ASCO 2024: Lung cancer highlights
Dr Antonio Passaro – European Institute of Oncology, Milan, Italy
Dr Matthew Krebs – The Christie NHS Foundation Trust, Manchester, UK
Prof Raffaele Califano – The Christie NHS Foundation Trust, Manchester, UK
Prof Nicolas Girard – Institut Curie, Paris, France
Dr Anna Minchom – Institute of Cancer Research, London, UK
AP: Hello to everyone and welcome to this ecancer webinar reporting on all the data presented during the ASCO meeting focussing on lung cancer highlights. I’m very delighted today to co-chair this webinar with Dr Matthew Krebs from The Christie Hospital, Manchester, and with a very exciting panel of colleagues. Today we’ll discuss a lot of the abstracts starting with three different sessions: the first with Professor Nicolas Girard from the Institut Curie in Paris discussing EGFR inhibitors, where we are moving to; the second presentation with Professor Raffaele Califano from The Christie Hospital Foundation, Manchester, UK; and the third with Dr Anna Minchom from London Royal Marsden Institute of Cancer Research discussing updates on non-genomic driven lung cancer. Then we’ll be able to discuss all of these kinds of analyses with an interesting open discussion. So in this slide you can see all the three main presentations with our guests and then at the end we have an open discussion. So I welcome you and ask all of you to send us your questions via chat box or news@ecancer.org and we’ll discuss in a [??] with all our guests of today. Now it’s my pleasure to introduce Professor Girard to start with the first presentation – EGFR inhibitors, where are we moving to? Thank you.
EGFR inhibitors – where are we moving to?
NG: Thank you for the nice introduction. We have seen during the ASCO 2024 meeting many data regarding the management of patients with non-small cell lung cancer and common EGFR mutations and also uncommon mutations. But many data in the metastatic setting and very interesting data in locally advanced unresectable disease from the LAURA trial. But let’s start with metastatic disease and we have seen updates from the MARIPOSA trial which compared amivantamab plus lazertinib to osimertinib in patients with metastatic non-small cell lung cancer, common EGFR mutations and were treatment naïve. Those patients were randomised to receive amivantamab plus lazertinib or osimertinib and we also had this lazertinib control group. The primary endpoint was progression free survival and this was achieved, demonstrating the benefit of amivantamab plus lazertinib versus osimertinib.
This was presented last year at the ESMO meeting, moving from a median PFS of 17 months with osimertinib, which is in line with the landmark data, to 23.7 months with amivantamab plus lazertinib. The key question now is what is the best patient group where amivantamab/lazertinib is working versus osimertinib. We are in a setting where we have many options now in the first line setting – osimertinib as single agent, osimertinib plus chemotherapy, the FLAURA2 trial, and amivantamab plus lazertinib based on the MARIPOSA trial.
So the data that have been presented during the meeting looked at those patients with poor prognosis, meaning patients with liver metastases, this is one of the key subgroups in this exploratory analysis. We know that liver metastasis is associated with a worse outcome in those patients and we looked in the MARIPOSA trial what were the outcomes with amivantamab plus lazertinib versus osimertinib alone. You can see on the screen this data. Osimertinib is doing 11 months median in patients with baseline liver metastases; amivantamab plus lazertinib is doing 18 months. This is a hazard ratio of 0.58. We have also data in patients without baseline liver metastases and you can see that those patients are actually doing better – median OS with osimertinib 18 months and with amivantamab plus lazertinib 24 months, a hazard ratio of 0.74. So we have a benefit with amivantamab plus lazertinib in patients with and without baseline liver metastases. So this is very interesting data showing that the magnitude of benefit is actually similar in those two groups.
Another group of interest are those patients with positive ctDNA at baseline. So many ctDNA analyses have been done on the MARIPOSA trial. This allows first to identify obviously the primary mutations, common EGFR mutations – exon 19 or L858R – but also co-mutations, especially in the TP53 gene which are known to be associated with a worse outcome with EGFR TKI. We have a similar proportion of patients with TP53 co-mutation in the amivantamab plus lazertinib and the osimertinib arms.
The second point with ctDNA is that is allows a kind of quantification detection of ctDNA at baseline which is obviously associated with a higher tumour burden, more metastasis and is also known as being associated with lower PFS and OS in those patients.
So if we look at this data with patients with TP53 co-mutation it shows probably that those patients with a TP53 co-mutation have a higher frequency of brain metastases, liver metastases, so clearly suggesting that these patients have more aggressive disease. Again, looking at the efficacy of amivantamab plus lazertinib versus osimertinib in patients with TP53 co-mutation and without such co-mutation you can see that the magnitude of benefit is again pretty similar, showing that those patients with or without co-mutation, with or without liver metastases are benefitting equally from amivantamab plus lazertinib versus osimertinib.
Looking at those patients with detectable EGFR mutation at baseline, it’s about 70% of those patients have metastatic disease. So it’s expected that we have a high proportion of patients with positive ctDNA. Again, it is associated with what we see in the clinic, this sense of aggressive disease, multiple metastases, liver metastases and it is more frequent in patients with detectable ctDNA. Again, in those patients with detectable baseline ctDNA we have a significant benefit with amivantamab plus lazertinib versus osimertinib, which is also true in patients without detectable ctDNA at baseline.
So finally the third pattern with ctDNA is to look at the clearance of ctDNA along with treatment, especially in the first weeks. Again, patients without cleared EGFR mutant ctDNA at week 9 probably suggesting more aggressive disease, primary resistance to treatment, are doing better with amivantamab plus lazertinib versus osimertinib. Look at those patients without ctDNA clearance at week 9 with osimertinib, it’s only 9 months. We know that in the clinic we have some patients which have a very limited duration of efficacy of osimertinib.
