Welcome to The Lung Cancer Sessions brought to you by the Video Journal of Oncology (VJOncology).
This exclusive roundtable discussion features leading lung cancer experts Sanjay Popat(The Royal Marsden NHS Foundation Trust, London, UK),David Baldwin(Nottingham University Hospitals NHS Trust), Fiona McDonald(The Royal Marsden NHS Foundation Trust, London, UK) and Riyaz Shah(Maidstone and Tunbridge Wells NHS Trust, Maidstone, UK).
Watch this on-demand feature to learn about the exciting lung cancer advances from ESMO & WCLC 2021 and gain exclusive insight into the UK lung cancer screening (UKLS) trial, Lung ART and the role of adjuvant radiotherapy in N2+ NSCLC, as well as the delivery of concominant chemoradiation in the UK. The panel also dissect the latest data in resected NSCLC including IMpower010 and consider the implementation of molecular testing in the multidisciplinary team. In addition, they evaluate the long-awaited POSEIDON data in the metastatic setting and discuss interesting trials in HER2-mutated NSCLC.
“There was a non-significant trend. In fact, a 31% reduction in lung cancer-specific mortality and about a similar trend in the all-cause mortality reduction. And if you looked at the lump that the mortality from cancer as a whole was diagnosed over the whole seven-year follow-up period, and there was a significant reduction in the lung cancer deaths. So that’s very encouraging and there are a whole lot of other things that are published in the paper that fit very nicely.”
[On cardiac toxicity with adjuvant radiotherapy] “We don’t know what the optimal constraints should be, even in definitive chemo-rad, curative chemo-rad to the chest involving the mediastinum… You know, you’re weighing up the risk of cardiac tox over the risk of not controlling an active lung cancer. Obviously that’s different in the adjuvant setting, but we don’t have any new dose constraints as yet to apply.”
– Fiona McDonald
Resected NSCLC & molecular testing
“What outcome is the right outcome in an adjuvant trial? You know, when I was training, we classically thought of overall survival. We classically thought that you give an adjuvant treatment to cure more people. So is disease-free survival an appropriate surrogate? So, I think there’s a big existential discussion that needs to be had. And because all these trials are reporting DFS outcomes, OS will come at some point, but we’re still years away.”
– Riyaz Shah
POSEIDON & HER2-mutated NSCLC
“I think that our treatment decisions are very complicated now, and this is part of the great education that we need to do across the UK as you’ve rightly said, across other disciplines to look at cases, to look at tips, hints, and tips of what works and what doesn’t work. And you know, where we’ve had success with one particular type of practice where another type of practice would have done something different. And part of this is adverse event management, understanding what the adverse events are, picking them up early and instituting the appropriate actions.”
– Sanjay Popat
Watch Full Video
SECTION 1: UK lung cancer screening
SANJAY POPAT: Well, hello everyone. My name is Sanjay Popat. I’m a consultant medical oncologist at the Royal Marsden Hospital and professor of thoracic oncology at the Institute of Cancer Research. It’s my great pleasure to welcome all of you to this episode of the lung cancer sessions, hosted by VJOncology. Today, we have a round table discussion held between myself and excellent experts in the UK. We have my colleague, Dr David Baldwin, who’s Professor of Respiratory Medicine at Nottingham. We have my colleague, Dr Fiona McDonald, who is Consultant Clinical Oncologist at the Royal Marsden Hospital, and my colleague, Dr Riyaz Shah, who is Consultant Medical Oncologist at the Kent Cancer Center.
All of you around the table here have been involved with some of the key data that have been presented over the last year. And what I wanted to share with the audience, this occasion, is some of the key data that has been presented at recent congresses, specifically the World Lung Conference, and also the ESMO Virtual Conference, just to highlight some of that key data and put it into UK perspective for the UK audience to bear in mind when seeing their patients in day-to-day practice.
So I’ll start off with David. David, I just want to focus in on the screening strategy that’s going on in the UK and the latest data that was presented at the World Lung Conference. So my understanding is that John Field and colleagues presented updates on the UK lung screening study and in parallel, this was also published in one of the Lancet Journal papers as well. David, do you wanna just give us a brief summary of how the study was designed and its key take home messages?
