ASCO 2024: Highlights and analysis

Share :
Published: 4 Jun 2024
Views: 2413
Rating:
Save
Assoc Prof Bishal Gyawali - Queen's University, Kingston, Canada

Associate Professor Bishal Gyawali talks to ecancer reporter Saira Ahmed about several presentations from ASCO 2024; on topics such as patient-centric trials, lung cancer, melanoma, GI cancers and breast cancer.

The presentations referred to are:
 

Commentary on ASCO 2024's theme
Patient-centric trials at ASCO 2024
Osimertinib after definitive chemoradiotherapy (CRT) in patients with unresectable stage III epidermal growth factor receptor-mutated (EGFRm) NSCLC: Primary results of the phase 3 LAURA study
ADRIATIC: Durvalumab (D) as consolidation treatment (tx) for patients (pts) with limited-stage small-cell lung cancer (LS-SCLC)
Lorlatinib vs crizotinib in treatment-naive patients with advanced ALK+ non-small cell lung cancer: 5-year progression-free survival and safety from the CROWN study
KRYSTAL-12: Phase 3 study of adagrasib versus docetaxel in patients with previously treated advanced/metastatic non-small cell lung cancer (NSCLC) harboring a KRASG12C mutation
Sacituzumab govitecan vs docetaxel in patients with metastatic non-small cell lung cancer (mNSCLC) previously treated with platinum (PT)-based chemotherapy (chemo) and PD(L)-1inhibitors (IO): Primary results from the phase 3 EVOKE-01 study
Neoadjuvant nivolumab plus ipilimumab versus adjuvant nivolumab in macroscopic, resectable stage III melanoma: The phase 3 NADINA trial
Prospective randomized multicenter phase III trial comparing perioperative chemotherapy (FLOT protocol) to neoadjuvant chemoradiation (CROSS protocol) in patients with adenocarcinoma of the esophagus (ESOPEC trial))
Chemotherapy and liver transplantation versus chemotherapy alone in patients with definitively unresectable colorectal liver metastases: A prospective multicentric randomized trial (TRANSMET)
Trastuzumab deruxtecan vs physician's choice of chemotherapy in patients with hormone receptor-positive (HR+), human epidermal growth factor receptor 2 (HER2)-low or HER2-ultralow metastatic breast cancer (mBC) with prior endocrine therapy (ET): Primary results from DESTINY-Breast06 (DB-06)
First-line inavolisib/placebo + palbociclib + fulvestrant (Inavo/Pbo+Palbo+Fulv) in patients (pts) with PIK3CA-mutated, hormone receptor-positive, HER2‑negative locally advanced/metastatic breast cancer who relapsed during/within 12 months (mo) of adjuvant endocrine therapy completion: INAVO120 Phase III randomized trial additional analyses
Prognostic utility of ctDNA detection in the monarchE trial of adjuvant abemaciclib plus endocrine therapy (ET) in HR+, HER2-, node-positive, high-risk early breast cancer (EBC)

 

ASCO 2024: Highlights and analysis

Assoc Prof Bishal Gyawali - Queen's University, Kingston, Canada

Saira Ahmed – ecancer reporter

Hi, and welcome to ecancer’s highlight and analysis video from ASCO 2024. We have Professor Bishal Gyawali here today to discuss these highlights with us. Thank you Professor.

Thank you very much for having me again.

Thank you. So my first question to you is what are your thoughts on this year’s ASCO theme?

Yes, I think I’m very happy with this year’s ASCO theme because the theme this year was From Cure to Care. We usually during these big meetings focus on cure, that means newer drugs, newer innovations but we should also not forget that only those innovations help that can actually reach the patient. You might have fancy innovations but if patients can’t access it then they are not going to save lives. So this year’s focus on caring for the patient, you might not always be able to cure it, especially in oncology, but you can always care for the patient. So I really love the theme.

In the opening ceremony the President made an inspiring speech about why her presidential theme focussed on caring in relation to curing for the patient. The ASCO CEO, Dr Cliff Hudis, in his opening speech, he made a remarkable statement, giving a shout-out, actually, to cancer groundshot, a philosophy that we have been advocating for for a long time. I think it was in ecancer that I wrote about cancer groundshot for the very first time back in 2016. After that we have been talking about cancer groundshot in a number of venues – two years ago ASCO had provided us an entire session on cancer groundshot – but it was really inspiring to hear from the ASCO CEO this year in his opening ceremony remarks why we need to also focus on cancer groundshot type interventions in relation to cancer moonshot type interventions. So these should go hand in hand.

