ASCO 2026: Highlights and analysis

Prof Bishal Gyawali – Queen’s University, Kingston, Canada

Hello everyone, this is Dr Bishal Gyawali from Queen’s University, Kingston, Canada. I’m very happy to be back with you all at the ASCO 2026 Annual Meeting. I’m here right after the plenaries and I’m here to bring you the highlights from ASCO 2026, like I do every year. I think this is now my 11th year of doing this so thank you for joining me every year.

I’m in a very good mood because for the first time at an ASCO plenary I gave a standing ovation and it was very well deserved. I don’t get to do this very often. As I said, this was my first time doing this and it was very well deserved. I’m talking about the new drug for pancreatic cancer called daraxonrasib, it’s a KRAS inhibitor, and this is one of the first real advances in the treatment of metastatic pancreatic cancer disease. As some of you may know, back in 2024 I wrote a paper in Lancet Oncology called ‘Treatment of metastatic pancreatic cancer: 25 years of innovation without any real benefit for patients’, but that has now changed. I even joked about it on Twitter saying I need to change the title and make it 27 years with one real progress.

I’m talking about the RASolute 302 trial that tested daraxonrasib versus chemotherapy of physician’s choice for second-line treatment of metastatic pancreatic cancer. I really could not find much fault with this trial, this was a very well done trial. The control arm allowed physicians to use whatever chemo they wanted, which usually was gemcitabine/Abraxane or FOLFIRINOX or even FOLFOX or 5FU plus nanoliposomal irinotecan; so I can’t find fault with the control arm. The primary endpoint, one of the co-primary endpoints, was overall survival. They included patients with both KRAS mutation but also no KRAS mutation but the majority of the patients had KRAS mutation. And the results look fantastic – the overall survival almost doubled, so from 6.6 months in the control arm with chemotherapy, with daraxonrasib it is 13.2 months with a hazard ratio of 0.4; the 95% confidence interval was from 0.3 to 0.53. This is as good as it can get. The progression free survival also improved. It should be approved and it should be the standard of care and patients should start having access to this drug as soon as it is possible.

Having said that, we should also not get too carried away in that the overall survival with this drug still is 13 months so we should still improve but this is like breaking the frontier and hopefully there will be more and better drugs coming out in the future, more combination therapies. But this is exciting in the sense that it broke that barrier for the first time. I was very impressed with the publication of this paper as well, as well as the presentation at the plenary session, because this publication was important in that it presented data not only for overall survival and PFS but also on quality of life simultaneously. Usually we see that we have to wait for a long time to get quality of life data but in this case they presented it simultaneously and it’s remarkable. It is important to note that there are some important and unique toxicities with this drug, especially the dermatological toxicities – rashes in 86% of the patients, diarrhoea in 58% of the patients. So there are some important side effects. So when I saw those side effects I thought that patients would have a miserable quality of life but they did measure quality of life both for pain and global quality of life scales and in fact there is improvement in quality of life too. The time to deterioration of quality of life with regards to pain was 3.7 months versus 9 months and for global quality of life it was 2.4 versus 5.6 months. So these are fantastic results – it improved survival, it improved quality of life and this is what we want to see.

So if I have only one issue, it is I want to know whether crossover was allowed and if these results are with crossover in the control arm or without crossover in the control arm. But nothing to take away from the trial, these results are good and it should be practice changing.

Because I started from the RASolute 302 trial, I guess let’s wrap up the GI cancer trials first before we move on to other major trials from this meeting.  Let’s talk about the HORIZON-CRC01 trial. This is the trial of an antibody-drug conjugate called trastuzumab rezetecan, this is a trial coming from China in HER2+ metastatic colorectal cancer patients who are chemo refractory, as in those patients who have progressed on 5FU, oxaliplatin and irinotecan and compared with TAS-102 or regorafenib or fruquintinib. They even allowed a HER2 drug to be used previously and it showed an improvement in progression free survival of 2.8 versus 5.5 months with a hazard ratio of 0.33.

