Andrew Katims: Thanks for having me here. So, thank you so much for having me, I'm excited about our recent publication, which Dr. Chang mentioned was just published in the Journal of Urology Predictors and patterns of Non-urothelial recurrence after nephroureterectomy for upper tract urothelial carcinoma. So, we know that about 25% of patients with upper tract urothelial carcinoma will have distant metastasis within the first five years after definitive surgery with radical nephroureterectomy. Patients that do recur generally have a poor prognosis. About 80% of patients will die from their disease within two years of that recurrence. Upper tract urothelial carcinoma, as you know, is a rare disease and a lot of the information is extrapolated from bladder cancer. But as we've learned more about upper tract disease, we know that it's really a distinct entity with its own features and associated outcomes. There have been some small studies that suggest lymphovascular invasion, stage, grade, lymph node involvement, tumor necrosis and tumor architecture are associated with a recurrence.
There has been one study showing that liver and bone metastasis have the worst prognosis, but there is an unclear role of the timing of recurrence and survival based off of site. We know in bladder cancer that patients with late recurrences seem to have a more indolent course and perhaps a better outcome, but that has not been shown yet in upper tract disease. So, the purpose of our study was to use a large multi-institutional database. We used seven institutions participated in this to define the timing and patterns of non-urothelial recurrence and also to validate some of these previously identified risk factors.
Getting into our results a bit, we looked at the site of recurrence and the associated survival with each site of recurrence. Our cohort consisted of over 2,000 patients and 454 in total developed a non-urothelial recurrence. If you look at the timeframe of recurrences, which we'll get to in a slide or two, about two years, 24 months after surgery, 19% developed in non-urothelial recurrence. Most commonly patients developed recurrence at multiple non-urothelial sites at once. So, for example, the lung and the bone, the lung and the liver, lymph nodes, and the liver, for example. Second most common site was retroperitoneal lymph nodes, followed by solitary lung metastases. We then compared the site of recurrence and the risk of death from disease, and this was compared to the other sites of recurrence. So, when patients just had recurrences in the lung, it looked like they had improved survival compared to patients that had recurrences outside of the lung, other sites.
This was also true for solitary lymph node metastases, and not surprisingly, patients with multiple sites of disease at once had the worst survival. We then looked at some of the risk factors for recurrence and tried to validate some of those previously identified factors. We found very clear risk factors of a large tumor size and high grade at the time of surgery being indicative of a risk factor for recurrence. A few noteworthy things for us from this study. Lymphovascular invasion was not significant, although this has been reported to be a significant risk factor, and we believe the reason for this was because of how critical stage was. Stage was by far and away, the strongest predictor for recurrence and this was true even after receipt of neoadjuvant or adjuvant chemotherapy. If patients were not able to get to a lower stage, they had a very significant risk of recurrence.
Finally, getting into the timing. Here, you can see a cumulative incidence chart of the site and the timing of disease. So, recurrences were very common in the first 24 months, first two years after surgery. Again, accounting for about 19% of patients in this study. In our study period, there was no time point which we could see a leveling out or a slowing down of recurrences. One of our secondary goals was to see if we could decipher a de-escalation point of surveillance imaging. But unfortunately, within our study recurrences continued until the end of the study period. We also found importantly that the timing of recurrence did not affect survival. So, patients that recurred prior to 24 months compared to those that recurred after 24 months had equally poor prognosis.
So to summarize, recurrences are fairly common and occurred about in 20% of our patient population, most often occurring at multiple sites. Concurrently, solitary lung and lymph node recurrences had better survival compared to other sites, and multiple recurrence sites had the worst outcomes. In our study, risk factors included large tumor size and high grade disease with by far and away, the strongest risk factor being the stage of disease and that was true regardless of the receipt of chemotherapy or not, and there were no survival differences based off of the timing of recurrence. Thank you.
Sam Chang: Andrew, great presentation. Obviously, we have a great idea that if you have a recurrence outside either the contralateral upper tract or the bladder, these patients are in trouble. So, a couple of questions that come to mind. First, is there, or did you find relationship with a urothelial recurrence and it's impact or prediction of a non-urothelial recurrence and perhaps even outcomes as well?
Andrew Katims: Yeah, so that was something that we did not specifically look at. We did monitor for urothelial recurrences in this study, although it was not a primary outcome because we understand that this is a huge burden for patients and more common than non-urothelial recurrences. So, the most I can say about our data is that we found in our study period, there was a two years, 34% of patients had intravesical or contralateral ureter recurrence, but we did not compare it to find if there was a survival benefit or a risk for non-urothelial sites of recurrence.
