Let’s move on for the sake of time to adjuvant immunotherapy. Heather, you’ve been leading the charge on this, your first author on the paper in [

] with adjuvant atezolizumab, the IMpower010 trial [NCT02486718]. Can you walk us through that data? You know it well and just what we saw with adjuvant immunotherapy.

: There have been 4 large, randomized trials looking at adjuvant immune therapy. This is the first with data and randomized patients who had completely resected stage IB to stage IIIA disease. They all had to get chemotherapy and then randomized to get atezolizumab for a year, so only a year or just best supportive care. The study was designed with a sort of hierarchical testing with the first look to be the stage II to IIIA patients who had some PD-L1 expression. That was the vast majority of the patients. There were only about 10% to 12% who had stage IB, so most had stage II and IIIA. And when we look at the PD-L1 expression, it was in over half the patients, so 55%. In that subgroup, the disease-free survival [DFS] hazard ratio is 0.66, and that led to FDA [US Food and Drug Administration] approval for that group of patients, which I think was the right thing to do. We’re still waiting for maturity with the stage IB and obviously need to see overall survival [OS]. Is that hazard ratio as good as what we saw with osimertinib in ADAURA? No, but the difference is with immunetherapy.When we think about the PACIFIC [trial; NCT02125461], their disease-free survival translated into an overall survival, which was parallel. When we’re talking about stopping the targeted TKIs [tyrosine kinase inhibitors], we don’t know that that persists and maybe staying on a drug lifelong, isn’t a bad thing. It’s a different way of thinking about it. The other thing that’s important to note with the IMpower010 data is that that benefit was really driven primarily by patients who had the higher PD-L1 expression on their tumors. In that subgroup with high PD-L1 greater than 50% that progression-free survival hazard ratio is 0.43. That is really good. With chemotherapy, it’s 0.86 for disease-free survival. That OS was similar too. That’s exciting. We also have the PEARLS trial [ NCT02504372], which was with adjuvant pembrolizumab. That was also presented just a few months back. Their overall hazard ratio for DFS is similar. It’s like 0.76 in all-comers, regardless of PD-L1 expression, but what makes no sense is it didn’t matter what the PD-L1 was so there wasn’t increased benefit with high PD-L1. It was not. It was not as good. That is a little odd. Then when you look at the data and the patients who did have 

 mutations, it’s also confusing because they seem to benefit in both trials, but in IMpower010only if they had PD-L1 expression. Adding to our earlier conversation in the subset that had an 

translocation, even if they had super high PD-L1, no benefit at all. If you gave them immune therapy, you weren’t helping them. That’s where we get back to we got to look for everything because then there are other drivers like 

, where you know immune therapy’s more likely to work. We have to really know the full story of the tumor before making decisions, but I’m super excited about immune therapy in the adjuvant setting. We’re obviously waiting to see more OS data. The initial hints were good in both studies.

You mentioned the PD-L1­–high group benefited more. Are you comfortable giving this? Is there a patient you wouldn’t give this to and are you comfortable giving this to the 1 to 49s?

: I don’t give it to patients who do not have PD-L1 expression in their tumor. I am not giving it to patients who have an 

 translocation, or another driver that is a driver that in metastatic tells us immune therapy’s less likely to work. We don’t have definitive adjuvant data there because there weren’t big enough numbers to look at and we didn’t do that full testing but I get nervous that I don’t think I’m really helping them. I will give it to the 1 to 49, but with discussion and saying it like this;“We don’t know how much benefit there is. It was trended but it wasn’t statistically positive in that group and I think we need longer-term follow-up.” We need to see what that OS looks like. We need to remember in metastatic. We had trials where there was no DFS benefit and there was OS benefit. There’s just a lot to understand and then there are more trials to read out.

I think we’ve got a lot of questions still to answer. I think you’ve done a nice job answering some of these questions from your leadership in the trial. I’ll ask another unfair question, which is a patient who has high PD-L1, let’s say 90% heavy smoker, but doesn’t want chemotherapy. Your trial mandated chemotherapy. Would you have any equipoise for that patient to go on single-agent?

: The PEARLS trial didn’t require chemotherapy. From there we have some data but it didn’t seem to matter as much. I think the combination of chemo plus immune therapy is going to be the better approach in the adjuvant setting, just like it is in the metastatic setting most of the time. Those trials are being designed, not done yet, and neoadjuvant, of course, which we will talk about more, that’s a combination.

Videos

Expert perspectives on use of immunotherapy in the adjuvant setting of early-stage NSCLC in light of data from IMpower010 and PEARLS.

