Thank you all for coming tonight. I know it's Friday night and you're delaying your dinner for a short time to hear about Corvus. We presented two really interesting presentations today that I want to review with you. This meeting is webcast. I also appreciate the people staying up, listening to us online. Appreciate that attention. We're very pleased to have with us this evening Dr. Jason Luke from the University of Pittsburgh, who just delivered a marvelous paper just a couple of hours ago, an oral presentation on our anti-CD73 antibody called CPI-006, which I'll get back to in a second. We also have members of the Corvus team here, Dr. Mehrdad Mobasher, our CMO, Dr. Suresh Mahabhashyam, our clinical director, Dr. Stephen Willingham, our senior scientist, and Drew Hotson, also a senior scientist.
Drew and Stephen are pioneers in the Adenosine Gene Signature, which is now, no question, being used by everybody. These guys deserve a lot of credit for the innovative and creative and hard work they've done on that, which is being, I think, confirmed by many groups. What I'd like to do tonight is I'm going to go a little off script. I would like to start by reviewing the poster that Dr. Willingham presented, and then we'll talk a little bit about Dr. Luke's presentation, the oral presentation. I want to start with the conclusion. There are two conclusions from today's work that I want you to remember. Number one is that we have a biomarker called the Adenosine Gene Signature that is very robust, very easy to measure, and appears to be successfully predicting who responds to adenosine blockade.
Furthermore, that predictive biomarker picks out patients who don't respond well to the standard therapies. Let me repeat that. We have a predictive biomarker that predicts who responds to our treatment. It predicts against who responds to other standard therapies, setting up beautiful potential clinical trial opportunities. That's the first thing I want you to remember. That's the first conclusion. After the next conclusion, you can leave. The second conclusion is that CPI-006, the anti-CD73 antibody, is a very unique antibody. There are a lot of companies with anti-CD73s, and they all block adenosine. No one, to my knowledge, has an antibody that blocks adenosine and activates this CD73 ligand as a stimulation of the immune system. The immunomodulatory stimulatory properties is something that's unique to 006, to my knowledge.
When you think about that, I know you've all heard about the analogy that checkpoints or inhibitors are about taking your foot off the brake. The immune system has a brake. We take your foot off the brake. For years, people have been thinking about putting your foot on the gas. 006, and here's the conclusion, 006 does the following. It takes your foot off the brake by blocking adenosine, and it steps on the gas by its immunomodulatory effect. I do that when I drive, and I have a Porsche. That's really cool when you can do that. It's true. I have a Porsche.
Again, those are the two conclusions, a predictive biomarker that picks the patients who respond to our therapy and picks against those who respond to standard therapy, and an agent with dual properties that are ideally tuned for the immune system, taking your foot off the brake and stepping on the gas at the same time. You can all leave now. No, don't leave because I want to show you some of the information behind that. Really quickly, what I wanted to do is run through just the very important points, particularly about this Adenosine signature and renal cell cancer and prostate cancer, because there's been some, I think, lack of clarity about this. Let's go to the next slide. Oh, I guess I can control that.
This was the title of Dr. Willingham's poster, Adenosine and AMP stands for adenosine monophosphate, Gene Expression Profiles. That's what this poster was about. I won't dwell on this. This has been presented at other meetings. It is in press now in Cancer Discovery. It will be accompanied by an editorial, I'm told. This is the set of genes that Dr. Stephen Willingham and Dr. Drew Hotson identified. Now, this is about 12 or 13 genes, and I won't dwell on them. We've got this down to eight that have very good clinical correlations. These eight genes are interesting. Interesting because they're myeloid cell-derived. Most people were thinking T cells, but it turns out, as you're hearing at meetings more recently, myeloid cells are very important in the tumor microenvironment. Stephen did this very simple experiment. He took white blood cells from people.
He threw in adenosine. He said, "What genes are expressed that are associated with the immunosuppression?" They're myeloid genes. Now, interestingly, right around this time that we were finding this in the laboratory, Genentech published a paper, published in Nature Medicine 2018 in July, I think, where they, in a randomized phase II with SUTENT, atezolizumab, atezoli, Avastin, described a gene signature which they called the myeloid gene signature, which in this study was a biomarker of lack of response, non-responsiveness to atezolizumab. atezolizumab is an anti-PD-L1. It turned out that the myeloid gene signature and the Adenosine Gene Signature are identical. They did this independently, not thinking about adenosine. They treated patients in a randomized trial, 305 patients. They identified a set of genes that had prognostic value, and they found eight genes identical to Corvus' Adenosine Gene Signature that predict a bad response.
Let me show you an example. I'm going to try to keep this simple. This is their progression-free survival curve for myeloid low. Think of that now as Adenosine Signature negative, adenosine low. In that group, atezolizumab, which is the gold, has a PFS up here, 17-month median. Okay. If you were myeloid high, which is adenosine high, the PFS was 2.9 months, the median PFS. 17 months to 2.9. You don't need to be a statistician to figure that out. The hazard ratios are, instead of being one point something better than SUTENT, instead of atezolizumab being better than SUTENT, it was worse. Okay. There's a lot more details in here that I don't have time to get into. The bottom line is the myeloid signature is identical to the Adenosine Signature. If you got that, you don't respond to a PD-1.
