Immunotherapy for Lung Cancers: First-Line and Beyond

H. Jack West, MD; Edward B. Garon, MD; Hossein Borghaei, DO


November 03, 2017

Editor's note: This roundtable discussion on important issues in immunotherapy for lung cancer was filmed at the American Society of Clinical Oncology (ASCO) 2017 annual meeting in Chicago on June 4.

H. Jack West, MD: Hi. I'm Dr Jack West from Swedish Cancer Institute in Seattle, Washington. Joining me today for Medscape Oncology Insights are two colleagues who are experts in immunotherapy for non-small cell lung cancer (NSCLC) and the whole field of thoracic oncology. First, I have Dr Hossein Borghaei from Fox Chase Cancer Center in Philadelphia and Dr Eddie Garon from University of California, Los Angeles. Thanks for joining me.

Edward B. Garon, MD: Thank you.

Hossein Borghaei, DO: Thank you.

Pembrolizumab for First-Line Therapy in NSCLC

Dr West: The field [of NSCLC] has been very dynamic recently. Let's start with how we would approach a patient in the first-line setting. In late 2016, pembrolizumab was approved for first-line treatment of patients with high programmed death-ligand 1 (PD-L1) expression, which affects about 30% of patients regardless of whether they have squamous or nonsquamous histology. At the same time, we had a presentation of phase 2 data in the KEYNOTE-021 G cohort on 123 patients.[1] Now [pembrolizumab] is approved as an option for patients with nonsquamous histology regardless of PD-L1 expression.

What do you think of this option? Is it something you are going to be inclined to widely adopt or selectively adopt? Also, how do you think it's going to affect the broader general oncology community? I'll start with you, Eddie.

Dr Garon: It's an excellent question and one that we colleagues have been debating amongst ourselves for a while. As you talk to different people, you get very strong opinions that differ from one another. In my practice, I must admit that I'm very cautious of these data. There are reasons to be cautious of the data based on the fact that this was a small study.

The bigger question to me is: What is achieved by using both chemotherapy and pembrolizumab in this case, or any other programmed death-1 (PD-1) or PD-L1 inhibitor, together rather than sequencing them one after the other? With the available data, I would argue at least that while response rate and progression-free survival (PFS) have improved with the combination, there are no data that survival has improved, and until I see a benefit in survival, my concern is that the combination basically adds more time on drug and more toxicity for more people without a resultant survival benefit. For me, that does not hit the bar that says it should be used as part of my practice in any routine way. Again, that is something that people disagree about.

[M]y concern is that the combination basically adds more time on drug and more toxicity for more people without a resultant survival benefit.

Dr Borghaei: I absolutely agree with you. For the standard-of-care practice, for anybody with 50% expression of PD-L1, I think I am going with single-agent pembrolizumab.

Dr West: That is an important question in itself, right?

Dr Borghaei: Absolutely.

Dr West: First of all, if you do not need to test, should you still test?

Dr Borghaei: I do not think we're at the place where we can say that we do not need to test. I absolutely support testing PD-L1 at the time of initial diagnosis, much in the same way we are doing all the molecular testing. As far as I'm concerned, for a patient who walks in through the door with an initial diagnosis, we do the whole battery of tests that now includes PD-L1. If they hit that 50% magical mark for studies, then as far as I'm concerned, pembrolizumab is the standard of care. For full disclosure, I participated in the KEYNOTE study you mentioned.

Dr West: We all have a perspective here.

Dr Borghaei: Exactly. From a scientific point of view and a clinical point of view, it is a combination that should have been investigated, and it should be investigated further. There are multiple phase 3 studies with combinations of different chemotherapy drugs and different immuno-oncology (IO) agents that are eventually going to be reported in the near future, so we will have some more clarity about the data.

I agree with you, Ed, that at this point, I do not consider the combination of chemotherapy with an IO agent as the standard of care in my practice. For greater than 50% [expression of PD-L1], I prefer to use pembrolizumab alone. For less than that, I use standard platinum chemotherapy in the absence of a clinical trial or molecular targets. At the time of progression, I have a discussion [with the patient] about using IO in a second-line setting based on the published randomized phase 3 studies. I cannot argue with the fact that we have to wait for the survival data.

