SP: Hello everyone, my name is Sanjay Popat, I’m a medical oncologist in London. Welcome to this ecancer expert discussion. Today’s topic is about EGFR exon 20 insertion mutant patients with advanced non-small cell lung cancer. We’ve all just returned from the virtual World Lung Cancer Congress armed with the latest data. I’m joined by my expert colleagues here, Dr Michael Thomas from Heidelberg, my colleague Marina Garassino from Milan, my colleague Rosario Campelo from Galicia and also my colleague Enriqueta Felip from Barcelona. I hope we’ve got plenty to discuss.
The first thing is about these EGFR exon 20 insertions – how will we actually find these in Europe? For the UK NGS is really not very common, it has only just started and many centres are using PCR-based techniques, and I worry that we’re not picking up these patients enough. Michael, how are things in Germany?
MT: Well, we have a broad scope on EGFR alterations. It’s very important, as you already mentioned, to have an in-depth analysis and here NGS is the best way to approach that. So, saying that, in our institution we’ve employed that already for several years and it’s broadly spread over Germany in a network which is quality assured. So this provides a good opportunity to get familiar with EGFR alterations and in particular exon 20. Here, as you mentioned, the exon 20 insertion is an important one which accounts for around 50-60% of exon 20 alterations.
SP: Thank you and what’s the situation in Italy, Marina?
MG: We are in a very jeopardised situation because there are academic centres doing NGS to all patients but there are small centres around the country where they just do the test for available drugs. So there is a situation and the number, also, of NGS platforms in Italy, they are not enough to cover all the landscape.
SP: I think we have this problem everywhere in Europe. And Spain, Enriqueta, how is the landscape of genotyping for EGFR?
EF: The situation is pretty similar, so there are a number of institutions that, yes, we are doing NGS but in general NGS is not really highly implemented for patients with non-small cell lung cancers. So there are a lot of patients that are tested with PCR-based technologies. Even in some institutions we are doing both – we are starting with EGFR, ALK by immunohistochemistry roles and PD-L1 and then for those patients that are negative we perform NGS. And we have the results but sometimes not just before starting treatment.
SP: Thank you. Rosario, we had data at the World Lung Congress from a couple of datasets, from Flatiron and AACR Project GENIE. Do you want to comment on how common these EGFR exon 20 insertions are?
RC: Thank you so much for the courtesy. I feel that these presentations have been very useful and they showed the increasing interest we have in this kind of genetic alteration. We have new drugs, we have very appealing drugs, very active drugs, so there is this interest in finding these alterations. What we have seen is that NGS can detect with more frequency these alterations, even more than fifty times compared to conventional PCR. Also we can detect different variants from the genetic alteration. It’s going to be important, probably, in the near future how these different variants can predict the efficacy of a specific targeted agent. So very interesting data, real world data, from the World Conference in Lung Cancer, yes.
SP: Thank you. Michael, at the moment many tests just simply say exon 20 insertion positive or negative or mutation present or absent. Is that adequate or do we really need to know the genotype to best treat these patients? What are your thoughts?
MT: For the time being of course it would be important to really receive the information – is there an exon 20 insertion, yes or not, because treatment options are rising at the horizon. When looking in the future, and Rosario already mentioned that, we get to know the molecular landscape on the distribution of the different alterations. For instance, we could differentiate between those ones which are in the loop following the C-helix, or those ones in the C-helix. Of course this is a preponderance for loop following C-helix, this accounts for around 90% or more than 90% of alterations and here the conventional TKIs do not work. Conventional TKIs might work sometimes in certain defined alterations in the C-helix so it’s important to know it for these reasons. In the future there might be even the opportunity to sequence treatments or to target a certain alteration with a certain type of treatment like TKI or antibody. So, for the future it might matter to know that landscape and for the time being it’s very important to get known at all is there an exon 20 insertion yes or no.
SP: Thank you. My question then is to Marina: how are these patients currently being treated? In the UK if they are identified they generally get platinum doublet chemotherapy, platinum pemetrexed; the combination with pembrolizumab or immuno-chemotherapy is generally not given because these are excluded from the 189 reimbursement. But in the UK we do have access to the IMpower150 regime up front and some of these patients are being treated with that. What is the landscape in Italy at the moment?
