But the idea here is when you do these single-cell sequencing experiments, essentially what happens is that you can actually resolve all the different cell types within the tissues. And the other component is, rather than us defining what cell types are there, we actually let the transcriptional profiles of the individual cells define what cells are there. And so when we did this and looked at differences between localized prostate cancer, hormone-sensitive prostate cancer and castration-resistant disease, we could actually look at how the cellular compositions might actually change over time. And the area that we found that was most changed actually is in the myeloid compartment. And one of the things to be aware about is myeloid cells have historically been very hard to characterize, because there are so many different markers. But when you do these single cell types of techniques, you can see with the different colors that are there, different populations that we can actually identify in that context. And so one population that we drill down on is these tumor-associated macrophages, or these TAMs, that express high levels of SPP1.
This is the gene that actually encodes for osteopontin, but in this case, we're actually using it as a way just to mark a specific myeloid population. And a couple points just to highlight is, if you look, these heat maps are basically showing relative abundance, these SPP1-high TAMs being in this dotted line, as prostate cancer progresses, actually the frequency of these myeloid cells actually go up. And so this is just a quantitation of this in this dotted box. But the other thing that I want to point out is that if you look like within the TME, they could represent about 20% of the myeloid cells within prostate cancer. And so we could actually see whether or not these cells are suppressive, and I'm not going to go into all those gory experiments since this work is published, but one way just to look at a high level is to apply an immune-suppression gene signature. And you can see these SPP1-high TAMs are suppressive, but just to highlight that there are also other myeloid cells that potentially could also and probably do contribute to the suppressive TME. Now, what type of mechanism could these cells suppress by? And again, one of the mechanisms that we identified in the context of these myeloid cells are molecules that are relevant in the adenosine pathway.
And I just want to point out, if we actually measure the amount of extracellular adenosine, you can see with these myeloid populations actually measurable levels in the supernatants, and we know that this is potently immunosuppressive. And this is actually a targetable pathway, in that we have drugs that can hit different nodes. And so we actually treated patients, again, with atezolizumab, this anti-PD-L1 antibody, in combination with, in this case, an inhibitor to the receptor for adenosine A2A. And you can see now we're actually seeing some clinical responses when we combine these together. And this is just an example of one of the clinical responses. But I think the important part, if you look at this, I mean, is this a glass half empty, or a glass half full? It's really not these impressive waterfalls that we've seen so far. But an important aspect was, if we actually looked at the biopsies from these patients, comparing responders to non-responders, the patients who were responding had a high frequency of these SPP1 high myeloid cells, whereas the non-responders did not. And so what we would argue is that, in these patients, these non-responders, there's probably some other mechanisms at play that are mediating therapeutic resistance.
Now, we wanted to think about ways we could combine this. And obviously, thinking about, this meeting, a trial that we performed with Tom when I was at UCSF, combined lutetium with pembrolizumab. And in this case, what we did was we actually just administered a single dose of lutetium PSMA, and then we looked at some different schedules of adding on an anti-PD-1 antibody. And what we had published a few years back is, even with this single dose, you can actually see some responses that actually could be quite durable over time. And we also described immune responses in these patients actually associated with this improved clinical outcome. And so now what we've been doing is actually interrogating biopsies from these patients. And I just want to highlight, this work is led by Anusha and is unpublished, but if we actually look in biopsies, these SPP1-hi TAMs sort of in this dotted line here, you can see in the non-responders, in this case with a lutetium PSMA and an anti-PD-1 antibody, the non-responders actually have high frequencies of these suppressive myeloid cells, whereas the responders didn't. Again, suggesting that these cells are important, and we should think about how we could target that. Now, in the final couple of minutes of my talk, I wanted to actually touch upon another immunotherapy modality that Shahneen also discussed in terms of these T-cell engagers. And as many of you are probably aware, we now have multiple T-cell engagers that have significant clinical activity in men with prostate cancer. And so this is one of them targeting STEAP1.
This is xaluritamig. And so one end binds STEAP1 on the tumor cells, the other end binds CD3 on the T cells. And the way I describe this drug to my patients is it's kind of like Velcro. You're sticking the T cell to the tumor cell with the goal that those T cells will fire and kill the tumors. And this is an established therapeutic modality, primarily in hematologic malignancies, but now in small-cell cancers with a DLL3 T-cell engager. And so with this molecule in phase 1 clinical trials, you can see the waterfall plots, a significant proportion of patients with PSA responses, as well as with RECIST responses. And you can just see like the percentages here at this high dose, 60% of the patients having a PSA-50, and at high doses, somewhere on the order of 40% of the patients actually having partial responses. And this was a patient that I treated on this trial when I was at UCSF, just to make this point in terms of the types of responses we can actually see, which are actually, can be pretty deep. But with these T-cell engagers, because we're turning the immune system up to 11, we basically have toxicities associated with it, that we call cytokine release syndrome. And this is the CRS that's highlighted there. This is just a tornado plot showing side effects associated with this xaluritamig.
And you can see cytokine release in the vast majority of the patients who are treated this way. And so these create challenges in terms of pre-medicating the patients and monitoring them, and being able to manage these toxicities as they occur. And so, this has sort of led to some next-generation T-cell engagers of which this VIR-5500 is one of them, and Johann de Bono presented this data just at ASCO GU a few weeks ago. And the idea around this is rather than having a molecule that can actually bind the tumor antigen and the T-cells, what if we masked both of those domains, so that they couldn't bind anything until this drug got to a tumor microenvironment where there would be proteases that could cleave those masks off. And so with this drug, what was reported at ASCO GU is, again, high rates of response, but importantly, very low rates of toxicity. In fact, the typical immune suppression that we have to give as pre-medication for these T-cell engagers, we don't have to give with this drug. And we now have this clinical trial open at the Fred Hutch, and actually some of you in the audience have been referring patients to that, and so we're looking forward to gaining more experience on this trial. So conclusions, immune checkpoint inhibitors by themselves have limited efficacy in prostate cancer. Immunosuppressive myeloid cells can restrict the activity of immune checkpoints.
Targeting myeloid-mediated mechanisms of resistance can sensitize some patients to this approach, and T-cell engagers can have significant clinical efficacy in prostate cancer. And I think the hope is by targeting these different mechanisms of therapeutic resistance, we can really realize the full potential of immunotherapy and prostate cancer. So, I'll close there. Thanks.