Ashish Kamat: Welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat from MD Anderson Cancer Center in Houston. And it's my pleasure to welcome Dr. Matt Galsky, who is a Professor of Medicine and Acting Chief of the Division of Hem-Onc at the Icahn School of Medicine at Mount Sinai. Matt needs no introduction. Dr. Galsky has done a lot of sentinel work in bladder cancer, has been involved with multiple different research groups and review groups and bladder cancer advocacy groups, and Matt and I also have been working over the last 18 months or so on updating the guidelines when it comes to immunotherapy for bladder cancer. And essentially there's no better person than I thought to invite to provide you, the audience, with a quick update as to what's new in metastatic urothelial cancer. So, Dr. Galsky, welcome.

Matthew Galsky: Thank you so much. It's my pleasure to be here. And I was hoping to take you through a whirlwind tour of metastatic urothelial cancer. Okay. So the take-home point number one is that first-line chemotherapy remains standard of care, and the role of PD-L1 testing based decision-making remains quite complicated. And I'll tell you what I mean by that.

So several Phase III trials have read out in the first-line treatment of metastatic urothelial cancer in the past few years. This is really more Phase III trials than we've seen in this space in probably more than a decade. These trials asked two major questions. Should we give chemotherapy alone for first-line treatment of metastatic urothelial cancer? Should we give single-agent PD-1 one or PD-L1 blockade, or immune checkpoint blockade doublet therapy, or should we give chemotherapy plus immune checkpoint blockade? And to make a long story short, the chemotherapy based combinations.

And so far the combination of PD-L1 blockade plus CTLA-4 blockade have not been a home run in terms of replacing frontline chemotherapy. And so chemotherapy alone, platinum-based chemotherapy, remains the standard of care. But we have this dilemma, right, because the label indication for first-line treatment of cisplatin-ineligible patients with metastatic urothelial cancer still includes single-agent PD-1 or PD-L1 blockade in patients who have high levels of PD-L1 expression.

So should we be using PD-L1 as a biomarker in this space? These three trials somewhat addressed this question. So here's PD-1 or PD-L1 blockade versus chemotherapy in DANUBE, KEYNOTE-361, IMvigor 130, in the intent to treat population. So this is regardless of tumor PD-L1 expression. These are the all-comers, you can see that the Kaplan-Meier curves look relatively similar for overall survival across all three of these studies. The initial portions of the curve favor chemotherapy, and then subsequently crossover.

So so far, no definitive role for single-agent immune checkpoint blockade in all-comers versus platinum-based chemotherapy. However, overall during the course of the survival curves, the ultimate outcomes are pretty similar.

What about PD-L1 testing? Well, this gets complicated because the assays that are used with each program are completely different. So when we talk about PD-L1 testing, we're not really talking about one thing and you might as well be testing for a different protein across these different assays because you're identifying different patient populations.

What do I mean by this? Well, here are the results of the PD-L1 subsets of these three randomized Phase III studies. You can see DANUBE, KEYNOTE-361, and IMvigor 130. These are the overall survival curves. And you'll notice that they look a bit different. DANUBE and KEYNOTE-361 look pretty similar to the intent to treat population, but the IMvigor 130 study looks a little bit different with the survival curve generally favoring immune checkpoint blockade. However, this does not reach statistical significance, and this is just an exploratory endpoint.

So why are we seeing these differences? Well, there are a bunch of possible reasons, but one of them is this biomarker is identifying different patient populations from 24% of patients being PD-L1 "high" in one study, to 60% in another. So clearly different biomarkers. We can't use these interchangeably.

Take-home point number two is that switch maintenance immune checkpoint blockade improves outcomes and really has become the standard of care. There have been two randomized studies in this setting, a smaller randomized Phase II study with switch maintenance pembrolizumab, and a large randomized Phase III study with definitive overall survival endpoints using avelumab. So this is four to six cycles of platinum-based chemotherapy. And then instead of waiting for patients to recur, those patients with stable disease are then immediately placed on immune checkpoint blockade. You can see the overall survival curve from JAVELIN Bladder 100 showing improved survival with this approach. And this has become standard of care.

How about other treatments? That's take-home number three, antibody-drug conjugates are active in the "third-line" metastatic setting in urothelial cancer. And we have two antibody-drug conjugates that are on the block. One of them has already received FDA approval, and that is enfortumab vedotin, which is an ADC directed against the protein nectin, and the payload is MMAE. And there's an antibody-drug conjugate for which we have already a lot of data in urothelial cancer sets, which is sacituzmab govitecan, directly against TROP2, with a payload of SN38.

