AN: Welcome everyone, welcome to the ASCO 2019 annual meeting. It’s a pleasure to be here and discussing and introducing the field of bladder cancer treatment which is being revolutionised by new treatments and new paradigms of treatment for the patients. It’s a pleasure for me to be here with esteemed colleagues, Petros Grivas, Professor Petros Grivas, from Seattle, Arlene Siefker-Radtke from MD Anderson and Rob Jones from Glasgow. I will start first with a very general question to all of you – which is, according to you, the most intriguing challenge today that the patient may experience in receiving any kind of treatment if they are diagnosed with advanced bladder cancer therapy? Petros?

PG: Well, there are a number of challenges. Unfortunately we have advanced urothelial cancer which is usually universally a lethal disease, most patients succumb to it. So I think trying to have potential for cure in this advanced setting has been a major challenge over the years. We have these outlier responders to immunotherapy that they may live for a long time with a good quality of life but most patients eventually have recurrence of disease. So identifying patients with the use of clinically useful biomarkers and trying to select the right treatment for the right patients at the right time has been a big obstacle so far. If we identify those biomarkers with the emergence of new therapies and new agents we may have a proportion of patients who might deliver long life with a good quality of life and we do have some of them with immunotherapy. I think we’ll try to increase that pool of patients. Arlene, any comments on that?

ASR: I think it’s true that our patients have challenges now, just because of the options available to them. In the past we only gave them cisplatin based chemotherapy but now patients are coming in saying, ‘I don’t want toxicity, I want something that’s better than chemotherapy.’ Honestly, chemotherapy does have a bit of a bad name due to the toxicity associated with it and the fact that the majority of patients are not candidates for cisplatin based therapy. So I have patients coming in where the initial decision is do they take chemo or do they want immunotherapy. What makes the field so much fun right now are really the plethora of trials looking at combination strategies, combining immunotherapy with novel agents, with chemotherapeutics, with these targeted therapies, not only erdafitinib but enfortumab vedotin. The next challenge is how do we give these treatments in the best possible fashion, picking the best patients for each treatment and head into a personalised medicine strategy?

AN: Rob?

RJ: Yes, you’re right but there is still a huge unmet need. Don’t forget there are some patients who are durable responders even to first line chemotherapy; there are some who are durable responders to immunotherapy but there is still the majority of patients who have an appalling trajectory of disease. We forget about the ones who don’t even get through the medical oncologist’s door because they’re so sick with the disease at diagnosis. So you’re right, the trials are essential because we desperately need new therapies because there’s still a huge proportion of patients who really have no active treatment available to them.

AN: Yes, the issue of patient selection is crucial. You, Petros, have mentioned the fact that we may select patients according to biomarker and Arlene has led a very important trial with erdafitinib we will focus on in a few minutes, but the point here is how can we move forward beyond the immunotherapy results? On the one side we have an all comer strategy with agents like antiangiogenic drugs and we will hear from this ASCO meeting which may be the results by combining these agents with first line chemotherapy. And on the other side there is biomarker directed strategy and the first drug in this class to have provided the important results has been erdafitinib. Which is your view in general practice? Would you recommend a strategy of biomarker selected patients based on the available data or we are still supporting first the strategy of all comers first and then moving to biomarker selected patients?