DAVID BALDWIN: Yes, so the study is the only screening study that represents a true population approach. And that was the design process was the only way we could get this funded was to design it in a way that the UK National Screening Committee ran screening interventions, which is essentially approach the whole population of the eligible age and ask them if they want to be screened. And then select them on the basis of whether they are at a high enough risk or not. And hence the UKLS study is the only one that gives us an idea about what actually happens when you approach the entire population as its been in a given age range and see just how many people end up saying yes to a trial. So that was one unique aspect of this study. And it shows that essentially that’s not the way to do things. So, and that’s translated now into implementation work in the UK where we’re absolutely not going to approach the whole population. We’re going to approach people on the basis of their smoking history that’s in our excellent primary care records and maybe some other methods of improving participation. The other unique, or more or less unique, element to this study, since it is the same design as some other non-lung cancer screening studies, is the single screen design. So, you would never just do a single screening in a screening program because the accumulated cancers that have, that occur over the next few years are the way you save lives. The only reason you do a single screen is to just check to see whether the intervention works or not in terms of reducing mortality. But it also allows you to measure things like overdiagnosis rate and a whole lot of other things that are important in implementation of screening. So this was a single screen versus no screen at all.
And there are a few other things that were part of the study including smoking cessation, very, very important, applied equally to both arms. And then very importantly, a downstream protocol for management of findings, which was basically on the basis of the NICE lung cancer guidelines for downstream management, very strict, at each center and sort of policed during the study. And then also a nodule management protocol, which was based broadly speaking on the NELSON study, modified for the single screen. We’ve led to a lot of repeat scans, in fact, a 50% repeat scan rate because of the very low threshold we sat at that point for the fear of missing any cancers.
So that’s given us some very, very useful information that’s already been published in a few different publications from UKLS. This latest publication is something that I know that some of the senior authors never really wanted to do because it, the study is not powered for mortality outcomes. So it’s a sort of brave thing to do. It wasn’t always the intention to try and publish a meta-analysis that included those of the other European trials and that’s exactly what’s happened. And what you can say, as a bottom line, from this latest publication was presented at the World Lung Cancer Conference, as you say, is now published in Lancet Regional Health Europe is that there was a non-significant trend. In fact, a 31% reduction in lung cancer-specific mortality and about a similar trend in the all-cause mortality reduction. And if you looked at the lump that the mortality from cancer as a whole was diagnosed over the whole seven-year follow-up period, and there was a significant reduction in the lung cancer deaths. So that’s very encouraging and there are a whole lot of other things that are published in the paper that fit very nicely. Often when you look at these screening trials, there’s always something that doesn’t really fit very well. And this fits perfectly. It’s one of the few trials where including NLST and NELSON, where, you know, you can actually see something straight away that there’s slightly anomaly. You can’t see that very much in the UKLS at all. So the late stage disease is reduced. Early-stage detection is very high. There’s a slight increase in incidence over the seven years equating to a maximum overdiagnosis rate of 15%, but where the fit, whether seven years is truly long enough in this context, we don’t actually know. And certainly the very long-term follow-up NLST showed that the overdiagnosis rate came down even further.
But the other concern of course, related to UK practices, that this is, the UKLS represents one of the most modern screening trials in terms of technology. More modern than NELSON even and the risk of overdiagnosis when you’ve got really highly sensitive, superb radiologists, as we did with the UKLS, is even greater. And the management of that stage of maybe the soft solid and part solid nodules, it was not quite as clear as it is now. So there could well be an overdiagnosis rate in UKLS, which is probably above of what we get now. In terms of how it influences the whole process, the meta-analysis confirmed that if you look at all of the mortality, the trials that reported mortality in CT screening across the world, then you get a 16% overall reduction in lung cancer-specific mortality and a 3% reduction in all-cause mortality. Bear in mind that these are trials with variable screening round numbers and length of time in which you screen the patients. And that will have a small influence on the disease-specific mortality and the all-cause mortality in particular. So you have to be really careful about saying, this is what we see in clinical practice. As it happens, if you calculate what you might see in all-cause mortality from ONS data, UK ONS data, the proportion of patients that die from lung cancer, it does turn out to be around about 3%, if you have a 100% participation rate. Which is essentially the similar sort of thing that you measure in a trial. So it’s interesting that those two figures are the same. I’m not quite clear why, but that would be some work for the future. In terms of the implementation now, we’ve had a number of studies that have looked at meta-analyses of mortality benefit and there’s no real argument that CT screening reduces lung cancer mortality. There is a little bit of concern that we haven’t got trials big enough to show an all-cause mortality benefit. But if you look at all of the breast cancer screening trials, all of the colorectal cancer screening trials, and the survivor cancer screening trials, none of them have shown an all-cause mortality benefit. It’s simply a power issue. And, you know, you need a trial that has got at least 45,000 patients in each arm to have a chance of detecting all-cause mortality because of the proportion of deaths that are contributed.
