Update on Urine Markers for Hematuria and Bladder Cancer - Yair Lotan
June 1, 2021
Yair Lotan, MD, Professor of Urology, Chief of Urologic Oncology, and holder of the Helen J. and Robert S. Strauss Professorship in Urology at UT Southwestern Medical Center
Ashish Kamat, MD, MBBS, Professor, Department of Urology, Division of Surgery, University of Texas MD Anderson Cancer Center, President, International Bladder Cancer Group (IBCG), Houston, Texas
Evaluation of Cxbladder and Adjudication of Atypical Cytology and Equivocal Cystoscopy - Badrinath Konety
EAU 2020: Urinary Markers in Low-Grade Non-Muscle Invasive Bladder Cancer: Ready to Stop Cystoscopies
Testing A Novel Molecular Urine Assay in Patients with Equivocal Urine Cytology - Expert Commentary
Urinary Biomarkers in Bladder Cancer: A Review of the Current Landscape and Future Directions - Beyond the Abstract
Ashish Kamat: Hello and welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat from Houston MD Anderson Cancer Center. And it's my distinct pleasure and privilege to welcome today a friend, colleague, and an expert in the field, Dr. Yair Lotan, who is Professor of Urology at UT Southwestern Medical Center in Dallas. He's Chief of Urologic Oncology and holds a distinguished chair in honor of Dr. Roehrborn as well.
Dr. Lotan obviously needs no introduction in the field of bladder cancer and is a true expert in all things to do with bladder cancer, but especially when it comes to urinary markers. He is in many ways, actually, my go-to person when I have a question. And it was just fitting to invite you, Yair, to present to our audience what you think is new and novel and an update in urine markers for bladder cancer. So Yair, the stage is yours.
Yair Lotan: Thank you very much for inviting me to give an update on urine markers for hematuria and bladder cancer. This is an area that I'm quite passionate about, and I will try to give some focus slides on some new studies that have come out and then some commentary about the topic.
Here's my declaration of interest. One of the first things I'd like to discuss is the challenge with cytology. People in urology often use cytology as a backup to cystoscopy, especially to detect patients who have carcinoma in situ. And we know that it's critical for diagnosis to try to find malignant cells, but there's a limited sensitivity when exfoliated tumor cells are rare or confounded by reactive or generative, or reparative cells shed from the urothelium, especially in patients who are getting intravesical therapies.
This study that came out in December of 2020 looks at a single-cell sequencing enabled assay for detection of rare malignant cells in the urine. And they looked at the Warburg effect, which is an effect that commonly happens with metabolism. And we find that cancers have an increased glycolytic rate and they have overexpression of glycolysis-related genes. And the first enzymatic step of glycolysis is catalyzed by hexokinase, and there are two isoforms in mammals. One is expressed in normal adult tissues, but HK2 is over-expressed primarily in cancers, and it was one of the main mechanisms that we use for PET scans.
In a paper that was published this December, Dr. Wang, et al. use an assay to detect tumor cells in the urine by immunostaining for HK2, pan-cytokeratin stain in DAP1. They then isolated these cells and performed a single-cell detection of FGFR3 and TERT mutations or copy number variation. So the thought was to figure out what cells in the urine are actually likely to be cancer, and then do genetic mutation analysis to try to find out if they're in fact cancer cells.
So it looks at cytology, but in a much more sophisticated manner. They did a training set in eight bladder cancers, but they were able to look at 385 cells and developed a threshold for specificity of over 90%.
So the scientists used a training set of eight patients and were able to look at 385 cells. And they sequenced these to establish a threshold for achieving more than 90% specificity. They then tested a blinded patient group of 384 patients, including bladder cancer and benign conditions, and found the sensitivity of 90% for detecting cancer, a specificity of 88%, and a positive predictive value of 83%.
I think this is significant because we all know that the sensitivity for cytology can be very low, especially in patients with low-grade disease. And so it will be interesting to see if this type of assay or a single-cell sequencing, which is really emerging as a hot topic, will lead to better assays for detecting bladder cancer.
The next study I would like to talk about is a validation of the inquiry of multiplex bead assay immunoassay. And this really is novel because it is a protein-based diagnostic assay to measure the relative abundance of multiple proteins associated with bladder cancer. They use Luminex technology and a custom immunoassay to look at 10 different urinary analytes. And we're familiar with different protein tests like NMP-22. But unlike that test, this is really the first one to look at many, many different types of proteins at the same time. And they're developing a bladder cancer-associated signature. This paper is really more of a methodology paper because they really tried to determine a lot of the factors that would have impact on the precision of this assay, including sensitivity, specificity, dynamic range, and detection threshold. And it's also important when you're doing a multiplex assay to look across your activity of proteins, develop detection of what minimal amount of protein you can find, and determine reproducibility.