Adjuvant IO Therapy in Early-Stage NSCLC: The IMpower010 and PEARLS Trials

Let’s move away from chemotherapy and talk about more exciting targeted therapy, where we are seeing huge disease-free survival [DFS] advantages. Jonathan, can you give us an overview of the ADAURA [study; NCT02511106] data? It was practice-changing and it is the first targeted therapy approved in the adjuvant setting for lung cancer. There’s been a lot of work put into this, and it kind of hit the mark on every level,

: The ADAURA trial with adjuvant osimertinib is one of the pivotal trials in 

-mutant non–small cell lung cancer. Based on the robust progression-free survival and overall survival [OS] benefit shown in the FLAURA study in the metastatic setting of osimertinib compared to earlier generation, first-generation EGFR TKI [tyrosine kinase inhibitors], this was bringing osimertinib to the adjuvant setting looking at stage IB to IIIA 

-mutant non–small cell lung cancer, the more common 

 exon19 deletion and L858R mutations. Patients with surgically resectable stage IB to IIIA 

-mutant non–small cell lung cancer with those common mutations randomized to receive osimertinib vs observation osimertinib for 3 years. The primary end point was in stage II to IIIA patients by the AJCC 7th edition [The American Joint Committee on Cancer: the 7th edition, effective for cancer diagnosed on or after January 1, 2010] of lung cancer staging, and there it showed a robust disease-free survival benefit that was quite striking. We still don’t have overall survival data that’s mature but the trend was in the right direction and so I consider it a practice-changing study despite not having the overall survival benefit mature results yet- because of the magnitude of DFS benefit. I do offer it to my patients certainly for IB to IIIA 

-mutant non–small cell lung cancer, and also with a robust but not quite as good a hazard ratio as stage IB; being not as advanced that makes sense. It still looks quite good and I do offer it to those patients as well but with a more nuanced discussion.

I will speak for myself and say I was overwhelmed with that hazard ratio. I hope we see an overall survival advantage. It’s tough for me to believe we won’t. We have an early data cut for central nervous system [CNS] metastases and not surprisingly there were way more CNS metastases in the placebo arm than there were in the osimertinib arm. It’s hard for me to believe that that in and of itself won’t translate into some sort of OS, but as Heather mentioned, some of the other data from China and Japan shows the curves diverge. At least TKI vs chemotherapy, after the 2 or 3 years of TKI, those curves come together, so the question is the duration of osimertinib in the adjuvant setting. I’ll ask my third unfair question to Jonathan. You have a patient come to your office, stage IB tumor, it’s resected and has a 

 exon19, doesn’t want chemotherapy, just wants the [osimertinib]. Are you OK just giving the [osimertinib]? We know those patients that didn’t get chemotherapy in the ADAURA trial did benefit. Are you comfortable doing that?

: I have a discussion with the patient and I make it clear that adjuvant cisplatin-based chemotherapy, as of now, is the only treatment in the adjuvant setting that has an overall survival benefit. With a patient we are clear cut, no positive N1 stage II disease, stage III, I say there is no substitute for cisplatin- based chemotherapy. But even after having that discussion and emphasizing that, not every patient wants adjuvant cisplatin-based chemotherapy and in that situation, I would offer adjuvant osimertinib.

And 3 years of osimertinib? And then call it a day?

: For now, yeah. I would follow data—until there are updates, I would follow the trial data. I think this is going to be an evolving set of research questions that we’ll spend years sorting out, things like duration, the role of circulating tumor DNA, and looking at MRD [minimal residual disease]. That’s all going to take years to sort out, but for now, I’ll follow the trial data.

: I wish we’d have 1 year, 3 years is a long time. Osimertinib is a well-tolerated drug, but by year 2, it’s not easy. There are real adverse effects, and there’s a cost to society, right? This is $150,000, if not more, per year and it should be taken into account. Maybe it should be lifelong. People argue this all the time. It’s the amivantamab story for adjuvant GIST [gastrointestinal tumors]. Are you just kicking the can down the road and what should it be?

Advances in Testing and Targeted Therapy for NSCLC: Translating Evidence to Clinical Practice 

Adjuvant Osimertinib in Early-Stage NSCLC: The ADAURA Trial

Heather, what are your rules for adjuvant chemotherapy? What’s the size cutoff and how do you do this at an everyday practice?

: As Alex said, it’s definitely a shared decision-making discussion. The 4-cm number, for better or worse, is what we have. If someone has a 3.8 cm tumor that had some aggressive features to it, and they really want to do everything possible, I will offer them chemotherapy. But if it’s a 4.5cm tumor, no nodes involved and someone is older and somewhat frail, I probably won’t.In both cases,I will really talk about and definitely try to frame it in terms of how is this really going to impact the patient in terms of cure and so, I use that at about 5%— changing cure. Because when you start talking about hazard ratios you lose almost everybody, but if you can put it in terms of 5% difference in survival at 5 years, then people can weigh is that worth 3, 4 months of chemo and we can weigh, “Am I going to cause permanent kidney damage?” Which is going to impact them within 5 years, among other things. So, it’s kind of weighing all of it, just like we do with the radiation as well.

