Sam Chang: Hi, my name is Sam Chang. I'm a Urologic Surgeon at Vanderbilt University in Nashville, Tennessee. And we are quite fortunate to have Dr. Alex Zlotta, who is a superstar in urologic oncology, really is one of the true leaders and has been for decades, actually. He currently is a professor and the Howard Sokolowski Chair of Urologic Oncology Research.

He, as many of you all know, is at the University of Toronto and really has made significant inroads in different aspects of bladder cancer care. He and his team have recently actually published in the Journal of Urology, an examination of what we thought was a more of a perhaps nuisance tumor, the low grade non-muscle invasive bladder cancer. And his work, I think, actually will be eye opening for many of us. So, Alex, first of all, thank you for all the contributions you've made in the past. We know you will continue to do so. But thanks for spending some time with us, and we look forward to your presentation.

Alexandre Zlotta: Thanks, Sam, and it's a really great pleasure to be with you. I am very happy, as you mentioned, to talk about a disease that we usually consider as indolent, and to look at the recurrence risk, the metastatic potential, and the length of surveillance for low grade non-muscle invasive bladder cancer.

And as you mentioned, this has been just published in the J. Urol, with Eunice Villegas and an amazing student who did a lot of the heavy lifting here in terms of data and also in collaboration with some centers, I explained, in France and even in Vancouver. And so we all know that non-muscle invasive low grade is what I call hugely with patients pussycats.

But aren't they out there just wolves in sheep's clothing? And how can we actually detect them? And so there's, as we discussed, a widespread assumption that Ta low grades never, ever progress to potentially lethal disease, that very few patients are at risk of developing metastatic disease or die from the disease.

But the reality on the ground is that the long term follow up data are really limited. We just don't know. We don't even know how long we really should follow those patients. Many people say five years and then shared decision-making, in the UK actually even one year. But we don't know what's going on when people have been free of disease for five years and what happens down the road.

And so the aim of this study was basically to look at the incidence of late recurrences, the risk of metastatic disease and death from bladder cancer in patients who started with low grade Ta in primary. So we accrued to do so over 500 patients with a long follow up at the University Health Network in Toronto, both Ta low grade and PUNLMP. And we had multiple endpoints to assess their evolution.

Interestingly enough, we found a couple of cases that were low grade metastatic. And Sam, I'm sure you're going to tell me that can't happen. Your pathologist was drunk. No, he wasn't. They double-checked everything. These were low grade that became metastatic low grade. And we did RNA sequencing on these tumors. And even more, in order to understand differences, we actually had a pathology review of a subset of those cases.

And so the first table shows the classical distribution of those 500 patients, nothing really out of the norm. But what strikes right off the bat here is the amount of progression to muscle invasive or nodal or metastatic disease. 30 patients out of 500 did progress. 20 became metastatic. And again, these were primary Ta low grade. It doesn't mean that they didn't evolve through a higher grade or had at one point something in between. But the primary tumor was the start of the journey.

Grade progression from low to high grade was seen in 100 patients, and stage progression in 57 patients. Interestingly enough, 15 out of those 520 patients, that's 3%, died from the disease. And interestingly enough, even in patients who were free of any recurrence for five years, still another 10% still developed bladder cancer within 20 years. Although the silver lining here in reassuring was that very, very few of those patients who had a related relapse ever died from bladder cancer.

And in this series with a 10 year median follow up, which is kind of rare. As I mentioned, we also did RNA sequencing. You see here examples of a primary low grade tumor and the met that was low grade, that was reviewed by pathologists. And what you see by doing RNA sequencing is that there's a disconnect between the phenotype that looks low grade and the genotype, which is in line with some more aggressive disease.

Now, to try to understand whether the pathology differences in interpretation could play a role, we took three world stars in pathology, Theo van der Kwast, John Sweet, and Eva Comperat in Paris. And look at the pathetic kappa of agreement for grade 1, 2, and 3, 1973, only 0.41. And you would say, wow, there's no way on Earth that those pathologists would even disagree between low grade and high grade. Yes, they did. They never always agreed with the kappa in nearing 80%.

But still, far from perfect. Interestingly enough, the low grade Ta cases were reviewed and were really agreed upon as low grade. So the conclusion was that our study is sobering and challenges the assumption that Ta low grade never progresses to muscle invasive or metastatic disease. There is a small amount of wolf, as you mentioned, in sheep's clothing.

There's still 10% of people who recur after five years, although they were recurrence free for the first five years. However, as we mentioned, the risk of mortality in those patients is low. And we really think that efforts and as you know, there's a lot of push in AI for the time being, are needed to minimize interobserver variability. Surprisingly, even experts disagree between low grade and high grade. There is clearly a need to identify wolf in sheep's clothing. So that basically was the summary of this study.

Sam Chang: Alex, incredibly sobering, as you say. But I think very important because not that I always like manuscripts that support my kind of clinical experience, but there's undoubtedly these patients as a subset. I was just surprised with how high a number there was, not only with the long term recurrence, but with that risk of metastatic disease. Tell me your initial thoughts of the data, because I think it's really true, I just didn't know it had the scope.

Alexandre Zlotta: So, I mean, I'm going to tell you something. I'm sure you and me are just very young urologists. We know each other for a couple of decades only. And I have to say that in my career, every time when I had a hypothesis, I was wrong. It doesn't mean that you don't find something interesting, I was plain wrong.

