Immunotherapy for Prostate Cancer
SESSION 2: Immunotherapy for Prostate Cancer
Moderator: Nour Shobaki (University of Pennsylvania)
Pasritamig, a Differentiated, KLK2-Targeted T Cell Engager for Prostate Cancer
Charles Drake (Johnson & Johnson Innovative Medicine)
View the Transcript Below:
Immunotherapy for Prostate Cancer
Nour Shobaki, PharmD, PhD [00:00:12] Hello, everyone. I am Nour Shobaki from the University of Pennsylvania, and it’s my pleasure to welcome you here for this session on immune therapy and T-cell-based innovations for prostate cancer. All right, our next speaker is Dr. Charles Drake, Vice President of Immuno-Oncology at J&J Innovative Medicine. He’s presenting today regarding KLK2 T-Cell Engager for Prostate Cancer. Welcome.
Charles Drake, MD, PhD [00:00:45] You know, so it’s good to be back. I have to say, like, I haven’t gotten to speak since I moved to a company, and it’s finally a nice chance to share some of the work that’s been done at the company, and also with some of our academic collaborators. These are my disclosures. The company actually does pay me, surprisingly. I’m gonna discuss mostly this drug that’s in phase one and now actually in phase three development. The other disclosure I have to say is I’ve been working on immunotherapy for a long time. As I prepared for this particular talk, I went back to my slides from PCF the first time I got to speak. Orally it was 2006, and I talked about vaccines, and then I talked about vaccines and I talk about checkpoints, then I talked about check points and vaccines, and then I talked about other combinations with radiotherapy. But I think this is different. I think finally we have therapies that can benefit the majority of prostate cancer patients. And it’s not just from J&J, it’s from other companies as well, which I’ll try to share. So, what are these drugs? These drugs are called T cell engagers. They’re not BiTEs, okay. You shouldn’t say BiTEs. BiTEs are little, tiny things like Blinatumomab that are just small pieces of antibodies glued together with amino acids, five amino acid linkers. These are T cell engagers, and what they are is they’re bi-specific antibodies in which one antibody grabs a T cell through its CD3 receptor, and the other grabs a tumor. And I hear this all the time wrong, so now you’ll know not to say it wrong. What people sometimes say in the clinic is, oh, these drugs drag T cells to the tumor, excellent. That’s actually not true. The principle of action or the mechanism of action of these drugs is that the CD3 binder twists or actually induces an activation change in the T cell engager that leads to activation. That’s how they work. And that’s also why you get side effects of cytokine release syndrome because you’re activating a significant fraction of T cells in a treated patient. So, when I was at the – I moved from Hopkins, right, to Columbia and I really believe in these drugs and so, we opened three clinical trials. We opened a Harpoon trial. We opened the J&J PSMA trial and one other one and they all failed. So, looking a little bit grim until this happened. This was presented by my friend Kevin Kelly at ESMO in 2023, now almost two years ago. And honestly, this was a game changer for the field of immunotherapy and prostate cancer. What they showed is an Amgen drug. This is not our drug. I’m fair and balanced, right? They showed actually really nice activity. This is a beautiful waterfall for a STEAP1 targeted CD3 T cell engager. We were all there actually. All my buddies from J&J we were sitting in the front row. We were just like, wow, this is amazing actually. The problem with T cell engagers actually, and I’ll show you how we address this, is cytokine release syndrome. That is when you activate a lot of T cells, you release cytokines and you get symptoms of hypotension, tachycardia, hypoxia. It’s bad actually. And it makes these drugs not be outpatient drugs. And this is not picking on AMG 509 here Xaluritamig, J&J was the first company to commercialize a T cell engager. It was in multiple myeloma. It’s against BCMA. And we have the same exact thing. We have two drugs in multiple myeloma, one against BCMA, one against GPRC5D. This is a common feature of these drugs. And it’s a challenge that we’ve largely, to be honest with you, overcome with the Pasritamig drug. So, what is Pasritamig? So, it has a different antigen than PSMA. PSMA is the most commonly targeted. You’ve heard it over and over. My friend Saul talked about PSCA, a little bit of variation there. But the KLK2 antigen is maybe the most specific solid tumor antigen that we’ve seen. It’s really only expressed in the prostate and in prostate cancer. It’s fascinating. It’s actually a secreted protein, and the original binders were created by a small company called Diaprost and licensed to Johnson& Johnson to start the development program. Diaprost included PCF-funded investigators, of course. Other antigens like PSMA are expressed in normal tissues. PSMA, as you’re well aware, is expressed in salivary glands, in the kidney, in nerve sheaths. And STEAP1 also has a little bit broader expression, including some muscular tissues, which probably explains some of its side effects. But KLK2 is very, very, very specific. KLK2 also has a really nice activation profile. And I think that you should take this as evidence that immunotherapy can work in prostate cancer, actually can work often, frequently and often. And the thing that’s interesting is, and I think Saul showed the best pictures, in prostate tumors there’s not a lot of T cells, right? So how do you think this is working? We don’t know for sure, but since we see responses in multiple tumors, what it must be doing is bringing T cells into cold tumors, right, because if it only worked in tumors that already had T cells it would not be able to have this kind of activity, actually. We see activity even in liver mets, which is surprising. We see about a 40% PSA-50 response. These drugs, and I’ll talk a little bit more about this at the end. Developing these drugs is the Zen master of early development programs, right? So, it’s really hard. So first of all, all these drugs are given with step ups. So small dose, maybe one step up, maybe two step ups, and I’ll show you in a second that the route of administration really matters. And so, this is, honestly, this is slightly empirical and it requires persistence from an early development group. So, these data were presented at ASCO by Capucine Baldini and published in JCO. Mark Stein, my friend, was the first author. This shows some of the many, many cohorts. So how did we address the cytokine release syndrome, actually? It turned out, to be honest, and again, I’m just being completely open, slightly by accident. So, the dose was being escalated, escalated. Most drugs that are being developed now, folks are trying to make them sub-q for more facile administration. And you know that Pembro and Nevo are both moving to sub-q. So, this started out as sub- q. Patient friendly, outpatient, we’ll all be happy. But what happened is we couldn’t get to the doses that we thought we need sub- Q. And dosing was switched from sub- q to IV to achieve a higher dose level. And when that was done, actually, the cytokine release syndrome, as you can see by the obvious arrow, dropped dramatically. The drug was already pretty good, right? So, there was already not that much cytokine release syndrome. Only a touch of grade two. I can’t really read exactly on the bottom. It was like 20 or 30%. But when we switched the drug from sub-q to IV, cytokine release syndrome went to no grade two, zero, and maybe like 8% grade one. On the top right, something you’ll never see in a phase one trial on this phase one trial at the recommended phase two dose and schedule 40% of patients had no A’s like zero. Right. And like that’s not what you see in phase one trials ever actually it’s remarkably well tolerated and a little bit more information on what happens. There’s still some injection related reactions. Fever treated with Tylenol but it’s remarkably well-tolerated. So that’s the route of administration right. What about this schedule. So, when you’re in a drug company, right, what is the idea? The idea is exposure, right? So, we have to have more exposure, so more area under the curve, giving the drug more frequently, and of course, everything starts out like, oh, by the way, this is, sorry, I forgot the slide. So just to show that the low cytokine release rate in this trial is a bit of a unicorn. The drug on the far left is ours. This is a PSMA, this another PSMA. This is the Harpoon drug. This is a PSMA from Amgen, the earlier one. This is JANX drug, 007. So, most drugs, and again, this is common across T-cell engagers in multiple cancers, as you see, cytokine release, syndrome, Pasritamig. Ather recommended, phase two dose in schedule is a bit of an outlier. So, what about frequency? So again, I mentioned that when you’re a drug company, you’re trying to increase area under the curve, exposure, these things like that. Victor Villalobos was the compound development team leader. We have a lot of acronyms. So, he was a CDTL and he and the team made a very astute observation, somewhat serendipitously, but it was astute. So basically, we’re treating patients every week sub-Q and there were patients who had to hold the dose or something happening logistically and what Victor notices in at least a couple of instances is patient who moved from getting the drug every week to every three weeks, they had dramatic responses. So, the question was, what happens if we