Biology and Genomic Distinctions in Metastatic Urothelial Carcinoma and Upper Tract Urothelial Cancers - Andrew Hsieh - Bishoy Faltas - Matthew Galsky

August 2, 2019

Petros Grivas hosts Matthew Galsky, Bishoy Faltas and Andrew Hsieh in a discussion about their research programs and recent findings that drive closer to providing personalized medicine in the treatments in metastatic urothelial carcinoma and upper tract urothelial cancers.  Groundbreaking work is discussed in this podcast and the three platform papers that discuss this evidence in more detail are linked below as related content.  


Petros Grivas, MD Ph.D., Associate Professor, Clinical Director of Genitourinary Cancers Program, University of Washington, Associate Member, Clinical Research Division, Fred Hutchinson Cancer Research Center

Matthew Galsky, MD, Director, Genitourinary Medical Oncology and Associate Medical Director at the Cancer Clinical Trials Office at Mount Sinai Health System and The Tisch Cancer Institute

Andrew Hsieh, MD, Assistant Member, Human Biology Division, Fred Hutch, Assistant Member, Clinical Research Division, Fred Hutch, Assistant Professor, Department of Medicine, University of Washington, Affiliate Faculty, Genome Sciences, University of Washington, Dr. Andrew Hsieh is a physician-scientist who balances the clinical care of prostate and bladder cancer patients with running a research laboratory at Fred Hutch. He studies how cells build proteins — and how cancer cells can co-opt this process to better grow and spread. Through these studies, his team aims to develop new drugs that can selectively target these changes and ultimately impact the lives of patients with cancer.

Bishoy M. Faltas, MD, Director of Bladder Cancer Research, Englander Institute for Precision Medicine, Weill Cornell Medicine

Read the Full Video Transcript

Petros Grivas: Hello, I'm Petros Grivas. I'm a medical oncologist at the Seattle Cancer Care Alliance and Associate Professor at the University of Washington, also a member of the Fred Hutchinson Cancer Research Center. I'm very honored today and excited to host three remarkable scientists and researchers in the field of bladder cancer. We're having Dr. Andrew Hsieh. Andrew is a physician-scientist and Assistant Professor at the Fred Hutchinson Cancer Research Center. Has done tremendous work in the field, mainly in interrogation of the biology of bladder and prostate cancer and he's also a medical oncologist in patients. Andrew, welcome.

Andrew Hsieh: Thank you. Good to be here.

Petros Grivas: We also have Dr. Bishoy Faltas, another great colleague and a major figure in the field. Bishoy is also a medical oncologist at Weill Cornell in New York. He's running a lab there and he's also seeing patients as well. Bishoy, welcome to the group.

Bishoy Faltas: Thank you, Petros.

Petros Grivas: And we also have Dr. Matthew Galsky, who is Director of the Bladder Cancer Program and Professor of Medicine and Urology at Mount Sinai Icahn School of Medicine. Welcome, Matt.

Matthew Galsky: Thank you.

Petros Grivas: So, I'm really happy to talk to all of you today. We have emerging data in the field of bladder cancer and upper tract urothelial cancer and all of you recently published some remarkable, I would argue landmark papers in the field. So, I would like to start the discussion by having you just give a brief outline of the papers that were recently published, because there are some common features there I think to impact the field in a similar way.

And I will start with Andrew. Just give us an outline of a manuscript that was recently published from the University of Washington, outlining the importance of the research autopsy program that we have here and particular take-home points from the paper to start us off.

Andrew Hsieh: Sure. Thank you very much, Petros. So, we here at the University of Washington and the Fred Hutch have developed a bladder cancer-focused rapid-autopsy program, which enables us to harvest a very late stage of bladder cancer tissues, multiple samples from the same patients after they pass away. And we started this program in 2015. It has gone on till now. What we decided to do after the first seven patients was to conduct a whole exome sequencing on the first seven patients, which included tumors from... multiple tumors from the patient's body, as well as the primary tumor,  compared to germline.

