ASCO 2020: Latest developments in EGFR-mutated non-small cell lung cancer

Bookmark and Share
Published: 2 Jun 2020
Views: 4899
Prof Benjamin Besse and Dr Alex Spira

Prof Benjamin Besse (Institut Gustave Roussy, Villejuif, France) and Dr Alex Spira (Virginia Cancer Specialists Research Institute, Fairfax, USA) discuss the latest developments in EGFR-mutated non-small cell lung cancer (NSCLC) presented at ASCO 2020.

Prof Besse begins by outlining the results of ADARURA study, which investigated the use of osimertinib as an adjuvant therapy in patients with stage IB-IIIA EGFR mutation-positive (EGFRm) NSCLC after complete tumour resection. Dr Spira states that the disease-free survival (DFS) achieved in these patients is profound and potentially practice-changing, but acknowledges there is some debate surrounding the immaturity of the overall survival data.

The experts also highlight the latest results of CHRYSALLIS trial, which is an ongoing phase I trial investigating amivantamab, an anti-EGFR-MET bispecific antibody, in NSCLC patients with EGFR exon 20 mutations.

Prof Besse and Dr Spira conclude by discussing the outcomes in patients with advanced NSCLC and PDL-1 expression treated with immune checkpoint inhibitor monotherapy, along with the evolving role of liquid biopsies.

This programme has been supported by an unrestricted educational grant from Janssen Pharmaceutica (A Johnson & Johnson Company).

BB:         Welcome to ecancer. We have fresh news from ASCO and I’m very happy to discuss some of the main topics with Dr Spira, Virginia Cancer Specialists. I’m Benjamin Besse, I’m a thoracic oncologist at Gustave Roussy near Paris, France. We might start with the plenary session. Lung was at the plenary session this year with the adjuvant study ADAURA. As you know, in Caucasian patients around 10% have EGFR mutation and it’s around 50% in the Asian population. So, one of the big questions was could an EGFR TKI impact the outcome in the adjuvant setting when these patients had resected non-small cell lung cancer. So ADAURA was randomising three years of osimertinib versus placebo in 682 patients and the outcome is quite impressive but the primary endpoint is DFS in the stage 2 and stage 3. The hazard ratio was 0.17 which is very, very profound. There is a snapshot on the overall survival but it’s clearly too soon to have any data on overall survival. So the study is clearly positive regarding its primary endpoint and all the discussion around it is is DFS a relevant endpoint in the adjuvant setting or should we wait for OS because what we want to do in the adjuvant setting is cure more. What do you think, Alex?

AS:         That’s a great question, Ben, and I think everybody is going to be debating this until we get the final overall survival statistics. But clearly the question is what do we do now for our patients. The DFS was beyond impressive with a hazard ratio that may never have been seen before in the lung cancer scenario. I think there are some criticisms, they are mild criticisms, that the use of adjuvant chemotherapy was not 100% clear or stratified. That being said, the DFS was so profound and, given the trend to OS, in my mind and I think in most people’s minds, we’ll move those stage 2/3 patients resected to getting three years of adjuvant osimertinib. This is a tough population, not everybody who relapses will even be able to get therapy and certainly talking to patients about the fact that you’re preventing their disease from relapsing is an important outcome. We might be wrong and we’ve been wrong before and we all agree that overall survival should be the ultimate endpoint in an adjuvant study because ultimately that’s what it is. But I think these statistics are so profound we have to move forward with this. Some may agree, some may want to wait for more but overall, especially given the fact that it’s a relatively low toxicity drug, it is an oral drug, here in the United States the biggest concern will be financial toxicity because it’s not a cheap drug. There are also, obviously, other questions and do you really need three years, is two years? Do you need five? Do you need it indefinitely like we treat CML patients? They’re on TKIs lifelong. Obviously a different scenario so there are a lot more questions we’ll get some answers to, some we never will be and some will be debated, but I do think this is practice changing as of now.

