ASCO GU 2023: Latest updates in mRCC
Dr Javier Puente – Hospital Clinico San Carlos, Madrid, Spain
Dr Bernard Escudier – Institut Gustave Roussy, Villejuif, France
Prof Stéphane Oudard – Georges Pompidou Hospital, Paris, France
Prof Viktor Grünwald – University Hospital Essen, Essen, Germany
Prof Toni Choueiri – Dana-Farber Cancer Institute, Boston, USA
JP: Hello, welcome and thank you for joining us in this online expert panel discussion as a part of the ecancer educational programme entitled Latest Developments in the Treatment of Kidney Cancer. I’m Dr Javier Puente, I’m a medical oncologist at the Hospital Clinico San Carlos in Madrid in Spain and it is my privilege to welcome you to this round table discussion meeting held just after attending the ASCO GU 2023 at San Francisco where important therapeutic advance has taken place. For these discussions we have four outstanding and well-recognised medical oncologists, all of them with a high expertise in the treatment of kidney cancer. So let me introduce all of them before the discussion – first, Toni Choueiri from Dana-Farber, here in the USA. Thank you for coming.
TC: Thank you, thank you for having me.
JP: Dr Viktor Grünwald from Germany, thank you for joining us.
VG: My pleasure.
JP: Dr Stéphane Oudard from Paris, France, thank you for coming.
SO: My pleasure too.
JP: And Dr Bernard Escudier from Paris, France, thank you for joining us.
BE: Happy to be here.
JP: Thank you, all of you, for staying with us. The objective in this meeting is to try to make an overview about the difference abstracts that have been presented during this ASCO meeting and try to gain feedback about what is the clinical impact that we have in our daily clinical practice. So we can start with the two studies that have been presented today about focussing the biomarkers in kidney cancer – the BIONIKK study and the exploratory analysis of the CheckMate 9ER. So, Stéphane, I think that you have been involved in the BIONIKK study, can you explain the study and the result?
SO: Yes, thank you. So just to summarise, the BIONIKK study was to look at frozen tumour tissue and to perform transcriptomic analysis on around 200 patients and to separate those patients between four groups, ccrcc1 to ccrcc4. Just to remind you that in ccrcc1 and ccrcc4 are those tumours which are either without any tumour lymphocyte infiltration and 4 which was highly infiltrated by T-cells. For the two types of tumour it seemed that immunotherapy and especially ipilimumab/nivolumab works much better than nivolumab alone. Ccrcc2, which is a kind of vascularised tumour type, it seemed that TKIs do as well as immunotherapy. In fact, the group 2 could be divided into two subgroups where one subgroup is highly vascularised which do respond very highly to TKIs, and the other group do respond very well to immunotherapy. So we could predict and study those patients based on this transcriptomic signature in order either to propose immunotherapy or TKI or the association of the two drugs so far. So now, looking at second line therapy, because now we have a long follow-up of around four years so far, for the group of patients who do have ccrcc2 subtype if you go for TKI it works very well even after immunotherapy and most of the patients received cabozantinib and the rate of efficacy was around 60-70% which is good for those patients.
JP: Do you think that this kind of approach in trials could help us in the future with other drugs?
SO: First of all, it was based on frozen tumour tissue which is quite complicated to do now. So now we can do it on FFPE tumour type 2 tumours and in the near future routinely we will be able maybe to a specific biomarker to allow either to go for TKI or immunotherapy or both.
JP: Toni, you have presented the data coming from the CheckMate 9ER, could you make a brief summary of this data?
TC: Yes, I think the presentation was intriguing because it’s the first time a PD-1-based combination with a VEGF inhibitor not nivolumab/ipilimumab is presented. We looked first at what was done before, especially the gene expression signature. Why? Because there are a lot of clinical trials being built prospectively based on this signature. We did not find the signature to be predictive – you have cabozantinib, it’s not bevacizumab or axitinib; you have PD-1 inhibitor nivolumab, it’s not a PD-L1 inhibitor like the other signature. But this was quite disappointing; it’s still first results. The same thing for the usual biomarkers or the ones we labelled in the past ‘biomarkers’ – PD-L1, cMET, CD8 – we didn’t see the same effect. One of my worries is that we are moving at full speed with triplet, whether that’s going to be the standard or not, which makes developing biomarkers to one class of agents, in my opinion, harder and harder.
JP: It’s a little disappointing for a biomarker predictive result in our daily clinical practice but what can we do next? What can we do in the future for looking for these kinds of biomarkers that the patients with lung cancer or breast cancer or colorectal are having? What could we do in the future in this field of biomarkers?
TC: Kidney cancer, what we have learned so far, it’s different.
TC: The TMB doesn’t apply, the CD8 doesn’t apply, the PD-L1, which has been at least less controversial, like in lung cancer and others or in bladder cancer we know the data, it does not really much apply. While continuing to collect and I remember the time I started these tissues were not collected, we are pushing the envelope towards a bit more advanced technology – spatial technologies, single cell sequencing etc. But that brings me back to what Stéphane said, that these are very difficult to do in practice. But we can use them as a discovery mechanism so we can focus then on something we can do that is more readily available.
JP: Yes. We can move to the first line setting and we have at least in the updated analysis of the CheckMate 9ER in a poster; we have also the sub-analysis of the COSMIC-330 trial. But what are your thoughts about the presentation by Tom Powles this morning? And focus in the pooled subgroup analysis.
BE: I think we had very conflicting presentations this morning which were a little bit difficult to interpret here. One, why in the poor risk patients in the COSMIC study does the benefit of cabozantinib seem to be so little while cabozantinib has been shown to be active by Toni in the past in poor risk patients? That’s something which is somewhat intriguing, although we have known for a long period of time that in the poor risk patients VEGF agents are not very active, so the limited activity is proved. The second thing, and I think we should have a view on this abstract too, is that what Mike Atkins presented was that nivolumab can be active in good risk patients which is somewhere against the guidelines that we have, at least in Europe, and which is not intuitive here because that’s patients where usually the VEGF signature is the most predominant, especially in the IMmotion study. So it makes a lot of noise about what we knew in the past. If we go back to COSMIC my personal view is that, so far, we don’t have any reason to use triplet until we see OS benefit because just increasing PFS is just because we add cabozantinib, which makes sense. But if we don’t see any OS benefit in the OS signal I don’t think we should use triplet, at least with this triplet.
JP: Absolutely, yes. You have mentioned the abstract presented by Dr Atkins in the GU16 trial using nivolumab up front; they incorporated the concept of the treatment free survival. Viktor, what are your thoughts about that? Do you think that this is relevant in the daily clinical practice?
VG: I think it’s relevant because I think that’s what we do, isn’t it? We treat and the hope is really to achieve the state where you don’t have to treat anymore, so you can withhold therapy and you still maintain response. That’s a very good outcome, actually. So it’s very hard to measure and we heard this discussion during the day in how to measure exactly and what does it mean with those numbers when you talk about treatment free intervals. But I would say it’s my reality, it’s what I do with my patients and we try to stop therapy and withhold it. The question is what is the risk? You addressed it, Toni, and said how many patients did we have progression at different sites, maybe at sites that then turn out to be symptomatic? It could be seen as more difficult to treat sites that have a lot of impact for quality of life in patients and I think that’s something that you need to take into consideration when you embark on such a journey, that’s quite important.
JP: When you interrupt the treadmill because of a grade 3 or grade 4 toxicity and you prolong the TFS, do you think that this kind of progression that you mentioned during the discussions about bone mets or visceral metastases is common or is it something…?
TC: You know, it’s really not common, I agree with Dr Atkins. But not common doesn’t mean doesn’t exist and it doesn’t mean for one patient in your practice, even, that disease was controlled and now the disease is completely not controlled and growing in sites that were not causing problems. So I think we need to be very careful when we stop treatment. When you have immune related AEs that are severe enough and patients have a deep response this is a reason to stop because of safety. This is actually what happened in CheckMate 214 – that was not built in, nivolumab was continued until progression. Patients that stopped, they had a reason, immune related AE was a reason why they stopped. Sometimes they wanted to just stop treatment and continue. So it’s a bit of a biased approach compared to practice.
JP: This is something that we learned with a TKI alone because there are some studies when you can stop the treatment with a TKI alone and maybe then re-challenge with the same therapy, it could be useful in some cases. We have to learn about that.
TC: Absolutely. I think also the panel will agree with me here that CONTACT-03 and TiNivo-2 where we re-challenge or continue the same pathway, I would say, as close as possible pending the effect that the drug is going to be different anyway, so how much? I think it’s going to shed another piece of evidence about continuing hitting that pathway.
BE: I think there is something we never take into account in our models which is, one, tumour burden and that’s probably important when we speak about patients.
JP: Absolutely, yes.
BE: And, second, tumour growth or aggressiveness. So we probably have to find a way to incorporate this characteristic into our models which is very difficult, difficult in database and so on. But I’m sure it’s not the same: if you have a patient where the tumour is growing within three months rapidly it’s not the same as the very long disease we have in some good risk patients.
JP: And finally in this part of the discussion in the first line, Toni, you have presented in a poster the updated analysis of the CheckMate 9ER at three years.
TC: Yes. It was Dr Burotto.
JP: Any thoughts about that?
TC: No, I think Dr Mauricio Burotto from Chile presented that.
JP: Yes, in a poster.
TC: It was part of the work… actually it’s this afternoon in oral presentation and a poster. I think the data is good; you don’t want to present at every meeting another update but here we have 44 months, I think that’s important.
JP: Three years, yes.
TC: For patients are there any new safety signals long term and the answer is no. We have a median OS that went up by almost a year; CRs that are still over 10% which I like it when VEGF and IO are together.
JP: Even in favourable risk 40%.
TC: Absolutely, so over 10% the CR, so I think that’s good. We started, when you look at OS and cabozantinib/nivolumab, we started maybe to see a tail of the curve. Of course we say we need biomarkers and you get the biomarker analysis that was negative.
JP: Yes, so let’s move to the second line because we have another abstract that has been presented today – big topic, the CaboPoint study, you have been involved in that? Can you mention, can you make a brief summary to the audience?
VG: When we started to build that study it was really about the issue that we have, cabozantinib being built in to our guidelines, it has usually strong recommendations as second-line treatment. But the data that was generated, it was really generated after TKI failure so that was the sequence: TKI failure and then you use cabozantinib. So the question is really if you have a changing treatment landscape using IO combinations up front what’s the delivery, really, of cabozantinib in that sense? That’s why we really did that study and we’ve investigated two cohorts, one of which is after ipilimumab/nivolumab because I think we just lack evidence here for cabozantinib, or prospective evidence. Then there is the TKI/IO cohort that has been investigated. Overall it was quite good, we had about 30% response rate in the total cohort. It was a bit higher for those that had been previously treated with ipilimumab/nivolumab, it was 32%, it was 25% for those that had TKI/IO previous exposure. So I think it’s very solid.
JP: Absolutely. This is another abstract presented and a poster from the UK in the same way of cabozantinib activity in second line after TKI plus immunotherapy. But I remember the retrospective analysis of the TKI activity with cabozantinib or with axitinib and in the nivolumab/ipilimumab population the median PFS was around one year or ten months but in patients that had been receiving a TKI plus IO, in a retrospective analysis, the efficacy was lower in the retrospective. For the first time you have prospective data that this is similar in high activity.
VG: Yes, it’s quite exciting isn’t it because it’s the opposite of what you would expect. Because if you treat somebody that is TKI naïve you expect more punch, basically, from the TKI. But the caveat is really we have to wait for PFS. This is the interim analysis, we have shown responses but we are just lacking the time-to-event curves, the progression free survival. I think we have to wait to see how they look like in order to make more sense.
JP: Absolutely. There is another abstract evaluating the activity of another second line like lenvatinib plus everolimus in real world data. What is the treatment landscape in these situations? Obviously you use cabozantinib plus nivolumab in the first line based on the CheckMate; what are the other options?
TC: I mean, the TKI that hasn’t been received, it’s as simple as that which is not great but we don’t have a phase III for everything. Lenvatinib is married, and still going to be married, to everolimus when it’s not married to pembrolizumab, that’s the fact. So I tend to follow guidelines and when I want I give them together which makes cabozantinib and everolimus a bit earlier. Having said so, lenvatinib/everolimus does have activity. Another drug also approved in the US which is the only case to my knowledge of comparison of a modern TKI to another VEGF TKI is tivozonib. Tivozonib is well tolerated, higher activity than sorafenib and especially in PD-1 inhibitor treated patients. Now it is being used with nivolumab with the phase I actually done at Gustave Roussy; it’s being used with nivolumab post-IO to see if continuing nivolumab or PD-1 inhibitor makes any sense.
BE: One of the big problems we have, I guess, for TKI is that we have two very strong TKIs which are cabozantinib and lenvatinib, that’s probably the strongest one. The question is should we use them in second line or can we keep them for third or fourth line and that’s the big question, especially for patients who receive nivolumab/ipilimumab. We had a question in one of the sessions this morning, what do you use after nivolumab/ipilimumab, and I think in many patients jumping immediately to a strong TKI might be a mistake because we still have a good activity of tivozanib or axitinib at that point and keeping cabozantinib or lenvatinib for later lines makes, for me, a lot of sense in some relatively indolent disease.
TC: I struggle with this, I think it’s not as easy. Bernard is my mentor so I blindly say yes but on the other hand I see patients that are not able… that I thought for sure were going to get a second line that, for whatever reason, are not able to get second line. I wish I do. So I think we are struggling with this.
JP: I think that many of us are recruiting patients in these trials where they are combining cabozantinib plus immunotherapy or another…
TC: Yes, after PD-1, yes.
JP: What are your thoughts about that, what are your feelings about that? I’m recruiting patients on that and I have my… In my experience I think that maybe I didn’t see a great response on that.
TC: I think the primary endpoint, at least in CONTACT-03, is PFS. I think it’s overall going to be the whole package what we accept – PFS, response and OS. Patients that progress on PD-1 and VEGF TKI maybe you have enough events to not blame the power here for OS. I’ll be honest, I do not know.
BE: I don’t know either but one big point is that I’m absolutely sure that PD-L1 is not as good as PD-1 in our… So I will wait more from the TiNivo than from the CONTACT; I’m not very optimistic about the CONTACT, I may be wrong. I think TiNivo will better answer the question. If both of them are negative then we probably will know that continuing PD-1 blockade is not the thing to do.
TC: The other thing I would add, both are going to enrol mostly, CONTACT-03 for sure, TiNivo also, patients with metastatic disease. The conundrum here is after progression if pembrolizumab becomes really the only adjuvant; it’s apparently the only one where we have an OS benefit, it’s going to be used more and more, what do you do after progression on adjuvant pembrolizumab? Is it the same biology? Would you re-challenge and when? I think that this is an open question.
JP: I think that we have enough time to review another abstract, for example, even in the adjuvant setting I think we have a poster evaluating the UCLA staging system.
TC: Yes, the UISS. These classifications, how to call high risk patients, every one of us here, their institution have their own way to do it. With KEYNOTE-564 we did what we thought is the most reasonable thing, it ended up well. But no matter how you dice this… UISS we get a lot, when we went to the AUA, that you have to look at UISS. So we collected… we couldn’t look at the Leibovich score because we don’t have tumour necrosis and, really, tumour necrosis is not mentioned in all pathology reports. Even the USPS and pathologist NGU, sometimes, despite almost mandated, things are not mentioned – tumour necrosis is one. But we did the UISS, two things – one, most of the patients were intermediate high by UISS, like by KEYNOTE-564 criteria, and, second, the benefit we saw in the 564 criteria was the same in UISS.
JP: I think that many of us are using the inclusion criteria and trying to use it with pembrolizumab in the adjuvant setting.
BE: But still we have a lot of patients referred, they want to receive pembrolizumab, of course. We really have a very different patient population and I think the benefit is mostly in high risk patients, so metastatic patients, M1 patients, that’s probably the patient population, but mostly high risk patients. So I’m always anxious when a patient with a small T3a, just because there was a little bit of margin or whatever, and grade 2 tumour wants to receive pembrolizumab. I don’t think that’s the best way to do it but…
TC: We looked at that, the hazard ratio was less than 1.0 but it’s a smaller… so, of course, that’s different than M1NED. You add to it, let’s say, in addition that we don’t have no tumour necrosis and a tumour less than 4cm and I bet you then it becomes really a conversation.
TC: But I agree. What we did was KEYNOTE-022 which is version 2.0, is that, one, we accepted microscopic margins that are positive because there was no reason to exclude. Second, this M1NED one year was arbitrary. There is work from when Bernard and I were in Cleveland that up to two years is okay. After two years it’s usually… so we included those two and I think it’s going to be a bit more inclusive in general.
JP: Finally, the abstract that I want to review is the nice presentation this morning about the management of brain mets. Maybe, Stéphane, you can mention what is your feeling on that?
SO: Yes, it was interesting because we had the incidence of brain mets, around 8%, which leads to the question of whether or not we need to perform a CT scan just to say whether or not the patients do have brain metastases and how to handle those patients so far. They found that it seemed that immunotherapy worked maybe better than TKIs but in daily practice TKIs also, especially cabozantinib, do work well. So we need to find out in a prospective way how it works with these brain metastases. They found out that also we need to perform a local treatment such as neurosurgery or radiation therapy at the brain metastasis level which maybe increases overall survival, which is a good point for the patients. They found out that now, from a decade ago, they have more neurosurgery than radiation therapy, maybe in order to prevent the total brain irradiation so far and maintain a good quality of life, so it’s a good point.
JP: It is a good point because obviously those patients have a good overall survival, can still be alive a long time and you reduce the side effects of this whole radiation because you reduce these mental disorders that they can suffer.
TC: I think we have to support the French study. The French always take on very difficult studies like the phase II of cabozantinib, why is that? And brain metastases untreated. I’m lucky at my centre, I have access to SRS but for the rest of the world there are many places you do not have access or SRS is delayed. So if we have therapy now that single agent PD-1 and single agent older TKI have failed us in brain metastasis, if we have agents that can control the brain metastasis while maybe we can get more localised treatment is very important. So this is a cry for help to support not the French soccer team but the French CABRAMET study.
SO: You’re right, that was done, yes.
BE: Yes, it’s important.
JP: I think that we have reviewed all the most important abstracts that have been presented. Thank you, all of you, for your practical insights. I will say thank you, all of you, for being here and thank you ecancer for their support to make this kind of meeting. Thank you so much.