Genetics in Upper Tract Urothelial Cancer, What Urologists Should Know – Eugene J. Pietzak

July 22, 2021

Eugene Pietzak details the molecular landscape of upper tract urothelial carcinoma and highlights practical clinical implications of recent findings of the molecular biology of upper tract urothelial cancer, with a particular focus on what is relevant to the urologist in this conversation with Sam Chang.   Upper tract urothelial cancer comprises a very small proportion of all urothelial cancer. it has many similarities, but also has distinct clinical management and distinct molecular characteristics as well. Dr. Pietzak highlights several molecular profiling studies and recent genomic analyses which have identified a high incidence of potentially actionable genomic alterations. He shares an optimistic outlook for future novel precision therapy approaches in this disease space.


Eugene J. Pietzak, MD, Assistant Attending Surgeon, Memorial Sloan Kettering Cancer Center, Department of Surgery, Urologic Oncology Service, Assistant Professor, Weill Cornell Medicine, New York, NY

Sam S. Chang, MD, MBA, Patricia and Rodes Hart Endowed Chair of Urologic Surgery Professor Department of Urology at Vanderbilt University Medical Center, Nashville, TN

Read the Full Video Transcript

Sam Chang: Hello, everyone. I'm Sam Chang, urologist at Vanderbilt University in Nashville, Tennessee. And I am happy to actually have an opportunity to share some knowledge with you today. That presentation actually given by Dr. Eugene Pietzak. Dr. Pietzak is an Assistant Professor and Assistant Attending Surgeon at Memorial Sloan Kettering Cancer Center. And he's really led an effort from Memorial in terms of evaluation of urothelial carcinoma involving the upper tract. Some key findings there. As well as this differentiation or perhaps similarities with bladder cancer in patients as well. So without further ado, Eugene, thank you so much for your insight and look forward to this presentation.

Eugene Pietzak Thank you. And thank you so much for the invitation. I'm looking forward to talking afterwards as well. So here are my disclosures. None of them are relevant for this presentation.

So what I was tasked with is sort of focusing on what are the practical clinical implications of some of the findings we've recently discovered, our group and others have discovered, for upper tract urothelial cancer molecular biology. And really what is relevant to the urologist both currently, as well as in the near future.

And so, as we know, upper tract urothelial cancer comprises a very small proportion of all urothelial cancer. And it has many similarities, but just like the clinical management is distinct from bladder, the molecular characteristics are as well. So this is the outline for my talk and with no further ado, I'll first go into the genomic landscape of upper tract urothelial cancer. And we see similar genetic mutations to what we see with bladder cancer, but with varying frequencies between the two. And these differences in the frequency remain present even after adjusting for clinical and pathologic stages.

So there are some uniqueness to the two. And I just highlight this figure here that that demonstrates FGFR3 tends to be enriched in upper tract urothelial cancer, RB1, P53, ERBB2 enriched in bladder. And some of the potential relevance for this down the road may be an earlier paper led by Aditya Bagrodia when he was a fellow in the Solit Lab, looked at these characteristics, looking at classifying patients based off their FGFR3 status, as well as their P53 MDM2, which is related to the P53 pathway, and whether they were mutant or wild-type. And as can be seen here quite distinctly, is if you have a FGFR3 and wild type for P53-MDM2, the prognosis after NephU is rather good. However, if you're FGFR3 wild-type and have a P53 or MDM2 with NephU alone, without any adjuvant treatment, there is a substantially high risk of recurrence, as well as progression of disease.

This paper was just recently published and looked at a slightly more intricate risk stratification. But some of that relationship is still seen with FGFR3 mutated tumors doing better, where P53-MDM2 mutated tumors do worse. So at some point down the stream, these DNA based subtypes may be helpful in sort of guiding the selection for treatment.

And one of the hottest debated topics currently in upper tract, as we're all familiar with the disease, is the discussion of neoadjuvant versus adjuvant. And obviously this would be the focus of a talk all by itself, but this is a space that is quite exciting. And the molecular genomics may be able to help guide that one way or another.

