This was a fantastic discussion, very robust. I am going to transition [to a new topic] and ask Chaitali to give us her thoughts. We mentioned how far back the combination therapy goes in stage III disease—first induction and then concurrent, and then issues around consolidation. We established that concurrent chemoradiotherapy with platinum-based therapy was the standard of care but then we had a bit of a lull where nothing was happening until once again, moving immunotherapy into earlier stages of disease, in this case, stage III with a specific trial and then Chaitali at ASCO [the 2021 American Society of Clinical Oncology Annual Meeting], recently we saw the 5-year survival update, so give us your perspective on the PACIFIC trial. How that’s changed things and just your thoughts about it.

: To me, the PACIFIC trial was practice changing when it was approved in February 2018. Since then, the standard of care has been if [a patients] has a good performance status, irrespective of PD-1 status or histology, after completing concurrent chemotherapy radiation therapy (RT), and go on to durvalumab maintenance for about a year. We knew the data was initially presented for 2 years with excellent hazard ratios for OS [overall survival] and PFS [progression-free survival]. At ASCO, we saw an update for the 5-year OS and PFS and the numbers look pretty good. The hazard ratio was .72 for OS and PFS was even better, it was .55. If you look at what was the percentage of people with OS as 60, those numbers were very high and impressive with 42% of people still holding good at 5-year mark. There are no new safety concerns, so I think this effect remains the standard of care and the 5-year overall survival data is impressive and tells us how well people do in stage III unresectable situation.

I was taught that when you have a positive phase 3 trial, you do the same thing in clinical practice that the trial did. One of the issues that always comes up is how important it is to receive the full year of durvalumab because if we look at the current situation, there are many patients who are dropping off and not getting the full year. Sometimes it’s for a good reason and sometimes it’s for not such a good reason. Your thoughts on the importance of the 1-year data, and do you share my opinion that it’s important in practice to do therapies how they showed benefit on the phase 3 trial?

: As much as possible I try to follow how it was done in the clinical trial, but the real-world experience is that once you are 6, 8, 9 months into it, it can get hot, just like how we saw with the mutation when your initial post radiation. It’s a similar situation that we all saw from the data from PACIFIC, that if you got it within the 2 weeks of radiation that those are the patients who did the best but how practical is that? How often do you actually get to get people started and completed in 2 weeks? It’s not hypothetical, so I do agree that as much as possible I would like to get patients to a 1-year mark. In practicality, how often does that happen? I would say I easily have a dropout rate something from 10% to 25% that don't get to the year.

 Does PDL-1 status mean anything to you in the stage III setting?

: If you are going to do this then you know it was. It was in the study it was irrespective of PD-1. They did do an exploratory subset analysis in which there was not much benefit if your PD-1 was less than 1% but having said that, there's caveats to that it was a subset analysis, so right now, irrespective of PD-1, I would offer them the maintenance durvalumab.

And of course, the indication in the US is agnostic to PDL-1 status but with our European colleagues, they have a PDL-1 positive in 25%, correct? Correct me if I am wrong.

So, put a little bit of handcuffs in—that’s based on a unplanned, retrospective analysis that wasn’t protected by stratification. We know that not every patient had PDL-1 measure in the trial.

: Also important, 28% to 30% of people never had PD-1 measured, so you don't know what those numbers would look like if they were.

Videos

Chaitali Nangia, MD, of the Patty and George Hoag Cancer Center, reacts to the significance of the 5-year overall survival data for the PACIFIC trial evaluating the use of durvalumab after chemoradiation as treatment for inoperable stage III non–small cell lung cancer. 

Durvalumab for Unresectable Stage III NSCLC 

 Before we move on, I want to go back to Neal. Essentially what we’ve done here is take what we used to do in the second line, which was immuno-monotherapy, and graft it on to standard first-line regimens. We combined first- and second-line standards in this setting. We’ve had a nice discussion about how successful that strategy was, but what do you do afterward? What do you do now in the second-line setting for these patients? The vast majority of patients are getting [chemotherapy/immunotherapy] combinations in the first-line setting. What do you do afterward?

 Thanks, Mark. Of course we’re trying to use clinical trials to figure out how to overcome immune resistance. When that’s not possible, for patients who have seen single-agent IO [immunotherapy], we’re going back to a platinum doublet, where patients who received chemotherapy/IO initially, we’re falling back on docetaxel, sadly. But I think that some of the progress we’ve made in 

 tumors often do respond to immunotherapy, but they often eventually progress. The advent of targeted therapy for 

 G12C gives us another option. Hopefully we’ll have more molecular therapies for some of these immune responsive tumors.