And so this immunoassay is really the first one that is using multiple proteins. And there are a few that are looking at multiple RNAs. For example, the Cxbladder assay does that. But there are no currently available protein multiplex tests. And at this time the test is really ready to be evaluated clinically. And so there really aren't any results yet, and that's something to look forward to in the future. But I think the main significance of this paper is that they really talk about a lot of the different steps that need to be taken in order to develop an accurate assay. Whether or not they're clinically valid is really obviously the next step. But if anybody's interested in methodology, this is a good paper to read.
I'd like to talk about a few currently available markers that have some new publications this past year. The first is looking at the Xpert Bladder Cancer Detection test, which measures five mRNA targets. And this was the validation of this test in hematuria patients. And it really evaluated 828 patients and found a sensitivity of close to 80%, sensitivity for high-grade of 90%, a very high negative predictive value at 98%, and a specificity, which was quite good, about 84%.
In this study, they compared to cytology UroVysion and found a higher sensitivity and negative predictive value for the Xpert test. But cytology still had the highest specificity, close to 97%. But overall, the marker performed very, very well for detection of high-grade disease. And the specificity was quite high, which is important to reduce the number of false-positive results.
There were several studies that evaluated the ADXBLADDER test that come from Arquer. And this is a urine test that is based on the detection of MCM5, which is a DNA licensing factor, which is expressed all cells capable of dividing. The expression is usually restricted to basal stem cells, however, malignancy MCM5 is expressed throughout the urothelium. And obviously, the benefit of this test compared to normal cells is that normal cells are not dividing as rapidly.
In this study that was published by Dudderidge in European Urologic Oncology, they did a blinded prospective study of 856 patients across seven centers, all of them with hematuria. Bladder cancer prevalence was 8.6%, which is relatively high for microscopic blood and closer to what you would see in patients with gross hematuria.
When evaluating the sensitivity for all tumors, it was around 73%, 65% for non-muscle-invasive disease, and 100% for muscle-invasive cancers. Low-grades were lower than high-grade with about 55% sensitivity for low-grade and 86 for high-grade, and 88% for carcinoma in situ. Patients who had any times of invasion like pT1 where 97% sensitivity. For pTa, it was lower at 53%. And this type of finding is pretty common for most markers.
When comparing patients with visible to microscopic blood, you found a similar sensitivity around 75%, similar specificity around 70 to 72%. The prevalence of disease, as expected, was higher for visible or gross hematuria, around 12% compared to microscopic blood, which was 4%. And this directly impacts the positive predictive value. Unfortunately, as you can see for most markers, because the prevalence of disease is relatively low, even with a high sensitivity and specificity, the positive predictive value is relatively low, about 10% from microscopic blood, 26% for gross blood. But the negative predictive value is quite high, 95% for gross hematuria and almost 99% for microscopic hematuria.
In the second study looking at the use of this marker for surveillance, they published the multicenter blinded study in 21 centers with 1,400 patients. This was published by Roupret, et al. in the Journal of Urology. 127 patients had recurrence and the overall sensitivity was 44%, a little bit lower than in the hematuria population, 38% for pTa, 75% for T1, and 100% for T2. Low-grade was 30%, high-grade was 73%. So not quite as good as in the hematuria population, but not necessarily that different than some of the other markers.
Before I move on to this next slide, I think it's important to point out that it's difficult to compare one marker to another when they have independent perspective studies, and there are really no good randomized trials where each marker is assessed prospectively in a large fashion, even though many of them do report results compared to other markers.
This next study is looking at the use of EpiCheck in patients who have cytology. And I specifically focus on patients with atypical and suspicious cytology. There have been several studies looking at other markers, such as UroVysion and the Cxbladder Monitor, looking to see whether or not you can use them as adjunctive measures to help clarify patients with atypical cytology.
And so within this study, there were some patients who had atypical cytology. There were 68 patients in this study out of the 375. And what you see is that when the Episcore is less than 60, 96% of the patients with atypical cytology had a biopsy negative, and there were two patients that were missed who had atypical cytology and ended up having cancer when the Episcore was less than 60. It's only about 4% false-negative rate. On the other hand, when the Episcore was more than 60, 43% had a biopsy positive and 12 were diagnosed with non-high-grade urothelial carcinoma.