And so this is just an anecdote from a clinical trial that was run by Jonathan Coleman and Dean Bajorin here of a split dose gemcitabine cisplatinum. This was a plenary talk that was given at the AUA 2019 when we were doing them in person. Looking forward to the one coming up relatively soon. And this was Nathan Wong gave the talk when he was a fellow in the Coleman lab. And this is a patient based off their pretreatment biopsy, had a ERCC2 hotspot mutation just like would be predicted to be sensitizing in bladder cancer for lower tract bladder cancer, received gemcis, and then had a complete pathological response. And as you can see, there's an obvious filling defects seen on this imaging study. So a little bit more provocative than anything else. And this is a area that's actively being investigated, but the same sort of predictors of response to chemotherapy that are being worked out in bladder cancer may also be relevant in upper tract urothelial cancer as well.

But of course, with upper tract, as we all know, who do these biopsies, do these procedures, that there are some tissue limitations. And again, this is a paper from Aditya Bagrodia and then the group at UT Southwestern, showing that there is concordance between the biopsies and the NephU specimens. However, getting and obtaining those tissues can be very problematic.

And again, referring to that previous paper that was published about two weeks ago, what they showed is the feasibility of looking at the urinary sediment from voided urine specimens. And it has excellent concordance between the urine specimen and the subsequent upper tract tumor tissue as well. So this may be an avenue. These liquid biopsies may become even more relevant and upper tract where it's difficult to get biopsies compared to the bladder itself. So this may be something coming down the pipeline relatively soon.

And of course, FGFR3 is enriched in upper tract as compared to bladder cancer. And this is a potentially actionable alteration. There are some FGFR inhibitors that are clinically available. This is a subgroup analysis that is suggestive that upper tract may be more sensitive than bladder, although only eight upper tract patients were analyzed in this group. So it's more of hypothesis generating than anything else. However, both neoadjuvant, as well as adjuvant, clinical trials of FGFR inhibitors are underway, including one that's being run by [inaudible 00:06:22] and the SUO CTC as an adjuvant trial. And there's a new adjuvant trial of FGFR inhibitor that's run out of MD Anderson. So we may have more information relatively soon on where these potentially actionable treatments fall into this spectrum.

And so shifting gears and focusing on Lynch Syndrome. As many of us have been taught over the years that upper tract urothelial cancer is one of the Lynch Syndrome associated cancers that are sort of less appreciated than colon cancer and some of the others, but increasingly recognized and increasingly important that urologist are aware of this association and are looking and thinking about this association. And so the interest clinically for this, besides the obvious implications for screening from other cancers, looking for colonoscopies, family members, et cetera, is also that immunotherapy, maybe very sensitive, may be very active, in patients with mismatch repair protein deficiency that's seen in Lynch syndrome. And these are some papers looking at various checkpoint inhibitors for these patients.

And again, a bit of an anecdote, but this is one of my patients that I shared with Dr. Min Yuen Teo, one of the medical oncologists here at MSK that I'm fortunate to work with, is a patient that actually during the pandemic that we saw who had this very large, mid-ureteral mass with known Lynch syndrome. And given the situation of the pandemic and the OR availability, et cetera, we certainly typically treat with neoadjuvant chemotherapy as our institutional protocol here with gemcis for those eligible. But because of the nuance of this and her Lynch syndrome and her desire to avoid chemotherapy, she was actually treated with pembrolizumab neoadjuvantly.

And then when we took her for her nephroureterectomy, there was still residual mass there that was actually active on the PET scan. But the subsequent NephU specimen was just basically chronic inflammation and immune infiltrates. And there was no residual disease, from what was previously very bulky, high grade urothelial cancer. So again, just shown to be a little bit more provocative than anything else.

However, it's important to recognize, although upper tract urothelial cancer is Lynch associated, most cases are sporadic. And as can be seen here, while the Lynch patients have this hyper mutated phenotype, the sporadic cases are generally lower, have a lower mutational burden, and may not do as well with immune checkpoint inhibitor. Although, there's obviously other factors than tumor mutational burden.

But two groups, Dr. Bishoy Faltas over at Cornell, as well as the group at a University of Washington on autopsy study basically demonstrated pretty consistently that upper tract urothelial cancer tends to have a lower tumor mutational burden than is seen in lower tract bladder cancer. And this may potentially explain or contribute to the differences that we're seeing in this subgroup analysis. Again, this is hypothesis generating than anything else, but we're all very excited about the recent presentation at ASCO and the publication of adjuvant nivolumab as potentially practice changing. But in the subgroup analysis, the benefit from nivolumab was predominantly seen in patients with bladder cancer rather than upper tract urothelial cancers. Again, hypothesis generating, but TMB tumor mutational burden, may be partially explaining this difference.