Dr Garon: One thing I would like to comment on is something you said that I think is important for the audience. Many of us, myself included, have taken part in some of these trials. We participate in a trial because it is a question that deserves to be answered and one that is important. And this is a great example. My concern in this data set is that, to date, we have not seen a survival advantage. If the trial that you are involved with or trials that any of us are involved in in this sort of combination space show that there is a very significant survival advantage, then that is a very different story.

Significance of the KEYNOTE-021 G Cohort Survival Trend

Dr West: Absolutely. Let's just double back a bit. At the American Society for Clinical Oncology 2017 annual meeting, we saw an updated data set on the KEYNOTE-021 G cohort.[2] It showed a nonsignificant trend in the direction of more favorable survival. We also know that in the trial there was crossover but not complete crossover. How impressed are or aren't you with this? Does this change anything for you?

Dr Garon: To be honest, this is an issue that I become very concerned about recently—how we evaluate clinical trial data. Traditionally, what happened was that you put together a protocol with a statistical analysis plan. I know many of the watchers are now glazing over as we are talking about the math of this.

You are going to look at a certain time, and that is important because certainly if the Atlanta Falcons could have ended the Super Bowl a little bit earlier, they would have been quite happy. There has to be some point where you look at what actually happened.

If you allow [investigators]—and this is a data set that I didn't happen to be part of, though there are plenty of similar data sets that I was and have been a part of—to continually cut and recut and recut the data, I think they become a little bit hard to interpret. Looking at an updated data set, in a study you already know is positive at a later time, and showing that it looks a little bit better does not change my opinion much.

Dr Borghaei: I agree with that. I do not think it's changing my opinion about the combination. Remember, a lot of these studies have survival as a secondary endpoint. To some extent, because there is a sequence of events that is going to happen in the future, it's okay to go back and look at it. However, it does not increase the statistical significance. It does not necessarily mean it's a practice-changing study at this point. It is still a 60-patients-per-arm clinical trial. There are plenty of phase 2 studies in lung cancer that end up being disastrously bad in a phase 3 trial.

Dr West: In literal terms, because there is US Food and Drug Administration (FDA) approval, it will be practice changing.

Dr Borghaei: I agree.

Dr West: I personally think that many general oncologists or nonspecialists in thoracic oncology are going to be more inclined to give it up front and not worry about testing results or think about it too much. That is not necessarily optimal for the majority of patients, but I do think it's going to have a bigger impact than it will among the three of us.

Dr Borghaei: It's possible, and I agree with that. But I think a lot of general oncologists are also aware that having that gold standard randomized phase 3 trial is important. I'm not 100% sure that everybody is going abandon PD-L1 testing and say, "Anybody with nonsquamous histology is going to get the combination, and we're done with testing PD-L1, and that's the end of it."

I'm hopeful that people are going to look at the data critically—maybe not as in detail as you and I are doing right now—and say, "This is a phase 2 study with no survival advantage. Should we sequence it, as opposed to giving everything upfront?"

I think part of the concern about the approval was that this was seemingly a rush to judgment.

Was FDA Approval a "Rush to Judgment"?

Dr West: I think part of the concern about the approval was that this was seemingly a rush to judgment with 123 patients, when we have an avalanche of data coming in the next 6-12 months. Multiple trials will have at least preliminary data presented, or press releases in some, and larger data sets that we'll be able to look at, including KEYNOTE-189, which is basically the same question but in a larger format. It seems a little early to be making the decisions based on phase 2 work.

Dr Borghaei: I'm not at the FDA. I've never worked there. I'm not sure what criteria they consider for [approval], but if you look at it from maybe a patient advocacy point of view, is having a better response rate more important than waiting for the survival data? We are talking about patients with incurable metastatic lung cancer. Maybe there are other factors that went into the decision.

Dr Garon: Rather than considering it a rush to judgment, I would say it is more an issue of what the important endpoint is for you. If the important endpoint for you is PFS, I feel reasonably confident, based on the KEYNOTE-021 G cohort, that adding pembrolizumab to chemotherapy increases PFS. Personally, that endpoint is not enough for me to change my practice pattern. Overall survival would [need to] be the endpoint.

Dr West: I think that's what the FDA was trying to get behind—not delaying what was seen to be an overwhelming benefit, even if it was not everyone's preferred endpoint. It was a very commanding improvement in PFS as well as response rate with the KEYNOTE-021 G cohort data. It's fair to ask the question. I don't know that it clearly has to be a major survival benefit if its trend is right, or is even about the same, but you have twice the response rate and a much longer PFS.

