Good afternoon, and thank you for joining us at Guggenheim's inaugural Healthcare Innovations Conference. I am Debjit, one of the therapeutic analysts here, and joining me today is Samrat Kulkarni, CEO and Chairman of CRISPR Therapeutics. Thank you for your time, Sam.
Thank you for having us.
Maybe let's start with a very quick overview, given how broad the pipeline has become currently.
Yeah, happy to. We're in the 10th year as a company. We were started with the promise of taking this powerful technology called CRISPR and making medicines with it that could be transformational for patients suffering from very serious diseases, and where we stand today, we're proud to have accomplished a very important milestone in the decades of biomedical innovation, which is to get the first approved CRISPR medicine in the world for CASGEVY, which is in sickle cell and thalassemia, and that is being launched right now by our partner, Vertex. We can talk more about that. But what's also equally what I'm equally proud of is that we've created a diverse pipeline, broad and diverse pipeline beyond CASGEVY that spans ex vivo and in vivo.
So on the ex vivo side, we have a best-in-class allogeneic CAR T targeted towards CD19, CTX112 that we're developing in both oncology and autoimmune indications. And on the in vivo side, we have direct injection, IV injection that could lead to liver editing for targets like LP(a) and ANGPTL3. I think LP(a) will be a very important target in the pharmaceutical space in the coming years. And that's just the high-level glimpse of the pipeline. We have a regenerative medicine portfolio targeted towards type 1 diabetes. We have other we're also going after solid tumors with advanced CAR Ts targeted towards CD70 and other targets that we're going after as well with our in vivo pipeline. So all in all, a very broad pipeline with different theses around using CRISPR, but targeted towards creating best-in-class medicines.
So why don't we start on the CASGEVY side? The launch is clearly gathering some traction. How much visibility do you have into 2025, or this is literally handed over to Vertex?
Vertex is leading the global commercialization, and we're proud to have Vertex as a partner, and they're doing an amazing job getting this to patients around the world. You saw that the first patient was treated commercially in the third quarter, and that was in the Middle East. And it just shows the reach and breadth of what we're doing here with the commercial enterprise. We have visibility into the launch. And what's interesting about the launch is there's a time lag, but it's actually one of those that where the growth can be compounded because you have growth because you have more ATCs activated. And we're about halfway into activating all the ATCs we plan to activate, or a little over half the way. And then we're going to have acceleration because each of the ATCs are going to treat more patients.
So I think what we've seen so far is sort of exponential growth in the number of cells, patients who've had cells collected. But as we get more ATCs, you're going to see a greater number of patients with cells collected. And generally, what we saw in the clinical trials was that a very large fraction of the patients, or almost every patient that got their cells collected, went on to get treatment.
Got it. So we talked to Vertex about this a little bit, especially given the recent news coming out of Beam with the unfortunate death of a patient primarily because of busulfan. How are you thinking about non-myeloablative conditioning regimens, and how involved is CRISPR in that process?
Yeah, I think the way we think about the life cycle of CASGEVY and our sickle and thalassemia franchise is that we think that launching CASGEVY in its current form actually addresses the need for a significant number of patients. I think in the U.S. alone, we have 25,000 patients that have very severe disease that need a medicine now. In the Middle East, there's an equal number of patients that have both sickle cell disease and thalassemia. And in Europe, you've seen very large populations of thalassemia. What's been surprising in certain countries like France and the U.K. is the anticipated number of patients. We anticipated fewer patients than there actually may be in terms of sickle cell disease. So there's a lot more sickle cell patients in some of these countries than we originally thought. Now, that's the initial population.
But what we're working on is to say, can we get a targeted or gentler conditioning agent that can expand the market threefold? It's hard for us to comment on someone else's data like Beam, but it's unusual, even with Busulfan conditioning, to have something like that happen. There are databases of 1,900 and 2,900 sickle cell patients who've gone through transplants, and Busulfan has shown to be an effective conditioning agent for transplant and relatively safe. But that said, I think on the targeted conditioning efforts, we're parallel processing CRISPR efforts together with Vertex efforts. And we're going to put them all side by side and say, what's the best agent, and put that into play with CASGEVY. And assuming if the data work out, I think that'll not only expand the market, but it'll be beneficial for both CRISPR, for both Vertex and us.
It accrues equally to both parties if the targeted conditioning works out.
Got it. So even if it's the Vertex program that gets nominated, your royalty structure on CASGEVY still remains the same?
Yeah, it's not a royalty structure. It's actually a profit sharing.
Sorry, your profit sharing agreement still remains the same.
Yeah, it remains the same. And eventually, what I suspect is eight to 10 years from now, we're talking about an in vivo solution that directly edits in the bone marrow for sickle cell and thalassemia. And in that case, actually goes back to a 50/50 structure if that works out.
Okay. So let's switch to CTX112. The two new edits seem to have done something. You clearly saw a different PK profile and maybe even a different PD profile. But the response rates were not that different compared to the CTX110 program. Was that primarily because of a more advanced population here than what we treated before? So any thoughts on what we should be expecting, baseline perspective at ASH?