My only thing about this trial is that I’m not sure whether TAS-102 plus bevacizumab was allowed, they did not say anything about the use of VEGF inhibitors or even EGFR inhibitors for KRAS wild patients; they used only KRAS wild patients in this trial. But this seems like a promising drug but more data are needed on this. But I’m glad that the control arm included fruquintinib which is one of the latest treatments that has been approved.

I was very happy to see the results of the PUMP trial coming from the Netherlands. Usually we get very nice, important, pragmatic trials from the Dutch group, funded by public funding, and the PUMP trial was also one of those examples. This was for patients with colorectal cancer with liver metastasis who underwent surgery for liver metastasis. This was a trial testing whether an adjuvant hepatic artery infusion pump for these patients improved any outcomes. They used patients with low risk and there have been a couple of trials of hepatic artery infusion pumps in the past and they have not been positive, especially for overall survival. This trial showed that, even in this PUMP trial, even progression free survival was not improved and overall survival was also not improved. So I think this should settle the question of the utility of hepatic artery infusion pumps, especially in this population, because hepatic artery infusion pump is not a very easy intervention, it requires a sophisticated surgery plus intervention and it is burdensome on the patient. We now have three trials, none of which have shown improvement in overall survival, but we still have so many centres around the world who do this as a routine practice. So I think this should give us pause and we should not recommend hepatic artery infusion pumps for patients without having any solid evidence that it is going to improve outcomes.

Also at the GI oral session, Tuesday morning, the results of the EPISODE-III trial was published. It’s the trial of low-dose aspirin in the adjuvant setting. We now have a number of trials of low-dose aspirin in the adjuvant setting and there are a couple of caveats like the only trial that is obviously positive is the ALASCCA trial for the PI3K positive patient population. The EPISODE-III trial did not segregate by the PI3K status, maybe those results are coming up but this is in the overall population. It did not improve disease free survival, the hazard ratio was 0.84. This is a Japanese trial. But at this point I think we should stop doing more trials of low-dose aspirin in this patient population because we have had a number of trials that are all negative and one positive trial is in the PI3K positive population. So I guess a better exercise to do would be to collect PI3K positive, PI3K status, biomarker status, from these patients across these trials and pool the data and answer definitively whether or not this is in the benefit of patients. There are some variances in these trials in terms of the dose of the aspirin used but at this time it does not change any of our practice.

Also I would like to talk about a very interesting study that was presented from China. This is not a randomised trial, this is more of a prospective observational study for patients with MSI high colorectal cancer that were treated with an anti-PD1 antibody and that led to CCR, complete clinical response. They were managed non-operatively and for those patients who underwent this non-operative treatment, they looked at differences in outcomes for patients who continued to receive maintenance PD-1 versus no maintenance and only observation. They actually found that there was no difference in outcomes. So, as I said, this is not a randomised trial so there are obvious limitations and biases, but this is thought provoking in that in many disease sites we keep treating patients for a long time, especially with immunotherapy, but is that long duration of treatment even needed? They showed that for patients who continued on maintenance anti-PD1 therapy there was an increase in the incidence of immune-related adverse events without any benefit in terms of outcomes. So that felt quite thought provoking.

Now, moving on from GI, I would like to talk about one of the plenary trials in dedifferentiated liposarcoma. This was a trial of abemaciclib versus placebo. Unlike the RASolute 302 trial, with this trial I have a number of issues, the biggest of which is the use of the control arm – placebo. These are patients with liposarcoma, the standard of care should have been doxorubicin in first line and trabectedin in the second line and that’s what other trials in this space have done and some of them have failed. The abemaciclib trial instead compared against placebo, and I find it very uncomfortable to use the placebo arm, both for the first-line and second-line setting.  Interestingly, the progression free survival was the primary endpoint in this trial. Now that boggles my mind because if you are testing against placebo, PFS makes absolutely no sense. Against doing nothing, giving any drug is obviously going to increase your progression free survival. That feels like a no-brainer. I think it is pretty difficult to lose against placebo for a PFS endpoint.