PG: It’s a good question, I think a little depends on the patient and the context and the setting of disease. The development of biomarkers has been a very, very challenging process overall and also in urothelial cancer and the challenge is to have biomarkers with clinical utility, that they change the management and the outcome of patients. Right now in urothelial cancer in advanced disease we have only a couple of biomarkers that have some clinical use, one is the use of PD-L1 expression by immunohistochemistry in the first line setting in patients who are unfit, ineligible to get cisplatin. The PD-L1 expression may help them, potentially help the providers select between chemotherapy and checkpoint inhibition, however, we do not have phase III randomised trial data to make this determination even if we use PD-L1 expression. What we know is if PD-L1 expression is low in this first line setting I would personally try to use chemotherapy over checkpoint inhibition based on the interim data from phase III ongoing trials. Having said that, patients who are not eligible to get any chemotherapy, even carboplatin, they may default to get checkpoint inhibitors in the first line setting. Overall we need clinical trials to answer those questions. The second biomarker is derived from the clinical trial that Arlene and you have been involved with with erdafitinib, an FGF receptor genomic alteration. We have now almost for a month or six weeks an FDA approved option, erdafitinib, based on the phase II trial that you led which actually allows patients to have access to an FGF receptor inhibitor, tyrosine kinase inhibitor, if their tumour exhibits FGF receptor 2 or 3 mutations or fusions. I think that’s an interesting point and the question we have to answer is if we have a patient who has progressed after platinum based chemotherapy and they don’t have a clinical trial available and they have an FGF receptor genomic alteration, do you go for that agent or do you go for immunotherapy, the debate is still open. I would argue, based on the existing data that the chance of response to FGF receptor inhibitor may be more than 30-40%; the chance of response to checkpoint inhibition having the FGF receptor mutation or fusion is much lower, maybe less than 10% in that particular subset. So you may make a case to go for targeted therapy first in that particular context.

AN: Arlene?

ASR: I think you make a very valid point, just using the objective response rates to pick which patients benefit the most from which treatment. There have also been arguments based on biology – do FGFR altered tumours behave differently from non-FGFR altered tumours? They appear more immunologically cold, there has been some data suggesting they don’t respond as well to immune checkpoint inhibitors. Conversely, there’s been some data suggesting they do respond to immune checkpoint inhibitors. So our current strategies using biomarkers are still relatively crude. With FGF that does give us a fixed marker to look at, mutations don’t change, and what we have to learn with erdafitinib is how to properly place the therapy. Is it better to give it before immunotherapy? Is it better to give it combined with immunotherapy? Is it better to give it instead of chemotherapy? So there are a lot of unanswered questions that we will find answers to as a result of current clinical trials. Then you add in all of the additional biomarkers looking at immune checkpoint inhibition. PD-L1, I would argue, is a pretty crude marker, it’s dynamic, if you give treatment you can enhance expression. So what does that mean? Does the dynamism impact the response rate? Is it a fixed level up front? Then you start looking at CD8 positive TILs, CD3 positive TILs, all of the other immune and inflammatory markers, tumour mutation burden. There’s clearly a lot that we need to learn before we can reasonably use these markers to pick the best long-term treatment strategy.

PG: And also the tumour microenvironment, like the TGF-β, the EMT, gene expression profiling and also the immunosuppression sometimes.

RJ: Can I ask just a really practical question. Andrea and I come from the other side of the Atlantic, erdafitinib is not approved in Europe. At what point in real life practice as opposed to a trials practice, at what point would you be actually testing for the FGFR aberrations?

ASR: I test as soon as I know that patient has an incurable or highly recurrent cancer because when we know it is going…
RJ: So prior to first line therapy?

ASR: Exactly. Well, I’ve tested in patients who have node positive disease after neoadjuvant chemo and surgery because those patients have a very high likelihood of relapse, especially if they have multiple lymph nodes and extranodal extension. Their likelihood of relapse is probably over 70-90%. So that group of patients I will treat after neoadjuvant, arguably metastatic or stage 4. And then any patient who presents up front with stage 4 disease I will test them right away.

RJ: So you’ll be sending off the PD-L1 testing and the FGFR testing at the same time? Obviously they’re different platforms.

ASR: I actually haven’t done much PD-L1 testing, mostly because we have clinical trials that we can utilise in our patients. The other issue we face, well at the moment we have two biomarker tests – PD-L1 and mutations. What happens down the road if we have ten other biomarkers including mutation burden, Lynch syndrome profiling, gene expression profiling? If we have to send that one piece of tissue to five different places to get a result, well we’re going to have to pick which markers are the most appropriate because nobody has that much tissue. There are limitations in the tissue as well. Hopefully the field will improve and we’ll be able to go with smaller and smaller samples to do more on these very small, very valuable tissue specimens.

