotherapy, this is a suitable regimen. I was disappointed when the FDA didn’t approve the PD-L1 negative population. This would be a great option for this patient population where we typically see little benefit of the addition of single-agent immunotherapy. It would be interesting.

If you allow for that comparison again, chemoimmunotherapy, ipilimumab/nivolumab looked good in comparison. Those were 2 arms, and it was originally scolded for being a complex trial, but in the end it turned out to be one that carried a number of valuable lessons, which is ipilimumab/nivolumab can be competitive in the PD-L1 positive space as a combination, it can be competitive against chemoimmunotherapy in the PD-L1 negative space. This is going to be hard to argue that this wouldn’t be a strong consideration for every patient in the PD-L1 continuum.

 When we talk about this combined PD-1 and CTLA-4 blockade approach, one of the concerns people typically have is about tolerability and safety of this combination. My view, looking at both the CheckMate-227 and 9LA data, is that the combinations are tolerable, and early recognition of adverse events related to autoimmune phenomena can be managed, and patients can go on to complete therapy. Martin, we saw some results of patient-reported outcomes for the CheckMate 9LA trial recently. Can you talk to us about that?

 CheckMate-227 has been, at least for non–small cell lung cancer, in a new dosing regimen. This is ipilimumab, 1 mg/kg body weight. We started at 3 mg/kg body weight in the melanoma setting, and in the adjuvant setting, even 10 mg/kg, and now at 1 mg/kg in non–small cell lung cancer. This is a dosing sweet spot, where you maintain an immune response, but your adverse effect profiles look substantially better, especially the GI [gastrointestinal] toxicities are now in the single digits in the grade 3/4 events, which is substantially lower than what we’ve seen with higher doses before. The 1 mg/kg body weight makes this a completely new setup.

This also relates to the fact that we must reconsider the reputation of this agent. This is an immune-sensitizer. This relates to the patient-reported outcomes, which are substantially better in terms of general quality of life for patients compared to chemotherapy, which is not surprising that if you prevent the cytotoxic adverse effects that are not liked by patients as well as clinical deterioration, that you would have patients that are overall doing better. My impression of this is that this is a well-tolerated regimen both in the combination setting with chemotherapy, where you have a limited addition of chemotherapy, as well as in single agent, and should be given a second look based on the new dosing.

: I would like to hear the opinions from some of our other participants on this panel about where they see a role for the I/O-I/O approach in non–small cell lung cancer in the frontline setting, given the data and the approvals we’ve seen with that. Steve, starting with you.

 It’s compelling. There are a lot of reasons why this is an appealing regimen. When I introduce this regimen or discuss it with my patients, while I will mention that this is a chemotherapy-free regimen, and there’s a lot of interest in that, I’m quick to follow that it’s not toxicity-free, and the grade 3 adverse event rate for CheckMate-227 is very similar to the chemotherapy arm. They’re just different toxicities. You won’t see necessarily the myelosuppression, the alopecia, the nausea that we may see with cytotoxic chemotherapy, but there are still a lot of other toxicities we need to be aware of. I need to find the right patients to use that.

There are a couple of reasons why I might go to a dual checkpoint blockade regimen up front. One would be, as Martin alluded to, that hazard ratio in the PD-L1 negative is compelling. You had mentioned earlier breaking things down by histology. If you look at PD-L1 negative squamous cell carcinoma, the outcomes with chemotherapy/I/O are not good. The CTLA-4 in PD-1, in that arena, arguably has better outcomes across trials. The hazard ratio was much better in the PD-L1 negative than we expected, and that was in part due to underperformance of the chemotherapy, but the chemotherapy arm did worse. The CTLA-4 in PD-1 arm really did the same for PD-L1 positive and negative with a very comparable survival. A very impressive duration of response, it’s compelling.

If all things are equal and you start with dual checkpoint blockade, and if it doesn’t work, if you’re not in that group where you get that long-term survival, now my second-line option is platinum doublet chemotherapy, which I would consider an upgrade over standard second-line docetaxel-based therapies.

 It adds a line of therapy for you when you use the I/O-I/O combination.

 Christine, I see you nodding. I take it that you share that viewpoint that Steve just mentioned.

I absolutely do. We also at the University of Chicago have had an investigator-initiated trial looking at combination ipilimumab/nivolumab and stereotactic body radiotherapy, or SBRT, and that was presented at ASCO [the American Society of Clinical Oncology annual meeting] this year in abstract form. Because of that, I’ve had a lot of comfort giving ipilimumab/nivolumab and managing the toxicities, and also have seen, exactly as Steve was just referring to, the benefits that that PD-L1 population with a 0% expression of PD-L1, seeing the benefits that they can obtain with this dual checkpoint blockade is quite impressive.