And it was a very interesting study because what it enabled us to do, was to delineate intra and interpatient differences at the genomic level. We specifically studied single-nucleotide variations as well, copy number changes that were either shared or different between patients.

Fortuitously, it turned out that our first set of patients actually evenly divided between upper tract urothelial carcinoma and lower tract urothelial carcinoma. So, it gave us an opportunity to look at some of the biological differences between these two cancer states in urothelial carcinoma. And really, what we're able to do is to actually understand intra and inter heterogeneity differences as they pertain to SNV and CNV differences, or also the to make some inferences on the origins of metastatic bladder cancers.

I think one of the important things to note about the study is that of course, it was a small study. There were always seven patients involved, although we did have multiple samplings from the same patients, is that something obviously that we're working on right now.

Petros Grivas: Thank you so much, Andrew. But can you comment a little bit about the infrastructure needed overall and what are the elements that need to be in place? You referred to that... just for the audience understand exactly what it takes to get this effort done.

Andrew Hsieh: Absolutely. So this program really is an example of the team nature of how science is being done in the Pacific Northwest. As you know, there has been historically a very well known rapid autopsy program in prostate cancer here at the University of Washington, Fred Hutch, which has been in operation for over 20 years.

About four years ago, Doctors Bruce Montgomery, as well as Dr. Jonathan Wright, started talking about the possibility of doing a rapid-autopsy program here, focused on bladder cancer, because of just the obvious need for that type of program in the world. And that to be honest, was one of the major pull factors for why I decided to join faculty here in 2014. So, within the year, I think one of the big changes that made a big difference was that we actually had the goodwill of the existing prostate cancer rapid-autopsy program. They had actually donated a year's worth of work to this, to enable us to collect basically the first three patients. And what that meant was that we have a team of technicians that are on call 24/7 that allows us to basically... When a patient passes away, to bring the body here to the University of Washington, to do the autopsy and then collect all the tissues, prep all the tissues, and then, of course, deliver the patient's body back to the family.

So, it's a huge undertaking. And through the... basically the goodwill of the prostate cancer rapid-autopsy program, we're able to get seed funding from donors, as well as some seed funding from the Fred Hutch. And that made a huge difference.

So, that's really what nucleated this program around 2014, 2015.

Petros Grivas: Thank you so much, Andrew. This is a really exciting effort over the last few years and it's very, very nice to see the deliverable product from that to work.

I would like to move to Bishoy now. As I mentioned, Dr. Faltas, he's a Professor at Weill Cornell University. Bishoy, your team has done significant work in the field and you recently published this really exciting paper in Nature Communication, outlining the biology of upper tract disease and try to compare it with lower tract disease and very, very... I would say, in line with the paper that Dr. Hsieh mentioned before. Can you take us through that take-home points of the study and the key findings?

Bishoy Faltas: Sure. Thank you, Petros. So, first of all, I'd like to congratulate Andrew and you, and the rapid-autopsy program. We also have a rapid-autopsy program here at Weill Cornell and I completely agree that it is a significant undertaking and it takes a huge team effort from pathologists to clinicians, to the oncologists, and ultimately to the the patients and the goodwill of their family and their understanding of the importance of what we can learn about the biology of the disease from dissecting these tumors, and then returning the bodies of their loved ones to the family.

So, I think it's an amazing program and it's a tool that has been underutilized for a long time. But I'm really glad to see that we're starting to utilize that... those kinds of studies to learn more about the biology of the tumor.

In terms of our paper that we recently published in Nature Communications about upper tract urothelial carcinoma, so it's a really a collaborative effort with Olivier Elemento, Brian Robinson, who's a pathologist at Weill Cornell, and Panos Vlachostergios, who was a medical oncologist at Weill Cornell. And this is a multi-year effort. It's also a collaboration with Surena Matin and Seth Lerner because as you know, these tumors are actually fairly uncommon. So, to collect the number of samples that would inform us about the biology of these tumors, we needed to really try to get every tumor that we can get our hands on.

