Longitudinal GFR Trends After Neoadjuvant Chemotherapy Prior to Nephroureterectomy for Upper Tract Urothelial Carcinoma - Surena Matin & Craig Labbate

January 24, 2023

In a discussion led by Sam Chang, Surena Matin, and Craig Labbate present a study on nephroureterectomy's impact on renal function in patients with upper tract urothelial carcinoma. The retrospective review of 152 patients treated with neoadjuvant chemotherapy and nephroureterectomy over 20 years showed an average GFR decline of 22 milliliters per minute, stabilizing one to three months post-surgery. Differences in outcomes based on chemotherapy regimens were noted, with 79% receiving cisplatin-based chemotherapy. An exploratory analysis connected a GFR rise during neoadjuvant treatment with worse overall survival. The conversation also delves into a multi-institutional nomogram project to predict post-nephroureterectomy GFR and emphasizes personalized care, recognizing the need for precision in treatment planning and considerations about patients' eligibility for post-surgery treatments.


Surena Matin, MD, Professor, Department of Urology, Division of Surgery, The University of Texas, MD Anderson Cancer Center

Craig Labbate, MD, Urologist, Fellow, The University of Texas, MD Anderson Cancer Center

Sam S. Chang, MD, MBA, Patricia and Rodes Hart Endowed Chair of Urologic Surgery Professor Department of Urology at Vanderbilt University Medical Center

Read the Full Video Transcript

Sam Chang: Hello, everyone. My name is Sam Chang. I'm a urologist in Nashville, Tennessee, and we have the distinct honor today to be joined by two leading physicians from the MD Anderson Cancer Center. First, someone who needs no introduction, Dr. Surena Matin, who is a professor of urology. He has talked with us before on different aspects of care for those patients with upper tract urothelial carcinoma. We are in addition joined by Dr. Craig Labbate. Craig is currently a fellow at the MD Anderson Cancer Center, and he's going to be sharing some slides looking at the impact that nephroureterectomy has on renal function. And so Craig, I'll turn it over to you.

Craig Labbate: All right. So the title of our manuscript was "The Longitudinal GFR Trends After Neoadjuvant Chemotherapy Prior to Nephroureterectomy for Upper Tract Urothelial Carcinoma." The primary objective of our study was to describe the natural history of renal function in patients treated with chemotherapy and then nephroureterectomy for upper tract tumors. And our secondary objectives in the paper were to evaluate the differences in their post-op renal function, stratify their chemotherapy type, and to explore whether GFR change during neoadjuvant chemotherapy could be a prognostic or predictive biomarker.

Our study design was a single institution retrospective review over the last 20 years. We included 152 patients who had undergone neoadjuvant chemotherapy, followed by nephroureterectomy and had longitudinal renal function outcomes. We included patients who had at least three cycles of cytotoxic chemotherapy. Of those patients, 79% of them had a cisplatin-based chemotherapy regimen.

As you can see by the figure below, we measured GFR with the 2009 CKD-EPI equation, and we took GFRs up to six weeks before the start of their chemotherapy; a second endpoint between chemotherapy and nephroureterectomy; and then their postoperative GFRs were taken at their first post-op visit between one and three months after their surgery, and then within one month of their 12 month follow up. Patients were excluded if they had not received three cycles of chemotherapy, if their chemotherapy was unknown, if their systemic therapy was an immunotherapy, and if they had a solitary kidney and became anephric as a result of their surgery.

Our primary figure is this line graph showing the longitudinal renal function outcomes stratified by the type of chemotherapy they received. In black, patients who had gotten a cisplatin-based chemotherapy had a higher pre-therapy GFR, which declined more rapidly after surgery and had a slightly higher postoperative baseline. And those with a non-cisplatin-based therapy actually showed a slight rise in their GFR during surgery or during the chemotherapy and ended up at a slightly lower GFR nadir after therapy was complete.

Another way to look at the longitudinal renal function is their renal function stratified by their KDIGO chronic kidney disease grouping. So grouping's CKD 1 and 2 are in green, yellow is CKD 3a, orange is CKD 3b, and red is CKD 4. You can see as they go through therapy, a larger percentage of patients end up in CKD3, although only 20% of patients ended up with GFR less than 30, that's CKD 4, after therapy. As part of our secondary objective, we wanted to look at how GFR changed during their chemotherapy. And this is a waterfall plot of all the patients from the study with their GFR change before chemotherapy to after chemotherapy before surgery.

And we can see on the right, there is a subset of patients who have a GFR rise during chemotherapy, and these patients have a larger percentage of those with non-cisplatin-based chemotherapies. And so we looked at this subpopulation with interest to see whether there were any predictors of if you found patients who had a GFR change during chemotherapy, would that change their long-term outcomes or their postoperative course? And so here we show in our exploratory analysis that for patients who had a GFR gain greater than 10% around their chemotherapy, there's a higher percentage of patients who have invasive disease on their final pathology. And it even translates to a worse overall survival over the period after nephroureterectomy. However, we temper our results by looking only at patients who would receive at least three cycles of cisplatin, which we considered to be the appropriate dose of cisplatin.

