MK:        Hello and welcome to this ecancer discussion looking at EGFR exon 20 mutated non-small cell lung cancer and recent advances that have been presented this year at ASCO 2021 and implications for the treatment of our patients. My name is Dr Mathew Krebs, I’m a medical oncologist at The Christie Hospital and University of Manchester, United Kingdom. I work in early phase drug development with an interest in particular for lung cancer. I’m joined today by a great set of colleagues to discuss these recent data and I’ll just invite them to introduce themselves for the purpose of the meeting. So let me go round the room, Dr Viteri.

SV:         Hi, thank you very much to the organisers for inviting me and it’s a pleasure to meet these colleagues to discuss about this fascinating topic. My name is Dr Santiago Viteri, I am a medical oncologist devoted to lung cancer treatment and molecular oncology. I work in Barcelona at the Instituto Oncológico Dr Rosell led by Rafael Rosell, a great pioneer in the investigation of EGFR mutations. As I said, happy to be here with you.

MK:        Thank you. And Dr Gadgeel.

SG:         Thank you very much, Dr Krebs. Yes, I’m also very glad to be in this discussion. My name is Dr Shirish Gadgeel, I’m a medical oncologist and Chief of Division of Haematology and Oncology at Henry Ford Cancer Institute and the Ford Health System in Detroit, Michigan. Thank you.

MK:        And Dr Cho.

BCC:       Hi everyone, I’m very glad to discuss about new developments in EGFR exon 20 this year. I’m Byoung Chul Cho, lung medical oncologist and Chief of the Lung Cancer Centre, Yonsei Cancer Centre, South Korea.

MK:        Thank you, it’s really great to have you all here for this discussion. I’m really looking forward to discussing these data. It’s exciting times for EGFR exon 20 insertion, the management of these patients. There are a number of emerging exciting therapies coming through. Perhaps one of the first things is we’re beginning to learn a little bit around the natural history of these mutations – the prognostic impact of having an EGFR exon 20 insertion compared with other EGFR mutations. There have been a couple of real world data studies presented at ASCO this year and I’d perhaps just like to focus on those first of all, just to get a feel for what is the standard management that’s going on at the moment from those analyses and how that really compares with your own practice currently. So perhaps, Dr Viteri, can we start maybe with the abstract that was presented from France around some real world data of EGFR exon 20 insertion in the context of the first line treatment setting?

SV:         Yes. Of course, whenever we face these driver mutations that we don’t know very well yet the first thing we need to do is to try to best characterise them to know the population, what kind of patients, what is the real frequency in our settings etc. So this is the main aim of this abstract, this effort from the French investigators. They analysed the data from a data platform that collects information, retrospective data, from patients that includes more than almost 15,000 non-small cell lung cancer patients with the treatment information etc. They interrogated the database and found EGFR mutated patients and then in this group they detected those who had the exon 20 insertions. We know that the exon 20 insertion is the third mutation in frequency; first of all the frequent ones, the sensitive ones that we know, exon 19 deletion and then the point insertion in exon 21. The third mutation in frequency is the exon 20 insertions. It’s very important to know that there are many types of exon 20 insertions at different variance. This is more or less, what we know from previous series, around 5% to 30% of EGFR patients.

In fact, in this work they found 61 patients accounting for 3.9% and the characteristics of these patients, of course the number of patients is small but they found more female patients, smoking or non-smoking status was even and most of the patients were diagnosed with advanced disease. The age was not so young as the typical EGFR mutations that we are familiar with. The treatments, with regard to treatment, most of them were treated with platinum-based chemotherapy, some EGFR TKIs, the first or second or even third generation EGFR TKIs were used but the results were poor, which is something that we already know, with a short overall survival especially if we compare them with EGFR sensitive mutations.

MK:        Yes, the data are beginning to show that they perhaps behave similarly to a patient that doesn’t have any sort of driver alteration. But it’s interesting that a certain proportion of patients were treated with the standard tyrosine kinase inhibitors. I wonder, Dr Gadgeel, do you want to comment on that? Why do you think that is and what would be your own practice for these patients?