: I think it’s a great point. I think if there’s ever shared decision-making that needs to happen, it’s with the role of adjuvant chemotherapy for resected disease and sharing with them what the real benefit is in the context of toxicity. So, speaking of toxicity, Alex, cisplatin, carboplatin? The data— interestingly enough—that 4-cm cutoff was from a carboplatin-adjuvant study that was negative.

: But followed up with other subanalysis.

That’s it, so, most of the data we have prospectively is cisplatin based. Is there a role with carboplatin?

: So, off the record. On the record, I’m a cisplatin guy. I mean, the data is with cisplatin, the original IL data. I remember when it came outbecause I was actually writing a review with my old mentor, your institution, we had to change the review because it came out a month after, but it was all cisplatin-based therapy. Now, if I have an 80-year-old [female] patient, with a creatinine of 1.2 and a 4-cm tumor, I’m not giving it. But if I have an 80-year-old who’s got an N1-resected, or even N2-resected disease, I’m all about the carboplatin. I think you have to make these real judgments.

’ve heard a lot, and we’ll talk about the neoadjuvant data, a lot of carboplatin used in that neoadjuvant study, and some people have extrapolated, say “OK, it’s time to start using just adjuvant carboplatin.”

: You know, it’s interesting because, if you look from a purist standpoint, it should be cisplatin. That’s what it is. But all the neoadjuvant studies do allow you to use carboplatin,which, is surprising because, as part of a clinical trial, the gold standard should be the gold standard.

: I think in the neoadjuvant setting with the carboplatin, you can see what you’re doing. In the adjuvant setting, you’re flying blind, right? And that’s where that cisplatin, where, even in the metastatic setting, we’ve proven in multiple studies cisplatin’s better than carboplatin, from a survival benefit, but it’s small, and so we don’t think that toxicity difference is worth it. But in adjuvant, where you’re only given maybe a 5% and if you give carboplatinit’s maybe 3, and then, why even give it? So, it’s not that I haven’t done it sometimes [laughter] but, you know …

: But in that patient, that 80-year-old, with elevated, mildly elevated creatinine with a 4-cm tumor, if I can’t give cisplatin for a 4-cm tumor; I’m probably not going to give carboplatin.

: No, I agree. And talking, in a resected…

: It’s like… you guys asking me about sutures, but….

I’m talking about this as a pair in the neoadjuvant setting with IO [immunotherapy], isn’t the carboplatin arms all performing better, or is it just me?

: There aren’t any cisplatin arms in the neoadjuvant - the only randomized data we have really was CheckMate816 [NCT02998528], which one…

: It’s carboplatin-based. We don’t have cisplatin-based studies in the neoadjuvant setting yet. We will, but they’re not done yet.

I think the only time —you’re right, as a purist, I think adjuvant cisplatin is the way to go—there was some data from a SEER [surveillance, epidemiology, and end result] analysis a few years, maybe 5, 10 years back showing that carboplatin performed almost as well as cisplatin—and it is less toxic. On the record I don’t give it, off the record I will tell you there are times that I would give it to the right patient.

Shared insight on the role and selection of adjuvant chemotherapy for patients with early-stage non–small cell lung cancer.

Selecting Appropriate Adjuvant Chemotherapy for Early-Stage NSCLC 

We're going to move on now to the more exciting, well, curing patients. We are able to cure patients in the stage IV setting, but not as commonly as we are in the early stage. All these drugs are starting to move into the early stage, thankfully. We have targeted therapies that are now starting to move into early stage, of course, we've got immunotherapy, as Heather will talk about, moving into the early stage in the adjuvant setting and then the neoadjuvant explosion of data that's coming out with neoadjuvant chemo-IO [immunotherapy plus chemotherapy] is pretty remarkable. But, let's just start with every day clinic and the surgical approach for early-stage patients. Jessica, just how do you select these patients? What's the conversation for these patients that are potentially surgically resectable?

: In surgical patients, most stage I and stage II patients, by almost definition, have resectable disease, and so our decisions are all about physiology, and most of them—that's to be able to determine whether they're operative or not. That all kind of changes in stage III. In stage III, a lot of what determines whether someone is operable or not is the extent of nodal involvement. We've learned a lot over the past 10 years or so about using things like pneumonectomies in that population and trying to do a little bit of picking and choosing who we operate on based upon that. And I think that's really important - the physiologic assessment comes into play, because it is the one stage, stage III, where we know we have to give them systemic therapy also, and your surgery has to pair well with systemic therapy. And if not, we're not going to cure patients. So that's always been the dance of, can I operate, I want to operate, and I also want to make sure they get through their systemic therapy.