The President, Dr Lynn Schuchter, in her presidential speech during the opening ceremony, she of course emphasised on the importance of caring for the patient, but she also highlighted the importance of communicating with the patient very well. Because one of the important components of caring is proper communication and we are not doing as well as we think we are doing in terms of communicating about diagnosis, prognosis, treatment, everything with the patient. So one of the examples that she gave was when we say a drug has 20% response rate, patients usually hear or understand as 20% cure rate which is not true. So, similarly we talk about progression free survival all the time; when we say progression free survival patients assume it means survival which is not the case again. So there is a problem with communicating with the patients. So I was very inspired with not only the presidential theme but especially with the opening ceremony that happened at this year’s ASCO.

In fact, congruent with that theme, if you look at the trials that were presented in the plenary session, the majority of the trials that were discussed at the plenary session were especially focussed on how we can improve care for our patients. One of the examples that I want to give is the third plenary abstract which is a trial that, among patients with lung cancer, compared the delivery of palliative care through telehealth versus in-person care. It is a very patient-centred trial for delivering palliative care – do you actually need to come to the centre in person or you can do it remotely by using telehealth from home. The primary endpoint for this trial was quality of life as measured by the FACT Lung score at 24 weeks. This trial was an equivalence trial, so this trial showed that delivering palliative healthcare using telehealth was equivalent to delivering palliative care for patients with lung cancer through an in-person visit. So that means we don’t need to trouble our patients and ask them to come for an in-person visit all the time which is a patient-centred trial and I’m glad that it went into plenary.

But there are a couple of caveats about this. One is that because it was an equivalence trial they had to enrol more than 1,200 patients into this trial, which is a huge number of patients. It takes more money, more resources, to conduct such a big trial. But my opinion is that you don’t even need to do an equivalence trial, you could have done a non-inferiority trial. Because if the assumption is that patients can get this treatment virtually, it’s palliative care treatment virtually, do we actually need to show that it is equivalent to an in-person visit because we know that certain aspects of patients’ quality of life are obviously going to be improved? For example, what we call time toxicity – the fact that they don’t have to travel all that distance, come, find parking, pay for the parking, that time is saved and that in itself is a meaningful outcome. So I think if we had done just a non-inferiority trial it could have required a lesser number of patients and so the trial would have been less intensive. That’s not a criticism on the outcome of the trial but it’s a criticism on the design of the trial because we could have shown the same thing with lesser resources.

The other thing is I had assumed that it would not just be equivalent but it would in fact be superior. I assumed that patients’ quality of life, if they got virtual visit for palliative care instead of in-person, it could even be superior. But why did we see equivalence and not superiority is probably the tool, that FACT Lung tool, that we used to measure the quality of life for these patients, it was not actually designed for this, it was designed for drug interventions. So it’s probably not capturing all these dimensions that matter to patients but is not captured by this tool.

There was actually a complementary trial to this, another trial of palliative care, but this was called a stepped-up approach to palliative care. Again, patients with lung cancer, instead of seeing them regularly, seeing them once after diagnosis and then seeing them again when required and that is when the treatment changed or when they got hospitalised. This is an important trial because we usually say patients need to get palliative care early but we should also be mindful of the resources that it takes, the healthcare resources – palliative care physicians are limited in number and they are busy. So if we are just sending all the patients to palliative care all the time, even those patients who are otherwise fine and probably do not need such frequent palliative care interventions, then can we just send them once so that they get in touch with the palliative care team and then send them again in the future when they need it, rather than sending them frequently. So this was, appropriately, a non-inferiority trial and it proved non-inferiority. So this was also a good result.

Thank you. One question for you, you said that you’ve seen more patient-centric trials at this year’s ASCO. Do you think patient advocacy has a role to play with this or are there any other factors?