So this is a summary of the data that have been presented. Very nice data showing the benefit of amivantamab plus lazertinib in all these subgroups which is of interest, obviously, suggesting that those patients with aggressive disease, predicted poor outcomes, are actually doing better with the combination. What is also of interest is that there is no subgroup where you do not find the benefit, meaning that it’s very hard to select one group or another that would benefit also more from osimertinib. It’s not the case, the benefit with amivantamab plus lazertinib is observed in all those subgroups. These are the conclusions of this presentation.
So, moving forward again with amivantamab plus lazertinib, obviously this is a kind of escalation in the intensity of the treatment, we are combining a TKI, lazertinib, plus amivantamab which is historically administered intravenously. So this is something, obviously, that is more complex for the patients. There is a development of subcutaneous amivantamab in this setting, first line, based on the PALOMA-2 study, so a multi-cohort study. PALOMA-2 enrolled some patients with first-line amivantamab plus lazertinib with amivantamab being delivered subcutaneously, so every two weeks, combined with lazertinib. As you may know, a higher risk of venous thromboembolism has been observed in MARIPOSA-2 with the combination of amivantamab plus lazertinib, so prophylactic anti-coagulation has been recommended for the first 3-4 months of treatment. This was implemented in one of these cohorts of patients receiving amivantamab subcutaneous plus lazertinib in the first-line setting for common EGFR mutations. So exactly the same population as MARIPOSA.
These are the patient characteristics, kind of expected. What is interesting in this study is that we have a similar pharmacokinetic as expected based on historical data with intravenous amivantamab and we have a response rate to amivantamab plus lazertinib that is, again, in line with what was observed in MARIPOSA with intravenous amivantamab, suggesting that subcutaneous amivantamab obviously has a similar efficacy as intravenous amivantamab. These are the waterfall plots, again expected – 75% response rate – and many ongoing responses from these patients for whom we expect PFS to be something like a median of two years, as in MARIPOSA.
In terms of safety, quite interesting to see that prophylactic anticoagulation, as recommended, is actually reducing by half the risk of VTE so it’s working as expected but it’s good to know. This is part now of the recommendations with the use of amivantamab plus lazertinib. What is also of interest in this cohort is that we have probably a lower rate of severe cutaneous toxicity which is probably in line with the better proactive management of these side effects by the investigators. Finally, with subcutaneous amivantamab we do not have those frequent infusion-related reactions that were reported with amivantamab intravenously. It’s probably a good way to reduce that risk.
In line with those data from the phase II PALOMA-2 trial in the first line setting, we have seen the data from the randomised study that compared amivantamab plus lazertinib with amivantamab subcutaneous versus amivantamab intravenous. A phase III trial, 418 patients. Here we are not exactly in the MARIPOSA setting, we are more like in the late, late line setting, so patients who had an exposure to the TKI and then chemotherapy and receiving the combination in the third-line setting, similar to what was done in the historical phase I/II study CHRYSALIS. So that is the phase III comparing the pharmacokinetics of subcutaneous amivantamab versus intravenous amivantamab. [??] patients in this setting and, as you can see, the primary endpoint was met, meaning that there is no inferior exposure to amivantamab with a subcutaneous formulation versus an intravenous formulation. So this is a demonstration of the equivalence between the two formulations.
We have here some data regarding efficacy in the late-line setting, a partial response of 30%, which is quite interesting, with a duration of response of 11 months, even higher with a subcutaneous formulation which is, to me, interesting, even if it’s exploratory. The same thing for subcutaneous amivantamab in terms of PFS versus intravenous amivantamab. It’s not the primary endpoint, it’s a numeric difference, but interesting to see that it’s working quite well in this setting.
With subcutaneous amivantamab we have a reduction in the risk of infusion related reactions. The global sense is that of better safety. Also in this study 80% of the patients had this prophylactic anticoagulation, leading to a reduction in the risk of VTE events, so it’s not part of the clinical practice with amivantamab to administer this prophylactic anticoagulation for the first month of treatment. Subcutaneous is obviously something that is more convenient for the hospital teams and the patient because it’s very short and it can be done very easily in the hospital or even at home at some point.
Finally, the fourth trial that I wanted to discuss tonight post-ASCO are those results in EGFR mutant non-small cell lung cancer, but not the metastatic setting, the locally advanced setting. In those patients the standard of care for unresectable locally advanced non-small cell lung cancer is chemoradiotherapy. Usually in wildtype tumours consolidation with durvalumab is administered for one year, this is the PACIFIC regimen and it’s now standard of care, given the benefit of immunotherapy as consolidation after CRT. But in patients with EGFR mutant non-small cell lung cancer it doesn’t work; durvalumab is not bringing any kind of benefit. We have data from the PACIFIC trial subgroup analysis, we have also some retrospective data suggesting that there is no benefit with durvalumab consolidation in those patients with EGFR mutant non-small cell lung cancer after CRT.
So LAURA is a very different strategy. In those patients with locally advanced unresectable non-small cell lung cancer, common EGFR mutation, after CRT those patients were randomised to receive osimertinib consolidation versus placebo. 216 patients. Very interesting to see that here the treatment with osimertinib consolidation is not delivered for a predetermined duration; it’s continuous administration of osimertinib until toxicity or disease progression. The primary endpoint was PFS and it was met in this trial with a huge benefit with osimertinib versus placebo. You can see the median – 39 months versus 5.6 months – a hazard ratio of 0.16. So this is a major difference and a major benefit for those patients. It’s mostly based on the protection against CNS disease which is quite frequent in those patients. In this trial brain imaging was done on a regular basis, probably explaining the poor performance in terms of PFS of the placebo arm.
But osimertinib is working to protect against recurrences; the benefit is observed across all the prespecified subgroups and it’s also increasing probably more rapid responses after CRT which, to me, is difficult to interpret, given the effect of radiation in those patients.