 It’s very nice after many years of not having a 

strategy, at least for a portion of the patients with 

 [mutations], we do have a new option, which the recent data updated at ASCO [American Society of Clinical Oncology annual meeting] on the CodeBreaK 100 trial looked pretty promising. The only thing I would add is don’t forget about anti-VEGF in the second-line setting. I do use a lot of ramucirumab with docetaxel in this particular setting.

Second-line treatment recommendations for patients with stage IV non–small cell lung cancer who experience disease progression after immunotherapy. 

Second-Line Therapy for Stage IV NSCLC 

 I want to transition to stage III here in a few minutes, but I wanted to go back and ask Roy. In our careers, we haven’t seen a step up like we’ve seen with the PD-1/PD-L1 inhibitors and to a certain extent the CTLA-4 data that we’ve just been talking about. It’s been such a change in the standard of care and the potential outlook, even being able to talk about 5-year survivors, which we never did historically. What’s your prediction for what’s next, Roy? Do you have a favorite target or a strategy that’s coming up, or what are your thoughts with regard to the next big breakthrough, if you will?

 As a great philosopher said, Mark, it’s hard to make predictions, especially about the future. But with that said, it is amazing, when at least 3 of us started this, 1 year of survival was something with a 20%, 30% [chance], and if we had 2 years, “wow.” It is nice to see what we’re having. We have proof of concept that immunotherapy works, and I would even dare say we can cure people with immunotherapy. The problem is clearly it’s only 15%, 20%. Every 5-year study that’s coming out, there’s a bimodal curve and there’s a tail, but it’s 15%, 20%, at least in the second line. In the front line it may be 30%, 35%, or more. So the question is, can we do better? I think there are 2 things to think about: primary resistance and acquired resistance. There are clearly some tumor cells that are not hot, they’re not inflamed, in fact I would say that’s more than half of them. If they’re not in that environment, no matter how much you block PD-1 or PD-L1, it’s not going to matter because you don’t have any T cells there.

I would predict we’re going to see therapies in the next few years to try to increase that, whether it be TIL [tumor-inflitrating lymphocytes], take the cells out, ex-vivo, and put them back, or agents that might alter the microenvironment, whether it be a molecule that has a CD3 on the other end, a T-cell CD3, or some sort of interleukin, IL-10, IL-15 has been looked at, IL-18. I think we’re going to see more and more of that. Then I think we also have to realize that some of these tumors might not have PD-1/PD-L1. It might not be a universal mechanism, or there might be other checkpoints, TIGIT. There’s another drug we’re studying here at Yale [Cancer Center], Siglec-15. That has to be looked at as well. But then of course it’s the whole resistance. I think we’ve all been happy to see patients who do well, and then they are resistant to immunotherapy. What you do then? I’m going to give you a thought for the future. We’ve spent 20-plus years, it was 1997 when we did the first EGFR studies, and we’re still figuring out resistance. We have to do that now with immunotherapy, and we have to personalize it, and that’s not what we’re doing, Mark. We’re just putting combinations together. So I would predict 10 years from now, or maybe even 5, we’re getting a biopsy of the tumor and the microenvironment, and we’re understanding what are the inhibitory factors, what are the inhibitory T cells, what’s missing. Then we’re putting a cocktail together in a personalized way. But the proof of concept is there, and it’s amazing to me. I would have never thought 20 years ago that we would have seen this much progress.

 I would completely agree with you. It has been amazing, and I agree with you, I think we are curing some patients. You and I stared at a time when our counsel would be, “You have a treatable disease but not a curable disease.” I remember when everyone thought it was a big deal when Alan Sandler, [MD,] got up at ASCO [American Society of Clinical Oncology annual meeting], and there was a 1-year survival rate from the ECOG 4599 trial that exceeded 50%. We thought that was a milestone at that time, and that was just 15, 16 years ago, so we have come a long way. But also, to your point, we have a long way to go in understanding all of this and what do we do next.

Future projections regarding the role of novel I-O therapies as frontline treatment for patients with metastatic non–small cell lung cancer. 

Frontline Treatment of Stage IV NSCLC: What’s Next? 

 Let me ask Sandip, one of the things that we haven’t talked a lot about is the [adverse] effects of immunotherapy [IO], although many of us feel that as a general class of drugs, these IO agents are generally pretty well tolerated in this setting. What’s the message for our audience about immune-related adverse events [IRAEs] and in their recognition and management? Do you have any golden pearls you can share with us, Sandip?

