Neoadjuvant Vs. Adjuvant Chemotherapy in Upper Tract Urothelial Carcinoma - Alison Birtle & Surena Matin
Surena F. Matin, MD, Professor, Department of Urology, Division of Surgery, The University of Texas MD Anderson Cancer Center, Houston, Texas
Alison J. Birtle, MD, MBBS, MRCP, FRCR, BSc, Consultant Clinical Oncologist, The Rosemere Cancer Centre, Preston, Lancashire Teaching Hospitals, NHS Foundation, UK
Ashish Kamat, MD, MBBS Professor of Urology and Wayne B. Duddleston Professor of Cancer Research at MD Anderson Cancer Center in Houston, Texas.
Watch: The OLYMPUS Study: Mitomycin Gel for Low-grade Upper Tract Urothelial Cancer - Seth Lerner
Watch: Moving the Field Forward, Upper Tract Urothelial Carcinoma, The POUT Study - Alison Birtle
Read: ASCO GU 2019: Challenges of Therapy in Upper-Tract Urothelial Carcinoma
Read: ASCO GU 2020: The Efficacy of Two Cycles of Neoadjuvant Chemotherapy for Upper Tract Urothelial Carcinoma Patients.
Read: CUOS 2019: High-grade Upper Tract Urothelial Carcinoma – Neoadjuvant Chemotherapy
Ashish Kamat: Hello and welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat, Professor of Urology and Cancer Research at MD Anderson Cancer Center in Houston. I'm joined today by two of my friends and colleagues who are true experts in the field, Dr. Alison Birtle who is Consultant Clinical Oncologist and Senior Lecturer at the Rosemere Cancer Centre in Preston, UK and is Chief Investigator of the POUT study that we've been hearing so much about, and Dr. Surena Matin, who is right here in Houston at MD Anderson Cancer Center. He is Professor of Urologic Oncology, heads our minimally invasive process here at MD Anderson, and is a true expert in upper tract urothelial carcinoma.
Thank you both for taking the time to speak with us today. For today's educational session, each one of you has been given one viewpoint of a commonly debated clinical scenario. In this case, "Should we use neoadjuvant or adjuvant therapy when it comes to upper tract urothelial carcinoma?" You will present your arguments to support this viewpoint and we'll then follow it with a discussion, which I'm sure will be very lively. Alison, you have the stage.
Alison Birtle: Thank you very much for that. I think adjuvant chemotherapy should be the standard of care for many patients with upper tract urothelial carcinomas. These are my disclosures.
The evidence that we have before the POUT trial came along was absolutely minimal. What we had were retrospective studies that were all really small and with huge confidence intervals, or we had collaborations across the world and none of them were particularly helpful. Then we had two papers two years ago. One of them said adjuvant chemotherapy, it's no good. The other one said, oh yes it is. Where were we? We know that transitional cell carcinoma, it's pretty similar in upper tract urothelial carcinoma as in bladder cancer, although with some differences. We know that in bladder cancer, cisplatin-based combination chemotherapy given neoadjuvant does make a difference, but this isn't bladder cancer; it's upper tract and it's difficult to get accurate preoperative staging that's absolutely definitive.
If you give neoadjuvant chemotherapy in this setting, it can result in over-treatments. A nephroureterectomy is much less morbid than a cystectomy, and adjuvant studies with bladder cancer have failed to recruit because of the morbidity of cystectomy.
We hear this the whole time; two kidneys must be better than one for giving platinum-based chemotherapy, and that largely seems to be based on this particular study from Shahrokh Shariat and his group. This looks at patients with the Galsky criteria of a GFR of greater than 60. In real life, we all know that we give cisplatin down to GFRs of 50 milligrams per meter squared and we might even use it down to 45 and give fractionated cisplatin. We can certainly use carboplatin for patients with a GFR between 30 and 45.
Also, we mentioned definitive histology. I've got two studies here that in these studies patients were thought to have bond or upper tract tumors based on imaging. In this particular one, 13% didn't have TCC, and in this one, which is Tim O'Brien's automate C study, 10% of patients either had no cancer or they had non-TCC pathology. Can we say even in the best of hands that we have enough evidence to say even with cytology and hydronephrosis that these are definitely locally advanced disease? Because the prognosis for pT1 tumors is so much better and do we really want to over-treat that group of patients? Also, we want to have studies that patients and clinicians want to do, and before we did POUT, we asked the experts. We asked the patients from focus groups and we asked all the clinicians across the UK and they unanimously supported the idea of an adjuvant, not a neoadjuvant study.