So a reasonable positive predictive value. And if one uses the Episcore to help determine who should have a biopsy, you'd be able to avoid biopsy in a lot of patients while only missing 4% of patients with high-grade urothelial carcinoma. Whether or not it's worth doing a biopsy for everybody or not, for atypical cytology, I really don't think so. At our institution, very few patients with atypical cytology actually ended up having a cancer. However, it is very institution-dependent because it is very much based on the pathologist and how aggressive they call patients suspicious or atypical.
Similarly in suspicious cytology, they only had 37 patients, but if the Episcore was negative in those patients, none of them had high-grade urothelial carcinoma. On the other hand, more than half the patients who were positive, with an Episcore over 60, did have cancer. So it did seem to be a helpful adjunct test in that population.
The new microhematuria guidelines came out this past year, and I think it's important that urologists familiarize themselves, as the new guideline did change the risk stratifications for patients with hematuria. But I do want to focus on one of the points, which is that clinicians should not use cytology or urine-based tumor markers in the initial evaluation for patients with hematuria. And they can get cytology in patients with persistent microhematuria after a negative workup of irritative voiding symptoms or risk factors for carcinoma in situ.
So I think it is important to note that even though there are markers that can be used in hematuria, at this time, based on the current evidence, the guidelines did not recommend routine use. There is a randomized trial currently going on at six centers. UT Southwestern is one of these centers. And the goal of this randomized trial is to determine whether or not a marker can be used in the evaluation of patients.
There are two arms, and I know this figure is a bit complicated, but the main point is that there's a control arm where the patients get standard of care. In the test arm, which is the marker arm, patients are divided into not low risk or low risk. Patients who are not low risk are getting standard of care. And then patients who are low risk randomized to a marker-based approach where they do not get cystoscopy to a standard of care approach. And the goal is to see whether or not it's safe to use the marker to avoid cystoscopy in patients who are low risk for cancer. And I think it's important to have randomized trials start determining whether or not there is a clinical role for markers in the evaluation of patients with hematuria.
Thank you. That's the end of my presentation. And I look forward to any questions or commentary with Dr. Kamat.
Ashish Kamat: Thank you, Yair, for that. That's a very insightful and succinct discussion on the urinary markers that are available and in development for patients with bladder cancer and hematuria. Of all the people that have been working in this field, I think you have put in a lot of thought into what works, what doesn't work. And for the benefit of our audience, if you sort of think of the patient's journey, so when the urologist or even for the PCPs that might be listening into this, when they see a patient starting out with symptoms and then hematuria with a microscopic of gross, and then when they're referred to urologist, would you be able to take our audience through their journey, and at each interval, sort of give us your views in which would be the best assay to use that we currently have, and which would be the best that's being developed?
Yair Lotan: Sure. So maybe we need to take a step back and actually look at the role of markers currently because I think that we've had the benefit of 20 years of different markers being developed over time. And I'd have to say that, quite honestly, none of the markers have really established themselves as routine use in clinical care. And I think there are several challenges with incorporating markers. And I think when you start off with a patient who either has visible blood or microscopic blood, and you're a primary care physician trying to decide what to do with this patient, we know that right now, many patients aren't getting referred to urology for a variety of reasons. And I think maybe the most common reason is that there's still a relatively low yield of bladder cancer, especially in patients with microscopic blood.
The new hematuria guidelines really try to start addressing this issue by giving different recommendations to low, intermediate, and high-risk patients. But still, the majority of patients are recommended to see a urologist. And I think there could be a role for markers right now in terms of trying to figure out who really needs cystoscopy. While urologists do over a million cystoscopies a year in the US, the fact is, for a variety of reasons, benign and cancer detection, patients don't feel that comfortable or thrilled with doing a cystoscopy, which is so invasive to them. And we probably do need to find out if markers can help in this role.
Do I think there is a marker that I would recommend right now? No. And that's mainly because there really hasn't been a validation of use. As I discussed in one of my slides, we are doing a randomized trial looking at incorporating a marker in the evaluation of patients with hematuria. I do think that there is a role because I think a lot of patients would prefer to give urine than to have cystoscopy if they have microscopic blood. But I don't feel like right now I can make a specific recommendation.
However, a lot of the markers have similar performance in terms of sensitivity. And I do think that there could be several different markers that could play this role. But again, I think validation is important. And I do think we need to incorporate randomized trials in the selection of these patients.
As far as surveillance of bladder cancer patients, I think a lot of how we do surveillance also depends on the risk for the patient. We obviously do less cystoscopies in low-risk patients, more cystoscopies in high-risk patients. And I think that for intermediate-risk patients, some of it depends on what kind of risk they are.
You published a paper looking at intermediate-risk and showing that some of them are on the higher end of the spectrum, some on the lower end of the spectrum, and I agree. A patient who recurs once every two or three years is intermediate-risk, but I don't feel that obliged to do as many cystoscopies as a patient who may have a small high-grade tumor or somebody who has 10 tumors in their bladder.