give it less frequently? And again, the phase one team really did a nice job of saying, let’s do the experiment. Let’s try Q3 and Q6 week, actually. And it worked, actually, so these are just comparing IV Q3 versus Q6. You double the PSA response, on the bottom line of the top table, you can see that you double the PSA-response rate at 12 weeks. So why is this? This is the immunology part. So, I’m the immunology guy, right? We have a translational research group. And I said to them, I said, you know, we have to try to figure out why Q3 week looks clinically to be superior to every three weeks or every week. And so, what I told them actually was that when you activate T cells too frequently, what you do is you induce activation induced cell death. You can measure this with caspase and other markers. And they said, You’re wrong. What’s happening when you’re giving it too frequently is you’re getting T cell exhaustion. So, we decided to test that they generated a 50-color flow panel focusing on T cells and performed the following experiments which I’ll sort of summarize here. So, on the top you can see what’s supposed to happen when you activate a T cell. It’s supposed go into activated, brief activation, then memory. But when you have chronic antigen activation in either infectious disease or in cancer, you become exhausted. But exhausting doesn’t happen all at once, okay? Exhaustion happens in stages. So, there’s a set of cells that are called progenitor exhausted. They express TCF-1, and these are the cells that we think we save with anti-PD-1 in kidney and bladder cancer. On the bottom, you can see terminally exhausted cells. They lose TCF-1. They express activation induced cell death markers like caspase and G2HAX, and then they die, actually. So, the idea was maybe we have more of these progenitor exhausted that can be saved with the CD3 redirector in the less frequent scheduling. These are the samples that we took. So, this is like a lot of work. So, we got samples from Q1 week, Q3 week, Q6 weeks, these are peripheral blood samples by the way. We don’t have enough tissue samples to do the same experiment. I know somebody’s gonna ask that later. But the translational group did a lot of work on analyzing these cells. I’m only gonna show you like a handful of markers because it’s like too many data. But these are the data and it was actually crystal clear. So, the bottom line is when you increase the dosing, particularly from Q3 to Q week, you have more of these rescuable progenitor cells. It’s like the p-value is ludicrous, right? It’s five e to the negative six. This doesn’t usually happen in immunology, by the way, these kinds of p-values, especially from patients, heterogeneous patients in a clinical trial. You also see more activation as expressed by CD69. They humored me. There was a little bit more AICD in the Q3 and Q1, but it’s not as significant as this. I’m not even gonna show it, actually. But it’s really true. And the bottom line is when we give these kinds of drugs, immune drugs, we should think a little bit, not about coverage and AUC, but maybe a little more about what makes sense immunologically. So where are we going with this next? Actually, I’m gonna show you two quick slides. So, for full T cell activation, you need signal one and signal two. Signal two is provided by CD28 or 4-1BB or maybe CD2. So, we also generated a PSMA by CD28 molecule. Bilal is going to talk a lot more about this, but the idea is if you get full T cell activation, you can maybe get memory. We hope to get deeper responses and more persistent responses. This is an ongoing trial. Everybody mentioned, and I think my friend, Sumit asked about the tumor microenvironment. Yes, we understand. The toxicity profile of Pasritamig allows us to do combinations, and so we’re doing some of these now. So, a simple one is to combine with docetaxel. And do some death by different mechanisms. This is ongoing. We like to combine it, guess what, with a radio ligand. We particularly would like to combine with an alpha emitter because we think alpha emitters might affect the TME differentially. Heterogeneity of the tumor. We have antibody-drug conjugates. Of course, we’ll combine with an antiandrogen and ARPI. More exciting, this is my favorite one, actually. So, we’re gonna publish this someday, but we have these, Larry Fong just published a beautiful paper in Nature, if any of you have seen it, about prostate cancer showing that. The TME is altered by adenosine and M2 macrophages, and adenosine plays a major role. We happen to have, for other purposes, a really potent A2A inhibitor. So, we’re hoping to combine Pasritamig with this A2A inhibitor, we have lots of pre-clinical data. It’s actually really cool. So, you do these experiments, right? So, you put tumor cells in a