That's why we went with this. We've seen these slides before and these are the pivotal studies through POUT, a Phase III randomized study. This was a pragmatic study because UTUC is a low incidence study so you need to be all-encompassing. You can see what the pathology was and these patients were then randomized with adverse pathology on a one-to-one basis between surveillance or platinum-based chemotherapy based on GFR. The choice of chemotherapy was just based on GFR. It was carboplatin and gemcitabine if your GFR was less than 50. If it was over 50, gemcitabine and cisplatin. The study was actually really ambitious. We were looking for a 15% difference between the two groups and three-year disease-free survival. The important thing is we had, as with every study, independent stopping rules, and these were met at two years, not at three.
This is the concert diagram, 260 patients. You'll see there was great compliance with treatments and these are the preplanned pre-specified subgroups, which is really important, these were not ad hoc subgroups. And you can see from this that the majority of the patients had T3 disease and two-thirds of patients were randomized to receive cisplatin rather than carboplatin-based chemotherapy. In both groups of patients, whether it's cisplatin or carboplatin, 71% of patients got all four cycles of treatment with a small number switching between cisplatin/carboplatin during treatments.
This is a chemotherapy study, so there were chemotherapy adverse events. None of them are unexpected and it was safe to give. These are the money slides. These are updated based on the publication from the Lancet published last month and you'll see that at three years there was a 25% difference between the two groups in terms of disease-free survival and the median disease-free survival was about 29 months for the surveillance arm versus not reached in the chemotherapy arm with a whopping hazard ratio of 0.45.
In those pre-specified preplanned subgroups, it was the same survival advantage. We did a test for interaction that tells you that the primary endpoint is the same across all of those subgroups. They're respective of stage, nodal status, what type of chemotherapy. Note the numbers are smaller, so the confidence intervals are wider. Metastasis-free survival, again, a really significant improvement of 17%. We wanted to have good quality of life. It dips a bit while you're on treatment, but then it goes back to normal afterward. This is the first quality of life data as well. We will get overall survival data. I'll present that when it's mature.
In summary, Phase III 261 patients large study, first RCT in this area. It is the standard to beat and I'm just so grateful to every person that got behind the trial and to our funders. Thank you.
Ashish Kamat: That was excellent, Alison. Once again, congratulations on getting such an ambitious and difficult trial completed. Surena, you have the stage now.
Surena Matin: Well, thank you so much for this invitation to speak about neoadjuvant and as well discuss the benefit of adjuvant chemotherapy. These are my disclosure slides and I was really glad to hear Dr. Birtle talk about the role of kidney function and as well the difficulty in assessing the risk of disease because these are really fundamental issues to discuss when we talk about neoadjuvant or adjuvant treatment of upper tract urothelial carcinoma. As we've heard, there is a significant challenge with these patients. Many of them present with reduced kidney function and after nephroureterectomy, if it's done reflexively, then the majority of these patients can be precluded from getting adequate amounts of cisplatin chemotherapy. There are substantial data that have been presented in this regard. Note that with the POUT trial, patients who had a GFR less than 30 after surgery were excluded and we really don't know what proportion these were of the total population.
In regard to risk stratification, one of the major challenges we have is that there's a big gap in regard to knowing whether the disease is muscle-invasive or more than muscle-invasive by nature of the limitations that we have with ureteroscopy, which just provides a mucosal biopsy and imaging, which may not even show fat invasion. We've recognized that for a long time and so clinically many of us were interested in looking at risk stratification rather than staging because we think it can be more accurate. One of the first nomograms, actually we've evolved from an initial one that was published by Vitaly Margulis was this one in trying to predict non-organ confined disease. Out of 23 relevant variables that we looked at, all available clinically, these are the four that were relevant.
Now Vitaly separately wanted to also look at predicting the risk of recurrence preoperatively and basically what we were able to identify were seven independently predictive factors of recurrence and you can see with this nomogram that there are seven of these factors and you can calculate the two-year relapse-free probability as well as the five-year. We have more objective methods of assessing risk now provided that a biopsy is done, which most of us in the field would argue should be done in most cases.
Now several years ago we presented the first systematic evaluation of neoadjuvant chemotherapy based on our institutional effort to systematically approach this, and what we were able to show was a significant improvement in overall survival as well as disease-specific survival. You can see that neoadjuvant chemotherapy was a very strong factor in regard to survival with a contribution with lymph nodes as well. Now we were very interested in getting follow-up data for this and this has been submitted for publication. We have experience with 126 patients and 10-year follow-up and basically the numbers in regard to disease-specific survival, metastasis-free survival all hold up. You can see that at 10 years overall survival drops off significantly in large part contributed by the fact that these are older patients, to begin with.