So I think this trial was designed from the outset to be a positive trial, regardless of how active or inactive the drug is. Interestingly, if you look at the median PFS for the placebo arm it’s 1.5 months, that’s like six weeks. That is the duration of the scan, that is when the patients will get the first scan. So the first scan there is progression. Of course there is progression because these patients should have been treated but they got placebo instead. So even in the first scan itself there was disease progression, which makes it even more questionable and borderline unethical to give these patients placebo. Of course there is no difference in overall survival as of now and there was crossover as well which also I think is not appropriate in this case. The patients when they progressed they should have received doxorubicin or trabectedin; for a drug that is being tested for the first time in a disease setting crossover is not indicated.

So I have been seeing these issues about crossover in many of our trials. Where crossover should not have happened crossover is being allowed and where crossover should have happened crossover is being prohibited, and we should speak out against these issues with crossover. Again, placebo control, just PFS, patients in the control arm progressing at their first scan and still being considered eligible for placebo at the outset. That means these were not patients with indolent disease, they have immediately progressed. So that did not feel like a rightly designed trial or a meaningful outcome.

Moving on to lung cancer trials, there were plenty of lung cancer trials, important lung cancer trials, presented at this year’s ASCO. First I want to start with the CROWN trial 7-year update. This is a trial of lorlatinib versus crizotinib first line and very exciting results. 7-year PFS rates of 55% versus only 3% in the crizotinib arm. This is a remarkable difference, quite impressive. The median PFS is still unreached after seven years of update so this is very exciting. But one caveat is that palliative radiation therapy was allowed for oligoprogression in this trial but that oligoprogression that made people do palliative RT didn’t trigger PFS events, I was not very sure about that. Because if that did not count as progression then there is a component of radiation therapy leading to that long-term disease control and leading to these long-term PFS results. But, regardless, these are fantastic results, the difference is remarkable. It’s good to see that in such an otherwise lethal disease there are not enough events reported or have happened yet for an overall survival analysis to be done. So these are all good news.

But one important thing that I found in the publication that I think some people may not have paid attention to is the impact of dose reduction. They started with 100mg then for patients, based on toxicities, went to 75mg and to 50mg. But regardless of this dose reduction there was no impact on efficacy outcomes and that is something worth thinking about because it has implications for toxicities as well as financial toxicity. So I think there is a role now with these remarkable results to test if we can get away with a lower dose. Maybe the patients don’t need a 100mg dose, maybe everybody can get away with a 50mg dose which would be half of the therapeutic burden. So that is something worth studying.

The OptiTROP-Lung05 trial of sacituzumab tirumotecan plus pembro versus pembro alone, again the median was not reached versus 5.7 months. Hazard ratio 0.35, confidence interval going from 0.26 to 0.47. This trial included both non-squamous and squamous histology, there was increased toxicity too. There have been some controversies about this trial because this trial included patients with any PD-L1 positive population and almost 60% of them were probably between 1-49% PD-L1 group. The standard of care for these patients is pembro plus chemo but in this trial the control arm was pembro alone. If we are being statistically pure, the control arm is defensible in the sense that there is no randomised data showing pembro plus chemo is better than pembro – the pembro plus chemo trials initially they tested against chemo, they did not test against pembro – so there is no hard data, to be honest, to say pembro plus chemo is better than pembro in this patient population. It’s the subgroup from the pembro alone trials that show limited efficacy in the 1-49% population as opposed to the >50% population.  But I was glad to see that people are questioning the control arm and people are asking these questions about the control arms. Yes, I think one of the best ways to answer is the control arm ethical is what would you do for these patients if it was your patient or if it was you as a patient? I think that answer is obvious that the majority of us would do pembro plus chemo for these patients. So, yes, it is very appropriate to question the control arm.

