Ashish Kamat: Hello everybody. And welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat, professor of urological oncology at MD Anderson Cancer Center. And it's a distinct pleasure to, first of all, welcome Dr. John Sfakianos, who is a newly appointed associate professor. Congratulations, John, at the Icahn School of Medicine at Mount Sinai, and another big. Congratulations to you for being awarded the prestigious award, BCAN  Research Innovation Award for this year's think tank. But this is always a great opportunity for you to share what your work is and what your award is about with the wider audience that UroToday can read. So with that John, the stage is yours.

John Sfakianos: Thank you. Thank you for the introduction Dr. Kamat, thank you for the congratulations. It's actually really an honor to receive this award. And I'm really excited about all the work that can get done because of this award and better understanding non-muscle invasive bio-cancer and BCG resistance, which you have a tremendous amount of research in the past on, and I'm sure, still doing great work on that, very much, same subject. For me, reading a lot of your work and a lot of work for mothers in the past, really trying to understand how BCG works, what happens in the microenvironment when the tumor's exposed to BCG, has always been almost a very... I've been always very curious about it, just because there's so much out there in the literature about it. And yet we really don't understand or have really good mechanisms of action.

And so this is just a schematic or cartoon about what we think happens in this microenvironment, or I should say in cancer, microenvironments. And what we think happens is you have all these different immune cells, all these different immune lineages that get invited into the microenvironment because of different cytokines that are secreted, not only from the tumor cells themselves, but from other immune cells that come into the microenvironment as well. And so there are endogenous and there are invited immune cells and immune lineages that come in. And when we give somebody BCG, we really stimulate this pathway and stimulate many of these different pathways to occur, to allow for the actual anticancer and antitumor activity to occur. To try to make sure that all the cancers are eradicated and then the actual mechanisms of memory occur, where we can prevent these tumors from coming back.

But yet which parts and which specific mechanisms are really still not quite understood that come into bladder cancer? And how these different immune lineages are recruited into the actual microenvironment and then once they get there, the interactions and so forth. And while a lot of your work and a lot of others have shown the different cytokines, IL-12, IL-6, trail and all these other different cytokines that can be present. We still don't know exactly the interactions between these and the actual immune cells, the tumor cells, the benign cells, et cetera. And so really our work set out to start to better define and to refine these mechanisms, to understand what's happening inside the bladder. And this is a schematic, I think I've shown this actually on UroToday, previously, with my collaborators that we had. And this is what we think is happening in general, where BCG is actually causing every single bladder, every single tumor to have an actual inflammatory response, pretty much mitigated by interferon-gamma.

And what ends up happening is that everybody, or the different bladders, in different patients, and in different tumors will then react to this interferon-gamma in different ways. And the continued stimulation that some patients may have because of the continuous intravesical installations, either for induction or maintenance, that itself, can actually then lead to significant dysregulation and a lot of different changes that can occur to lead to tumor resistance. So, I would just pose a question then maybe maintenance is good for some patients, but really bad for other patients. Maybe six doses are really good for some patients but maybe we only need two doses in other patients. And we haven't really dug into those questions, into those details to really understand that yet, and that's what our primary goal was. And the ability to do this, now, today has actually become a little easier, but a lot more expensive.

And the reason for that is because of these really novel technologies. And so traditionally we had bulk RNA sequencing, which was great because you can understand what the gene expression profiling were in the entire tumor, but you couldn't tell exactly where the different expression profiles were coming from, which different cells, which different lineages. We then moved into single-cell RNA sequence sequencing, which was also great because now you can actually understand the different clusters, the different lineages, the different cells that are actually in the microenvironment. What are they doing? What mechanisms and what mechanics are turned on or turned off in those cells, to allow the different actions that they can admit to happen in the actual microenvironment? That's great because we're learning more, but that still wasn't good enough. And now at this point, we have something called spatial transcriptomics or 10x Visium transcriptomics, which allows us to get to about a three to five-cell layer level on an actual parafin slide.

