Sam Chang: Hello everyone. My name is Sam Chang. I'm a urologist in Nashville, Tennessee. And we are very fortunate today to have Dr. Kyle Rose. Kyle is a senior fellow in urological oncology at the Moffitt Cancer Center, and he is actually headed this late summer to join the Ochsner Medical Center, and will be one of their urologic attendings, and will be focused on urologic oncology.

We have him today to discuss actually, one of the late breaking abstracts at this year's GU ASCO conference in February of 2023. And he's going to actually describe their findings within their cancer center, looking at cell-free urinary DNA, and it's possible predictive impact on those patients with non-muscle invasive bladder cancer.

So, Dr. Rose, thank you so much for joining us. I've followed your young career for actually several years now, and looking forward to great, great things.

Kyle Rose: Thank you, sir. Thank you for having me. Thank you for the opportunity to present this work and sort of walk through our data, which we're very excited to share some of this pilot data at GU ASCO, and has really been a culmination of the last two years of work here at Moffitt, under the wing of Dr. Lee, and our collaboration with Predicine.

I think in the last five to seven years, ctDNA, or cell-free DNA, has really taken off, especially in urothelial carcinoma. There's been several landmark papers looking at the metastatic setting, muscle invasive bladder cancer, even using ctDNA to guide adjuvant immunotherapy. So this is a sort of the next frontier in a pioneering realm in the biomarker studies of bladder cancer.

But one area that's particularly interesting, is not just circulating tumor DNA that's found in the blood, but what is shed into the urine, which is cell-free urine DNA or utDNA. What's interesting about this is that, it can be shed directly from the cells into the urine, either from the upper tracts or from the bladder in less invasive cancer. So we're able to identify that.

And two, in advanced or invasive cancers, they can be circulated from the blood through the kidney, and then dumped into the urine. And so, for a process like non-muscle invasive bladder cancer, this turns out to be a very promising study for biomarkers and genomics.

This is a figure showing that, that the urine and the serum can both be leveraged in these patients with urothelial carcinoma. And it really does serve as a true liquid biopsy. Meaning that we can detect various types of mutations, or tumor genomic assays, such as tumor specific alterations, TMB, shifting clonality during treatment, or even just the total cell-free tumor DNA burden, and monitor that over time.

Where for us this really becomes exciting, and our clinical question that drove this pilot. And truly this is more, I would say, more of a feasibility study, before we get too into the woods on the clinical outcomes and the clinical implications. This is really a feasibility question for us, in that, patients with high risk non-invasive bladder cancer, that are undergoing a repeat TURBT as standard of care, so not changing any sort of clinical care parameters. Can we use the urine to precisely detect tumor alterations or mutations that are present, either in the index or the repeat TURBT sample, or both? And what are the implications if we can do so?

In order to do that, we performed a prospective study, in which we performed whole exome sequencing across 20,000 G genes in these high risk non-invasive bladder cancer patients. We used whole exome sequencing on the index and the repeat TURBT. And then for the urine itself, we use an ultra-deep sequencing through the PredicineBEACON platform. We considered the urinary tumor DNA to be positive. And keep in mind, that's the tumor prior to the repeat TURBT. Positive if we had two or more concordant variants that were seen on the study.

And so, the first aspect of the study is looking at the correlation, or concordance, between the index and the repeat TURBT tumor looking at the whole exome sequencing. In the 11 patients of the feasibility study, we found 73%, or eight of those 11 patients, to have tumor present in the repeat TURBT. And we found a very high concordance between the index and the repeat TURBT. We did have a higher median range of mutations identified in the index, versus slightly lower in the repeat TURBT. And we can see here on the oncoplot, that many of the most common genes that are detected were, in fact, detected in both. And obviously, this is focusing on the patients that had tumor on the repeat TURBT, not necessarily those patients that are T0.

And then again, further looking into that whole exome sequencing. Looking at the concordance rates, obviously these are lower numbers, but we found was that concordance rate was highest among patients with invasive and aggressive cancer, especially those that were upstaged from non-muscle invasive to muscle invasive cancer, and lowest in patients that are harboring CIS on their repeat TURBT specimen.

