Radical Cystectomy Versus Trimodality Therapy for Muscle-Invasive Bladder Cancer: A Multi-Institutional Propensity Score Matched and Weighted Analysis - Jason Efstathiou & Alexandre Zlotta

June 22, 2023

Leslie Ballas, Jason Efstathiou, and Alexandre Zlotta discuss the study Radical Cystectomy Versus Trimodality Therapy for Muscle-Invasive Bladder Cancer: A Multi-Institutional Propensity Score Matched and Weighted Analysis'. The research suggests TMT is as effective as radical cystectomy, exhibiting no significant disparity in five-year survival rates. This implies TMT could be a valuable alternative for certain patients, not just those where surgery isn't viable due to severe comorbidities. Findings show that patients requiring salvage cystectomy have survival rates similar to those opting for upfront cystectomy. Despite these promising results, Zlotta advises careful patient selection and surgical expertise when considering TMT due to potential post-radiation complications. Both Efstathiou and Zlotta endorse multidisciplinary, patient-centered care and call for more research into improving treatment procedures, enhancing systemic therapies, and developing personalized treatment through biomarkers.


Jason Efstathiou, MD, DPhil, Massachusetts General Hospital, Boston, MA

Alexandre Zlotta, MD, PhD, FRCSC, Mount Sinai Hospital, Princess Margaret Cancer Centre, Toronto, ON

Leslie Ballas, MD, Cedars-Sinai Medical Center, Los Angeles, CA

Read the Full Video Transcript

Leslie Ballas: Hello, I'm Leslie Ballas. I'm a radiation oncologist at Cedar-Sinai in Los Angeles and I am very excited to be joined today by both Dr. Jason Efstathiou and Alex Zlotta, who are here to discuss with us their recent publication in Lancet Oncology, Radical cystectomy versus trimodality therapy for muscle-invasive bladder cancer: a multi-institutional propensity score matched and weighted analysis. First, I just want to say thank you to both of you. I'm so excited to get to chat with you about this really quite incredible publication and a multi-institutional effort. So if you guys wouldn't mind introducing yourselves. So why don't we start with Jason.

Jason Efstathiou: Hi Leslie, great to be with you. I'm Jason Efstathiou from Massachusetts General Hospital where I'm a radiation oncologist.

Alexandre Zlotta: I'm Alex Zlotta, a uro-oncologist at Mount Sinai Hospital in Princess Margaret Cancer Center in Toronto.

Leslie Ballas: So why don't we start with Jason, you, sort of reviewing with us the publication and the findings that you want to highlight.

Jason Efstathiou: Thanks, Leslie. So this is the paper that Leslie is referring to and I just want to make it very clear, this is our paper, Leslie, you're the second author on there. And this was a wonderful collaboration between Princess Margaret, Mount Sinai in Toronto, USC in Los Angeles, and Sia Daneshmand, who is a high volume uro-oncologist at USC and the Massachusetts General Hospital in Boston, which has a long history in trimodality therapy. So an absolutely wonderful collaboration. Just to give background, we know that radical cystectomy has been considered a standard treatment for muscle-invasive bladder cancer, but there's been a paradigm shift towards organ preservation in many different cancer sites. And there is an option in muscle-invasive bladder cancer and that's called trimodality therapy, which is a combination of a maximal TURBT followed by chemoradiation.

And yes, this is not necessarily an option for all muscle-invasive bladder cancer patients, but for select ones, and we'll get into those criteria in a little bit, but there's been a lot of growing evidence supporting TMT. It's accepted by many national international guidelines, but still it's not being as commonly used as one might think and it has often been restricted to patients with more significant comorbidities where surgery is really not an ideal option and part of the reason why it's not widely used is a lack of very good comparative studies. Prior randomized trial between chemoradiation versus cystectomy in the UK closed due to poor accrual. And so in the absence of any good completed accrued level one data, we sought to provide what we hope is the best available evidence comparing outcomes between TMT and radical cystectomy.

So the cohort includes 722 patients with muscle-invasive disease at the three institutions that we've discussed. The tumors were less than seven centimeters. There was either no or only unilateral hydro and no multifocal extensive or diffused CIS. We had two independent statisticians do these analyses and they used two very robust statistical techniques. One was propensity score matching with three to one matching and the other was inverse probability treatment weighting. And these are robust techniques when trying to sort of recreate randomized trial designs. You can see the covariates that were adjusted for and the primary endpoint of interest was metastasis free survival. We performed this as an intent to treat analysis. We performed many sensitivity analysis and one really important point is that 30% of the muscle-invasive bladder cancer patients who underwent radical cystectomy at the contributing institutions were actually candidates for this study and thus candidates for trimodality therapy. So that's an important number to keep in mind, that maybe it's up to 30% of muscle invasive bladder cancer patients who would be good candidates for TMT.