So now you actually have the three-dimensional, you have the resolution of to up to three cells and now you can actually see the different immune cells, the different immune lineages, the areas of tumor cells or the different tumor cells that are grouped together in the different parts of the actual slide or the different tumors, and now actually get really refined interactions, tumor immune cell, and what's happening. And the best part of that is now you actually can do this on a stack of slides. So you can cut a tumor into 10, 15 different slides. You can do the spatial transcriptomics on one slide and then you can do protein analyses on the others to actually validate what you're finding in the expression profile. So from there, we were actually able to do some preliminary data analyses on BCG-resistant cells. So this is a group of four BCG-resistant non-invasive bladder tumors.

The median time to recurrence for these is 3.6 months, so these are really unresponsive resistant tumors. And we can actually do a variety of different analyses where we can actually see where the CD4 T cells are in location to actual tumor cells. So you can see hot and cold, so you can see where the actual hot areas of the tumor are, compared to the cold areas. You can then do a variety of different other analyses with the specific markers of interest where you can actually look within these areas of, let's say hot versus cold, what is happening. And interestingly enough, what we identified was this variability in expression, which is an MHC class 1 protein. And these areas that you can see where there's high HLA-E actually correlate with areas that actually have a significant amount of cytotoxic cells. And these cytotoxic cells are actually producing cytokines that help recruitment of the different immune lineages to those specific areas.

But you have other parts of these tumors that are actually cold. So this is a cold area, cold tumor that doesn't have any T-cell surrounding it and doesn't have any set of toxic type of activity around it. And the idea here is that we think that there's actually program cells. There are programmed tumor cells that will recruit the different immune lineages to that area of the tumor, that will allow these areas to build resistance or to exhaust immune cells, to allow other areas to survive. It's actually a hypothesis that's been shown in melanoma cells with these exact different cytokines, CXCL9, 10 11, and CXCR3. And now we're actually seeing this same type of activity in BCGM responsive bladder cancer. And so with these type of really refined analyses, we hope to be able to expand our cohorts and to validate some of these findings that we're seeing.

And the other interesting part is that we actually can then look at the actual tumor cell, immune lineage-specific interactions, based on the different subtypes. So we can actually take parts of these tumors that we've done spatial sequencing on, and then be able to cluster the different cells based on the different subtypes. And then do neighbors analyses to be able to actually look and see which cells are closest to the basal cells, which cells are closest to the p53, in terms of immune cells, and within these immune cells, we can actually see, are they exhausted or are they active? Are they trying to kill? Or they're dormant because of high levels of the different exhaustion markers. Again, this is very interesting because you can see things like basal cells, which really, if we use the same sort of heat map type of scale, the basal cells really don't have many T-cells next to them.

But obviously, we would expect the immune lineages of immune subtypes to have a variety of different immune cells near them. But luminal cells can have CD8+T cells that are close by. You can see p53 cells have a variety of NKT cells that are nearby. And so, now we are starting to get a very refined, really magnified look at these different subtype of cells, along with the different cells that are in their neighborhood and what they're doing to help us better understand how we can actually target or attack these BCG resistant cells. But furthermore, I think we're going to start learning a lot more about how the immune cells become exhausted in these tumors, to allow us to be able to target them better in the future. Lastly, I think the most important thing here is that we're going to be able to take a subsequent slide and do something called image mass cytometry, which is a very high-level immunohistochemistry, where you have different antibodies attached to different metals.

You can get up to, I think our panel is up to 40 different protein antibodies now. So you can actually validate these different immune cells or immune expression profile of immune cells that we're seeing on the transcriptomic level, on the protein level. So you can see CD3 cells, you can see NK cells, and CK5/6 to differentiate the different subtypes along with the HLA-E and CD49 in this image. So you can very nicely look at these different clusters where you can see that there's high HLA-E, high CD49a, in the area of where you see either basal subtype or luminal subtype, et cetera. And this is one representative image from one tumor so we really can't say a lot about this, other than it's going to give us a lot of information, but hopefully, with the entire cohort state, we'll be able to get a lot of very high-level data.