dish, you put T cells in the dish, you put in Pasritamig or another T, and then they kill, right. You put in adenosine, and there’s adenosine analog, it’s called NECA, and they don’t kill at all. From any donor, from any T cell engager. You block it with this A2A inhibitor or other A2A inhibitors, and the T cells can kill again. So, the preclinical data for this combination are particularly strong, and this is a particularly exciting combination. We also believe in regulatory T cells. So, I’ll once again mention that if you take these out of your for ease of this product, you get better CAR-Ts in patients. We would like to deplete them, but really in the clinic, there’s not a good T-reg depletor that we’ve come across yet. So, take home message. T cell engagers work. I think immunotherapy for solid tumors is being changed. Tarlatamab is FDA approved. These are four of the ones that are in development. Prostate cancer, I especially like the name 007 for the JANX drug. Our data show, and I think others will find the same thing, that the frequency matters. The route of administration matters, but it’s empiric. And I just put this note. So, the company Johnson & Johnson Innovative Medicine developed a drug for EGFR mutated lung cancer called Amivantamab. It was developed IV. And it works, it’s a good drug. It’s FDA approved as a survival advantage. But late in the development pathway, it was tested sub-Q, and that one was better sub-Q. It’s less toxic and slightly more active. So, I don’t think you can know empirically about that. Pembrolizumab is the canonical immunotherapy nivolumab as well. And the reason is not because you give it alone, right? I haven’t treated a patient with Pembro mono in years. It’s given with other things. And so, the, I think the feature of Pasritamig we hope to exploit is that can be combined. And then finally, as you can all imagine, we wanna move these agents early in our disease phenotype. We know that biochemical relapse is a second chance for cure. These are the acknowledgements, and there’s a lot of them. I would just like to go through a few. The discovery group is led by my friend Brent Rupnow. He’s actually here, Theresa who, stars are people who left, right? So, Theresa helped develop. Fei Shen, who’s also starred, was helping develop, and now he’s more working in early development on the CD8 program. And also, he gave me some very pretty slides that I included. Translational research was done by Kristin and Ruchi, a lot of work, and also my flow course, Sarah and Michael. Early development could not possibly be more honored here. They worked for years on this, and they actually were willing to do the hard work and do many, many cohorts and test many patients. Site PIs, Karen Autio, my friend, is the one who presented this at ESMO recently. Diaprost, and of course the PCF, and then patients and their families. So, I’m 43 seconds over, which is not too bad, and thank you for your time.
Nour Shobaki, PharmD, PhD [00:16:38] Can I actually start with the first question?
Charles Drake, MD, PhD [00:16:40] Of course.
Nour Shobaki, PharmD, PhD [00:16:41] So you mentioned at the end a combination between KLK2, CD3, and PSMA CD28. How is the safety profile? I’m asking because as far as I know, a general clinical trial with CD28, PSMA resulted in systemic toxicity.
Charles Drake, MD, PhD [00:16:55] Yeah, so combining CD28, so CD28s are scary, right? The super agonist from T Gen scared us all. We’re currently in dose escalation, and we’re really working that out. So, we really haven’t revealed anything on that yet. But I can say that we hope to move that combination forward. How’s that? Sorry.
Unknown [00:17:20] I mean, fantastic talk. And I mean interesting you kind of partly confirmed my cynical impression, which is that drug schedules are driven by companies’ marketing departments, not by their scientific department. And what was so acute about this was that actually a de-escalated schedule was better than escalated schedule. And I’d be interested in your opinion as to how general this phenomenon is. Because if you look up, say, for example, what the half-life of Durvalumab is, it’s pretty much the same as the interval between doses. So, you’re actually progressively accumulating it. And from what you’ve just said, that might be precisely the wrong thing to do with a lot of these drugs. And it’s quite striking that in the NIAGARA trial, for example, I know that’s bladder cancer, only half the patients got to the end of the schedule, but there was a survival advantage. And so, the suspicion is that it’s whatever you did up front that matters. So how much are companies like yours and how much should we as academics be fiddling around with these schedules rather than going with kind of a chemo paradigm with immunotherapies?
