KK: Hello everyone and welcome to this ecancer discussion which will be a quite broad conversation with my medical oncologist colleagues on the subject of EGFR and EGFR-targeted therapies and some new developments in that area, including some things that are featured in abstracts here at the World Lung Cancer Congress in Vienna. So our first topic, really, is a fairly general one considering new treatments that are available in the EGFR mutation space, something that we’ve been quite familiar with over quite a period of time but there are now some newer drugs and newer approaches that have come along. These feature amivantamab, which is a different type of therapy and approach in this area, and lazertinib, which is a new generation, third generation, tyrosine kinase inhibitor. So there are a couple of abstracts in the programme looking at this combination in first line but also in later line therapy. I wonder if you could give me your perspective on the use of this combination in general but thinking about those two abstracts. Can I start with you Melina?

MM: Sure, absolutely. The first abstract to discuss is the combination of amivantamab and lazertinib in the frontline setting. So that is an update on some previous data and really they still actually weren’t able to measure the median progression free survival or the median duration of response after a median follow-up of about 22 months. So showing really that people are having durable responses, are responding to this combination therapy. Of course in the EGFR space people do have long, durable responses and it will be seen if compared to something like osimertinib in the first line that this is a viable approach, this combination. 

AT: That’s the interesting thing about studying this combination in the first line, that it’s a space where we already have an established, very strong therapy. So what this data is looking to do is to see can we do better in terms of the number of responses, and I had to read the abstract a couple of times with really 100% of patients responding to the treatment with a relatively small number of patients in the trial at this point but very, very high levels of response which is something that we don’t see very often. We always see some patients who don’t respond. Also the duration of response, we’re going to keep following these patients, we’re going to see what the activity is in the central nervous system. Interestingly, the co-occurrence of p53 mutations is documented in the population and those patients also seem to be responding very well. So it’s definitely an interesting approach to keep looking at.

KK: I guess the co-occurrence of the p53 mutations is what you would expect but, yes, the fact that they are responding perhaps as well, small numbers, is a hopeful sign. Jordi, what’s your take on this?

JRM: I think that it’s interesting to have this combination because you try to prevent one of the major mechanisms of resistance, that is the MET amplification. However, for instance, it’s really promising the results but I’m not completely sure that treating up-front with everything in combination is really better than the sequential treatment approach. I think that it’s important also in this trial or in this abstract is the role of the circulating tumour DNA to try to select the patients that really need to escalate, or perhaps de-escalate, the treatment. This is a more clever, from my point of view, strategy than perhaps not giving a more toxic combination for all the patients. Perhaps it will be more interesting in the future to try to adapt these kinds of strategies according to the escalate, if in the DNA there is not a clearance, or de-escalate, something like that. I think it’s more interesting. It’s true that it’s promising because we are blocking one of the major mechanisms of bypass after a third generation alone but I would really like to know if this really impacts the outcome. It’s true that the duration of response is longer, it seems longer, than osimertinib alone because after 22 months there are no patients with a real progression. But, as you said, it’s a phase I but always when we combine it’s really sure that we are really improving the outcome of patients and balancing with the toxicity. Probably in the future, as my colleague said, it’s more important to select the patients for the best treatment up front and perhaps not all EGFR mutations are the same and this is important to try to select the patients up front.

KK: You mentioned about adverse events, what was your feeling about the adverse events that were reported with this combination?

JRM: Always patients with oncogenic driven tumours, they are really happy to have just oral treatments and they want to delay the intravenous treatment and go to the hospitals. Amivantamab, during the first month you should apply this drug weekly and then every two weeks so it means that you should have more visits. Also there is the infusion reaction and it’s important and you should have some training to really react if you have some kind of reaction. So I think that there will be a curve of learning to manage this drug and perhaps there will be an initial increase in the toxicity based on this infusion reaction. Also it will be really important to have the patient’s perspective – do you want to have this treatment that perhaps gives you more response but similar progression free survival and you should go every two weeks to the hospital? This is the same when we combine TKI with chemotherapy or TKI with antiangiogenic agents. I think it’s something important also, the patient’s perspective in these decisions.

KK: One of the issues that we have to consider when we are thinking about patients particularly who have relapsed on osimertinib is whether newer treatments such as amivantamab and lazertinib may be combined with chemotherapy. There was an abstract here at WCLC that considered that particular combination with quite impressive results. Would you like to give us your perspective on that particular piece of work?

MM: The combination of amivantamab, lazertinib, carboplatin and pemetrexed is given to patients after disease progression on a TKI. It could have been a first or second generation TKI or osimertinib. They found an impressive 50% overall response rate. I think that it’s interesting, it certainly brings up some of the issues of whether to sequence these therapies or whether to give them all together, knowing that 50% of patients post-osimertinib don’t have a resistance mechanism identified. It is an interesting approach.