So in conclusion, I think that the TME highly influences the immune cell function. Whether it's endogenous or exogenous, whether it's being releasing cytokines to recruit different immune lineages or whether it's the manipulation of the transient immune cells that we need to better understand, in order to be able to target or manipulate BCG therapy. And so the subtypes, I think, are going to be very interesting because we've looked at subtypes at the level of bulk sequencing or bulk RNA, which doesn't necessarily tell us exactly their activity within the cell. So here we're going to be able to look at the specific subtypes and how they actually interact with the environment. And then for me, this is very interesting because similar to some of the other work that we've done in the past, I think subtyping is a very active process where these tumor cells can change very quickly.

And so to be able to look at this in a much more refined level, I think, it's going to give us a lot of understanding of tumor cells. And of course, the new technologies like spatial IMC are going to be game-changing in the efforts to help understand the mechanisms of action of either BCG, BCG resistance or these non-muscle invasive tumors. One thing that I would be really excited about... To use this in the future or obviously for the variant histologies, for things like microcapillary and just to understand why are those tumors cells so different than traditional urothelial or the other variants. So I'll stop there. And I want to say once again, thank you for the opportunity to present this and thank you very much, Dr. Kamat.

Ashish Kamat: Great, John. And again, once again, congratulations on finding a question that you're passionate about and then finding the appropriate tools to investigate. And like you said, you have these assays now that you've put together in such a way that you can understand the whole TME. You can understand what's going on, not just in snapshots, but like you said, you're pointing to do this in the longitudinal sequential, same patient cohort with different tumors. That's great. I know it's early and you have some preliminary data on this, but do you have any thoughts on how this will interact with what's now also another hot topic, which is the microbiome and how you might incorporate your studies with understanding the microbiome and how that interacts with the TME?

John Sfakianos: Yeah, so we've actually had a lot of conversations about the microbiome and I think to... The point is the microbiome is going to influence what the memory or the activity of the endogenous immune cells, and even actually the exogenous. Because if there is exposure to certain microbiomes or certain bacterium or even viruses, we have a very interesting story around different viruses that we've been exposed in our lives and how that actually builds memory for our immune cells in our immune system, that actually help strengthen their interaction to BCG when exposed. And so to that point, I think the microbiome is actually going to be very important for education of the immune system and the endogenous immune system and that's what allows some responders to respond better and so forth and so on. So, that's really exciting. We are starting to do some microbiome stuff as well to try to bring this together.

Ashish Kamat: And just one more question before I let you escape back to your busy day, are you also using this to understand in upfront patients? Not one's that have had recurrences, but in upfront how this might correlate with actually response and hence maybe a counseling tool or an assay for predicting who might do well with immunotherapy?

John Sfakianos: So our goal is to actually do this in BCG-naive tumors in the cohort of resistant and sensitive patients. So we can actually have the upfront data in the naive setting to allow us to help get some biomarker data. And what we do and what we have been doing is using this data, the transcriptomics and the protein data to guide us in making a different analysis called Olink, which allows us to look at 96 analytes in serum and urine. And so using the transcriptomics, you can actually make different series or different sets of Olink panels, which will then allow you to look in the urine and look into the blood of patients.

So we actually have some interesting data on Olink, which combines a lot of the work that you've done in the past, which is combining a lot of the different cytokines that we think we're seeing, that it may be important for resistance upfront. So we are getting those urines, so we're combining it all together to actually help better understand which patients, which ones will benefit from BCG and which ones want to help for counseling. I think we're a long way from there, but that's the end goal. The end goal is to really get those two buckets and say, "We think you should not get BCG. Let's figure something else out for you because you're likely going to be resistant."

Ashish Kamat: It's great to see this work coming out from your lab, John. I mean, this is great. Just a few years ago, I had put together a panel of international experts at the request of the EAU and we went deep down to figure out how or when we can actually predict who will respond to BCG. And looking at all the data that was out there just three or four years ago, the only things that we could use were great in stage. So we certainly are a long way from it, but hopefully, your work will get us there quickly. Once again, congratulations and looking forward to seeing you in a few days and hats off to you.

John Sfakianos: Thank you very much. See you soon. Thank you again.