But the important thing to take home from this is that germline testing should be considered for all patients with upper tract urothelial cancer. It's generally seen in about 10 to 20% of upper tract urothelial cancers. The NCCN guidelines mentioned it for patients younger than 60 or patients with a family history of colon or endometrial cancer. Our general practice here is to offer tumor and normal sequencing for germline testing and tumor testing for all patients with upper tract urothelial cancer. And I think that's going to be increasingly recognized as being important.

Finally, looking at the implications of what are called dropdown metastasis. We know patients with upper tract urothelial cancer, they have a substantially high risk of developing subsequent bladder recurrences. Whether the patients get ureteroscopy or not, that risk is present. However, ureteroscopy certainly increases that risk. When we look at the clonality, how similar those tumors are for individuals that have primary upper tract urothelial cancer, and then develop bladder recurrences, they are highly clonally related, suggested that they are essentially drop down metastases, cells seeding from the upper tracks down to the bladder tracts, down to the bladder itself.

And so we do know that there are some randomized trials that support the use of intravesical chemotherapy after NephU, given at various different chemotherapies, given at various different time points. We also know that the utilization of these approaches are relatively low.

There are several groups looking at the use of interoperative intravesical chemotherapy. So at the time of nephroureterectomy. And so at least on this retrospective review by the UT Southwestern group, and others, suggest that it may be as effective as postoperative intravesical chemotherapy. And there is an ongoing Phase 2 trial that's led by Vig Packiam that I just wanted to highlight as well, which is currently ongoing.

So the clinical implications for this as a urologist, I think, the idea of at the time of NephU, clipping the ureter low as low down as possible with minimal manipulation of the kidney at first, to prevent some of these drop down metastases, is important. I know that there are some individuals that are concerned with clipping and then causing subsequent hydronephrosis. I tell you that in our experience here, at least, that is not a problem. Generally we try to clip the ureter before manipulating the high limits [inaudible 00:12:44] the first move we typically do after mobilizing the colon and exposing it.

Obviously should go without saying at this point in 2021, that we should always take the bladder cuff. And then the role of peri-operative intravesical chemotherapy has been demonstrated in randomized trials in interoperative intravesical chemotherapy during NephU may be as effective or potentially be more effective and improve utilization.

And then this is, again, just more to be provocative than anything else, but what about at the time of your ureteroscopy? And so, just like we know at the time of TURBT, giving an installation of perioperative chemotherapy reduces that risk of recurrence, predominantly in low grade tumors. But many of these drop down metastases tend to be FGFR3 papillary type tumors. But this may be an interesting future cooperative group study to potentially explore, if we could sort of get rid of these drop down metastases upstream even before NephU, or for patients that are undergoing nephron sparing procedures.

And then the last point, and just very briefly, is just like there's establishment of molecular subtypes in bladder cancer, those are also increasingly recognized and upper tract urothelial cancer. The difference is that although the subtypes tend to be similar, upper tract urothelial cancer tends to be more luminal, or luminal papillary.

And again, this is demonstrated by a Dr. Faltas' group that they tend to be more in the T-cell depleted group, which may have implications for immunotherapy. There tends to be a decrease in some of the interferon gamma signaling pathways for tumors that are T-cell depleted that has an up regulation of FGFR3 gene expression in a re-analysis of the [inaudible 00:14:29] data, as well as the checkmate 275, although FGFR3 mutations were not associated with a decreased response to checkpoint inhibitor, so those patients shouldn't be excluded from that treatment. They also saw a similar difference in some of the gene expression patterns as well for these immune signatures. So something to keep in mind as we proceed on.

And then just going back to that more recently published a paper that is quite interesting, and almost like a TCGA like type of paper as described by Dr. McConkey and the Editorial, they well-correlated some of the DNA based subtypes to some of the RNA based subtypes, the expression of molecular subtypes as well. So a lot to sort of dig through in that and a good future direction.

