Revolutionizing UTUC Risk Stratification: The Promise of Urinary and Plasma Biomarkers - Helen Hougen

March 14, 2024

Sam Chang converses with Helen Hougen about her presentation on predictive biomarkers for upper tract urothelial carcinoma (UTUC) at the SUO, emphasizing their critical role in risk stratification. They delve into the limitations of current diagnostic tools like ureteroscopic biopsies, which often provide imperfect information and pose significant patient burdens. Highlighting the promise of minimally invasive biomarkers, they discuss two studies on cell-free DNA, revealing its potential in accurately diagnosing and predicting treatment outcomes for UTUC. Dr. Hougen's enthusiasm for integrating artificial intelligence to extract deeper insights from existing patient data points towards a future where diagnosis and treatment of UTUC are both less invasive and more precise.


Helen Hougen, MD, Urologic Oncologist, The University of Iowa, Carver College of Medicine, Iowa City, IA

Sam S. Chang, MD, MBA, Urologist, Vanderbilt University Medical Center, Nashville, TN

Read the Full Video Transcript

Sam Chang: Hello everyone. My name is Sam Chang. I'm a urologist at Vanderbilt University in Nashville, Tennessee, and we have the honor of having Dr. Helen Hougen from the University of Iowa, the Department of Urology there. She's an assistant professor. I got to meet Helen. She probably doesn't remember this, but I remember meeting Helen as she was applying through fellowships and actually completed a fellowship with one of our former fellows, Chad Rich, down at the University of Miami. And I've always admired her research efforts, her integrity, her honesty, and her humility, which is a tough combination when you talk with urologic oncologists. So she has them all in spades.

We're going to focus on an SUO presentation that she gave recently, looking at perhaps the use of predictive biomarkers in looking at risk stratification with upper tract urothelial carcinoma. So it's an area that we focus on here, at least I do in UroToday, looking at different types of treatment options, different types of diagnostic options, looking at upper tract urothelial carcinoma.

So, Dr. Hougen, thank you so much for spending some time with us, and we look forward to having you give us an idea and an update regarding these predictive biomarkers.

Helen Hougen: Great. Thank you so much for having me. It's truly an honor to speak with you and also to hopefully learn from you, who I really admire as a giant in the field. So thank you.

I'd like to spend some time to talk about the role of biomarkers for upper tract urothelial carcinoma and risk stratification. Good biomarkers provide useful clinical information and are obtained in a minimally invasive manner. Right now, in upper tract urothelial carcinoma, the gold standard for obtaining tissue biopsies leaves a lot to be desired. Ureteroscopic biopsies have two main disadvantages. First, they provide imperfect information. For example, one meta-analysis that compared ureteroscopic biopsies to final nephroureterectomy biopsies showed that ureteroscopic biopsies understage at a rate of up to 46% and undergrade at a rate of up to 32%.

And second, the process of obtaining these biopsies places a high burden on our patients, from the procedural burden with anesthesia, the cost, the time, and the travel, to complications that are inherent to any surgical procedures, as well as the well-known risk of lower urinary tract seeding. And due to these limitations, we're not accurately selecting patients for effective therapies such as neoadjuvant chemotherapy, which we know has good activity in high-risk disease. And with more accurate staging, we can also potentially avoid the risk of overtreating a patient with a nephroureterectomy instead of providing appropriate nephron-sparing therapies to avoid the medical and the financial burden of chronic kidney disease.

Right now, I think we have really good tools to treat upper tract urothelial carcinoma, from effective neoadjuvant and adjuvant chemotherapy to good endoscopic and chemoablative options, to extirpative surgeries when we need them. So now, we need better information to guide us in terms of which tools we should use. We need good biomarkers that provide accurate diagnosis, that are predictive of who would benefit from certain treatments, as well as prognostic so we can better risk stratify our patients.

And right now, the two most commonly used biomarkers in upper tract urothelial carcinoma are imperfect. There's cytology, which can tell us if a disease is high or low grade but overall is a poor predictor of invasion or high-grade disease, and FISH, or fluorescent in situ hybridization, which can be more sensitive for diagnostics overall but doesn't necessarily help us distinguish between high versus low-grade disease.

And now, moving on to an exciting frontier of cell-free DNA, which are extracellular DNA from either the blood in the form of circulating tumor DNA or urine in the form of urine tumor DNA. These markers are easily available, and they also have a short half-life, which captures the "real-time biology" of the tumor. And we know in bladder urothelial carcinoma, tumor DNA can reflect disease burden and predict recurrence.