No, to answer your question, in general patient advocacy absolutely has a big role to play but I am not exactly sure whether what we saw in this year’s ASCO is directly a result of patient advocacy. I can’t say for sure. But this is congruent with the presidential theme. So I think it also matters who the President is for that particular year and what the President’s vision is for ASCO and the oncology community as a whole. So this year’s President, Dr Lynn Schuchter, her vision was about caring for the patient, about palliative care, about supportive care, in addition to curing for the patient and that’s why I think it mattered. Some years we have seen that the theme is innovation or breakthrough and in those years you see that most of the plenary session abstracts are all about new shiny molecules, new innovation.

So do you have any other plenary lung cancer abstracts that you’d like to talk about?

Yes. This year the ASCO was almost like ASCO Lung, there were plenty of lung cancer trials, very important trials, presented. If we focus on the plenary lung cancer abstracts, I think the most controversial was the LAURA trial which is a trial of osimertinib versus placebo for patients with stage 3 lung cancer after chemoradiation among patients who have EGFR mutation positive. First the results, the results were super impressive – the median PFS was 39 months versus 6 months with a hazard ratio of 0.16 which is one of the best results that we have seen. But, having said that, why I mentioned that it is controversial is for a couple of reasons. Of course these results are super impressive but if you look at overall survival it’s not mature yet, it’s only 20% mature at this time but overall survival is not significant.

So that brings us to our discussion that we do every single year because it’s the same problem is PFS enough, should we be looking for OS? In fact the discussant made an argument that OS is not required, PFS should be enough because if these patients progress then they will have several quality of life issues, they will have symptoms. But there is a flipside to this argument. One is that all PFS events are not created equal. There can be some PFS that is not symptomatic at all; there can be some progression that is very much symptomatic. How do we delineate that? Either we could look at symptomatic PFS instead of just looking at PFS or we could measure quality of life and show that the quality of life throughout actually improves compared to not giving this drug. Because if it actually leads to these symptoms then that means it leads to detrimental quality of life, that means it should be captured by our quality of life tool. So instead of making an argument that because PFS has improved quality of life is better, we need to actually prove that quality of life is better because we have shown that PFS is not a good surrogate for quality of life.

But the second argument is here there is no fixed duration of treatment, unlike other adjuvant or curative setting treatments. In this trial osimertinib was given indefinitely and I’ll tie my other argument with this indefinite treatment. The other argument here is that because if you look at the placebo arm patients the placebo arm patients are doing quite poorly. Usually for stage 3 you see that there is some percentage of patients who actually do well without any further treatment. There is some percentage of the patients that are supposed to be cured without any further treatment after definitive treatment with chemoradiation. But here you see that they are doing quite poorly and the fact that PET scan was not mandatory for patients to enrol in this trial, that means it makes us question whether these patients were already sub-clinically metastatic to begin with. Because if these patients were not stage 3 and if they were stage 4 to begin with and we did not stage them properly and we enrolled stage 4 patients in a stage 3 trial, then basically this becomes a trial of not stage 3 therapy but this becomes a trial of early versus delayed treatment for stage 4 patients.  We know that for stage 4 osimertinib is the standard of care, we already know that – osimertinib is the best treatment for EGFR positive stage 4 patients. So the biggest concern is whether these patients have been under-staged. This is an important question because if we are going to treat them indefinitely throughout their life, then is treating all of them at stage 3 important or can we only treat those who become stage 4 when they become stage 4? Because overall survival is the same. If you are not gaining any overall survival but just increasing the length of treatment, that’s a substantial amount of time for anyone to be on a treatment that has diarrhoea, that has rashes, that has all these side effects, for no improvement in survival. So survival is going to be important.

When we are talking about survival in these trials we need to talk about crossover. So in this trial actually the crossover rate was not unsatisfactory like we have seen in several other trials. We criticised the ADAURA trial a couple of years ago but in this trial there was an 80% rate of crossover which is a satisfactory percentage. In some other trials we have seen in a similar setting more than 90% crossover rates so there is scope for improvement but 80% is not bad. So the inference is if OS, now it’s only 20% matured this time, if OS is not significant in this trial, then can we definitely say that for a trial with potentially under-staged patients based on PFS alone, should we change our practice, given the fact that we are going to use this treatment throughout their life with time toxicity, financial toxicity, all those side effects, just based on PFS without OS or without quality of life? So that is my concern with this trial.

What do you think is the future for this trial then? What does the future look like?

