JL: Hi everybody, welcome to this ecancer update from ASCO 2022 in kidney cancer. I’m joined today by three eminent friends and colleagues. My name is James Larkin, I’m a medical oncologist at the Royal Marsden in London and I will go to my colleagues now so they can all introduce themselves. So, first of all, Jens Bedke.
JB: Hello everybody, my name is Jens Bedke. I’m a urologist with a special focus on medical treatment of genitourinary cancers and I’m based in Tübingen, that’s a university town in the southern part of Germany.
JL: Thanks very much. Toni?
TC: Toni Choueiri, I’m the only US representative here, from Dana-Farber Cancer Institute and Harvard in Boston. I do GU tumours with a focus on kidney cancer.
JL: And Bernard?
BE: I’m Bernard Escudier, I’m a medical oncologist in Gustave Roussy, close to Paris, and I am focussed on kidney cancer for many years now.
JL: Excellent, thank you all. Jens, I’m going to ask you to kick off and we’re going to think to start off with a little bit about advanced disease and have a bit of discussion there and then we will move on to adjuvant after that. So, Jens, the question is at this year’s ASCO in advanced disease in kidney cancer what did you think overall and what caught your eye? What do you think people would be interested to hear us talking about?
JB: Well, James, we’ve got several big trials which are now approved of the TKI/IO based combinations and also the ipilimumab nivolumab that has been presented at this year’s ASCO meeting. I personally found the quality of life analysis presented from David Cella quite interesting from the CheckMate 214 trial. They made a different approach and analysed the FKSI-19 score from baseline, change and correlation to overall survival and progression free survival. They also made a longitudinal approach of the change of the FKSI-19 score to overall survival and PFS. At the end they demonstrated that if you have a five point improvement in local score that you have a 31% risk reduction of death. It turned a little bit out for me that quality of life is very, very important in the treatment of metastatic disease and that it correlates with PFS and with overall survival. To be a little bit provocative here, maybe we can skip the PFS in the future and just make imaging of the patient [inaudible] and quality of life. So that’s one of the messages from the CheckMate 214 quality of life analysis.
JL: I was just going to say, so do you think the quality of life tools are robust enough to rely on them for trying to work out whether the treatments are helpful for patients? Whereas, with all due respect, OS is obviously a hard endpoint and PFS, although it’s got lots of problems, is tried and tested.
JB: Absolutely. Of course we cannot rely only on quality of life, you need the imaging. But, anyway, it puts quality of life more and more in the centre, especially if you talk about combinations nowadays and maybe triplets in the future.
JL: Thanks. Bernard, what do you think about this issue?
BE: I am not a big fan of quality of life, also I agree with James, this abstract was interesting because it’s relatively new to look at correlation between quality of life and outcome. But in my opinion patients who are doing well are those who live longer which is quite obvious for me. If you look at first line, one interesting observation has been from the CheckMate 9ER, maybe, about the depth of response. Because one question we all have is what’s happened with these near complete remission patients. What came from this analysis was that probably patients with CR or with more than 80% PR are quite the same in terms of overall survival. Maybe that probably gives us an idea of whether CR is so important in terms but deep response is probably important here. So, for me, it was maybe one of the most interesting studies which has been reported in terms of first line, knowing the fact that it has not been a great year for first line. It will probably come at ESMO but not this time.
JL: So, Bernard, can I provoke you on that? We’ve talked about this recently, actually at the meeting, as well. Could that mean that for checkpoint inhibitor containing regimens you don’t necessarily need to have a CR? I’m thinking back now with cytokines, years ago, and the importance of CRs. Is it different or don’t we know?
BE: We know for a long time that some PR, a long-lasting PR, may be almost CR and that may be those PR which are more than 80% or more, 75%. If you go back to our patients with cytokines that’s probably the case. So it gives us a good idea of what’s going on with these patients with very good PR.
JL: Just a supplementary question before I bring Toni in, do you think that means when we’re designing trials and so on and so forth we need to really be thinking about driving hard into really good PRs as a measurement of efficacy of, let’s say, new treatments or combinations or something like that at a relatively early stage of testing drugs or not?