However, I do think there might be a role for markers to try to space out the interval in some patients. So rather than doing cystoscopy every three months, maybe we can do cystoscopy every six months in some patients. And there could be markers that we could incorporate in that role. A lot of the markers do have a high-negative predictive value, especially when looking at high-grade disease. And I think that's the critical thing. We know that most patients with low-grade disease won't die from their disease. They're unlikely to developed progression or metastases. And so looking less often and using a marker to alternate with cystoscopy does make sense. We don't have prospective studies that have proven this concept, and I would strongly encourage companies to do these studies. But I definitely think from a performance standpoint, many of the currently available markers could play that role.
A second issue that a lot of urologists face is that they often will see indeterminant patches or have atypical cytology as an outcome. And then there is a point of anxiety from either the urologist or the patient, or both, about what was that red area cancer? Should I biopsy? What if it happens to the patient who is on BCG? Do they have BCG-unresponsive disease or not? I think these are situations where it can make a clinical difference whether or not you do take a patient to the OR to do a biopsy.
On the other hand, there are a lot of false-positive results for our cystoscopy and even for cytology. And when it's atypical, that we don't want to operate on every patient who's over the age of 80 or even their high 70s, who smoked a lot. For them, the risk of anesthesia is not negligible. You could do office biopsy in some of those cases. You can do enhanced cystoscopy, or you can use a marker. And there's definitely evidence from a variety of markers now, whether or not it's UroVysion, whether or not it's Cxbladder Monitor, or whether or not it's EpiCheck, where you could try to use a marker to try to identify which patients with atypical cystoscopy or atypical cytology really merit a biopsy. Because those tests, in conjunction with what you saw or with cytology, significantly increase the positive predictive value. So it will minimize the number of unnecessary biopsies patients undergo.
As far as the routine use of these markers for surveillance, I have a bit more of a challenge because the PPV for a lot of these markers ranges between 10 and 20%. And that, in comparison to cytology or the positive predictive value is over 80%, means that if you take every patient with a positive marker to the operating room to biopsy them, only one in five patients will actually have cancer. And that means you'll put patients through a lot of unnecessary biopsies. So I'm a bit hesitant to recommend one specific marker for routine use and surveillance because I think many urologists will find that they have a lot more false-positive results. And that obviously also increases anxiety for patients, increases the number of unnecessary procedures, and increases costs to the healthcare system.
Ashish Kamat: That's very wise advice. And I'm glad you brought up the fact about cytology and its potential limitations. You have also done a large amount of work when it comes to clarifying the results of cytology. And with the new Paris reporting system being adopted by most academic centers and hopefully a lot of community centers as well, have you noticed any change in the percentage of the so-called atypical cytology? And what is your current go-to clarifying test for those instances when you still get cytology results from the lab, which really doesn't help much clinically?
Yair Lotan: Sure. Well, I think it's important that pathologists will stay up to date with the current recommendations for our interpretation of cytology because that really impacts us as urologists a lot. We're the ones who have to convey those results back to a patient. It often is inconvenient, because you do your cystoscopy. It's clear. You tell the patient, "Hey, good news. We didn't see anything. I'll see you back in three months or six months." And then a week later, the results come back and I have to call the patient, say, "By the way, the pathologist wasn't thrilled with what he saw. There's some atypia and now we have to decide what to do."
And I think the Paris new interpretations really have helped. We have had less atypia. I do notice a lot of variability between even my own pathologists, knowing how they interpret cells. And I definitely think that no matter what we do as urologists, the intravesical therapies do make it more challenging for the pathologist in interpreting cells, because I think reactive cells are difficult for pathologists. And especially in patients who've had multiple resections, maybe we still see sloughing of granulation tissue or things even three months after an event, or patients who get intravesical chemotherapy, I think they have abnormal-looking cells.
And so they do sometimes call patients atypical more often than I would like. We have had for many years a reflex protocol using the UroVysion FISH assay. As I mentioned, it's not the only assay that could be used as a reflex tool. But the nice thing about it is that I don't have to get more urine from the patient. It can be used off the original cytology specimen.
If you're going to use a different assay, such as a Cxbladder test, you would need a separate urine sample. It's not that difficult to get another urine sample from a patient, but there is an inconvenience factor of the patient coming in to bring it in. And our pathologists routinely will do a reflex assay and notify me of the results without me having to put in a separate order. But again, that's the main reason I'm using it, not because I think it has better performance characteristics than the other markers. And if one wanted to get another fresh urine from the patient, then that would be perfectly reasonable. And there are a variety of other assays they could use.