In fact, the response rate in that group, which isn't in the paper, but deducing it from hazard ratios and so forth, I figure it's around single-digit numbers, if that. That's being optimistic. That's actually not in the paper. myeloid gene signature high, Adenosine Gene Signature high is a bad prognosis for PD-1. Is that clear? Drew Hotson being as smart as he is, he goes to the TCGA database, and he asks the same questions. What happens if you apply the Adenosine Gene Signature to these patients, 350 patients or so, actually 500 patients. Look at the survival curve. If you're Adenosine Gene Signature negative or low, you have a much better survival than if you're positive. The P value on there is 10 to the minus 4, approximately.
In the TCGA database, where people are treated with everything, we don't know what they're treated with, but they're treated with whatever the standard of care is. If you were Adenosine Signature positive, you did statistically significantly worse in terms of survival. Genentech shows Myeloid Signature high, you do worse with PD-1s. TCGA says you do worse. Conclusion, it's a bad prognostic indicator for standard therapies and for PD-1s. Adenosine Signature, which we now call the AdenoSig, positive patients are an unfavorable group. It's not that we're picking better patients. We're picking worse patients. They are worse patients. Let me say it a third time. They are worse patients. Okay. They are more likely to respond to ciforadenant, our A2A antagonist. Okay. Let's go through the data. This is not frontline therapy.
This is third, fourth, fifth, sixth-line therapy, patients with renal cell cancer. This is a waterfall plot on about 32 patients where we've done the signature, comparing the Adenosine Gene Signature high versus low. You can see in the high group, whether it's monotherapy or combination therapy, that's where all our responders are. When I say responders, I mean tumor reduction. Even if you look at it by RECIST criteria, it's 17% versus 0% in the other group. Okay? Seventeen versus zero, you can do the statistics on that. It'll be significant. In fact, it's 0.008. What do I mean by that? Just to look at some examples, you can take off your glasses to see this. This is one of the patients, actually. There's a mass that goes from the mediastinum to the chest wall. I'm an oncologist, and so is Jason.
When you have a mass that goes from the mediastinum to the chest wall, that's not good cancer. That's a bad mass. This patient has been on treatment for a couple of years now and has almost complete, if not complete resolution of that mass. To summarize here, these are patients who have median three prior treatments. 85% were resistant to IO, not all of them, and I'll show you details in a second. The Adenosine Signature statistically significantly predicts responders to our treatment. Right? Adenosine Signature high, Myeloid high, those are bad patients. I already showed you that with two examples, Genentech randomized trial, prospective trial, and historical from the database. Adenosine Signature patients are bad with standard therapies, but they respond to our treatment.
That has to do with the molecular mechanisms of action of the drugs. Let's look at who these patients are a little more carefully. This is the Adenosine Signature high versus low from our study. This is the progression-free survival curves. We like to see everybody cured, but we get particularly interested when we see lines like that. Lines on PFS curves or plateaus mean maybe you are curing some people or maybe you're at least producing long-term control. There's not a continuous decay. What's interesting here is that all of those patients, all the long-term responders or disease control are Adenosine Signature high, and of course, the lows, that curve goes down to zero. Compatible or consistent with what I've just told you.
In our publication that's coming out, we also look at survival in these patients. We have very good survival. Who are these patients? There are six patients on that tail. Are they good patients? Did we just by accident come across really good patients? Marathon runners who just got diagnosed yesterday with renal cell cancer? No, we didn't. I already told you that the Adenosine signature is bad, these patients, if you look at them, some monotherapy in the solid green, some combination with atezo. At least four of them were on an anti-PD-1 and had PD on it. By the way, I don't have the times of the exact PDs, it has been almost a rule. These patients are on PDs and progress well beyond their PD because doctors usually have nothing else to give them, they progress more.
Anyway, you can see that all of these patients, it was a requirement that you had to have PD before you got on the trial. My point here is that these guys had failed the PD-1. Even the ones that didn't fail, and I'll show you an example of that in a moment, had PD coming onto the study. What we're interested in is, as we track the longevity, the durability of the response, does it exceed their last treatment? This is something we did with Rituxan that was really cool. This really was a key thing in Rituxan development. I was there for this. We compared early on how long was the remission with our Rituxan versus their last therapy. The general rule in oncology is every subsequent therapy gets shorter duration. It's not absolute, but it's usually true.
We're looking at these patients with the expectation that we'll start to see very long durability on that, and our PFS is telling us that. Okay. Well, here's one of the patients now. This is the mass I showed you before, but here's what's interesting about this guy. Well, first of all, it's interesting that a mass that big goes away. I don't care what your treatment is, that's a good result. That's big-time disease. What's interesting here is that this patient was diagnosed in April of 2016 and starts on our protocol in July 2017. What's that, 15 months or something? That patient has been through sunitinib and axitinib. When you have a tumor grow that much and fail two lines of therapy in one year, that's not a good tumor. That is not a good tumor. You would agree with that.