 I think these therapies are remarkably effective immunotherapies that we’ve been discussing, but they don’t come without toxicity, and the double-edged sword of the immune response is friendly fire against the healthy tissue and you can get autoimmune toxicities. In lung cancer the toxicity that’s probably the most vexing in our practices is pneumonitis. Immune-mediated pneumonitis, it often mimics various other entities that our patients with lung cancer are particularly prone to, whether it’s a COPD [chronic obstructive pulmonary disease] flare; an infection, especially most recently with COVID-19; pulmonary embolism; disease progression. So these patients invariably get a CT chest shot to try to better understand, and the problem with a lot of immune-related adverse events is there is not a specific blood test that tells you that this is an immune-related phenomena. You’ll get a liver function test, or you’ll get a CT that shows that something is wrong. You’ll try to rule everything else out, but the problem with almost all our immune-related adverse events is that they are diagnoses of exclusion. With that, you’re ruling out other things, and you’re empirically beginning therapy.

My humble 2 pieces of advice related to IRAEs are, No. 1, have a high index of suspicion. If the patient is more short of breath than usual or is starting to have diarrhea, rule out 

, but be prepared to treat this as immune colitis. The second is start the steroids early, especially once you’ve ruled out infection, which you can often do quite rapidly with a procalcitonin, for example. Or treat concurrently with the steroids and antibiotics for a really severe case of let’s say colitis while you’re waiting for the results to come back and waiting for specialty expertise because I’ve never seen anyone regret giving steroids too soon. I’ve only seen people regret not starting steroids soon enough. Then finally, there was a nice abstract at ASCO [American Society of Clinical Oncology annual meeting]…that looked at what to do if steroids don’t work. I think the message there is if you have immune-related colitis, meaning you’ve ruled out 

, which unfortunately can be common in this patient population, infliximab or another anti-TNF [tumor necrosis factor] can be efficacious. But above and beyond that, our options start to become increasingly potentially problematic for patients, including drugs that block T cells, such as mycophenolate. You may use drugs that block movement of immune cells into normal tissues, like vedolizumab, and I think it gets really complicated at that point. But the one point I should make about steroids is this isn’t your 5-day COPD burst. If you’re treating an immune-related adverse event, you want to have at least a 3-week steroid taper, and probably closer to a month, maybe slightly over a month because if you don’t, the likelihood of this immune-related adverse event coming back, whatever it is, is quite high.

 We presented some data at the recent ASCO that having a low-grade immune adverse event might be a good thing in this setting. Of course, it was also true on the control arm of the analysis of IMpower150, 130, 132, so any thoughts on this? Because I do agree with your comment that understanding the spectrum of toxicity can affect almost any organ and to recognize that, hopefully when it’s grade 1 or 2, be early in treating it and preventing it from being 3 or 4. And that was kind of the message from that analysis.

 Absolutely, and it’s a great summary of a very complicated data set. I think the broad theme across the studies you mentioned, as well as other, especially some retrospective analyses, is that the development of an immune-related adverse event means the immunotherapy is doing something to the immune response. And as you’re elevating the temperature, so to speak, of the immune system, hopefully more of that heat is directed at the cancer, but some can be directed at friendly organs as well, and so there is this correlation that I think has been seen across multiple data bases. I’m not a 100% there, but I think most of us have seen it, that most of those patients who develop IRAEs often are the ones who get benefit. That does not mean you have to get one of these autoimmune toxicities to benefit from immunotherapy, but what it does mean is that for those patients who do develop an immune-related adverse event, the question about rechallenge often involves the severity of the toxicity, how life threatening it was, how salvageable it was with steroids that are immunosuppressive. But if you also know that that patient who had that immune toxicity has a pretty good prognosis long term, it means you can take a break. You can watch with serial CT scans and see if the immune system has taken care of the cancer as well as it caused toxicity on normal epithelium. I think that’s probably where that data set is most relevant to most people’s clinical practice.

Recommendations for managing immune-related adverse events associated with checkpoint inhibitors used to treat stage IV non–small cell lung cancer.

Immune-Related Adverse Events in NSCLC

 Martin, let’s get your thoughts on PD-L1. We did see, from the EMPOWER-Lung 1 trial, getting a bit more granular with regard to the PD-L1 group greater than 50%, dividing it up into, I think, 50% to 60%, 60% to 80%, and 80% and above, and your perspective on that. But then we’ve heard a lot of buzz about TMB [tumor mutational burden]. There were a couple of presentations at ASCO [American Society of Clinical Oncology annual meeting], some co-mutations like 

, and these sorts of things. Let’s start with PD-L1, what are your thoughts there? Then delve into the TMB and the other co-mutation aspects of this.

