Sam Chang: Hi, I'm Sam Chang. I'm a urologist at Vanderbilt University Medical Center, and we are joined by Kieren Lewis, who is and will be a rising star in the field of urology. He is at Cleveland Clinic currently. He was the first author of a urology practice article and with Vanderbilt being one of the centers of excellence in UroToday for upper tract urothelial carcinoma. This was actually a very important and I think interesting article looking at developing a way to predict actually renal function, and this was with a novel way based upon the Cleveland Clinic data, and so Kieren, thank you so much and look forward to your presentation.

Kieran Lewis: Thank you, Dr. Chang and then yeah, thank you for having me here today. I'm really excited to talk about this paper we just published. Okay, hello everybody. Yes, today we'll be discussing the prediction of renal function after nephroureterectomy using a split renal function-based approach and its implications for adjuvant versus neoadjuvant chemotherapy. So in 2022, the results from the POUT trial were published. This was a phase-three open-label RCT assessing the efficacy of a systemic platinum-based chemotherapy following nephroureterectomy in patients with locally advanced upper tract urothelial carcinoma, and the study found that adjuvant gemcitabine platinum chemotherapy improved disease-free survival in patients with locally advanced disease. Following this, the AUA guideline statements for UTUC in 2023 stated that clinicians should offer platinum-based adjuvant chemo to patients with advanced pathological staging with UTUC after nephroureterectomy or ureterectomy. However, there's a caveat to this, and that is that upfront nephroureterectomy renders many patients ineligible for full-dose cisplatin therapy due to concerns regarding nephrotoxicity.

Recognizing this, there's an additional AUA guideline statement that says that clinicians should offer cisplatin-based neoadjuvant chemo to patients undergoing nephroureterectomy with high-risk UTUC, particularly those whose post-op GFR is expected to be less than 60 following surgery. It will say that many clinicians extend to cisplatin eligibility to GFR thresholds closer to the 45, 50 threshold. So the combination of these two guideline statements create sort of a treatment algorithm. In patients with high-risk UTUC, if they're expected to have a GFR less than 50 after surgery, then they should first receive neoadjuvant chemotherapy then followed by nephroureterectomy. However, if they're anticipated to have a GFR greater than 50 after surgery, they can likely safely undergo nephroureterectomy. If they're found to have non-advanced disease, they can then undergo surveillance. However, if they're found to have more advanced pathological staging, then they can then receive an adjuvant chemotherapy. Now, one might ask, what is the benefit of going this route with the upfront nephroureterectomy? Well, for one, overtreatment is avoided in patients with non-muscle-invasive disease and treatment is really only given those who are likely to benefit, and also there's no delay in patients receiving any definitive surgical treatment. And you may also notice that there is an important inflection point in this diagram and that's that it requires us to predict GFR following nephroureterectomy.

So how exactly can we do that? That's what some of my research in the Campbell Group at Cleveland Clinic has focused on. Our research has developed a simple split renal function-based approach to predict new baseline GFR after radical nephrectomy in kidney cancer patients. I'll kind of walk us through this model here. So we have predicted new baseline GFR equals 1.25, which is the average renal functional compensation of the contralateral kidney following nephrectomy. We have preoperative GFR times the split renal function of the contralateral kidney. Now, this can either be measured by nuclear renal scan, which I think most are probably most familiar with or parenchymal volume analysis. Perhaps some are less familiar with that, so I'll walk us through what this means. Now, the central assumption behind this method is that the relative parenchymal volume of each kidney is equal to the relative function of each kidney. So how do we do this? We can measure the ipsilateral volume of the kidney and the ipsilateral tumor volume, and then we can measure the contralateral volume. We can then use some simple calculations to calculate the contralateral split renal function based on parenchymal volumes, and what our prior research has shown is that this parenchymal volume-based approach of split renal function estimation is superior to nuclear renal scan for the accurate prediction of new baseline GFR after radical nephrectomy in kidney cancer patients.