SG:         Yes, I believe among community oncologists the knowledge about exon 20 EGFR mutations has evolved. I suspect as initially there was so much attention to exon 19 and, as Dr Viteri mentioned, the point mutation in exon 21, L858R, which are highly sensitive to all the available EGFR TKIs, particularly to osimertinib. So there are occasions, particularly in the past, when oncologists may not have appreciated that exon 20 EGFR mutations are a distinct population and are tumours that usually are not responsive to the currently available EGFR TKIs. As was demonstrated in this database, the duration of treatment was extremely short with these EGFR TKIs, almost as if they had no benefit which is what is shown in preclinical studies and in the clinical database. There is a much greater appreciation now with all the data that is emerging for exon 20 that when EGFR mutation is identified that we understand what specific mutation exists in the patient’s tumour.

I also believe that there is a more practical aspect to why there is a greater recognition of exon 20. In the past several institutions across all countries did point mutation analysis so there would be specific mutation analysis for exon 19 and the point mutation exon 21. But as we have gone to multi-gene assessments, and particularly to NGS testing, we are much more frequently identifying these exon 20 insertion mutations. So all the more imperative that we are aware of these mutations and their unique clinical course and their lack of responses to EGFR TKIs.

MK:        Yes, I think that’s a really great point and there are differences also between the different panels that are used as to how many exon 20 insertions are actually analysed. It is worth just highlighting there’s one very specific exon 20 insertion that is highly sensitive to standard EGFR TKIs, I think it’s the A763-764 insertion which accounts for 4-5% of the exon 20 insertions overall. But, aside from that, as you say, they are not responsive to standard TKIs. There was also a second study presented with some real world data, this time looking at the use of amivantamab in the CHRYSALIS study compared with data pulled from three US databases in terms of standard of care practice in the second line setting. I wonder, Dr Cho, if you might comment on that study for us and tell us some of the highlights and pros and cons of these analyses.

BCC:       This is a real world dataset, so they compared CHRYSALIS, one dataset who were previously treated with platinum-based chemotherapy, a 1R patient population, with a pooled real world dataset of 125 patients. This is a retrospective analysis and there are several caveats because this is a retrospective analysis. But in this dataset the key message here is the patient with EGFR exon 20 insertion has a poor prognosis in terms of progression free survival and overall survival. So the patients who were treated with amivantamab compared to a pooled real world dataset showed longer progression free survival and overall survival in this dataset.

MK:        Perhaps you can comment on why these data are important to look at the real world comparisons rather than randomised controlled trials?

BCC:       Because EGFR exon 20 accounts for only 10% of all EGFR mutant NSCLC. It’s not that large a patient population like common EGFR mutations so it’s very hard to conduct a large-scale randomised phase III study in the second line setting. So given the recent approval of amivantamab in the post-platinum setting, it’s important to conduct efficacy of amivantamab in patients previously treated with combination chemotherapy with a real world dataset. So in this dataset it clearly suggests that amivantamab is an effective treatment option after progression on platinum-based combination chemotherapy in EGFR exon 20 insertion positive advanced NSCLC.

SG:         Pardon me, Dr Krebs, I was just going to add two points to the excellent points that Dr Cho made. One is that this data was quite relevant since amivantamab has received recent approval in the US for patients with exon 20 insertion EGFR mutation positive non-small cell lung cancer. But it is specified in patients post-platinum-based chemotherapy so this is a little bit unique in that several of our targeted agents have been approved irrespective of the line of therapy so this agent has been specifically approved only in patients who had received prior platinum therapy because that was the group of patients that were enrolled in CHRYSALIS. So it was important to define the outcomes of exon 20 insertion patients in the real world who had prior received platinum-based therapy. As was shown in the French data, this was more about patients that were identified, newly diagnosed and then how did they have an outcome. So to have specifically patients with second line therapy was the unique characteristic of this abstract.

The other thing I would add is that this may become one way of documenting benefits of targeted agents in groups of patients where they are quite uncommon, such as NTRK or others, because it is so challenging to do randomised studies. So these real world databases across the globe, not just in the US, are going to become extremely relevant in terms of defining the efficacy of these new agents.

BCC:       One more comment for this abstract – given the lack of effective treatment, particularly in patients after progression on platinum-based chemotherapy, in the real world setting there were various different treatment strategies including single agent chemotherapy, immunotherapy, tyrosine kinase inhibitor and other therapeutic options. Combined efficacy from these diverse treatment options, progression free survival in pooled real world data was only 2.9 months. So, given the recent approval of amivantamab and excellent efficacy data of these agents in post-platinum EGFR exon 20, these real world data clearly suggest that amivantamab should be a standard of care in post-platinum EGFR exon 20 insertion NSCLC.

