ESMO 2023: Highlights and analysis

Dr Bishal Gyawali - Queen's University, Kingston, Canada

Hello everyone, this is Dr Bishal Gyawali from Queen’s University, Kingston, Canada and it’s my big pleasure today to provide a roundup of the major abstracts that were presented at the ESMO 2023 Annual Meeting. Actually, I’m right now in Kathmandu so it’s really interesting to provide the ESMO roundup video from the Himalayan country Kathmandu where we are here for the ecancer 2023 Research Methods Workshop.

At this year’s ESMO Annual Meeting several important abstracts were presented and I want to collect them into separate themes. The first theme that I want to discuss is about perioperative immunotherapy versus neoadjuvant immunotherapy in lung cancer because several abstracts were presented at this year’s ESMO meeting that talked about the role of perioperative immunotherapy, for example the AEGEAN trial which tested durvalumab, four cycles before surgery and twelve cycles after surgery, and saw an improvement in event free survival and pathological complete response rates; there is no OS yet. The CheckMate 77T trial neoadjuvant nivolumab plus chemo followed by surgery followed by adjuvant nivolumab – again an improvement in event free survival. The KEYNOTE-671 trial – again perioperative pembrolizumab. So all these trials are testing the role of perioperative treatments in the context of resectable non-small cell lung cancer and we are seeing an improvement in event free survival and, in the case of the pembrolizumab data, also an improvement in overall survival.

But the question that we need to ask here is what is the comparative role of the adjuvant component of the treatment here? Because we have data from the CheckMate 816 trial that showed just three cycles of neoadjuvant nivolumab in combination with chemo would improve event free survival. So if just three cycles of neoadjuvant nivolumab will improve event free survival, do you actually need the adjuvant component that all these trials are testing? So I personally think that the added benefit of the adjuvant component of the immunotherapy is probably negligible and we can omit it, but this is a hypothesis that needs to be tested. Our trials are not being done to ask the right questions. We need to have trials that are asking the question of the extra benefit of the adjuvant component of the treatment. So we need to have a trial testing neoadjuvant immunotherapy plus chemotherapy followed by surgery followed by the adjuvant component of immunotherapy, as is being done currently, versus neoadjuvant immunotherapy alone without the adjuvant component. Only those types of trials will give us the answer about what the role of the extra adjuvant component of the treatment is.

This is important to discern because that extra adjuvant component is extra cost, extra toxicities and an extra healthcare burden for the system and for the patients. So I hope that we will have those trials that will answer these questions but, for now, we have data from CheckMate 816 showing that the neoadjuvant nivolumab alone improves event free survival so I’m more inclined to use that in my practice.

The second theme I want to highlight from other lung cancer trials is the trials that show improvement in progression free survival without improving overall survival. For example, let’s talk about the PAPILLON trial which is the trial of amivantamab plus chemotherapy in exon 20 mutation positive non-small cell lung cancer patients. Improved PFS, a very impressive hazard ratio of 0.4 but so far overall survival is not significant. They highlight this by saying that there were 66% crossover rates from the control arm to amivantamab upon progression and probably that’s why overall survival is not significant. For example, in the LIBRETTO-431 trial of the RET inhibitor selpercatinib in RET positive non-small cell lung cancer, again progression free survival improved substantially, more than double, a hazard ratio of 0.46, but, again, overall survival so far is not significant. Again, there is a 60% crossover rate so people are explaining that this crossover has diluted overall survival.

But in both these cases what you can actually argue is if by having crossover the progression free survival benefit vanishes and the overall survival becomes not significant or the overall survival does not improve, that then probably means that you can give the treatment second line and have the same survival. You don’t necessarily need to give these treatments first line, that would be the correct interpretation. Substantial PFS does not translate to OS because of crossover, that means you can give the treatment second line and you do not need to expose all patients with these treatments first line.

The other theme is in cases where there is PFS and absolutely no OS and not even explained by crossover. For example, let’s talk about the MARIPOSA trial of amivantamab plus lazertinib versus osimertinib first line, or the MARIPOSA-2 trial also falls into the same bucket. In these cases it’s very attractive to explain that PFS benefit in itself is substantial and the PFS benefit itself is practice changing, but I don’t think so. If the OS does not improve and PFS only improves and that is not even explained by crossover, that means the drug is not beneficial for the patients because it did not improve survival. The drug should improve quality of life in those cases but quality of life data we have not seen and we do not know. So in the absence of quality of life benefit there is no way to justify using these treatments based on PFS alone when OS is not significant. Not to mention the side effects like these drugs like amivantamab plus lazertinib in the MARIPOSA trial had a 37% rate of venous thromboembolism. 10% of the patients had to discontinue treatment because of side effects so these are not easy to tolerate drugs. So we need to see either OS or quality of life benefit.

