Highlights from ASCO 2020

Bookmark and Share
Published: 5 Jun 2020
Views: 4101
Rating:
Save
Dr Bishal Gyawali - Queen's University, Kingston, Canada

Assistant Prof Bishal Gyawali speaks online to ecancer about the highlights of the 2020 American Society of Clinical Oncology (ASCO) Annual Meeting.

Prof Gyawali opens covering some high profile trials where he discusses two controversies in the the plenary sessions including ADAURA, JAVELIN, and KEYNOTE-177.

He then touches upon a few of the important negative trials, and policy relevant trials, covering breast, prostate, gastric, and hepatocellular cancer. Prof Gyawali looks through the lens of a global oncologist, and whether some of these treatment options are viable on a global scale.

He finalises by talking generally about the ASCO meeting's changes this year, highlighting the huge increase in global oncology trials which can aid practice in LMICs, he briefly covers a few of these.

All opinions expressed in this video are the speakers own and do not reflect that of any institution.

Today I’m here to share some of my feelings about some key research and trials that came out of ASCO 2020, the first virtual meeting of its kind. As I do every year with ecancer, today I’m just going to bring you a roundup of the key take home messages from this ASCO meeting.

Firstly, let me discuss a couple of high profile trials, the trials that led to the biggest debate on social media and the trials that are a little controversial about whether or not they are practice changing and what my take is on those trials. Let’s just start with one of the plenary trials called the ADAURA trial of adjuvant osimertinib versus placebo in EGFR mutation positive non-small cell lung cancer. This was a trial in which they tested adjuvant osimertinib for three years versus placebo and found that adjuvant osimertinib significantly improved disease free survival with the quite impressive hazard ratio of 0.17. Now, if you take it at face value this seems like this should be a practice changing trial; this looks like a very important trial. You very rarely see a hazard ratio of 0.17 and especially in lung cancer. But if we look into the details of that trial then there are a couple of things that are of concern. First, the control arm patients did not necessarily get chemo, so getting chemo was not mandated and, in fact, more than half of the patients did not get chemo in the control arm which is not the standard of care. The standard of care for patients who have undergone surgery for non-small cell lung cancer stage 1 to stage 3 is to get adjuvant chemo. Now, how could some patients in the control arm not get adjuvant chemo, I will never be able to understand that, that’s not a standard of care. So this is a trial that is not comparing against the standard of care, it is comparing against placebo. Some of the patients might have undergone chemo but some of the patients are not getting chemo. So the control arm is not the ideal control arm.

I really can’t imagine patients who underwent surgery for lung cancer not getting anything at all and just getting placebo because they are part of this trial, a, and, b, the overall survival data are not mature and the overall survival, whatever data we have at this moment, does not show that… it shows very parallel running OS graphs, the Kaplan-Meier graphs. Now, without having overall survival data I think it’s difficult to recommend any adjuvant treatment for any disease setting; it’s not only about lung cancer. Because the bar for giving an adjuvant therapy is much higher than the bar for giving any treatment in the advanced setting. In adjuvant therapy, because a number of patients will have already been cured, for example, if we take stage 1b patients that were included in this trial, half of them don’t even relapse. So we are asking all patients to take three years of therapy, of this expensive therapy, the cost of osimertinib is somewhere between US$15-20,000 a month, and we are asking patients to take this drug for three years, 36 months. It’s not only about cost, we are asking patients to undertake a therapy for three years, although people say the toxicity is manageable, and things like that, the toxicity is still there, grade 1, grade 2 toxicities and the toxicities are lasting for three years. So even if it’s not grade 4 toxicity it’s grade 2 diarrhoea for three years, that’s quite a big deal. The most important point is we don’t know what the overall survival is going to be and we don’t know whether the patient was going to relapse or not in the first place. The patient probably would have been cured after surgery and chemo so the patient may never need the drug. And if the patient did relapse, osimertinib is already approved as a good first line agent when the disease relapses, so it’s a good first line agent. So you have to compare overall survival with this approach versus patients getting the control arm actually should be surgery and adjuvant chemo and then followed until they relapse and at the time of relapse give them osimertinib first line. So whether giving osimertinib up front as adjuvant for three years improves overall survival versus this approach, that should have been the ideal comparison. So that is not the case here so we can’t say, even if there is an impressive hazard ratio for disease free survival, we don’t know whether it’s going to improve overall survival or not and unless we know that, versus a standard control, I don’t think we can recommend this therapy for every people who have undergone resection for non-small cell lung cancer.

We should definitely continue to give the adjuvant chemo, that’s my takeaway. With regards to whether or not we should be giving adjuvant osimertinib, I would wait for the mature overall survival data to come. There have been a lot of examples in oncology, not only in lung cancer, where first disease free survival, impressive hazard ratios in disease free survival, did not translate to benefit in overall survival. Sometimes it can translate to harm in overall survival, we have seen that, say bevacizumab in colorectal cancer – pretty good disease free survival we saw early on but then that translated to worse outcomes later. So it’s still too early to recommend this.

