Read the Full Video Transcript

Sam Chang: Hi, my name is Sam Chang, a urologist in Nashville, Tennessee, at Vanderbilt University Medical Center. And I have a lot of wonderful colleagues that I get the pleasure to work with and learn from every day. And one of the true highlights that I have is actually working with Dr. Alan Tan.

Alan's an Associate Professor here in the Hematology-Oncology Division and really leads our bladder cancer initiatives within our medical oncology department. And we're quite fortunate to have him. He's going to highlight some work that was presented at ASCO GU 2025, looking at molecular profiling in the comprehensive evaluation of bladder cancer, not only in treatment but also in surveillance.

And so I wanted to welcome Alan and thank him for, first of all, keeping me out of trouble almost every day, but secondly, for giving us some highlights of what he presented at the 2025 meeting. So, Alan, we'll turn it over to you.

Alan Tan: Yeah, thank you so much, Sam. Always a pleasure to join with you. And good to see you at San Francisco, too.

So as Sam mentioned, we're going to be talking about muscle-invasive bladder cancer (MIBC). And this is a cancer that's near and dear to my heart. I think the future is bright in MIBC. And we're going to be talking about the role of circulating tumor DNA (ctDNA), along with urinary tract DNA (utDNA), in MIBC.

So just a refresher. I know this is a hot topic across the field, but just a refresher that circulating tumor DNA has a very short half-life. It's less than two hours or 2.5 hours. So it's really a good real-time snapshot of what's actually going on inside a patient's system. And it's actually cleared by the urinary tract, too. So this makes it really nice to actually measure utDNA as well, especially when we're looking at surveilling patients with urinary tract cancer.

So ctDNA is very prognostic at different time points. This is a highlight of the Christensen paper back in 2019, which really told us that patients that have ctDNA positivity at baseline have a worse prognosis. And especially if they're going to get radical cystectomy and they get neoadjuvant chemotherapy ahead of time, which is the standard of care, by the way,

If you have ctDNA positivity at that time, you actually have a worse outcome. And the chances of relapse systemically are very high. And also, if you do surveillance on these patients and you have persistent ctDNA positivity, these patients are associated with a worse outcome.

So another thing I want to also mention is ctDNA kinetics is very predictive as well. This is a patient that we share here at Vanderbilt that underwent neoadjuvant chemotherapy. Couldn't tolerate gemcitabine-cisplatin, so we had switched him over to
enfortumab vedotin-pembrolizumab. This patient had a little bit of trouble with skin reactions, a little bit of autoimmune hepatitis or liver inflammation, and actually had to go on some intermittent prednisone.

So this may have affected or impacted this patient's outcome. But as you can see here, ctDNA tumor fraction is quite high, almost 1,000 parts per million at baseline. And as she goes under neoadjuvant chemotherapy, she has some significant decrease, all the way down to 23 parts per million. But pay attention to what happens here.

We repeated her ctDNA right before her surgery, and it actually has an upward inflection or persistence of ctDNA. So really consistent with what Christensen showed—that ctDNA positivity is a bad actor. And when we did the radical cystectomy, the patient had lymph node positivity and also ovarian spread.

On the contrary, here's a patient that actually has ctDNA clearance after one cycle of EV-pembro. This patient has metastatic disease to the lymph nodes and has EV-pembro. And we all know from EV 302, really high response rates, 30% complete response rate systemically. So you can imagine in the perioperative setting, you might see something very similar.

So he got four cycles of EV-pembro. And I just spoke with his surgeon. Surgery went well. Did not have any clinical evidence of disease. And we were waiting for pathology to see if there's a complete pathological response.

So can ctDNA help inform bladder preservation? So we see these new regimens, EV-pembro, EV-Durva/Treme. Enfortumab vedotin is a strong antibody-drug conjugate that we're going to probably see higher pCR rates, maybe probably high CCR rates (complete clinical response rates). And I think this is going to give us pause.

Our patients are going to want to preserve their bladders more. Are they going to refuse radical cystectomy? I think we're going to see what comes of that in EV 304 when that results in.

But we have a sneak preview that a similar type of ADC that targets HER2, disitamab vedotin, was presented at GU ASCO recently. And we're seeing 64% pCR rates. This is really unprecedented. Just for comparison, with chemo alone or even in the NIAGARA trial, we had 37% pCR rates.

So I think we're going to find that bladder preservation can happen just with systemic treatment alone and maybe even with intravesical treatment. Maybe we don't need radiation anymore. So here's an example of a patient that failed BCG, had muscle-invasive bladder cancer, and was referred for trimodality therapy. But then we did a PET scan, and he had a significant amount of lymph nodes involved.

So I recommended using EV plus pembrolizumab (EV + pembro). He had a really high ctDNA fraction of over 3,000 at baseline. And after only one cycle of EV + pembro, we cleared ctDNA, and we have sustained clearance. He's 80 years old, so we don't want to overtreat these patients.

As you can see right now, even with four cycles of EV + pembro, we have some asymptomatic findings in the lung. Even though he had a complete clinical response, we have pneumonitis that's incidental. So I think further treatment may be futile in that patient.