 I agree. I think the granularity of PD-L1 is going to really shape the selection of therapy in the first-line setting. Greater than 50% is no longer a binary call. I think we’ve seen an incremental benefit in 10% increases going from 50% closer to 100%. I think this also helps in the selection of patients…deescalate to a single-agent PD-1. We have seen, obviously, the combinatorial strategies providing benefit across the PD-L1 spectrum. We’ve had an 11% increase in overall survival at 4 years in combination with ipilimumab. I’ve seen a similar signal with the addition of TIGIT. We are certainly moving forward in getting PD-L1 refined in terms of its application. I would say that low PD-L1, greater than 1% but below 50%, is one where I find the KEYNOTE-042 regimen unsuitable, simply because it wasn’t positive as an overall trial. Addition of chemotherapy or a second immunotherapy agent here is helpful. My take on it is PD-L1 really decides about the intensity of immunotherapy. The closer to 100%, the more likely I would be willing to give a single-agent PD-1. But when in doubt, at 50%, even 60%, I would still consider a combinatorial strategy the right way moving forward.

 And your thoughts on TMB? The first question is, do we know how to measure it?

 It is very difficult, and especially now that liquid biopsy has started to offer TMB assessments as well. It tells you how all over the place it is. If you have synchronous measurements, you can see that there is a high degree of discrepancy with liquid biopsy, typically overestimating tumor mutational burden. I was originally very excited about tumor mutational burden when the first analysis of CheckMate 227 was released in 2018. It has since probably lost a little of its impact. We’ve seen data at AACR [the American Association for Cancer Research annual meeting] this year, showing positive predictive impact in CheckMate 9LA for TMB. I like to think of this as an add-on, so if I have a TMB-high, PD-L1–high level, I have higher expectations, but I don’t use [it] beyond PD-L1 as a true frequent selection marker. The secondary mutations, those are exploratory analyses, and I don’t find them quite ready for prime time for selection against immunotherapy, even though they’ve been looked at. I think it was very interesting to see that in combination with some of the KRAS inhibitors that they would…respond well, but I can’t factor in 

 mutations just yet as a marker that selects against immunotherapy in any way. I think it’s interesting to see, my expectations are certainly lower but I wouldn’t change my overall take based on the secondary mutations unless we have more prospective data on it.

 I would agree. They are interesting data, helping to identify patients who may not be benefiting as much from immunotherapy approaches and maybe would benefit more from different approaches. What those approaches are, I don’t think we know.

Martin Dietrich, MD, PhD, shares his thoughts on PD-L1 status, tumor mutational burden, and other predictive and prognostic biomarkers when making treatment decisions for patients with stage IV non–small cell lung cancer. 

Frontline Treatment of Stage IV NSCLC: Biomarkers 

I want to bring Chaitali into the discussion. We’ve just had excellent reviews of all these trials that we now have FDA approvals for, but it gets complicated a little in terms of choice of therapy. I just want to get your perspective and then put it in context. We’ve kind of alluded to patients who have a high symptom burden, high tumor burden versus those—we see so many patients diagnosed with what I would refer to as low-volume, relatively asymptomatic stage IV disease. Do you view them differently in your choices in terms of how you incorporate all of this data?

 Absolutely. When we are in this zone when you have chemotherapy, chemotherapy–I/O [immuno-oncology] combinations as options for you, and I/O–I/O combinations, patient and disease factors completely weigh in. If you have, for example, an individual who isn’t chemotherapy amenable or who has the chemosensitivity, then that’s the one that I might consider a chemotherapy-free I/O–I/O combination. At the same time, if you have bulky disease, then that’s the one—symptomatic, painful. As Sandip mentioned, I would definitely go with chemotherapy–I/O combination. At the same time, if I have very low-volume disease, oligometastatic disease, PD-L1 is high, that’s the one I’ll go with a single-agent I/O. Which single-agent I/O? We have three, and I’m not partial to one. Any of those would be a good combination. Then you start teasing out information. As we just saw from ASCO [American Society of Clinical Oncology Annual Meeting] 2021, if you have squamous histology, and PD-L1 is low or less than one percent, then that’s when you want to consider the ipilimumab-nivolumab combination. If you have somebody with stable CNS [central nervous system] metastasis, at least based on evidence, you can consider cemiplimab if PD-L1 is higher than 50%. You start looking into the data to help you. But ultimately, the patient factors and disease factors both weigh in.