This effort actually builds on recent studies from John Sfakianos, David Solis and others over the... Jonathan Coleman over the past few years that actually focused on ... and Surena actually, that actually focused on the more or less the genomic aspects of urothelial carcinoma. And these studies showed a high prevalence of FGFR3 alterations in these tumors and we were able to confirm that in our sample from Weill Cornell. And the focus of our study was really to try to understand how these genomic alterations shape the immune contexture and the transcriptome of upper tract urothelial carcinoma.

And we can look at this in a... whether you're a lumper or a splitter, you can look at this in a lot of different ways, whether upper tract is different from lower tract urothelial carcinoma or weather, the way I would actually prefer to look at it, that upper tract represents a distinct spot on the continuum of a lower tract urothelial carcinoma and not so much a difference, but a distinction.

Now when we looked at upper tract urothelial carcinoma, we found that, as I mentioned, that there is an enrichment of FGFR3 alterations. And that is actually quite interesting, given that we also found that the prevalent mutational process that was responsible for for the mutational profile of upper tract was the APOBEC induced mutagenesis mutational signature. And this is quite interesting given recent work by [Mali Aprotinin 00:10:32] and others... and [Holarthritis 00:10:35] that was actually published in European Urology, showing that the APOBEC mediated mutagenesis process may actually be responsible for some of the FGFR3 hotspot mutations, such as the S249C mutation.

So that was actually quite interesting to us. Now for a very long time, there has been an association between upper tract urothelial carcinoma and Lynch Syndrome, which results from a general loss of function mutations in canonical MMR genes, such as MLH1, PMS2, MSH2 and MSH6.

So, this association is definitely true, but actually, the truth of the matter is, that most of upper tract urothelial carcinomas are sporadic, meaning that they're not associated in with Lynch Syndrome or with these germline mutations. So, we wanted to actually look at these sporadic upper tract urothelial carcinomas and see whether we can see also some dysregulation in these... whether there is a dysregulation of the expression of these genes in sporadic upper tract urothelial carcinoma. And it has been reported before, several years ago, using other methods such as immunohistochemistry, that there may be indeed a decrease in the expression of these genes in upper tract urothelial carcinoma, but... and that somehow that may have something to do with this known association with Lynch.

But at that time we didn't really have very sensitive technologies to understand whether... to diagnose a latent Lynch if you will, and also to differentiate sporadic from Lynch-associated upper tract urothelial carcinoma.

So, we thought that this area... There's a gray area here that would benefit from clarification. Now the real litmus test in our minds for whether you have a defective DNA mismatch repair processes is whether you really see an increase in microsatellite instability. And nowadays from whole-exome sequencing data you can measure that using a tool essentially called MSI-sensor. And when we looked at the upper tract versus this [J 00:12:54] lower tract or bladder cancer data, we found that surprisingly the majority of them, or almost all of them, were really microsatellite stable. In other words, that decrease in the levels of these genes did not actually translate into increased microsatellite instability.

And when we looked at the tumor mutational burden, we found that the tumor mutational burden was in fact lower in upper tract urothelial carcinoma compared to bladder cancer. Definitely not higher. And this was in concordant with data from Andrew and also with other data from Dr. Solit's group. So, I think was actually quite interesting to us.

And then, the next part that we looked at was really how do these upper tract urothelial carcinomas cluster within the subtypes, the expression-based subtypes of bladder cancer. And we looked at this in a lot of different ways using the UNC classifiers, the TCGA classifier, the MD Anderson classifier. And we found that no matter how we looked at it, they always seemed to be clustering with the luminal subtype. And if you use the TCGA classifier, they would classify with the luminol papillary subtype. And we looked at this using two different statistical methods and we found that this seemed to be true using different classifiers and using two completely different statistical procedures.

And this suggests to us that this is actually a stable and probably a defining feature of upper tract urothelial carcinoma. And this goes along with the prevalence of FGFR3 alterations in these tumors, which coincides with luminol papillary subtype.