So if this patient has a complete clinical response, can we ensure that this is sustained? I don't think the field really understands what's happening right now. Do we need to consolidate these patients? So you can imagine something intravesical being entertained there—maybe chemoradiation. But further systemic treatment probably is not the answer, especially in somebody that has significant toxicity.

So we had some experience partnering with a company called Predicine, which looks at a lot of ctDNA and utDNA offerings. And so here's a study that they presented at GU ASCO 2022, where they looked at toripalimab, which is a PD-1 inhibitor, in the neoadjuvant setting.

As you can see, the patients that had positive utDNA and had a decrease in their utDNA actually showed responses within the bladder. In fact, it actually showed that it was associated with pathologic responses. And MRD was identified in up to 60% of these patients.

So in ESMO 2023, we actually presented a poster that looked at a real-world cohort of 35 patients. And here's an example of a patient matched with ctDNA. You can see that ctDNA and utDNA correlate really nicely. Essentially, the ctDNA in the plasma cleared, but the utDNA was persistently positive.

And what did that actually mean? When he had a radical cystectomy, he actually still had disease in his bladder. So that kind of informed us that we didn't have a complete clinical response, even without going in and looking with a cystoscopy.

Furthermore, we can actually see genomic alterations within the bladder using urine, right? The PredicineBEACON assay doesn't actually require the submission of tissue. You can actually get, with 40 mL of urine, enough DNA to genomically profile a patient. And as you can see, you can get mutations that are actually relevant in bladder cancer, such as FGFR3 and ERBB2, which translates to HER2, as well as BRCA mutations, right? We all know that platinum sensitivity is increased with the expression of BRCA.

And so I think this might help inform the future role of utDNA and ctDNA together as a complementary way of surveying patients and deciding who's actually going to be able to keep their bladders and who's probably better off with a radical cystectomy. We're far away from getting there, but I think this is the first step.

And so that brings me to the poster that we just presented at GU ASCO in San Francisco a couple of weeks ago. We took those 36 patients from our first cohort, and we were able to do what's called whole-genome DNA methylation. EPIC is a whole-genome methylation profiling method, which really doesn't require tissue.

The turnaround time for something like this is 7 to 10 days, and I think that's really important when we're talking about something where we need answers right away—such as upper tract urothelial cancer, where it's really tough to get tissue and tough to get the diagnosis sometimes clinically.

And then we also added a whole-transcriptome profile, looking at RNA as well. And we think that this might actually add to the picture and enrich this population for biomarkers to really understand what's happening biologically but also to inform who truly has residual disease.

And so we looked at the kinetics of three patients using this Predicine complete picture, and we saw that it tracked very nicely. All patients who had a treatment response actually had a decrease in their tumor fraction.

And so I'll end with this: ctDNA and utDNA—at least ctDNA—are currently being used by many providers. But I think the combination of this and utDNA can actually help inform us about who can actually do bladder preservation with systemic therapy, intravesical therapy, chemoradiation, etc. So I propose that, in the future, we use biomarkers, especially for patients who don't have a complete clinical response after four cycles of neoadjuvant chemotherapy.

Can we use an FGFR mutation that we find in the urine to guide treatment with TAR-210? Can we use HER2 mutation profiling from the bladder to determine whether someone should receive HER2-targeted therapy? So with that, I'll stop there and see if there are any questions.

Sam Chang: Alan, that was great. I think it really gives us an idea of where we're headed. I guess the toughest questions for an audience just learning about this will be that there are probably complex answers to simple questions. So, simple question one—and I've asked you this before—how about other cancers, other solid tumors, breast cancer, colorectal cancer, lung cancer? Are we way behind in bladder cancer, or is everyone still in this exploratory phase?

Alan Tan: Yeah, I think—

Sam Chang: Can you tell me where we're at?

Alan Tan: So, colorectal cancer is the solid tumor that kind of pioneered this. Actually, I don’t think that’s even fair. MRD testing actually came from hematologic cancers. When I used to treat hematologic cancers, like multiple myeloma or chronic myelogenous leukemia, we were talking about curing cancers by clearing ctDNA and achieving a complete molecular response.

So to see that actually translate into solid tumors is really exciting because we're overtreating a lot of people in the adjuvant setting just to save a fraction of them. And in colorectal cancer, they’re already using this as standard of care for stage II colon cancer. Usually, you don’t give chemo, but maybe 15% of these patients are MRD-positive, and they are destined to recur with metastatic disease. So maybe those patients do get adjuvant FOLFOX, and we can spare them from developing metastatic disease.

Sam Chang: So for urologists that treat bladder cancer or treat kidney cancer, let's say we just want some preliminary type of data for those who are listening and watching. If I have a muscle-invasive bladder cancer that has not metastasized—not that we clinically evaluate, but in a patient that has muscle-invasive bladder cancer that is not metastasized—how often is there detectable circulating tumor DNA in the serum?

Is it half the time? Is it two-thirds of the time? Is it basically never? Give me an idea about that so that people have an idea of, all right, where is this going to be useful? And then along the same lines, how about noninvasive cancers? Do any of them have circulating tumor DNA that is, in fact, just non-muscle invasive? Or do we know yet?