And in rather large single-arm Phase II studies in patients who already had chemotherapy and already had immune checkpoint blockade, these drugs are associated with single-agent response rates: 44% with enfortumab, 27% with sacituzmab. Respectable compared to therapies that we've had in the second group setting in this disease in the past. Some of these responses are quite prolonged. And so these are certainly important new tools in our armamentarium. They have different side effect profiles.

And finally, take-home point number four, the adjuvant space. The adjuvant space in muscle-invasive bladder cancer is evolving rapidly. So we have three large randomized trials that are testing or have tested this concept of, should we give immune checkpoint blockade in the adjuvant setting to patients who have residual cancer in their surgical specimen after receiving neoadjuvant therapy, or for patients who have locally advanced disease on pathology but were not candidates versus platinum-based neoadjuvant therapy because they're cisplatin-ineligible? And so three trials similarly designed with some important nuances and we have some results.

So the IMvigor-010 study did not meet its primary endpoint. That was presented at ASCO 2020. I'm going to present some intriguing data that's recently emerged from the IMvigor-010. CheckMate 274 did meet its primary endpoints, and that's going to be presented at ASCO GU 2021. And the AMBASSADOR study continues to evolve.

So what do we know about IMvigor-010? This trial mentioned is a randomized study of adjuvant atezolizumab, did not meet its primary endpoint, but there was a pre-specified analysis based on ctDNA in this trial, and this data was presented by Tom Powles at the ESMO IO symposium. And what the investigators did was look at whether or not patients had minimal residual disease detected by ctDNA on cycle one, day one of therapy, and looked at outcomes based on whether or not this MRD assay was positive. Here you can see the prognostic impact of having positive ctDNA in this study in the observation arm, both for disease-free survival and overall survival, and really a powerful prognostic factor in this setting.

What about treatment, though? Is this predictive? And here you can see that in patients who have positive ctDNA, both disease-free survival and overall survival, the positive patients seem to benefit from adjuvant immune checkpoint blockade, while the negative patients maybe not so much. And so this is an important potential tool to refine decision-making in the adjuvant setting. That certainly needs to be prospectively validated before implemented in the clinic, but it really changes the way that we look at this space potentially in the longterm. Those are the take-home points and thank you for your attention.

Ashish Kamat: That was excellent, Matt, and I did time you. You were true to your word, so thank you. Let me ask you a couple of questions and maybe you could go backward from the points that are covered. Take your take-home point number four, as far as the adjuvant space and it evolving. So if you look at the data that's emerged in the trials that have essentially readout per se, how would you compare and contrast this with the body of literature that we have when it comes to adjuvant chemotherapy, number one? And if you would postulate, for example, the role of measuring circulating tumor burden, in this case circulating tumor DNA, when it comes to immunotherapy versus chemotherapy? Just your thoughts on the differences that we see between responses in chemo and immunotherapy.

Matthew Galsky: Sure. So a few thoughts in that regard. One is that, and I think this is maybe a point that isn't receiving as much attention as it should, that these adjuvant studies are testing a different patient population than we've tested in adjuvant populations in the past. Prior adjuvant studies with chemotherapy have been in patients who can receive cisplatin-based chemotherapy, not in patients who've had neoadjuvant therapy and had residual disease, not in cisplatin-ineligible patients. So probably the only recent potentially practice-changing study that is encompassed in these studies is POUT because POUT obviously included some patients who were cisplatin-ineligible, but that's limited to upper tract disease, a small subset of these patients. Otherwise, these are different patient populations that haven't been addressed before. This is an unmet need in a new space.

Regarding ctDNA, I think we have to separate this notion of patients who need treatment from patients who benefit from treatment. And these are both major hurdles in terms of curing muscle-invasive bladder cancer. I think ctDNA potentially can help us identify who needs treatment but does not necessarily solve the problem of who benefits from treatment. And I think just like we see in the metastatic setting, a ctDNA positive patient, one might benefit from adjuvant immune checkpoint blockade, one might benefit from adjuvant chemotherapy, one might benefit from either of those approaches. And until we determine that, it's going to be hard to truly apply a precision approach to the adjuvant setting.

Ashish Kamat: You think single-cell sequencing and assays of that nature are potentially poised to actually help us in that regard?

Matthew Galsky: I think single-cell sequencing certainly has a long way to go into becoming a clinical diagnostic just because of the cost and the feasibility. I think learnings from single-cell sequencing studies by all means can be translated into precision assays for use in the clinic.

Ashish Kamat: Great. Yeah, then if you go to your take-home three, which is the ADC's essentially, as a class, for example, EV, which targets nectin-4 which is ubiquitous in urothelial cancer, as a class these drugs appear, at least on paper, to be able to change the course of disease even in earlier stages. And I know you focused on metastatic, but if you could highlight what are your thoughts on using EV as a bladder sparing therapy, for example, in patients that have muscle-invasive disease and are not able or willing to undergo radical cystectomy, say combining it with local consolidation with radiation? What are some of your thoughts in that area?