If it was pembro plus chemo then these results would have been even more impressive but I was also happy to see that for the first time immunotherapy, or pembrolizumab, was, in a sense, beaten. There were results better than immunotherapy in a lung cancer trial because I was getting fed up of seeing more and more me-too trials of me-too immunotherapy drugs, still continuing to use chemo control sometimes. So it was good to see a new combination improving upon pembro alone.

Let’s move on to the WU-KONG28 trial of a new drug called sunvozertinib. This is specifically for the EGFR exon 20 positive population. For this population the previous standard of care has been amivantamab plus chemo. This is a single agent, sunvozertinib, and this trial showed that for these patients, compared to chemo, PFS was improved from 7.5 months to 10.3 months. This trial allowed crossover but for an earlier line therapy trial because this is a drug that is already approved for later lines. Sunvozertinib is not a new drug in the sense that it has been approved for second and subsequent lines. This is a trial testing if it is effective in first line, but for first line trials I don’t think PFS is enough because the question is not whether we should use this, the question is whether we should use this now, at first line, for treatment naïve patients. So the answer to that question should always be overall survival against using it later. So crossover being allowed in this trial was fantastic, I am glad that crossover was allowed because if crossover was not allowed we would not have answered that question. But now that crossover is allowed we should look after overall survival results and if overall survival results are positive then this should be practice changing. How it compares against amivantamab plus chemo, I think it will be a second option to have and physicians and patients can discuss based on the risks and benefits profile.

Now let’s move on to the plenary trials for lung cancer. First the HARMONi-6 trial, it’s the trial of ivonescimab plus chemotherapy versus tislelizumab plus chemotherapy. As I mentioned earlier for the OptiTROP trial, that was the first trial that I saw against pembro and I wished there was pembro plus chemo as a control arm, here we have that. The control arm is immunotherapy plus chemotherapy so there are absolutely no concerns about the control arm here. Tislelizumab is an immunotherapy, like pembrolizumab, plus chemo, so that is the ideal control. The experimental drug here is ivonescimab. Ivonescimab is a bispecific antibody with both immunotherapy and VEGF components there. It improved median overall survival, so this is not PFS that we are talking about, and this is what we want to see instead of the [?? 19:49] PFS results. This is overall survival 23.7 months versus 27.9 months with a hazard ratio of 0.66 with a relatively stringent p-value cutoff for the OS boundary.

So I think this is a very encouraging result, there are concerns about side effects here. The incidence of severe AE has increased from 59% to 69% and haemorrhage seems to be a big concern here. They included non-squamous and squamous patients, in squamous patients haemorrhage has always been a concern with VEGF inhibitors. In this trial the incidence of severe haemorrhage was 1% versus 3%. The only limitation here was that this trial included patients only from China and we don’t know how it fares with patients outside of China but I don’t see why it would act differently in a population outside of China.

But at least this is pretty exciting, that pembro plus chemo has remained the control arm, or the standard of care for a long while and we finally have something that improves on that, not only for the surrogate endpoint of PFS but also for a survival endpoint.

The other plenary lung cancer trial was the LIBRETTO-432 trial, it’s the trial for RET positive patients in the adjuvant setting, stage 1B-3A, of adjuvant selpercatinib versus placebo after they have completed their surgery and adjuvant chemotherapy for three years. An impressive hazard ratio of 0.17 for EFS, a two-year EFS rate of 91% versus 61%. Previously with the ADAURA trial we have asked about what is the role of DFS if there is no OS and later on overall survival results were also positive and improved, so this hazard ratio of 0.17 for the LIBRETTO-432 trial, I believe that the overall survival will also be positive and this is a very impressive hazard ratio. My only concern here is that the duration of treatment is for three years; there is no data for why it is three years so we need to be able to tease this apart further – if we can get away with a lower dose, if we can get away with a shorter duration of treatment. That is where we should focus on because three years is a pretty long therapeutic burden for patients.