Ashish Kamat: So if you were advising a community practice that didn't have the benefit of the pathologist doing this reflex test the way you did, is there still, in your estimation, a preference to use UroVysion, or would you pretty much say that they could pick any of the current markers that are out there? What's your sense, based on the literature to date?
Yair Lotan: Well, in the guidelines there is data for UroVysion and for ImmunoCyt, which I don't think is necessarily commercially available, last year there was a paper looking at Cxbladder Monitor and showing that it could help adjudicate cytology. In this past year, there was one looking at EpiCheck. And so a lot of the markers seem to do a similar job in terms of performance for this setting. And I think a lot of them would probably be fairly equivalent, even though I don't think there's going to be really good comparison studies showing one is superior to another.
Ashish Kamat: And again, there are so many different markers that are out there. There's this confusion in the minds of people sometimes just based on the marketing that companies put forward as to why this test is better than the other. And hearing from someone like yourself that there's not that much to pick between the different markers is reassuring to those that sometimes don't have access to certain urinary markers, and have to go with whatever's available.
On the other hand, there has also been a push, and this is periodic and I'm sure you've seen this, and then it gets exposed, news media and then it goes away, of people using multiple markers in the same patient at the same time. I personally have always strongly advocated against that, but I'd be interested to hear your views. Would you ever recommend using more than one marker at a given time for a particular patient?
Yair Lotan: No. I really don't know how that would really be beneficial to the patient. One could argue just from a mathematical standpoint that if you want it to boost sensitivity, you could use a protein marker and an RNA marker. And one would know that it's possible that cancers carry different characteristics and that one marker would be positive and the other one would be negative and vice versa. That'd be fine. But by definition, you will also lower specificity. And specificity is really one of the biggest problems because a low specificity means a lot more false positives. And so if you made a decision, I'm going to use two different markers, then you would say, okay, you have to decide if am I going to treat it as a positive result if either is positive? Or am I going to treat it as a positive result if only both are positive? And that type of calculation, I think, just can get very confusing in the minds of anybody.
I think the first decision you actually to make is, am I going to do anything about the marker if I even order it? I really don't like ordering tests that aren't going to change my management. So for example, if I look in the bladder with white light, and then I don't routinely use the urine markers currently, other than as a reflex test. So you might ask, what would I do if the UroVysion is positive? Generally speaking, I do two things. One, it might bring him back for a blue light cystoscopy in the office because I have a flexible blue light in my office that will help detect additional carcinoma in situ. And also will assess upper tract imaging because I think the use of upper tract imaging is highly variable. I don't use it routinely for low-grade disease, I use it every one to two years for our high-grade disease. So sometimes I haven't had a recent upper tract imaging, I'll do that if I have a positive marker.
But I think if you just use it routinely, you have to ask yourself, what am I going to do if I looked in the bladder, I didn't see anything, and I have a positive marker? I think the answer is not to go to the operating room on every patient. That will be far exceeding the need and will put a lot of patients at risk for unnecessary anesthesia and potential complications. And so when people are using a marker, one or multiple, they really need to know how they're going to interpret the results and how it's going to impact patient care. But in general, I think it would be very challenging to use more than one marker routinely, and use it in a clinically useful way.
Ashish Kamat: Thanks for clarifying that. I'm glad you and I are in agreement there. Just to clarify a few things for the audience, just as we're wrapping up the segment, and I don't want to paraphrase or put words in your mouth so correct me if I'm wrong, you did state that you get cytology in all high-risk patients and the lab uses a reflex UroVysion for those that are atypical. Is that accurate?
Yair Lotan: Yes. And the use of cytology, I think in low-grade patients, I think that's not routinely recommended. But I also don't think it's unreasonable to use it in cases where you might worry that somebody might have carcinoma in situ or progression of disease.
Ashish Kamat: Great. That was going to be my next clarifying statement. And again, I'll just make the statement for the benefit of our audience. You correct it or expand on it. So for a low-risk, first-time patient that has just a solitary tumor and no recurrence, there's no reason to do a cytology on those patients. But in those patients that have maybe recurrent tumors or you're worried that they're more than just the low-risk, or the lower-risk, intermediate-risk patients, there you may use cytology on an ad hoc as-needed basis, correct?
Yair Lotan: Yes, that's correct.
Ashish Kamat: Excellent. Great. Yair, as always, chatting with you on markers, I learned a lot. So I do want to thank you for taking the time off from your busy schedule to spend some time with us. Stay safe and stay well.
Yair Lotan: Thank you so much. Appreciate your time and thanks for inviting me.