There are two recent studies that looked at upper tract urothelial carcinoma. One study out of China examined urinary tumor DNA in patients prior to surgery. The study demonstrated good concordance between urine and tissue variants. As you can see here in this heat map, one column is a patient, and one row is an alteration, and the red triangles indicate tissue sample positivity and yellow indicate urine. And so, you can see there's really good concordance. They also looked at methylation levels of a transcription promoter called ONECUT2, in addition to a DNA panel of 17 genes. The group had developed models based on the gene mutation level, methylation level, age, and specific status of one transcription promoter. And the AUC of their final model, which included methylation status, age, and TERT promoter status, had a really incredible sensitivity of 94% and AUC of 0.96.

Notably, the purpose of the study, the primary goal, was for upper tract urothelial carcinoma diagnosis, but in their gene panels, there were certain signatures that were more common in high-grade or muscle-invasive disease that may serve as signals that can be useful for risk-stratifying patients.

Another study out of Moffitt Cancer Center looked at 30 patients prior to extirpative surgery who did not receive neoadjuvant chemotherapy with upper tract urothelial carcinoma. They looked at a 152-gene panel sequenced from plasma cell-free DNA, and they found that more variants were found in invasive samples compared with noninvasive samples, and that two or greater plasma DNA, ctDNA variants, and a copy number burden higher than 6.5 predicted invasive staging. Interestingly, adding certain clinical factors did not improve their predictive power. And as expected, patients with circulating tumor DNA positivity preoperatively had worse survival outcomes than those who did not have ctDNA positivity.

And as we can imagine, a future use that has been shown in bladder urothelial carcinoma is potentially using circulating tumor DNA dynamics to associate with treatment response.

And so, looking ahead, we can imagine a situation where we can use a combination of imaging and liquid biopsies to diagnose a patient with upper tract urothelial disease. We can then treat them with neoadjuvant chemotherapy if it's appropriate. We could use liquid biopsies to assess a response, consider ablation versus extirpative surgery to consolidate their treatment, and then use the liquid biopsy more as markers to monitor their response. And ultimately, our goal should be to decrease the invasiveness and increase the accuracy of our diagnostic tools by incorporating these promising biomarkers for patients with upper tract urothelial carcinoma.

Thank you.

Sam Chang: Thanks a lot, Helen. Everybody wants to find that magic marker that will help us not only diagnostically but also predict therapeutic response, to either escalate or deescalate care.

So, let's look at... Looking at your data, you had a set of urinary markers, and you had a set with the plasma cell-free DNA. So let's start off with the urinary marker evaluation. That was done more for perhaps being able to predict who had upper tract disease, correct? And so you had rates of greater than 90%, which was very... Tell me a little bit more about that cohort. Were those patients with suspected abnormalities on CT scan imaging, people who had no disease? Tell me a little bit about that cohort. And where do you think we need to go next with urinary markers to diagnose upper tract disease?

Helen Hougen: So, that study was a very interesting study where I think they took patients who had hematuria and found upper tract urothelial carcinoma in a subset, and then did urinary and tissue markers for diagnosis. And in terms of where we're going, I think one of the things that I'm always curious to explore is how we can more accurately localize disease location in terms of using, for example, urinary DNA or urinary markers. For example, at the University of Iowa, we often do localizing urine studies, look at cytology and FISH, with the idea that finding out exactly where the cancer is and is not will be a huge part of how we treat patients with the goal of organ sparing, especially as we have more and more treatments such as in the chemo ablation field. I think that ability for markers to localize from the urine will be very important.

Sam Chang: So as you all evaluate suspected lesions, you do tend to then do ureteral washings or renal pelvic washings and compare them to the bladder wash cytology. Tell me a little bit about the actual practice of evaluation of someone with a suspected upper tract lesion. What do you do now in terms of evaluation and diagnosis?

Helen Hougen: Do you mean in terms of patients who you suspect have urothelial carcinoma?

Sam Chang: Correct. Yeah, correct.

Helen Hougen: Okay. Yeah. I think right now, our clinical practice is probably pretty common to other places, where if you have a suspicion from a selective cytology, then we have to examine ureteroscopically in a relatively invasive manner. But I think if we can fine-tune and expand the role of urinary tumor DNA, then that would probably be the most exciting field.

Sam Chang: Got it. And that'll also then help, especially in times where we cannot get tissue or where it's too dangerous, or it's too bloody, we haven't sampled it well, or it's a lower pole lesion, very difficult to access, etc. It would be great to have a surrogate, a marker in the urine that could tell us definitively. Now, if it can tell us low grade versus high grade, that even gives us more information.

So then let's look at the plasma data from Dr. Huelster in the European Urology article looking at the plasma presentation of circulating tumor DNA. And those were patients that had, again, known upper tract lesions, is that correct?

Helen Hougen: Yes. Yeah, 30 patients prior to nephroureterectomy.

Sam Chang: And then so overall of the 30, how many in fact actually had lesions or had circulating tumor DNA found prior to the surgery?

Helen Hougen: That, I'm not sure. I don't remember the specifics from that paper.