If I have to predict, of course it will get approved for this setting, people will start using it. That’s the way oncology has been working. But I think as a community we need to ask these critical questions and not just jump into conclusions without asking these questions. If you talk about policy, several countries in the world can recommend them that, okay, you should spend your limited resources to give osimertinib to everyone based on this trial result. I think no, for many countries in the world these are potentially under-staged patients and if OS is the same then of course they can save that treatment for patients who become stage 4 rather than giving it to everyone when they are stage 3. So there are bigger policy questions.

Irrespective of which specialist are and when you become a specialist in a narrow field, when you become an EGFR expert in the world, then you look only at EGFR and EGFR is your whole universe. So for a hammer everything looks like a nail. But as a physician we are taking care of the entire patient and we are also responsible scientists and we are also responsible policy makers. So when we think in terms of larger population-level outcomes we can’t be thinking about just that one narrow field and celebrate every little thing.

Moving on to another lung cancer plenary, there was the ADRIATIC trial of durvalumab as a consolidation treatment versus placebo in limited stage small cell lung cancer. Here I have only one criticism is that the patients in the placebo arm, they apparently did not get durvalumab when they progressed so the crossover was apparently allowed but it was pretty low, the crossover rates, from what I understand. It has not been published so I don’t know the details but the crossover rate seems to be very minimal. Except for that one caveat, actually I’m quite happy with these results for limited stage small cell lung cancer. Small cell lung cancer has not had too many good treatments for a long, long time and here we are seeing median OS improvement of almost two years which is quite a remarkable achievement. We usually don’t get to talk about OS improvements in terms of years, we usually talk in terms of weeks, sometimes months, but here we are talking in terms of years which is a pretty good result. But I’m disappointed that there was not a higher rate of crossover.

This theme comes again and again and I have written a paper about it in Nature review Clinical Oncology last year immediately after ASCO because this was a common theme last year at ASCO as well, that patients in the control arm were deprived of the standard of care when they progressed. Some of these trials, what they say is crossover would be allowed, crossover would be permitted, but that does not improve crossover rate. We need to mandate crossover. If you say crossover is allowed then the onus is on the physician and the patient to look after these drugs and find them. But as a sponsor if you can provide these drugs to the experimental arm, you can also provide them to the control arm. You should mandate crossover so that we don’t have to talk about, ‘Oh, this is only 20% crossover, is it enough or not?’ The crossover would be the best possible crossover rate. Of course it won’t be 100% all the time because some patients drop off, some patients develop toxicity. But crossover should not be just allowed or permitted, it should be made mandatory in these trials.

Would you like to discuss the CROWN trial?

Yes, there were other important lung cancer trials at this year’s ASCO outside of the plenary. So the CROWN trial, it’s not a new trial; the CROWN results have been published before. But what was new at this ASCO is the five-year results and the five-year results actually are pretty impressive. So, for background, this is a trial of lorlatinib versus crizotinib, both are ALK inhibitors. The median PFS was not reached after five years so this was an unprecedented milestone for advanced non-small cell lung cancer to not have a median PFS reached after five years. So this is a fantastic milestone but the problem here is the control arm, the control arm is crizotinib. Crizotinib is an old ALK inhibitor that has been beaten so many times, it’s like beating the same kid again and again. There was the ALEX trial, alectinib beat crizotinib very well and became the new standard of care, but we still ran a trial with lorlotanib versus crizotinib. Even brigatinib had beaten crizotinib. So alectinib has beaten crizotinib, brigatinib has beaten crizotinib, how many times should we beat the same kid again? It’s not even ethical. When you know that crizotinib is bad and there are other better drugs out there, how can you consent a patient in a trial and say you can get randomised to get crizotinib which we already know is very bad?

As I said, I can only express disappointment; unless the whole community takes a stand and voices concerns against this, it’s not going to change. The IRBs need to change the approval, the regulatory agencies need to take a step, the health assessment bodies need to take a step, the zonal entities need to take a step. But we physicians have a big moral responsibility; if we start celebrating everything then how are things going to change? They are not. Yes, this is an amazing result, median PFS not reached after five years, wow, but poor crizotinib, poor patients in the control arm – these are actual lives. So I think this is an amazing result but I also think that the control arm was unethical.

And what are your thoughts on the KRYSTAL trial then?