You raise a good point, and I want to transition to talking a little more about PD-L1 in a bit more granularity. But let me ask Sandip and maybe Roy [Herbst]. We saw a presentation by the FDA group at ASCO this year looking at this group of one to 49% and looking at the use of I/O monotherapy. Of course, we do have an indication for pembrolizumab in that subset of patients versus chemotherapy–I/O. This was an analysis of phase 3 trials that had been done. The take-home message that I took from that was that chemotherapy–I/O is the standard of care in that population. I’ve been reluctant to use monotherapy in this population, but I want to get your perspective, Sandip, on that analysis.

 I completely agree. If we look back to the KEYNOTE-042 presentation, which was a couple of ASCOs ago as a plenary, we saw the 1% to 49% data looked quite similar to the FDA analysis. To me, it was practice reinforcing that for a patient within the PD-L1 strata of 1% to 49%, that a combination approach—whether that’s chemotherapy plus PD-1 inhibitor, chemotherapy–PD-1 inhibitor–antiangiogenic inhibitor, CTLA4 inhibitor—makes sense. In terms of the active clinical trials in this area, TIGIT-based combinations with PD-1 are very reasonable. But the bottom line that I took away from KEYNOTE-042 and the most recent presentation from the FDA is that patients with PD-L1 1% to 49% are best served by a combinatorial strategy. What that best combination is in 2022 and 2023, we still probably won’t know. But the idea is that it should be some combination, so try to maximize the benefit for these patients.

I agree. In my opinion, that combination, at least until we prove otherwise, should be chemotherapy in that population too.

Considerations for selecting frontline therapy for metastatic non–small cell lung cancer based on various new data and treatment approvals in the field of immunotherapy. 

Frontline Treatment Decisions for Stage IV NSCLC 

As you point out, we’ve had a flurry of wealth with all of these positive trials, most of which have led to FDA approval, so these options are out there for patients. One of the options, I’m going to ask Neal Ready to comment on his perspective with regard to what we call the chemotherapy-free combination of the nivolumab and ipilimumab, that stems out of the CheckMate 227 trial. We recently had an update with further follow-up at ASCO [American Society of Clinical Oncology Annual Meeting] 2021. Neal, give us your thoughts on how this fits into the landscape of everything Sandip {Patel] just reviewed very nicely for us.

 Thank you, Mark. As Roy pointed out, it’s very impressive that we’re seeing with I/O [immuno-oncology] therapy really durable responses that we didn’t see with chemotherapy before. We’re seeing five-year survival, and the thought has been that, with combination immunotherapy, that we would get a higher number of durable responses and really push up our four- and five-year survival rate, which is our goal. Nivolumab and ipilimumab, with the addition of a CTLA4 inhibitor, is a logical choice. In a large, randomized phase 3 trial, nivolumab and ipilimumab gave superior survival compared with chemotherapy alone in the PD-L1–positive group. That has been a durable benefit.

At the most recent ASCO 2021, the four-year survival rate, or overall survival, duration of responses, progression-free survival, have all held up to show a promising result for combo immunotherapy. This was planned for the PD-L1–positive group based on CheckMate 012, which was a big phase 1 trial. In that trial, we were looking to get the right dose. The ipilimumab, 1 mg/kg every 6 weeks, turned out to be a feasible, safe dose in lung cancer. That trial was phase 1, so we were trying to look at other populations, groups of patients who might not benefit from single-agent I/O.

In the cohort of 77 patients that we used to identify biomarkers for planning CheckMate 227, there were 12 

-mutated patients. They undoubtedly were in the PD-L1 zero and low group. It led to CheckMate 227 being planned for PD-L1–positive because it really looked like nivolumab-ipilimumab was better on the high PD-L1. But when we look at the big phase 2 trials, CheckMate 568 and CheckMate 227, where 

 were excluded, it really looks like it was in the low–PD-L1 group, either PD-L1 less than 1%, or PD-L1 low, where we saw the biggest benefit. As my partner says, if your PD-L1 is low, you need tumor-infiltrating lymphocytes. But if your PD-L1 is really high, you have them. You don’t need CTLA4 to get them. In a large trial that Merck ran of pembrolizumab with or without ipilimumab, there was no benefit. In the high–PD-L1 group, over 50%, you probably don’t need the ipilimumab. PD-1 checkpoint antibody is enough. But from my perspective, in patients who look like they should have antigens that the immune system could identify—PD-L1 less than 1%, PD-L1–positive, and especially if tumor mutational burden is high—those are the patients who can benefit from combined I/O therapy. That’s how I look at things.

