Vidit Sharma: Yeah, I appreciate the introduction and I appreciate your interest in our study. My name is Vidit Sharma. Like you said, I'm a urologic oncologist at Mayo Clinic in Rochester. And we recently published this study titled The Development and External Validation of a Local Pelvic Recurrence Risk Score After Radical Cystectomy: Identifying the Ideal Candidate for Adjuvant Radiation Clinical Trials. This was published in the Gold Journal in 2025 and it was led by our first author, Dr. Matthew Lee. He's a PGY-4 stellar urology resident going into urologic oncology, so I just wanted to give him the credit he deserves for this great work. So really, I'll start with the introduction. Local recurrence is defined as a recurrence in their cystectomy bed, pelvic lymph nodes, or the pelvic soft tissue such as a vaginal cuff. And in studies, it's actually recently common, 15 to 20% in the largest series, with an overall survival, historically, that's been relatively poor after local recurrence of eight to nine months. Adjuvant systemic therapy really has not been shown to reliably prevent local recurrence, even though it does reduce the risk of metastatic disease. There have been historic randomized clinical trials that have looked at adjuvant radiation, but these have been largely in the squamous cell carcinoma population, and until recently, there were weak retrospective studies or smaller trials in the urothelial carcinoma population.
So the main recent study that resulted was actually the BART study, which was bladder adjuvant radiotherapy study. And in this study, they randomized patients in India to adjuvant radiation versus observation with a relatively broad inclusion criteria. So really any patient with T3 or higher disease, node-positive disease, positive margins, less than 10 nodes dissected, or those patients who were T3 before neoadjuvant chemotherapy. And in this population, they had about 150 patients randomized. Adjuvant radiation actually prevented local recurrence in about 50% of patients. So the effect size was about 50%. So it reduced the local recurrence risk from 24% to roughly 13% with a hazard ratio of 0.43, which is relatively impressive results for this endpoint. So our objective was to create a local recurrence risk score to stratify patients in terms of the local recurrence risk and really create a usable, simple, user-friendly score that doesn't need to be memorized and can be used in clinical practice. To do that, we queried our prospectively maintained institutional radical cystectomy registry from years 1988 to 2022. We identified adults undergoing cystectomy for urothelial carcinoma, excluded patients with prior metastatic disease, pure variant histology, no follow-up imaging, and excluded patients in which the local recurrence variable was not populated by our registry.
The primary outcome was time to local recurrence and we used univariable and multivariable Cox proportional hazard models to create a local recurrence risk score. And then we actually externally validated that risk score from Milan, courtesy of Dr. Marco Moschini from there. So here is our baseline cohort characteristics. We had approximately 1,200 patients, and of these patients, 227 developed a local recurrence. About 42% of our patients were T3 or higher disease. 21% were node-positive disease and the median follow-up after cystectomy was 4.4 years, so relatively standard high-risk cystectomy cohort. And then if we look at our local recurrence-free survival, the median time to local recurrence was 11 months. The two-year risk of local recurrence was 17%. The five-year risk was 22%. So really the majority of local recurrences are occurring in that first two-year window. Very small percentage are occurring after two years. And notably, about half of patients with local recurrence had a distant metastasis, either concurrently diagnosed or diagnosed prior to the local recurrence. When we looked at factors associated with local recurrence in our Cox regression model, we present the univariable and multivariable analyses here. And on the right, the multivariable analyses are shown and any variable with a P value of less than 0.1 is highlighted. These variables were then put into a local recurrence risk score to create our final score.
But notably, one variable I'd like to highlight is that patients who had neoadjuvant or adjuvant chemotherapy actually were not found to have a significantly lower risk of local recurrence in our multivariable model. So when we take these five variables and we put them into a separate regression model, we can then come up with regression coefficients to then calculate our risk score. We basically chose a simple intuitive integer-based risk score so you don't have to memorize nomograms. And basically you get one point if you had T2 disease, two points if you had T3 disease, three points if you had T4 disease, and then one point for either node positivity, positive ureteral, urethral margins, positive radial margins, or if you had less than 20 lymph nodes dissected, you also get one point. So this is an important finding of our study that removing less than 20 nodes actually does increase your risk of local recurrence in our data. When we looked at how well this score stratified patients as far as the risk of local recurrence, the data is shown here. Our C index was 0.73, which is reasonable, and patients with zero points had a 2% five-year risk of local recurrence, whereas patients with three points had 33% five-year risk and four or more points had a 52% five-year risk. And notably, there's a reasonable number of patients who fall into the three-point category, so 22.3% of patients, and actually 12% of patients approximately, fall into four or more points. So these aren't rare cohorts, they're actually a substantial number of patients. And then we externally validated this with data from Milan.