It's Adenosine Gene Signature positive, and this guy goes on ciforadenant and atezolizumab and has nearly complete resolution of that mass. That's a very good response. A very good response. Okay. This is a paper that just came out last week in The Lancet Oncology, and I just put this up here for some comparators. This is axitinib in renal cell cancer after you've failed IO. A new disease has emerged. What does renal cell cancer look like after you fail an IO? By the way, they're all going to fail it. It's like we didn't know about penicillin resistance until we had penicillin. What this paper showed from Brian Rini, who was the senior author, was this 40 patients, 76% had failed two or one therapy, an IO plus something else, and about half of them were one prior therapy.
The median PFS was 8.8 months. Not that good. It drops off pretty quickly. ORR was pretty high, though, 45%, which shows another thing about immuno-oncology, and I don't know why people aren't understanding this because ORR is good. We all like to get CRs and PRs, but durability's important. These are not durable. The same thing is true of chemotherapy, by the way, in lung cancer, renal cancer, and things like that. You get good response rates. The response rate to TAXOTERE in lung cancer is just as good as atezolizumab or pembrolizumab. The problem is it doesn't last, and what makes pembrolizumab and atezolizumab good and why it got approved for second-line lung cancer, because of the tail on the curve. Now, what we want to do is, of course, keep increasing the tail. Now here's what you need to understand about this curve.
It's bad enough. There's no tail on this curve. Okay? You have to realize this is adenosine signature positive and negative together. If we were to do a trial against the adenosine signature negative, which, by the way, I should have mentioned, is about 50% of patients. The breakdown of adenosine positive, negative, or myeloid positive, negative is about 50%. It was 49% in the Genentech study. It's about 60% in our hands. That curve is worse. Okay? It ain't going to get better. It's worse. Okay. The summary of the adenosine gene signature work is the adenosine signature, what we call now AdenoSig, and the myeloid signature are identical. By the way, that predictor has now been confirmed by two other large pharmaceutical companies, and I understand today at our poster, by several small biotech companies.
They're going to try to modify it and tell you that it's better and stuff like that, but basically, we have laid the foundation for that. Adenosine Gene Signature high patients have a poor response to PD-1, PD-L1 therapies. These are the conclusions. Adenosine Gene Signature and myeloid gene signature are identical. They predict poor response to PD-1s. AdenoSig high patients have a worse survival. That's the TCGA. AdenoSig high patients most likely to respond to ciforadenant. That's the waterfall plots I just showed you, which by the way, over time is holding up. We have yet to see a response in an Adenosine Gene Signature low patient. Adenosine Gene Signature high patients have prolonged progression-free survival versus low when treated with ciforadenant. Those are the conclusions. That's the key takeaway from Dr. Willingham's poster. The Adenosine Gene Signature is not limited to renal cell cancer.
We have looked and are looking at other cancers. This is again, data from TCGA, and this is the Adenosine Gene Signature. Each dot, of course, is a patient biopsy. Here are the cancers on the bottom there. As you can see that some tumors are high, some tumors are lower, some are in between, et cetera. Renal cell cancer is in between. Where is renal cell cancer? Right here. Said it's 50/50. Some are higher. Where is esophageal, interestingly, is very high. Gastric's pretty high. We intend to apply this signature to other tumors as well. Now, where the cutoff will be and things like that remains to be determined. There very well could be different cutoffs and thresholds for what's positive and what's negative, but it's not limited to renal cell cancer. Okay.
I want to talk about one of the other tumors that we're interested in besides renal, is prostate cancer. Prostate cancer has the Adenosine Signature. It's about 50/50, roughly. We have been treating everyone. We're collecting Adenosine Signature data, we're still not gating on it because we need to and want to collect negative data as well. This is a patient, again, failed their usual hormones and chemotherapy. I just want to show you some examples of what we're seeing in prostate cancer, and we will be presenting a paper at GU ASCO in February where we'll update this, I think maybe have 20 patients or so. Here's a patient who had a PSA. This is before treatment. His PSA is rising.
It goes up to about 100. He starts on this treatment here. It promptly falls over several cycles, goes down to less than one. Concomitant with that drop in PSA is a very large 6-centimeter peri-aortic mass that goes away. It's almost normal. Here's another prostate cancer patient who has also an elevated PSA rising to about 15. He starts on treatment. PSA drops to under five. PSA of around four is normal, if you're on hormones, it really should be close to zero. Here's while on treatment, this is a man who's still on treatment, getting treated right now. I think he's only on his third cycle, he already, after two cycles, has, Oh, sorry, I got to go this way, a reduction in peri-aortic mass, peri-aortic adenopathy. Compare that to that and that to that.
Tumor has three dimensions, so you're cutting through it in two places. That's prostate cancer. In the biomarker-defined population, Adenosine Gene Signature positive patients, the response rate in third-line or more renal cell cancer by RECIST is 17%. That's third, fourth-line treatment, and we'll do better than that in the disease control. People who don't quite make RECIST criteria is way more than that. Half the patients are getting some kind of response, and the PFS has a tail. That's the message. In prostate, we're going to look at the same thing. Okay. CD73 blocks the conversion of AMP to adenosine. Dr. Willingham and Dr. Hotson, who are far smarter than I am, asked the question, well, if you block AMP, what does AMP do? I'll skip a lot of the details, but this is a NanoString assay of in vitro, again, on blood.
What Stephen and Drew did was show that basically AMP generates the Adenosine Signature. How does it do that, they asked. It turns out that AMP can bind to the A2A receptor. If you throw in CD73, it doesn't block the Adenosine Signature from coming out because you still have a backup of adenosine. Because AMP can also bind to the adenosine receptor, albeit at a slightly lower affinity. If you put ciforadenant in there, you block it. The conclusion, and I won't go through all the details of this, the conclusion of Stephen and Drew's work is that if you just block CD73, you'll get a buildup of AMP. AMP can do what adenosine does based on its gene expression profile, and ciforadenant blocks AMP and adenosine. It's not surprising. Ciforadenant is sitting right in there in the pocket.
It's a very, very potent drug. Very selective and very potent. Okay. That means you want to give both if you're using CD73 as a blocking agent. Of course, our CD73, I told you, does a lot more than that. It steps on the gas as well as taking your foot off the brake. Okay. I'm not going to spend time on the conclusion. I think I've drilled it quite a bit. I do want to now turn it over to Jason. Jason, I think you can whip through, because they just heard your talk a couple of minutes ago. Just focus in on a couple of things, and if you don't, I'll correct you.
All right. I'll speak loud, turn on the mic so it picks up. I'm just going to go through the talk relatively quickly because we want to have time to go ahead and ask questions.
Exactly.
Hopefully everyone can hear me now. Quick, I don't want to really go over the background here. We presented data at ASCO. There we go. Okay, sorry. We presented data at ASCO describing some of these immunomodulatory effects and initial data to suggest that the immunomodulatory effects were independent of the adenosine axis. It dovetails in with what he was talking about. We're blocking the adenosine axis, but additionally, there are these immunomodulatory effects. Really, what we had shown before that's important is that we saw activation markers go up on B cells, and we saw very interesting patterns of lymphocytes trafficking in and out of the peripheral circulation.
Those were very interesting findings, and we've continued this clinical trial and continued to study those phenomenon, both as the monotherapy to block this interaction, but additionally to facilitate some of these immunomodulatory effects and in combination with ciforadenant. How have we updated this? We've looked more at some of the effects on B cells, which is to say that we see impacts on B-cell activation. This is looking at a marker of activation comparing CPI-006 to another in-class CD73 inhibitor, and you can see that it does not do this. Okay? Over time, you see an increase in a number of markers on B-cell surface that are associated with activation of B cells, and this eventually leads to B cells becoming plasma cells and secreting antibodies into the circulation, which presumably could also theoretically have anti-tumor effect.
The clinical trial is ongoing as of monotherapy, combination with ciforadenant, as well as there are future cohorts that are described here. I think it's important to point out this trial is predefined with those expansion cohorts in the tumor types that were just alluded to. This is important when you see the results a little bit later. We are specifically interested in these populations relative, potentially, to the total population. It's a phase I clinical trial, pretty standard. The drug is given every three weeks, as well as the ciforadenant. We have not integrated the biomarkers yet that were just described. May do that over time, but just to facilitate the dose escalation, have not done that yet. These are patient characteristics. The thing to point out here, four lines of prior therapy in both the monotherapy and the combination arm.
If you look at the tumor types that have been included, they're predominantly refractory GI tumors in the dose escalation. These are tumor types, generally speaking, where we have not seen impact of immunotherapy, so keep that in mind when you consider clinical results early in dose escalation. Including, however, PD-1 refractory renal and non-small cell lung cancer. These are data somewhat similar to previous data that we've presented, again, showing that the serum PK is dose proportional.
As you go up, you get more exposure, that you can engage the target in the peripheral blood on CD73 in peripheral B cells at doses in blood above about six milligrams per kilogram, and that we're engaging the target in a specific fashion in the tissue using this cross-blocking antibody, such that at doses of 18 milligrams per kilogram, we think we're getting to the dose that should completely infiltrate the tumor and block the target. These are some of the effects that we saw in terms of these changes in B and T cells in the peripheral blood. There appears to be a CD73-dependent impact of this. These are B cells from patients that are CD73 positive. When these patients are exposed to CPI-006, we see very quickly a rapid reduction in the peripheral blood of these cells.
This does not take place in the CD73 negative population where nothing happens. These B-cell dynamics are characterized as an early drop and then a return, basically to a near return to stable state. This is a stable state, but not as quite to the baseline. We see similar changes in T cell dynamics, which we have not studied as much, but here we see the drop, and then they come back to the steady state that they were previously at. Interestingly, seeing those changes, then we're interested in what do those cells, what do they look like when they come back? They were regular-looking B cells. What do they look like when they come back?