KK: Anyone else have any comment on that one?

JRM: I think that’s something to explore in the future – what patients can benefit from escalating this treatment perhaps. It will be a more dynamic approach – start with amivantamab lazertinib and then, based on ctDNA, perhaps add the chemotherapy for those that do not respond. But I think it’s a good strategy also to combine.

KK: So we’ve considered the combination in first line, there was also that abstract with lazertinib by itself online. Were you struck in any way by that abstract and what it might point to?

MM: Lazertinib on its own, certainly we expect it to perform in a similar way to osimertinib and I think it did continue to do that. There still wasn’t a lot of follow-up in that abstract so that data is something we’ll look forward to seeing. 

KK: It looked as expected, I suppose you could put it that way. What about the combination in later line therapy, the other abstract that was featured here? How do you see this fitting in to the routine treatment of patients with EGFR mutations where certainly osimertinib has a large chunk of, let’s say, the market? But we need options for our patients after the almost inevitable failure takes place with osimertinib. How do you view this combination of amivantamab and lazertinib in that setting? Amanda?

AT: It’s a very interesting option for patients because patients who have done well on targeted therapies often want to stay on targeted therapies. There is always the option of switching to chemotherapy or to a multi-drug chemotherapy/IO or VEGF combination. But at least the responses that I’ve been getting from patients are very strong, that their preference is to stay on a targeted therapy. When we think about the biology of the disease it’s very appealing to keep working in a targeted way against progression. If we look at the response rate there’s going to be a bit of a learning curve because I don’t have the exact number but I think it was around 30% overall response rate which is a strong signal but it’s not enough for patients who are really progressing rapidly. Our practice in osimertinib monotherapy for a long time has been to really try to keep patients on drug for as long as we can, even if it means treatment beyond progression and even if it means local therapy of oligometastatic disease. So oftentimes we don’t have a lot of time to give a treatment that might not result in response for 70% of patients. So we’ll have to think about maybe positioning this treatment at an earlier point in progression and not when the liver is really in need of a response.

KK: So I suppose it’s inevitable I’m going to ask this question of you as a pathologist and thinking about the response rate as not being as high. On the one hand we know that response rates, generally speaking, in later lines of therapy are never as high as in first line with targeted agents. On the other hand, we are giving two targeted agents in a population where we at least know one of the targets but we don’t necessarily know why those patients failed. One of those drugs has action against MET but we don’t actually know whether the patients who do respond are patients who have MET amplification as a resistance mechanism or not. So I think that does ask some questions. This is an obvious, perhaps, question but should we not be looking for these things? Should we not be trying to target these combinations a bit more appropriately?

AT: I think we should be. I don’t want to make it sound like 28%, I think it was, is a bad overall response rate. It’s an excellent overall response rate and it’s coupled with a very high stable disease rate. So disease control is very, very good; the waterfall plots in even late line amivantamab lazertinib look very, very good. What I’m getting at is our practice up until this point where we haven’t had very many non-chemotherapy options, or any non-chemotherapy options, for these patients has been to really keep mono-TKI treatment going for as long as we can and then switch to chemotherapy. I think the more good options we have after mono-TKI treatment, the more ready we have to be to switch at a time point where the disease is maybe not progressing as rapidly as it had been in our previous practice. I don’t know, what do you think?

JRM: I think that if we accept that oncogenic drivers should be treated with a precision approach based on oncogenic drivers, the same concept should be continued at the time of progressive disease. We know that from the ALK-positive population that have sequential TKIs, because they are different next generations that block previous mechanisms of resistance, are good or are better. Also, in this combination of lazertinib amivantamab it was really interesting in ESMO last year, or in ASCO, I don’t remember, presented data in the population enriched for a MET amplification. The response rate reached 47% and the median progression free survival was seven months. For me, and you as a pathologist, it was really amazing that there was just an immunohistochemistry score and in the patients with an immunohistochemistry score according to the EGFR and MET the response rate was 90% and median progression free survival of one year. It gives us an idea that probably a personalised treatment approach based on a genomic profile is better than the standard chemotherapy. But in second line this is not already proven after osimertinib because all the trials or clinical trials comparing a personalised approach blocking MET versus chemotherapy are pending. Again, another question – should we treat these patients as an agnostic approval with chemotherapy or a personalised approach blinded or based on the person. I think it’s really important and challenging questions today. I completely agree with my colleagues but I think that as a concept it would be better if we applied targeted therapies based on a molecular target, otherwise it’s an agnostic approach and perhaps 50% of patients respond or that only will have toxicity with any benefit. So it’s no good for patients. 

