Read the Full Video Transcript

Ashish Kamat: Hello everyone. And welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat, Professor of Urologic Oncology and Cancer Research at MD Anderson Cancer Center in Houston. It is my pleasure to welcome today Dr. Bishoy Faltas, who is an Assistant Professor of Medicine, Cell and Developmental Biology, and the Director of Bladder Cancer Research at the Englander Institute for Precision Medicine, and the Gellert Family Research Scholar at NY Presbyterian Weill Cornell Medicine. Bishoy, you have a lot of titles, and I could spend the whole day introducing you, but essentially what I want to do is just introduce you as "the thinking man" that does genomics research in medicine as it pertains to urothelial carcinoma. With that, I will hand the stage over to you.

Bishoy Faltas: Thank you very much, Ashish. I'm very, very happy to be here today and honored to be introduced by you. I very much appreciate having that title of "the thinking man" because I think that actually really means a lot to me. So with that, hopefully, I will talk to you today about the genomics of upper tract urothelial carcinoma.

These are my disclosures. And this is an outline of my talk today. So I'm hoping that over the next few minutes, I will give you an overview of the main advances in our understanding of the biology of upper tract urothelial carcinoma or UTUC and its relationship with bladder urothelial carcinoma or BUC. I will describe the genomic and transcriptomic features of UTUC, its relationship with defective mismatch repair and the data, and the clonal relatedness between synchronous and metachronous UTUC and BUC. I will also show you how we are now leveraging single-cell technologies to understand UTUC biology.

First, a brief introduction to the clinical phenotypes of UTUC. We know that five to 10% of all urothelial carcinomas occur in the upper tract. We also know that UTUC has stronger epidemiologic associations with arsenic exposure, aristolochic acids, and other environmental exposures.

UTUC commonly presents with an advanced T-stage with muscle invasion in up to 60% of cases versus 20% to 30% of BUCs. UTUC patients can have more advanced T-stage or presentation, but actually similar outcomes to BUC after controlling for the stage.

The muscularis propria of the upper tract is thinner than in the bladder wall, which partially explains the advanced stage of presentation. However, from an embryological standpoint, the upper tract also arises from the ureteric bud, which is a mesodermal structure, as it ascends to meet the metanephric blastema. In contrast, we know that most of the bladder arises from an endodermal origin, supporting the notion that cancers arising from the bladder and the upper tract could be very different at a fundamental biological level.

So by applying the lens of genomics to UTUC, we have learned a lot recently about the biological features that set it apart from BUC. These are two of the first studies that look at some of these differences, both published in European Urology. The MSKCC paper showed a relatively low prevalence of TP53 mutations and a high prevalence of FGFR3 mutations and fusions. The MD Anderson paper showed similar findings with respect to the prevalence of TP53 and FGFR3 mutations in upper tract urothelial carcinoma. We published our experience here at Weill Cornell in 2019 in Nature Communications, and we used whole-exome sequencing of upper tract urothelial carcinoma tumors, compared them to the TCGA BUC cohorts, and our results also confirmed significant differences in the prevalence of TP53 mutations.

We then used RNA sequencing to try to understand how these genomic alterations in UTUC translate into transcriptomic and phenotypic differences. We started to understand where UTUC would fit within the spectrum of RNA expression subtypes, which were originally used to characterize BUC, that are illustrated on this slide. We discovered that UTUC is predominantly luminal using the BASE47 classifier. The TCGA classifier refined this observation further, showing a clear enrichment of the luminal papillary subtype in UTUC. This was a combined cohort of patients from Weill Cornell, Baylor, and MD Anderson, and we were glad to see that this finding was independently validated in a different cohort from Memorial Sloan Kettering by Dr. Jonathan Coleman and Dr. David Solit's groups. If you look at the bottom, you will see that the majority of UTUCs are luminal using BASE47 and luminal papillary using the consensus classifier.

This is some of our unpublished data that was recently presented at a USCAP meeting, where we looked at a separate UTUC cohort from Weill Cornell, including matched primary and metastatic samples. And again, we saw that most primary UTUCs are luminal papillary, as we discovered, and there are some early signals suggesting that this subtype distribution may be different in metastatic tumors.