So overall, very encouraging, no surprises, I suppose. And from the point of view of the negotiation we’re having at the moment, it’s still all around cost-effectiveness, is this a cost-effective intervention or not? And I think it is, but we need to need to be clear about that. And we also need to quantify what the ICER is for CT screening, because that will enable us to decide on exactly who we offer the intervention to.
SANJAY POPAT: David, these decisions on cost-effectiveness are currently in the hands of the National Screening Committee and, or is it another committee? And what’s the timeframe that we in the UK might find out about the decisions that might have been made?
DAVID BALDWIN: Yeah, so it’s firmly in the hands, in terms of a national program, funded in the normal way through ministerial approval, and then section 7A in the NHS England at least, basically delivered by the NHS, but quality assured by the UK Health Security body, that it replaces Public Health England. If that happens, then it will mean a quality assured national program with funding pretty much guaranteed, but they do need to prove that the ICER is acceptable. And the current process is there is a revised cost-effectiveness analysis going on, acknowledging that the initial one that was done back in 20, 2006, then followed by an update in 2018. Both were not well calibrated models. And when I say not well calibrated, really quite significantly different to what we see in clinical practice or in screen detected cancers. So that’s ongoing, and we are currently in a situation where we provided the best data possible, I think, in terms of resource use and treatments, contemporary survival during the same time periods. Pretty much the envy of the world when you say this is what we have. And that is going to be put into the model. So I fully anticipate a very accurate model which will not just determine what happens in the UK. I think that if we get this through, and we produce a result that will be highly influential in Europe and elsewhere. I think most of the very detailed cost-effectiveness analyses that have been done have overestimated the cost of CT screening and underestimated the alternatives, including the Australian evaluation and the latest EUnet evaluation, which was very detailed, but seems to have some fundamental issues with the way they’ve done the cost-effectiveness analysis. So all good, I think, but the timeframe is you, is there needs to be a recommendation by the end of February from the UK National Screening Committee via a initial stakeholder view, which I’m sure, you know, everybody will be invited to input into. So the final recommendation comes out there, the stakeholder period for about three months and then by June next year they should recommend yes or no to a national screening program.
SANJAY POPAT: Thank you, David. And I think we all watch this space with anxious intent, keeping a very beady eye on what’s happening.
SANJAY POPAT: I wanted to switch focus of attention now to established lung cancer, particularly in the radical setting and bring in Fiona, my colleague to the discussion. One of the areas of controversy we’ve had for many years has been on patients with N2-positive non-small cell lung cancer, Stage 3, which has been resected, either found to be incidentally involved at N2 at the time of surgery or with established disease, which has been resected. One of the great questions has been the role of adjuvant radiotherapy in this setting. This was answered in part by the Lung ART trial, which our French colleagues presented at the ESMO Meeting last year. And we had an update of the Lung ART dataset, which my colleague, Dr Cécile Le Pechoux presented at the ESMO Meeting. Fiona, do you just want summarize about the Lung ART trial, the findings and its impact in the UK and perhaps, you know, what the most recent updated data we’re seeing.
FIONA MCDONALD: Yes. So the data, so the trial is looking at where we find unexpected nodes at the time of surgery in the end to middle mediastinal space of what the role of post-operative radiation therapy is because from historical studies, the jury was out was a bit of a gray area. The initial results we had of the Lung ART trial, at ESMO last year, showed that there wasn’t a disease-free survival advantage between the two arms adding in the post-operative radiotherapy or not adding it in. It did show that there was reduced mediastinal relapse quite significantly so in those that had the radiation, but there was an excess number of deaths, which, you know, if you look at radiation in general, in the thorax, we know from RTOG 0617, that we’re worried about heart doses. So all kind of fits that we would see excess deaths with radiation over the mediastinum. And so the recent update was looking at patterns of relapse, in particular patterns of the mediastinal relapse. And it did fit with the kind of nodal maps we were following on what stations we should be covering with our radiation fields. So the most common relapse for left-sided tumors, I mean, it was all kind of like station 2, station 4 and 7, slight difference between left- and right-sided tumors. So kind of what we would expect to see, but good to see that data. What it did highlight, prognostic factors wise, what was back to the question of the quality of the surgery, on how many lymph nodes stations were sampled? All of that information we kind of know from before that it helps us define our field that helps us then reduce the relapse. But overall it doesn’t add, the latest stage, doesn’t add any more to the fact we didn’t see disease-free survival advantage.