KK: I do find it hard to believe, perhaps, that the biology should be so very different in the context of relapsed disease. Sure, there are some changes in the tumours but the principle of the target being there and it being targeted by the drug and we know very well from the very earliest data that if the target is not there the patients really do not do well at all on these targeted drugs. So why that principle should not happen in the relapsed setting I’m struggling to understand.

JRM: It’s really difficult to take a biopsy and then looking at biopsies is more difficult also for MET amplification. You know, your preclinical practice…

KK: It is but I’m going to be a little bit provocative here, it’s always difficult to take a biopsy.

JRM: I understand.

KK: And if you actually look at least at the reported data, and of course real world is sometimes different from reported, the biopsy success rate reported in the relapsed setting is about 70-80%. But if you actually look in real-world data in walk-in clinic type performance, that’s about the successful biopsy rate in our patients at first line. So I’m not sure how much more difficult it really is but then I’m not the person having to do the biopsy and face the patient so I do accept…

JRM: I think it’s not a difficulty, it’s that despite we have results we don’t have any of these treatments approved. So we only have the opportunity to enrol these patients in a clinical trial. So I think it’s really important to have post-osimertinib or post-lazertinib, post-furmonertinib the biopsy or liquid biopsy to know the mechanisms of acquired resistance and improve our knowledge. But, based on that, you don’t have any standard treatment approved, at least in Europe, for this population. So some physicians perhaps say, ‘Okay, I’ll start with chemotherapy.’ It’s not the feasibility to perform the biopsies, what do you do with these results?

KK: But if there were data and there were approvals, because we all know that ultimately approvals are the things that drive practice, do you think that we would be able to achieve that repeat biomarker assessment? Clearly it is easier for the patient to do the liquid biopsy but we also know that MET amplification is actually not well assessed by liquid biopsy. So would you expect, that may be the better word, would you expect the data to drive a change in perception and make post- or relapsed biopsy more acceptable?

MM: I think there’s mounting evidence that we should be doing tissue biopsies after disease progression on osimertinib, first of all looking for the 10% of patients that have transformation to small cell which you can’t get on a liquid biopsy, and then looking for MET amplification either by FISH or IHC. Certainly there is more data now for IHC in terms of response to amivantamab and lazertinib. So that is my practice, actually, I do try to get a tissue biopsy and a liquid biopsy in every patient that has disease progression post-osimertinib.

KK: Interesting. So perhaps we should now move on to a more specific part of the EGFR story. We’ve been discussing new treatments or what to do, how to introduce new treatments in the broader group or, perhaps more accurately, in the group of so-called common mutations because some of these drugs are not approved in some of the rarer mutations. But a particularly rare, or not so rare actually, but another group of EGFR mutations, a large and complex family of exon 20 insertions, has risen in our consciousness in the last couple of years as a result of new therapy being available for a group of mutations which hitherto were largely, although not exclusively, considered to be resistant to the currently available drugs. So let’s start our discussion about exon 20 insertions with a fairly obvious introductory question – do you see these patients in your population? Do you get reports of exon 20 insertion mutations? Do they cross your desk?

MM: I think that’s a great question because we know that if you’re doing hotspot testing with PCR you’re not going to find these patients so it reinforces the need, really, for next generation sequencing. So because we’re doing that broadly in all patients diagnosed with non-squamous non-small cell lung cancer we do see these patients now. But I think that’s a huge problem identifying them up front. 

KK: Indeed. I think that has probably been one of the biggest messages early on. I suppose it’s a truism across the board – you only find things that you look for. We do need to have this broader perspective to find these mutations. I suppose the imperative previously was not quite there because we didn’t really have anything we could do for those patients beyond standard chemotherapy. But now things have moved on and it is a bit different. Amanda, what’s your take on these in terms of how they come to you? They are reported to you, I assume, you’re seeing these patients?

AT: Definitely. So they are part of our NGS panel and, in Munich at least, we have a two-tiered approach to molecular testing. We always order NGS testing up-front at the time of first diagnosis but we know that we have a lag time of about three weeks until we get results. So parallel to initial NGS testing we do PCR hotspot testing for the common EGFR mutations and for ALK fusion we do FISH and immunohistochemistry as a screening for a FISH so that by the end of week one I have the common EGFR mutations and I have ALK testing. Then for patients with high disease burden I can go ahead and if those are both wildtypes start them on chemoimmunotherapy. So very often in the third or fourth week after the first biopsy I’ll get my full NGS report and then I’ll see the result of the exon 20 analysis. But, like my colleagues said, you only find the things you look for and if you don’t look for this you won’t find it. There are some other alterations that you look for and you don’t find. Personally I have found one NTRK fusion and I know people who have found no NTRKs.

KK: Absolutely. 