And then, finally we try to understand what is really the immune contexture... what does the immune contexture of these tumors look like? And we developed, we looked at these tumors and we looked at the top 170 genes that were deferentially dysregulated between the T-cell inflamed and the T-cell depleted clusters. Dr. Randy Sweis from the University of Chicago has done something similar before in lower track tumors and that we developed a similar, but actually a different set of genes that look at these T-cell high and T-cell depleted clusters. And we found that, to our surprise, most of the upper tracks actually clustered in the T-cell depleted subgroup.

And these observations were similar in our cohort and also to the MD Anderson Baylor cohort. So, this seemed to be true no matter what. No matter which cohort we looked at, there was actually a third dataset that we had looked at using a different type of expression profiling and this seemed to also hold up in this data.

So... And then finally, we looked at whether FGFR3 can really be coordinating all these features together, and we noticed that the FGFR3 expression is really an outlier in these tumors. And that's not necessarily driven completely by mutations or fusions. So, some of these cases actually do not have mutations or fusions that we could identify. So, we don't really completely understand what drives the high FGFR3 expression in these tumors.

But it was certainly a very obvious feature in these tumors. And then we looked at a dataset using shRNA in a bladder cancer cell line. And we found that knocked down of FGFR3 using shRNA, results in up-regulation of genes and the Interferon Gamma response pathway, namely BST2 and another gene called IRF9, which is a master transcription factor that drives the response to interferons. And did a pathway analysis. We saw that all these pathways related to response to Interferon Gamma were up-regulated in this. I'm very glad to see that data from Dr. Galsky is also... Galsky's papers are also converging on this point.

And then finally, we asked whether we could replicate this using the FGFR3 inhibitor Erdafitinib and we found that we could actually replicate the same effect showing the up-regulation of BST2 and actually IRF9, which is not shown in the paper using this pharmacological FGFR3 inhibitor.

Petros Grivas: Thank you, Bishoy. This is really, really exciting paper and thank you for outlining all the key messages in this very, very long work. Before I ask Matt a few questions, let me ask Andrew, his take on the luminal subtype of upper tract disease and if you can place the two papers together in context, Andrew, of any thoughts about next steps and the biology of disease and potential, I would say treatment implications, upper versus lower tract and also your comment about biopsy of metastatic versus primary tumors.

Andrew Hsieh: Oh, that's a lot of questions. So, I think... So, in terms of the luminal phenotype, that's something that we're very interested in looking at, particularly across metastatic tumor samples. I think Bishoy's finding was very exciting given the fact that, we know that these subtypes do have implications for clinical outcomes. So, it is interesting to us obviously that, with upper tract disease the clinical outcomes are different, right? And with this luminal phenotype, I think it helps redefine that particular subtype as it pertains to lower tract versus upper tract disease.

In terms of our particular study, unfortunately we haven't done... We're in the process of doing the RNA-Seq analysis. What I'm particularly interested in is this concept that, we mostly define these subtypes on prime... These subtypes have been defined on primary tumors.

And the big question is, what do these subtypes look like within an entire patient? Right? If we had five different tumors from different areas in the body, how would these look? I think there are some implications. Bishoy has done some previous work with their rapid-autopsy programs, kind of giving a flavor of that. And I think there's a lot more to study in that context. I was really excited actually to see Bishoy's findings with the TMB because the only implication... the only findings before that were really from the Moss paper published in the EU in 2017 showing that... It was really in the discussion section where they said, "Hey, the TMB seems to be lower in UTUC," but I think the most definitive finding of that really is Dr. Bishoy's paper, which really shows... Sorry, Bishoy Faltas' paper, that basically shows that there is actually a decrease in the TMB.

We were very interested that we also saw decreased TMB in our small cohort and wanted to ask the question about MSI. So we used the exact same MSI measuring pipeline. And we also saw that none of our UTUC patients had indications of MSI or any mutations to Lynch Syndrome genes. So, I think the findings are very solid and it's great to see this agreement between these three studies.

I think it's important to mention that there have been some recent studies and Bishoy, I'm very curious about your thoughts about this one on the Audited paper, right? That looked at some UTUCs as well, which suggests there might be increased TMB. However, there, I don't think there was actually a distinction between MSI high and low and that context was all-comers. And to me, that's one of the defining features and maybe why our findings were different from that.