I think the key thing to remember is there were some exclusions from the trials of the portion of patients that never went into the trial are those with R1 resection and those with extracapsular spread. So we don’t have data in that setting. They were felt that they should be excluded from the trial because we should be offering them postoperative radiotherapy. So I don’t think we’ve got any data to suggest we shouldn’t still be doing that. But then now it’s a question, if you’ve got a complete resection, no extracapsular spread that you know, that we should be thinking twice about offering it. And then obviously the world has moved on. So we’ve got other trials coming out in that space, as you know, with the adjuvant immunotherapy after chemotherapy instead, and those trials haven’t got post-op radiation in so IMpower010 and then others will follow. So whether radiation in that setting gets entirely replaced by IO going forward, that may be the case, but I think we’re still left with having to consider it for patients with incomplete resections, R1 resection and also patients with extracapsular spread.
SANJAY POPAT: And one of the major concerns about adjuvant radiotherapy has been perhaps the cardiac toxicity, which may, have been accounting for some of the mortality identified. So if that is going to be given do radiation oncologists need to change the radiation practice to ensure that there is a greater concern given to the base of the heart or the cardiac vessels?
FIONA MCDONALD: That’s a very good question. And the answer is we don’t know what the optimal constraints should be, even in definitive chemo-rad, curative chemo-rad to the chest involving the mediastinum. So since 0617 highlighted this as an issue, there’s a lot of research going on in that area with lots of kind of cardiac investigations at baseline and monitoring that really trying to define, not only which are the most at risk parts of the heart, but also what the dose constraint should be, and whether it’s realistic, even in the definitive setting. You know, you’re weighing up the risk of cardiac tox over the risk of not controlling an active lung cancer. Obviously that’s different in the adjuvant setting, but we don’t have any new dose constraints as yet to apply. But also the dose to the heart isn’t solely just about dose to the heart. Dose to the heart also means dose to lymphocytes cause your blood pool is processing through it and how that impacts on survival and outcomes. But also the heart is intrinsically linked to the lungs. So cardiac tox and lung tox are kind of interlinked, and that interplay is something we don’t understand terribly well either.
SANJAY POPAT: Thank you Fiona. I mean the other setting, which has been problematic for the UK has been the delivery of concomitant chemoradiation for many years and the National Lung Cancer audit and the Royal College audit have demonstrated relatively low rates of concomitant chemoradiation uptake, for stage 3 lung cancer. Now with the publication of the PACIFIC trial, we now have a long-term data demonstration, a big impact of durvalumab. There’s been a shift in recent years to increase the numbers of patients with concomitant chemoradiation. At the ESMO Meeting, you’re a co-author of the PACIFIC-R data, Fiona, which was the real-world outcome, a retrospective chart review of the outcome of patients treated with durvalumab, that had received other concomitant chemoradiation, or sequential chemoradiation, as this was available for some patients on a compassionate use basis. Do you want to just talk to us about what your data showed and did that compare to what the PACIFIC Study showed and specifically, really the, what happened to the sequential chemoradiation group?
FIONA MCDONALD: So off the PACIFIC, I mean the PACIFIC-R obviously is an enormous study, a large initiative, as you say, based on the compassionate access program that came out almost immediately after the initial PACIFIC data was presented. Over a thousand patients, but only a small proportion, about 169 I think, had sequential chemoradiotherapy in that. So relatively small out of the cohort, but we’re starting to get some safety analysis in those patients. We haven’t got as far as outcome analysis yet, but we will. There was some data from the GEMSTONE study using a different IO agent that allowed sequential patients as well. It came out in ESMO, suggesting that the safety, and at the moment, the early efficacy looks promising similar to the concurrent patients. We’ve also got PACIFIC-6 ongoing. There was no data in the current ESMO that I’m aware of, but there was some data at ELCC earlier this year, looking at safety of durvalumab in the sequential setting, single-arm study observational. So I think we are going to get more data for the sequential patients going forward.
I mean, I think historically we know our rates have been lower than other comparable countries across Europe. And I think there are a number of reasons for that. I mean, they’re always going to be a proportion of patients who are not fit for the concurrent approach, probably a small proportion. And for those we want to know what the role of durvalumab is for the less fit patients. There are also a really important group of patients who just have too large a volume. I’ve seen two within the last three week, last three weeks where I cannot get the volume in up front, but they are otherwise fit enough, PD-L1 status, all the other markers are there for them going down that pathway. And at the moment, if we need enough shrinkage with the chemo upfront, we can’t offer them durvalumab.