The KRYSTAL trial was the trial of the KRAS inhibitor adagrasib versus docetaxel. It’s an open-label trial so it’s not placebo controlled and it’s against docetaxel as a last line therapy in non-small cell lung cancer. These people die, they have very poor prognoses. So even in such a setting are we happy with PFS as the endpoint? We should not be. I understand getting a PFS endpoint in a first-line setting and arguing that it takes a long time to measure OS but over here it does not take such a long time. These are patients with very poor prognoses and we should not settle with PFS. And in an open-label trial against docetaxel when we already know that sotorasib, another similar KRAS inhibitor, actually had similarly fake positive signals with PFS but it could not improve OS.

So the randomisation here was 2:1, I don’t think that was the right randomisation equation, it should have been 1:1. Now, there are reasons why I think that and it might be a little too much discussion for today but I think that shows that investigators have already decided that this drug is going to be better, which is not the case. So I don’t think this is a practice-changing trial at all, it should not be called a practice-changing trial.

Then there was the EVOKE-01 trial of sacituzimab govetecan versus docetaxel, again docetaxel is the control arm. Actually for all these trials some people argue that docetaxel is not the right control arm because people argue that docetaxel plus ramucirumab should be the control arm. I don’t think that’s the case because ramucirumab is a pretty low magnitude of benefit drug. Like when the ramucirumab trial was published we argued that it is a very bad drug and it should not be the standard of care. So therefore docetaxel alone as the control arm, I’m okay with that. I’m not saying that this control arm is bad. We can’t have it both ways, you can’t say that ramucirumab is a bad drug and it should not be the standard of care and then you can’t say well why is it not in the control arm? It’s okay ramucirumab not being in the control arm.

But the problem with this trial is the reporting of the trial. There is so much spin in the reporting of the trial, in its press release, in its publication. OS was not improved, that’s the end of the story. If your OS is not improved you should say, ‘We could not improve survival.’ But there are all sorts of things to not say that. So what they ended up saying was, ‘Although it was not statistically significant there is still some numerical advantage,’ and then some random subgroup analysis saying that, ‘Probably in this particular subgroup there is lots of advantage.’ No, that’s not how science works. Anyways I guess that’s all about lung cancer trials.

We can talk about the rest of the plenary trials like in melanoma. Let’s talk about good trials.

Yes, moving on.

So in melanoma the results of the NADINA trial were presented for stage 3 melanoma. This is a randomised trial where they randomised the control arm to receive the standard treatment which is surgery followed by one year of nivolumab and in the experimental arm patients received two cycles of nivolumab plus ipilimumab and after receiving those two neoadjuvant cycles they got surgery and after surgery if the patients had a major pathological response they got nothing. So those two cycles before surgery, that was the entire systemic therapy they got. But if they did not have a major pathological response they continued to receive nivolumab to complete one year of nivolumab. The primary endpoint here was event free survival and it’s still a very short follow-up time, it’s only 10 months but the one-year EFS rates were 84% versus 57% and this was significant.

I think this is a practice-changing trial and I think we need to encourage more of such trials because if we think about all… I was talking about patient-centredness and how some of the plenary trials are patient centred. This is a patient-centred trial because what this allows us to do is for these patients who get a major pathological response after getting two cycles of nivolumab plus iplimumab, that means without this data you would have treated them for one year but now they are treated only for six weeks, just two cycles and they are done. So it saves money, it saves time, it saves all the side effects, it saves the whole therapeutic burden. These are the types of trials we want to see more.

There are still a lot of remaining questions. For example, does it have to be nivolumab plus ipilimumab because when you combine nivolumab and ipilimumab together the toxicity rate increases, so could we do only nivolumab and not do ipilimumab? You can de-escalate further for even other patients who did not achieve a major pathological response, should you actually continue with the whole one year or can you stop it after six years? So there are other important questions that can be answered but I think this was a fantastic patient-centred trial and this should be practice changing.

But this also has lessons for other tumour sites because in lung cancer especially we have seen a couple of perioperative trials. There is a trial of nivolumab plus ipilimumab neoadjuvant and it improves EFS, I think OS results are still pending. It improves EFS and similarly it’s just a limited number of cycles. But the other trials, they are perioperative so neoadjuvant plus adjuvant for a long duration and people claim that we should do that long duration treatment when in fact you have evidence that shows that only neoadjuvant alone can improve the outcome substantially. So what does that adjuvant component add? Nobody knows. So lung cancer is the field where this strategy, similar de-escalation trials, should be tested immediately. We have written about it in Nature reviews Clinical Oncology three or four months ago – is the adjuvant portion even adding anything and isn’t there a burden of proof for us to show that the adjuvant portion, giving that adjuvant portion, is beneficial over doing neoadjuvant alone? That evidence we don’t have yet in lung cancer.