And if anyone's actually interested in this topic any further, I would encourage you to check out this recent paper written by Peter Reese, who's one of the clinical fellows now here at MSK, who wrote this paper sort of summarizing some of the clinical implications. And I believe that you're familiar with them, Dr. Chang. He tells me you taught him everything he knows-

Sam Chang: And now he's learning everything that's correct. That's perfect. Now let's just look at, have a discussion, here real quick, Gene. Because that was a great presentation, as you unshare your sides. As you were talking about things, a couple of things just came to mind regarding next steps, as well as, what should we be doing now? So let's talk about something you didn't talk about at all, which is there is some concern and some studies that say that people with invasive bladder cancer or with hydronephrosis with bladder cancer, those that have stents are more likely to develop upper tract recurrences down the line. And there's some controversy, et cetera, et cetera. Do you think number one, are you aware of any, I'm not, aware of any research showing, are they genetically similar, these bladder tumors that go up into the upper tract? I don't know if there's any studies. But do you think there's some seeding then going in the reverse fashion with these stents in place?

Eugene Pietzak I think that's a great question. And I do agree. I would certainly be concerned. I mean, I'm concerned with that. If there's tumor down in the bladder, pushing it back up. We know that the reverse studies, looking at bladder and then looking at development of upper tract recurrences, tend to be less clonally related. But I believe you're correct. I've not seen any data looking whether or not, whether a patient's had a stent or PCN made any difference one way or another, but yeah.

Sam Chang: Okay. So write that down. That's your next project. Get Peter Reese on that. Because you guys and Anderson and people who have large [inaudible 00:17:30], I don't think that would be too difficult to look at, especially since you know there's going to be 20% of those that have hydronephrosis, maybe 10 or 15% of your cystectomies have a stent in place. So I think that's just a great thought as you think about things.

While you are considering that kind of clonal relationship between your upper tract and your lower tract recurrence, it seems to me, at least hypothetically, that the peri-operative use of chemotherapy, which is now currently being studied [inaudible 00:18:07], as you pointed out, Dr. Packiam's study from looking at Gemini and gemcitabine peri-operative, what do you do? Because I've asked this a few times and people give very different answers.

I'll be honest. I actually give the gemcitabine as they go to sleep, as the catheter goes in. And I leave it in for an hour or two. And by that time, because I'm slow, I'm close to the kidney. But no, I leave it in while I've... Basically almost immediately. And we do clip the ureters relatively soon. And honestly, I tend to clip it wherever I can get it easily. Especially for a renal pelvis tumor, I'm not as diligent about going as distally as possible, but it makes sense. What do you do in terms of peri-operative or peri-surgical treatment in terms of intravesical chemotherapy? When do you give it and how do you give it?

Eugene Pietzak Yeah, so my own practice patterns, I actually have been giving it at the time of ureteroscopy as well. So that's, again, there's no data to support that, but just based off some of the molecular characteristics, it's easy enough to do. And oftentimes these patients will have bladder tumors as well. Those patients were excluded for those intravesical randomized trials, but oftentimes you take a look in, and there's a small little papillary tumor that you're [inaudible 00:19:29] at the same time, treating a known upper tract urothelial cancer. So my personal practice patterns have been to also give it a time of ureteroscopy.

I've attempted multiple different ways to give it at the time of NephU. I generally, to me, at least, it makes more sense to give it as close to the time of surgery as possible. I have instilled it before the case. I have instilled it after the case, as well, after we confirm that the bladder cuff is sufficiently closed. That, to me, makes more sense than giving it at the time of the catheter removal, at least, because we know-

Sam Chang: I agree. I agree. I agree. I mean, if we think theoretically that this is what's going on, it just seems to make the most sense. But the data doesn't, like you said, doesn't necessarily support that. But I think that's a great point.

Along those lines, in terms of seeding, say you have someone that you want to further evaluate their upper tracts, but their upper tracts look fine on imaging. But you, for some reason, perhaps there's a question of a cytology, different things. And that patient has a known bladder tumor. I tend to actually evaluate the upper tract first before I disrupt the bladder tumor. Does that make sense? Is that something you normally do as well?

Eugene Pietzak Yeah, I think we're sort of in a data free zone there, in terms of that practical management. I think a lot of it depends on what the tumor is or what it looks like. Just like your practice patterns as well, oftentimes you're going in for restaging 2RBT and tumor's ideally debulked. Maybe there's a small amount of residual tumor. Those patients, I would typically evaluate their upper tracts at the same time, take a look up and then complete it. If there's residual tumor there, we'll clean it up a little bit and then throw up the scope.