That is where things are. We will have a lot more data on the question and potential answers for combination chemotherapy and immunotherapy, with the understanding that it's a very viable alternative to select your highest PD-L1, or best predictive value for high PD-L1, and spare them chemotherapy and start with immunotherapy. Sequential chemotherapy followed by immunotherapy is another fair option.

Do You Delay Immunotherapy for Certain Populations?

Dr West: Are there patients for whom you are inclined to relegate immunotherapy to a significantly later line? Is it never smokers or patients with a driver mutation? Who do you deselect immunotherapy for?

Dr Borghaei: This is another really good question that we do not have a lot of clinical data to answer, but we can extrapolate data from some of the reported phase 3 studies in the second-line setting.

In my opinion, if somebody comes in with a molecularly driven tumor, such as epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK)—for all of those patients, I'm not offering them single-agent IO in the second-line setting. I make sure they have exhausted the tyrosine kinase inhibitor (TKI) options. I prefer platinum doublet chemotherapy if they do not have a TKI option over single-agent IO. The only exception I make is if I have a rationally designed IO combination study asking a reasonable question, I'm willing to consider a particular patient for that particular trial. As standard practice, I have not done it.

We are beginning to test everybody's tumors to see if there is any PD-L1 expression. Many community oncologists in my area have asked me: What do you do with a patient with an EGFR mutation, ≥50% PD-L1 expression, and no other options? I do not have a clear answer, but I would likely say that that it's probably safe to try it under very close observation and that if there is any sign of progression, the patient should be switched to something different. I would fully admit that I do not have a data set in support of that approach.

Dr Garon: First of all, it's a little easy for us in this situation because we tend to have a bunch of combined immunotherapy protocols, and that tends to be our default answer. For people in practice, this is, of course, more difficult because they are having to make a clinical decision.

A very important issue is that your care of a patient is in some ways a partnership between you as the physician and the patient. For instance, I'm pretty confident that a patient with an EGFR mutation who has no particular factor making them more likely to respond to an immune checkpoint inhibitor would be better off with, for instance, docetaxel/ramucirumab after all other therapy; in other words, after platinum-based chemotherapy and all their TKI options are completed. That being said, after having this conversation with many patients, that is not what my practice pattern ends up being for the most part because that is not what the patient wants, and I do not have level 1 evidence saying that is the wrong thing to do. The patient does get a say, of course.

Dr West: I think we all agree that in patients with a driver mutation (eg, EGFR, ALK, ROS), our first-line approach is going to be targeted therapy directed at that driver. It's important to note that in nearly all of these trials, patients with a known EGFR mutation or ALK rearrangement were not eligible and were directed presumably towards targeted therapy. When you have exhausted those targeted therapy options, do you think a little more about chemotherapy/immunotherapy, or do you still favor a sequential approach?

After Targeted Therapy Is Exhausted, What Next?

Dr Garon: Full disclosure: We do have a clinical trial in patients with PD-L1 expression receiving front-line pembrolizumab instead of the initial TKI. That is in a clinical trial context; I would not do that as my standard of care at this point.

Whether or not that subgroup will do better on chemotherapy/immunotherapy combinations is a great question I am hopeful some of the phase 3 trials can answer if they include patients with those mutations. The problem is that because [investigators] have largely looked for front-line indications rather than taking patients after a TKI, they have been excluded. That important question may not be so easy to answer.

Dr Borghaei: I agree. That is a problem.

Dr West: One of the challenges we face in the clinic is that patients read about [immunotherapy] in the news, and there are incredible amounts of attention and hype about it. It's important to distinguish for those patients: "Don't trade in your EGFR inhibitor for immunotherapy; that would be a mistake here."

Dr Borghaei: Right, that is where the physician responsibility lies. We have to spend a little bit more time explaining, "Look, this is not in your best interests at this point because you have a really good viable option, and all the hype about immunotherapy might not necessarily apply to you." Yes, patient engagement is really important, but to some extent in the limited time that we have in the office, we have to do some level of education for somebody who is newly diagnosed with an EGFR mutation. I think it's a mistake to put them on an IO agent outside of a clinical trial.

When (If Ever) Is Upfront Chemotherapy/IO Considered?