Interestingly, at the level of the individual patient, we found that UTUC tumors could switch between molecular subtypes during disease progression. We even see that different metastases in the same patient could have different molecular subtypes. This finding has obviously important implications for the treatment of patients with metastatic UTUC.

To understand the immune contexture of UTUC tumors, we developed a 170 gene classifier comprising key immune genes, including CD8, and we found that 87% of UTUC tumors in our cohorts were T-cell depleted. This was associated with significantly higher PPARgamma and FGFR3 expression, suggesting that FGFR3 signaling potentially coordinates the luminal papillary and T-cell depleted subtypes.

To interrogate the functional link between FGFR3 signaling and the T-cell depleted phenotype that we observed in UTUC, we looked at RNA expression data from urothelial cancer cells with genetic FGFR3 knockdown. We discovered that this knockdown resulted in the upregulation of several interferon-gamma response genes, including BST-2 or tetherin, which is a viral restriction protein.

So we wanted to see if we could replicate this effect pharmacologically using a small molecule FGFR3 inhibitor, and a post-doc in my lab treated three UC cell lines harboring activating FGFR3 fusions with erdafitinib. And sure enough, we found a statistically significant increase in BST-2 and interferon response factor nine or IRF9. This suggests that FGFR3 inhibition is a potential therapeutic strategy for reversing this immune cell-depleted state. As you know, there are several ongoing trials with FGFR3 inhibitors that will test this hypothesis.

Now I'd like to switch gears a little bit to the topic of UTUC and mismatch repair deficiency. As we know, there is a well-known association between Lynch syndrome, which is caused by germline loss of function mutations in the canonical mismatch repair genes with UTUC. We know that Lynch syndrome patients have an increased risk, up to the 22-fold increased risk of developing UTUC over the general population. These patients tend to be younger, and it is important to recognize that these UTUC patients with Lynch syndromes will have a higher tumor mutational burden and microsatellite instability, which has therapeutic implications for treatment with immune checkpoint inhibitors. However, it is also important to note and to understand that the majority of UTUCs are sporadic, meaning they are not associated with Lynch syndrome.

So interestingly, even when we look at these sporadic UTUC patients, the mRNA and protein levels of the MLH1, PMS2, MSH2, and MSH6 are actually significantly lower compared to sporadic bladder cancers. However, I'd like to emphasize that this downregulation does not necessarily translate into an increase in microsatellite instability. As you can see here, we found that the tumor mutational burden was actually lower in our sporadic UTUC tumors compared to bladder cancers, and the MSI sensor scores were not statistically different. This likely occurs because a complete loss of function of one of these mismatch repair proteins is required to impart an impairment in DNA mismatch repairability and result in microsatellite instability.

Again, this was confirmed in the MSK cohort. So here you can see that the TMB of sporadic UTUCs that are shown in yellow is actually quite low. There was another study from Andrew Hsieh's group that confirmed lower TMB in metastatic sporadic UTUC cases compared to bladder cancer. Taken together, this data supports that the vast majority of sporadic UTUC cases are not hypermutated or microsatellite unstable.

We know that intravesical recurrence occurs in between 22% and 47% of UTUC patients. We also know that patients with primary UCB have an increased risk of future UTUC. So what can we learn about the clonal relationships between UTUC and BUC in these patients?

There was an interesting paper published in 2019, which asked the question, does BUC occur as a result of intraluminal seeding from prior UTUC, or is it a second primary in the setting of a toxin-induced field cancerization effect? What they show very nicely is that in patients who developed UTUC first followed by BUC, that there was a high degree of clonal relatedness at the molecular level between these tumors, and this essentially suggests that these bladder tumors were dropped metastases from the upper tract. Whereas in patients who present with BUC first and then go on to develop UTUC, the degree of clonal relatedness is reduced, suggesting that these tumors arise independently within a cancerization field defect.

In the last few minutes, I'd like to share with you some data from our latest project, which aims to understand UTUC at the single-cell resolution. This is a collaboration between my lab, Olivier Elemento, and Juan Miguel Mosquera at our institution. We used a technique called imaging mass cytometry or tissue cyTOF, which uses antibodies conjugated with lanthanide metals, allowing highly multiplexed imaging at a single-cell level. We applied this to our cohort of upper tract urothelial carcinomas, and as you can see here, this method allows us to map out these spatial relationships between various components of the tumor at single-cell resolution. For example, we can discern cancer, which is double-stained for keratin 5 and pan-keratin in the green and the blue channels, and vimentin positive fibroblasts in the red channel. We can then switch to different channels on the same image, and we can see the spatial distribution of different subsets of immune cells based on their respective surface markers. As you can see here, the CD3+ T-cells are infiltrating the tumor, but also some of the stromal areas.

So we interrogated 11 UTUC tumors, and we actually looked at the matched primary and metastatic tumors, and overall, we looked at more than 350,000 individual single-cells. As you can see here in this UMAT plot, we can then use the marker intensity to construct joint spaces that reflect the phenotypic similarity between the different cell types in the immune microenvironment. We are able to study the distribution of tumor cell markers, such as keratin 5, and gamma 3, immune markers, such as CD4, and CD8, and pathway-specific markers, such as PD-L1 in different tumors and microenvironment cell compartments.

Once we quantify this, we can begin to understand the interactions between the UTUC cells and other cell types in the microenvironment, such as endothelial cells, fibroblasts, and muscle cells. We also perform bulk RNA sequencing of the same upper tract tumors. As you can see here, we were able to validate that the RNA signature that we previously derived predicts T-cell depletion in UTUC concordant with protein expression that we see on the tissue cyTOF imaging.

So in summary, our understanding of UTUC biology is rapidly advancing. Most UTUCs are luminal papillary and have a T-cell depleted immune contexture. Sporadic UTUCs are microsatellite stable and do not necessarily have a high TMB. UCB recurrences after UTUC are clonal drop metastases, and high FGFR3 expression is enriched in UTUC and the T-cell depleted immune microenvironment, which might imply sensitivity to FGFR3 inhibition. And then finally, I would like to say that these single-cell and spatial transcriptomic and proteomic techniques are deepening our understanding of UTUC biology.

Thank you.

Ashish Kamat: Great. Thank you so much, Bishoy. That was a very succinct compression of all the months and years of work that you have done into one presentation, so I compliment you on that. If I might, let me ask you first sort of a broad question that people often kind of ask and are wrestling with. Based on all the work that you've done and your review of the literature, are they the same disease or are they different? What's your bottom line answer there? Bladder cancers and upper tract tumors. Are they the same? Different?

Bishoy Faltas: I've used different versions of that question. People talk about identical, fraternal, disparate twins. I think still, there is no easy way to answer that question. I think it depends on essentially why you are asking. It depends on what's the purpose of the question. So from a biological standpoint, there are definitely some important distinctions as we've seen. I would foresee that in the future, UTUC targeted treatment strategies that are informed by a deeper understanding of biology are on the horizon. On the other hand, we all know that UTUC is, from a practical standpoint, is a rare tumor. From a practical standpoint trying to come up with trials that focus only on the upper tract, urothelial carcinoma patients can sometimes be challenging because of a lack of statistical power. Also, if we're looking at different subtypes within UTUC, then that makes it even harder to do that.

So I think for the purpose of clinical trials, as long as we are aware of these differences and maybe stratify by upper tract versus bladder cancer, I don't think that it's wrong to consider them together. However, I think that as we understand more about the biology of the disease and then translate this knowledge to treatments that five, 10 years from now that we would have, again, UTUC targeted treatment strategies that may be significantly different from what we do for bladder cancer patients. Maybe it's the sequence of treatments or maybe it's dependent on a biomarker or so on.

Ashish Kamat: That's a very good point you made, and I'm glad you made that because the context in which people are asking that question is very relevant. Is it a fundamental research question? In which case, obviously, we can do deep dives like you are doing into the single-cell level even and try to figure out what's going on at the basic transcriptome, proteomic level, and immune-enrichment level.  But at the broader level when we are actually looking at the patient in front of us in the clinic, the question sometimes is a little bit different.

So with that in mind, let's switch gears a little bit and focus now on the clinician that is faced with a patient that has synchronous bladder cancer and an upper tract tumor. Based on, again, the work that you have done, where the upper tracts tend to be more luminal papillary, the bladders may not follow the same architecture, et cetera, and the fact that gene expression profiling and molecular profiling of tumors is not still available in many places and not really standard of care. What are some of the caveats that you would bring up to the clinician who is considering doing this to direct the treatment of his or her patient?