If someone has a significant multifocal bladder tumors and there's something that's concerning on the upper tracts, I may probably, more oftentimes than not, favor treating the bladder first or going up and taking a look. [inaudible 00:21:44] it looks like it's low grade and we want to sort of minimize the risk of anesthesia, et cetera, then I'd potentially consider treating at the same time.

But my tendency has actually been to resect the bladder tumors, flush out the bladder a bunch of times, make sure that it's as well-irrigated as possible. There's not any debris floating around. Almost to OCD type of extent, cleaning it up.

Sam Chang: Right. Right.

Eugene Pietzak Or, then instrumenting the upper tract because-

Sam Chang: Actually in thinking about things, that actually makes a lot of sense. You basically are, as best you can, making your bladder as sterile as you can from a tumor cell standpoint before manipulating it. I think that makes a lot of sense.

Last question then for the urologist, because you gave just a wonderful outline of, look at these genetic changes that we seem to be associated with benefit from therapy. Your pembrolizumab for your upper tract with mismatch repair, with MSH2 presentation and the FGF3 mutation patient, et cetera.

When are we, and places at Memorial, you're very unique in terms of being able to get this kind of tissue mutation evaluation, get germline tasting. It's something that can be done relatively easily. And you guys do it well and do it on a regular basis. For someone practicing now, what should they be doing when they get a biopsy of an upper tract tumor? Should they be trying to work with insurance companies and getting basically DNA testing of their tissue themselves, pushing their pathologist to help with that? Or should they at this point still just depend upon clinical grade stage, clinical variables? What do you think? Should urologists start pushing that genetic evaluation envelope?

Eugene Pietzak Yeah, I mean, I think the future is definitely going to be towards next generation sequencing with these large panels that allows both the tumor, as well as germline testing. I think right now it's mostly for research purposes, with the caveat that if these patients aren't getting germline testing routinely, then doing simple IHC, immunohistochemical staining, for some of the MMR proteins, I think that that should be something that should be highly considered for any of the upper tract tumors. And that's a patient that if they have loss of one of those protein expression, then that's a patient that definitely needs germline counseling and testing.

Sam Chang: Yeah. The blinker should go off. Yeah, last follow-up. I lied. I said I was going to have one more. [inaudible 00:24:36] that was my last question. But it's not. This is the last question. Put your kind of your predicting hat on. I mean, you probably predicted that the Knicks would beat the Hawks and that was not even close. But put on your thinking hat, and come up with, when will we really, for upper tract disease, be looking at individualizing and personalizing care based upon their genetic or tumor [inaudible 00:25:11] findings. Are we two years away? Are we 20 years away? Five years away? Just give me an idea.

Eugene Pietzak I think. Great question. Obviously, my crystal ball is always a little bit cloudy. But I would imagine probably within the next five to 10 years. And it's such an area where there's such a problem with clinical under staging and issues, uncertainty about what the appropriate approach is, neoadjuvant treatment, adjuvant treatment, et cetera, that I think the addition of some genomic biomarkers would be of such incredible value in this disease space that I hope within the next five to 10 years, we're at least integrating some of these markers within the standard clinical and pathological features that we're already sort of dependent on, and unfortunately are somewhat unreal.

Sam Chang: Well, thank you so much, Gene. I'll tell everyone that's viewing this: some of those key findings and some of that leadership will be led by Dr. Pietzak. Really, really appreciate his time with us today. And really, I think as much as anything, I know all our patients appreciate the inroads that you have all made at Memorial Sloan Kettering in terms of, what are the next steps and what can we do to further individualize and improve patient care? So thanks again, Gene, as always. Very much appreciate it.

Eugene Pietzak Thank you very much for the invitation. Thank you [inaudible 00:26:37] today. And of course, I just want to also thank my mentors in this disease space, Jonathan Coleman who's a major leader in this field, as well as David Solit as well. We're sort of the foundation of everything that we're doing here at MSK. And I agree, this is a great space for collaboration in the future as well. So looking forward to doing all that.

Sam Chang: Great. Great. Thank you.