So a benefit in terms of PFS, a reduction in the risk of CNS disease and, to me, this is a major endpoint in the clinic. We do not have overall survival yet in this disease, in this trial, this is something that we can discuss with the panel. To me it’s still changing practice because protecting the vast majority of patients from progression, protecting against CNS disease, is something that is needed in the clinic for those patients. We have seen a similar pattern with ADAURA in resectable non-small cell lung cancer, EGFR positive, with three years of adjuvant osimertinib.
In terms of side effects, no clear increase in radiation pneumonitis, a higher rate overall but a limited number of grade 3 or more side effects. Then, obviously, the expected side effects from osimertinib in those patients. So, to me, and I agree with the conclusions of the authors, osimertinib is probably the new standard of care for these patients in the consolidation setting after CRT.
So this is what I wanted to share with you, some additional data from MARIPOSA showing the benefit of amivantamab/lazertinib in all these patients with EGFR mutant non-small cell lung cancer, even those with poor prognoses; subcutaneous amivantamab, probably changing the practice and facilitating the integration of amivantamab in the strategy; and then the LAURA trial with consolidation osimertinib. Thank you.
Panel discussion
AP: Thank you very much Nicolas for this very incredible presentation considering the weight of the data that were presented during ASCO. I’m very delighted to open this first part of discussion co-chaired with Dr Matthew Krebs, discussing with all of you this data and also understanding how we can translate these kinds of results while waiting for the approval in clinical practice. So I have immediately a question for you, Nicolas, so you are very involved in the trial in the EGFR setting. Of course, the presentation of the subgroup analysis by Enriqueta Felip, so a high-risk patient population in MARIPOSA, show very interesting data. So this was the first big trial, more than 1,000 patients, with a predefined subgroup analysis focused on a high-risk patient population. So do you think that this kind of analysis would be a key factor also in understanding and balancing the future clinical approval and selection, also balancing the FLAURA2?
NG: That’s a very good question. Probably FLAURA2 is even more aggressive in terms of the escalation process for intensification of the treatment of those patients because FLAURA2 has chemotherapy and obviously it’s probably a higher burden. With amivantamab plus lazertinib the burden is mostly cutaneous toxicities and we need to have a better learning of how to proactively manage those side effects. There is a learning curve for sure with amivantamab plus lazertinib. It’s a chemo-free regimen which allows to deliver chemotherapy later in the strategy. Probably, to me, it has to become standard of care for the first-line treatment of those patients.
Probably what we have seen from FLAURA2, we have not seen so far data on liver metastases, ctDNA and so on, those subgroups but we have seen some data in CNS disease. In FLAURA2 those patients probably also had some kind of radiation at baseline. We have seen complete responses in 70% of the patients with chemotherapy plus osimertinib so clearly we have very good efficacy, intracranial efficacy, of chemotherapy plus osimertinib in those patients. 40% of the patients enrolled in all those trials had CNS disease at baseline. So we need to obviously see more data with amivantamab plus lazertinib in terms of intracranial response and intracranial efficacy for sure, but clearly it’s a matter of escalation in the treatment. Still some patients will receive osimertinib, for example elderly patients, patients with poor general status and comorbidities. Many patients should be eligible to the MARIPOSA regimen and probably some patients will receive FLAURA2, especially those patients with very aggressive disease.
AP: Thank you Nicolas.
MK: Nicolas, you’ve covered a whole spectrum of possibilities there. I think it’s quite remarkable that we’ve gone from single agent osimertinib to now these studies showing really, potentially, very beneficial combinations. I think it’s not entirely clear yet how we’ll differentiate. As you say, the brain metastases may give some guidance. What do you think about the overall survival data? Do we need to wait for overall survival? Are there any overall survival updates yet for the MARIPOSA regimen?
NG: I think it’s good to have overall survival. We have a trend towards a benefit in FLAURA2, the same thing in MARIPOSA. So at the end it will be very interesting to see how the PFS benefit translates into an OS benefit. But the fact is that beyond first line things are moving pretty quickly and we have new options in the second-line setting – MARPOSA-2, chemotherapy plus amivantamab. We have obviously chemotherapy possibly associated with new agents such as this bispecific PD-1/VEGFR antibody, the ivonescimab. We have seen data from the HARMONi-A trial at ASCO. Patritumab deruxtecan, we are expecting the data for this trial that compared patritumab, an ADC targeting HER3, versus chemotherapy. We have also all the strategies which are focussing on one mechanism of resistance such as MET activation with some agents such as savolitinib being assessed in patients with MET activation on the re-biopsy after osimertinib.
MK: Just going to the LAURA data, we’ve got a couple of questions coming in but just before we come to that, it looks pretty promising to use adjuvant osimertinib for that patient group. How would you see the first line metastatic treatment for those patients once they progress?
NG: That’s a very good question, it was the same question in PACIFIC actually. Here in LAURA we are randomising patients after CRT so we don’t really know that much about CRT by itself. We know it could be sequential or concomitant but this trial started after chemoRT. I’ve seen many tweets after the presentation looking at what was the quality of this chemoradiotherapy and how patients were selected. That’s a good question, for sure, and obviously the recommendation is to have a PET CT and brain imaging before starting CRT to make sure that the patient is not a metastatic patient. But still, I don’t think that so many patients with occult metastatic disease or metastatic disease that was not diagnosed have been enrolled in LAURA because if a patient has liver metastases and you miss the liver metastases or you miss the CNS disease, you will start chemoRT but it will have a limited effect on metastasis at the end. So I would expect these patients actually not to be enrolled in the trial at the time of assessment for the enrolment because I do not expect to have complete responses on CNS metastasis or liver metastases in those patients. Those patients are not enrolled in LAURA.
Obviously the control arm is performing very bad, 5 months, but I guess this is related to the fact that brain MRI were done every eight weeks in this study which was not the case in PACIFIC where the median PFS for patients with EGFR mutant non-small cell lung cancer was something like nine months. Here it’s only five months but we have those systematic MRIs and we know that EGFR disease is progressing in the brain.