One of the other things that I love about this regimen is that time to chemotherapy gets pushed out. You get to save an amazing second-line option of carboplatin/pemetrexed, which has been the backbone of management for these patients for the prior 10 to 20 years. There are a lot of reasons to consider it as a frontline option.

Treatment Options in PD-L1 Positive Non–Small Cell Lung Cancer

We learned about another interesting study in the frontline setting that looked at a newer checkpoint inhibitor, cemiplimab. Steve, what did that study report and what were the implications?

You’re referring to the EMPOWER-Lung 1 study. This is looking at a PD-1 inhibitor, cemiplimab, vs standard chemotherapy as the frontline treatment for PD-L1 positive non–small cell lung cancer. A randomization to a PD-1 inhibitor vs chemotherapy may sound familiar to the KEYNOTE studies we’ve seen, like KEYNOTE-024. What we saw were in line with those results, that PD-1 blockade with cemiplimab improves survival compared to chemotherapy. We saw an improvement in PFS, [progression-free survival] but most importantly, an improvement in survival, the hazard ratio there in the PD-L1 high subset at 0.57. Cemiplimab had a better response rate, a better PFS, better overall survival, and a better toxicity profile, a winner across the board. This was despite having a high crossover rate, in fact, higher than what we saw in KEYNOTE-024. The crossover rate in this study was 74%, reflective of the time where the drug is more widely available. What we’re seeing with EMPOWER-Lung 1 reaffirms what we’ve seen with other PD-1 inhibitors in this space.

Another PD-1 inhibitor, it’s probably the third drug. We have PEMBRO [pembrolizumab] and ATEZO [atezolizumab] already approved in this space, and cemiplimab shows data. Having seen these data, Christine, do you notice any remarkable differences between one drug vs the other?

There aren’t dramatic differences that I’m appreciating between the 3 drugs and what we’ve seen in the trials. I have a tough time moving away from what has become standard as pembrolizumab in the frontline setting to seeing a large advantage of cemiplimab over what we already have.

One interesting feature about this study was there was a high number of patients with squamous cell carcinoma in the EMPOWER study, about 45%. Is there a difference in efficacy of I/O [immunotherapy] agents between squamous and nonsquamous histology? Martin and Firas, I would be curious to hear your thoughts on this.

Squamous cell carcinoma is a more immunotherapy-sensitive disease. The correlation with smoking, higher TMB [tumor mutational burden], and inflammatory markers is probably more stringent than for adenocarcinomas. They all exclude 

, but I’m not certain that they all have a full biomarker assessment. Are all PD-1s made the same? I know that we have questions about the PD-L1s. But I have to say that the data, despite the crossover rate, the hazard ratios are better, the absolute numbers are better, so I don’t want to put the topic to rest. I’ve looked back at other disease spaces where cemiplimab was performing quite well, but there may be differences between the individual PD-1 inhibitors. Cemiplimab was designed following nivolumab, it’s a second-generation PD-1, and there may be a difference. But squamous cell carcinoma is more set up to provide a meaningful benefit for immunotherapy, especially when you don’t properly select out all the oncogenic driver adenocarcinomas that may be involved in the study.

: I agree with Martin. If you practice where I practice in the tobacco belt, we see a lot of squamous cell carcinoma. There is a difference in responses between adenocarcinomas and the squamous cell carcinomas, regardless of the difference between PD-1 inhibitors. When you combine immune checkpoint inhibitors, PD-L1 status plays a lesser role, as well as histology plays a lesser role. We’ll be talking about some of the combination checkpoint inhibitors. When you’re doing more than 1 immune checkpoint inhibitor, you tend to overcome some of those differences based on PD-L1 status or histology. To me with this study, as Stephen alluded to, the crossover rate was high, and despite that, the survival advantage was still seen. It might argue that if you delayed initiation of immunotherapy, then patients may not do as well. You can add a checkpoint inhibitor down the road after they’re done with chemotherapy, but we’re seeing more evidence that the earlier you adopt immunotherapy, the better the outcomes are.

: I agree with that. If this trial had read out in 2017, it would have changed the world. In 2020, the world has already changed, and to me that lessens the impact. I would absolutely be in favor of an approval here. I don’t know if there’s a lot for me to change my practice. We tend to be creatures of habit, and inertia is a little difficult to overcome. I’ve got my work pathways set up, I’ve got prior authorization routes set up, my EMR [electronic medical record] is set up. I don’t know if I see data that are compelling enough to make me change my practice, but I’d absolutely support combinations building on this or using this as a control in a randomized trial.

Steve, when you said the world has changed in 2020, I thought that was the most understated statement of the year. But we will take that. One thing I found interesting, when you look at these trials, and going back to Christine’s point, what’s different from one another, what’s fascinating is these are trials done with different drugs, in different settings, albeit they’re all patients with first-line non–small cell lung cancer. What you end up seeing is the hazard ratios for PFS and survival almost come together across these drugs, even though we know that there are intrinsic differences from one drug to the other. The consistency of the results is something that gives us confidence that giving frontline therapy for a high PD-L1 patient with a single-agent checkpoint approach is very strong.