Neal Edward Ready, MD, PhD, describes which patients with stage IV non–small cell lung cancer may benefit from the combination of ipilimumab and nivolumab in the frontline treatment setting. 

Frontline Treatment of Stage IV NSCLC: IPI/NIVO 

 Before I transition to Sandip, as we move from monotherapy to the combinations, give me a sense, Roy: In your practice, do all patients over 50% get monotherapy? Or are there those patients you think are better served with chemotherapy–I/O [immuno-oncology]? Do you see a difference in these—say, 50% to 60% PD-L1–positive versus 90% to 100% PD-L1–positive populations?

 That’s a loaded question. Even in the 50%, the response rate is, what, 43%? Even a high PD-L1, over 50%, have a primary resistance. Yes, we have to start thinking about combinations. In the clinical setting, without a protocol, I’m at a place where the protocols are flying back and forth every minute. But without a protocol, I still try to give the immunotherapy alone to the 50% of patients unless they’re having significant symptoms, and I’m worried they need an immediate response. Then I might add the chemotherapy to the immunotherapy. I will then add chemotherapy later on if they’re having a less-than-robust response. But of course, there are so many protocols looking at what to do in that 50% group, the No. 1 being the SKYSCRAPER-01 trial, looking at TIGIT, another checkpoint [inhibitor], perhaps you can do that. There’s been some thought that antiangiogenic, or VEGF inhibitors, might have an effect on the immune microenvironment, so there are some trials. But in most cases, if we can avoid chemotherapy, especially during this last year with COVID-19, it’s nice to be able to avoid making someone neutropenic.

Sandip, give us your thoughts on the chemo–I/O. Obviously, if you’re committed to using chemo–I/O, perhaps the PD-L1 status is a little less meaningful. But what are your thoughts on the major trials—KEYNOTE-189, KEYNOTE-407, IMpower—and how they’ve changed the landscape.

 To follow up on Dr. Herbst’s excellent discussion, previously we were talking about single-agent I/O for patients with PD-L1 high, typically greater than 50%. It depends on the assay for a given agent. The question is, for patients who are PD-L1 low, or even PD-L1 high, greater than 50% but have highly symptomatic burden, they’re admitted to the hospital with very large, painful lesions. They need an immediate response. The question then becomes, could you combine chemotherapy with the newest immunotherapy? Though we’re a decade into this immunotherapy era, we try to maximize response. One point about the Kaplan-Meier survival curves we’ve seen for I/O versus even chemotherapy alone is that in a biomarker-selected population for PD-L1 greater than 50%, some patients don’t benefit. As mentioned earlier, it’s hard to salvage, especially if they have highly symptomatic tumors.

The thought behind combining chemotherapy with immunotherapy is the basis of multiple frontline studies that allow for utilization of immunotherapy independent of PD-L1 status. As Dr. Herbst mentioned, there are opportunities for I/O monotherapy for patients who are PD-L1 high. The first two involve pembrolizumab: carboplatin-pemetrexed-pembrolizumab, for the KEYNOTE-189 status for patients with nonsquamous. It’s important to mention that these are patients without 

 mutations across all these studies. For the majority of patients with targeted mutations, such as 

, who should undergo appropriate molecular sequencing, their best treatment is targeted therapy up front. In these studies, 

 were tested for. If positive, the patients went on appropriate targeted therapy. But if negative, they were allowed to be randomized to chemotherapy versus chemotherapy-immunotherapy. The point to make about KEYNOTE-189, which is in non-squamous histology, is that we have a tail of the survival curve. It’s really impressive to see responses, even in PD-L1–negative patients. For PD-L1 zero, there’s an overall survival advantage when you combine chemotherapy with immunotherapy in this setting.

For the majority of the frontline studies, the duration of therapy was two years. There is some controversy around what to do at the two-year mark, and this is an ongoing question. That’s part of the personalization of medicine, is that discussion of what to do. But the idea that we’re seeing durable responses, both in squamous, which is the KEYNOTE-407 study with carboplatin-taxane, or in the KEYNOTE-189 study, these really represent the backbone of the treatment of non–small cell lung cancer. We’ve also seen benefit from the integration of immunotherapy with angiogenesis inhibitors, so the IMpower150 study looked at carboplatin, paclitaxel, bevacizumab, and atezolizumab in combination. One aspect of this study is that it allowed patients who had prior EGFR-ALK treatment. Essentially, they finished their targeted therapy but were going on to this chemotherapy-immunotherapy combination.