And we see that the correlation we see in the initial cohort falls out of statistical significance. So as we reach our conclusions, we showed that the GFR declined, on average, 22 milliliters per minute after both neoadjuvant chemotherapy and nephroureterectomy, and that this GFR decline stabilizes at one to three months and stays similar at 12 months after surgery. GFR changes are dependent on the chemotherapy regimen. Patients who receive cisplatin-based chemotherapy, in general, have higher GFRs than those who do not and that their postoperative GFR is significantly higher than the non-cisplatin. However, the clinical significance may be different.

CKD is highly prevalent after completing therapy for upper tract urothelial cancer, showing 85% have CKD 3 after completing surgery and neoadjuvant chemotherapy. And on our exploratory analysis, which showed that the GFR rise during neoadjuvant was associated with invasive tumor staging and worse overall survival, and that this effect was most pronounced in those who did not receive cisplatin-based neoadjuvant chemotherapy.

Sam Chang: Craig, that's great. Let's look at your last point, which to me was the most provocative. And then you tended to temper it as you went through your presentation. Do you really make something out of this kind of transient improvement in GFR being associated with worse disease? You think that was just not necessarily a spurious finding, but one that was just, "Hey, this is something different. We can't really base a lot on this." What is your take on it?

Craig Labbate: I think it's exploratory with a capital E in that I don't know if we can make strong conclusions at this point, just because it seemed like the effect was pronounced in patients who did not receive what has been shown in trials for neoadjuvant chemotherapy and that these patients oftentimes can be transitioned to cisplatin-based chemotherapy as a result of this rise in GFR, which may be the biggest thing to see long term is whether these patients who can then transition to a cisplatin-based regimen if they have the same survival effect as those who started with cisplatin-based therapy.

Sam Chang: So let's bump it up to Surena. Surena, is there currently, I know this was a review of I guess almost a decade, I think of maybe even longer. I can't remember if it was 2010 to 2019 or 2000-2019. So a lot of different chemotherapy regimens. Currently at MD Anderson, and obviously there's a discussion of neoadjuvant versus adjuvant, currently in those patients that you decide to do neoadjuvant therapy outside of any trials that you might have going on, what's the current chemotherapy regimen that you all are advocating in the neoadjuvant setting?

Surena Matin: Yeah, our medical oncologists, I think primarily their primary go-to would be dose-dense MVAC, but obviously it's highly individualized. So in those who maybe can't get all of the other combinations, if they have heart disease and can't get adriamycin, for example, gem-cis would be the other common regimen that's given. Part of what they have done for many years has actually been this, what they would call "nephron-sparing regimen" in those who have poor kidney function to begin with. And so we knew that, going in, that these patients had some improvement of kidney function when they received that regimen. Usually it's GTA, gemcitabine, Taxol, adriamycin.

And so that was an observation that we've made for several years here where we saw that they would have improvement on this and then some of them would be able to switch to cisplatin. What we didn't expect was this association with those patients basically doing more poorly. Now, it'd just be selection. We've known for a long time that those who have worse kidney function have worse outcomes and have worse disease. Now, who knows what the chicken-and-egg situation here is, how they're associated we don't really fully understand, and consequently we don't fully understand this phenomenon, either. But it does open up doors, as Craig said, to explore further to look into it. So that's a long answer to your short question, but yeah-

Sam Chang: It clarifies things I think quite a bit because clearly there could be multiple different reasons why, at the time of nephroureterectomy, you've got worse staging. I think selection, just as you said, made a big difference, but it really does kind of open up the doors of what to evaluate next. The other thing that really strikes me is your very first point here. In patients that have received neoadjuvant chemotherapy, are these patients that we think are fit enough to get chemotherapy or we've chosen? That decline of GFR of 22 milliliters per minute is significant, really makes me consider our patients post-nephroureterectomy.

I know as surgeons we're always concerned about after the nephroureterectomy, what can we give our patients in the adjuvant setting? And clearly the impact prior to nephroureterectomy is quite significant. I guess when you guys look at your huge experience and huge database, in those patients, and this is off this study, but in those patients that have gotten adjuvant therapy or have been able to get adjuvant therapy, I guess that's the first question is how many just roughly you think, either Craig or Surena, regarding those patients that underwent nephroureterectomy that you would then say, "We're no longer platinum eligible?" Is it half, is it 25%? Obviously it's a huge and widely heterogeneous cohort, but roughly what do you think in terms of the impact of the nephroureterectomy on your average patient total population?

Surena Matin: Yeah, it has a huge impact, of course. Quite honestly, Sam, I was surprised by this 22% average decline. I thought it was going to be more. Now, of course, what we don't have here is the range or the confidence intervals, and those can be quite significant in some patients. And they burn a few nephrons with cisplatin sometimes, as well. And so in some ways, I'm happily surprised that it's only 22.