I was hoping that we can do a nice study where we can reassure absolutely all our patients, listen, we'll look for 10 years, nothing happens. We don't need to bother you. That was exactly what I had in mind. Fair enough, exactly the opposite happened. And so this is somehow in line, if you remember, the large study by Steve Williams. If you really look into the data, there was also 3% of mortality. It's buried in the manuscript. It was not highlighted because it was about the cost.

If you look at the Swedish registry, 3%, so I think we are exactly in line. Peter Black and I  just published another paper in European Neurology showing actually 10%, 9.6% of people recurring after five years. So we haven't reinvented the wheel here. We may have had a longer follow up than others, and maybe more pathology review in those kind of things and RNA sequencing simply more granular. But I think these data are, sadly enough, in line with the reality.

Sam Chang: Yeah, I think your point early on when you said we previously had so little data beyond five years. We just made the assumption in the literature that because of that, and since nothing was written, it was safe to do the shared decision-making, et cetera. And it's my experience that there were always these outliers, and I thought, OK, they're only outliers, but this percentage is so significant.

Now, the next question is, are we going to be—and you've started showing that in terms of your RNA-seq data, et cetera—are we going to be able to predict those patients up front? And so tell me your thoughts about that. And what should we be studying next to try to help detect these late recurrence people, these people more at risk for developing and progressing?

Alexandre Zlotta: I don't want to be pessimistic, but I do think it's really, really challenging. And I'm going to tell you why. First of all, even from the clinical data, we were unable to underscore any type of clinical features that were in keeping with a risk of later progression. That was the first thing. The second thing, I think we all agree that it's completely unrealistic to do RNA sequencing on every Ta low grade. That's absolutely ridiculous. We agree on that.

Now, we could have systems where we're going to have low grade and high grade, which are real low grade and high grade, including maybe the 1973 classification, which is very good. But because of the huge amount of variability in the grade 2 mainly, it's not very useful. And so maybe AI will help a little bit there, where I would like to bring this to your attention and our attention. And I know how much you've been working in the field as well is that we have this big program of TMT in Toronto, the bladder preservation, of course.

And what has struck me over these last 20 years is that you have patients who start with a T2 high grade disease treated with TMT, and you have 20% of those patients who recur superficially non-muscle invasive and only low grade. And it's obvious that the genetic hit was the same to start with, that the risk factors. And something drives them to something and to the other.

And I think that as long as we don't understand what happens at the time of the progression from low grade to high grade, and the nitty gritty of bladder cancer, I think we're going to have a little bit of challenges to predict in a reasonable and cost-effective manner. I know it's not very positive, but I'd rather be objective than stupidly optimistic.

Sam Chang: Yeah, I think your point is you're exactly right. I mean, you take 100 people and 70, 80, 85 are going to do exactly what we think. And the resources to find those 10, 12, 8, 15, it's going to be right now is too challenging. But I think your point regarding AI, better evaluation, looking at thousands, thousands and hundreds of thousands of cases as opposed to our experience—we currently base upon it—I think will be really, really informative and important.

This is a great time to highlight your extensive role and your important role in bladder preservation therapy. I mean, the work that you all have done that you've helped lead really has moved our field forward in terms of this. We used to have this dichotomy. It's either cystectomy or not.

Now, with the systemic therapies and now with understanding a little bit more disease mechanism, better evaluation, I just wanted to take this chance to applaud all your efforts regarding that. When looking then at your patients now with the high grade Ta tumors, haven't had a recurrence, you're at the three, four, five year mark, what are you going to tell your patients now? And how do you follow them now?

Alexandre Zlotta: I would say I think that the AUA and CUA guidelines are pretty good. And I basically—you established all these guidelines, and you were the first author, as you know. Without long term follow-up, a little bit common sense, I think you and me also know one thing is that when we started, life expectancy of patients was barely 80 years old. People had an average age of diagnosis at 70, 71. They barely even made it on average to 10 years. And so we literally had no clue because people met their life expectancy. Now, how many patients now are diagnosed in their 50s and 60s—

Sam Chang: Incredibly good point.

Alexandre Zlotta: —and they get to live up to 90? And how many patients don't we even operate well into their 80s and were—and so I think we have to revisit and get more data. I think it's the place of—your place at Vandy and other places to try to bring things to get more and more data. That's what we need.

I do believe that in Ta low grade five years, no recurrence, it's reasonable, as you mentioned, to have a shared decision-making. Listen, you can have a 10% but unlikely to claim your life. Would you like to carry on? For high grade disease, I'm uncertain that there is any follow-up period that is set in stone. I just saw a patient—I remember, I did a Ta low grade multifocal about 19 years ago. She presented with carcinoma in situ multifocal 10 years, and so I just think we don't know; we need to gather data.

Sam Chang: Now, really, really good point, and your point about our increasing longevity. I mean, for me, you're still young, but for me, those ages look younger and younger as far as I'm concerned in terms of longevity. So I think it is very important. And that shared decision-making—it's a term that we use too much—but it really, I think, is important in this situation where we have patients who are reliable versus not reliable, long distances, et cetera. So a lot of things that come into play.

But Alex, thank you so much for spending some time with us. And gosh, every time you put something out, it's always of interest, it's always evidence-backed. And so again, I really want to express our field's appreciation for all the work that you've done. And we look forward to your next contribution.

Alexandre Zlotta: Likewise. Thanks, Sam. It's really great to be with you this morning.