Bishoy Faltas: That's an excellent question. So just let me clarify the question. Are the caveats related to molecular testing?

Ashish Kamat: Not so much molecular testing, Bishoy, because I didn't want to go into that. I mean, I wanted to talk about, well, assuming that the clinician has tissue from the patient in front of him and can send it off and get the results back, which one would, in your opinion, be driving the care of the patient? The bladder molecular profile or the upper tract profile, and how would you go about making that differentiation?

Bishoy Faltas: Not an easy question, but thank you for that question, Ashish. In my mind, I think a lot of it sort of depends on the stage. Even if it's an advanced stage and a patient has a TMB or microsatellite unstable and more so a microsatellite unstable tumor, TMB high or microsatellite unstable. So a patient with a Lynch syndrome, for example, or even with a somatic loss of one of these DNA mismatch repair genes that end up having a high microsatellite sensor score or a high TMB that sometimes reflects microsatellite instability, those patients should be considered for immunotherapy perhaps earlier than in our typical sequencing of agents where we would go for chemotherapy first if the patients are eligible and then have immunotherapy as the second line.

So in those patients, I think it would be reasonable to perhaps move immunotherapy to the frontline setting, and there are ongoing concepts for trials that are trying to test that concept, as you know, systemically to see if we should have a different sequence of events or sequencing of treatment agents for these patients. So that's one important, I think in my mind today, that would be the most practical thing to look at.

If we go to sort of more clinically localized stages, I think that may change things a little bit because then we are talking about adjuvant or neoadjuvant settings. Combining bladder and upper tract in this context can become complicated. Adjuvant or neoadjuvant immunotherapy in patients with upper tract or bladder cancers with microsatellite instability may be a very attractive concept, and I think it remains to be tested.

Ashish Kamat: Right. But again, just to remind the audience, most sporadic UTUCs do not have MSI or high TMB, so that is something to keep in mind. You raised an important point about the upper tract tumors and the bladder cancer recurrences and the clonal drop metastases with the upper tract and the bladder in that sequence. Your thoughts on the fact that patients with bladder tumors are at a much higher risk of developing upper tract tumors and how that biology would play out.

Bishoy Faltas: That's a great question. So I think we have known for some time about this field cancerization, which presumably underlies the development of cancers from the upper tract and the lower tract that is occurring may be sequentially or at the same time without having a direct spread as in drop metastases. This cancerization could be environmental, and I'm sure that there is a component of that. So if there is a carcinogen that's from cigarette smoke, for example, that is excreted in urine and that urine comes in contact with the entirety of the urothelial tract or the entirety of the urothelium, then I think that can cause a widespread cancerization.

But some of it has probably, I'm starting to think has a lot to do with the genetic makeup of the patient. So as you know, in the past two years, we are starting to learn a lot more about the role of germline variants in bladder cancer patients, upper tract urothelial cancer patients, aside from the Lynch-related genes. And we are discovering that perhaps between 10% to 20% or even higher than that, depending on the definition of pathogenic variance and the genes that are included, up to 20% of urothelial cancer patients have germline variants in DNA repair genes. One could easily envision a situation where this field cancerization is occurring because of these germline variants resulting in defective DNA repair. It doesn't have to be in the mismatch repair pathway. It can be homologous recombination deficiency or other DNA repair pathways that over years results in field cancerization and the development of multiple cancers, including urothelial cancers at multiple locations.

Ashish Kamat: So Bishoy, again, you and I could chat about this forever, but in the interest of time, let me wrap this up. But I do want to give you kind of the final word. So briefly, some of your high-level closing thoughts for our audience.

Bishoy Faltas: I think we need to continue to understand the biology of this disease because even though we've made tremendous progress, I think we still have to learn more. I think we probably would know that we have learned enough once we get to the point where we can translate some of these scientific advances in our understanding to actionable treatment strategies for our patients.

Ashish Kamat: Great. Once again, Bishoy, I want to thank you for taking the time from your busy schedule to spend time with us. This is very, very important, and it kind of hurts me that we haven't been able to meet for quite some time. But until this is all done, stay safe and stay well.

Bishoy Faltas: Same to you, Ashish. Thank you so much. I hope that we will be able to meet in person very soon.

Ashish Kamat: Absolutely.