Moving on to another plenary trial and this time from GI cancer, the ESOPEC trial. This was also a bit of a controversial trial, especially in the radiation oncology community. This was a trial that compared the FLOT regimen with the CROSS regimen. So FLOT is the regimen of three chemotherapy drugs – 5FU plus oxaliplatin plus docetaxel – and CROSS is neoadjuvant chemoradiation with paclitaxel plus carboplatin. So for adenocarcinoma of the oesophagus these two were equally standard of care so they compared these two standards of care against each other. They showed that the FLOT regimen improved outcomes. The primary endpoint here was overall survival and the median overall survival was substantially improved with 66 months with FLOT versus 37 months with CROSS; the three year overall survival rates improved.

But the controversy here is that the patients who got the CROSS regimen underperformed compared to what we have seen in the original CROSS trial. Underperformed in the sense that first the percentage of the patients who could complete the treatment as planned was pretty poor and, second, the outcomes, this median OS of 37 months, is less than what we had seen for these patients in the CROSS protocol, in the Neo-AEGIS trial, the other trials that employed this regimen. So what do we make of it? Clearly, despite all those caveats we have seen in a phase III randomised trial that FLOT has improved outcomes substantially compared to CROSS so this should be practice changing. But some of the caveats are that because FLOT is a quite intense regimen many patients would not tolerate this. So for patients who can’t tolerate this CROSS is still the standard of care. The other caveat is that of adjuvant therapy when these trials were done, at the time we did not have any adjuvant immunotherapy as the standard of care. Now we have adjuvant nivolumab for patients who have undergone the CROSS regimen. So the unanswered questions are can you give adjuvant nivolumab for patients without a pathological complete response after FLOT? Although that evidence came from patients who had received CROSS, can we do that?

The other question is some of my colleagues from low-income countries made an excellent point here is that many countries don’t even have radiotherapy machines and some of the countries have a limited number of radiotherapy machines and the patient volume is huge. So in that case this provides excellent evidence that you could use the FLOT regimen which does not require radiotherapy. But there are several unanswered questions especially why the patients in this particular trial performed so poorly on the CROSS regimen compared to what we had seen from more than a decade ago.

I think that concludes our plenary but in GI cancer there was one more interesting trial called the TRANSMET trial. So this is a trial for patients with colorectal cancer, advanced colorectal cancer with metastasis in the liver, only the liver, nowhere else. So for these patients this was a trial that compared chemotherapy alone versus chemotherapy plus liver transplantation which is a pretty intense intervention. And the results were amazing in that the five year OS rates were 73% versus 9%, a hazard ratio of 0.16. We have never seen a hazard ratio that good in colorectal cancer. The statistical plan of this trial, they actually attempted to find a survival difference of 40% in five years which is huge and they powered the trial accordingly. They thought they could enrol only 50 patients in each arm and still see such a huge difference.

Now, what that means is if the intervention is this good you actually don’t need too many patients to show a difference. For an intervention like liver transplant it’s a huge intervention, it’s not like giving chemo. On the flip side the argument is that if we are proposing such a big intervention then the OS benefit should actually be this good otherwise why do it? You don’t want to do liver transplantation to have survival improved by three months, six months. If you are doing liver transplant you actually want to see a huge amount of survival difference and that’s what they showed. But the question is how do we find livers? You can’t make liver in the lab and give it by injection. So, yes, liver transplant seems to be…

That’s a big caveat in this.

Yes. So we need to find a way to make livers. Maybe AI will make it, maybe we’ll find more livers through…

3D printing maybe.

3D printing.

Moving on to our final section, do you have any highlights from breast cancer that you’d like to talk about?

Yes, there were a couple of important trials from breast cancer, not in the plenary, which were presented at this year’s ASCO. Now I’m making this joke because every year at ASCO there are at least one or two trials from breast cancer. So this year we had the results from the DESTINY-Breast06 which is a trial of trastuzumab deruxtecan, an antibody-drug conjugate, for HER2 low or ultra-low tumours that are hormone receptor positive and have already received a hormone therapy. It showed a PFS improvement of 5 months, OS is not significant.