MK:        Yes, and even despite the limitations of these real world data analyses they are important data. I agree, Dr Gadgeel, it’s quite refreshing that the regulatory authorities now are willing to look at and accept these data as an alternative to randomised controlled data where it just simply isn’t possible to obtain these rare populations. So NTRK is a great example of where we’ve seen that. Both amivantamab and some of the newer therapies which have also been presented at ASCO this year, and we’ll come on to just in a moment, tend to be in that post-platinum setting. I don’t know if anyone wants to comment around whether we should be looking at these therapies in the first line setting? I appreciate it tends to be the later stages we start and then bring it forward but perhaps Dr Viteri, do you want to comment?

SV:         Yes, if we think about what happened with common EGFR mutations at the beginning we had the same dilemma. Platinum-based chemotherapy has been our backbone and standard treatment for many kinds of patients during decades and it’s difficult to get rid of them, these drugs. We know that they are not specific, we know that they are not the solution of the disease, but we know that they are helpful. So we shouldn’t easily just forget them and run to the new drug in the block but the truth is that whenever we have been able to identify a driver mutation, and we have been able to design a drug that specifically targets this driver, the results have been outstanding and very easily overcome the benefits of regular chemotherapy, not only in the efficacy setting, also in the safety or in the side effects that usually are much better tolerated for patients. If you ask patients you know what will be the answer – most of them prefer anything but chemotherapy.

But the truth is that, for example, in the amivantamab example that we are discussing now there is a clinical trial ongoing, PAPILLON, which is first line exon 20 EGFR comparing platinum-based chemotherapy against amivantamab plus chemotherapy which is trying to determine if the addition of the two agents or the two approaches can be feasible for these kinds of patients.

MK:        Yes, that will be a very important trial. So just on this theme of tolerability and thinking about some of the other agents coming through, there are a few oral agents we’ve seen presented. We’ve had mobocertinib and TAS6417 and another compound as well presented which we can come to. But in terms of all these, and also throwing amivantamab into the mix, we have to weigh up how efficacious are these drugs, what are the tolerability profiles etc. So just with all that in mind and thinking about how on earth are we going to navigate exon 20 emerging therapies in years to come if they’re all approved, perhaps we can focus for a moment on some of these data. Dr Gadgeel, do you want to talk us through the mobocertinib data that was presented in platinum pre-treated patients and the EXCLAIM expansion cohort?

SG:         Yes, in this particular trial we have seen the data from mobocertinib in the past but this was an update combining two specific populations from the phase I/II trial of mobocertinib. There were 114 patients in the prior platinum cohort and then there was a more broader cohort called the EXCLAIM cohort of about 96 patients. The update showed both responses as assessed by investigators as well as independent review. The independent review response rates were somewhat more modest at about 25-28% whereas the investigator assessed responses were in the range of 32-35%. The median progression free survival holds up at about 7.3 months, which is what was reported in prior presentations of this database. The median overall survival was 24 months in this patient population. There was efficacy observed with this drug across all exon 20 insertions so benefit was not restricted to specific mutations.

The limitation continues to be the toxicity that is observed with many of these agents. The primary limitation of many of these EGFR TKIs is that the difference that we see in inhibiting mutant EGFR with regards to exon 19 and exon 21 L858R mutation as opposed to wildtype EGFR inhibition is somewhat much more narrow when it comes to exon 20 in that the doses required to inhibit exon 20 also end up inhibiting wildtype EGFR to some extent. Therefore that lack of selectivity leads to toxicities that I suspect limit the tolerability of these agents, particularly in terms of diarrhoea, stomatitis and rash.

One other aspect observed in this database was CNS was identified as a common site of progressive disease in this patient population. So, unlike the efficacy of osimertinib against CNS metastases in patients with sensitive mutations, there may be some limitation in terms of CNS efficacy with a drug like mobocertinib. Whether that is because of lack of ability of the drug to cross the blood-brain barrier or it is just that the toxicities limit the ability to dose the drug at an effective dose is unclear. The dose reduction rate was not very high, 20-25%, and the dose discontinuation rate due to AE was not very high. Nonetheless I do believe that the toxicities limit the ability to tolerate these drugs.