To do that, we looked at these B cells and then looked at differentiation in class switch markers from pretreatment to early on treatment, where you can see that basically everything starts to go away in the peripheral blood, and then on treatment, where we see an increase in staining for CD27 consistent with this class switch memory B cell, so starting to be able to make antibodies. In terms of looking at changes over time, we see that on treatment, eventually we see more memory B cells becoming present, again, consistent with the potential to make antibody. I think perhaps maybe the most interestingly is that when we look, so there's a lot of these changes. They come back, they're phenotypically different, and the question becomes, are they clonal cells that are coming back?
In other words, have they seen an antigen, and are they coming back looking for that antigen? To do that, we look from pre-treatment to on treatment, so week six, of changes in the immunoglobulin structure. These are a little bit complicated kinds of graphs to look at, but simplistically, this is baseline, this is on treatment. What we see is that on treatment, we see this differential abundance of new B-cell clones at very, very high frequencies. Up to one in 100 B cells in the peripheral blood, which, if you're not super into immunology, is an astronomically high number of B cells in the peripheral blood that are specific for one thing. We've seen this actually in the majority of patients so far.
This is a very interesting phenomenon that I think in and of itself has a lot of potential for discovery, both just in the general scientific sense as well as in the drug sense. This phenotype has not been associated with any serum immunoglobulin changes. This is important because if you were making a lot of memory cells, you could start getting hypergammaglobulinemia, but we don't see that. We think that maybe that's because the antigen is not ubiquitously present once these cells come back to the circulation. These are outcomes data of patients treated on the clinical trial. What you can see here is the doses going down this way from six to 12, 18, 24. Remember that at 18 milligrams per kilogram is the recommended phase II dose.
Really only below here are the patients getting the dose that we think actually probably has the potential to truly infiltrate the tumor, and make a big difference. Again, be cognizant of the fact this is a dose escalation study. There's a lot of refractory colorectal cancer in here, which these patients are participating, but these are not the population we're necessarily looking for. In that regard, in the pre-specified disease settings, remember of renal cell, prostate cancer and lung cancer, here we do see tumor regression in these patient populations. This is a relatively early look. These patients haven't been on treatment very long, and there aren't that many patients who have been treated yet at a dose we think is likely to be effective. We have, however, seen clinical evidence of activity.
This is a case report of a patient with prostate cancer, got monotherapy. You can see they had tumor reduction, and while that may not look like the most amazing thing, in prostate cancer, we don't see that much tumor on a CT scan. Rather, patients have bone lesions, and we follow the PSA number. This guy was in pretty rough shape, on a high amount of narcotics, and very quickly after starting the drug, had a tremendous improvement in performance status and reduction in his pain syndrome. That PSA number has basically been about stable throughout. This guy's been on treatment now 19 cycles. Similarly, a patient with kidney cancer got combination with ciforadenant. He already had had TKI for VEGF, multiple lines of immunotherapy, lung metastases here, gone on treatment.
This patient also has continued on treatment and just continues to do very well. We're very excited about this, again, in these diseases, which we already know adenosine blockade can be an important part of the treatment. These data all sort of indicate this phenomenon of an immunomodulatory impact of CPI-006. There are cells floating around. They're exposed to this. We totally don't understand how this works yet, but you start to take on this activated B-cell phenotype, and due to other data that we don't show here, we think probably these cells leave the circulation to go into lymph nodes, where they continue to differentiate either into plasmablasts or come back out as a memory B cell, and such that over time, there's the potential, like I said, to start to look for antibodies that these cells might be making.
These are the conclusions. CPI-006 obviously blocks conversion of AMP to adenosine, but additionally, independent of that, has these immunomodulatory properties. Inducing B cells, class switching, secretion of immunoglobulin, and generation of memory cells. We see that exposure to the drug causes increase in activation markers, CD69 and others, consistent with antigen presentation. Take-home message here, the dose that we think is the recommended phase II dose is 18 milligrams per kilogram. There haven't been that many patients treated at that dose yet. We're excited to see what the data will start to look like as patients are getting that efficacious dose. We do see these rapid changes in peripheral blood for T and B cells, and when they return to the circulation, they have a different phenotype and have expanded now in a clonal fashion.
We think this suggests an induction of adaptive humoral immunity, and we've seen tumor regression in patients with prostate and kidney cancer. Again, I'll highlight that these returning B cells after treatment with CPI-006 seem to have recognized an antigen. What is that antigen? Well, we're going to try to look and find out, and that really is a platform now to potentially identify novel, potentially anti-tumor antibodies, and that in and of itself is very, very exciting. Thank you very much.
Thank you, Jason. Perfect. I know Jason has to run to another meeting, but maybe we can get any questions? We'll open it up for questions now for Jason or me or anyone else. Don't be bashful.
Viren.
Yes, Viren.
Richard, a question. How did you identify the 18 mgs per kg dose as a go-forward dose? I think you had three patients that were treated with the 24 mgs per kg.