Dr West: Are there patients for whom you would potentially welcome this, such as patients with poor performance status, certain social issues, or who had significant progression between their initial CT and the PET CT after their biopsy? Patients who may not do well 8 weeks from now if they do not respond? Is there a subset of patients you would be inclined to give chemotherapy/immunotherapy all upfront and hope for the best? We have all said that it's not our general preference, but are there patients for whom you would be inclined to say you would?

This is the art of medicine. This is not all science. Not everything is going to be captured in data forms and clinical trials. Patients do not follow the rules or read the textbook.

Dr Borghaei: There are always cases like that. We can all think of cases in our clinics where we would really like to hit patients with whatever we have because we worry that we only have one chance because of the rapid progression and nature of the disease.

This is the art of medicine. This is not all science. Not everything is going to be captured in data forms and clinical trials. Patients do not follow the rules or read the textbook. For the individual sitting in front of you with their hopes and fears and the disease, you have to make a clinical judgment and say, "This is the patient I'm going to use the combination in." I'm perfectly happy with that kind of an approach. I will probably do that.

Dr Garon: That that is probably true for me as well. The real trick here is not to change what our view of that sort of "disaster" is when we treat with everything right now because this option is available. This is going to be a trick that I have in my practice, and I imagine that people in their own practices will have as well.

IO Combinations

Dr West: In the near future, we are going to have some data on IO combinations: immunotherapy with a PD-1 or PD-L1 checkpoint inhibitor as well as CTLA-4. There have been toxicity challenges; but, in general, we have seen that you can de-escalate the CTLA-4 component, and it's more manageable. How excited are you about this to potentially displace or even to add to chemotherapy, in addition to the options we already have?

Dr Garon: I end up being the killjoy on this one. It's not that I'm not hopeful that this will be something better because, of course, I am. I will be very concerned until I see randomized data showing that this is a better and reasonably tolerated approach before I embrace it. I have had significant fears as to whether patient selection issues are driving the relatively favorable results we see with the combination rather than the single agent. The nice thing about having this as a concern is, unlike some of the other things we have mentioned, that I will get an answer. There will be randomized data, and hopefully that will help me get over my fear that if results are positive, they were based on patient selection.

Dr Borghaei: Again, for full disclosure, I have been participating in a lot of trials using these combinations with CTLA-4 antibodies. I share your concerns. Toxicity has been an issue, but now with the new dosing and schedule it's a little bit less of an issue but not completely gone.

I do agree that patient selection probably does play a role, as this does in many early stages of development of any new combination. I completely agree with you that the randomized study, which is ongoing, is the best way to figure out if this combination has any leg to replace anything that we are doing, or if it is just simply going to add to what we are doing.

Dr Garon: Just to clarify, patient selection is okay if you are looking at a patient and saying that this patient cannot get the combination. In clinical practice, we would do that all the time. Until I see randomized data, the question I have is: Are the differences we see in nonrandomized patients driven entirely by patient selection, or are these therapies really better? I do very much look forward to seeing those randomized data.

Dr West: I will reassure you that I'm very much in your camp here. We are all interested in this, but there are major barriers. Patients who go on [study] are a selected population. They are the most motivated and probably more fit than the rank and file folks at the Department of Veterans Affairs and the broader community. Centers where they are treated have more experience and more support. It's been a steep learning curve for everyone, and I'm afraid that that would likely be recapitulated in the broader communities where the patients have less support and the doctors have less experience.

There is going to likely be some significant gulf between selected phase 1 and 2 trials and broader phase 3 experience. There is another gulf between patients who go on phase 3 trials and the 81-year-olds in the community who may have an ill-advised recommendation to get what was on the trials.

There is a lot of reason to be cautious. In January of 2017, two press releases about results with these approaches were most notable for what they did not say. They both basically said, "We will let you know later; there is nothing to say right now." No one sits on great news. We will see what happens, but we really have not learned anything now. I take that with a guarded pessimism or wariness. As you say, we will see.

Then there are the practical societal issues. The only thing potentially more challenging than ongoing immunotherapy for everyone with one checkpoint inhibitor is combination immunotherapy for everybody. That is something we cannot even begin to fathom right now.

Dr Borghaei: I agree. Those are really good points.