AP: Thank you. So, Anna, I have a question for you. You were part of the development of subcutaneous amivantamab. In the presentation, Nicolas, we saw very interesting data in the PALOMA-2 and particularly in PALOMA-3 that is a part of the development and the reimbursement [??] in Europe, considering that in EMA the [??] have submitted the report for MARIPOSA and PALOMA-3. But the question from my side is what is your evaluation to understand an improved advantage in the efficacy? I know that this is a very one billion question but it’s very interesting because we saw very interesting improvement [??] in progression free survival and overall survival. It’s very interesting also translating this potential activity in the first line.
AM: I thought you might ask that. The first thing that springs to mind is if you’re seeing improved efficacy in a subcutaneous administration is are you getting better exposure to drug? So is the tolerability better so patients are having it for longer and not coming off because they had intolerable toxicity? But I don’t think that was the case; the dose intensity looked broadly similar. Which leads to the second question – is there something about the subcutaneous administration that means that it’s working better? Again, in the Twittersphere there’s lots of discussion about whether it could be something about the fact it’s going into the lymphatic so it’s having more of an immunological mediated mechanism of action. Amivantamab does have some immune mediated actions so could it be something to do with that? We just don’t know. Or could it simply be just that this trial was not designed with efficacy as its primary endpoint and that’s something we need to look at properly in a trial designed to compare efficacy outcomes. So intriguing but I don’t think we’re going to get a definite answer on that any time soon.
NG: The global sense is that subcutaneous amivantamab will be prominent in the prescription at the end because it’s easier for the patient, for the staff and the nurses. We have a lower risk of infusion related reactions so at the end it will be subcutaneous for sure.
MK: I think the other thing that’s just worth commenting on is the incidence of the thromboembolic events as well. With the prophylactic anticoagulation it did seem to be really along the levels you’d expect to see for this patient group with any treatment. It didn’t look to be particularly high incidence. So I thought that was quite reassuring as well.
AM: Yes, and asking a patient to have thromboembolic prophylaxis for the first four months of therapy, is that going to be something that patients are willing to take? If you’re moving from an intravenous regular attendance to a subcutaneous regular attendance perhaps a thromboembolic prophylaxis doesn’t seem like too much of an ask for those patients if you’re getting the outcomes you want.
AP: I have only the last question for Raffaele. So we discussed a lot of new data in the first line, new data in the second line, also from PALOMA-2 and PALOMA-3 and also LAURA suggests that in this particular disease, EGFR positive disease, our patients need to be treated with a drug specific for the tyrosine kinase [??]. So what’s your feeling in the future, considering this data, to postpone, to anticipate a combination strategy both in the early stage, so we have data from ADAURA, LAURA, and [??] in metastatic disease?
RC: Thank you, that’s the million dollar question. Are you thinking about chemotherapy/osimertinib or chemotherapy plus something else in the adjuvant setting concurrently instead of sequential adjuvant chemotherapy followed by osimertinib? Is that what you’re intending to?
AP: Yes, both this and also a potential combination or amivantamab/lazertinib or osimertinib in the early stage.
RC: In drug development, as you all know very well, we tend to move from later stage to locally advanced into the early stages as we go along. So I wouldn’t be surprised if we’re starting to see studies evaluating these strategies also in the adjuvant or neoadjuvant setting. Of course we do have neoadjuvant strategies in EGFR TKIs with studies currently enrolling. So there has to be a trade-off between, and we were discussing about LAURA, indefinite osimertinib plus concurrent chemoradiation. How many patients will actually be on it indefinitely? My experience in the adjuvant setting is that osimertinib in the adjuvant setting seems to be slightly less well tolerated and in the advanced setting there are more patients needing a dose reduction and that’s only for a certain amount of years in the adjuvant setting, three years. LAURA is potentially indefinitely so that’s something else. Going back to your question, if we are talking about adjuvant strategies, new sorts, or neoadjuvant strategies, how do we trade off between the duration of the strategies and patients’ convenience and, actually, the maximum biological treatment that we need to guarantee the outcome, that’s quite tricky at the moment. That’s the reason why there needs to be a very strong built-in or bolt-on translational component because we still need to think about who we should de-escalate instead of treating everybody with the same cocktail. But it’s easier said than done at the moment.
MK: Fantastic, thanks very much. So there was some really interesting discussion there around our first topic. In the interests of time we will move on to our second talk. So delighted to formally introduce Professor Califano, medical oncologist, lung oncologist, from The Christie Hospital in Manchester, UK, who is going to give us updates on genomic driven non-small cell lung cancer, so EGFR and beyond. So Raffaele, over to you.
Updates on genomic driven NSCLC
RC: Thank you very much Matt and good evening to all our colleagues who are joining us online. I will only have three abstracts to present and I will focus on giving you an update on genomic driven non-small cell lung cancer, not only on EGFR. The first abstract I would like to present is a phase III randomised clinical trial evaluating ivonescimab, which is a dual VEGFR and anti-PD-1 antibody combined with chemotherapy versus chemotherapy alone in patients with EGFR mutant non-small cell lung cancer who have progressed after a first or second or third generation EGFR TKI. This is the HARMONi-A study.
So patients who were enrolled in HARMONi-A were patients with advanced non-small cell lung cancer which harboured common EGFR-activated mutations and they had experienced disease progression in their cancer either after first generation or second generation EGFR TKI and then osimertinib or straight after osimertinib. Patients were randomised to receive carboplatin, pemetrexed chemotherapy plus placebo versus carboplatin, pemetrexed plus ivonescimab for four cycles then moving to a maintenance phase of pemetrexed/ivonescimab every three weeks or pemetrexed/placebo every three weeks up to a total of two years. The primary endpoint was progression free survival by independent radiological review.