Let’s segue into talking about the I/O-I/O combination approach, where we have seen some results in the past few months that are interesting. I can talk about the CheckMate-227 trial, and then I’ll turn it to Martin to talk about the 9LA trial, which are both strategies that combine ipilimumab with nivolumab. I had the pleasure of presenting the CheckMate-227 data at ASCO [the American Society of Clinical Oncology 2020 annual meeting]. This was a trial that enrolled patients with frontline non–small cell lung cancer and randomized them to various treatments based on the PD-L1 expression level. Patients who had PD-L1 expression greater than 1% were randomized to chemotherapy alone vs NIVO [nivolumab] alone vs IPI [ipilimumab] plus nivolumab, and the PD-L1 less than 1% group, the randomization was to chemotherapy alone versus chemotherapy plus nivolumab versus ipilimumab plus nivolumab.

When we look at the comparison of ipilimumab plus nivolumab versus chemotherapy, there was a statistically significant improvement in overall survival, both in the PD-L1 high group and the PD-L1 negative group. The survival hazard ratio was even more favorable in the PD-L1 negative group at 0.62.

We presented longer-term follow-up data from the trial, and the number that comes to my mind is the 3-year survival rate for patients treated with ipilimumab and nivolumab was 33%, both in the PD-L1 high and low group, which is remarkable considering that not too long ago, we were not talking about 3-year survival rates for patients with lung cancer. The combination of ipilimumab and nivolumab is now an FDA-approved regimen. It’s approved for PD-L1 positive patients in the US; NCCN [National Comprehensive Cancer Network] guidelines recommend it for both positive and negative patients.

Durability of response with ipilimumab and nivolumab in that trial was almost 3- to 4-fold higher compared to chemotherapy. If patients had a response with chemotherapy, the duration of response was 6-and-a-half months. With ipilimumab and nivolumab, the response duration was almost 23 months. This suggests that patients who derive response with this combined immunotherapy approach have the ability to maintain that benefit for a lot longer.

Treatment Efficacy in Non–Small Cell Lung Cancer

, “Updates in Immunotherapy for Thoracic Malignancies.” I’m Dr Suresh Ramalingam from Winship Cancer Institute of Emory University in Atlanta, Georgia. Joining me today for this discussion are my colleagues, Dr Firas Badin from Baptist Health in Lexington, Kentucky; Dr Christine Bestvina from the University of Chicago in Illinois; Dr Martin Dietrich from the University of Central Florida in Orlando; and Dr Stephen Liu from Georgetown University School of Medicine in Washington, DC. Welcome everyone to this program.

Today we’re going to discuss a number of topics, including immunotherapy-based treatments in the up-front treatment of non–small cell lung cancer, recent developments in small cell lung cancer, and finally we will finish up with mesothelioma, a disease where there has not been a lot of progress in 2 decades.

We’ve had a number of recent meetings that have provided us with key insights into how we use this exciting group of drugs for treatment of our patients. We’ve had conclusion of ASCO 2020 [the American Society of Clinical Oncology annual meeting], the World Conference on Lung Cancer presidential virtual session, and more recently, the ESMO [European Society for Medical Oncology] 2020 conference. Now because these were all virtual conferences, we have not had a chance to talk face to face, but certainly all of us have been interacting through various platforms to bring to our clinics the exciting advances from these practices.

Let’s begin with frontline treatment of non–small cell lung cancer. This is the setting where we know molecular testing is done for every patient with non–small cell lung cancer, specifically the patients with nonsquamous non–small cell lung cancer. We want to ensure that every patient with a targetable driver mutation is treated with targeted therapy, and there are at least 7 driver mutations with the FDA-approved targeted approaches. Assuming that those patients are not going to be candidates for frontline chemotherapy, we’re going to focus this discussion on patients who don’t have a driver mutation to start with. With that in mind, I’m going to invite Dr Stephen Liu to talk about a study that was reported at ESMO recently, which was the combination of nivolumab with chemotherapy and bevacizumab.

This was a randomized trial for patients who didn’t have 

 alterations, and there are a lot of parallels between this and the approved IMpower150 regimen. That regimen builds on the backbone of carboplatin/paclitaxel/bevacizumab, showing that when you add atezolizumab, we improved outcomes. This was a similar approach, except instead of the PD-L1 antibody atezolizumab, this used the PD-1 inhibitor nivolumab. Patients all received standard carboplatin, paclitaxel, and bevacizumab, with a simple 1:1 randomization to concurrent nivolumab or placebo. After induction therapy, they then received maintenance bevacizumab and continued the nivolumab or placebo maintenance therapy.