PG: We follow the same approach with Arlene. When I have a patient with incurable urothelial cancer I tend to send a panel to the genomic sequencing, and there are multiple potential vendors for that, and that might include PD-L1 testing but, as Arlene said, if I have a clinical trial I may favour that. But I try to get everything up front to have it available.

AN: Yes, the point is well taken and Rob raised a very important point of patient accessibility to new treatments or new tests. Don’t you think that this strategy, Petros, may further increase the disparity for patients across countries to receive or to get any new therapies or to get a new test or having their tumour analysed in a new way? Because, for example, there is still a disparity in patient access to standard immunotherapy in Europe, in many countries in Europe, compared to the United States. My fear is that the same scenario will be repeated also for erdafitinib or for enfortumab vedotin. So the patient access to new drugs will be the primary goal for governments, for regulatory people and also for us as physicians, as investigators.

PG: You raise a great point and I agree with Rob, this is a very important issue to address. I think the healthcare disparities can come across different countries, different continents, sometimes among people with different age. Access to care and access to medications is a big problem, especially in some countries outside the US. Overall we have examples not only in urothelial cancer but in other tumour types and other diseases that that has been an issue. One problem is availability, the other is cost and the question is how do you overcome that. There are countries where to get approval for a medication it’s a very long multi-step process and may take time, may take a year or two. So there is definitely a lag time compared to other countries. I think to address that should be a more global discussion with all the stakeholders and look at value-based care. Each country has different resources so to meet a certain value might be enough, adequate for some countries, but may not be adequate for some other countries with less resources. So this merits further discussion and one idea would be forums like ASCO or ESMO may be a good venue.

AN: Involving patient advocates also around the table.

PG: Involving patient advocacy and potentially industry and policy makers and many other stakeholders.

AN: Yes, very interesting. So let’s focus now a bit more in detail regarding the results that will be presented at this year’s ASCO meeting in bladder cancer. In particular let’s focus on the CALGB study in first line which is a large US phase III study which compared first line cisplatin gemcitabine with placebo or bevacizumab. The trial failed to meet the primary endpoint and I would like to know from you will we be able one day to overcome standard chemotherapy, cisplatin based chemotherapy, based also on this study. There was another study which was negative. The second question is how about the antiangiogenic strategy in bladder cancer patients which is your idea based on the bevacizumab data and based also on the RANGE study, the ramucirumab study, in the second line setting. Rob?

RJ: Yes, you’ve got to put this into the context of the time these studies were designed because this was before we’d really even heard of checkpoint inhibitors. This was a point in history where we really had made no real progress beyond cisplatin combination chemotherapy for getting on for thirty years. We’ve all been involved in multiple phase II studies and pretty much every cancer drug that has come along. I’ve been involved in several of antiangiogenic targeted therapies. They’ve all shown an element of a signal but not a strong signal. We’ve never really been able to select patients for antiangiogenic therapy. Then we’ve got now these two phase III trials and the first line one, the one we’re getting the results from at this meeting, the CALGB study which was gemcitabine cisplatin plus or minus bevacizumab, as you say, failed to show a survival advantage. Intriguingly, it did show a statistically significant progression free survival advantage as a secondary endpoint. I don’t think that’s going to be enough to change anybody’s practice. Then you’ve got the second line study which is docetaxel plus or minus ramucirumab. Again, a study essentially conducted in the checkpoint inhibitor naïve era and, of course, that met its primary endpoint which was PFS but not its secondary endpoint of OS. So I think this draws to a close the story about antiangiogenic therapy. The only thing that could save it is if we had a selection biomarker which would enable us to identify a group of patients who are going to do better with that than anything else.

AN: Arlene, you have investigated the bevacizumab combination in the neoadjuvant setting combined with MVAC chemotherapy which is your feeling in using this class of agents combined with chemotherapy in advanced bladder cancer patients.