There are a couple of caveats here. First, as always, OS versus PFS. Trastuzumab deruxtecan is already approved in second line or subsequent lines so do you need to use it up front, after hormone therapy of course, but do you need to use it up front or can you save it for later lines? If the OS remains the same and it’s just the PFS then you can very well make an argument that we can save it for subsequent lines and we don’t probably need to give it to everyone up front. That’s the correct interpretation for PFS positive, OS negative trials.

But the other caveat is crossover. How many patients got trastuzumab deruxtecan when they progressed from the control arm? Here I think the crossover rate was only 20%, that’s pretty low, pretty poor. That’s why I keep saying crossover should be mandatory – if you say crossover will be allowed or whatever you get after progression, we don’t care, then people won’t get it. So we need to make it mandatory and this should be 80%-90%, why is this 20%?

The control arm was chemotherapy of physician’s choice which is usually okay but in this trial, and some other breast cancer trials we have seen this, they say treatment of physician’s choice but they limit the choices to two or three chemotherapies. I think if we are saying physician’s choice we should just let them choose what they would normally choose without limiting their choices because there are a couple of options that were not included in this choice.

The other trial was the INAVO120 trial in PIK3CA positive patients first line, palbociclib plus fulvestrant control versus palbociclib plus fulvestrant plus inavolisib. Again PFS has substantially improved, more than doubled – 15 months versus 7 months. But we have seen similarly impressive PFS with alpelisib in a similar patient population. We have seen similar, not in the PIK3CA population, but we have seen something similar in the BOLERO trial with everolimus. What I mean by that is PFS improvement, substantial PFS improvement, is seen in other trials as well. So all these trials went on to have negative OS results, therefore even in this trial it’s very likely that this 7 versus 15 months will disappear when OS results come. So we should learn from history, a, and, b, if you look at the graph, the Kaplan-Meier graphs in this trial, there is huge informative censoring. Actually this informative censoring was a problem with the BOLERO trial, it was a problem with the SOLAR trial too. So this trial also has a huge amount of differential informative censoring so that should be accounted for when we are interpreting the results.

The last breast cancer trial that I want to talk about is the post MonarchE trial which is a trial among patients who have progressed on CDK4/6 inhibitors, giving them fulvestrant versus fulvestrant plus abemaciclib. Something was very interesting here. The PFS was significant but the PFS, if you look, investigator assessed PFS was significant by only 0.7 months. But if you look at the independent central review PFS it was significant by a bigger margin, I forget exactly how much. But if you look at the experimental arm PFS it was 6 months based on investigator assessment but it was 12 months based on independent assessment.

So why is there such a big difference based on whether the physician is assessing the PFS or an independent body is assessing the PFS? So we’ll need to look more into the publication but one explanation is this differential censoring in the experimental arm. Because the rules they use to apply this censoring is different. Actually, the discussant of this trial, she discussed about how informative censoring can lead to these differential outcomes but one statement that the discussant made was incorrect, is that overall informative censoring does not affect the interpretation of the results. It does. In fact, informative censoring is one of the most important reasons why we can see a false positive PFS even when there is no PFS signal. There have been some papers that actually prove that even if the drug is totally not, just based on its toxicity, it can show that it has PFS benefit if there is informative censoring.

You should also look at other trials that have been done in a similar space. There have been other trials where they have been testing the same hypothesis – can we give CDK4/6 inhibitors after the patient has already progressed on the same drug? I think that hypothesis has usually failed in several occasions. Like, can you give the same drug after the patient has already progressed on the same drug? We have tested that in bevacizumab – after progressing on bevacizumab can we give bevacizumab again in different tumour sites? After progressing on immunotherapy can you give immunotherapy again? After progressing on CDK4/6 can you give CDK4/6 again? I think even in those supposedly some of the positive trials, the magnitude of benefit is so small and I think the opportunity cost is huge. Instead of using the same product that has already failed, I think it’s far more important to test a new product.

Okay, thank you very much. On this note, thank you for doing this interview with us, Professor Gyawali.

Thank you very much for having me every year.

Thank you, and thank you to our audience for watching this video.