BE: We have to better define our CR, our patients with good response. I’m very upset about patients with studies where the CR rate of sunitinib is 7-8%, which is not the case. Our clinic, it’s always 1%. So when we get 8% CR in a study it just means that the definition of CR is not what I call CR actually. But if we add to these CR patients with very, very good partial response, that’s probably what we should reach in terms of endpoint.
JL: That’s great. Toni, quality of life, how important is it? Listen, if we’re talking about what might be a potentially curative treatment in metastatic disease, can we put up with some impairment of quality of life for part of the time the patients are on treatment if we’re really thinking about cure, at least in some patients?
TC: No, I do think this is a legitimate analysis from CheckMate 214 but it only applies to nivolumab ipilimumab. And it’s in hindsight because we know already the result published that nivolumab ipilimumab has better quality of life, remember, than sunitinib. I’m more interested, not in the baseline because in the baseline these are patients that have probably also good performance status, but in the longitudinal aspect of the study which refers to health related quality of life collected after randomisation while on study. That probably is more meaningful. Then we have to decide at what time – the first questionnaire or the second questionnaire. When the quality of life maybe drops then perhaps get imaging earlier. I’m still digesting the results here, I hope we can do it in further studies. To add on to Bernard, I do think there’s a significant value if we all come together and define PR1 across studies. Then what we could do, because CRs are rare, the definition is hard and we have a lot of discrepancy between investigator assessed and independent centre review, but if we define the larger group, maybe 20-22%, that have this CR plus deep responses then we could design a study, like we did in the past phase II, single arm where they could pass to a phase III. That is the major advantage and I urge the future studies not just to report CR and overall response rate, PR plus CR, but report this endpoint. But I think the most interesting thing, despite nothing that’s going to change the standard, is adjuvant.
JL: Benard, what did you think of Martin Voss? You know the impact of subsequent therapies?
BE: The question from Martin Voss’s poster which looked at what’s the impact of pembrolizumab lenvatinib sunitinib on subsequent therapy, is it going to change the second line treatment here and what came from the study is that there is no important effect, or negative effect, of pembrolizumab lenvatinib on subsequent therapy results here. Which means that the second line PFS is not decreased by giving pembrolizumab lenvatinib. So this just confirms that pembrolizumab lenvatinib is a good standard of care and probably will still be effective in the future. So I’m not sure that it provides so much difference after this abstract but that just supports the combination.
JL: Bernard, can I ask you a supplementary question? Do you think there’s much to choose between pembrolizumab axitinib, nivolumab cabozantinib and lenvatinib pembrolizumab? Those three regimens? Maybe something on toxicity, yes, but in terms of efficacy? I know it’s cross-trial comparisons and all that stuff but would you have a favourite regimen assuming that they’re available and reimbursed and everything else?
BE: In my opinion pembrolizumab lenvatinib and cabozantinib nivolumab are very active so I would say these two regimens are the most active regimens in terms of response. That’s actually my preferred regimen. In patients with relatively indolent disease I still wonder whether we should not choose pembrolizumab axitinib or even maybe axitinib monotherapy because those patients probably have very little benefit from IO and these patients benefit from having different lines of therapy. I’m pretty sure that starting with axitinib allows many lines of therapy afterwards. So if this patient could risk indolent disease I would still probably favour pembrolizumab axitinib. Outside of that pembrolizumab lenvatinib and nivolumab cabozantinib are very good regimens and that’s my two preferred regimens in terms of aggressive and bulky disease.
JL: That’s great, thank you. Jens, can I ask you the same question I asked Bernard first about preferred regimens of those three? Then perhaps any thoughts you’ve got on the Martin Voss abstract, if I may?
JB: Of course James. I totally agree with Bernard, we have pembrolizumab lenvatinib and cabozantinib nivolumab, two very, very active TKI/IO based regimens. I think the important thing is we should remember how it started. It started with axitinib pembrolizumab, we nearly forgot about axitinib avelumab in the past, and we moved forward. This step forward is of importance that the latest results of cabozantinib nivolumab and lenvatinib pembrolizumab improved our efficacy threshold in terms of PFS, objective response rate and also overall survival. Maybe related to the more selective TKI partner or an unknown mechanism, but that’s why I prefer these, both of them.