ASR: I think the reason these trials have failed to discern a survival benefit is the fact that we’re still treating all bladder cancer patients the same. These trials were not biomarker selected and we’re learning that bladder cancer is a very heterogeneous disease composed of multiple different cancers, some of which have loss of inflammatory pathways, some have a HIF signature and angiogenesis signature. As you mentioned, when we did a clinical trial of dose dense MVAC with bevacizumab it was based on earlier data showing that angiogenesis signals were associated with poor outcomes in the neoadjuvant setting. We gave bevacizumab with MVAC, it didn’t achieve any survival difference or improved response rate that we could ascertain. But when we looked at gene expression profiling we started to see that maybe there’s an angiogenesis signal in this basal chemo signature and this trial that we published in European Urology back in 2015 gave us some early hints that maybe we could use gene expression profiling to start predicting which patients may benefit from chemotherapy. When we looked at that group of basal tumours they really did extraordinarily well with the combination, leading me to wonder would a chemo VEGF treatment work better if we selected patients based on a basal signature.

RJ: Do we know, is there a plan to look at those signatures in the CALGB dataset? Is there a tissue collection in that study?

ASR: I know they had some tissue collection, I know there’s been a high interest in looking at some of the signalling to see if we can determine any potential benefit in that group of patients.

AN: Yes, unfortunately no biomarker data were presented at this meeting but, of course, I totally agree that this is the starting point for any analysis aimed at further refining the patient selection which is crucial for an antiangiogenesis approach. I think in general we cannot close the door to antiangiogenic pending the biomarker data. Petros?

PG: I agree with what Rob and Arlene mentioned. Overall patient selection has been a major challenge with lack of clinically useful biomarkers. We did trials before with antiangiogenic agents, single agents or combinations, but the biomarker component was lacking. I think at the end of next generation sequencing and gene expression profiling we have a better chance to identify those potential candidates. I think that there will be data in the future, not at this meeting, from the CALGB gemcitabine cisplatin bevacizumab trial, maybe trying to validate the data from Arlene and her group. But I think there are also other agents out there that might be considered. There was some study with Nick Vogelzang and his group are looking at lenvatinib combinations in patients with metastatic urothelial cancer. There was some promising signal, the question in that trial was combination with immunotherapy and angiogenesis or whether this combination may have some synergies because you prune the vasculature and you might potentially enhance trafficking of T-cells into the tumour microenvironment. There you might potentially have some immunomodulatory activity as well. So that may be worth looking at.

AN: And then the boundary between first line, maintenance therapy and second line is becoming more and more flexible, I would say. So let’s focus on the second study which will be presented in the oral session by Matt Galsky and colleagues, an academic study, investigatory study sponsored by the Hoosier oncology group which is a randomised study of maintenance pembrolizumab versus maintenance placebo after response to first line chemotherapy, either cisplatin based chemotherapy or carboplatin based chemotherapy. The study was positive for PFS. So I would like to hear from you which may be the role of maintenance immunotherapy in these patients, mainly due to the fact that we have large phase III studies again already recruiting patients which may provide combination immunotherapy and chemotherapy followed by a maintenance approach. How do you frame the results of a sequence, chemotherapy, IO, in the context of the available trials? Arlene?

ASR: This is a small trial but I think Dr Galsky deserves credit for doing it, showing that there is a potential benefit at least by PFS. Whether he’ll achieve a survival benefit isn’t clear, I think he has a chance at it. But we also have to consider patients will get immunotherapy downstream and we don’t know, it’s possible that as long as you get the immunotherapy maybe it doesn’t matter as much if you get it right away versus delayed immunotherapy. The other trials that have been ongoing from several pharmaceutical companies in the front line setting combining chemotherapy with immune checkpoint inhibition or following chemotherapy immediately, those do have a chance of being positive even from the additive effects of getting a good response to chemotherapy and then the potential benefit from immunotherapy. There’s also that potential benefit from synergy – if the release of antigen can enhance the immune response, if the induction of PD-L1 expression that we see after chemotherapy plays a role in benefit. So there’s a lot of reasons that many are thinking a chemo-IO combination or sequence will be the optimal strategy. I suspect we’re going to need the results of these large clinical trials since a smaller investigator initiated trial is limited by I think it was about a hundred patients that were enrolled onto that therapy.