Though the analysis was complicated, there were some interesting analyses that suggested benefit in these populations. This has represented a backbone, especially for driver mutation patients, and there were some provocative data related to liver and CNS [central nervous system] metastasis as well. The other study in the frontline space that integrates immunotherapy—Dr Herbst was mentioning some of the TIGIT-based combinations—is our original immune checkpoint blockade agent, ipilimumab, an anti-CTLA4 agent, in combination with a PD-1 agent, nivolumab, in combination with chemotherapy. This is the CheckMate 9LA study, in which two cycles of chemotherapy are given, and patients go on to immunotherapy alone.

We saw some updated results recently at ASCO [American Society of Clinical Oncology Annual Meeting] suggesting a durability of a benefit, and some provocative data in certain cohorts. For example, the PD-L1 negative cohort, the less than 1%, squamous histology, and then patients with CNS metastasis, seemed to confirm preferential benefit. But this is comparison with chemotherapy, not chemotherapy-immunotherapy.

Finally, we’ve seen some really tantalizing data with another CTLA4 agent, tremelimumab, which hasn’t found a home in combination with durvalumab and chemotherapy. This was teased as a press release about a month ago for which we don’t yet have the full data. The frontline space has become complicated, in addition to personalized medicine, above and beyond just looking for the appropriate driver mutations and making sure patients get appropriate driver mutation-directed therapy if they lack those driver mutations. Their optimal immunotherapeutic strategy will likely become less clear before it becomes more clear as we’re starting to integrate some of these novel therapeutics into the frontline space.

What’s the best role of CTLA4-based combinations, and how can we best understand the role of chemotherapy, especially if a patient is PD-L1 high? These are some of the outstanding questions in the field. We’re seeing five-year survival data, meaning three years after their last dose with a durability of response, so it’s a really phenomenal time for patients, and we just want to raise that bar further.

Something we never saw with chemotherapy by itself.

Drs Roy S. Herbst and Sandip P. Patel comment on patient candidacy for immunotherapy as monotherapy to treat PD-L1+ metastatic non–small cell lung cancer versus chemoimmunotherapy.

Frontline I-O Monotherapy Versus Combination Therapy for PD-L1+ Stage IV NSCLC

®, “Updates in Immunotherapy for Non–Small Cell Lung Cancer.” I’m Dr Mark Socinski from the AdventHealth Cancer Institute in Orlando, Florida. Joining me in this discussion are five of my colleagues. I’d like them to introduce themselves. I’ll start with Martin.

 Hello, I’m Dr. Martin Dietrich. I’m a medical oncologist at Florida Cancer Specialists & Research Institute at the University of Central Florida Cancer Center in Lake Nona.

 Hi, I’m Dr. Roy Herbst. I’m the chief of medical oncology and an associate director of translational research at the Yale Cancer Center in New Haven, Connecticut.

 Hi. I’m Chaitali Nangia, medical oncologist at Hoag Cancer Center in Newport Beach, California.

 Hi, everyone. I’m Sandip Patel. I’m a medical oncologist and an associate professor at the University of California San Diego where I lead our experimental therapeutics program, as well as our clinical trials office.

 Hi, I’m Neal Ready, a professor of medicine at Duke Medical Center in Durham, North Carolina.

Thank you all for joining today. Today, we’re going to be discussing a number of topics, including initial immunotherapy-based regimens for metastatic non–small cell lung cancer, consolidation immunotherapy in stage III disease, and perioperative regimens with immunotherapy in resectable non–small cell lung cancer. We’ll discuss the latest research in the field and the impact of recent clinical trials on making decisions around treatment selection. Let’s get started on our first topic.

I’m going to ask Roy to give us an overview of frontline immuno-oncology. Back in October 2016, KEYNOTE-024 led to the approval of pembrolizumab in the high expressers, greater than 50% PD-L1 positive. Since then, we’ve had the addition in this space of both atezolizumab and cemiplimab. I wanted Roy to give us his thoughts and overview of this population and these studies.

 Thanks, Mark. It’s great to see that results are being reproduced with other agents, and immunotherapy. I guess it’s more than 10 or 11 years that we’ve been using these agents in lung cancer, but really only in the last three or four, or maybe five, that we’ve had approved agents. Immunotherapy is a standard in the frontline therapy. We already had, going into this year, pembrolizumab from KEYNOTE-024 approved in frontline versus chemotherapy, and then atezolizumab beating chemotherapy, that being a PD-L1 inhibitor—pembrolizumab is a PD-1 inhibitor—in the IMpower110.