MK:        Yes, all really fantastic points. I agree about the CNS penetration, there does seem to be a difference compared with the standard TKIs that we’ve seen. Let’s just talk about some of the other data that were presented as well. Theoretically a more selective inhibitor is DZD9008. Perhaps, Dr Cho, do you want to tell us the data from that study? It’s relatively early days.

BCC:       DZD is also a potent selective EGFR exon 20 insertion inhibitor showing potent activity in in vitro models and in vivo models. In a phase I study they had a dose escalation and dose expansion cohort. They escalated those from 50mg qd to 400mg qd and in the dose expansion cohort they had two cohorts of 300mg qd and 200mg qd. The demographic of the patients was typical of EGFR exon 20 insertion in general and had a safety set of slightly over 100 patients, an efficacy set of a 56 patient cohort. DZD had generally well-tolerated safety issues  and had a grade 3 toxicity of 6.3% in 200mg qd and about 30% of grade 3 or higher adverse events in 300mg qd, suggesting that severe toxicity was correlated with the dose. The most common adverse events, as expected, were diarrhoea, rash and several GI toxicities. The vast majority of toxicities were grade 1 or 2 and grade 3 toxicity was not common with the DZD compound.

Anti-tumour activity of the DZD compound showed around 40% in the whole 53 efficacy set and efficacy was shown regardless of EGFR exon 20 insertion genotype and showed above 90% disease control rate. A response rate was shown across all dose levels but the confirmed objective response rate was reported at about 45% in the 200mg qd and about 40% in the 300mg qd dose. The duration of response and median progression free survival is still immature and we need to further follow-up with this compound.

MK:        That’s great, thank you. Just before we open it up to a bit further discussion I wonder if, Dr Viteri, you could tell us a little bit around the TAS6417 compound as well?

SV:         As we discussed, there are several compounds targeting the EGFR exon 20 insertions. This compound that formerly was known as CLN-081 is an oral EGFR TKI with broad activity against all kinds of mutations including the exon 20 insertion. It is supposed to have an attenuated activity against wildtype EGFR compared to the exon 20 insertion, making this potentially with the best safety profile, avoiding some of the side effects that we think are related with the wildtype inhibition, allowing us probably to use higher doses.

At ASCO were presented the interim results of a multi-centre phase I and II clinical trial particularly in advanced EGFR exon 20 patients. By that time of the communication of the results there were 37 patients that had been treated, most of them have received previous lines of therapy. They tested doses of 30mg up to 150mg twice a day and the most common side effects were, as expected, rash and diarrhoea and other typical side effects of EGFR inhibition such as stomatitis or dry skin. Grade 3 were not so common – 3% of diarrhoea, 5% of anaemia – so it seems that the drug was quite tolerated at this point. Regarding the efficacy, what we can see is that among the 25 patients that were evaluable, 10 had partial responses, meaning 40%, and 14 had stable disease, accounting for almost 96% of disease control rate which is quite promising.

MK:        Thank you, some really interesting compounds coming through. Amivantamab data were presented at World Lung this year with a response rate of around about 40%, relatively low grade 3 toxicity and a PFS of just over 8 months, I think. But obviously it’s an IV administration rather than an oral administration so, perhaps just to open this out, what are your thoughts around this emerging landscape and where next? Amivantamab is already FDA approved but I’m interested for your thoughts about where things are going here.

SG:         To start off with talking about amivantamab which is approved, as you mentioned. 40% response rate and also what appears to be a progression free survival, with a note of caution that this is a cross-trial comparison, that appears a tiny bit better than some of the PFS we have observed with TKIs, exceeding 8 months. Amivantamab does have a unique adverse effect of infusion reaction which almost appears to be universal with the first dose and one needs to be mindful of that but the good news is that it doesn’t appear to be the case after the second dose. But one needs to be mindful about this infusion reaction and measures to mitigate that.