So there’s, you know, there’s two populations of patients we do need the data for. I think what PACIFIC may have done in the UK, you know, there was an argument for saying maybe it was, you know, clinicians were concerned about concurrent. Maybe we’re more cautious with toxicity than some of our European colleagues, possibly some pathway issues. You know, maybe because we have, you know, the Med Onc, Rad Onc split, people prefer to give chemo in peripheral centers, move into radiation afterwards. So hopefully with the introduction of PACIFIC, specifically with concurrent, I’m hoping some of those pathway issues and clinician experience will have changed. So it could possibly be beneficial if the initial data was only in the concurrent setting as far as generally improving our concurrent rates. But I still think we need that data in the sequential rates for the patients who still won’t be suitable for concurrent.
SANJAY POPAT: The data that was presented by Dr Wu from the GEMSTONE study was with a PD-1 called sugemalimab, which is being further developed with Pfizer. You know, it wouldn’t surprise me if, you know, given that this is a positive study, this ultimately does result in regulatory approval. And as you say, PACIFIC-6 is recruiting and with the preliminary results we’ve got with PACIFIC-R, with GEMSTONE-301, you know, if you’re a betting man, you’d suspect that you’d have a positive benefit with durvalumab after chemoradiation, or at least with IO after sequential chemoradiation. Is this gonna undo years of effort that the UK Radiation Oncology Community have been doing to try and make sure that patients get concomitant chemoradiation by just taking the foot off the brake and saying, well, that’s fine, we can give sequential chemoradiation, and just allow immunotherapy to mop up the disease there afterwards. What’s your view of the future, Fiona?
FIONA MCDONALD: Maybe I’m more optimistic than the picture you’ve just painted. I don’t see why it should. I think if we’ve all got used to getting more concurrent and we’re all happy with the toxicity profile, we’ve always known the concurrent, the adjuvant IO space, we’ve always known that concurrent provides the better results than sequential. So, like I say, the important group of patients who aren’t suitable, I’d like to think that if these studies pan out they are accessing IO as well, but I would hate to think that if we move away from delivering the optimal standard of care and move away from concurrent, just because we’ve got some data on sequential.
SANJAY POPAT: Great, I think we can keep an eye on that space and see how that evolves. But you know, my take on this space is that immune checkpoint inhibitors following radical chemoradiation of whatever ilk are now confirmed, firmly embedded and I’m sure their utility will continue to be detailed over the course of the next few years, as trials read out.
SECTION 3: Resected NSCLC & molecular testing
SANJAY POPAT: I’d like to bring in Riyaz, my colleague to the conversation, just focusing still in the radical setting, with non-small cell lung cancer, there’s a lot going on in resected non-small cell lung cancer. We’ve heard about the role of adjuvant radiation from Fiona with the lung ART data. A year and a half ago, Dr Herbst presented at the ASCO update, the long awaited ADAURA clinical trial data in patients with EGFR-mutant resected non-small cell lung cancer with a phenomenal relapse-free survival hazard ratio of about 0.2, 0.19. And the first of the trials in the adjuvant setting was presented at ASCO.
The IMpower010 trial, looking at the role of adjuvant atezolizumab. This was a randomized phase III trial in which patients with stage 1B to 3A non-small cell lung cancer, who’d all undergone cisplatin-based platinum doublet chemotherapy for four cycles were then randomized to receive either one year of atezolizumab, or placebo with a primary endpoint of disease-free survival in a number of different datasets, according to their hierarchical testing strategy. Overall, we saw a strong disease-free survival hazard ratio of 0.43 in patients that were PD-L1 at 50% or more. That was further represented at ESMO.
And Riyaz, I wanted to know what your thoughts are on that dataset. I mean, for me, more than 50% TPS score resected non-small cell lung cancer, stage 2 to 3A, hazard ratio 0.43, that’s a pretty strong hazard ratio. And I think that’s clinically meaningful, but you know, what’s your take on that data?
RIYAZ SHAH: Well, I mean, it’s obviously the first of several trials to read out and it’s important to remember that durvalumab, pembrolizumab, and nivolumab are all currently being tested in a randomized setting. The big difference with all of those trials is their plus or minus chemo and IMpower010 mandates adjuvant chemotherapy. So this is the only trial that I know of that’s a big randomized registrational level study that mandates chemotherapy before randomization to one year of IO or not.