If we look at some of the data that's out there, both retrospectively and prospectively, and I'll just highlight the three prospective trials in this space, one was the single-arm Phase II intergroup trial. The other one was a study by Memorial Sloan Kettering and what you can see is that these show significant rates of complete remission as well as downstaging, which is consistent with our original publication and as well our subsequent publication that's coming up the ranks. Here is the data from the POUT trial, which also obviously very favorable. What I think we can take away from all this data is that look, chemotherapy works. We have shown this retrospectively, we are showing this prospectively, although not at the level of evidence we would like yet with neoadjuvant data. Adjuvant therapy, based on the data from the POUT trial, also shows that chemotherapy works for this disease.
For sure, it is a standard of care. There is no question that it sets the bar and it was an extremely well-done study, but I think what we need to do is have a nuanced approach where we think about what the postoperative GFR is going to be before we do surgery. We've talked about looking at the GFR before surgery, but really what we need to think about is what do we think the GFR is going to be afterward? If you have reasonable expectation that the GFR is going to be less than 50 and it's a high-risk patient, then that patient should really be considered for neoadjuvant chemotherapy before they go on to surgery. On the other hand, if there's reasonable expectation that the patient is going to have very good kidney function after surgery, by all means, we have very solid confident data that this patient could undergo initial surgery and subsequently based on that pathologic outcomes and risk stratification that way, go on to receive adjuvant. By the way, there's five different nomograms available that can help us assess that risk if by chance we're in a quandary about whether to offer adjuvant to that patient or not.
Now the advantages of the neoadjuvant approach are many. It's given in the presence of maximal renal function, surgical complications or delayed recovery aren't going to interfere with chemotherapy administration. There was a 5% dropout in POUT, and only 75% received all four cycles. We do tend to see better outcomes with surgery, with initial chemotherapy, and the other very interesting factor though that we're running into with these over a hundred patients also that we treated is that there are some that are non-responsive. In fact, what we take away from that is that we recognize immediately that they are, and at the first sign of recurrence, they are eligible for immunotherapy based on the fact that they were not responsive to cisplatin chemotherapy.
My argument really is that these are not mutually exclusive concepts. It just requires a little bit more of a nuanced approach and to consider neoadjuvant when you think the patient's kidney function is not going to be up to par after surgery. Thank you.
Ashish Kamat: Thank you, Surena. That was excellent as well. If I could turn the stage back to both of you and ask you, Alison, what is your sense about the paradigm proposed by Dr. Matin about using the predicted GFR to select patients who should not be considered for adjuvant chemotherapy?
Alison Birtle: I think it's interesting, but you were saying with the data that we had, not knowing the patients that were excluded from POUT because of a GFR of less than 30, we kept really detailed screening logs. We just haven't published the whole of the screening log data yet. In each, because it's a low incidence tumor, every patient that was suitable for POUT at one of the network MDTs was then recorded, and the reasons for them not subsequently going in was recorded. We will have that data. The main reason people didn't go into POUT was that either they didn't want chemotherapy because at the time surveillance was an acceptable standard of care, but also the majority of them that didn't go in was because they had pT1 disease and these were people that were thought off their imaging, preoperatively, and or biopsies and or cytology, to have pT2 disease or above.
That was the main reason, which is the whole concern about overtreatment with neoadjuvant chemotherapy. We know from the subgroup analysis, yes, the competency was a wiser, but we've got no reason to suppose in this particular setting that carboplatin doesn't work. We've got less patient numbers in that, which is why the confidence intervals are wider. That's why we did a preplan subgroup analysis. I guess the other thing to say is that we are looking at analyzing the translational data, so we've got obviously the nephro use specimens, we've got blood and we've got urine on patients who were in the main study, so we will be able to get more translational, hopefully, predictive and prognostic information from that to look at the patients that got chemotherapy that we could quickly, or those that were in the surveillance arm that didn't recur at all. Then we can tease out those patients.
As you were saying, you might want to give immunotherapy too in the future, but why wait until they recur? If we're able to predict this and get away without treating some patients and also offering IO therapy to others rather than chemotherapy, I think that the translational data from POUT is going to be just as important as the primary endpoints.