Yeah, I mean, we have, I think, a best-in-class allogeneic CAR T on our hands here with CTX112. Two things I'll point out when you compare the ORR is the ORR disclosed for CTX110 was at the optimized dose level. And here, we just reported ORR for all the dose levels that we treated, one-three . So the second thing is, with the increased capacity for autologous CAR Ts, you're generally getting sicker patients into trials now with the allogeneic CAR T or any experimental medicine in lymphoma. So if you put those two together, I think what you're seeing is definitely an improved profile. The CR rate also is pretty high. In fact, like for like, maybe better than autologous therapies, CAR T. We'll see how that holds up in terms of durability, et cetera, in the greater end.
But we're very encouraged with the early profile that we're seeing with CTX112.
So once you get your dose finalized, do you think you can circumvent the patient population in a sense you don't land up getting post-CAR T patients or post-auto CAR T patients?
Yeah, you know a very interesting dynamic, which I think the regulators are quite attuned to, is the fact that only about 25% of the patients that are eligible for auto CAR T are getting it because it just hasn't penetrated the community settings. And you're also seeing that the market's not expanding for auto CAR T. If you look at the latest quarter results and sales for Yescarta and Breyanzi, you're seeing market share dynamics, but you're not seeing an expansion of the market. That's because they've saturated the academic medical centers. So yes, in a way, for competition for enrollment in the academic medical center where the trials are being done, we're getting sicker patients. But the ultimate goal with these allogeneic CAR Ts is to get to community settings.
And the safety profile being better than auto CAR T, I think, will allow us to do that and get to community settings. And so we're talking about greenfield territory for about three quarters of the population that are eligible for these CAR Ts or deserve a CAR T. So I think there's plenty to go here. But we do need to demonstrate first that we're competitive relative to auto CAR T, and then the safety profile supports moving into community settings.
On the safety profile, you had grade 2 CRS events and, I think, grade 1 ICANS. How does that compare with, or what would be the ideal profile then? What's more troubling, ICANS or the grade 2, especially in the community setting?
Yeah, I mean, you want to see some CRS. I think it's generally a hallmark of efficacy, and it's actually encouraging that we're seeing a little more CRS than our CTX110. It just tells you that it's a proxy for expansion and efficacy of these cells. On the ICANS, I think grade 1 ICANS is manageable. Generally, I think we're getting better at managing these side effects of CAR T, and the use of tocilizumab, the regimen is pretty prescribed now or well thought through in terms of how to manage patients. So I think it can get to community settings. It still is lower than auto CAR T. There was this big, large meta study that just came out on all the auto CAR Ts.
And you're looking at almost, it's greater than 10% grade 2 ICANS with auto CAR Ts in the market, and I think 3% or 4% grade 3 ICANS. So I think it still is much safer from that standpoint with allogeneic CAR T. So we may have threaded the needle here in a way that allows us to be in pole position in this marketplace.
Is there a theoretical limit to the number of edits that can go into 112, or you sort of at the level that you want to be before you start to worry about T cell exhaustions and stuff?
No, I think when we came up with the edits to say we're going to do Regnase and TGFBR2, we looked at obvious edits like PD-1, et cetera. It turned out that this pairwise edit was the best for potency. There are other edits we can do in terms of taking the cells to more memory-like state or prevent Class II-based elimination of these CAR Ts, et cetera. Believe it or not, we have to work on these things well ahead of time. We now have the ability to, with a single cut in the CAR Ts, put in eight to nine edits. If you have a single cut, you're not going to have any translocations. It's a single cassette that expresses the CAR. It actually expresses cytokine and has like four or five RNA-based approaches to knock down genes that make them more potent.
So there's no theoretical limit. But I think there may be marginal utility beyond a certain point.
So that becomes a good segue into what would be the ideal product profile when you're thinking of autoimmune diseases. Let's start with SLE then.
Yeah, I think our competitive position relative to autologous therapies and T cell engagers is even stronger in autoimmune disease because we talked about threading the needle in terms of safety and efficacy in the oncology space to get to community settings, but if we have almost autologous-like efficacy or equal efficacy, but we're safer, and we can dose titrate and in the autoimmune setting, I think we have the situation where T cell engagers may not get durable responses because they're not eliminating the B cells or the autoreactive B cells that are deep tissue resident B cells and the autologous CAR Ts are doing that, but it's just too cumbersome to develop in autoimmune disease, as you're seeing with a lot of the companies in terms of opening up more sites or getting more patients in the trial.
We can move very fast, and we're going to be more potent than the T cell engagers. And it's a one-time treatment. So I think that's going to be a huge advantage. The other advantage that people don't appreciate as much, and our Chief Medical Officer, Naimish, has been speaking with a lot of these rheumatologists who are exposed to all these trials, taking patients off of cyclophosphamide or other treatment for about a month before they collect their T cells for autologous therapy is not an easy thing to do for these patients. They're like pretty sick patients. So it's not just the fact that autologous therapy is cumbersome. It's actually, from a course of treatment perspective, it's a little challenging to get some of these patients off treatment. So there's so many advantages to having allogeneic therapies.