BB:         And to put ADAURA into perspective , it was very interesting to see the update of the CTONG study. So it’s one adjuvant study in China that compared vinorelbine cisplatin to gefitinib and the fact is that the DFS at three years is positive. It was a primary endpoint and it’s only a study in stage 3 so it partially overlaps ADAURA but when you look at OS, because we have much more follow-up of this study, there is absolutely no sign of benefit in OS. This is something we have already seen if you take the metastatic setting and the OPTIMAL study, it was a TKI, a third generation TKI, against chemotherapy. I know this is the metastatic setting but just to give a perspective to the audience the hazard ratio was 0.16 for PFS so you really think it would impact the OS for these metastatic patients. In fact, the impact for OS was 0.21. It’s just that I agree with you that so far this is the data that we have but we really have to take a step back and wait for more results to see if those impact OS.

              Another exciting time for targets is exon 20 EGFR mutation. It’s a rare mutation, around 1% of all our patients, probably less than 10% of EGFR mutation, and so far the activity of the TKIs in this subgroup of patients was debated. A new drug, amivantamab, has a completely different mechanism of action. It’s a bispecific antibody against EGFR and MET and we have seen very interesting results from a phase II study in this subgroup of patients, so patients with exon 20 EGFR mutation, with a response rate of 36% and a PFS of 0.3 months in pretreated patients. Alex, what is your view on that study?

AS:         Amivantamab is an interesting molecule. As you pointed out, Ben, the EGFR exon 20 mutated space has become actually somewhat complicated. There are at least two or three drugs presented at ASCO this year that are all showing promise and this one is unique because it’s a monoclonal antibody and we haven’t really talked about monoclonals in the EGFR lung cancer mutated space, although there is an approval for colorectal cancer for a long time. It’s going to be very difficult to tease all these different drugs together but amivantamab has been a very exciting drug. It appears to have response rates that are higher than some of the other TKIs; it appears to be well tolerated. This is a very well-done study. For the first time we’re going to have drugs that target this molecule and I think that we all expect that within twelve months there’s probably going to be commercially available options. Given the relatively good tolerability, given the lack of other options, if you look at this study there was a clinical benefit rate that was 67% and up to 72% for patients who had platinum-based therapy, which is pretty remarkable for this area. Major toxicities, as expected for EGFR, are rash related which is a little bit of a throwback to the older EGFR molecules that we don’t use as much anymore. But I do think this is beyond promising, it’s going to be interesting to see how it plays out because it’s been a very complicated area with multiple different entries into this field. So we’re probably not going to know which one is better because obviously they’re not going to be compared head to head as they reach market. But it’s super-exciting and patients need to be found for these studies, we need to be testing for EGFR exon 20 insertion which is a big challenge. But as we do that and get these options it’s going to be a game changer for these patients.

BB:         And it should be reminded that we both co-authored the abstract and that there are other interesting drugs, particularly TKIs. There was an update on TAK-788 which is an oral compound with a response rate of 43%. Crizotinib was also updated with probably more toxicity and response rate a bit different so far. It would be very interesting to see how these different drugs will be either complementary or compete in this setting.

AS:         You know, Ben, in terms of novel drug development, I hate to say, it was almost the year of exon 20 because we had updated data on poziotinib. Poziotinib has been a little disappointing because the response rates haven’t kept up with where they initially were but we have updates on TAK-788. We have high dose osimertinib which actually was surprisingly, although in a small study, some interesting response rates, as well as the monoclonal antibody amivantamab. So it’s going to be a very complicated area and exciting. Again, another new, novel target.

BB:         It’s probably important to recall that for exon 20 mutation all the mutations are not similar compared to the classical activating mutation, exon 19 or exon 21, where it’s very homogeneous. Mutations in exon 20 are very heterogeneous and it might explain, the molecular analysis might explain why sometimes some drugs are more active than others. It could be very interesting to see if this is the case for amivantamab.

AS:         I think the only other things is we’re all so used to such high response rates to osimertinib in regular EGFR, I don’t want to say regular but in EGFR mutated, and across the board the exon 20 drugs are probably not going to be as good. They are still going to be great options, these drugs also have more side effects and we’ve gotten so lulled in to the fact that osimertinib is such a good drug. This is going to be a little bit different; exciting but different.