Then the other thing is about duration. With adjuvant therapy you don’t know how long you have to… This trial tested for three years but why stop at three years, then? You could easily ask that question. If the disease is under control for three years, maybe we should give it for five years, maybe we should give it for ten years just like adjuvant hormone therapy in breast cancer. But are we really going to go down that line? Are we really going to recommend adjuvant treatment such as osimertinib for a number of years without even knowing whether it’s going to help patients or not in terms of improving overall survival? So that is those ethical, economic issues don’t give me the impetus to just say, ‘Oh this is a wonderful trial and this should be practice changing.’ No, I’m not able to say that because of the reasons there that I mentioned.

The other controversial trial, let’s say, was another plenary, the JAVELIN-Bladder-100 trial, which tested maintenance avelumab after first line chemotherapy for patients with urothelial cancer. Again, why is this controversial? There is overall survival benefit and usually when you see overall survival benefit you don’t say that it is controversial. But right now, despite seeing overall survival benefit, the reason I’m saying this is controversial is because the control arm patients, only 43% of them got a checkpoint inhibitor later. So this is just a question of if they had received a checkpoint inhibitor later when they relapsed then probably they would have had the same survival and maybe that survival difference that we are seeing right now would have vanished. So, again, it’s a question of should we give an expensive and toxic treatment up front for all the patients when you could keep the same treatment when you need it, that is at the time of relapse, and you could maintain the same survival. We don’t know the answer to that question because there was no crossover and we need to see results from a stratified analysis based on patients who got a checkpoint inhibitor later and what was the overall survival difference for that group of patients versus those patients who never got a checkpoint inhibitor in their life. So that will help us to answer some of those concerns.

The other interesting thing is in this trial if we look at the progression free survival, the progression free survival difference is quite small, it’s 2 versus 3.7 months. Why this is important is that if we look at the trial design patients got chemo, platinum-based chemo, and then there was a 4-10 week period until they could start the trial. So that 4-10 weeks is in itself 2½ months and the progression free survival difference is 2 versus 3.7 months. So we should also know at what point, because 4-10 weeks is quite a long duration, we need to know at what point did patients start getting avelumab versus the control arm patients. If the patients on the treatment arm were started down the line towards the 10 week point but the patients in the control arm were started up the line towards the 4 weeks’ time, then that difference could be an artefact of just when we started the patients on treatment.

The other plenary that actually brought a lot of attention is the KEYNOTE-177 trial for MSI high colorectal cancer patients which showed an impressive progression free survival difference. It nearly prolonged the progression free survival by double – 8.2 versus 16.5 months with the use of pembrolizumab versus the chemotherapy arm – and that was quite impressive. The overall survival follow-up is ongoing but I was impressed with the hazard ratio and the toxicity profile of pembrolizumab versus the chemotherapy. There was some debate going on about why there was crossing over of the PFS curves towards the early period but I don’t think that’s very unusual for immunotherapy because we have seen such crossing over of PFS curves in lung cancer as well with immunotherapy. What that actually means is that immunotherapy takes a certain time to work and we might see some pseudo-progression as well, we need to see if that happens in colorectal cancer as well. If the OS curves cross early on then we need to be careful about it but crossing over of PFS curves is not a big deal.

Those were some of the controversial plenaries from the ASCO 2020 meeting but with regards to other important trials, not plenary but other important trials that were presented at ASCO 2020, I was impressed with the fact that ASCO gave a platform to a number of negative trials this year, including one in the plenary which was a trial of breast cancer – whether or not we should treat the primary tumour with a local therapy when the patient presents with stage 4 disease. This was already answered by a trial from India a couple of years ago that showed that doing surgery for the primary tumour was not of any benefit when the patient has stage 4 disease. The results were the same from this new trial that was presented at ASCO 2020 but nevertheless it was good to have a confirmation of the finding.

The other trials, one trial that was also published in The New England Journal simultaneously was the trial of a GnRH antagonist, relugolix, versus goserelin, a GnRH agonist for prostate cancer patients, advanced prostate cancer patients. Androgen deprivation therapy has been one of the key progress against prostate cancer but the problem with GnRH agonists is the initial testosterone flare that can aggravate the disease for a certain initial duration. A GnRH antagonist like degarelix doesn’t have that side effect so GnRH antagonists are a little more attractive but, for some reason, degarelix has not gained a lot of uptake in practice compared to a GnRH agonist. One reason is that degarelix is given by injection, it can be given monthly or three-monthly, and there are some injection reactions with the drug. So this new drug, relugolix, is an oral form of GnRH antagonist so it was pretty exciting. The good news is that it was tested as both a non-inferiority and superiority trial in an hierarchical design and it met both the criteria so that means relugolix was superior to GnRH agonist. So that was good news.