Sam Chang: Okay. Because clearly, your graph shows that if you had circulating tumor DNA before the nephroureterectomy, the outcomes tended to be worse. I thought your algorithm at the last slide that you presented I think would be very appealing to all of us. You have someone with circulating tumor DNA at the outset of diagnosis. Those are the patients perhaps you really want to give neoadjuvant chemotherapy prior to your nephroureterectomy and then be able to assess the response of the neoadjuvant chemotherapy. And then if they still have circulating tumor DNA, then you do the nephroureterectomy, check again, and see if that's cleared or not. And if it's cleared, you feel better. If not, perhaps you consider even more strongly either additional therapy or adjuvant therapy, etc. So I think that that stage is really not too far from now in looking at those patients that do or don't present circulating tumor DNA.

So what are you all doing now at Iowa when it comes to these patients where you have a lesion that you're planning on doing a nephroureterectomy? Are you starting to incorporate any of these markers?

Helen Hougen: Right now, we're still using just FISH and cytology. However, as you know, Dr. Mike O'Donnell is a huge fan of using intravesical as well as upper tract chemo ablation. In addition to Gem-doce, we have been incorporating more, for example, Jelmyto, even in patients perhaps with some high-grade disease for the chemoablative properties, especially if some lesions cannot be ablated completely ureteroscopically. And we're also incorporating, I would say, a pretty incredible and creative armamentarium of combination chemotherapies, including Gem-doce, including sometimes Valrubicin, Docetaxel, to try to treat and save renal units for these patients. I just looked back on Dr. Huelster's paper, and it looks like 11 patients out of the 30 had ctDNA positivity.

Sam Chang: Right. So in the patients that have ctDNA positivity, it's definitely helpful in terms of monitoring perhaps response, just as you said, to therapies, and then perhaps considering those patients to be those that you want to escalate treatment either with neoadjuvant or consider treatment afterwards. I was struck with your treatment considering Jelmyto for those patients with upper tract, but you all are also considering other retrograde treatments with ureteral catheters or through nephrostomy tubes. Tell me a little bit more about that.

Helen Hougen: Yeah, typically we place nephrostomy catheters and then essentially just drip whatever agents in, or in the case of Jelmyto, push it in and let it dwell at the appropriate time.

Sam Chang: I don't want to give away any of your findings, but have you all had some success with those upper tract treatments? We've struggled with BCG for CIS in the upper tract, and there's, I would say at best, mixed data. It can be quite dangerous. There clearly have been some that have responded. There clearly have been many that have not responded. If you could give us at least a teaser, and maybe in a year or two, we'll be able to get you back and tell us a little bit more about these upper tract combinations of chemotherapy agents. Can you give us an idea that there's perhaps some promise to these combinations?

Helen Hougen: Yes, absolutely. So in terms of the Jelmyto, I think in addition to some of the published data from, I think it was in European Urology Focus, I think it was Dr. Rose's publication in terms of looking at Jelmyto for high-grade disease. I think, at least based on our data, we find that about half of the patients do respond with high-grade disease and Jelmyto, even though technically the indication I think is for low grade. And in terms of the combination therapies, there have been some success, some of which our group has published in upper tract already using Gem/doce. And again, what at least we found is in Gem/doce, if that doesn't work, then we have some other treatments that may show promise as second line or potentially even third line treatments after Gem-doce.

Sam Chang: Well, that's really... I think the real movement with everything that we've done in oncology is an attempt at least at organ preservation, and in addition to deescalation or treatment when possible, escalating when we need to, but deescalating as well. And so, having regimens for those individuals that we really want to try to do nephron sparing, it's incredibly exciting. And to have more options only is helpful for our patients.

Helen Hougen: Absolutely. Yeah.

Sam Chang: Yeah. Helen, tell us next what you're looking at in terms of research. Anything exciting that you guys are looking at in the urothelial carcinoma world?

Helen Hougen: Yeah. I think like many other groups, the most exciting frontier, or one of the many exciting frontiers, is incorporating and harnessing the powers of artificial intelligence to essentially get more information out of the biopsies or imaging that we already have. And one potential use that other studies have already highlighted would be using a deep learning model to look at perhaps pathologic slides or radiographic studies that are already necessary for patient care for getting as much information as possible, but then using an AI model to get even more information to help us find essentially AI markers for pathologic and radiographic predictors of response and prognosis. So I think that's really the next very exciting frontier. I think we're just scratching the surface, and there's a lot more to do.

Sam Chang: Well, Helen, thank you so much for spending some time with us. I think that obviously your future is bright, and with the historical work that's been done at the University of Iowa, with your efforts, I think the team will continue to make great inroads into finding better ways to diagnose and treat our patients with urothelial carcinoma. So thanks again, and I look forward to seeing you continue your success.

Helen Hougen: Thank you very much for having me. It's been really fun to speak with you.