Treating Beyond Progression

Dr West: One of the other concepts that we are beginning to get a little data on and hear more about is the concept of treating beyond progression for patients with RECIST (response evaluation criteria in solid tumors) criteria. In reality, most of that, if not the vast majority, is still in the second-line setting, saying that patients treated beyond progression seemed to do better than we might expect and better than the arm that was not assigned immunotherapy.

How much does that contribute to your thinking? Does that apply in the first-line setting? Retrospectively, I'm afraid that the take-home message for this selected population is: If the scan looks better, keep going; and if the scan looks worse, keep going. Is this is a very applicable change in how we should be practicing oncology, or is it something that is still only for very selected patients?

There is progression, and then there is progression.

Dr Borghaei: I would say to proceed with extreme caution in this area. There is progression, and then there is progression. Somebody sitting in front of you who is losing weight, having more shortness of breath, having a hard time getting out of bed, and has a deteriorating performance status is clearly progressing regardless of how big or small the RECIST criteria or CT is. On the other hand, if you have a patient with [significant] disease who is, for the most part, stable or better with one or two areas that might be a little bit different and is clinically doing better, maybe you do want to consider treatment beyond progression with close follow-up.

That is one area that clinical trials have not addressed adequately. Which patient population has the possibility of receiving treatment beyond progression? In my practice, I do not do that very often, even though there is some idea that you might be able to get away with it with immunotherapy. I get nervous about that myself. I do not have a lot of patients who get treatment beyond progression because the disease is rapidly progressing, and I do not want to do that.

Dr West: There is a significant difference between having received second-line therapy where the options afterwards are very limited and not especially effective, and after first-line therapy, where if someone has never gotten chemotherapy, you would not want to necessarily obviate that.

Dr Borghaei: I still have a little bit of a problem saying, "We are in the second-line with an IO agent, the disease is progressing, and we really don't have any other good options." There is always an option of going to a different drug if the drug you are using is potentially going to cause significant toxicity, and it's clearly not working. We have to be careful with that kind of an approach: that just because we are afraid the next line of treatment is going to be docetaxel that we should avoid it at all costs. I do not think I like that approach.

Dr Garon: Just as you mentioned the learning curve for toxicity associated with dual checkpoint inhibition, I think there has been a learning curve with how to use these agents, particularly in lung cancer. Some of these issues of atypical radiographic findings are much more pronounced, for instance, in melanoma. Our group here does not really treat melanoma, but for most people who use these agents, their first exposure was in melanoma, where this may be more of a real issue.

Early after the approvals, I remember getting so frustrated when I would have consultations and patients would come in progressing on nivolumab for months because their doctor had told them, "It's going to get better in a little bit, you've just got to keep plugging through." I'm not getting those consultations to the same extent anymore. Just like there was a learning curve for me with these agents that generally happened in the context of clinical trials, I think that practitioners had that learning curve in the clinic. I am somewhat encouraged that people now see that if you have a patient a couple of months in who is feeling worse, and their scans are worse, the drug you are giving the patient is not working. Unfortunately, that is an issue we have learned to deal with in oncology.

Options for Second-Line Therapy

Dr West: There is also this new entity of patients who respond well for 18 months or 2 years on, say, first-line pembrolizumab, and then they progress. Obviously, chemotherapy-based treatments are an appropriate option, but how interested and inclined are you to try to perpetuate the immune response by either adding to it or switching agents? Certainly, I know physicians in the community are wondering whether we might be able to do better by adding a CTLA-4 inhibitor or switching from a PD-1 to a PD-L1 inhibitor in the second-line. Any thoughts on those data to guide people?

Dr Garon: This is a place where I feel like I have a lot of experience but not a lot of data. I think that the experience can be helpful. I would really categorize progression in those initial responders into two camps. One of them looks exactly like the progression on carboplatin/paclitaxel.

Dr West: Unequivocal.

Dr Garon: Yes, exactly. Then there are patients who essentially will have an escape lesion. I can tell you that the natural history of those are also unusual because sometimes on clinical trials, we had no choice but to just watch them. You could watch them over a long period of time. Sometimes they would get a little worse on each scan, and maybe they would stop growing for a couple of scans and then grow again. In those patients, we had great success with ablation or radiation and then continuing on with their therapy.

Dr West: Did any of them get resections and learn anything from that?

Dr Borghaei: Again, I agree with you. We have done a similar approach. It goes back to this concept of oligometastatic disease that we keep hearing about. If somebody is full-blown progressing, you know that from the scan and their clinical presentation. In our centers, for an isolated lesion here or there, surgical resection has been used to control disease locally; and for the lung cancer patient, more likely with local radiation.

