Alicia Morgans:  Hi, my name is Alicia Morgans. I'm a GU Medical Oncologist and Associate Professor of Medicine at Northwestern University in Chicago in the US. I'm so excited to have here with me today, Dr. Srikala Sridhar, who is an Associate Professor of Medicine at the University of Toronto at the Princess Margaret Cancer Center. So excited to have you here today with me, Kala.

Srikala Sridhar:  Thanks, Alicia. I'm really happy to be here and chat about all the exciting things that are happening in urothelial cancer today.

Alicia Morgans:  Wonderful. So Kala, I really wanted to talk with you about a beautiful presentation that you did at ESMO virtual meeting, 2020. You were looking at biomarkers that might help us predict response, particularly within that JAVELIN trial. Can you tell us a little bit about the JAVELIN trial and then what you looked into in terms of biomarkers?

Srikala Sridhar:  Yeah, absolutely. So I'll just recap quickly for you the JAVELIN trial. And I know you know this trial very well, and you've probably spoken to the lead author, who's Tom Powles, about this trial. But basically, this was a trial, it's recently been published in the New England Journal, and it randomized patients with metastatic urothelial cancer, specifically those that had stable disease or responding to treatment to either avelumab or to best supportive care. And that was given at 10 milligrams per kilogram, every two weeks. And in this trial, as you know, the median overall survival was 21.4 months for the avelumab plus best supportive care arm versus 14.3 months for the best supportive care arm alone. And that was a hazard ratio of 0.69 with a P value of less than 0.001.

So this was really, really exciting, probably the biggest study that has happened in metastatic urothelial cancer in, I'd say, easily 30 years. So that was sort of the premise of it, certainly very, very exciting. And so based on this study, we performed a preplanned exploratory analysis into, as you said, the biomarkers. So can we understand who we should be treating? Or the question I like to ask is really, who maybe we shouldn't be treating with these drugs. And so I think this is really, really an important evaluation going forwards in the field.

Alicia Morgans:  I completely agree. And I appreciated the way that you went through things in this trial, looking at things that we've already assessed. So things like PD-1, PD-L1, staining. In the tumor tissue, in the microenvironment area, looking at things like tumor mutation, tumor mutational burden. And then you also came up with some really interesting and thought-provoking immune cell signatures of expression and populations that may be associated with response. Can you tell us a little bit about what you found from the old markers and then what you started to explore?

Srikala Sridhar:  Sure. So the first analysis that we did as part of this biomarker study was really looking at PD-L1 status and how that may impact the outcome. And we looked at PD-L1 expression on tumor cells and on immune cells. And we found that neither expression on tumor cells nor immune cells fully predicted the overall survival benefit. So it's definitely not going to be that easy, I think looking forwards, right.

Alicia Morgans:  I completely agree. And we've seen this I think, in other settings where PD-L1, PD-1 expression is really not, is not involved or doesn't seem to be a perfect biomarker. I mean, I think that the initial atezolizumab study, the IMvigor study that didn't ultimately confirm what we expected in terms of PD-L1 staining was disappointing, but was a huge lesson to us that this was probably not going to be the sort of most meaningful biomarker that we thought that it would be. But you also looked at tumor mutational burden or TMB, which I think is interesting. And we've seen some presentations where individuals have actually combined TMB and PD-L1 staining to see if that would be more effective. And you guys looked at TMB in this data set as well, right?

Srikala Sridhar:  Yeah, exactly. So TMB is really a measure of the somatic mutation rate. So if it's high, it may predispose tumors to be better recognized by the immune system. So that's the basis of TMB and why it might be important, especially when we're testing immunotherapies. And so in this analysis, what we had done is we used a median of 7.66, and this is important because usually, the threshold is 10 mutations per megabase, but this was exploratory. And what we found that patients above the median showed an overall survival benefit with avelumab, but this was even in the subgroup defined as PD-L1 negative.

And then we also found that patients with TMB below the median showed an overall survival benefit if they were PD-L1 positive. So neither biomarker in this analysis at least was sufficient to define all patients who may benefit from avelumab. So the plot seems to be thickening, right? So I don't think we can say, look at PD-L1 alone. I don't think we can look at TMB alone. In combination, maybe we're getting a little bit more information, but it's still not enough to be able to perfectly identify the patients who we should be treating.

Alicia Morgans: Absolutely. But you and the team actually did some more exploratory analyses that were actually very new to me. I had not really thought about or looked at any of the signatures that you had described in terms of the specific mutations and their impact on outcomes. And then you looked also, I think, at different genes that maybe, when mutated, associated with survival. Can you tell us a little bit more about that?