AN: Rob?

RJ: Yes, I think one thing that has come out of actually both this trial we’re seeing and, of course, the avelumab phase III trial which is in the same setting of maintenance therapy, one thing that I find curious, both these trials have struggled to recruit in an era where immunotherapy has been an incredibly attractive option already. I think we need to understand a little bit more about what the patient population going into these studies actually is. Is it just an investigator barrier to recruitment or is there something about these patients? Because what it tells us is that actually patients maybe don’t want more therapy at the end of chemotherapy and maybe we do need to explore that a little bit further. But I guess the upfront combination studies should ride over that because the default from the get-go is you’re going to continue the immunotherapy until progression.

AN: Petros?

PG: It’s a very interesting field and dynamic field. We have, as you mentioned, four large randomised phase III trials – the DANUBE, CheckMate 901, IMvigor 130 and KEYNOTE-361 – all of those trials are ongoing in terms of the follow-up. We don’t have any data yet so it’s interesting to see the results of those four trials. We have the switch maintenance JAVELIN trial, which is switch maintenance avelumab versus best supportive care in that same context that Dr Galsky performed his trial. So all these trials will show some interesting data and those might inform the treatment landscape the next year or two. In terms of Dr Galsky’s trial, I agree with Arlene, he deserves congratulations for completing the study in a very challenging setting. Maintenance trials have always been hard to accrue for many logistical issues, I think. But it went through. There was an adjustment in the sample size of the study, it became a smaller study, and had a crossover design. So because of the potential lack of power and the crossover it may be difficult to discern an overall survival benefit.

RJ: I’ll accept it will still give us the clinically relevant question which is is early immunotherapy better than delayed because that is the standard of care. So that’s actually a positive feature of the trial. All the patients will have had access to immunotherapy upon progression so I don’t think we should worry too much about that.

PG: Absolutely, and that’s the question.

RJ: It will answer the question or it will give us a clue to that.

PG: It will give us a clue. And even with a small sample size and small power it will still give us a clue. The question is whether PFS among the scientific community is enough to result to regulatory approval. I think that remains a big question in the field and it’s a value-based question.

AN: I think it will need a trial designed specifically aimed to answer this question.

ASR: Ten years ago if we had trials that showed improvement in PFS we would have been dancing in the streets for bladder cancer because we really had nothing else going on. But now that we have trials showing survival benefit it does become more challenging to approve things based upon a PFS endpoint. So it provides very valuable data, helping us understand how immunotherapy works. I do worry it won’t meet a final endpoint of survival and as we move into the field of combination immunotherapy I suspect there will be a lot of unanswered questions whether immunotherapy up front versus switch maintenance will make a difference if we’re now going to go into combination strategies.

PG: And of course toxicity and cost can be also relevant points.

RJ: But what we do already know is that actually you can combine these immunotherapies with chemotherapy and certainly my experience is that patients really don’t have much problem with that. So actually from a patient perspective whether you have combination therapy or you have chemotherapy and then go straight into immunotherapy is probably not going to be that different in terms of toxicity.

ASR: Although patients have options, at least the patients I’ve seen. It’s pretty uncommon for me to put a patient on single agent immunotherapy just given the large number of clinical trials of combination strategies. So speaking to accrual that might be why a trial would have difficulty accruing.

AN: Okay. So please consider that we will hear also the results from enfortumab vedotin from the late breaking abstracts from Dan Petrylak that may be a further game changer on top of this discussion.

PG: Exciting data.

AN: Yes. Okay, time is up. I would like to thank the faculty and my esteemed colleagues, Petros Grivas, Arlene Siefker-Radtke and Rob Jones and all of you for your attention. Thank you and see you at next year’s ASCO meeting. Thank you.