Ashish Kamat: Thank you. Surena, I know one of the points that is often raised, and Alison brought this up as well, is that the patients with upper tract disease are very hard to stage accurately clinically, and you've done a lot of work to emphasize to the community that the staging is very, very important. It has to be done certain ways. Can you share with our audience some of the pearls on how to properly stage these patients and actually plug them into the nomograms?
Surena Matin: Sure. Yeah. Listen, Dr. Birtle pointed out one of the fallibilities of urologic practice, which still happens, which is that people look at imaging, they see a filling defect and then they take the patient to surgery. Of course, you can't predict what stage that is. It's not like you can see the muscle on a CT scan. You can't even see the muscle there that well on a CT for the bladder. Nevermind the microns, the muscle layer of the ureter, or the renal pelvis. This idea of staging really is fallible for upper tract disease, but it can play a part in the risk stratification. We do rely on the ureteroscopic biopsy and again, many of us more so over time, have come to regard that step as essential. There can be some cases maybe with a nonfunctioning kidney where you think you can skip it, but for the most part, we do want to biopsy because it does provide critically important information that is predictive; the biopsy grade.
Now in the past when we wanted to risk stratify, we just looked at the grade and we said, oh, it's high grade. There's a two-thirds chance that it's invasive, so let's give chemo. Dr. Birtle is absolutely right. Even if you biopsy and you just do the high grade, you're going to over-treat roughly about 30% of patients. We wanted to do better than that. It turns out the tumor architecture is one of the most reliable findings as well. When you're looking at this on ureteroscopy and it's not a papillary tumor, but a sessile tumor, that architectural presentation is predictive even postoperatively when you have the pathology available. That architecture plays a very important independent role in prediction.
In terms of systemic factors, when we plugged in a bunch of different lab variables that people have published on, nuclear lymphocyte ratio, albumin, sodium levels, all those things, all of those tend to drop out. The single strongest predictor turns out is actually hemoglobin levels and probably it's related to maybe insidiously more advanced disease than we can detect. GFR somehow seems to play a role as well. In one of the nomograms, GFR got kicked out and was supplanted by the other variables, but in the other variable that predicts recurrence, GFR actually plays a role. Again, there are multiple other studies that have shown that GFR is an independent predictor. Somehow there is a relationship between the disease and kidney function and it's something that's kind of interesting. If you look at the aristolochic acid exposed patients who do have nephropathy because of the Aristolochia, that seems to be a common element there. I'm not suggesting it's the same thing, but there is something to be said about this disease and kidney function that seemed to also be interrelated.
In regard to your question, another long-winded answer, basically it's that the concept of staging it is absolutely not reliable, but it should play a part in our risk stratification. We keep saying that maybe the molecular markers will help us risk stratify better, but in truth, that has not come up yet. It is something that we're all still interested in. Look, these nomograms have accuracies in the upper 70's, lower 80 percentiles, which is pretty good. If you look at the pathologic nomograms, those are also in the upper 70, 80 percentile range. They're not perfect, but they're somewhat better than what we can do based on biopsy alone or staging alone or imaging alone.
Alison Birtle: When you collect the data, we're looking back now at the preoperative creatinine and GFR's so that we can see what percentage drop there was and what percentage of patients who went into the carbo or might've been eligible for cisplatin, et cetera, and we've also got the data on the neutrophil-lymphocyte ratio. I wanted to ask you, do you think we're missing a trick? Do you think we should be doing more MRIs on these patients?
Surena Matin: I don't know. The problem with MRI is that it lacks the spatial resolution of CT, which is really what we need. On the other hand, MR has a really good contrast resolution. I know people have looked at this. There was one study from Japan that looked at the various different parametrics of MRI. I don't think that's gone anywhere, and I think the major problem we have, again, is the lack of spatial resolution that goes down to that sub-millimeter level essentially, or at least even millimeter level. Yeah. It's hard to know. There's a couple of different technologies that we've evaluated endoscopically, intraluminal ultrasound. There's confocal laser endomicroscopy, which is basically like an insight to biopsy so it doesn't tell you about invasion necessarily. There's another group that's looking at optical coherence tomography. That seemed promising, but again with bulky tumors, it just doesn't have great penetration.