The trial met the primary endpoint, demonstrating a longer progression free survival favouring chemotherapy plus ivonescimab with a median PFS of 7.1 months for the investigational arm versus approximately 5 months for the standard chemotherapy arm with a hazard ratio of 0.46. There was a consistent benefit seen across all the PFS subgroups, also in patients with or without liver metastases and independently on the kind of EGFR mutation. This is in particular focussing on the Kaplan-Meier showing the different activity by exon 19 deletion or exon 21. You can see from the Kaplan-Meier that there is a sustained benefit of chemotherapy/ivonescimab versus chemotherapy only in both kinds of mutation with a hazard ratio which is very similar, around 0.48 for the deletion 19 and 0.43 for exon 21.
The authors also looked at the PFS on the basis of presence or absence of brain metastasis and there seems to be a benefit also in patients with brain metastasis which is particularly relevant in patients whose tumour harbours EGFR-activated mutations, again with a longer progression free survival for the investigational arm over standard chemotherapy with a very similar hazard ratio.
There was a higher response rate for the chemotherapy/ivonescimab arm, a 50% response rate, with a disease control rate which was 93% and a longer median duration of response favouring the chemotherapy/ivonescimab arm at nearly 7 months versus 4 months for the standard arm.
The authors presented two overall survival Kaplan-Meiers. Both OS analyses were required by the Chinese regulatory authority and this is the first one with 30% data maturity, approximately 96 events. There is a trend towards a longer overall survival favouring the investigational arm. This is the subsequent overall survival with 52% data maturity, again with a trend towards a longer overall survival favouring chemotherapy/ivonescimab.
What about safety? Clearly patients who were randomised to the investigational arm had a higher incidence of serious adverse events. There was a higher incidence of treatment related adverse events which led to discontinuation of ivonescimab or placebo. It goes without saying we were expecting a higher incidence of immune related adverse events and the grade 3 and above incidence was 6.2% in the investigational arm together with a higher incidence of VEGF related adverse events again in the chemotherapy/ivonescimab. Looking at these adverse events of special interest in a more granular way, patients who were randomised to chemotherapy and ivonescimab had a higher incidence of proteinuria, which is a VEGF related adverse event, but also bleeding and hypertension which is all pretty common and standard with this kind of agent. The important thing is that the incidence of grade 3 or worse adverse events such as bleeding was zero and for thromboembolic adverse events again it was zero. So it was usually grade 1 and 2 the kind of adverse events of special interest that we saw in the investigational arm.
So this is an East Asian study and the authors conclude that the combination of ivonescimab plus chemotherapy certainly improved progression free survival in patients whose tumour had progressed after prior EGFR TKI with a benefit across the prespecified subgroups. This is now currently approved in China but the study has been expanded with a global study, which is the HARMONi study, where patients from different countries, including Europe and North America, will be randomised in a study with exactly the same design of HARMONi-A and I think that will be very important to evaluate the efficacy of this combination also in a predominantly Caucasian population.
Moving on, we’re moving to ALK gene rearrangement positive non-small cell lung cancer and I’d like to present to you the five-year follow-up data with the PFS data from the CROWN study that was presented by Professor Solomon at ASCO. Just to remind you, CROWN was a large phase III randomised clinical trial where patients with advanced and untreated non-small cell lung cancer harbouring an ALK gene rearrangement were randomised to receive lorlatinib or crizotinib. The trial did not allow crossover to lorlatinib for patients who were randomised to the crizotinib arm. The primary endpoint of the study was progression free survival by independent radiological review. We have seen the initial data cut-off showing that CROWN was a positive study with a longer progression free survival favouring lorlatinib. We’ve also seen the three-year PFS analysis. What I’m presenting here is the post-hoc analysis at five years. Bearing in mind that the primary endpoint this time is investigator assessed progression free survival as the independent radiological review stopped at the three-year analysis.
This is the PFS for the five-year follow-up. There was a sustained benefit in progression free survival favouring lorlatinib over crizotinib with 60% of the patients still progression free at five years with a median progression free survival not reached on lorlatinib and similar to the previously reported PFS at around nine months for crizotinib.
There was a sustained benefit across all the subgroups and I would like to point your attention towards the presence of brain metastasis yes or no. There was a clear impact of lorlatinib in the CNS metastatic disease where for patients who had brain metastasis at the outset, approximately 23% of the patients, the hazard ratio over crizotinib was 0.08, so excellent results for patients with metastatic disease. But lorlatinib worked exceptionally well with sustained benefit in progression free survival also in patients without baseline brain metastasis and you can see that in both datasets PFS was not reached. The hazard ratio was 0.08 as shown before for patients with baseline brain metastasis and went to 0.24, which is a pretty good hazard ratio, in patients without baseline brain metastasis.
There was a longer time to intracranial progression by investigators in patients who received lorlatinib over crizotinib with 92% of the patients being intracranially progression free at five years, which is an extraordinary achievement, I believe, with a hazard ratio of 0.06. This is also the intracranial progression in patients with or without brain metastasis, again clear benefit and excellent penetration of lorlatinib in the brain.
Looking at safety, the safety profile mirrored what was previously reported in the previous presentations and papers. Lorlatinib clearly has a different and distinct adverse event profile compared to the other first or second generation ALK inhibitors. We’ve seen more oedema, lipid [arrangement], GI toxicity, weight gain and also neuropathy when compared to crizotinib. Approximately a quarter of the patients had to have a dose reduction which was related to a treatment related adverse event. From a clinical point of view, it’s important for us clinicians managing this disease to understand if reducing the dose has an impact on patient outcomes. The authors reported on the progression free survival for patients… exactly answering this question, for patients who had to have a dose reduction in the first four months of treatment and there’s a clear answer there is that independently on the dose reduction there is still a benefit of lorlatinib over crizotinib in terms of progression free survival and also time to intracranial progression, which is very, very important in this disease group.
The authors also looked at how lorlatinib performed against crizotinib in two specific subgroups of patients with poor prognosis, looking at biomarkers such as the variants 1 and 3 but also the presence of the co-mutation of p53. We know from previous studies that patients, for example, with a variant 3 have a more aggressive disease and also patients with a co p53 mutation again have a more aggressive biological disease and potentially lower benefit from ALK inhibition. Looking at this dataset, it is reassuring to see that the benefit of lorlatinib was consistent also with the worse variant and with the presence of co-mutation of p53.