Matthew Galsky: I think that I'm certainly cautious about over-interpreting single-arm Phase two studies, and even larger ones, but I would say that the data with EV plus pembrolizumab is certainly exciting. It's hard to ignore. It's a small data set. It's a single-arm Phase II study, but it's hard to ignore. And when you see that proportion of patients having some degree of tumor shrinkage, it certainly raises an eyebrow. And so I think regimens like that certainly need to be moved earlier in the treatment setting to see if we can have a bigger impact, particularly on subsets of patients who traditionally have been excluded from perioperative systemic therapies as you mentioned.

Whether or not ultimately that can be translated into this concept of treatment with TUR plus systemic therapy and not removing the bladder, I think lots of us have trials in that space. That's a hard space to be able to show benefit or to change the paradigm, given the longstanding use of potentially curative surgery. But I think it's a goal that we should continue to try and achieve.

Ashish Kamat: Absolutely. And of course, even using ADCs in a non-muscle-invasive setting, clearly as an intravesical formulation holds promise. So that's a class we'll all be looking at very, very closely.

Coming to your switch maintenance. I mean, kudos to you for doing all the work and to Tom for the Javelin 100 study, the data was extremely impressive, but a little bit varied as well. And when you look at this side of the pond and that side of the pond, any thoughts from your perspective that you could share with the audience as to compare and contrast between your work and the Javelin 100 results?

Matthew Galsky: Our study was a randomized Phase II study, a much smaller study with overall survival as a secondary endpoint. It certainly wasn't powered to detect a survival improvement. And so when there's discussion about the lack of survival improvement in our study, I certainly don't recognize that as a definitive conclusion from our study by any means.

We set out to determine whether or not this approach improves progression-free survival. The trial was powered for that and it achieved that. So I think there are more similarities between the trials than differences, to be honest. We did include crossover in our study, and there is some concern about whether or not exposure to subsequent immune checkpoint blockade might have a conduit for differences in survival.

And I think that even within the context of a randomized study, we recognize that, unfortunately, some of our patients who developed disease progression could not be crossed over in time because of the rapidity of their disease. So I'm very respectful of the natural history of this disease. And I don't think that all of our patients have an option to pursue subsequent therapies, which is one of the reasons why switch maintenance is so attractive.

Ashish Kamat: Very, very, very important point. And it's made it into the guidelines in a very rapid fashion. I think something about being stuck at home and COVID allowed the community to do that as well.

And then lastly, your take-home number one. I'm glad you put that as your first take-home, which is first-line chemotherapy or modern standard of care in these patients. There are lots of folks, especially overseas, that I'll bump into and we're having talks, and obviously tele-talks right now. They go, "Oh yeah, I started all my patients on immunotherapy upfront because it's so much easier for them, and I reserve chemotherapy for if they don't respond to checkpoint inhibition, and it's so much better for my patients." And these are well-respected medical oncologists in their own right. What would you say, other than the data you presented, what would you say as a practical response to that statement? I mean, again, not so much science because we know what the science says, but what would you caution those people?

Matthew Galsky: In terms of starting with immune checkpoint blockade?

Ashish Kamat: And then switching to chemo, yes.

Matthew Galsky: Yeah, I mean, I think that if one looks at the Kaplan-Meier curves from the three randomized studies that I showed, then one becomes pretty cognizant that we might not be able to select the patients who are going to fare best from that approach, and we might miss the window of opportunity to get some control of the disease. So I think in the past there was some theoretical discussions about that being a concern. Now we have three randomized Phase III trials showing the same exact thing. I think the wild card here is whether or not PD-L1 testing still offers a benefit in terms of selecting those patients. And there we see mixed results from the studies, depending on the biomarker. And it just speaks to the importance of getting the biomarker right, which is not simple by any means.

Ashish Kamat: Matt, thank you again for taking the time and spending it with us at UroToday's Bladder Cancer Center of Excellence. I want to give you the last word and maybe if you could share with our audience in your typical succinct manner, what are one or two or three of the exciting developments that you are looking forward to in the next 18 months?

Matthew Galsky: I'm mostly excited about what can potentially be achieved in the perioperative space. I think that's where, as a medical oncologist focused on bladder cancer, I have the potential to make the biggest impact. And I think that the opportunity to really change the outlook for patients in that disease state is potentially remarkable if we can apply all of these tools in the right way to maximize benefit and minimize the burden of side effects, and the financial burden as well.

Ashish Kamat: Right. Once again, Matt, thank you so much. Stay well, stay safe, and hopefully, we will see each other in person this year.

Matthew Galsky: Thank you. Looking forward to it.