We saw at ESMO 2020, that the addition of nivolumab to the carboplatin/paclitaxel/bevacizumab backbone did improve progression-free survival [PFS], a notable improvement there. We had a hazard ratio for PFS of 0.56, clearly improving outcomes. It’s immature yet for survival. We did not yet see a survival benefit—a hazard ratio of 0.85 with a pretty wide confidence interval—but showing some signal that when you add nivolumab to the carboplatin/paclitaxel/bevacizumab backbone, we do improve outcomes.

This whole concept of combining antiangiogenic or anti-VEGF targeted therapies with immune checkpoint inhibition is something that’s widely pursued not just in lung cancer, but in many other cancers as well. We saw another abstract in the same lines, which was the combination of atezolizumab and bevacizumab. What did you learn from that study?

This one fell under the radar, a smaller study presented by Takashi Seto, MD, PhD, but an impressive end point. This was a single-arm phase 2 study, a modest trial, only 40 patients. This trial enrolled patients with PD-L1 high expression non–small cell lung cancer. No randomization, single arm, all patients receiving atezolizumab, the standard PD-L1 inhibitor, but given concurrent bevacizumab. What we saw in these early end points was a high response rate, over 60%. For reference, when we look at PD-1 inhibitors alone in some of our best studies for PD-L1 high, we’re looking at response rates in the upper 30% range. Seeing a response rate of 64% I thought was quite compelling. The PFS, again, immature, a modest study, but we have a 1-year PFS rate of 55%, which is quite impressive. It’s unfortunate this was a single-arm study. I would have loved to have seen a second arm with atezolizumab monotherapy to really see how this stands up. But those high response rates suggest that there may be synergy between these 2 pathways.

Interesting. Christine, when you look at these 2 studies and think about what these data mean to our current landscape in lung cancer, how do you see this? Is this exciting for you? Is this something that tells you there’s more work to be done?

:I do think there’s more work to be done as Stephen alluded to, particularly the trialed atezolizumab and bevacizumab was not a randomized trial. I would like to see that confirmed with randomization against atezolizumab, which as the authors pointed out, is something that’s planned. But the response rate demonstrated in that study is extremely exciting to me. When we talk about triplet therapy, chemotherapy plus immunotherapy in the patients with high PD-L1 expression, one of the arguments that providers will make for using the triplet over monotherapy is if there’s a high response rate that’s needed, you may choose the triplet. But with seeing this high response rate with the addition of bevacizumab to atezolizumab, I wonder if we’ll be able to use angiogenesis inhibitors in a similar fashion, while still sparing more traditional chemotherapy agents.

When you talk about 4-drug combinations, Martin, I would be curious to hear about your thoughts. What can you tell us about the safety profile of these regimens?

There are several considerations. I still have difficulties finding the role for bevacizumab in the first-line setting, and I personally haven’t used it much. There’s a cumulative toxicity profile, they stay within their lane in terms of additional adverse effects, and I’m concerned about cost. But I would like to understand in the first-line setting what’s the contribution to efficacy. The trial atezolizumab plus bevacizumab was interesting, and there’s a lot of thought even from other disease spaces, that the antiangiogenesis may be contributing to an anti–PD-1 effect. But this was a small study with a very PD-L1 enriched population. There was a tremendous contribution from patients who had 75%-plus PD-L1 expression. It’s hard to interpret that information in the context. I would consider the thought that Stephen brought up, that this is a study that is thought-provoking but it needs confirmation in a more controlled fashion. I don’t have a big role for either the nivolumab- or the atezolizumab-based bevacizumab trials at this point.

Thank you. Firas, anything to add there with regard to this strategy from your standpoint?

I agree with my colleagues. They are interesting data. In our thoracic clinic, we’re moving away from single-agent PD-1 inhibitors into combination treatments, but we haven’t found the best dancing partner yet with PD-1 inhibitors. Is it anti-VEGF treatment? Is it anti–CTLA-4? Is it chemotherapy? I think it’s exciting. Nobody knows the data yet. One of the questions that came to my mind is, is there a particular subgroup of patients that might need a quadruplet, because as Martin reviewed, it’s not only the adverse effects, but also the financial toxicity about giving 4 drugs all at once for those patients. We know from previous studies that patients with liver metastases might do better with quadruplet regimens, with anti-VEGF plus anti–PD-1. We need to wait and see and have more data. But it’s exciting to see these kinds of combinations.

Christine Bestvina, MD, University of Chicago

Articles

Understanding Immune Response in Non–Small Cell Lung Cancer

Chemotherapy and Immunotherapy in Non–Small Cell Lung Cancer

Treatment Options in PD-L1 Positive Non–Small Cell Lung Cancer

Treatment Efficacy in Non–Small Cell Lung Cancer

Frontline Treatment in Non–Small Cell Lung Cancer