Bishoy Faltas: Yes, absolutely. I think that's a great observation. So, actually, I think that if you actually look at the data carefully and I'd be curious to hear if they would agree with that, that actually their data is concordant with ours because this, as you mentioned, this was all-comers. And actually, this included Lynch Syndrome patients.

Andrew Hsieh: Exactly.

Bishoy Faltas: So, if you look at data, only the patients who had Lynch Syndrome were the patients who really had high TMB. The patients with sporadic upper tract, which are the majority of patients, had a low TMB, or let me say a not high TMB. So, I think that you can easily reconcile that with the data from your paper and from our paper. And I think, this is fairly definitive data that if you were to find a high TMB, then that should actually trigger a testing for... should trigger testing for undiagnosed Lynch, because in patients who have sporadic upper tract urothelial carcinoma, it'd be pretty uncommon for them to have a very high TMB unless there is a loss of... a second hit if you will, with a complete loss of function.

And the reason we think that there is this difference is I think that... The fundamental difference between these loss of function mutations that occur in the germline and the dysregulation, if you will, at the expression level that happens in somatic upper tract urothelial carcinoma, is that you really need a complete loss of the protein, not just a downregulation in order to have microsatellite instability.

Andrew Hsieh: Mm-hmm (affiamative) Mm-hmm (affiamative)

Petros Grivas: Pretty interesting points guys and I... Every time I discuss with both of you, I really, really enjoy the implications of an understanding of bladder cancer and upper tract urothelial cancer biology. Let's move on and ask Dr. Galsky. He recently published a very important paper evaluating mechanisms regarding a response to checkpoint inhibitors PD-1, and PD-L1 inhibitors in metastatic urothelial cancer.

There has been this notion that patients with FGF receptor three activating mutation or fusion might have a more immunosuppressive microenvironment that could impair responses to checkpoint inhibition. And this has come through preclinical data and some clinical observation.

Matt, do you want to take us to the main findings of your study and potential implications for the treatment of those patients and clinical trial designs?

Matthew Galsky: Sure, absolutely. So, as has been mentioned already, there's been this observation that tumors that are FGFR3 mutated or tumors with increased FGFR3 expression associated with decreased signatures of T-cell infiltration, Bishoy nicely showed this in his paper and it's been shown in some others as well, including really the TCGA papers. That's been interpreted as these tumors being less responsive to immune checkpoint blockade. And I think there are a couple of dangers making that conceptual leap.

One of those dangers is that an association certainly doesn't mean causality. And this, the second danger and of course, and hindsight is 20/20, right? But we know much more about this now than we did in the past when some of these initial observations were made. But lack of T-cell infiltration based on gene signatures or based on immunohistochemistry does not preclude a response to immune checkpoint blockade, nor does the presence of T-cell infiltration guarantee a response to immune checkpoint blockade.

So, those tumors might be potentially less responsive to treatment overall but certainly can respond. And one of the reasons that they can respond or one of the observations is that some other biomarkers, predictive of response and resistance to immune checkpoint blockades like TMB are very poorly correlated with these measures of T-cell infiltration. So, TMB high tumors have a higher likelihood of response.

T-cell infiltrated high tumors have a higher likelihood of response, but there's actually a very poor correlation between those two measures in individual tumors, very nicely shown in a paper from David Kaufman's group at Merck, in a paper, I Believe in Science in the last year. I Believe in Science.

And that's an observation that other people have made as well. So, this notion that FGFR3 mutant tumors might be less responsive to immune checkpoint blockade is an extrapolation and certainly something that should be tested directly now that we have the tools to test it directly. So, we took two large clinical trial data sets IMvigor210, which is a phase two study of Atezolizumab, and CheckMate 275 which is a study, phase two study of Nivolumab and looked at the correlation between FGFR3 mutations and response. And we found no correlation. Wild type tumors and mutant tumors responded very similarly.