Alan Tan: No, that's a great question. And I think we do know preliminarily, and I have my own biases, perhaps. But I think if you look at the IMvigor010 study, which was an adjuvant study, the detection rate wasn't 100% in patients who had recurrences. And they only did one time point at baseline and another time point at cycle three, day one.

So the sensitivity increases when you keep tracking this over time. And we might be redefining adjuvant therapy eventually because I don't think it really matters when you start that adjuvant treatment, as long as you detect it when the molecular trace is there. We're seeing from other studies, like Tombolo, that you can actually clear ctDNA, and it will be associated with good outcomes too.

But to answer your question, how often are we seeing this? I think with the newer tests, they're actually mimicking the real world of what you see in recurrence. So maybe half the time, eventually, you'll have positivity because we know patients who have stage II muscle-invasive bladder cancer—it sounds like it's not that bad. Stage II is not as bad as Stage III and IV.

But I think it's actually misrepresenting the severity of this cancer to our patients because stage II colorectal cancer is probably not that bad, right? I agree. But stage II muscle-invasive bladder cancer is pretty bad. The recurrence rate is about 50% to 60%, depending on which data sets you look at. But they do poorly whether they have radical cystectomy or bladder preservation. But I bet that's going to be improving now that we have really good antibody-drug conjugate combinations.

Sam Chang: And then with noninvasive bladder cancer, the vast majority should not have circulating tumor DNA. And if they do—well, that is a trigger for us as we struggle with deciding who we should be aggressive with for upfront cystectomy. And I mean, you know, our tendency has been, OK, we treat. We really don't give neoadjuvant chemotherapy.

But in those patients that we are worried about but haven't proven to have muscle-invasive disease, if they had measurable circulating tumor DNA, that's exactly the patient I'd want to know about—to escalate care prior to actually doing some type of localized therapy. So I think that's—I mean, just as you said, the ability to de-escalate and escalate is going to become increasingly important as we try to personalize care for each of these patients.

Tell me about the urinary tumor DNA (utDNA). I am struggling to try to figure out the best way to determine this because we know that within non-muscle invasive disease, tumors can be quite heterogeneous within the bladder. And so I see the advantages of getting multiple biopsies, having different tissue samples, and looking at them.

But I love the off-the-shelf idea—just to be able to have something relatively quickly from the urine itself as opposed to getting tissue, etc. Tell me where you think the landscape is going to go with urinary tumor DNA, how we're going to use it, and where we're going to use it?

Alan Tan: Yeah, that's a great question, Sam. So across genomic testing, we always say that tissue is the issue. Tissue is always superior in terms of having a tumor-informed test. And the downside of tumor-informed testing is that it takes a long time. You have to do whole-genome sequencing, which takes about four weeks at least.

And frankly, if we're trying to make major decisions, we probably can't wait that long all the time. But with utDNA, the fact that you're just using a urine sample—since we all have to urinate—it's noninvasive. It's actually capturing the heterogeneity of that tumor. We know that if we biopsy a sample in a few areas, we're only getting that snapshot of where we biopsy. Same with Gleason 7 prostate cancer—then you do a radical prostatectomy, and oh, there's some Gleason 9 in there too.

So I think the future is doing both, right? If you want the full picture, that's why we developed PredicineComplete—because we're adding more pieces to the complete puzzle. And you might say, well, that's going to be very expensive, right? So I think eventually, we're going to use both in the beginning. And we're going to get our initial answer really quickly from something like methylation.

Seven to ten days later, you're going to get your answer. If this is upper tract cancer, let's start treating. And then later on, once we actually know what we get from tumor-informed and non-tumor-informed testing in parallel, we'll run these together, and we'll probably pick the best of both, right? And maybe it is the tumor-informed one.

But the nice thing about non-tumor-informed testing is that you can actually capture the evolution of the cancer. Cancers can resist our treatments. They can transform into something else. And so I think doing both at different time points is probably the best complete answer. But ultimately, we don’t know yet. And that's why we're doing these studies.

Sam Chang: Yeah, Alan, I just wanted to emphasize to everyone—we're in this age of really an explosion of therapies for not only advanced disease but also for non-muscle invasive disease. It’s not too dissimilar from the evolution of treatments for advanced prostate cancer or advanced kidney cancer.

We’ve really opened up the field to study and expand options for patients. But the ability that you and your team have, in terms of being able to determine the timing of that therapy and the appropriateness of the length of therapy, really, really will allow for great strides forward—not only in patient recovery and beating their tumors but also in quality of life, cost, and the overall burden of therapy.

So I'm incredibly excited about all the work that you've done. We are all very, very fortunate to have you with us. And we look forward to getting a follow-up in another year or so—learning more from what you've found with the Predicine test and from your work with the Alliance as well.

And good luck with everything. Like I said, we’re all very, very fortunate.

Alan Tan: Thanks so much, Sam. And it wouldn’t be possible without urologists like you who are on board with these collaborations. So thank you so much.