MG: The situation is very similar to the UK but we are allowed to use the KEYNOTE-189 regimen that potentially excludes just exon 19 and exon 20. What we know is that in general patients with driver mutations do not respond very well to immunotherapy but, again, we do not have data for this particular population with exon 20. Today the scenario is evolving; I think that we do not have an answer yet because we have multiple new drugs giving a response rate but, again, the response rate is in the range of 40-50%. So I think that in the near future we will have to clarify if to start with chemo, IO or with the targeted agents. I believe also that the safety of the new drugs, the PFS of the new drugs together with the response rate, will clarify in the future what will be the best strategy.
SP: Thanks Marina, I’ll ask Enriqueta next, we had some exciting data with these new drugs. One of the drugs that we’ve all been looking forward to seeing some mature data is mobocertinib, previously known as TAK-788. Dr Zhou presented data from the EXCLAIM ongoing trial. Do you want just to summarise what some of that data showed, Enriqueta?
EF: I think the data looks very good. So in this situation, post-platinum chemotherapy, the agent was associated with a response rate of approximately 30%, 36% according to the independent review committee, 35% according to the investigator, and a median progression free survival of around 7 months. So this is something that is changing because our experience with TKIs in these patients with exon 20 insertions is really limited with PFS with approved drugs of only 2 or 2.3 months. So I think really good information in these patients with a difficult alteration to target.
SP: One of the things that I noticed in the presentation, there’s quite a high adverse event rate. I think I saw about a 20% grade 3 diarrhoea rate. That reminds me a lot of our experience with afatinib – we’ve seen that sort of rate before. Rosario, what are your thoughts on the adverse events that we saw with mobocertinib?
RC: This is one of the most important aspects of this new agent – how are we going to manage the toxicity adverse events, since the activity since vibramycin. I am really worried about these grade 3 adverse events because it can negatively impact in terms of quality of life. Maybe lower doses, we have to check how this works with a lower dose and how we can implement management of this toxicity but it can be a challenge, in my experience it has been a challenge to treat these patients. We have to take into account that these patients have been treated with many lines of therapy previously so maybe the situation, the clinical situation, is not as smooth as the first line setting. But, again, I think that toxicity is going to be a challenge with these new agents.
SP: Thank you. I don’t think I saw any data by genotype, Michael. We’d previously discussed the specific genotype and perhaps maybe a difference in binding of the kinase and that might result in a heterogeneity of responses. We saw a median PFS of about 7.3 months but the duration of response was 17.5 months. To me, that implied that if you responded you responded very well but overall there might be some heterogeneity. Do you think I’m reading too much into the data and it’s a statistical quirk or do you think there is some heterogeneity in outcomes?
MT: I think you nicely picked the piece of the data which gives consideration to think further forward. This is the difference between the mobocertinib trial, the duration of response and if you compare it to amivantamab, for instance, there we have a readout which is not so long. Which gives a little bit the impression, okay, what are those responders and which types of insertions are responding to mobocertinib, we don’t know yet, it has not been analysed to my knowledge. But what we have seen, this is the reason why I mentioned it, we have in the C-helix some alterations which you can really pinpoint which is, for instance, A763. This one is nicely responding to the conventional TKIs, this is the only one, and the rest not. Knowing that, there might be a difference. So it’s quite encouraging to step further forward with this perspective that you just mentioned.
SP: We’re very much looking forward to some more data being presented in due course. Similarly, as you discussed, we had this other new dataset from the CHRYSALIS study presented by my colleague, Dr Sabari. This is using the EGFR MET bispecific drug amivantamab with one arm activating EGFR, the other arm MET. Marina, we heard the extension cohort efficacy data with, again, a similar response rate of 40%, duration of response just under a year, progression free survival of 8.3 months. What is your take on that data? How do you think that will fit in our paradigm?
MG: Yes, so we have now multiple possibilities because we have poziotinib, we have mobocertinib, we have amivantamab. So the results are interesting and 8.3 is the best progression free survival that we had. The toxicity profile seems also better compared to the other compounds. So clearly we will have to see more data, we will have to understand when to use this bispecific before, in the beginning of the story of the disease or in the second part, but the progression free survival together with the better toxicity profile can be something very important for our patients. I agree with Michael that we have to create also education around the family of exon 20 because they are not all the same. So maybe also in the near future we will see some particular subgroup that benefit more and they will go immediately with the drug. So we are accumulating data in the process.