Dr West: If you have patients with more diffuse progression, and it's not amenable, or it's not one or two lesions, can you get the genie back in the bottle with another immune intervention?

Dr Borghaei: I do not know if we have the data on that. That is another risky area without any clinical trials. If you have somebody on front-line pembrolizumab who is progressing, by adding a CTLA-4 antibody in the absence of data, you might end up doing more harm than good. I do not think I would like that approach.

Also, I do not think there is a lot of clinical or scientific rationale to switch from a PD-1 to a PD-L1 inhibitor and expect to get a response. I do not think that has been proven either preclinically or clinically to be a useful maneuver for rescuing these patients.

Dr West: To be clear, that is not my angle. People are asking about this, and I think it's being done.

Dr Borghaei: Absolutely; I get the same questions all the time.

Dr West: Outside of any clear guidance, a lot of people just want to go back to the well.

Let's turn to the second-line setting, which may be narrowing because there are going to be a lot more people getting immunotherapy first-line, whether it's by selection or not. Do you see any clinically significant differences among the agents available now or anything being tested? Is anything likely to bubble up as a checkpoint inhibitor of choice in the second-line, or is it all just pick and choose based on your own comfort level, whatever the insurer prefers, etc?

Dr Garon: From an approval perspective, you cannot give pembrolizumab to someone who does not have staining for PD-L1, and nivolumab has more frequent dosing. Other than that, I view them as somewhat interchangeable.

Dr Borghaei: Me too. I treat them pretty much as you said. For people who live locally and do not mind coming in every 2 weeks, we might use an agent that is given every 2 weeks. For somebody who comes in from 4 hours away, I might choose an agent that is given less frequently. The PD-L1 testing, obviously, is important in a second-line setting.

Dr West: Yes, for pembrolizumab.

Dr Borghaei: Exactly.

Dr West: I have heard that some guidelines might prefer one over another based on cost. This may emerge as a differentiating point if not a strong one. When you have what are essentially interchangeable choices, if guidelines say you should choose one over another because it costs $1000 or $2000 less a month, I do not think a lot of us would get on the phone and have a tantrum over that.

Dr Borghaei: Absolutely not.

Dr Garon: I wish they would have more of a competition. That would be a good thing for availability, were that to happen. You are right; some pathways have preferred one over another based on cost, but even when you look at them, the amount of difference is fairly small.

Does Drug Cost Affect Your Decisions?

Dr West: Whether we are talking about stage III or stage IV disease, the reality is that we are now working with drugs that are costing $15,000 a month and combinations costing $20,000 or $25,000 per month at least. As a global question, do we need to see robust real survival differences just because of that as compared with adding docetaxel or something?

Dr Garon: That is a hard question because our system is somewhat unusual in that we do not understand exactly who sees that cost. If we felt like we were acting as salesmen in the clinic, we might be more uncomfortable with the cost with respect to the benefits. Right now, whether it's right or not, I do not feel that way. I do not want to feel that way. I have generally not necessarily made that the way I make decisions.

Dr West: I think that is the way most people feel, but the problem is that somebody is paying for it. Everybody is paying for it. No one in the room is getting affected by it. Therefore, everyone who is making decisions does not have real skin in the game financially or, if anything, has other incentives. It's a challenge that we really do not address that much. We would be forced to be somewhat schizophrenic and choose between our best duty to a patient versus potentially competing interests to society, and it would be nice to not be forced into that dichotomy.

Dr Borghaei: I agree; that is a difficult question.

Immunotherapy for Small Cell Lung Cancer

Dr West: We are also getting a little trickle and a little more information about immunotherapy for small cell lung cancer (SCLC). In fact, it's now in the NCCN Guidelines[3] as an option in previously treated relapsed SCLC with either nivolumab or the combination [of nivolumab and ipilimumab]. This was not based on a large amount of data[4,5]; data were certainly very limited. The current standard of topotecan is not beloved by any stretch and not very active.

How do you see this playing out? Do you see that as an appealing option? Because the reality is that it's in the NCCN Guidelines, and it's commercially available. The potential is for a more durable response if you get one. Where does it rank for you?

Dr Borghaei: In small cell refractory disease, I would describe my practice as desperate. I need something that works or something that is a little bit better than the topotecan that we have available right now.