18 gave us complete saturation in the plasma on B cells and T cells for three weeks. We did biopsies on patients. When we look at the biopsy data, the 18 gives us complete saturation of the antigen in the tumor. We don't need to go to 24. We've gone to 24. It's not necessary.
Got it. The Adenosine Gene Signature, when are you going to incorporate that in the expansion phase?
We are going to incorporate that in the expansion phase, but I think, don't forget, this is now a different mechanism of action than CPI-444, so we're going to have to look and see what the relationships are. We're going to do Adenosine Signature now on everything we do, basically. This is a different drug, and different tumors and things will need to be. Everybody else is going to be doing it. You can actually get the Myeloid Signature, I think, starting in January. You can call up Foundation Medicine and get it. I know doctors are already doing it with renal cell cancer. Again, Corvus has set the standard. Yes.
Yes, just another question on Adenosine Signature. Looking at your TCGA slide, the prostate cancer is not really remarkable as far as standing out on the Adenosine Signature yet. I think the prostate activity with some of these drugs is one of the more compelling things, given PD-1 monotherapy in that indication. Just curious how you rectify that.
Well, let's look at your statement there, because I might challenge that a little bit, if I may be so audacious.
Unless I just was looking incorrectly.
No.
Prostate adenocarcinoma there.
Where's prostate? I got to find it. Here it is. Okay. Here's your signal. This is the line. That's the cutoff for kidney cancer. The cutoff for prostate cancer, we don't know where that is. It's analogous to PD-L1. What's the PD-L1 level you should look for in lung cancer? Some people say 1%, some people say 5%. It depends on your assay and so forth. I wouldn't draw any conclusion on this. We put this line here just because that's what we know for kidney cancer. Where'd it go? Prostate cancer, what's interesting is there is a pretty big spectrum, and I don't know what percentage are going to be high. Maybe the line is here. Okay. Prostate makes perfect sense for many reasons.
That's why you're asking the questions because of prostatic acid phosphatase, and there's presumably a lot of adenosine being made in the prostate for reasons that are not clear. Now, whether that is the reason why this treatment appears active in prostate, who knows? Be careful where you draw your cutoffs. In the biomarker game, where you draw your cutoffs takes some work.
Got it. Thanks.
I would challenge that assumption. You would agree, right?
Yeah.
Okay. My guest speaker left. Okay, I guess it's all right. Yes.
This is pertaining to the Adenosine-positive patients. Clearly, something positive is going on for these patients receiving a support, and to me, it looks like there's a need for T cells that need to be activated and directed to the tumor site, say perhaps to get a high response rate. You have shown us that these patients do not respond to PD-1s well, but have you considered adding PD-1 with ciforadenant and see what happens? Any thought, hypothesis, translational data that you can share would be immensely helpful. Thank you.
Yeah. Well, first of all, the trial we have ongoing now, Tosh, is it? The studies we're doing now is combination of atezolizumab in renal. The ongoing study is the atezolizumab plus ciforadenant. We're doing that, the idea being to activate T cells and so forth. You're correct, maybe the treatment is limited by insufficient T cells in tumors. That's a possibility. It looks like the Adenosine Gene Signature identifies myeloid cells, and those myeloid cells actually secrete things that keep the T cells out. It could be that A2A blockade is removing that function. Okay. Anyway, to summarize, we are giving it with a PD-L1 in this case. Look, the reason for doing monotherapy was not to prove that monotherapy is better than anything. People always have that as a misconception. We weren't trying to prove that. We were developing a drug.
We are developing a drug, and you need to give it as a monotherapy. I know you got a lot of other companies who you follow, they go right to combinations, and they're never going to learn anything, unless it's knocking it out of the park, and even then, they're not going to learn anything. The reason we know our biomarker work is from the monotherapy work. If I give it together with another drug, then the biomarker work becomes useless, your safety becomes useless, and your efficacy becomes useless. I don't know why. As an oncologist, this is Oncology 101. Combination therapy is not a new idea in oncology. It started in 1957, I believe, with combination chemotherapy for leukemia at the National Cancer Institute. Maybe it was 1960.
It was taking drugs that you knew what the safety and efficacy of the individual components were, and you put them together. That's what you have to do. For us, it makes a lot of sense. The biology here is pretty straightforward. In Genentech's work, the myeloid gene signature, Adenosine Gene Signature predicts poor response to PD-1. Our drug works in that group. That suggests, doesn't prove, but suggests that at least one important resistance mechanism to anti-PD-1s is adenosine in the tumor microenvironment. It makes perfect sense to give both of those together because you're going to do your PD-1 and you're going to block the other resistance pathway. That's what we do. That makes perfect sense. That's a good question, though. All right, any other questions?
I did want to just, of course, some of the stuff on the CPI-006, just for those who aren't familiar with how the immune system works. You got these B cells and T cells. Actually, you got other cells too, I know. These B cells and T cells are kind of roaming around your bloodstream, and they're going in and out of lymph nodes and spleen and stuff, and they're your kind of scouts. The bloodstream is just a highway to get from one place to another. The lymphocytes leave the bloodstream, they go in the lymph nodes, then they get into the lymphatics. The lymphatics dump back into the bloodstream in something called the thoracic duct. They all kind of wind up that way. They're just roaming around looking for stuff to do.
Once in a while, if you had an infection in your finger, let's say, the macrophages and neutrophils carry that antigens from that infection to a local lymph node, maybe here, epitrochlear, axilla, under your armpit. When those lymphocytes go in there, and they find a fit to that antigen, they get activated. They don't leave. They don't leave because if they left, they couldn't do anything. They get activated, and then they stay there because they express something called CD69, which keeps them there. That's one of the markers that Jason referred to. CD69 keeps the lymphocytes from egressing. Okay? They get stuck in the lymph node if they're responding to that antigen. They proliferate, divide, and divide. A process, a very interesting process happens where the immunoglobulin genes mutate, and the B cells with a better fit get driven.
The ones that don't fit, poorer fit, they just go on and die. The better fit ones make antibody. They become plasma cells. They make antibody. Eventually, the antibody, of course, is soluble. That gets out into your bloodstream and goes everywhere. Eventually, the antigen goes away. There's no more antigen. You're cured, right? This is what happens when you get an infection, flu or something. Those B cells don't all just die then because if they did, you wouldn't remember yet you got that infection. We know that the immune system has memory. A portion of those cells that are becoming activated kind of go off back into the blood and go to other parts of the body, including other lymph nodes, and those are those memory B cells.
Memory, the next time you see that antigen, you're starting way ahead of the curve. You've already got a B cell with a much higher affinity. Instead of having one B cell, you got millions of them, and you can mount a more profound, faster immune response. That's the basis of vaccination. That's why you get a flu shot or any other shot. It's that the next time you see that antigen, you'll mount a response so fast the infection doesn't have a chance because it takes a few days for a virus to get going. If one virus goes in your body, you're not sick 10 minutes later. It takes a few days. That's how you get these illnesses prevented. This is called clonal selection. Macfarlane Burnet won a Nobel Prize for that, an Australian.
The B cell clones that have the best fit will be selected and expand. That's why we don't see hypergammaglobulinemia. You might ask the question, "Why? Geez, you're just activating every B cell, making antibody in every B cell. You just make a ton of antibody." That doesn't happen because there's clonal selection. Only the B cells that have an antigen that will continue to drive it will survive. What are those B cells reacting with? I hypothesize, we hypothesize, but don't have proof yet, that it's possible that those antibodies are directed against tumor antigens. It's also possible that the patient got infected with a virus this morning, and it's reacting with whatever happened to be getting into his body or her body. We don't know yet, but there's technology now that enables us to look for that.
We've got some preliminary interesting work where maybe we see serologically some signs of that. We know tumors have neoantigens. We know some patients have tumor regression, and we know we're having this profound effect on B cells. It's a reasonable hypothesis to look for antibodies. Those antibodies will be incredibly interesting. Antibodies are good drugs. I think you'd agree with that, right? Come on. We can talk about T cells and stuff and CAR T, but compared to an antibody? Come on. Antibodies are cool. They're great drugs, and they work if you have the right antibody. We think this could be very valuable. The T cells are probably important as well. Jason was asked the question. We haven't gone as far on the T cell story, just we haven't had time to do it.
We see changes in T cells. We see CD73s on T cells, subset of T cells. We see clonality, change in clonalities, but we just really haven't gotten to a point to check it out yet. Yes, another question.
One question. Yeah. You said AMP itself can bind adenosine receptors of one-
A2A.
Yeah. One I assume is A2A, not A2B.
A2B, Stephen?
It has both C and both.
Yeah. Okay. The downstream results are the same, cyclic AMP suppression.
That's what the Adenosine signature is.
Yeah.
That's how he found the signature.
Right.
Is he throws AMP in and sees the same genes.
The phosphate is just hanging out. It's not preventing the adenosine molecule from binding its receptor.
AMP is binding.
AMP is binding.
Yeah, because Stephen's smart, he has a non-hydrolyzable form of AMP where the phosphate can't. That's a good question. You could say, "Well, how do you know the phosphate's not being cleaved off?" He knows that because there's a non-hydrolyzable AMP where the phosphate doesn't get cleaved off, that does the same thing. It's able to fit in the pocket. We've looked at some molecular models. It's able to fit in there. The Ki for it is not as strong as adenosine. It's low enough. It's what? 100 nanomolar, 150 nanomolar? Yeah, that sounds right. Yeah. Yeah, you hear a lot of numbers about adenosine and all that stuff. Adenosine turns out to be really hard to measure. When people tell you they're measuring adenosine, unless they're putting a catheter in and doing it steady state over hours, do not believe any of that.
Just don't believe it. If they're even telling it to you, punch them in the nose for wasting your time. Understand, so it fits in there, just not as well. ciforadenant will outcompete it, obviously, depending on concentrations. Reversible inhibitor. Okay, any other questions? Yes, Matt.