Moving on from lung to breast, first of all in breast cancer we saw the final overall survival results from the KEYNOTE-522 trial of pembro in the curative setting. 7-year OS from 77% to 85%, this is a gain of 8% in 7-year overall survival rates and I wanted to highlight this because, especially in breast cancer, people make an argument that you can’t improve overall survival so we should be happy with DFS alone, EFS alone. Or even if we look at the adjuvant CDK4/6 inhibitor trials, the overall survival improves by 1%, 1.2%, 1.6%, then we are supposed to celebrate that. But, no, if the drug is good and if it is effective it improves overall survival remarkably well, like in KEYNOTE-522 you have a survival difference at seven years of 8%. You compare that with the CDK4/6 inhibitor trials and you will see what I’m talking about. So don’t settle with 1.6% OS gains.

Another important trial was the OPTIMA trial. This is for patients with ER+/HER2- early breast cancer, including patients with 0-9 lymph nodes and if lymph nodes were negative, even tumours up to more than 3cm were included. The question this trial was asking was a very patient-centred question in that for these patients if chemo is not needed then they should not be getting chemo if endocrine therapy alone is effective. But we used to do that only for low risk patients to segregate, even if they are low risk, whether or not they need chemo. But this is asking the question on the other end – even if they look like high risk maybe we can spare some of them with chemo if they don’t need it. So they used the Prosigna test and if the ROR score was less than 60 they randomised patients to get endocrine therapy alone and if it was more than 60 then they got both endocrine therapy plus chemotherapy. This was appropriately designed as a non-inferiority trial and the primary endpoint, the invasive breast cancer free survival at five years, it hit the non-inferiority criteria – 91.5% versus 90.4% – in the overall population. In the less than 60 score population it was 94.9%; in the more than 60 score population it was 93.7%. So it clearly met the non-inferiority criteria and there was not even heterogeneity in terms of menopausal status or nodal status. So I think it is practice changing and for patients who have these high risk features with lymph nodes and tumour size I think this can be a standard of care, by using this test and using the risk of recurrence scores to decide whether or not they can be spared chemotherapy. So this is a meaningful trial.

The other trial, on the other hand, called the lidERA trial, it randomised patients to get adjuvant SERD which is, in this case, giredestrant versus endocrine therapy alone for similar ER+/HER2- patients. It showed an IDFS benefit but I don’t think this should be a practice-changing trial because, again, we don’t have OS benefit here, it’s just IDFS benefit. The therapeutic burden here is immense, it’s for five years. This is five years of treatment and I don’t think it is meaningful to offer five years of an expensive treatment with side effects for patients without any solid evidence that it improves overall survival.

I was surprised that nobody… I don’t think the control arm was bad, the control arm was okay because I think that adjuvant CDK4/6 inhibitors don’t add much, so I can’t argue on both ends, but I was surprised to see that even people who fiercely advocate for adjuvant CDK4/6 inhibitors did not complain about this control arm. Because if they think that adjuvant CDK4/6 inhibitor is standard of care, then they should have advocated for adjuvant CDK4/6 inhibitor to be added to endocrine therapy in this control arm. Anyways, just an observation.

Moving on to the final histology, GU cancers. There were a number of important trials. Firstly let’s start with the TALAPRO-3 trial which is in the metastatic castration-sensitive prostate cancer space for patients with homologous recombinant repair positive patients. This was a trial of talazoparib, which is a PARP inhibitor, against the control of androgen deprivation therapy plus enzalutamide. The primary endpoint was radiographic PFS which was positive with a hazard ratio of 0.48. A number of issues with this trial is that, a, the control arm did not allow for triplet therapy, as in did not allow for docetaxel to be added. I think the protocol changed midway to not allow docetaxel anymore. And, more importantly, crossover was not allowed and I find not allowing crossover… and that’s what I was referring to earlier, as in where crossover should have happened and did not happen and where crossover that should not have happened happened. Here crossover should have been mandated because in a later line setting PARP inhibitors are already standard of care for these patients. If you look at what post-protocol treatments these patients got in the control arm, I saw that only 27% of them got olaparib which is the only PARP inhibitor listed there. They did not get any other PARP inhibitor, only 27% got olaparib. That is a very low number that is a very, very low percentage. So there should have been crossover in this trial, not allowing crossover was not the right thing.

In fact, if you look at the post-protocol treatment, 26% of them are receiving abiraterone and 13% of them are receiving enzalutamide. They just progressed on enzalutamide. The control arm received ADT plus enzalutamide, they progressed on it, and, again, 13% of them received enzalutamide. That does not make much sense. Or even 26% of them receiving abiraterone, that also does not make sense. So I have issues with this post-protocol treatment here.

The other trial, before we move on to the PROTEUS, which was the plenary, let me just share a couple of thoughts about the RAMPART trial. The RAMPART trial is in the renal cell cancer space, adjuvant durvalumab versus placebo. Here adjuvant durvalumab failed to improve DFS and the reason I bring this up is now we have four different immunotherapies that are tested in the adjuvant renal cell cancer space, we have adjuvant nivo which failed, adjuvant atezo which failed, adjuvant pembro which was positive, and now adjuvant durvalumab which is negative. I don’t think there is any meaningful difference between these four drugs, these are all PD-1/PD-L1 inhibitors. When you have four trials in a space and three of them are negative and one of them is positive, I think that we need to look at the totality of the evidence. Even in that adjuvant pembro trial, we need to look at what percentage of the patients in the control arm got pembro when they relapsed and whether the overall survival benefit stands when you correct for that lack of crossover for patients in the control. So I just bring this up because I find it striking that after four immunotherapy trials, three of them are negative and one of them is positive.

Now, finally, going on to the PROTEUS trial which was in the plenary, these are patients with high-risk localised prostate cancer who received a radical prostatectomy and the standard of care is radical prostatectomy alone, after that you just do surveillance and follow-up. This trial tested whether giving them ADT plus apalutamide, both in the neoadjuvant and adjuvant phase, six months each, so a total of 12 months of treatment, improves outcomes. But the control arm was not surveillance, here the control arm was ADT plus placebo. The speaker mentioned that this was done to maintain blinding of the trial which makes sense but I want to highlight a couple of things in this trial. One is that the primary endpoint, there were two primary endpoints, one of them is the pathological complete response rate but it’s not just the pathological complete response, it’s pathological complete response plus MRD rate but we need to remember that this is not like our pathological complete response in lung or breast where it is actually pathological complete response. Here anything less than DP2 and less than 5mm was considered as pathological complete response [?? 31:35] MRD so all of them were not exactly pathCR. Even then although what was highlighted was the odds ratio of 10, you need to look at the exact numbers, it’s just 1% versus 9%. So it’s pretty low, the pathologic complete response rate, and there have not been too many neoadjuvant trials in this setting so we don’t know what the ideal pathological complete response rate would be in prostate cancer. But this looks pretty low, just 9%. The other important primary endpoint was metastasis-free survival and it improved by 5% almost over five years – 73.5% versus 78.2%, a hazard ratio of 0.8, 95% confidence interval from 0.67 to 0.96.

Now, let’s talk about the issues. First the control arm. Instead of surveillance it was ADT plus placebo and does that matter? In terms of looking at efficacy I think it actually might even raise the bar somewhat that patients got ADT. But if we are looking at quality of life impact, because that is going to be very important here because we are going to treat patients for 12 months with a treatment where the standard of care is to give none of this treatment, so the quality of life impact is quite important. Unfortunately, quality of life has not been reported so without having quality of life data it is very hard to draw conclusions on that. But what impact that 12 months of ADT has on quality of life and does that masquerade the new proposed standard of care – ADT plus apalutamide – quality of life impact, because the control arm also had a detrimental impact on quality of life because they got ADT instead of getting nothing. So that is what we should keep in mind when we are looking at the quality of life results which, unfortunately, we don’t have yet, unlike the daraxonrasib trial which reported quality of life.