Together with that, the authors also described what happened at a biological and molecular biological level for patients who were treated with lorlatinib. I think it was very important to see that in terms of mechanism of resistance there was no new single or compound ALK mutation but the biology completely switched and the progression was due to activation of bypass tracks such as MEK inhibition or BTK pathway or other molecular pathway. This clearly shows how lorlatinib was able to wipe out all the ALK resistant clones and the biology had to switch to a different mechanism of survival.
So the authors conclude that this five-year follow-up study in the CROWN study showed that there is a consistent benefit for lorlatinib over crizotinib with a median progression free survival that still has not been reached with 92% of the patients still being intracranial progression free at five years which is unprecedented. This reinforces the message that lorlatinib is certainly one of the very effective first-line ALK inhibitors that we can use in clinical practice.
I will then move to the third and last abstract. Moving again to a different kind of molecular aberration, this time KRASG12C mutations. I’m going to present you the data from the KRYSTAL-12 study which was presented by Professor Tony Mok. This is a phase III study looking at adagrasib versus docetaxel in patients with pre-treated non-small cell lung cancer harbouring a KRASG12C mutation. These patients had to be treated with platinum-based chemotherapy and also anti-PD-1 or anti-PD-L1 treatment. Patients were randomised in a 2:1 fashion. The primary endpoint was progression free survival by independent review.
The study was a positive study; there was a longer progression free survival favouring adagrasib over docetaxel – 5.5 months versus 3.8 months with a hazard ratio of 0.58 which was very similar to the hazard ratio reported in the PFS by investigators. Looking at the subgroup analysis, pretty much consistent benefit favouring adagrasib over docetaxel. Perhaps if you want to be quite specific you can look at the brain metastasis subgroup where the hazard ratio in patients with brain metastasis at baseline crosses the equivalence line. But, to me, a piece of very important information is the fact that this an effective strategy in patients who had chemoimmunotherapy concurrently or had chemotherapy first followed by an anti-PD-1/anti-PD-L1 at progression. There was a higher response rate favouring adagrasib over docetaxel with the study reporting a 9% response rate for chemotherapy whilst the overall response rate for adagrasib was 32% and it was also a longer median duration of response in the adagrasib arm. There was also a report of intracranial response and efficacy with a higher intracranial response at 24% in the adagrasib arm versus only 11% in the docetaxel arm with a disease control rate intracranially for the KRAS inhibitor at 64%.
The authors also reported on quality of life and this is the time to deterioration using the Lung Cancer Symptom Scale. Again, it took longer for patients on adagrasib to witness a worsening of the deterioration of their lung cancer symptoms when compared to docetaxel. Looking at safety, there was certainly a higher incidence of treatment related adverse events which led to dose reduction in the adagrasib arm, also dose interruptions, but the number of toxic deaths was pretty similar and that is reassuring.
On the basis of KRYSTAL-12, the authors concluded that adagrasib is now an effective and safe option for patients who are seeing their cancer progressing after chemotherapy and immunotherapy where they are at the point of choice with docetaxel. It seems to be certainly higher efficacy in terms of PFS, response rate but also intracranial response rate.
Having discussed these three abstracts I would like to hand over to Dr Matt Krebs who will lead on the discussion. Thank you very much.
Panel discussion
MK: Fantastic, thank you for that very clear summary of three really interesting talks and updates at ASCO. Lots of potential discussion points here. I’ll kick us off and please, Antonio and others, please, please chip in. So just taking the ivonescimab first of all, I find this intriguing. So the data look quite promising but in the context of immunotherapy not being effective for the EGFR mutation population, so KEYNOTE-789, CheckMate 722 data that clearly show immunotherapy isn’t beneficial, we’ve now got a bispecific antibody against PD-1 and VEGF that looks promising. So my first question is do we think this is all just the VEGF driven component of this or do you think there is a PD-1 component that’s additive? I’m just interested in your thoughts about how this is working.
RC: It’s difficult to say. First of all I’d like to remind the audience that this was an East Asian study only so we may need to see how this combination performs in a Caucasian, or certainly non- predominantly East Asian population. But I was thinking exactly the same. When I saw the results and you go back historically to IMpower150 and then the later on studies, IMpower150 was a different chemotherapy backbone with carboplatin/paclitaxel and obviously evaluating atezolizumab and bevacizumab. That's a harsh chemotherapy, it’s not an easy chemotherapy to tolerate in terms of adverse events. So this is the standard carboplatin/pemetrexed which is the one I would use for patients where trials options are not available so far with no other approval. We’ll at the point of chemotherapy use carboplatin/pemetrexed.
The short answer is I don’t know. We know that VEGF inhibition and antiangiogenic may play a role in terms of immunotherapy resistance and therefore when you combine the immunotherapy together with a VEGF inhibition you may enhance the activity of the immunotherapy, potentially manipulating the vasculature but also the microenvironment. So, to me, this are very important results. I don't know if it's just the compound which is better than what we’ve previously seen. What I reserve to see, though, is to see the phase III randomised clinical trial which is now planned to be launched globally in exactly the same setting. Because at that point you will have evidence also in the non-East Asian population.
MK: Yes, it’s intriguing. I guess you can only be a bit speculative about this but I’m just wondering where do you think this combination might fit in relation to the MARIPOSA-2 combination in terms of chemotherapy, amivantamab and lazertinib? What are your thoughts? It may be a bit difficult to answer.