And then, looking at the urinary tumor DNA. And this was really the key question for us was, can we use the urinary tumor DNA as, not as a surrogate, but as a complimentary study, to see if it's feasible to represent the repeat TURBT specimen? And so, we found several of the most frequently found genes in non-muscle invasive bladder cancer in the repeat TURBT specimen, including TP53, PIK3CA, or RB1 and MYC.

Then looking at the utDNA tumor fraction. And for urinary tumor DNA, this was using the BEACON probe. It was calculating tumor fraction based on the frequencies of all the probes that were positively detected. Each probe mutation frequency was weighted by the frequency of the original index tissue sample. And what we found is that, patients who had disease present on the repeat TURBT had a higher utDNA tumor fraction. And then that final question of, can we use urinary tumor DNA as a surrogate? Or what is the accuracy, if we use the repeat TURBT as a gold standard? We can change the area under the curve, but at optimal sensitivity at 75%, the specificity of urinary tumor DNA in these high-risk non muscle-invasive bladder cancer patients approaches 100%. At target sensitivity you have 90%. The specificity obviously, decreases. Clinically, what this means is that, in these patients, there's nobody that we would want to miss, or want to forego, a repeat TURBT potentially. We would never want a false negative. We would much rather have a false positive in that setting as repeat TURBT is standard of care.

And so in conclusion, we found a high concordance rate of genomic alterations between the index and the repeat TURBT. We found that urinary tumor DNA is a promising biomarker, especially as a liquid biopsy, to detect the genomic alterations done on both the index and the repeat TURBT samples. I think the very big picture of this is not to overstate the clinical outcomes, and really outline this as a feasibility study, that has important implications in future studies, with larger cohorts, with longer term follow up data, to see if urinary tumor DNA can risk stratify these patients prior to repeat TURBT.

And then briefly, I always have to say thank you to our patients, physicians, mentors, collaborators, here at Moffitt Cancer Center, and with Predicine, who have just been phenomenal to work with, and have really been on the cutting edge, enabling us to do this important work.

Sam Chang: Kyle, that's fantastic. Let's start right there with Predicine, because I'm not familiar with Predicine. There are lots of different ways to examine, basically, cell-free DNA. They're commercial suppliers, they're private, they're lots of different ways. What is unique about the Predicine kind of evaluation process that you've done? And is there any specific change to this focus for urinary evaluation?

Kyle Rose: Well, being totally transparent, so a lot of the details is, the details of their work is proprietary. But from what we have discussed, I think you're absolutely right, that this is sort of an emerging field, a new frontier, and there's no true standardization of what we can call, how we call things urinary tumor DNA positive, or circulating tumor DNA positive. And everybody, not that everyone has their own rules, but right now, there are several avenues that we can explore.

What I really think set Predicine apart, and what really gave us the most interest, is that they were looking at cell-free DNA in the urine, and not every company out there was pursuing this, and they really had the resources and the science to back it up. And they just had a phenomenal team. They were incredibly responsive to work with. The other thing I think really was, actually in this study, is that this was not necessarily running a small panel gene expression study on specific tumors. This was looking across tens of thousands of genes from the primary tumor, and creating a bespoke panel, or a personalized panel, for these patients for their urine.

Sam Chang: So Kyle, next question. In evaluating this as kind of a promising technique, were there certain genetic changes perhaps in the urine? Let's just focus on the index lesion. Were there certain genetic changes that, or mutations, that weren't necessarily picked up as often? You would've hoped that you'd have pretty much 100% concordance, between your tumor genetic changes in your urinary sample. Were there certain ones that tended to be missed, or was it just a random? I know you can't tell much with small numbers, but did you notice any differences there, in terms of where you were off a bit with the urinary evaluation?

Kyle Rose: Well, fortunately for the urinary evaluation, there were a few patients that had positive urinary tumor DNA, that was detected in the urine prior to repeat TURBT. And even in one patient that was T0, that did not have tumor. And so there's several possibilities there, but the one that really stands out in our mind is, is there tumor that is visible? Even if it might be CIS, or a very small lesion that is being shed into the urine. It could possibly even be from the upper tract. That we're not seeing with conventional white light cystoscopy during the repeat TURBT.

Sam Chang: Right.

Kyle Rose: And I think, that is something that needs to be teased out. That's something that needs to be studied further under larger cohorts. And I think, longer follow up will ultimately answer that.