Now, we wanted to ask, can we apply this parenchymal volume-based approach to those undergoing radical nephroureterectomy for upper tract malignancy? And that brings us to the current study. So objective of our current study was to develop a split renal function-based approach for predicting new baseline GFR after nephroureterectomy with the goal of informing optimal timing of systemic chemotherapy in patients with UTUC. We had 352 patients managed with nephroureterectomy between 2013 and 2023 who had available contrast CT imaging, and there was an additional challenge in this population. That's that patients with UTUC have a much higher prevalence of hydronephrosis and infiltrative renal masses compared to radical nephrectomy patients. We hypothesized that this could impact the accuracy of a parenchymal volume-based approach for split renal function estimation, largely because hydronephrosis may distort the parenchymal volume function relationship. So to address this challenge, we modified our methodology by incorporating the degree of parenchymal enhancement on contrast imaging. So our new method, which we named PVA+, takes into account both the relative parenchymal volume and the relative enhancement of each kidney into the split renal function estimation. So you can see here we have an ipsilateral kidney that is obstructed and it's taking up less contrast than the more healthy contralateral kidney.

We can just directly measure the average Hounsfield unit of each kidney and then calculate a relative degree of enhancement between the ipsilateral and contralateral kidney. In this case, that's a 1.6 ratio, and we can directly apply that to our model. What we found was that by incorporating degree of contrast enhancement into our model, PVA+ was actually most accurate for predicting new baseline GFR after a nephroureterectomy outcompeting the parenchymal volume alone approach, the non-split renal function-based approach, and the nuclear renal scan-based approach. The benefit of this PVA+ model was most pronounced in those who had hydronephrosis, and other benefits of this approach is that PVA+ can be done at point-of-care with readily available software in less than five minutes, requiring only routine preoperative imaging to complete the study. So now briefly to finish things off, I just wanted to demonstrate the clinical application of our model. So in this situation, we have a patient with high-risk UTUC. We used some of their preoperative information and the contralateral split renal function to estimate a GFR following nephroureterectomy. Theirs in this case is 55, so that suggests they can safely undergo upfront nephroureterectomy. In this case, we find that they have more advanced pathological staging and their observed GFR was similar to our predicted GFR. So in this circumstance, the patient followed the upfront nephroureterectomy followed by adjuvant chemotherapy pathway, ensuring that only once pathologically confirmed to be locally invasive would they receive chemotherapy, and with that, I'd be happy to take any questions.

Sam Chang: Kieren, great presentation. Really appreciate the careful clinical application of your formula to give us an idea of really who should get perhaps neoadjuvant versus adjuvant. So tell me the nitty-gritty regarding this. Is this something you said it's easy in terms of software? How do you all actually apply that? Did you set the software up, tell your radiologist to actually do this calculation? Is it something you did on your own? Tell us a little bit about that.

Kieran Lewis: Yeah, absolutely. So for the study, myself and other researchers did the measurements and we have a 3D automated software that really does it for us. We can input the images and within very little manual editing with less than five minutes, you would have a split renal function estimation. In actual practice, at our institution at least, you can just submit an order on Epic to the radiologist and then they can measure the parenchymal volumes for you, and then you can do a simple measurement.

Sam Chang: Calculation.

Kieran Lewis: Exactly, absolutely, which is very fantastic and makes the really seamlessly integrated into this clinical pathway.

Sam Chang: Right, and so as the clinicians have this idea regarding renal function, obviously that in addition to tumor characteristics, patient desires, etc., that'll only hopefully better inform our decision-making as we attempt to determine what's better for that particular patient. I think really a big step forward. Kieren, great work that you and the colleagues at the Cleveland Clinic did. We look forward to you guys perhaps using this prospectively and telling us how accurate it's been over time and really look forward to other future contributions that I know you'll be making, and good luck to you always.

Kieran Lewis: Thank you, Dr. Chang. Yeah, thank you for having me here. It's been an honor, thank you.