RC: Yes, potentially we’re going from no options at progression from a first line EGFR TKI to a number of options and that would be for us to try and select the patients who should receive option one or option two. So at present the closest approval that we’re going to get is the MARIPOSA-2 regimen which is likely to be carboplatin, pemetrexed plus amivantamab which, again, likely to be the subcutaneous formulation which we have discussed beforehand. I still don’t know how we should select patients who should go into chemotherapy plus VEGF inhibition and immunotherapy. There have been discussions in the past about potential use of VEGF in patients with brain metastasis because of the potential anti-oedema effect, that speculation. At some point we may need to see the data and the answer will only come from a randomised head-to-head study which may not happen. But if we have also this triplet approved, which is realistically addressing two different pathways, the VEGF and the immunotherapy, as clinicians treating lung cancer, we will need to sit down and try and scratch our heads on how to select who is getting what.
MK: Just before we move on to ALK, just perhaps a question about your own practice, perhaps to everyone as well. With so many changes now with potential options and potential options second line with combination treatment that may not matter what the resistance mechanism is from your previous therapy, is there still a place for rebiopsying patients, liquid biopsy on progression, to look for resistance mechanisms?
RC: I still feel that if it is feasible and the patient is willing we should rebiopsy. So certainly at the moment we don’t have a lot of data on rebiopsy post, let’s say, the MARIPOSA regimen and I don’t think I’ve seen a lot of data on the FLAURA regimen. But for patients who are exposed to osimertinib in the first line setting albeit, for example, in the FLAURA regimen where you get chemotherapy, I still would like to see if there’s a switch to squamous or small cell or if you are able to identify any bypass tracks, you know MET, we can then target at that stage. I think important data will need to be presented also on the post-amivantamab/lazertinib. I know that a number of patients can’t or won’t have a rebiopsy and then at that point we’ll have this sort of allcomers regimen which can be used off the shelf. That’s my personal view.
MK: Yes, I think it still has a place for those reasons. Yes, I agree. Okay, let’s come to lorlatinib, so is this going to change our standard of care?
RC: The drug is incredibly effective, that’s undeniable. We have known for a while that one of the strengths of lorlatinib is the CNS penetration, perhaps sometimes too much CNS penetration in terms of the adverse events that you can see from a neurological point of view and the neuropathy for example. You need to bear in mind that this study was launched and designed a bit later on than the ALEX study or brigatinib which were presented a while before. So we also got better at managing metastatic disease in the brain in terms of neuro-oncology MDT. The data is very good, it’s undeniable, and I was impressed to see that kind of intracranial progression free survival. So there are very few patients with intracranial events and actually very few patients with the cranial events from year 4 to year 5 or from year 3 to year 5. I really would like to see the overall survival data because that may also make you think if potentially this is something as a strategy that you would need to reserve for patients with brain metastasis at the outset compared to, let’s say, doing a second generation ALK inhibitor and then switching to lorlatinib at progression. Because it has to be said that the adverse events are different for lorlatinib compared to brigatinib and alectinib. It would be also relevant to see if these patients are going to be on the drug for years how sustainable it is to be on lorlatinib for three, five years, six years. We have seen that a quarter of patients will need a dose reduction. So, to me, it’s a balance between efficacy and adverse events, even lower grade, because you want your patient to be on an ALK TKI for a long period of time. I still feel that the overall survival data, if we’ll have it at some point, will be important to make a judgement call. I don’t know what other colleagues think about it.
MK: I was perhaps just going to come to Nicolas as well, just to bring in a question we’ve had from the audience, and a fair question. Was the control arm right for that study? Should that have been…?
RC: Yes, correct. I don’t know if Nicolas wants to answer because otherwise I’m answering all the questions.
NG: Yes, that’s a very good question. We would have preferred to have a control arm with a second generation ALK TKI for sure. It’s not the case but still if we do cross-trial comparison my sense is that lorlatinib is probably doing better. But five year 60% PFS, that’s really good, and the efficacy of lorlatinib post-alectinib or post-brigatinib is probably lower. I agree that the side effects are a little bit different, probably also the side effects are different in the first line versus the late line setting in patients with a history of CNS disease, a history of radiation and we know that it may impact also the cognitive abilities. So that’s for sure really interesting data. I agree with you that we need also to continue the follow-up of these patients to also understand what are the sequences of treatment after lorlatinib. Probably chemotherapy is an opportunity for these patients but we have also fourth generation ALK TKIs that are now being investigated. So that’s obviously some opportunities for the patient.
AP: Yes, I think that this is a very interesting topic discussed by Raffaele and many questions are ongoing. We are receiving many questions but it’s time to move on and it’s my pleasure to introduce Anna Minchom discussing the third part of this webinar on updates on non-genomic driven lung cancer. Please, Anna, the stage is yours.
Updates on non-genomic driven lung cancer
AM: So I’m going to present two abstracts from ASCO. The first is a change of tack from non-small cell lung cancer to small cell lung cancer. This is the ADRIATIC study. Small cell lung cancer is a very poor prognostic cancer, as we all know, and in the metastatic setting has a median overall survival of around 12 months. For limited stage our current standard care is chemoradiotherapy plus or minus prophylactic cranial irradiation. The ADRIATIC study was a phase III study looking at adjuvant durvalumab, a PD-1 inhibitor, and tremelimumab as consolidation therapy following that chemoradiotherapy. The tremelimumab I’m not going to discuss any more because that data wasn’t presented at ASCO.
Patients were eligible if they had early stage non-resectable small cell lung cancer who had not progressed after concurrent platinum-based chemoradiotherapy. They were randomised to receive chemoradiotherapy plus placebo or chemoradiotherapy plus durvalumab for 24 months, for two years. The patients were randomised first in a 1:1:1 for the different arms, the three arms, and then just between durvalumab and placebo. They were stratified according to stage and whether they received prophylactic cranial irradiation. It was dual primary endpoints for progression free survival and overall survival.