SP: We’re living in exciting times, we have new drugs coming through. With amivantamab, Enriqueta, we have some unusual toxicities. We have the EGFR effects with skin toxicities, we also have the MET affects with pedal edema but also this drug has infusion reactions and it has to be currently given intravenously every two weeks. So what are your thoughts on the adverse events, bearing in mind the other drug which is also targeting the same population, whilst oral, has issues with diarrhoea as its main problem?
EF: I have to say that the infusion reaction is something that we have to face but in our experience and what was presented is that they are really mild and these reactions are not impacting on the treatment. So my experience with amivantamab, and I have participated in the trial, is that the agent is well tolerated in general. I have not seen a lot of anti-MET toxicity in our patients and, yes, there is some kind of rash but as a medical oncologist I am treating with other EGFR TKIs, we are used to treating this toxicity. So infusion reactions but not really important and will not impact the treatment outcome in the patient.
SP: Great, thank you. Rosario, we had an update on the ZENITH20 trial, this is a poziotinib trial. Poziotinib is another EGFR kinase inhibitor. I would say that the data of the once daily dosing that we’ve seen so far have lower response rates. Here we had about a 15% response rate reported and a lesser progression free survival – 4.2 months. It’s also quite a toxic drug with a high rate of grade 3 diarrhoea, about 30%. How do you think this drug, poziotinib, is comparing with the other options that are coming through – mobocertinib and amivantamab?
RC: With the data we have so far I think the main problem we have with poziotinib is the less efficacy in terms of response rate and PFS, as you have mentioned. But overall I would say that the toxicity profile is the main problem we have with this agent. We have to wait for the results testing new doses but so far I would say that amivantamab or mobocertinib can do better than poziotinib.
SP: Thanks. Michael, one of the problems these patients have as the natural history of disease progresses is brain metastasis. We see this quite frequently in patients with EGFR mutation. I worry that with amivantamab being an antibody we’re not going to get much brain penetration. With TAK-788, mobocertinib, we had lower responses in patients with CNS metastasis, we haven’t seen any brain specific data. Is this a vulnerability across all of these drugs that we’re going to have to be concerned about or do you think this will just settle down in the end?
MT: Of course this is an important point. As we know, at the key point of diagnosis we get around 20% of patients with brain mets in stage 4 disease. Over time, if you follow over the disease trajectory, it might increase up to 40%. We have learnt from the EXCLAIM data that have been shown, as you mentioned, that in patients with brain mets that there are responses, something around 18-20% in this subgroup with amivantamab. Nothing in this direction has been shown, it’s difficult because it’s an antibody and we know that this might not penetrate the blood-brain barrier. So this is something where we would have to establish management strategies. Personally I would perceive, as we know from other molecularly altered types of disease there might be a lot of brain mets but not so large. Here we even start to employ stereotactic radiotherapy even up to 8-10 brain mets if the size is less than 2cm. So brain management matters; mobocertinib shows some activity and it’s important to keep this in mind in the future.
SP: Thank you very much. So if we’re thinking about where we’re going in this field, Marina, at the moment all of these datasets have been in patients who have progressed post-platinum. We’re seeing some first line trials with these newer agents ongoing. We have the PAPILLON trial in the first line setting of chemotherapy, carboplatin, pemetrexed with amivantamab. And also with mobocertinib we have a first line trial of mobocertinib versus chemotherapy. Do you think these drugs are really suited for the first line setting or we need to optimise their function better in the second line setting? What are your thoughts?
MG: Yes, this is a good question because we know that they do not have a 70-80% response rate like for exon 19 and exon 21. So really we would like to increase the number of responses in these patients. So maybe one drug could be not enough and this is why are so important the combination trials. Maybe in the future when we will understand better the biology behind the exon 20, maybe I foresee that there will be a group with the single agent and maybe another group with a combination with chemo or something like that. So in the near future maybe we will be able to personalise also the rarer family.