We're very limited by technology and I'm not sure exactly what the next incremental advance is going to be for us in terms of doing a better job to identify the really bad players. Translationally, we are finding out that in terms of the basal and the luminal components, those do seem to have very different outcomes. With the luminal and upper tract, mostly is luminal, but we do see a significant subset that are basal. Particularly with the p53 mutations, those are pretty bad players and we see many of those being non-responders to neoadjuvant chemo. Absolutely. I think in those patients, immunotherapy is probably the way to go. In the US, we can't use it as adjuvant. Not approved for that. That's why I said in terms of waiting for recurrence, but absolutely. I think in regard to further avenues of adjunct therapy, CPIs, potentially in the adjuvant setting, in these types of patients is going to, is going to be very important.
Alison Birtle: I think when we look at the other adjuvant studies that have been out there when we look at the pembro and the atezolizumab loosen up nivo studies, upper tract tumors were allowed in, but they were added on at the end when they weren't recruiting for the cystectomy cohort. It's not going to be powered enough to be able to look at the upper tract tumors there because it was no more than 10% of the overall cohort. It's going to be underpowered really. Obviously IO is more interesting because increased MSI and things in upper tract tumors and FGFR receptor inhibitors are also going to be equally interesting in the group because of the increased number of mutations in UTUC.
Surena Matin: Absolutely. It should be very interesting. We do get concerned that the luminal subtype may be somewhat immunologically cold, but then again after chemotherapy, probably the bets are off. Who knows. I think that's all very interesting, but I have full faith that if anybody's going to be able to pull off that trial, it's going to be the United Kingdom and you all. There's about five randomized control trials in upper tract cancer and I think at least two of them are from the UK. Congratulations to you all for having done that. I do think you deserve a lot of credit for having been able to design it and carry it off.
Alison Birtle: Thank you. Obviously what we really need is when we're looking for POUT II, which is going to be a really imaginatively titled POUT II, is we'd love you to get on board with that.
Surena Matin: Absolutely. We'd love to.
Ashish Kamat: This has really been a great discussion and I've enjoyed listening to the two of you as well. You answered a lot of the questions that we had. In closing, if you would each put forward how you counsel patients that present to you with high-grade invasive looking tumors that present to your clinic. Maybe, Dr. Matin, I'll have you go first. How would you counsel your patient? What should we be telling them?
Surena Matin: First I want to make absolutely clear that there's no concern for lymph node disease and that patient would definitely go to chemotherapy first, but for the patient who appears there's concern for locally invasive disease, I plug their numbers into the nomogram, I look at their GFR, look at the contralateral kidney. If I'm unsure about my cognitive guess as to what their postop GFR is going to be, I'll get a nuclear renal-gram to be more objective about it. Then based on that data, just like I showed you, basically, if it looks like they're going to have a great GFR afterward, I give them the option of neoadjuvant versus adjuvant. Many patients, in that setting, given the choice, will want to try to go for the adjuvant because potentially it may help them avoid chemo. Although I think generally we've gotten pretty good at this point that we can predict it pretty good. For that patient whose GFR is going to be compromised, I make a very strong recommendation for neoadjuvant.
Ashish Kamat: Great, and Dr. Birtle?
Alison Birtle: I guess I do things a bit differently in that we tell patients now that... I mean obviously it needs to look operable. We're not talking about downstaging chemotherapy for inoperable disease because people often get confused between neoadjuvant treatment and downstaging treatment both in bladder cancer and in UTUC, which is an operable disease. I say to patients that at the moment, yes, there's enthusiasm around neoadjuvant treatments and obviously peri-operative chemotherapy is what we're really looking at, but the Phase III data at the moment, so the strongest data that we've got is, with adjuvant treatment. If their GFR looks reasonable and then we can predict that it's going to be in an area where we'll be able to safely deliver either cisplatin or carboplatin postoperatively, then we go with that.
At least now we've got information on chemotherapy, which means that we have got a standard of care because beforehand we did a survey across the UK and there was no standard of care and I'm sure it was the same elsewhere. People say, well you can have chemotherapy if you want it or you can have surveillance if you want it, but we don't know if it's a good thing. Whereas now for all of us, we know that we've got strong evidence for chemotherapy. It means that we'll be recruiting more patients together because we'll have far fewer patients choosing surveillance because surveillance isn't the standard of care. That will help all of us, really, I think in recruiting to these low incidence tumors.
Ashish Kamat: To summarize, you counsel your patients to undergo surgery and then do adjuvant chemotherapy?
Alison Birtle: If it's operable, yeah.
Ashish Kamat: Okay, great. That was really enlightening and we could go on forever, but in the interest of time, we do have to close. I do want to thank both of you for taking the time to be part of this very important educational activity. Stay safe and stay well.