I think in the onco community, there is a huge philosophical debate going on at the moment, kicked off by ADAURA and only stoked, this fire has simply been stoked with these adjuvant trials about. Well, what outcome is the right outcome in an adjuvant trial? You know, when I was training, we classically thought of overall survival. We classically thought that you give an adjuvant treatment to cure more people. So is disease-free survival an appropriate surrogate? So, I think there’s a big existential discussion that needs to be had. And because all these trials are reporting DFS outcomes, OS will come at some point, but we’re still years away.
On the face of it, the IMpower010 trial is a very straightforward, well conducted study done to registration standard. It shows a very clear benefit in the overall population for the addition of IO. You know, in the overall population, the hazard ratio is 0.79, and it’s significant, the confidence interval doesn’t cross one. But when you see the PD-L1 subgroups, you can very clearly see an extremely strong benefit to the greater than 50% of patients. Now, the hierarchical design of this study is very important and I think it’s important for the viewers to understand that these designs are very common now in complex pharma-funded studies and they’re structured around the chances of getting a registration from a regulatory authority. So the original trial, the original first endpoint is PD-L1-positive stage 2 to 3A. Now the trial includes stage 3, stage 1B to 3A and it includes patients of all PD-L1. But the primary initial hierarchical design was to just look at greater than 1% PD-L1 stage 2 to 3A and that’s positive. Then when you look at all PD-L1 in the same stage 2 to 3A, it’s still positive for DFS. Then the next one was to include the 1Bs. And although it’s positive, statistically in that group, you didn’t meet the pre-specified endpoint of the actual trial design.
So I think it is likely that the licensing will be in PD-L1-positive stage 2 to 3A. That would be my hunch. I don’t actually have any hard data to support that, but my onco hunch is that that’s the group we’ll go in. And then that means that basically PD-L1-negative patients, we will just accept that this doesn’t have efficacy in them. And I think that’s a fair thing to say because the point estimate in the forest plot for that is literally at one. So I don’t think it works in PD-L1-negative patients. The PD-L1 1 to 49 is interesting. I mean, in the O, I mean, at the end of the day, the point estimates are not that great. They’re just to the left of one, just, and the confidence intervals are wide and they cross one and it’s not significant. So I think there’s a reasonable argument to say that the data supporting its use in 1 to 49 is very weak, indeed. Although it will more than likely get licensed in that setting. And it more than likely as a result of license will get through on a project all based type approval. So we will see a blue tech that says, this is an approved treatment in any PD-L1 greater than 1%. So I think we really, we need to have a discussion, I think, as oncologists as to, well, what do we think of this actual data? Are we.
SANJAY POPAT: Yeah, I think that’s right. I mean, the insofar as the benefit is very much driven by the PD-L1 more than 50% with a hazard ratio of 0.43 and the 1 to 49% in the 2 to 3A population, the hazard ratio for DFS was 0.87, which is really not as impressive as you would like for DFS as an early endpoint. So I think that is really where the debate is as to whether that really is clinically meaningful or not.
I’m gonna bring David into this discussion because really these drugs are going to come through. We already have adjuvant osimertinib, which is now funded by NHS England, at least whilst NICE decide whether they want to continue funding it or not. And Riyaz is quite correct, cause I’m sure we will have MHRA licensing for the 1% and above and whether the reimbursement for the more than 50% or greater, you know, less than 50% comes through. I think we’ll just have to see what the reimbursement strategy is for the UK. But, you know, are MDTs geared up for PD-L1 testing and EGFR mutation testing in the radical setting and following up on that result to make sure that patients don’t miss out in seeing an oncologist, because at the moment it may be, well, the patient has had a stage 2 tumor resected. There may not be fit. Many MDTs would say, well, they don’t need to see the oncologist and that patient then just gets followed up. Whereas you don’t necessarily have the EGFR mutation result back yet and it could be that that patient may benefit from osimertinib or it may be strongly PD-L1 positive, may benefit from atezolizumab. How is that going to work out in day-to-day MDT practice?
DAVID BALDWIN: That’s a good question. I mean, I think I’d like to say that it ought to be covered by the process that has been set up because the peer review processes, as you know, asks about whether you discuss your post-operative patients in your MDT, which you should be doing, everyone should be discussing their post-operative patients. And that’s the time when you pick these things up, and you implement guideline-driven management.
SANJAY POPAT: – I guess that’s true and I hope that that’s the case. The challenge will be that the molecular results usually take two to three weeks after the request has gone off to get some sort of result. Now of course, you may, there may be a lag time of two to three weeks between surgery to present that case. In which case you can present everything as a fait accompli. But you know, these days with VATS surgery, very quick recoveries, we’re discussing our patients far quicker postoperatively than before. I think that the message is we need to get the full picture of the tumor, which includes the moleculars. Is that a message that’s resounding with the respiratory community David?
DAVID BALDWIN: Well, I think it’s one of those things that is, it needs the education element. Which BTOG provides very ably and I think the sort of thing that needs to be discussed much more as the evidence comes through. But, you know, as more and more treatments become available, this will become a key factor in how we, the messages we send it back to the various MDTs and, you know, it’s not always easy for the peripheral MDTs, where the surgery has been done at a big center and they’re discussing the post-op patients and whether that is sufficient and whether they should be done centrally, I don’t know. I mean, the whole argument about rapidity, of particularly the molecular tests. So in PD-L1 is not a problem generally because it’s done locally, but nowadays with the genomic centers, we’re really running into trouble about turnaround times. This all needs to be in the mix. And it’s helpful because you know, it’s another argument to say that the genomic centers have got to get their act together a bit more and have this and get rid of this three-week turnaround time that they’re talking about at the moment for simple, relatively simple tasks, like EGFR testing. But it is a problem and it does need addressing and I think we’ll be on to that.
SANJAY POPAT: Thank you, David. Riyaz.
RIYAZ SHAH: Yeah. I think one of the issues that often gets missed is, as David alluded to patients, there’s a central surgical MDT. There may be a local regional local MDT. The biopsy may be, many of these patients have a pre-diagnostic, pre-surgical biopsy that if the center does reflex testing, the unit for example, will have been reflexively tested. So patients often get presented at MDMs with a molecular result and no one really knows cause they’re in hospital A, that hospital B actually did PD-L1 and EGFR and ALK, you know, six, two months ago and those results are not always the same. They can be different. So you could have a different PD-L1 result on a diagnostic biopsy and a different PD-L1 result on a surgical specimen. And so this is a potential minefield of, and it’s very easy to miss patients or patients to miss out on potential therapeutic options.
SANJAY POPAT: So I think that the take home message of this section is really, we just need to make sure that as the MDT, we’ve got our processes set up to make sure that the biomarkers are fully captured.
SECTION 4: POSEIDON & HER2-mutated NSCLC
SANJAY POPAT: Riyaz, I wanted to come back to you and just focus on the metastatic setting at well lung and the presidential setting session, we had the POSEIDON, long-awaited POSEIDON trial results. This was a first-line metastatic trial across all histologies and stage 4 disease of histology-specific chemotherapy, histology-specific chemotherapy plus durvalumab, followed by durvalumab maintenance, or the same together with tremelimumab, durva and treme together with chemotherapy with maintenance was only one treme followed by durva maintenance. The primary endpoint of overall survival was only positive after hierarchical testing in the durva/treme plus chemo group. And then even in that, the benefit was only seen in the non-squamous population. So we have a trial where you have a positive outcome in non-squamous patients with combination chemo, durva and treme. How does that fit into the UK when we have so many options? Is this a meaningful option for our, for the UK?
RIYAZ SHAH: Yeah, I mean, the thing about the POSEIDON Study is it brings something to the table and what it brings to the table is a well-conducted large registrational study, where there’s an arm having chemotherapy, there’s an arm having a checkpoint inhibitor with chemotherapy, and then there’s an arm having dual IO CTLA-4 with checkpoint inhibitor plus chemo. And there aren’t really any trials that have done that in such a clean way. So I think it does add academic information.
We do have some data already for the use of CTLA-4 combined with checkpoint inhibitors. So we’ve got the CheckMate 9LA study and the CheckMate 227 study. The CheckMate 227 study is one of the biggest challenges I think, in thoracic oncology to understand. A very difficult to understand study, very complex presentations that have only ever presented subgroups within the trial. So I think what was nice about this as it just upfront randomizes patients as three arms and just presents the data, you know, as it is and what we see as exactly as you’ve alluded to. Is this of relevance to the UK? I mean, I think it’s difficult because we’re already in a very crowded space and I think KEYNOTE-189 and KEYNOTE-407 have really established chemo/IO using pembrolizumab as a recognized standard of care, we accept that there are multiple standard of cares, but once you get used to using a particular standard of care, there is really little incentive to change unless something is superior to that. And I don’t think there’s enough mature data to show anything long-term. I think as you say, there’s very little difference between single agent and dual IO and it seems to be restricted to the non-squamous patients. So I, having looked at POSEIDON, don’t lose sleep over treating my patients with standard chemo IO without CTLA-4. That for me is the message. It’s an interesting study. I like the fact that it was presented with just the raw data in a very understandable way, but it doesn’t really change treatment paradigms I think for us.