I have been using the combination. I am doing a little bit of patient selection here with somebody who has a better performance status and who can possibly tolerate toxicities with the combination. In my practice, we are using the combination for refractory disease based on the NCCN Guidelines and based on the small study that you are referring to that has been published.

I like the option. I like to be able to use it. Obviously having more data would be very helpful to guide us exactly where to go with the combination. Again, is there a way to choose the right patient who might benefit from it? We do not have any of that information, but I like having the option of using the combination.

Dr West: I'll turn to you and ask the same thing but also specifically, with the question of the combination of nivolumab and ipilimumab versus single agent in a population that can often be riddled with comorbidities and have a marginal performance status when they relapse.

In all honesty, it has been several years since I have used topotecan.

Dr Garon: As I mentioned earlier, I have been very reluctant to embrace combinations of PD-1 and PD-L1 inhibitors plus a CTLA-4 inhibitor until I see randomized data. I have not used nivolumab and ipilimumab together. I certainly have used nivolumab. In all honesty, it has been several years since I have used topotecan because I do not like giving it based on both the efficacy I'm used to seeing and the toxicity. That being said, we have not liked that regimen for a long time. We have been pretty confident that a whole host of things would beat it, but that has not turned out to be the case.

I would be terrified to see that happen this time because I certainly would rather be able to give my patients the checkpoint inhibitor, realizing that both of them are very unlikely to have a substantial benefit. But if they were to have a substantial benefit from nivolumab, it tends to last longer and be associated with less toxicity.

Dr West: I feel that we have such an unacceptably unimpressive standard that it could afford you to be more liberal. You would not necessarily need to have the same rigor because what are you going to lose? Not much. Like you, I have been terribly unenthused about topotecan. With those kinds of positive results, who needs negative ones?

Immunotherapy for Mesothelioma

Dr West: A very similar situation emerges in mesothelioma. We are starting to get some data. Various trials are largely showing response rates of 9%, 10%. It really is interesting to see some people spin it as, "It's active in mesothelioma; we should do this." It's also a situation where after first-line, we do not have a standard with meaningful efficacy. We are starting to see nivolumab versus nivolumab/ipilimumab in mesothelioma.[6]

How rigorous are you going to be? How much data are you going to want versus just seeing what it does?

Dr Garon: One thing that is different in mesothelioma is that there is no standard second-line therapy. There we have, in many respects, an easier time just embracing it. That terror I have that someday someone will announce that nivolumab did not beat topotecan is not there because there is nothing to beat.

Regarding the combination for which we do not have many data on yet, I do get nervous. Certainly, there was a randomized trial[8] with a different CTLA-4 inhibitor, tremelimumab, where it was pretty convincing that tremelimumab was no better than nothing in mesothelioma. I understand that there are issues about whether you are modulating the immune response and different aspects when you are combining with a PD-1 or PD-L1 inhibitor, but it still makes me a little nervous.

Dr West: There is a potential harm.

Dr Borghaei: There is always a potential for harm with these agents. That is why you have to be very careful.

Dr West: Are you encouraged enough by these data to offer it for your patients after first-line?

Dr Borghaei: In all honesty, and in the absence of a clinical trial, I have offered single agent to a couple of my patients. Anecdotally, I have seen some responses, so that has encouraged me to be a little bit braver about discussing it with a potential patient. But again, it's a data-free zone. We do not have the kind of data that we need to necessarily say that, based on a phase 3 randomized study, we should use one versus the other. We do not have that.

Dr West: I've done the same thing with a single agent; and, actually, my best patient benefiting has not had what would be categorized as a response. We should parenthetically mention that mesothelioma is not an easy one to assess for response.

Dr Borghaei: It's not.

Dr West: This is somebody who has had marginal shrinkage for a very long time, and her symptoms have decreased over time. That would not be captured in a response rate benefit, but to me and to her, there is no question that it's been very beneficial for a sustained period. This was somebody who was progressing convincingly and increasingly symptomatic after chemotherapy. This is a situation where prolonged stability could be a very meaningful benefit. This is also a situation where, as we have alluded to, there is plenty of room for good judgment.

Hoss and Eddie, thank you so much. I don't know that we have figured out all of these vagaries and options. The data will keep coming, and we will have more to talk about. Thanks a lot.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: