YL: Hello everyone, welcome to ecancer discussions. I’m Dr Loriot, I’m a medical oncologist here at Gustave Roussy in Paris. I am pleased to welcome Dr Scott Tagawa for the discussion around data on bladder cancer during this ESMO. Is it needed to introduce Dr Tagawa, Professor of Medicine and Urology at Weill Cornell Medicine? So thank you, Scott, for accepting this discussion. We will focus on FGFR alterations and inhibitors during this discussion but maybe to start off the discussion what are the most important clinical updates around bladder cancer treatment from ESMO for you?
ST: There have been a number of different, what I would call, practice-changing studies and presentations and publications in 2020 in general, some of which came at ESMO. What I would say overall is that frontline cisplatin and platinum in general is hard to beat so adding immunotherapy or even in subsets immunotherapy working better than platinum chemotherapy did not work. Luckily what we saw earlier this year is that switching maintenance from platinum-based chemotherapy to nivolumab, and a separate study for pembrolizumab, does work well. So there have been advances in terms of improving the overall survival of our patients with metastatic urothelial carcinoma. So that’s what I’d say at a very high level combining all the different studies together.
YL: Among the different therapies we have or will have in Europe, are the FGFR inhibitors. To you, what are the challenges and opportunities for the management of patients with FGFR alterations?
ST: One of the challenges initially for the average practising clinician is identifying the patients. This goes beyond identifying patients like we normally would for cisplatin based upon clinical characteristics. We need to look at tumour characteristics because if we look at the population of tumours that represent metastatic urothelial carcinoma, it’s the minority that will have an activating FGFR2 or 3 mutation or fusion and it actually is zero if we don’t test for them. So we need to get used to testing for them. We’ve had the potential luxury of not usually having to test for PD-L1 so it’s not ingrained in us like it is for, let’s say, non-small cell lung cancer due to testing. Now we’ve had access to erdafitinib in the United States for a while and hopefully that’s coming to Europe so beyond just on an academic research platform we need to get this ingrained to the general practising clinician. I don’t know if you were going to ask me this but luckily we can do this with two different types of platforms – tumour tissue itself or plasma, looking for cell-free DNA. There is a subtlety having to do with fusions that we would like to have some RNA built into the platform but luckily for urothelial carcinoma we are usually in the mutation realm rather than the fusion realm unlike, let’s say, cholangiocarcinoma. So it is easier in a genetic, DNA-only platform – we’ll still capture the majority that happen to be FGFR3 mutations.
YL: In daily practice do you test on the plasma, the cell-free DNA, or the tissue and what are the pros and the cons of the two approaches?
ST: I use both, particularly for urothelial carcinoma. The clear advantage of plasma is that it’s easy – I can see the patient sitting in front of me, I’m generally checking their CBC or chemistry anyway, I can just add that sample so it’s pretty easy. The other nice thing about it, although we have yet to find a true use for this, is the potential for serial assessments. So that’s what I would say. Luckily, if it actually works, if we’re talking about FGFR3 it does tend to be an early alteration so it’s not something that we have to worry about so much changing down the line, at least to the best of our knowledge. The downside to plasma is it is probably not quite as sensitive so if there is a positive result I believe that, if it is negative it’s possible that we might be missing something and that’s when I will go to tumour. Tumour would clearly give us the depth of coverage, the disadvantage is its heterogeneity – we’re looking at not only one tumour but one spot where the needle went within the tumour. Luckily, as we’re talking specifically about FGFR2 or 3 alterations they do tend to be early and persist so, in my mind, there’s a little less of that issue in terms of heterogeneity. But it does boil down to where that needle went in.
The other disadvantage of tumour, at least potentially is what’s available. So we can order in academic centres a metastatic biopsy, it’s not always the easiest, but it does appear that, at least if we’re looking at, let’s say, a cystectomy specimen or a nephroureterectomy specimen, if it was there, specifically talking about an FGFR3 point mutation, it probably also is going to be there down the line in metastatic disease. So there are pros and cons to both plasma and tumour testing. What is nice is that both are fairly widely available.
YL: Arlene Siefker-Radtke reported an analysis of cell-free DNA collected from the plasma in the BLC2001 study to try to identify patients who benefit the most from erdafitinib. So the question is what are your views on this work?
ST: As a general I like any study that includes biomarkers. This happens to be, as far as we know, a predictive biomarker but we don’t have so many that are not treated. But there may be certain mutations that result in better responses. I don’t know that that’s true. First of all, it’s a relatively small dataset of approximately100 patients and we don’t have randomised data yet. So I think that it’s hypothesis generating and we may be able to validate that down the line with a larger dataset, particularly a randomised dataset which is the confirmatory study. So I think it’s very nice to have so far but I wouldn’t exclude a patient from treatment, for instance.
YL: And there were some data on FGFR alterations in non-muscle invasive bladder cancer regarding the prevalence of the mutations. So what do you think about the role of FGFR screening mutations in this context in non-muscle invasive bladder cancer?
ST: The longer term dogma has been that for non-muscle invasive they’re in the majority rather than the minority of metastatic specimens and it may not actually be the majority but it is clearly in all of the datasets more common in non-muscle invasive disease. So that’s just a fact. It may be also enriched in upper tract, that I’m not so sure about but that might be also true. So I think that has implications for the use of FGFR targeting agents in earlier stages of disease because a major unmet need worldwide remains upper tract or targeted use intravesical therapy in bladder that is resistant to BCG. So again if we don’t identify those tumours we’re never going to be able to take the next step.
There is also an association with the subset that tends to be enriched for FGFR3 and whether we’re talking about TCGA subsets or other genomic subsets that might do less well, for instance, with PD-1/PD-L1 but that is not fully validated. But the more data that we get, the better it’s going to be, particularly as we’re now moving to an era of different choices in terms of therapy.
YL: Another question, the big challenge we have with targeted therapy is resistance, obviously. One strategy to try to overcome resistance is to combine targeted therapy with something else. We have updated data from the NORSE trial which combined cetrelimab, a PD-1 inhibitor, and erdafitinib. So what do you think about this combination in terms of activity and safety?
ST: It makes a lot of biologic sense, based on what I just said. So FGFR2/3 driven tumours are going to be enriched for FGFR2/3 targeted therapy and might be de-enriched or might be more resistant to single agent PD-1s, PD-L1s. So that combination does make sense. What we saw from the initial part of NORSE, which has been updated, is the tolerability where essentially the full dose of erdafitinib can be administered with a PD-1 inhibitor although those of us that have used it in trials and in the clinic in the United States will know that it’s not necessarily any multi-kinase inhibitor, or something that’s targeting FGFR1-4, is going to have toxicity. So even though we might want to up-titrate from 8 to 9 in those that don’t have a major phosphorus increase, that is a little less clear if that’s going to be the best tolerable strategy going forward in terms of a combination approach. But hopefully it’s actually not as important with the combination strategy. We know that from other combinations that we don’t necessarily need the full doses of each individual drug as there may be either additive or hopefully some synergy. The nice thing about NORSE is that it does include a randomised component so we’ll see what happens in the future.
YL: We saw very great data on biomarkers in the JAVELIN study, it was a big programme. So what do you think? What was your take-home message regarding this programme of biomarkers in this study?
ST: The take-home is that we still need more work, honestly. So one thing to remember about JAVELIN is that the tumours had to have at least stable disease to get in. So we can’t have anything there that truly says resistance to chemotherapy, immunotherapy because they didn’t get into that trial. That’s one aspect. The other aspect is it’s fairly clear that PD-1 or TMB alone or even a combination don’t tell the entire story. Those were the most ready for primetime but I would not call them primetime. The genomic signatures look very interesting to me but really hypothesis generating so we need to look at those in additional datasets. But if we don’t do the initial studies to generate the hypotheses we can then validate so it’s great to have that out there. It’s also great to know that in a large Pharma-driven study to be able to collect the studies and do the work is very important.
YL: I think we are on time so thank you, Scott, thank you for your very useful insights. Thank you for listening and see you next time at ASCO GU, ASCO, ESMO for hopefully new great data in the bladder cancer field. Thank you, Scott.
ST: I agree. Hopefully in person.