Oh, why do you think your CD73 has this immunomodulatory activity, and you don't see it with other CD73 mAbs?
Because of the way it was made. I've made a lot of antibodies. Trust me on this. When you make an antibody, you get for what you're screening. Your screening is what dictates what you find. Most people who are looking at anti-CD73 antibodies more recently were looking for blocking conversion of AMP to adenosine. That was their assay. They come up with antibodies that do that. This antibody was developed initially by Linda Thompson at Scripps. She discovered CD73, and she was looking at functional assays, lymphocyte adhesion, and properties like that. She was looking for something different. Why does it do both these things? First of all, we know for sure our antibody binds a different epitope. It doesn't cross-block with the other antibodies. It doesn't internalize. The other one's internalized.
It doesn't cross-react with mouse. The other antibodies, most of them cross-react with mouse. We know it's a different epitope, and it has this other property, and I think Jason showed the data. None of those other antibodies did it. In fact, there was a poster today. What's the name of that company, French company? Innate Pharma. Innate Pharma. Yeah. God only knows how they come up with this, they had 006, our antibody. I don't know where they got it, they had it on their chart, and they had their antibody, and they were showing that, "Gee, ours causes CD69 and theirs doesn't." They're making a good point out of that. I don't know what the point was actually, they confirmed what we did. What I really want to know is where they got our antibody. They could make it from our patent.
We did the same thing. We've made antibodies from our competitors, but I don't usually put them on charts and call them the same thing. That's a really bad thing to do. We did the same thing. We don't tell you what it is.
When we see data from other CD73 antibodies, the read-through is going to be harder because yours has this additional activity, correct?
Well, you mean other CD7.
Other CD73.
Well, we're predicting that anti-CD73 alone, if you're just blocking conversion of AMP to adenosine, without blocking the receptor-
Receptor
it's not going to be effective.
Okay.
That's what we predict. It'd be less effective.
Thanks.
That would be the prediction. Yeah. By the way, I would also predict that anybody making it on a small molecule is dead.
Sorry.
Yeah, a small molecule is never going to cause this immunomodulatory effect. There's just no way because a small molecule inhibitor of CD73 is trivial to make. You take AMP, and you modify it a little bit, and you've got an inhibitor. My chemistry group would do that by tomorrow morning if you wanted them to.
Yeah. Great work. I actually was just very curious on these. To your point where you said the CD73 just monotherapy inhibition, you predict that it's probably not going to be very effective by itself.
If it's just blocking adenosine.
Right.
Now, the immunomodulatory properties, that's adenosine independent. Maybe I forgot to mention that. All those properties are adenosine independent.
Yeah.
We're sure of that.
Yeah. There's a lot of other pathways where these AMP can build up.
Yeah.
Would it be more important to you blocking upstream or further downstream? For example, in this case, the A2A.
Yeah, that's a good question. Everybody asks that question. You're right. Today was a paper on CD38.
Right
can make adenosine directly. We've always thought that the best way to block adenosine is to block the receptor because I don't care where it comes from, I got it blocked, and that's still our working hypothesis. Remember, this functionality of CD73 was not new to us. We knew about that. We actually read the literature, and this is a paper in 1997 and so we knew that. We never were buying the CD73 story alone, which is why we never went after CD39.
Because I don't know how many pathways there are to adenosine. You can't block them all.
Yeah, no, you don't.
Having said that, we have a trial that's going to start shortly at Johns Hopkins in multiple myeloma, where we're giving our A2A blocker together with, what's it called? daratumumab, the myeloma drug. The anti-CD38 for myeloma, which I think is going to be a really interesting study.
Yeah.
Yeah. I would predict you have to give both together. Again, the immunomodulatory effects are adenosine independent. Frankly, I find that to be the most exciting thing about it and what it's doing to the immune system, B cells and T cells, and how we leverage that. Frankly, I think our anti-CD73, I'll go way out on a limb here now. It's not in our risk factors. It's a forward-looking statement for sure. I think our 006 is going to be really useful in infectious diseases. I think it's going to be great for viral infections where you don't have a good antibody response. There's a lot of viral infections like that, HIV being one, dengue virus being another. There are many like that, and maybe other infections as well.
I think this could be a great antibody to give as an adjuvant to enhance any immune response to any antigen. Don't forget, when you get a vaccination, it's not just the antigen. It's mixed together with an adjuvant. That's why it hurts when you get your flu shot or whatever. For flu, alum, which is the usual antigen. It's just a precipitate, an aluminum hydroxide. At least it used to be that. Maybe they've changed it. That's pretty simple. Again, for other more serious or difficult things, maybe this antibody could be a great adjuvant. We're talking about tumor therapy, tumor vaccination, infectious disease. If you want me to go further out on a limb, I'll draw you a connection to autoimmune disease, but maybe it's getting late. I've had too much to drink. No. Any other questions? Well, first of all, thanks, everyone, for coming.
A lot of fun talking with you tonight, and hope you enjoy the rest of your meeting. Thanks.