Sam Chang: And how about the impact on, I didn't see, I may have missed, but I didn't see data regarding either neoadjuvant or adjuvant systemic therapy of any kind and its impact on these recurrences. I may have missed it, but can you comment on that as well?
Andrew Katims: Absolutely. This was an important part, at least in the planning phase of this study, because the majority of patients are going to get some sort of neoadjuvant or adjuvant therapy, especially patients with muscle invasive or higher stage disease. We did do an analysis of this that did not end up in the manuscript because it was very... The outcomes really confused the picture a bit. Ultimately, what we found was that receipt of neoadjuvant or adjuvant chemotherapy essentially had no impact more than stage. So, the way we interpreted this was that essentially patients who had high stage disease, meaning T3, T4 disease, regardless of the receipt of systemic chemotherapy, even in the neoadjuvant setting or adjuvant setting, still had a very high risk of recurrence. I think we see this in other diseases too, especially with bladder cancer. If patients don't respond to neoadjuvant chemotherapy, their outcomes are worse. And I think that's the interpretation of this result as well.
Sam Chang: No, I think that makes a lot of sense. I think that there's going to be a subset of... With our current systemic treat... And this must change with ADC therapy, combination therapies that we're really looking at that look exciting, but with our chemotherapy in the past or platinum-based regimens, clearly there's a cohort of patients that would either low-volume disease, microscope disease, et cetera, we make a big difference. We may be over-treating a small cohort in the neoadjuvant setting, in the adjuvant setting, but then there's these unfortunate patients that have significant higher stage disease that are more likely to develop recurrent disease outside the urothelium that have a poor prognosis.
So, I don't think it's surprising in a sense that with our current or our immediate past, present systemic agents that you all examined that we're going to help a small proportion of patients, but the ones who have significant high-risk disease, those are patients obviously we worry about. I think some key messages here that we need to emphasize is almost 20% of these, basically 19% of these patients developed a non-urothelial recurrence, that is a significant percentage. Obviously it's retrospective, it's charred. There are those that we may have salvaged with non-nephrou... Et cetera, et cetera. But clearly these are patients well enough that we consider to do a nephroureterectomy and despite careful attention or treatment, et cetera, that we have a significant percentage. Do you agree with that? That's much higher than I would have passed on to our patients before.
Andrew Katims: Absolutely. It's a high percentage. I think that looking at our data, one thing that we could not control for, just based off the retrospective nature of this, was the role of lymph node dissection. We had a good number of patients recur in the lymph nodes, retroperitoneal lymph nodes, which in theory should be removed. Based off of our data, we had a true reporting of about 10% saying they did not do a lymph node dissection, although there were about 40% of patients included in the study had no information entered. So we presume that they did not have a lymph node dissection. That's about half of the patients, which is significant. But with that being said, still the most common cause of sites of recurrence were multiple at once.
Sam Chang: Multiple sites, yes.
Andrew Katims: Exactly. So, while that lymph node dissection probably helps a subset of people, it still will have a fairly high recurrence rate outside of the urothelium.
Sam Chang: Yeah, I think that it'll be, to me, fascinating as you all collect even more patients in the patient pool regarding that true impact of lymph node dissection. I mean, if you look at what we found with bladder cancer, if you look at what you've seen with esophageal, with gastric, with other, pancreatic cancers, that the lymph node dissection is helpful for staging, it's unclear the therapeutic benefit for sure. But I think definitely for a core of patients you probably provide some significant benefits. So, who those are and when we do it, et cetera well, is another question, which leads me to what's next. I mean, this is a big collaboration, big effort, true centers of excellence, where are you all going to go next as data accumulates?
Andrew Katims: One thing I would like to see from this study, and we kind of alluded to this within the study, is that our guidelines that were recently released by the AUA for upper tract disease, the timing and frequency of surveillance for patients after surgery is defined in there, but the evidence for that is not great. It's based off of expert opinion and I would love to see some more definitive timing and how often we should be scanning, where should we be scanning, how often should we be scoping these patients because it is a big burden on patients obviously, both the scanxiety as they say, as well as coming into the office frequently for cystoscopy. So, from this study, I would like to see some more work come out as to is there a true point when those recurrence risk decreases as we progress outwards and can we really deescalate surveillance of patients at any point or do we need to truly continue surveilling them?
Sam Chang: Yeah, no really good point. Really focusing on surveillance and screening when we really can make an impact and then deescalating when we really perhaps should not be adding onto the burden for these patients for what you're saying, the scanxiety, et cetera, all those things, and the social support costs, the financial cost, et cetera. So, Andrew, thanks so much for spending some time with us. Great and very important work, and we look to future publications from this big consortium.
Andrew Katims: Thank you so much for having me.