Ashish Kamat: It gives me great pleasure to welcome a dear friend and colleague, Matt Galsky, who is a Professor of Medicine at the Mount Sinai School in New York. Thank you, Matt, for taking the time and spending a few moments with us today.

Matt Galsky: Thanks for having me.

Ashish Kamat: There's a lot of stuff going on in bladder cancer, right? I mean, it's a chockfull of new developments and you've been instrumental in a lot of the progress that's been made. One of the things I wanted to pick your brain on is the whole field of adjuvant therapy in bladder cancer. What are some of your views and the most important developments in that field?

Matt Galsky: So, I think that the adjuvant space is really a critically important point where we can potentially favorably intervene on a patient's illness. But it's very complicated in terms of space as well because the value proposition there is that we're applying systemic treatment to a large proportion of patients and about 40 to 50% roughly don't need that treatment. They've already been cured. And we're exposing them to potential side effects where they might not benefit at all by definition. And so, it's complicated and I think that's in part why it's had somewhat of a checkered past in bladder cancer.

But I do think that there is a potential role there and I think it's a good space to explore some novel therapies. And immune checkpoint blockade in this setting makes a lot of sense for several reasons. Three large randomized trials ongoing, we've heard the first results in a press release from one of those studies recently, which I think has us scratching our heads maybe a little bit.

Ashish Kamat: About the press release, it has us scratching our heads because they didn't meet their primary endpoint. But at the same time, there is a hint of a benefit to the patients. What's your read on where that would potentially readout?

Matt Galsky: So, of course, what we're talking about is IMvigor010, which is a randomized study of adjuvant atezolizumab versus observation in patients with T3 or higher disease in the bladder who are ineligible to receive adjuvant cisplatin-based chemotherapy or patients who had T2 or higher disease after prior neoadjuvant therapy. So the primary endpoint of this study was disease-free survival. And the press release, we have very limited data, it didn't meet the primary endpoint. I think there are several potential explanations for that and it's hand waving a little bit until we see the data. But this is really our first foray in urothelial cancer in immune checkpoint blockade in the adjuvant setting. And I think quite simply one question that comes up is disease-free survival the right endpoint?

I've had patients on these studies anecdotally and I hate to rely on anecdotes to interpret an entire study. But I have had patients who've had a little bit of lymphadenopathy develop on study came off study, had those nodes irradiated and have gone on to have long-term disease-free survival. And so whether or not there is some pseudo-progression that occurs in this context, that's quite different than the metastatic setting because it would really lead to radiographic evidence of disease where there was none before. I think that's one potential complexity of these endpoints in this setting.

Ashish Kamat: Right. And early on you said these patients who are getting adjuvant therapy have potentially been cured in about 50, 55% of the time. And using pathology to try to predict who needs adjuvant therapy is the way we've done it as a community. There's obviously a push towards getting better predictive tools and better selection tools. Where do you think we are today and where should we be headed?

Matt Galsky: So I'm very hopeful about ctDNA as many people are conceptually in terms of measuring minimal residual disease. I think it makes sense probably more so than measuring properties of the primary tumor, which might tell us biologically who's at risk but actually isn't measuring the disease itself. So I'm fairly bullish about measuring ctDNA in this setting really for two reasons. One for risk stratification, and the other for the potential ability to measure the activity of therapies in this space in smaller studies where we can establish whether or not there's a signal before we move on to these very large randomized adjuvant studies. For those technologies to be beneficial in this setting obviously lots of hurdles need to be crossed, but in particular, they need to provide added value beyond what we achieve with pathologic staging. And pathologic node-positive disease does a pretty good job at risk stratifying patients. Of course, not everyone will recur but a lot of patients will recur. And so what we get from these new technologies needs to add to that and not be redundant.

Ashish Kamat: Right. So Laura Striscott has a study over in Denmark looking at ctDNA as entry criteria for patients who go on to get adjuvant IO therapy. It's small, but we are looking forward to seeing what the signal from there shows. You mentioned node-positive disease as a very important prognostic and predictive in some ways factor. How do you factor in prior therapy when you're counseling our patients as to whether they should consider adjuvant therapy?

Matt Galsky: So in patients who've had prior neoadjuvant therapy, is that what you're referring to?

Ashish Kamat: Correct.

Matt Galsky: We know that pathologic staging after prior neoadjuvant therapy is pretty good in terms of risk stratifying as well. And so unfortunately in that space, we don't have anything that's been proven to be effective. So my standard approach in that setting is observation. But I do think that's a really important space for drug development.

Ashish Kamat: And you have done a lot of work in trying to put pathologic staging or path CR out there in the forefront and you lead an FDA workshop that occurred recently. Can you share with our viewers the summary of all the work that went into that?

Matt Galsky: So that workshop as you know, was really to put the issues on the table. What do we know about pathologic complete response? What don't we know? And what do we need to know to allow us to make a decision as a field to advance the regulatory science in that space? And I think there were some important messages that came out of that workshop. One of which from the pathology side was that when we hear about a pathologic complete response, that might be different from a clinical perspective compared to a pathologist perspective in terms of if they looked very hard to find every last a potential cancer cell, would they find that? I think that's important in terms of making sure that we're defining pathologic complete response the same way across trials and that pathologists are analyzing the specimens the same way. Anytime we have an endpoint or a biomarker, we need it to be obviously as homogeneous as possible and reproducible as possible.

At the same time, we know from clinical studies that having a path CR in the context of the heterogeneity in terms of how that's assessed is still prognostically important. So I think there are some lessons to learn in terms of standardization, but we know that it's important on an individual patient level in terms of prognosis. I think in terms of what that means for drug development, I think it's critically important to include pathologic complete response in all the prospective studies and correlate that with time-to-event outcomes like event-free survival or overall survival really so that we could be confident that with future studies, improvements in pathologic complete response rate will indeed translate into those benefits.

Ashish Kamat: Yeah, and I think that's critical to keep in mind, especially now with the IO's being moved into the neoadjuvant space and we're seeing fairly impressive path CR in relatively small studies. How does that translate to chemo-induced path CR?

Matt Galsky: Absolutely.

Ashish Kamat: This has been great. Any closing thoughts on this pretty broad topic for our viewers?

Matt Galsky: So maybe one closing thought which is that the adjuvant cisplatin-based chemotherapy studies that were done over a few decades worth of time enrolled about a thousand patients collectively. And the three immune checkpoint blockade adjuvant studies that are enrolling over about a five year period of time will enroll about three times that many. And so there was concern in the field as to whether or not we can do these studies. And I think we've proven that in that's an advance in the field in and of itself. And obviously the next step is to have that translate into patient benefit.

Ashish Kamat: Absolutely. That's a great point you made. Thank you again so much, Matt, for taking the time.

Matt Galsky: Thank you.

Ashish Kamat: It's always a pleasure.

Matt Galsky: Thanks.