AT: So there are a lot of things out there you won’t find if you look for them but this is something you will find if you look for it, at least 1-2%, probably a little bit more than that.

KK: Actually when you do the math, as they say, they are a relatively common group. They’re the next common group of EGFR mutations after exon 19 deletions and L858R. If you break that down into your overall group they are on a par with ROS1 fusions and RET fusions, so a significant group of cases. So, Jordi, I guess also you’re seeing these patients?

JRM: Yes, the major issue is that in Spain next generation sequencing is not a standard of care and I can only have these patients if I can apply the next generation sequencing. I think this is a major limitation but this is another discussion, not today, but access to this platform to really genotype all the patients with lung adenocarcinoma to search for this kind of rare, not so rare as ROS and BRAF, and to give the opportunity to give personalised approaches.

KK: So we had an abstract here at WCLC describing the early access programme experience in exon 20 insertion mutations in Spain. Do you have any comment to make on that? Were you involved in this at all?

JRM: I was not directly involved in this access programme but I think this is really relevant to have real-world data of the efficacy because then you can see that probably the efficacy is lower because patients are not hyper-selected. Also the risk of discontinuation or toxicity is slightly higher than reported in clinical trials because it means that each centre that has patients can apply the drug. So it’s some kind of curve of learning to manage this toxicity that might be induced, that if you don’t manage properly the patients would say, ‘I don’t want to continue with this drug.’ So I think it’s important to have this data because it reflects that the drug works perhaps differently to the clinical trial, as obvious, but also that we need to learn to manage this drug. I think this is, for me, the most important message of this abstract.

KK: Do either of you two have any experience of using this drug in this particular setting?

MM: Amivantamab? Yes, we do have experience using it. I would agree that the toxicity, in particular the infusion related reaction, is a little bit scary at first. So learning to manage it, splitting the dose up between the first and the second day, knowing that it’s really usually not life-threatening or serious when managed appropriately with supportive measures, is really important but that comes with comfort and experience. So it’s definitely something that’s going to take some getting used to and will probably really only happen when it’s approved in the post-osimertinib setting when physicians are able to use it more broadly.

KK: Given your concerns, do you feel that it still looks more favourable than chemotherapy as an alternative or do you think it’s going to be something that you would definitely look like using?

MM: One of the interesting abstracts at World Lung was looking at the EGFR exon 20 patients and their response to different therapies. The thing that was notable to me about that is that certainly they have some response to chemotherapy but not a great response to chemotherapy; they have very little response to immunotherapy monotherapy. So having appropriate targeted therapies in this population is really important. I also think it’s interesting that we group them as a whole population because the mutations are quite different – they’re a heterogeneous group even within EGFR exon 20. So that’s something we’ll have to grapple with.

KK: So there was an abstract on that subject, describing the variations in the mutations. This really quite complex family of mutations sometimes gets broken down into the C-helix and then the loop following the helix. There have been statements that the C-helix mutations are actually responsive to our currently available TKIs, although those data appear to me, at least, to be a little bit I’m not so sure about it. Any of you want to comment – do you think that you might be nuancing your decisions of how to treat these exon 20 insertion patients according to the type of mutation or do you think you will treat them all in the same way? You can almost guess what the answer might be here but do you want to expand on that?

AT: They are interesting questions to ask and they’re really hard questions to actually answer with evidence in such small groups of patients. So it’s something that I think really shows that the smaller the group, the more international and the more broad in terms of regions we need to draw on to really group patients in cohorts and look at them. That’s something that around ALK we’ve been getting better and better at doing, where we have seven, in Germany at least, approved TKIs for the ALK setting and we know about different variants and we know about different types of fusion and variant 1 and variant 3 and with and without p53 and we start asking very, very detailed questions about which of these TKIs is better in which setting. I don’t think we’re quite there yet in the exon 20 situation and the way to get there is really by documenting large cohorts and looking at what people… Not waiting to start making our decisions until we know the answer because we don’t know the answer right now but rather to start treating our patients and registering them in cohorts and documenting them sharing that data and then looking back and saying on the basis of real-world data which patients benefitted from what.

KK: Because I guess individually we are going to see relatively few patients, relatively speaking.

AT: Absolutely.

KK: And then if we consider the number of mutations in that family it’s going to take a long time or a lot of patients before we really understand the granularity of it all. So international databases and things like this, registrations, whatever, will help greatly in trying to figure that out. So there are clearly a lot of things happening in the EGFR world still where we thought perhaps we’d solved all the problems, or most of them. Clearly there are new things happening and that’s good for our patients for sure. So I think we’ve had a very interesting discussion about a topic that has exercised us for a long time and I’d like to thank you all very much indeed for engaging and having a great discussion. Thanks a lot and enjoy the rest of the conference.

MM: Thank you.