Here you can see the distribution of propensity scores before and after matching. This was a well-matched cohort and on the right you can see the features after matching, the median age 71, 3 quarters male, which is common in bladder cancer, 10% had hydronephrosis, 60% got either neoadjuvant or adjuvant chemotherapy, full dose systemic therapy and 90% were clinical T2 after matching. So here's the primary endpoint that we looked at. Metastasis free survival, these are the propensity score matched results and you can see no difference in five-year survival probability between either RC or TMT. Absolutely overlapping results. By looking at the other statistical method IPTW, similarly, no difference in metastasis free survival.

And we performed a sensitivity analysis where we limited the radical cystectomy patients to only those who had received neoadjuvant chemotherapy and compared them to TMT. And as you can see, again, no difference in MFS. And this slide here I think really speaks to the generalizability of our results. So whether you were operated on in Toronto or Los Angeles or received chemo radiation in Toronto or Boston, results are exactly the same. So this is looking at MFS by treatment group and center and there's just no difference between the modalities of the treatment or the centers where the treatment was delivered.

And lastly, here's some important other pieces of data. Salvage cystectomy was performed in 13% of patients who underwent TMT. It's an important part of the paradigm of trimodality therapy that if there is a recurrence in the bladder, a muscle-invasive recurrence, it can still be salvaged and then that can still be very curative. Interestingly, in the trimodality therapy patients, those that required a salvage cystectomy due to a muscle-invasive recurrence in their bladder versus those that didn't require salvage cystectomy because they never had such a recurrence, survival was exactly the same in the long run. So salvage cystectomy contributes to the success of this treatment paradigm.

There were about 20% non-muscle invasive recurrences in the TMT patients. This argues for lifelong cystoscopic surveillance in these patients, right, recurrences can happen, they can happen even many years after treatment and those can be managed conservatively with TURBBTs and intravesical therapy. But it argues that these patients need to be followed closely. You see the final pathologic T stage here for the radical cystectomy patients on the third bullet point, and that just really speaks to what's well known in the bladder cancer community, that there is clinical pathologic stage discordance in bladder cancer. That's really important. We clinically under stage patients, and this is exactly what we would expect in a muscle-invasive cohort in terms of what you would see pathologically with 44% T3, T4, and a quarter of patients having node positive disease.

As you would expect in the radical cystectomy cohort, T2, patients that have hydro do worse than most patients that don't have hydro and higher T stage leads to worse outcomes than lower T stages. That's exactly what we would expect. It was sort of just a reality check and I would just argue that the results in this study were not due to suboptimal surgery, right? The cause-specific survival in the surgical cohort and the TMT cohort were well over 80%. The peri cystectomy mortality rate in the cystectomy patients was two and a half percent. That's very low. And that is what we see at centers of excellence, high volume surgical centers.

The median number of nodes removed in the cystectomy patients was 39. That's a lot. These are good surgeons like Dr. Zlotta, like Dr. Daneshmand, these are the top surgeons. This was great surgery and the positive surgical margin rate was only 1.1%. This was excellent surgery. And so it just highlights that these results weren't because it was suboptimal surgery, this was excellent surgery and TMT held its ground in these analyses. And so maybe I'll stop there and happy to discuss further.

Leslie Ballas: Thank you for sharing that and for highlighting the results from the paper. I'm curious, how did this collaboration start? How did you guys decide to look into this?

Jason Efstathiou: Alex and I have been friends for a while and we were, I believe at the EAU somewhere, it was probably five, six years ago. It was at some international meeting and we ran into each other and we said we should do this. And then at a subsequent beacon meeting we cornered Sia and you Leslie and said, hey, would you be interested in participating in contributing surgical patients? And I think that's really, we were all aware, and Alex and I were aware that better comparative data, trying to compare apples to apples was needed, right? And while this is still retrospective research and it's not perfect and it's not a randomized perspective trial, we did our best in this study to try to compare apples to apples and provide the best available evidence possible short of running what is not planned and is really hard to do as the UK showed us in a randomized trial.

Leslie Ballas: And Alex, what do you think makes this study different than other retrospective analyses? Jason mentioned comparing apples to apples. What does that mean and how is this different?

Alexandre Zlotta: I think that really what makes the difference here is that it's the granularity of the data that we brought, the fact that this was analyzed independently. Basically we decided as well to have two teams of statisticians looking with different methods completely independently so that there's absolutely as minimal bias as possible. And then because all of us have what we have multidisciplinary bladder cancer clinic, we truly wanted to mimic what would happen when we see patients and when they're offered one or the other option. And we truly basically try to get as close as it gets. Although it's a retrospective study, all the patients in our cohort, and I know in Boston as well are actually integrated in databases prospectively week after week and we monitor them. The analysis was retrospective, but we went somehow prospectively and used the best methodology in the absence of randomized study to come up with something which is helpful for patients and physicians.