SP: Thank you. Enriqueta, one of the options that my colleagues in the US seem to use quite a lot for these exon 20 insertions is high dose osimertinib off-label. Zofia Piotrowska presented a cohort of data at ASCO with about a 25% response rate. Obviously we’re used to using osimertinib, we like the adverse event profile, how does that stack up against this data? Do you think that’s really viable? We have a different regulatory and reimbursement environment in Europe, where do you think that is going to fit in?
EF: Yes, really it is that, at least in Spain, we are not able to give osimertinib at normal dose in this indication. I have to say that even the study presented, the results are modest so we need other treatment approaches in this group of patients. Yes, I think 25% you mentioned but PFS is short, so I think there is a need for new treatment strategies. We are not usually treating these patients in Spain and in Europe with osimertinib at a higher dose.
SP: So I think things will settle down over the course of the next few months as we await more data. Certainly we have these ongoing first line trials with mobocertinib and amivantamab and I, for one, am very interested to see how these come out. Do we think we are going to get any better patient selection to allow us to use one drug over the other? Michael, will we be doing NGS at baseline, as you’ve so well organised in Germany, and perhaps preferentially selecting those with MET amplification for amivantamab over mobocertinib? What’s your vision of the future?
MT: A good remark. I think the first line trials could help to elucidate in this direction and here it would be really good to have some paraffin embedded blocks in collection and do this analysis even potentially with co-mutations. What might be interesting, in the PAPILLON trial of course amivantamab, as Marina mentioned, is tested in conjunction with chemotherapy versus chemotherapy on its own. The question is, is it the type of molecular alteration or is it a certain prerequisite of the microenvironment already given up front? So it could be of interest to look on the molecular phenotype in conjunction with NanoString based analysis to have some proxies on the cellular environment. But this is grasped wildly in the future but it would be really good to have in depth analysis in the first line trials.
SP: Thank you. Marina, at the moment these drugs and these trials are fairly much staying in the post-platinum setting at the moment; this is where we expect there will be access in the near future. We know with osimertinib we have to be careful if patients are pre-primed with checkpoint inhibitors because of osimertinib triggering IO toxicities. Do you think we need to be worried about that with mobocertinib? I presume we don’t need to worry about that with amivantamab because it’s a monoclonal but what are your thoughts on that?
MG: So the answer is that we don’t know very well. But, anyway, combining immunotherapy with TKIs can be always dangerous. So at least the period of washout could be important and we have always to remember that with the driver mutations immunotherapy does not work very, very well. So I think that it will balance between the story of the patient, the possibility to have a washout and then again we will have an increase of data. But never combine at least poziotinib with immunotherapy because we saw in the past disasters.
SP: We stay away from that combination. Enriqueta, at the moment without access to these drugs some physicians are trying second generation EGFR kinase inhibitors like afatinib or dacomitinib for specific genotypes. Do you think that that practice will still stay embedded or that will disappear once we have these drugs available?
EF: I think if we don’t have any clinical trial at least in Spain what happens is that the patients are treated with first line chemotherapy, no Rosario? In some places chemotherapy plus immunotherapy but not in all patients. If in second line we don’t have any clinical trials or we don’t have access to any compassionate use, we can try with afatinib but we know that the response rate is less than 10%. So if we have available the possibility to treat the patient post-platinum with compassionate use with mobocertinib or amivantamab I think this could be the treatment of choice in our patients.
SP: And at the moment, Rosario, if we’re seeing a patient with metastatic disease, they have high PD-L1, 100%, should these patients be starting off with first line pembrolizumab? Is there data to support that?
RC: I don’t think we have enough data to support that. We have more respective data so in the benefit of treating these patients with conventional platinum-based chemotherapy and the data are not so bad with pemetrexed combinations. So, for me, it would be the best way to treat these patients in the first line setting in case I don’t have any clinical trial. So far I prefer to be cautious regarding the use of immunotherapy in this population. I like much more the idea of trying the antiangiogenic agents in this patient population. It could be an option but let’s see when we have more data before starting to do it in our daily clinical practice.
SP: Great, thank you very much. With that I’d like to thank my colleagues for joining me on this discussion. It’s been a fantastic discussion about this rare type of mutation, who these patients are, how we’re currently treating them and where we’re going with the latest data. I’d like to thank all of you for joining us for this excellent discussion. Thank you all very much.