Now we have the EMPOWER-Lung 1 trial, which is looking at cemiplimab, another PD-1 inhibitor. This trial, as you can imagine, was done pretty much in the rest of world because there were already approved agents for frontline therapy in the United States. It brought access to this drug in a large number of countries. Interestingly, to be in this trial, you had to have PD-L1 greater than 50%, or it’s also for high PD-L1. And you had to have had some smoking history. They also allowed patients with brain metastases, so it was a real-world population. But I was impressed that the hazard ratio was in the 0.5–0.6 range for survival, and with PFS [progression-free survival], the difference was quite significant, as well as response rate. The toxicities of this agent don’t look much different from what’s seen with some of the other immune-acting agents. What does this mean? It means that there’s another agent in the front line that can be used alone in the 50% positive patients or combined with other agents. I expect we’ll see that as well. But I’m reassured that we have three agents like this as single agents in the front line. Of course, we have combos as well, bringing nivolumab and ipilimumab into the fold. it’s good for patients. It gives us more options in the clinic.

Roy, as you point out, it excluded never-smokers from this trial, which may have enriched it for the squamous histology patients, who were well into the 40% or so. They’ve also included the hepatitis B and C population, at least who have chronic viral infections, although there were very few patients with this issue on the trial. How do you see this influencing choices? In my practice, you’ve been using pembrolizumab for so long. It’s not clear that this is necessarily better. It certainly doesn’t appear to be. It appears to be every bit as safe as pembrolizumab. I’m wondering how you put this into the current landscape.

 It’s reassuring. It looks about the same at this stage—not so early. It’s a bit more mature than atezolizumab data were when first presented. But of course, from Drs [Julie] Brahmer and [Martin] Reck and team, we have five-year data with KEYNOTE-024. The 5-year survival was 35%. That’s the gold standard. This drug looks, at this point, to be quite similar. Antibodies against PD-1 or PD-L1, this 1 against PD-1, might have slightly different epitopes that they recognize. Perhaps, they have slightly different half-lives. They’re not all the same. So far, this has also beat chemotherapy, a little bit different population.

This gets to preference. I’m used to using pembrolizumab in this setting. I probably still will. Is there a time when my hospital pharmacy officer might say, “Roy, we’ve made a deal, and we’re using this drug, and we feel it’s equivalent”? If that were to happen, 1 could say that probably is true, and we do that. My goal is to get immunotherapy to more patients. Mark, having done this for a long time, these drugs work only if people have access and can afford it. Certainly, having multiple choices are good. But right now, I’m still using pembrolizumab. We just published the KEYNOTE-010 second-line data, and those looks pretty good in the high group—20%, 25% at five years. Let’s keep following these data. I’m sure we’ll see them again at different congresses.

I want to get your thoughts on one of the issues: crossover rate on the chemotherapy arm. The crossover rate was in excess of 70% to cemiplimab. We still see, even though most patients are getting the immunotherapy second line, that the survival doesn’t catch up to you getting first line. Do you have any thoughts in that area? We saw the same thing with KEYNOTE-024. Dr. Brahmer presented some data along the same lines.

 Yeah, KEYNOTE-024. I discussed that at ESMO [European Society for Medical Oncology Congress] last year. But even at 66% crossover, and the survival didn’t get up to the 35 months. Maybe it approached 20 months or something like that. My feeling is your best shot with immunotherapy is to get it right away. Certainly, having used chemotherapy for quite some time, you’re never going to cure anyone or have that long-term tail of the curve with chemotherapy. What happens with a crossover is some patients don’t make it through. They progress; they get neutropenic; they have a pneumonia. Something happens during the first therapy. You can’t always rescue those patients. Frontline use of these drugs is where they should be. Of course, we’re seeing, from ASCO [American Society of Clinical Oncology Annual Meeting], adjuvant therapy having effects, so move these drugs earlier and earlier with a stronger, healthier patient. But it’s nice to know that if you don’t give it front line, perhaps you’ve got someone who can’t get immunotherapy for some reason, then you can always add it in later.

Roy S. Herbst, MD, PhD, of Yale Cancer Center, highlights treatment advances in stage IV non–small cell lung cancer, including cemiplimab as frontline monotherapy based on EMPOWER-Lung 1.

Neal Edward Ready, MD, PhD, Duke Medical Center

Articles