It also lowered the risk of cardiac side effects. So overall it is good news but, again, as a global oncologist if I have to put it into global perspective then I think relugolix is going to cost substantially compared to degarelix. Relugolix and degarelix both are GnRH antagonists and although relugolix is oral and there is good compliance, I think from the global oncology perspective degarelix should still be the choice because there is no data to say that relugolix is better than degarelix. Of course, both of them are GnRH antagonists and then you can give it once every three months, an injection once every three months should not be a big deal, and there will be confirmed compliance with an injection compared to oral medications which should be a big deal in low and middle income countries. So that’s my take on that.

There was one policy relevant trial presented, that is the trial of tucatinib, the HER2CLIMB trial, the trial of tucatinib in breast cancer, HER2 positive breast cancer. The policy relevance that I’m excited about this trial is that they included patients with breast cancer who had brain metastasis. So usually in most of the trials brain metastasis is an exclusion criteria – if a patient has brain metastasis they cannot be a part of the trial. But in this trial, this trial showed that patients with brain metastasis can very well be enrolled into a randomised controlled trial and can very well be a part of the trial. So rather than the efficacy of tucatinib, which we can debate again, or its relevance to low and middle income countries, the relevance of this trial for me, the take home message from this trial for me is that we should not be excluding patients with brain metastasis from routine randomised controlled trials.

A couple of important trials were presented in hepatocellular cancer – a trial of two new agents, donafenib in first line and apatinib in the second line, metastatic hepatocellular cancer. The good thing about these trials is that both of them came from China, now we have to take a look at what the cost will be. I’m hopeful that they will cost less than what these new drugs like imatinib cost for hepatocellular cancer in the West. So that was also of interest to me.

Of course I was also interested to see some of the presentations related to COVID-19 and cancer. For me, the question about whether or not patients with cancer who are getting chemotherapy are at a higher risk of dying from COVID-19 complications compared to patients who don’t have cancer or who were not getting chemotherapy for cancer, I think that that question has not yet been settled. We still need to continue with the efforts and collecting data that we have had. Special congratulations to my friends, a lot of friends whom I have at the COVID-19 and Cancer Consortium, they have done a huge effort and their publication in The Lancet was the result of a huge collaboration and we should be aiming for continuing such collaborations. Not only for COVID-19, I think we should take this lesson and we should continue collaborations for rare cancers, for example.

There were a number of negative trials that have been entered to ASCO 2020 and I think they are very relevant because negative trials tell us what not to do. One of them was the KEYNOTE-061 trial of pembrolizumab versus paclitaxel as a second line treatment in gastric cancer. It showed that there was no improvement in outcomes by using checkpoint inhibitor, pembrolizumab, versus classic chemo paclitaxel which costs much less than pembrolizumab and which we have been using for a long time. So newer is not always better, newer is not necessarily better.

The other thing was just having a target does not mean you can hit the target and improve outcomes. This we saw in the RTOG 1010 trial in which we tested the same hypothesis, adding Herceptin to the CROSS regimen for oesophageal cancer and this did not improve disease free survival or overall survival. So the bottom line is we need to keep testing the targets in a randomised fashion because if we are not testing these trials in a randomised trial then we would never know whether they were going to improve outcomes or not. So we need to keep testing in randomised trials but we should not be so excited and get ahead of ourselves that just having a target would make us believe that we can improve outcomes by having a drug against it.

And that should be a key take home message because we also saw trials for newer targeted agents like trastuzumab deruxtecan, which was presented for a number of tumour types, like gastric cancer, for example. This was just phase II trials and some of them were not even randomised trials and still we were getting pretty excited about, oh, this should be a game-changer, this should be the biggest new thing in this particular tumour type. We should not get ahead of ourselves, we should continue to remember the lessons that we have learned over the years and we need to continue to test them in a proper randomised controlled trial to make sure that they really improve outcomes or sometimes they can even worsen the outcomes and we need to save our patients from that.

So that should be the take home for nearly all single arm trials that were presented this year at ASCO. Even with phase II trials the response rate, we have a done a study that was published in The Journal of NCCN this January that the response rates from non-randomised trials actually is falsely exaggerated and the duration of response particularly, not response rate but duration of response. So the durable responses that we see in non-randomised trials could actually shrink when you do a randomised trial, much less about overall survival.

The other negative trial that I want to highlight is about the trial of, again, adding Herceptin for ductal carcinoma in situ. This was the NSABP B-43 trial of radiotherapy plus or minus Herceptin for DCIS. Again, adding Herceptin did not improve outcomes so this is another negative trial that actually informs our clinical practice.