When you look at FGFR3 at the impression level, there's also no correlation between response and... And so, I think at this time, certainly, there is no solid data to exclude patients with FGFR3 mutations from receiving immune checkpoint blockade.

But now comes this other part of this story, which is whether or not this was an association or whether or not there's anything causal between FGFR3 mutations and lack of T-cell infiltration in the first place. Because even though it's not an absolute correlate of response to immune checkpoint blockade, it still might offer insights into immunobiology. And we hypothesize that maybe the similar response rates are because FGFR3 mutant tumors might be linked to less T-cell infiltration, but might have an imbalance in some of these other predictive biomarkers. And what we found was that the FGFR3 mutant tumors or the FGFR3 overexpressing tumors compared to wild-type had similar TMBs, had lower T-cell infiltration and had a lower expression of some of these repressive stromal signatures.

And so overall, we think some of these features in the tumor microenvironment might be "balancing out". And that's why these tumors might respond similarly to immune checkpoint blockade. But we did see that there was less T-cell infiltration, at least based on this gene expression surrogate of T-cell infiltration. And similar to what Bishoy had done, we wanted to see if we could generate, at least some exploratory mechanistic data to support that. And by taking FGFR3 mutant bladder cancer cells in knocking down FGFR3 and doing RNA-Seq on those cells, yes we also see that there is an increase in Interferon Gamma signaling compared to the cell lines that have a FGFR3 mutant intact.

Interestingly, what we also see is that EMT related gene signatures in the FGFR3 mutant cell lines are lower and of course this only represents the cancer cells. We don't have that the tumor microenvironment in these systems. And so it's a little bit of an artificial system and we're making some leaps here. But it does support the clinical data in that FGFR3 mutant tumors might have a lower T-cell infiltration, because of lower Interferon Gamma signaling. But they might also be less stromally suppressed because of lower EMT related gene expression.

And if you modulate that pharmacologically, it is possible that you'll do more of the good stuff, but more of the bad stuff too. And clearly, that needs to be sorted out in future trials, combining FGFR3 inhibitors with immune checkpoint blockade.

Petros Grivas: That's really fantastic, Matt. And very insightful commentary in the heart of this paper. So, you think that the combination studies combining FGF receptor inhibitors and checkpoint inhibitors might have some signal? And do you see any potential for any triplet down the road? We don't have any triplet yet, but I think about modulating the microenvironment, anti-PD-1, PD-L1, plus FGF receptor inhibitor? Is that something that you think is feasible? Assuming safety.

Matthew Galsky: So, I think with regards to the rationale for the combination question, the rationale for the combinations at least in part in the past have been based on the fact that these tumors respond less well and perhaps modulating them will convert "cold tumors to hot tumors", et cetera. I think the rationale for combinations is actually even more compelling if these treatments are non-cross resistant.

And so, we haven't done a great job historically at taking drugs that didn't work in a population and making them work by giving them in combination with something else. And so, I think the fact that both of these classes of drugs work in our clinically non-cross resistant it seems is actually a very positive aspect to the rationale for these combination studies. And so I think if anything, it supports these studies moving forward. With regard to triplets, much more mechanistic data is certainly needed, but one could certainly think about FGFR3 inhibition, TGF-beta inhibition, and PD-1 or PD-L1 inhibition.

Petros Grivas: Thank you, Matt. This is really interesting data. Since we're running out of time, let me go around and ask the three of you for some key take-home messages for the audience about... Based on the data we discussed today on these three very exciting papers, how do you envision the near future of urothelial cancer upper and lower tract? And any last-minute thoughts that you have. I'll start with actually Matt and we'll go to Bishoy and then we'll finish with Andrew.

Matthew Galsky: So, I think that the understanding of the biology of upper tract versus lower tract tumors is really helping to reinterpret some of the data that we've seen in the past with the standard chemotherapy in urothelial cancer. And I think we'll markedly shape how we think about doing studies in the future. And my interpretation of all of this data is that upper tract tumors are enriched in a biology that is shared with a subset of bladder cancers and not necessarily that there's a distinct biology.