Sam Chang: Yes. Yeah, understood. Understood. Yeah.

Surena Matin: But that range can be significant. But yeah, for sure. Now I will say that our experience here is a little bit different because one of the messages we're trying to get out, and this project is part of this message, which is that as urologists we need to be a little bit more long-sighted. So when we first see a patient with upper tract disease, right then is the time to think about their post-nephro-u kidney function, not after the nephro-u. And so for me, part of my decision making and part of what I've been teaching and advocating with our group is that we want to risk stratify and at the same time evaluate their kidney function upfront and try to do the best we can estimating their function post-nephro-u, and then based on that information, make decisions on neoadjuvant versus adjuvant.

So the other thing is, I know there's the people are interested in pitting the two against one another, neoadjuvant and adjuvant. I don't see it that way. We know perioperative chemotherapy is effective, and so it really comes down to how do we deliver therapy more precisely? Right? And, of course, we can be most precise in the adjuvant setting when we have pathology, but we can't be that good about it if we're eliminating 80% of patients from getting that effective therapy. And so this has been my approach now. And so I think by using nomograms and estimating GFR post-op, and one of the other papers that we are in the process of preparing for submission now is actually a nomogram for upper tract patients. It's a multi-institutional project. We have over a thousand patients, and we've developed a nomogram to be able to predict what their post-nephro-u GFR is going to be. And again, calculating that upfront and based on that and the risk stratification of making decisions about what we're going to recommend more strongly.

Sam Chang: I think those results are going to be incredibly important and impactful. One of the things that I always hesitate when I see the European and the NCCN guidelines, to your point regarding risk of disease and low-risk versus high-risk, and whether you try to do nephron-sparing versus a nephroureterectomy, it doesn't really take into account the patient and the patient's actual nephron capability the impact of the different treatments. And so I really look forward to that because being able to, just as you say, understand upfront what patients may or may not be able to receive either pre-nephroureterectomy and afterward really makes a difference. So I think that'll be very, very impactful. While I've got you for the last couple of minutes, and it'll be a teaser for the nomogram, can you give us some ideas of what the ingredients are into that nomogram? Do you put in differential function, do you put in proteinuria? Just give a little tease for the audience of the factors that you guys take into account.

Surena Matin: Yeah, for sure. Craig, I'm going to turn to you in just a minute because we just looked at the abstract that we're submitting and so he can give you the four variables. But the other thing I will say that I've started doing more often is basically asking our radiologists to do CT volumetric studies. There's a really nice paper from Steve Gamble and the group at Cleveland Clinic that showed that that's a much better study than a MAG3 or DTPA nuclear renogram. And quite honestly, it's been a little frustrating working with that group as well because sometimes you just get weird results that you know for sure are not quite on the money, and it's been hard to understand why that happens. And so I think, as urologists, we're just so much better understanding of CT scans and of parenchyma. So I've actually been heading in that direction to estimate differential function, and I'll consider that.

Sam Chang: I love that. I'm glad, I always appreciate your honesty, Surena, because I've, similarly... We've had trouble in terms of delaying getting the imaging, but then secondly, interpretation of some of the findings are so off either with a DMSA, a DTPA, or even a simple MAG3 renogram in terms of function and what they're looking at and the curves that they draw. So we turn it over to you, Craig, in terms of kind of, you don't have to give us all variables, but give us a few that we should be looking at.

Craig Labbate: Yeah, so I think the primary variable is their preoperative GFR, so good GFR leads to better GFR afterward. There are some demographic variables like their age and diabetes status. And an interesting one is the volume of tumor they have as being a predictor of worse GFR afterward. So those are the main ones. And then some interesting negative findings where the proteinuria aspect did not pan out in our univariate analysis... And hydronephrosis, too, some of the things we would expect do not have the same statistical significance as others.

Surena Matin: How many clinical variables did we evaluate? Do you recall, roughly?

Craig Labbate: It was roughly 20 total variables including demographics, tumor, as well as some more renal-specific, like if they had had a renogram or proteinuria.

Surena Matin: Very good.

Sam Chang: Now I really look forward to that paper and, once it comes out, I look forward to inviting you guys again because I think that will really help us kind of formulate, just as Surena was mentioning, a plan that makes sense. I love your comment regarding it's not pitting against each other, it's personalizing and individualizing the optimal care for that patient, their disease process, their medical comorbidities. And so I think that would be immensely important as we take more and more care of these patients, as we have more multidisciplinary care that impacts their overall kind of disease response and therapeutic options.

So, Surena, as always, fantastic. Craig, it was great to be able to spend some time with you. And we look forward to your future publications, and good luck with the rest of your fellowship. And thanks again for spending some time with us. Thank you both.

Surena Matin: Thank you.

Craig Labbate: Thank you, Sam, very much.