JL: Great, and what did you make of the Martin Voss abstract? Did it make you think differently about this in any way or not? Or did you just think, as Bernard said, that confirms it’s a good regimen, right, which we kind of knew already?
JB: It did not change my mind in any way. We got the overall survival data and the hazard ratio for that to compare lenvatinib pembrolizumab versus sunitinib. What is the need or the impact of PFS2? If you have a missing overall survival benefit and you would like to question is there any reason or rationale to do a sequencing from a TKI monotherapy, that’s the comparator arm of sunitinib, then followed by a checkpoint inhibitor in second line treatment then it makes sense to give the PFS2 data. But if you have a proven overall survival benefit then it makes no sense because that is already included. The overall survival data includes the PFS2 data and, in addition, overall survival data also includes all the patients who never received a second line therapy. So if we have a look at PFS2 data we should never forget there’s a certain percentage of patients who are not included in the PFS2 analysis but who received a PFS1, so the lenvatinib pembrolizumab and the sunitinib, but were either lost to follow-up, were not able to receive second line therapy, died, or had any other cause. We all know that patients are lost from line to line therapy and this must always be kept in our minds if we interpret the PFS2 data. Better to compare the overall survival.
JL: Yes, so let’s move on to adjuvant, I think that’s a good time to do that. I might go back to you, Bernard, if I may, for adjuvant. So maybe some brief comments about where we were, let’s say, before ASCO, the new data at ASCO, impact on the field and maybe what you see for the next 6-12 months in your crystal ball?
BE: Before ASCO we had the pembrolizumab study reported by Toni which is certainly one of the most outstanding data we have. Still waiting for survival benefit and it was really well discussed in the educational session by the patients. [inaudible] and we discussed the points of your patients. They liked the overall survival benefit so we like to see that, but still pembrolizumab is certainly a very promising drug and I think it’s going to be positive. The EVEREST study has been reported too at this ASCO and although negative it does show that everolimus has still an activity. The hazard ratio was quite interesting, although not that significant, and it might be a drug to think about in selected patients based on genomic classification. But that was, for me, interesting and surprising data. I would have thought that the study would have been much more negative than that. So my crystal ball I would say we should have at ESMO maybe EMOTION-010 and CheckMate 914, so that might be a great ESMO in that view. We’ll see what’s going to happen.
JL: Thanks Bernard. So what about the subgroup analysis in EVEREST, the benefit being driven by the highest risk patients? Is that because they had early metastatic disease while you were seeing a benefit or would you just say, ‘Oh, it’s a subgroup analysis, we don’t really know’?
BE: No, I think everolimus has some activity in metastatic disease, we should not forget it. We have better drugs now but everolimus has activity and it makes sense that these patients with high risk disease who probably have micrometastatic disease, they benefit a little bit from everolimus. It makes sense for me.
JL: So, Jens, what do you think about adjuvant? It keeps coming up, doesn’t it, this whole business about overall survival, we haven’t seen it yet. Do you think we need to see it before we can prescribe the drug, it’s approved now? Or do you think we say it’s probably going to be positive so with that in mind we’re going to start prescribing it now? What do you do?
JB: I think we can make it simple, we can prescribe it right now. We’ve got a very, very strong signal for the DFS improvement. Of course we need the overall survival at the end and if that will change anything for the future then we’ll have to reconsider this indication. But now, as it stands, the KEYNOTE-564 with the positive and significant improvement of disease free survival with pembrolizumab clearly recommends the prescription. But maybe switch a little bit back to the EVEREST trial because it was very, very surprising, this trial. They were a little bit unlucky, to be honest, and I saw the hazard ratio and the p-value for that. What we learned from EVEREST, or what I learned from EVEREST, is that we have a clear indication for adjuvant treatment because that nicely demonstrated that the patients who had no adjuvant treatment with everolimus in comparison to the adjuvant treatment with everolimus, that there was a delta and there is a need for patients who are relapsing. Maybe the trial helps us a little bit to define the subgroups a little bit better. You cannot compare the adjuvant trials among each other and that EVEREST had different subgroups of patients who were included compared to 564, but in the high risk group of EVEREST the benefit was much more pronounced. I think this points a little bit towards the patient groups to be treated in the adjuvant setting.