But I do think that has implications in terms of clinical trial design. And I think that although we haven't done it historically, we need to start thinking about doing separate trials. If those are... could be feasible to do or we really have to think about molecular, we characterizing all the patients on these studies in a better way so that we stratify appropriately.

Petros Grivas: Thank you, Matt. Bishoy, what are your thoughts and take-home messages overall and the field going forward message?

Bishoy Faltas: Yeah. So, I think this is all very exciting. I think it is really, really, really interesting to really start understanding the biology of urothelial carcinoma in general, whether it's upper tract or bladder cancer. Because as all of you know, the revolution with the immune checkpoint inhibitor therapy was sort of fortuitous in a sense that the immune checkpoint inhibitors happened to work for upper tract urothelial carcinoma. But the biology of the disease itself has really been under... whether it's upper tract or lower tract, has been really understudied compared to other cancers for a very long time.

So, I think it's really amazing to see that we're starting to dig into the biology a little bit more. But it's really important to know that, because we're essentially trying to play catch up here is that we're just starting to scratch the surface of understanding the biology and the distinctions, for instance, between upper tract and lower tract, and whether they're parts of the continuum, or whether you know they're more different than similar, or whether they're more similar than different, and whether we can actually find treatments that can be targeted to upper tract that uses these differences to the patient's advantage to try to enrich for patients who are likely to respond to particular treatments and also to develop treatments that are targeted to particular subtypes, whether that's in bladder cancer or whether it's in upper tract urothelial carcinoma.

I would agree that we're also really starting to understand the interface between FGFR3 and Interferon Gamma signaling, for instance, which in of itself is a very interesting pathway with a lot of duality where it can really play a role in immune innovation, but also can play a role in immune surveillance, and have you know, pro-apoptotic, anti-apoptotic properties. So really, you're trying to understand what's the right balance to achieve with FGFR3 inhibitors and how can we really understand the biology so we can tip the balance to our advantage, to achieve deeper and longer-lasting responses in these patients.

Petros Grivas: Awesome. Andrew?

Andrew Hsieh: Yeah, I think one of the biggest take-home messages from our work has been this concept that when you look at a lesion within different metastatic examples within the same patient that quite a few of these, the majority of these are actually shared if these are actionable. Actionable mutations actually shared amongst all the tumors within a patient. And I think the immediate implication for this is that there could be some clinical utility of metastatic biopsies to help guide potential therapeutics, for instance, in the case of targeting FGFR3, as these inhibitors start coming online.

I think another big point though is that this data has to be taken with a grain of salt because of course 30% of biopsies will yield. Lesions are only present in one tumor, not shared amongst all of them. That would then suggest that we're not really getting a full picture.

But I'm very, very excited by new technologies that are being used in the bladder cancer field. For instance, cell-free DNA and how these may or may not be able to be reflective of what's going on within a patient's body in total. I think some of the work that you've done, as well as work by Alex Wyatt, points to the fact that CFDNA for targeted panels of genes may actually be very useful. And that's where I think the fact that we have rapid-autopsy programs like the one we have here or the one at Cornell, it should be able to help us answer these questions, because we can, at the same time sequence what's going on in the patient's entire body, and also compare that to what actually is elaborated in the context of CFDNA.

I'm very excited about that because that could have dramatic implications in terms of how we treat patients. And of course, all of this has to do with gaining a deeper understanding of the underlying biology of UTUC and LTUC. And without that, we're flying in the dark with these new very highly innovative tests like CF DNA or metastatic biopsy.

So, I'm very excited of what's going on on both sides, both clinically and of course, what's going on at the bench side or at the computer terminal, however, we think about it.

Petros Grivas: I fully agree with all of you guys and I cannot thank you enough for spending the time here with me today, covering three very important papers and also relevant literature in the field. I would like to thank the audience for staying with us all this time and thanks again all of you. Hopefully, I will have you again in the future in the Bladder Cancer Center of Excellence, and thanks to everybody for the attention.

Matthew Galsky: Thank you.

Bishoy Faltas: Thank you.

Andrew Hsieh: Thank you very much.