Charles Drake, MD, PhD [00:18:16] So here’s what I’ll tell you, actually. First of all, it’s not so much the marketing as the early development group has more use to small molecules in trying to achieve coverage, right? So, I think the marketing wasn’t so. But the data that we saw actually at ESMO from the group from Australia we just heard of really showed a really beautiful thing with RLT administration. And it was a little bit of a similar story, right. So, it was with advanced cell with the alpha emitter. And the idea was very similar to what we see. If you spread it out a little bit more and actually, you’re a little more adaptive, you got more activity. I think the thing to remember is that once you go to phase three, you don’t get a chance to go back, right? And so, I think that taking the time in that phase one space to explore alternative schedules really makes sense, particularly for immunotherapy, right. Your immune system wasn’t designed to be exposed to stuff forever, right, it was designed to see an infection, work really hard, clear the infection, turn into memory and then be happy, right And so giving immunotherapy more frequently might get more acute responses, but in the end, it might wind up hurting you through exhaustion. Great question, thanks.
Unknown [00:19:26] Thanks for the fascinating talk, you know, the data just looks very amazing. My question is on the nature of the secretion of KLK2, which the expression level is off the roof, the TPF is like 2,000, 3,000. Can that explain partially the limited toxicity when injecting IV or secreted KLK2 blocking the fact.
Charles Drake, MD, PhD [00:19:47] So it’s interesting, so Faye and the group actually published a nice paper in clinical cancer research that details a little bit, but the antibody binds to an epitope of KLK2 that’s in the process of being secreted, and it’s actually not synced, there’s no synch effect from soluble KLK2. So, the other thing about KLK2 that’s interesting is the number of receptors per tumor cell is quite a bit lower than PSMA, so it’s a lower expression and the antibody has a very limited epitope to which it can react. That might explain the low toxicity, but to be honest with you, you don’t have any other things to compare it to.
Unknown [00:20:27] Hey, Chuck. Super exciting. One question is going back to the memory point. Did you guys see evidence of induction of memory T-cell subsets with Pasritamig.
Charles Drake, MD, PhD [00:20:37] Yeah, so it’s a good question. So, there’s a ton of data from that flow, 50-color flow panel that we obviously couldn’t show here. And it does look like there’s induction of memory more with the extended dosing than with the compact dosing. We’re really working on getting all those data together in a more digestible form for a future publication. But yes, we did see more memory with the more extended dosage.
[00:21:00] Awesome. Thank you.
Charles Drake, MD, PhD [00:21:03] Thanks.
[00:21:04] Sorry, just a quick mechanistic question. Oh, sorry.
Unknown [00:21:10] Thank you, so it’s a question around, you know, this observation with immune checkpoint inhibitors that actually immune checkpoint inhibitors delivered in the morning results in better outcomes than if you have immune checkpoint inhibitors in the afternoon. And of course, our understanding of what drives that is unknown, maybe it’s tied into high cortisol levels in the morning as opposed to other times in the day. I suppose I’m just wondering if this paradigm is gonna be relevant for these sorts of therapies as they start to come in the clinic and maybe we need to be thinking about collecting this sort of information as well.
Charles Drake, MD, PhD [00:21:49] That’s a great point. And certainly, with our case report forms, we should be collecting like what time you got the drug, and we could start to correlate those data. This drug is in phase three, which is many hundreds of patients. And so those are data we can collect. The one thing that I can tell you, though, I didn’t really emphasize it, is this drug, it’s actually interesting. So, during the step-up doses, it’s given with steroids, right? And the steroids don’t really impact that. And then once it’s fully stepped up, then you don’t give any more steroids. I’ve been surprised actually, and we’ve gotten this question before, like, why don’t you give those steroids? I’m like, well, because we couldn’t give the drug, right? But on the other hand, I wonder if some of the diurnal effects aren’t due to the levels of corticosteroids in the body. Really, we’ve seen across other programs, too, that the corticosteroids don’t really seem to impact the effect of the CD3 effectors, and there’s going to be a talk from Amgen and my friend Judd, I think, tomorrow or the next day, and will see what he says about that steroid effect as well. Thank you very much.