Let me ask you another unfair question, and we'll talk about neoadjuvant chemo IO, but, with the advent of durvalumab in the PACIFIC trial [NCT02125461] for stage III disease, and seeing the outcomes that we've seen at 4 years, does that alter your perception of some stage III patients, whether they should be resected or not now that we have chemoradiotherapy followed by durvalumab, with all these great outcomes. We looked at this in our own data set and it did make a dent in some of it.

: I think it does, I think it's a real difference, to say that we have a population that we're not operating on, that we're giving them curative therapy, and it's got a 40%, 5 year survival, that's really impressive. And sometimes you do have to convince patients, I'm not saying we're not treating you or not working to cure you, and it does set the bar high for surgeons. So, the one thing we have to assume now or have to guarantee a stage III patient is a negative margin. If I don't have a negative margin, I can't get a 40% survival, no matter how much systemic therapy I give them.

: So that really makes that a really close thing that I need to be careful of, so I do think the durvalumab changed that.

: I would just ask Jessica, do you think the pendulum is going to swing back, now that we have the CheckMate 816 data [NCT02998528]?

: We know now at least an event-free survival benefit to immunotherapy with overall survival trending in the right direction, but don't have to fully report it.

: I think surgeons always questioned who is unresectable in the PACIFIC trial. Surgeons were not included. There was no definition, so I think there were a lot of people who may have been surgical candidates in there, and I think some of them will come back. The hope is that we will see a higher overall survival and that will, again, keep surgery in that stage 3IIIA discussion.

We've looked into it; even at [John] Hopkins [School of Medicine, Baltimore, Maryland,] I think some. I mean, with the PACIFIC data that’s come out, I do think that, as you've mentioned, if you can't get a negative margin, there's no way you're going to get that type of outcome that we saw in PACIFIC, and so, this did alter our perception. It goes back to the multidisciplinary tumor board too, having a collective discussion about these patients, upfront to decide what to do. Stage IIIA is a heterogeneous disease too, and, depending on who's in the room at that tumor board and what flavor of coffee they have. It may make a big difference with how the surgeon could, or how we approach these patients. Chaitali, can you give us your take on how adjuvant therapy is used at your institution? We know that adjuvant therapy—platinum-based adjuvant therapy is approved for patients in early-stage disease who are resected, we had some data before Lung ART [NCT00410683] to suggest that consolidative radiation or postoperative radiation therapy was important, maybe with N2 disease, or a positive margin, that may have changed, but just your approach on your clinical practice and some of the newer data we have with Lung ART.

: Well, in clinical practice, I think, it's very much a disciplinary approach. You look at all factors, like what are the margins of N2, and sometimes a patient will tell you or the surgeon will tell you, “I'm not so confident about this, though the pathologist is,” so, I think the decisions are best made in a multidisciplinary manner. Adjuvant chemotherapy is pretty standard, it's not something that you will not do, so, that does follow the stage III disease, and the decision for radiation, concurrent or sequential, is based on this margin status and the high-risk features. The update from the Lung ART was that, now that the PORT [NCT00880971] is done in the N2-positive disease, would modern chemotherapy, and modern radiation; it does show almost a 15% improvement in disease-free survival [DFS] at 3 years. I mean it is little bit more toxic, the late cardiopulmonary toxicity was slightly higher, but one can't negate the DFS advantage of PORT.

And I'm not sure if it showed it in OS and I think that putting the question, the role of postoperative radiation therapy.

: I think we use it a lot less than we used to, although I do have to say, that was older radiation techniques and I'm sure there was a lot of cardiac dosing that went on there, which probably is what impacted the overall survival. But, I do think, also, at least, maybe only in my mind, I don't think that local recurrences and systemic recurrences are the same anymore. I think we can salvage local recurrences so much better than we ever could before that, I think a lot of us are willing to hold off on that radiation for everyone with N2 disease and wait and see what happens.

: Jessica, are there any N2 patients where you would consider it post surgery, like an extracapsular extension or something like that?

: Definitely for a positive margin, that has to be done. And I think if someone had a bad extracapsular extension, or you felt like you were leaving the highest node and things like that, that I would consider it, but it's many fewer patients than it used to be, like a postage stamp, everyone got it.

Shifting focus to early-stage NSCLC, expert panelists define the role of resection and consider when it is appropriate to utilize adjuvant radiation therapy.

Heather A. Wakelee

Articles

Adjuvant IO Therapy in Early-Stage NSCLC: The IMpower010 and PEARLS Trials

Adjuvant Osimertinib in Early-Stage NSCLC: The ADAURA Trial

Selecting Appropriate Adjuvant Chemotherapy for Early-Stage NSCLC

Treatment for Early-Stage NSCLC: Resection and Adjuvant Radiation Therapy