Leslie Ballas: I think another thing that I think is really great is that all of the patients that you included from the surgical cohort were and would've been candidates for TMT. There was no bias for the patients being either too sick for cystectomy and therefore getting TMT. I mean 75% of patients had ECOG zero. So these were robust patients who got TMT as well as cystectomy. And so limiting the patients who were included from the surgical cohort to those who had no, or unilateral hydronephrosis, size criteria for the tumor, I think that also really makes this a very robust study and really does compare similar patients. So I think that's a really key component and something worth highlighting.

Jason Efstathiou: I think you're right Leslie, I think it's improperly clinically matched patients that are eligible for both treatments. Survival is comparable between RC and TMT. And I think that that's sort of a big message here. And the other really big message is often chemoradiation has been reserved to those patients that are not good surgical candidates, but these results in this study really support a trimodality therapy in the setting of multidisciplinary shared decision making where they've heard about all viable options. But trimodality therapy should be offered to all suitable candidates with muscle-invasive bladder cancer and not only to patients with significant comorbidities for whom surgery is not an option. I think that and that I'm basically reading from the conclusion of our paper, but I think Alex and I would feel very strongly that that's the case.

Leslie Ballas: With that kind of comment. So who are the patients that can and should be offered trimodality therapy? Anyone with T2 to T4 N0 disease I think?

Alexandre Zlotta: Yeah, so I think we run our clinic a little bit like Jason as well here in Toronto where we have defined and the data we're existing as Jason mentioned about the efficacy of TMT for specific patients. And so I have to say that in this particular study, we try to limit to what we call the ideal candidates correct tumors that you can fully resect, no multifocal, so you don't have a genetic instability and a risk of recurrence, which is increased, no multifocal CSS because we do know it's a surrogate. No bilateral hydro because we know that this always will understate.

And so what that study shows is that the patients that we see on a weekly basis that are doing super well after TMT are actually indeed an equal outcome as radical cystectomy. And it also mimics somehow some of the RTOG criteria. What I think the big next step will be is, okay, ideal candidates are doing fine, what's the sweet spots? How much can you push the envelope? Is it two tumors, three tumors? How much CI is next to it? So I still think that our study is the first step identifying the best candidates, but they're still probably other candidates or for one or the other option that need to be found too.

Leslie Ballas: I noticed that there was like 90% of patients or about 90% of patients once there was the upper matching, had T2 disease. And that's for both surgery and radiation obviously. And so where do you think the T3, T4 patients are? Are they all node positive and therefore are many of them node positive? I mean, why are we not seeing as many in this cohort?

Jason Efstathiou: Yeah, I can offer a couple of thoughts. One, 45% of those T2's are actually T3, T4. Number one, but because of the clinical stage, clinical pathologic stage discordance. But I think my thoughts are that trimodality therapy has historically sought acceptance, but in that process boxed itself a little bit into a corner of the perfect patient for TMT. And I think just to pick up on what Alex was saying, very important point that where is that sweet spot?

Can the eligibility be broader than what we offered in this study? And what we can tell you is that in our study, the T3, T4s, and as you mentioned after matching 90% were T2, but 10% were T3, T4. They did not fare worse with TMT compared to cystectomy. They did equally well. Now we would love more numbers of course, but that early signals telling us there wasn't anyone in our study that was faring worse with TMT compared to RC. That was obvious in all the kind of clinical pathologic variables that we would want to look at. And so maybe as Alex is suggesting, future research has to look at how broad can that eligibility be.

Leslie Ballas: Yeah, one of my favorite things about the paper, which Jason highlighted during his slides is that the patients, the 13% that need salvage cystectomy end up with cancer specific survival that is similar to those who did not end up getting salvaged cystectomy and sort of thereby similar to those who would've gotten cystectomy upfront. Alex, what do you think about patients getting TMT and if god forbid they're in that 13%, then they just go to salvage cystectomy, sort of this idea that you should always try to start with TMT, preserve the bladder and only those that then need salvage cystectomy go to cystectomy.

Alexandre Zlotta: Yeah, I think it's a good point. I think we have to be slightly careful though. And what I mean here is that we've defined what we think would be kind of ideal candidates here. Correct. And we know that in these ideal candidates, despite having multifocal CSS, still about 10 to 13% will have a recurrence, correct. And so although one could say, okay, why don't you just start with TMT and then if it fails, just carry on with cystectomy. There are two points to be mentioned. One, truly requires respectfully to be in centers where surgeons are very comfortable operating on post-radiation on a regular basis because if you do it once or twice a year, honestly speaking, you're going to have increased complications. And so that's one of the specific points. Two, it shows that our study is not pitting surgeons against radiation and medical oncologists.