Srikala Sridhar:  Yeah, sure. So I think one thing that was really interesting and as we dig deeper into the whole TMB story, was we actually looked at where the mutations were happening, what signatures were occurring as a result. And then secondly, in terms of what genes were involved. So in terms of types of mutations, we found that these signatures resulting from a C to T transition, if you go back and think about mutations that can happen in the DNA. So C to T transitions, in particular, appear to be important. And the other was DNA damage repair genes, and these have been thrown around for a while. Now that maybe that might impact too on response to treatment. So that was kind of interesting. So both the type and location, not just the degree of, or the level of mutations, but actually specifically where they're happening. So that was important.

And then we did do these interesting things like volcano plots that were looking at the relationship between gene expression level and hazard ratio. So high expression of genes on the left side of the volcano plots that I was showing were associated with lower hazard ratios. And in this way, we were able to sort of see what genes seem to be involved. And it came out that genes were involved both in the innate and the adaptive immune system were playing a role. So some of the genes like CD8, CXCL9, interferon-gamma, had been previously described in other immune checkpoint inhibitor studies, those came out to be kind of interesting.

And then we, we further looked at signatures that corresponded to immune cell types, which express the FC gamma receptors 2A and 3A. And these receptors, as you know, combine the FC region of IGG with higher or lower affinity. And so a lot of this is exploratory, but some of these high-affinity FC gamma receptors have been associated with improved anticancer response to things like cetuximab and trastuzumab. So that was part of the reason it'd be interesting to look at this and there perhaps is an association, again, it's early data, early news, but you know, so we'll have to look ahead and look further into this aspect, I think.

And then finally you mentioned looking at, we looked at the relationship between some immune gene signatures. So the JAVELIN Immuno 26 gene signature and the T-cell inflamed 18 gene signature, an outcome, and patients were then assigned subgroups based on whether their scores were above or below the median. And again, there seems to be a suggestion that these signatures may be broadly predictive, but again, early days, early days. So I caution us a little bit in that regard. And I guess finally, the other thing we looked at was in patients who had these signatures, which would have suggested that they would have responded to treatment and yet didn't, there were a few pathways beginning to come out that might suggest resistance such as angiogenesis, notch hedgehog signaling. So they may play an important role in resistance. So it's just a bit of a window into the future and I think it's very exciting.

Alicia Morgans:  I think this is one of the most exciting biomarker studies that I've seen in urothelial in some time because it's really exploring things that have not been described yet and giving us, as you said, a window into the future, where we might go next and these things, of course, can be prospectively integrated into the ongoing trials that we are starting. And we can really validate them in ways that are meaningful clinically and that actually demonstrate their predictive and prognostic value.

So I commend you and the team for this. I think that I don't know about you, but as a medical oncologist, I am not always as in tune with all the integral features that are involved in such a complex system as the immune system. And I think that it's phenomenal that you and the other investigators have really kind of gone outside of our safety zone, our comfort zone, I guess, and have really explored something so interesting. So as you think about this work, what would your overarching message be to folks who are reviewing this and really excited to think about where this might take us in the future?

Srikala Sridhar:  Yeah. I think I have a couple of comments to that. So first, I think I agree with you, this was an incredible body of biomarker work, really rigorous, really well done I'd say. The biomarker team is fantastic. And I think that we need to look at a very comprehensive, deep look, at all of our future trials, wherever possible, right? Because I think we learn, be it a positive trial, be it a negative trial, I think we need to invest that effort to really understand what's happening at the biomarker level.

I think more specifically with this study, we don't have all the answers yet, but I think it's setting the stage. And I think that we really can learn as to the next steps in terms of where do we go when someone has a disease that doesn't respond to treatment, what is our next step? Do they have a resistance mechanism, perhaps one of them that we've outlined here, perhaps a different one, but can we learn that, so then we can develop the next set of new drugs for this disease? It's been super exciting and super interesting. And I think it's just beginning, I think there's a lot more to come in this disease.

Alicia Morgans:  I agree. And we'll talk about some of those additional things or at least thoughts about old tricks and how we might use them again in another conversation. But I want to sincerely thank you for your fantastic work, your wonderful ESMO 2020 presentation, and for taking the time today to talk with me.

Srikala Sridhar:  Oh, thank you so much, Alicia, it's been a pleasure talking with you. It's been a pleasure being involved in this. And I think with my lead authors who, like Tom Powles and Petros Grivas and the entire team, I just feel so thrilled to be part of this major sort of transforming step in this disease. Thank you for having me.

Alicia Morgans:  Wonderful and keep up the good work.

Srikala Sridhar:  Thanks.