We had a cohort of 614 patients and this similarly risk-stratified patients with a C index of 0.75, very similar to our development cohort. So here is our conclusions. We created a simple, usable, relatively intuitive risk score with reasonable C indices, one point per criteria. And we did this because we want to try to identify the ideal patient for adjuvant radiation therapy. If we radiate everyone after surgery, we would have unnecessary toxicity and cost and treatment burden, but if we radiate no one, then we may not actually prevent these local recurrences. And so if we use the BART criteria, actually about 70% of our patients would have met criteria of adjuvant radiation, which many would think is too much. So using our adjuvant radiation selection local recurrence risk score, if we use a threshold of three points, that would apply to 33% of our cohort and these patients would normally without radiation have a roughly 33% risk of local recurrence at five years. And then using the BART trial hazard ratio of 0.4, this would translate to a number needed to treat of about five patients to prevent one recurrence. If we use a four-point threshold, this would apply to 12% of our cystectomy cohort, and using the BART trial hazard ratio, again, this would translate to a number needed to treat about three patients.
So these numbers may be more favorable, because the risk of GI toxicity after adjuvant radiation is still relatively understudied in bladder cancer, but there are data emerging and that grade three or higher events are roughly in the 10% range. So for these patients, at least the benefits are higher than those risks. And then furthermore, I do think this is a, I think, easy opportunity to identify patients for adjuvant radiation plus immunotherapy trials because patients may benefit from a dual-modality approach in the adjuvant setting. So thank you very much for your attention. I'll be happy to take any questions.
Ashish Kamat: So thanks very much. It was a nice, succinct presentation. The BART study that Moschini led at Memorial obviously show the benefit, but also one thing that it opened everybody's eyes to is the low toxicity, because people, and we've used radiation therapy in the bladder paradigm in the adjuvant setting, in the neoadjuvant setting over the years, but the toxicity issue is something that people worry about, especially after radical cystectomy because there's nothing there, there's just bowel. And hats off to his group for actually showing that the toxicity is very manageable. And then efforts such as yours, I think, are very important. Because it helps focus people's attention on the fact that local toxicity does have a treatment beyond trying to treat with systemic therapies. Now, one thing that struck me a little bit, and I'd love to hear your comments on this since you have the data and deeper insight into it, is the fact that the margin status had a score of one, which was less than the score for T3 with a negative margin. So comment a little bit upon why you think that is, because to me that goes a little bit against what I've always believed, that the positive margin is really more of a predictor of local recurrence in T3 disease with a negative margin.
Vidit Sharma: Yeah, I think that's an interesting observation. I think we generally think of positive radial margins and cystectomy as a very poor prognostic sign. And the T3 disease, at least our data shows, may have just as high of a risk of local tumor shed beyond the fat that we're not appreciating with perhaps micrometastatic regional deposits in that area, and the positive margins behave similarly in our data. And it's hard to know exactly why that is, but to me it shows that there's probably room to substratify T3 disease. So there's probably very low depth of penetration T3A disease that perhaps is not as high of a risk of local recurrence, but this may actually be driven by the really bulky T3 disease, with large extravesical extent that is driving this. We didn't have really the imaging segmentation of these tumors to classify that, but that would be a pretty interesting project to go and look and really individualize the risk of local recurrence based on imaging.
Ashish Kamat: Yeah, I think that'd be important, especially in today's day and age where we have pCR rates that are approaching 55, 60% with new adjuvant to EV pembro, and of course, the Niagara protocol. So clearly less patients are having bulky disease left behind on this specimen. And in that situation, I guess a T3 is completely different from the historic T3s and a positive margin might be even worse, right? Comment a little bit again on your finding that node-positive disease was as much a predictor of local recurrence as systemic recurrence, or did I misread your slide?
Vidit Sharma: Node-positive disease was a predictor of local recurrence and it's also a known predictor of distant recurrence. The reason, one of the reasons why node-positive was a predictor of local recurrence was because the definition of local recurrence included a regional lymph node recurrence, so that is probably the biggest driver of that. It's hard to know if node-positive disease predicted soft tissue deposits because the event rate drops, obviously.
Ashish Kamat: Okay. Okay, great, great. Thanks for clarifying on that, because I misread, I thought what you meant was a node-positive disease also predicted for non-nodal pelvic recurrence, which would be a different beast, which you said makes perfect sense. Overall, are you now at your center incorporating this into your recommendation for adjuvant radiation? How are you treating the data that you generated?
Vidit Sharma: Yeah, so we're basically using this data to basically create a working group with radiation oncology. Specifically, we have a radiation oncologist, Dr. Brian Traughber, who's interested in adjuvant radiation studies and our bladder cancer cystectomists to create a protocol of three or more, four or more points in our score, very reasonable to use adjuvant radiation or at least have a discussion with the radiation oncologist. So I think this data really shows that A, it's a real problem for a subset of patients, so perhaps brings attention to that problem, and B, now that there is encouraging data coming out that this local recurrence can be prevented with relatively low toxicity, like you said, we are trying to create a standardized pathway for adjuvant radiation. And we also are trying to use this data to approach pharmaceutical companies to see if they'd be interested in running a trial combining novel systemic therapies with adjuvant radiation.
Ashish Kamat: Great. Thank you so much, Dr. Sharma, for taking the time and congratulations.
Vidit Sharma: Yeah, thank you.