It's truly a collaboration where each of us are playing a specific role. And third, although in theory, technically you can perfectly do a neobladder up to TMT and sometimes it may work, we have to be realistic that with patients who are often about the age of 70, a cystectomy neobladder after radiation would end up probably with less functional outcomes than in a younger patient non in radiation. And therefore, very often one of the discussion is either people choose to have a neobladder upfront or they do TMT with the understanding that often due to their age if they fail, they have to kind of accept a non-continent diversion, although it's always technically feasible. And so slightly a little bit more in shades of gray than black and white.

Leslie Ballas: Yeah, I think those are really good points. We looked at our own experience when I was at USC with Dr. Daneshmand about doing neobladder after any sort of pelvic radiotherapy including prostate cancer treatment, which goes to a much higher dose obviously. And we found that in the well-selected, as you are saying, in the well-selected patient, it is feasible. And yes, of course it's technically feasible, we don't have exceptional functional outcomes for those patients, but I think that people understanding that it's not totally off the table is important as well. Okay, so if you could say one takeaway or a couple takeaways for practitioners from this study. Jason, you sort of summarized your results, but what is sort of the big picture takeaways for practitioners?

Jason Efstathiou: The big picture takeaways I think are, first of all, we have to acknowledge many muscle-invasive bladder cancer patients nationally are not getting curative treatment. And TMT can help fill the gap. There is an underserved and undertreated population and certainly there are options for treatment such as TMT. We know TMT is supported by numerous clinical guidelines, et cetera. But now we have really good, I think, robust data suggesting, as I said earlier, in clinically matched patients, survival is comparable and we know that over 85% of these patients are going to keep their native bladder. We also know from separate quality of life research that long-term quality of life is good. I think we know, and as Dr. Zlotta can eloquently talk to, trimodality therapy is not a non-surgical treatment that TURBT, that salvage cystectomy, as we've highlighted, are all important components to the success.

This requires multidisciplinary team-based care. And I think our collaboration and this paper highlights that value because ultimately we have to safeguard the autonomy of patients with bladder cancer and inform decision making by the patient, is key. But they can only have that if they're meeting with all the disciplines and have team-based care that are offering viable options. Again, not all patients are candidates for these treatments, but that are offering viable options for patients and patients that are suitable candidates should be offered TMT. And I think that this data and this paper support that notion.

Alexandre Zlotta: I would just echo what Jason said. I think the paper epitomizes the multidisciplinary nature of bladder cancer management and I think times are completely over where only one specialty deals, whether radiation or medical or surgeons, each case has to be discussed separately, individually. No one patient is the same. But the way that we and many other centers look at things that patient come with a disease, they're assessed by different teams. The different teams meet and decide whether a patient would be a better candidate for surgery, a better candidate for both, or better candidate for TMT. And once the teams have decided, we then come to the patient and family and propose the different options, bringing the pros and the cons and ultimately it reassures patients and people choose and the family, the patient after being fully informed, chooses what we know as of today is the best candidates. And we know that for those patients that Jason mentioned, I think it's fair to say that they should be proposed different options. Research will try to analyze how far this can be expanded and fine-tuned.

Jason Efstathiou: Yeah, I totally agree with Alex. And I think just to pick up on the research thought, yes, more research is needed, this paper was great, but there's a lot more to do and Alex has referred to kind of extending the eligibility criteria for TMT, but we need more research in optimizing clinical staging of the disease. But we've talked about the clinical under staging and really getting a good sense. And for example, MRIs are, I think underutilized and there's a lot of value in that we need to continue to work on optimizing the delivery of radiation therapy. There's hypo fractionation that's out there as options, there's adaptive planning, there's tumor boost dose escalation, there's all kinds of... There's maybe decreasing the number of fractions and making it therefore much more accessible to the population. And there's research looking at, can we squeeze all the radiation to five treatments and really make it an option that is broadly easier to deliver and available and addresses some of the financial toxicity patients experience and access to care.

We need to improve systemic therapies. And there's lots of trials going on, adding immunotherapy to the base of what we've talked about, the chemoradiation, does immunotherapy improve outcomes? And I think we're going to get that answer soon. And then we just need to keep working on biomarkers and validating biomarkers to guide bladder preservation therapy and for personalized treatment selection. Could it be that there are biomarkers out there that tell us, no, this patient should go for cystectomy and this one should go for chemo radiation? Well, there might be, and I think we need to support and continue to engage in such research. And the wonderful thing about this paper is we have three institutions that are intimately connected now and we can use it as a launching pad for much more good research to come.

Leslie Ballas: Well, thank you both so much for taking the time to talk with me today. I'm excited about the future of bladder cancer research after hearing you both speak. And I think that there is a lot of good information that is yet to come and exciting information that is yet to come. So again, thank you and I'll look forward to seeing both of you hopefully soon in person.

Jason Efstathiou: Thanks, Leslie.

Alexandre Zlotta: Thanks.