730 patients were randomised including over 260 to durvalumab and placebo. The baseline characteristics were well balanced. At this data cut-off, which happened earlier this year, there was a long duration of follow-up – over 37 months for overall survival and over 27 months for progression free survival. The headlines are that overall survival was significantly improved with durvalumab compared to placebo with a hazard ratio of 0.73, statistically significant, and a median overall survival of 55.9 months in the durvalumab arm and 33.3 months in the placebo containing arm. Progression free survival similarly was improved with a PFS of 16.6 versus 9.2. The subgroup analysis, although it’s not presented here, I can tell you was consistent across the predefined subgroups that were looked at. Some of the confidence intervals are fairly large given the quite small numbers of patients in the subgroup but nevertheless all on the right side of benefit with durvalumab.
In terms of toxicity from the durvalumab, the maximum grade 3 or 4 adverse events occurred in around 24% of patients in both arms but the rate of treatment discontinuation for these was higher in the durvalumab arm – 16.3% versus 10.6%. In terms of probably our most important toxicity in this setting which is pneumonitis, a combination of immunotherapy plus radiation to the thorax raises this as a very real risk. The rate of pneumonitis or radiation pneumonitis was 38% in durvalumab versus 30% in the placebo arm. But interestingly the high grade, grade 3 and 4, both about roughly similar, around 3% in both arms.
So the authors conclude that they have both a statistically significant and a clinically meaningful improvement in overall survival and progression free survival compared to placebo with durvalumab and I would agree. Salvaging these patients from metastatic disease is of primary importance to improve their outcomes and this is what we appear to be doing here. Also in this potentially quite poorly group it was reasonably well tolerated and, importantly, rates of high grade pneumonitis were similar. So the authors conclude this is a new standard of care which we can discuss but I believe it probably is.
The second paper I’m going to present is of EVOKE-01; this is of sacituzumab govitecan versus docetaxel. Sacituzumab govitecan is a Trop-2 targeting ADC with an SN-38 payload. Trop-2 is preferentially expressed in solid tumours including lung cancer. This is a schema of EVOKE-01. Patients with advanced non-small cell lung cancer who had progressed on platinum-based chemotherapy and PD-L1 or PD-1 containing regimes were randomly allocated to sacituzumab govitecan, which I’ll call SG, and docetaxel. Just over 600 patients were randomised, the primary endpoint was overall survival and patients were stratified according to histology and, this is an important one, response to last anti-PD-L1 containing regime. So this is a prespecified stratification.
The primary endpoint of overall survival showed no difference in the overall patient group with a median overall survival of 11.1 months in the SG arm and 9.8 in the docetaxel arm. Progression free survival, similarly, was very similar between the arms – SG and docetaxel around 4 months PFS – and overall response rate not statistically significant in comparison.
But here are the subgroup analyses and that response to anti-PD-L1 is given here. So patients were categorised as to whether they had progressive disease and stable disease as their best response to their prior PD-1/PD-L1 inhibitor and whether they achieved a complete or partial response. In the subgroup who only achieved progressive disease or stable disease as their best response, there did appear to be a benefit from the SG when compared to docetaxel. Here we have the overall survival broken down, so 11.8 months with the SG and 8.3 with docetaxel. They then broke down that group who had non-responsiveness to last immunotherapy into the squamous and the non-squamous groups. That overall survival benefit was seen in both groups, non-squamous and squamous.
We always need to look carefully at the adverse event profile of both chemotherapy and antibody drug conjugates. You can see here that there are broad similarities. In terms of rates of overall any grade toxicity we have 67% and 76%, forgive me, that’s grade 3 or 4. So over 50% of patients are getting high grade toxicity from the drugs. There seems to be, as you would expect, some increase in haematological side effects in the chemotherapy arms.
There were also attempts to look at patient related outcomes, in particular this symptom assessment questionnaire. The modality that was presented with this shortness of breath domain which seems to show that the burden of their symptoms was less in the SG arm with a deterioration that was slower to happen. Similarly with the NSCLC-SAQ total score. I’d say this is to be applauded – where we’re trying to work out the toxicity and burden of these quite toxic drugs, how the patient is feeling is really, really important.
So in conclusion this was a negative study, it did not meet its statistical significance for its primary endpoint of OS. However, in some prespecified subgroup analyses a benefit was seen. The safety profile was similar to docetaxel. So this drug is being evaluated in ongoing studies, including in a combination of immunotherapy in the first-line setting.
Panel discussion
AP: Thank you Anna. So discussing these very interesting trials, one, the triumph in the early stage of small cell lung cancer and the other potentially one of very exciting data but in a negative way, so failure in another ADC in patients with non-small cell lung cancer. So I think that we have not so much time for a big discussion after the three very interesting presentations but my question for you is around the EVOKE trial. So this was a very interesting trial, a new molecule, a very interesting ADC with a payload currently received for our patients with non-small cell lung cancer, also using very old-fashioned chemo, but the results are clearly negative if we consider the intent to treat population, the primary endpoint of overall survival. So what are your thoughts considering this trial and also the future of ADC in the field?
AM: I think docetaxel is not a great option for our patients but we need to be doing better than docetaxel and we’re clearly not seeing that across the board here. There probably is a small benefit in terms of tolerability compared to docetaxel but we’re still exposing patients to toxicity. So I don’t think that this will change our practice for our patients in an unselected group. They are active drugs and I think the future probably lies in really good patient selection and we don’t yet really know what that is. We’re not selecting for marker expression in this setting. We have that sniff of activity in patients who didn’t have a good response to prior immunotherapy but what is that about? What’s the biological reason behind that? If we can start to unpick that sort of question then we might be able to find a subgroup where we can get enough benefit from a drug to make the toxicity and intravenous administration worth it to be better than docetaxel.
AP: Nicolas, I see your face that you are…
NG: It’s disappointing for sure but we have learned from this trial in the second-line setting, probably not the best setting to assess a new drug because of the heterogeneity of patients, the fragility of patients and so on. So I’m a believer of first line with ADCs, so there are many trials ongoing in the first-line setting. We’ll have to see how these drugs are performing versus chemotherapy plus IO; combination is probabl