SANJAY POPAT: Yeah.
RIYAZ SHAH: For now.
SANJAY POPAT: And non-small cell lung cancer is becoming increasingly split into different molecular subgroups and we’re having more and more drugs for each of these different molecular subgroups. One of the most interesting subgroups is the HER2-mutant subtype of non-small cell lung cancer, particularly seen in never smokers with adenocarcinoma. There’s a new group class of drugs, the antibody-drug conjugates, which are being developed with third generation technology dropping off their toxic payload near the tumor cells. Perhaps one of the most interesting trials presented at ESMO was the HERTHENA-Lung01 one trial of trastuzumab deruxtecan which is an ADC targeting HER2, which my colleague, Bob Lee presented. This is a single arm phase II trial demonstrating a response rate of 55% in relapsed patients, median progression-free survival of 8.2 months and median overall survival of 17.8 months in patients who already had relapsed disease, which is really quite astronomical. Riyaz, can I have your thoughts on that data? I mean, for me, that’s basically data, which we should be using frontline. What are your thoughts that sort of data from T-DXd?
RIYAZ SHAH: Well, I think we’re all waiting to get access to this drug, frankly. I mean, with NGS technology coming to the fore, we’re increasingly picking up HER2 exon 20 insertion patients. I have a whole string of them actually under my care at the moment. And they’re being treated on conventional treatment pathways at the moment. There is a very small amount of data for T-DM1 in this group and that data doesn’t look great. So this data was very exciting and you know, it’s not currently available. It’s used in the breast cancer scenario, where I believe it may even be licensed. I’m not entirely sure. But to me of all the potential things targeting HER2 exon 20 insertions, this would be a thing that I would try and hold out for, for my patients. I very much hope a Named Patient Program becomes available so that we can start getting experience.
SANJAY POPAT: One of the main challenges with this drug is the pneumonitis risk and 25% of patients had any grade pneumonitis in this study and the pneumonitis didn’t necessarily start early. It occurred at any time point, perhaps there were some cases late pneumonitis and looking at clinical features that didn’t seem to be near obvious clinical features associated with risk of pneumonitis. These are mainly never smokers remember with good lungs going into this. So David, are we as a respiratory community already geared up to treating pneumonitis from checkpoint inhibitors, or is this going to add more burden to respiratory colleagues?
DAVID BALDWIN: Well, it undoubtedly will add more burden, particularly with the, you know, the post-COVID backlog of people that are requiring to be monitored at the moment with CT, with ongoing interstitial lung disease. And again, the uncertainty about whether treatment makes any difference. Often the time, as you know, with the pneumonitis, it’s trying to make sure there isn’t something else going on first before you then weighed in with them with immunosuppressant drugs or anti-inflammatory drugs. So I’m confident that most respiratory specialists can tackle this. It is going to be an extra burden.
One of the things to perhaps ask you back, actually, because as someone who’s on the sort of outside of this, in the sense of all of the very clever combination of drugs that are going on at the moment, there’s no doubt in my mind that quite often, there’s a fairly marginal difference between agents, individual agents and you know, that you just described a study, where dual IO – I was particularly underwhelmed with that study. So I’m glad I interpreted it correctly, but, you know, a lot of it can, it’s the way you manage the whole patient. And if you’re very experienced with one individual drug, as an oncologist, you will know a lot more about how to, you know, maybe alter drug regimes or whatever, to tailor it to the individual and get the best possible outcomes. And that could actually outweigh any advantage in going for a marginal benefit. How do you propose to sort this out because it’s getting really complicated and we’re just starting in London, I think aren’t, we really?
SANJAY POPAT: I think that’s right, David. I think that our treatment decisions are very complicated now, and this is part of the great education that we need to do across the UK as you’ve rightly said, across other disciplines to look at cases, to look at tips, hints, and tips of what works and what doesn’t work. And you know, where we’ve had success with one particular type of practice where another type of practice would have done something different. And part of this is adverse event management, understanding what the adverse events are, picking them up early and instituting the appropriate actions.
Chaps, we’ve been discussing a lot. We’ve had great discussions. We could go on for another hour, but I’d like to thank all of you for your learned input and your great discussions. David, Fiona, Riyaz, thank you very much for your time. And on behalf of Video Journal Oncology, thank you for watching. And I do hope you found this useful at, and I hope to be joining you again for further discussions at another time point. All the best, bye-bye.