There are some other trials that also fall into the same bucket, for example the trial of antibody-drug conjugate datopotamab deruxtecan versus docetaxel in second-line non-small cell lung cancer. Very small, less than 1 month, benefit in PFS, overall survival is not significant. So second-line treatment, less than a month benefit in PFS. This is not at all practice changing, this is not at all an encouraging trial. So we need to stop celebrating these types of results. The ATLAS trial, for example, atezolizumab plus bevacizumab plus chemo versus chemo as second line after TKI in patients with non-small cell lung cancer, again, some benefit in PFS but no improvement in OS. So such types of trials we should stop celebrating. These PFS only trials are meaningful only if they improve quality of life and we need to see data for that.

In the same context I want to add the trial called the ALINA trial which is, again, a lung cancer trial but this is for the ALK positive patients after surgery. This is a trial of adjuvant alectinib versus chemo and this shows, again, an impressive hazard ratio of 0.24 for disease free survival. But I have discussed a lot about the ADAURA trial in the past in the context of adjuvant osimertinib, all those caveats will apply to this ALINA trial as well. We need to know what percentage of the patients received ALK inhibitor when they progressed in order to interpret the results of this trial correctly.

There were a couple of urothelial cancer trials that were very encouraging that were presented at this year’s ESMO. The first is the trial of enfortumab vedotin plus pembrolizumab versus chemotherapy. A very meaningful improvement in PFS as well as OS. The overall survival nearly doubled from 16 months to 32 months with a hazard ratio of 0.47. So this is a practice-changing trial, this is a very meaningful result and the presentation of this trial also received celebratory applause from the audience, very deservedly. So it’s very encouraging to see that the standard of care for first-line treatment of urothelial cancer has changed. Let’s forget about the cost for a moment because I think it will take several years for this treatment to become accessible and affordable for patients in LMICs. We’re having this ecancer workshop right now in Kathmandu, our patients will have to wait several years, if not decades, to have access to this treatment. So that always bugs me but, from a scientific point of view, these data are very encouraging.

There was also the trial called CheckMate 901 which showed that adding nivolumab to chemotherapy in the first line also improves overall survival in urothelial cancer but the margin of difference was not as great as we saw with the enfortumab plus pembrolizumab trial. So if access is not a problem, enfortumab plus pembrolizumab should be the go-to treatment. There was also a trial of erdafitinib versus chemo as second-line treatment improving overall survival. So we need to think carefully about how to sequence these treatments.

But with regards to GU cancers, I also want to say a couple of things about the prostate cancer trial that tested lutetium PSMA versus androgen receptor inhibitor in a taxane naïve patient population. I was upset looking at the control arm of this trial. We have published this issue with substandard control arms in prostate cancer trials in our JCO paper last year as well. This trial does the same thing. The control arm patients had already progressed on abiraterone or enzalutamide and they’re still getting one or the other as the control arm treatment. That is substandard care. These patients in routine practice would not receive androgen receptor inhibitor after progressing already on androgen receptor inhibitor. So that is quite concerning and, even then, OS has not improved, it’s just the progression free survival. So we need to stop celebrating these trials that are using a substandard control arm.

Next I want to go to GI cancer trials. There was the trial of the KRAS inhibitor sotorasib plus panitumumab in patients with colorectal cancer as a last-line treatment. Improved progression free survival but did not improve overall survival which I don’t think should be considered practice changing at all, because as last line treatment you should, you must, improve survival. Patients have limited months to live to begin with as a last line treatment. If we can’t improve survival even in the last line of therapy then that drug is probably not meaningful.

In the same context I want to discuss about the HER2 positive patient population, the KEYNOTE 811 trial in upper GI cancer. Pembrolizumab plus Herceptin plus chemo versus Herceptin plus chemo in these HER2 positive gastroesophageal cancers. There was an improvement in progression free survival of less than two months: 10 months versus 8.1 months. But the discussant mentioned that there was no need to wait for overall survival and this should be immediately practice changing and this may substantially improve quality of life. I don’t agree at all. First, we can’t say it may improve quality of life – we have to prove that. Our study has already proven that you cannot guess quality of life based on progression free survival. We have several treatments that improve progression free survival but have a detrimental effect on quality of life. If we think that it improves quality of life we have to collect that data and we have to prove that it improves quality of life.

The progression free survival advantage here is less than two months. Overall survival has not improved or we don’t have the data yet. So how can we make an assertion that there is no need to wait for overall survival data and treatment should be approved on the basis of PFS alone? We don’t have OS, we don’t have quality of life so exactly on what basis can we make that recommendation? That is something that we need to think carefully, we don’t need to get on the bandwagon and celebrate every tiny improvement.

Finally, I would like to mention a couple of words about the SANO trial in oesophageal cancer. In oesophageal cancer having surgery is a big deal, it has a detrimental effect on the quality of life of patients. This was a pragmatic trial, a non-inferiority design trial, testing if we could avoid surgery for patients who received a complete clinical response after receiving neoadjuvant chemoradiation. They showed that, yes, non-inferiority can be proven but the hazard ratio is 1.14 with a 95% confidence interval going up to 1.78. So we need to carefully discuss with our patients if that is what they consider acceptable or non-inferior. But it was a pragmatic trial and it was good to see that researchers in the world are also working on answering these questions that are very meaningful for the patients.

So that’s the end of my ESMO 2023 highlights video. I’ll join you again after ASCO 2024.