Then I would like to talk about a couple of trials from geriatric oncology that showed us that not every practice-changing trial is a new drug trial. You could make a lot of difference by doing some simple interventions. There was an RCT of doing a geriatric assessment for elderly patients and just doing a geriatric assessment decreased the incidence of grade 3-5 toxicities which was a pretty nice observation and I like that trial.

I also liked the trial, I think it was called the PANDA trial, for colorectal cancer patients over the age of 70 stratified on the basis of G8 and geriatric screening score and tested whether or not we could omit oxaliplatin from a FOLFOX plus panitumumab regimen for [?? 21:55] patients. I think they did not make a statistical comparison but the median overall survival was pretty much similar for patients who did not get oxaliplatin versus those who got oxaliplatin based on this geriatric screening. So those were some good, informative trials.

I think I have never seen so much globally relevant trials presented at ASCO. Most of my colleagues working in low and middle income countries, they used to tell me that coming to ASCO was like an expensive trip to Chicago. The actual relevance to their clinical practice would be very small. It was nice for sightseeing and networking and everything but most of the trials that were presented at ASCO were about these new drugs that they never would have access to anyway. So there was very low value in that. And actually I think one of my colleagues from India, I think it was Dr Nirmal Raut who just posted on Twitter, I saw it a couple of hours ago, saying something like, ‘Going to ASCO is like going to a fashion brand store. You go inside and you get attracted by looking at a number of things but then you return without buying anything to take home.’ So, for global oncology colleagues that perfectly summarises how we feel about ASCO – you hear about this new PARP inhibitor, this FGFR inhibitor and things like that, you go back home and there is nothing to use there.

But actually at this ASCO there was a lot for global oncology colleagues to take home and a number of those trials came from low and middle income countries themselves. I was very, very happy with that, I’ll give you a couple of examples. One was the relevance of the same breast cancer surgery trial, local therapy trial, that I previously discussed from the plenary session. That is relevant globally – now we know that there is no point in offering surgery to patients with breast cancer that have presented at a stage 4. The second one was a head and neck cancer trial that came from the Tata Memorial Cancer Centre in India which showed that the use of metronomic chemotherapy was non-inferior to the use of IV chemo. So that should be practice-changing in my opinion. Especially the burden of head and neck cancer is more in low and middle income countries anyway so they have nicely shown that we could use metronomic chemotherapy plus celecoxib and it was non-inferior to using injection and platinum.

With regards to head and neck cancer there was one more study that came from Japan, the JCOG group, where they looked at weekly versus three-weekly cisplatin in combination with radiotherapy for locally advanced head and neck squamous cell carcinoma. Again, they could prove non-inferiority so we could give a lower dose weekly and preserve quality of life, lower toxicities. So these are the type of de-escalation practice changing trials that I am actually a big fan of because this translates to clinic tomorrow all over the world. This is about preserving efficacy and protecting patients’ quality of life.

The other one was, again, a surgery trial in head and neck cancer about the role of prophylactic sentinel lymph node dissection in oropharyngeal cancers. Again, this reiterated the findings from the Tata Memorial Centre a couple of years ago.

There was one nice poster about adding a PPI omeprazole to patients with triple negative breast cancer which seemed to have a good pathological complete response rate. Again the trial is a [?? 26:07] trial, I’d like to see a bigger RCT of this to confirm the findings.

One other impressive result that I wanted to share was the long-term follow-up results from three years versus one year of imatinib for gastrointestinal stromal tumours. An impressive hazard ratio of 0.55. So these are the types of trials that actually make a difference.

So the bottom line of the take home message from ASCO 2020 has been that it is becoming more globally relevant than before and I think ASCO should continue this trend. I would encourage ASCO to actually invite discussants from low and middle income countries to discuss these trials that are of global importance. It was very easy to do so this time because it was virtual anyway. So I think we should continue to improve in that direction, that is one feedback for ASCO on how to improve things in future. They are doing quite a good job, I appreciate that. I am additionally continually pushing to make it a more and more globally relevant conference.

The other take home message for us clinicians is that newer is not always better, we should always be careful of that. Just having a target and a drug to hit that target does not necessarily mean it is going to improve outcomes. There are a number of interventions including de-escalation of therapy that can improve patient outcomes and are not necessarily related to new drugs.

Finally, especially when we are talking about maintenance therapy or adjuvant therapy where we want to treat 100% of the patients with a drug indefinitely or for a number of years, I think it’s minimum to ask for overall survival gains. Without knowing that the drug is going to improve overall survival, especially in the adjuvant and maintenance therapy setting, we cannot settle for anything less because that is exposing 100% of patients to a drug without knowing that some of them might never need the drug and some of them who do need the drug up until relapse may do equally well by getting the drug at relapse. So we need to make sure that we improve overall survival, especially in those two settings. So those are my key take-home messages from this ASCO.