The other comment I would make is I really believe now we have several TKIs that are very efficacious, however, I do have a question about their ability to selectively inhibit exon 20 as opposed to wildtype EGFR. The consistent theme appears to be that we see a response rate, at least in the initial results, of 35-40% but as we expand the cohort or increase the number of patients that are enrolled, particularly at the phase II dose, or the recommended phase II dose, we are seeing that the efficacy drops down to maybe less than 30%. I suspect that is primarily because as you enrol more patients, particularly at the recommended phase II dose, we are seeing a little more wildtype EGFR activity resulting in toxicities that are somewhat challenging to manage. That’s a bigger limiting factor and at least in the datasets that I have seen, including of DZD9008 and the CLN-081 data, it looks comparable to the initial results that were reported with mobocertinib and poziotinib. It remains to be seen as we enrol more patients whether we will see that selectivity. That is emerging as a challenging factor in developing particularly TKIs for this group of patients. The hope is with DZD9008 in particular that there will be selectivity but again the data to me seems not too dissimilar from the initial reports of poziotinib and mobocertinib. As more data came out with those two drugs the efficacy did appear a little more modest than the initial reports. So let’s see what happens with DZD9008.

MK:        Yes, I agree. Any comments from Dr Viteri or Dr Cho?

BCC:       I totally agree, as Dr Gadgeel already mentioned for, for example, DZD9008. PK data suggest that the higher dose may be more beneficial in covering phospho-EGFR of cells carrying EGFR exon 20 insertion. But in the clinical setting a higher dose produced a higher incidence of grade 3 or worse EGFR inhibition related toxicity such as skin rash and diarrhoea so they could not go up to the higher doses so they selected 200mg qd and 300mg qd as the potential recommended phase II doses. So the tolerability is a big issue with small molecule TKIs with DZD and the CLN-081 compound too. But, interestingly enough, in one or two patients in each dataset these small molecules had a response after progression on amivantamab suggesting that these two agents may not have overlapping resistance mechanisms. So, at the end of the day, if we have several more TKIs in our clinic we may face an issue of sequencing of these agents for our patients.

MK:        Yes, indeed.

SV:         Yes, I totally agree with Dr Cho. The good news is to have different drugs with different mechanisms of action. We don’t know how important it could be that amivantamab also inhibits MET, this is a common resistance mechanism in EGFR mutations, at least in common ones. The sequence can be interesting and also the potential of the combination that also has been tested in ongoing trials.

MK:        Yes. On that note, actually, and just in the last couple of minutes there are a couple of trial in progress reports as well. One was around a subcutaneous administration of amivantamab rather than IV and one was a combination with lazertinib in various different settings. I wonder perhaps, would you like to comment, Dr Viteri, on those two abstracts as well?

SV:         Yes, particularly with amivantamab one of the challenges is that the administration is intravenous, as Dr Gadgeel mentioned, and the infusion reaction is one particular toxicity that is very specific to this drug. It requires some training to deal with it because it can be somehow scary for your patients and the doctors that are on their floor at the moment but it is usually manageable and it only occurs with the first dose. The other thing is that at least with the designed trials the administration is weekly during the first month and then every two weeks which is a little more uncomfortable and problematic for the organisation of our hospitals compared to orally taken drugs. So the intention to test the subcutaneous administration could help to resolve some of these problems. We don’t know if the infusion reactions will be lower, it’s suspected maybe, but also the time of administration to make it easier for the patients and teams to administer the drug. That’s very appealing for the near future. Also there is the possibility that new schedules of treatment would be tested with the drug, not every week or every two weeks.

Also, as we were speaking before, the combination with EGFR TKIs. EGFR TKIs have different mechanisms of action and you could potentially block the EGFR receptor in the extracellular domain with the antibody and in the tyrosine kinase domain enhancing the efficacy. The problem here could be the safety or the feasibility of this combination. We know that some of the newer EGFR TKIs have some problems because if we need to target the exon 20 we need potency that is difficult to manage. Also we don’t have to forget that we are also inhibiting MET so the overlapping of the MET toxicities plus the EGFR toxicities can be a problem and we are eager to know the results of this lazertinib amivantamab combination.

MK:        Okay, thank you. I think we could continue to discuss this for quite some time but we’ve run out of time. This has been a really great discussion. I’d just like to give great thanks to all three of you, Dr Viteri, Dr Gadgeel and Dr Cho. Thank you for your time, thank you for your interaction and thank you also to ecancer. Thank you. Bye bye.