BB:         The last abstract we would like to discuss is a very interesting one from the FDA. It’s been quite new that the FDA is about to reanalyse their own dataset, mainly from randomised trials. They did that for ten trials that tested single agent immunotherapy. So they have compared in all these trials that could be either in first line or in second line how PD-L1 expression could impact the outcome in the subset of patients that had high PD-L1 expression. When we say high we say PD-L1 percentage in more than 50% of the tumour cells. It has been quite interesting to see that if you compare the outcome of the very, very high PD-L1 expression, meaning more than 90% of the tumour cells that are stained, the outcome is much better, either in terms of PFS or OS, first line or second line. As an example, the hazard ratio for PFS in first line is 0.78 when you compare the very high PD-L1 to the PD-L1 50-89%. So it might fuel the debate of what is optimal in first line, should we use always single agent immunotherapy, should we use immunotherapy plus chemotherapy or immunotherapy plus immunotherapy in the near future? What do you think Alex?

AS:         This is, as you said Ben, an interesting thought and it kind of echoes what we’ve known all along that while we’ve put PD-L scores in the <1%, 1-49%, >50% basket it’s a spectrum and not a cut-off. What this tells us is that the higher you are the more likely you are to respond to checkpoint inhibitor therapy. That ultimately has the implication what do you do in first line – do you give somebody triplet therapy? We can debate whether or not triplet therapy, given some of the new data as well, that’s a separate discussion, right? But whether or not, for example, you give somebody carbo/pen/pembro or to give pembro alone. I think there’s data here that the higher the PD-L you are, the closer you are to 100%, you’re probably more likely, and this is as expected, that you’re more likely to respond to single agent checkpoint inhibitor therapy. That obviously needs to be debated – are there patients that you don’t think are going to get the second line? Are the patients going to need a response, need to throw everything at that? But if you’re looking at somebody, maybe they’re older, maybe they can’t do chemo, maybe they’re asymptomatic and their PD-L score is very high, 90-100%, I think this can justify, you could have justified it before but now with more ammunition, whether or not you can give them single agent pembrolizumab therapy rather than triplet therapy. So I think it enhances some of that. How that applies to some of the newer data, obviously nivolumab is now maybe part of for some people, should be part of for a lot of people, first line therapy. I don’t think we have time to go into that right now but I think as you look at single agent checkpoint inhibitor therapy this is, again, more data to say that it’s okay to give, especially in very high PD-L patients.

BB:         I agree, we have to learn more about the phenotypes of these patients that could not derive a benefit from single agent monotherapy. Maybe high tumour burden, maybe aggressive disease, rapid doubling times, this is probably also a factor that perhaps we’ll analyse in the future too. And discriminate IO, IO-IO or IO-chemotherapy but exciting times to come. Alex, have you heard anything exciting about liquid biopsies?

AS:         I don’t know if I’ve heard anything exciting or new but I think the exciting thing at ASCO this year has been all the new targeted therapies. We have new c-MET drugs, we have new RET drugs, we have new EGFR drugs. I think the take home message has to be we’re making sure we’re checking all of our patients. Part of that is liquid biopsies. You may not be able to get a test, you may not have enough tissue, and I think there’s adequate data to say now, for a long time, that by whatever means necessary we should be interrogating our tumour for all the appropriate mutations. Liquid biopsies, which have really come a long way, are an integral part of that. You can argue whether or not they should be done on everybody or not, part of it is a timing issue. Somebody walks in, you want all your data so the patient can start on therapy and, of course, be less anxious, but I do think they have a clear role. There’s also a role now in patients at progression. If you look at some of the newer molecules we know that there are secondary resistance mutations and picking that up while you can do a second biopsy is also easily done on a liquid biopsy. So we’ve been talking about this for a long time and it continues to be that we need to keep looking for these actionable mutations, again by whatever means necessary.

BB:         I agree, liquid biopsy is now a completely accepted tool, it’s used to screen the molecular abnormality, to screen resistance. I think at ASCO there was some fine-tuning around could it be a predictive marker reflecting, for example, the tumour burden with the allele frequency? Is the dynamic evolution of ctDNA in the blood after a few cycles predictive of benefit? I think we will fine-tune this for the use of the liquid biopsy. With that, I would like to thank you for your attention, listening to this ecancer expert to expert discussion.