JL: I might press you on that a little bit. We have regulatory approval and I would imagine reimbursement in Germany for pembrolizumab. Who are you treating, is it exactly as per the trial? What about the blurred lines around the edges because that’s always the question in clinic?
JB: We’ve got the approval here in Germany and we got the approval quickly for renal cell carcinoma and high risk renal cell carcinoma but what we learned is each trial has its own definition of a high risk population, intermediate to high risk population, and that also a new group like the M1 NED, the metastatic patients who have metastasectomy and have now no evidence of metastatic disease, it’s a new subgroup to be treated. I personally think that you always have to consider with the patient together in a shared decision making – what does the patient wish – and to discuss also side effects, potential side effects and treatment benefits of the adjuvant treatment with pembrolizumab. You cannot say we do not need the treatment in the intermediate high risk group or we can only do it in the high risk group, that’s a shared decision making with the patient. But if you would like to break it down, in my opinion, we should have a focus on patients which have a very high risk of relapse, that’s definitely T3b, T4, M1 or also the M1 NED subgroup.
JL: Now, Toni, your turn. Three things – what do you think about EVEREST? What do you think about RFS and OS in KEYNOTE-564 and how are you using pembrolizumab in clinic at the moment?
TC: Something that I will never do unless my life is on the line is climb Mount Everest. But now going back to the EVEREST study of everolimus, like Bernard, I was surprised about this result. We opened EVEREST, we included over 20 patients, and in hindsight 2020 if we included the same eligibility criteria as KEYNOTE-564 who said that the study won’t be positive? On the other hand, the hazard ratio for RFS will be less than KEYNOTE-564 and the OS really the hazard ratio is close to 1.0. I would love in the next version of this important academic study to learn more about OS in the highest risk patients. I think this will be important. All of us, as a community, come together and invest, at least clinically, into what constitutes high risk and continue to support trials such as the RAMPART trial that had CTLA-4 and a PD-L1 inhibitor, it doesn’t have a placebo arm. It’s the only, to my knowledge, study of cure IOs that is still open and accruing, right?
JL: Yes, you may be right. You may be right.
TC: Yes, but the other thing about pembrolizumab, I do think the signal is very obvious. We accepted DFS in other tumours, why do we have to be like that in kidney? In bladder cancer, which I argue is more aggressive, events happen earlier, we haven’t seen any OS from nivolumab at all, at least in the public domain I haven’t. The signal is strong. There is also an attempt to look at secondary endpoints that are not as important as OS but kind of make you feel more comfortable. To that end, during this ASCO meeting we have reported on three of these endpoints – distant metastasis free survival, time to first subsequent drug treatment and PFS2, which all support DFS and they were all statistically significant and, I would argue, clinically relevant because there was around 10% absolute difference between placebo and pembrolizumab.
JL: So you think there’s going to be an OS benefit?
TC: OS benefit? Yes. I do, and I’ll tell you why, not that I know. If you look at adjuvant therapies in other diseases, when you start the treatment the curves went back like this. I’ll give you an example of colorectal cancer with FOLFOX Avastin etc. There was separation and then six months later they are closed. Here you have 30 months of follow-up and patients after a year of follow-up and if you look at the curves they continue to be like this, the hazard ratio 0.52. I think it will but who knows? I think we had 66 events in ASCO GU and we need, my understanding, 132. So the last thing we want, and we were criticised, rightly so, for that, you can blame the EMA and Bernard, is every year or nine months update with 20 more events, 15 events to a meeting after a meeting. We’re trying to avoid that.
JL: So are most of your patients who aren’t going on to clinical trials in clinic who are suitable for adjuvant pembrolizumab, are most of them having it?
TC: Yes. I follow KEYNOTE 564 criteria for eligibility anyway religiously. There are situations, I have to tell you, where let’s say it’s not M1 NED for 12 months, it’s 13 months, what do you do with the patient? Or these very high risk young patients that are papillary, let’s say, that we know from KEYNOTE-427 the response rate to metastatic papillary is 25%, less than the 36%. What do you do? I think there’s a discussion and if I’m that patient I do want it. On the other hand, chromophobe, lower risk, let’s say T2 low grade, three years from resection, I would definitely not do that.