The second issue is that of metastasis-free survival. The metastasis-free survival, initially when the trial was designed it included the conventional imaging-based metastasis-free survival but then later the protocol was changed to also allow for PSMA-directed metastasis to count as an event. That, I think, has diluted the results because metastasis-free survival is a surrogate for overall survival but that statement holds true for conventional imaging-based metastasis-free survival. For PSMA-based metastasis-free survival it’s not a good surrogate endpoint; pathCR is even a weaker surrogate endpoint. In fact, they did a sensitivity analysis using only the classical definition of metastasis-free survival and for that the results were not significant.

So I have issues with the endpoint as well but, more importantly, for the sake of argument let’s assume that 5% improvement in metastasis-free survival is true. First, we don’t have overall survival results and we can’t discount lack of overall survival results here because we are adding one year of extra therapeutic burden with ADT plus apalutamide in these patients. We could argue that it is just one year of treatment, just one year of treatment, and if it is going to delay the time until you have to get your next treatment or if it is going to improve the cure rate then that is meaningful. But we need to look at the magnitude of benefit too. The magnitude of benefit here, we are not talking about hazard ratios of 0.4, 0.3, 0.15, like we are doing in other cases. Here we are talking about a hazard ratio of 0.8, the 95% confidence interval goes up to 0.96. In fact, if you look at the Kaplan-Meier graphs up until the 36 months timepoint it is hard to say there are two lines. Only after that they somewhat start to differ.

Even if we take that metastasis-free survival result at face value, it’s a difference of 5% over five years, that means almost 74% of the patients have been completely metastasis free over five years by surgery alone, they did not need anything. By giving them an extra year of ADT plus apalutamide, and both ADT’s burdens should also be accounted for because there was not a third arm of no treatment, the control arm was ADT and this treatment is ADT plus apalutamide. I wish there was a third control arm without any ADT. So that therapeutic burden needs to be accounted for because it would mean one year of extra treatment and that would improve those rates from 74% to 78%, 73.5% to 78.2%. So we are making an extra five patients metastasis free over five years by this one year of extra treatment. So that should be accounted for. As in, even if we did that, 22% of the patients are going to have metastasis regardless, even after that one year of treatment. It’s not like we give that one year of ADT plus apalutamide and everyone becomes metastasis free; 22% of the patients are going to have metastasis regardless and 74% of the patients are not going to have metastasis regardless. So is that benefit worth the therapeutic burden is something that we need to carefully consider about.

So those are my overall thoughts about the most important practice-changing trials from this ASCO Annual Meeting. In addition to that I also want to highlight a couple of other observations. Most importantly, at this ASCO Annual Meeting at the opening ceremony, as well as we had a fireside chat between the ASCO President, Dr Eric Small, and David Kessler, the grief expert. These moments were very human and very poignant in that we were talking about death and dying and handling grief. We usually try to hide these things and not talk about these things but it is important for oncologists, for patients, for the society to open up about these things. It’s okay to talk about grief and dying and loss and bereavement. I’m very thankful to Dr Eric Small for opening up about his own story but also allowing these topics to become a part of the ASCO Annual Meeting and discussion so that people feel okay to talk about these things. So I really wanted to highlight that.

So a lot of similar, not necessarily important, trials but meaningful research are being presented. Overall, in my opinion, the bottom line is we need to be critical about these results, we need to think about how to improve the quality of care and we need to try to do the best for our patients without necessarily buying the hype. But also being hopeful where it is deserved and trying to put a critical eye on hype. As I say, we should never outsource critical thinking, we should think for ourselves and we should be able to discern the hope from the hype.

Thank you everyone, and I guess I’ll see you back after ESMO for the ESMO highlights.