Sam Chang: So then, two follow up questions to that. Why do you think CIS, you mentioned this, the concordance, in terms of repeat TUR was really the highest, this is a separate question from the urinary, but basically your index lesion and your repeat TURBT, the concordance was quite high, in terms of mutational changes. But the least so with CIS. It was higher with a higher risk in muscle invasive. Why do you think that's the case?

Kyle Rose: I would be speculating, and I wouldn't have the data to necessarily back this up. But CIS being a multifocal tumor, that can appear in several areas of the bladder, especially for patients undergoing a repeat TURBT. In patients that come in without CIS, and they have a high grade T1 lesion, we know exactly the area that we're resecting. We can see it very clearly. We obviously, do a full cystoscopy, to look and make sure there's no other areas. But I think, in terms of the accuracy and the location of that tumor, for patients without CIS, it's much more consistent. And so, we know where we're going to be sampling from. And so, I think that there can certainly be a sporadic nature to the location of the tumors, and possibly even the clonality.

Sam Chang: Got it. So then two simple questions, because I'm familiar with simple. Have we looked at cell-free tumor DNA in those patients without tumors? So specifically, without bladder cancer tumors, to see what is found, if no really genetic change, that's number one. Number two, what does that also look like in a patient say, with a large kidney cancer? So combination of not only shed in the urine, but also directly shed in the urine. Any conjectures or thoughts, or any data regarding that?

Kyle Rose: Yeah, absolutely. And as a short aside, we recently just published our Nature Reviews paper on circulating tumor DNA and cell-free DNA in urothelial carcinoma, which should be coming out, which addresses a lot of the feasibility for patients, and where cell-free DNA originated from. Because as you mentioned, we're really specifying that this is cell-free tumor DNA. Which is why we don't necessarily use the broad category. But cell-free DNA, the majority of it is derived from our white blood cells in our circulation. And things like trauma or exercise can increase that sporadically. And so, there have certainly been historic studies looking at control patients with no known evidence of cancer. Now looking at the urine specifically, I'd have to go back and check. I don't know that there have been specific urine studies, but serum plasma, certainly. In fact, we have to use, if you saw on my first flow diagram or the outline of the study, we do have to collect a blood sample in order to do germline testing.

Sam Chang: And then evaluate. Right?

Kyle Rose: Right. Right. Exactly. Now, what's interesting though, is the question about renal cell carcinoma. We had Beth Green, a former Vanderbilt resident, she actually published on circulating tumor DNA in RCC versus UCC. And from what I understand, it's admittedly not my field of expertise, but renal cell carcinoma does not appear to be as high of a shedder into the urine, or into the serum, compared to bladder cancer, upper tract cancer. And so, there's certainly a lower propensity for detection of ctDNA in renal cell carcinoma.

Sam Chang: Yeah. It is truly, I mean, we are the evolution of not only the biomarkers that are out there, but the evaluation of which biomarkers are going to be predictive, and how they're actually going to be evaluated, I think is constantly changing and constantly evolving. And I think your work really starts the process of, okay, we've looked at metastatic disease, now we've looked at invasive disease, and now we can even consider it in non-muscle invasive disease, and with actually, a urinary marker. So I think this represents exciting, exciting work.

So where do we go next, Kyle? Once you make the move to New Orleans, where were we going to go next with this research?

Kyle Rose: So I think broadening and strengthening this data, what I would really love to see, and what we would love to create evidence for, is using the urinary tumor DNA as a surrogate for minimal residual disease. And I think that has important implications for both clinical practice and for clinical trials. Especially, as you know, in the non-invasive bladder cancer realm, where not all of the BCG unresponsive trials have a for cause biopsy, necessarily. I would love to see urinary tumor DNA as a method of monitoring response to therapeutics, especially as more come down the pipeline, in the BCG unresponsive. And that's just one example, but that's the one that I think of in my head. So I certainly see that as an area with promise.

Sam Chang: Well, incredibly exciting. And as I said at the beginning of this, we really look forward to your future contributions, because I know how much of a scientific leader you'll be, and you have been already in your young career, and we look forward to it. And any way we can support it, look forward to actually having you talk again to UroToday and our audience, regarding your next findings. So good luck with things, and thanks so much.

Kyle Rose: Thank you so much, Dr. Chang, for having me.