JL: We can have a whole separate discussion about this, it’s really interesting, about the groups of patients that aren’t in the trial and what we try to do in clinic when we have these patients in front of us. But I have to say I would agree with you on both of those points. The other thing, obviously, stage 4 NED is quite heterogeneous and, like Jens said as well, a lot of this comes back to sitting down with the patient and talking everything through.
JL: Thank you very much indeed. So I think we’ll wrap up that section there unless either of you have got any more you want to say?
BE: I would have something to say. We did not speak today about two drugs which are probably interesting in renal which are belzutifan and the AXL inhibitor which has been presented. I’m still concerned about what we are going to do with these drugs, especially the AXL inhibitor. So it was presented in combination with cabozantinib with a response rate which was quite compelling. But, for me, it just means that either it’s phase I data with biases to explain this response rate or it means that cabozantinib is not a real AXL inhibitor. So there is somewhere here concern about what we should do. If you look at belzutifan in the reverse way which is actually tested in combination with lenvatinib as an example, why should we combined two drugs which block VEGF? And that’s still a concern for these two combinations which are actually in development. So I’m interested in what’s going to happen with this combination.
JB: Maybe if I’m allowed another word, one abstract which was very, very interesting was the ctDNA measurement in the neoadjuvant treated patients. It was sunitinib that was given in this neoadjuvant treatment but we’re all talking about biomarkers and measurements and predictive tools and I found it very, very interesting that, first of all, it was possible to determine the ctDNA in the individual patient with such an approach. The authors demonstrated that more than 80% it was possible to get a ctDNA signature then to correlate that to the response. I just would like to outline that because it illustrates to us that we have a way of biomarkers to go forward.
JL: No, I thought that was interesting as well. The numbers were pretty small and so on and so forth but there are lots of other tumour types where ctDNA is really emerging as potentially useful. It has been a difficult area, honestly, in kidney cancer in the past. So, as you say Jens, that’s absolutely worth highlighting. Getting back to belzutifan, I guess we’re waiting to see, aren’t we, whether or not it’s going to be superior to everolimus in a randomised trial. But to your point, Bernard, do you think we’re going to see activity which is over and above or non-cross-resistant with VEGF inhibitors? That’s one question in my mind, if you see what I mean. So let’s just say, for the sake of example, you get a trial of everolimus versus belzutifan versus your choice of VEGF inhibitor, what do you think it might look like?
BE: What I can tell you from the patients I enrolled in the study is that those patients who had received cabozantinib before entering the study have very little activity of belzutifan while patients who had not received cabozantinib before are probably more active with belzutifan. So there is probably some cross-resistance, at least in part and that’s, for me, a problem. It’s still a VEGF inhibitor by the mechanism of action it has. I don’t know if it has other receptors which are interesting for kidney cancer but most of the activity is coming from VEGF. It’s a safe drug, certainly, outside of anaemia it’s a safe drug, easy to manage, but I’m not really impressed by the activity outside of the VEGF disease which is really a different mechanism of action.
JL: Indeed. Jens, what do you think about belzutifan? A similar sort of question for you, really – crystal ball again – is it going to beat everolimus or not, do you think?
JB: No, I think maybe if there is some cross-resistance between VEGF inhibition and belzutifan we definitely have a new mode of action which is used here in the aRCC patients. At the beginning we talked a lot about CR and depth of response and we all know that belzutifan, so far as the data stands, has a pronounced benefit in disease stabilisation and a little bit in response rate. That disease stabilisation takes time to convert into response. That might be an interesting discussion and question for the future – what really is the impact on depth of response of belzutifan, of this respective agent. If compared in the 005 trial, belzutifan versus everolimus, well the crystal ball but I think that it had its chance to beat the everolimus in that setting. But I agree with what Bernard has said with the combination with a VEGF targeting tyrosine kinase inhibitor like cabozantinib, that would be of interest for the future, how that reads out.
JL: Thank you.