Within the allogeneic cell therapy space, I'm convinced that we have the best-in-class construct in medicine out there.
So the durability has come up as an issue with the autologous CAR Ts in autoimmune diseases. Do you think that's going to be a lingering challenge even with the allo stuff or your edits sort of?
I mean, the durability, if you put it in context, is not bad. We had one-off patients that weren't as durable. I think the expectation was 100% durability based on the early Georg data. And then when there's subsequent data where you had some one or two patients come off that weren't durable, everyone said this is the end of it, right? Because the notion is to do autologous therapy and go through the mechanics of collecting cells and expensive therapy, you all need to be almost perfect in terms of the data. So are allogeneic cell therapies going to have 100% durability? I'm not sure. It may not be. You'll have some patients that are not durable, but we'll have pretty good durability is my sense. And then will T cell engagers have 100% durability? Probably not. I mean, in fact, maybe lower than allogeneic cell therapies.
So I think there's going to be some sort of titration there. But I think if you can get 70%, 80% of patients into durable remissions for an extended period of time and the fact that we can, again, redose a year or two later, that's a pretty powerful product profile.
When should we expect the first SLE data?
We haven't committed to that yet. I think we're going to have a number of data readouts coming soon. We did say that we're going to have data. We'll have more data, obviously, at ASH for 112 in oncology. We'll most likely have data for 112 in autoimmune next year. We also have data for in vivo therapies next year, LPA and ANGPTL3. And we'll have data for our solid tumor programs as well next year. So lots coming in the next six to nine to 12 months. But we haven't said exactly when the data events are going to be.
Got it. So let's pivot over to the cardiovascular side of things. There still is a perception that maybe genome editing is not the best thing in cardiovascular medicine. I'm just wondering what your perspectives are because Verve clearly struggled with the initial PCSK9 as a target. Or maybe it's a more target-specific issue.
Yeah, I mean, one thing I'll tell you is, when we speak with all the leading KOLs in the space, you can name, you probably know them all by first-name basis. They are all very excited about gene editing for these cardiovascular diseases because, one, the compliance in these populations is very low. People who should be taking medicine, sometimes it's like as low as 30% compliance. Second, if you look at these siRNA therapies, you're talking about the prospect of taking once every three months or once every six months injection for 40, 50 years of your life if you start in the 40s, if you have high Lp(a), for instance, and you don't know what the safety is going to be, and you get this sawtooth effect as your dose wanes.
So I think if you can do a one-and-done in a safe manner, I think there's going to be tremendous support behind it from all the leading KOLs and people who are doing these clinical trials. I think for the general population, it's like our sickle cell drug. People ask me, well, what patients are more likely to get it first? Or if you look at your commercial sales, what are the patients that are raising their hands? A lot of it is attitudinal. I think if you take a group, if you take a room of 100 patients, I bet there'll be some segmentation of people who will say, you know what, I'll just try the one-and-done and be done with it, and I don't have to worry about it anymore, versus some that'll say, you know what, I'm not so sure. I'll start taking these six-monthly injections.
Or some that'll say, you know what, I don't mind sticking myself with the sub-Q injection every two weeks. So there's going to be some segmentation of all this. But in a market like Lp(a), which you've highlighted, by the way, I think it was one of the best reports on Lp(a) that I've seen and very comprehensive, the opportunity set is so big that even if you get a third of that market, this is a tremendous blockbuster.
So to de-risk it, it's more about safety than efficacy, right? So how much follow-up do you think the community would need before completely saying, hey, this is a safe approach? And that's the sort of pivot transformational movement or moment for genome editing.
Yeah, I mean, if you think about safety, there are two aspects of the safety. One is the acute safety, how much liver enzyme elevation is there, and does it go away? Is it really transient? Does it go away in three days, four days, or whatever? And then have you measured any sort of off-target effects from a genome perspective? I think those two are going to get answered very quickly. Within the next two years, we'll know what the safety package looks like. And in fact, it's going to be the acute liver. The liver tox is going to be safer than siRNA, right, because you only have one dose. I think there's an open question. People will say, OK, what if very low Lp(a) is not good for you in the long run, da, da, da. But that's the same risk as any other medicine you're developing.
Of course, the only advantage is you can take them off that medicine if there is such an issue. But I think with everything we've seen with natural history data and people who have low LPA walking around, they're normal. You don't know what the role of LPA is in early-stage development. But low LPA doesn't hurt later in life. So I don't expect there to be anything that would be a big surprise in terms of long-term effects of low LPA, for instance, only benefits. But it'll be a gradual unfolding. But I think on the acute safety side, I think you'll know pretty quickly.
So on Lp(a), to your point about we don't know how low can you go, right? So as you think through your dose escalation, are you shooting for 80%, 90% of the hepatocytes getting edited, or are you shooting for something lower just to give you that therapeutic window so you don't the aha moment, which would come after HORIZON reads out or after the Amgen program reads out?
Yeah. No, I think at this point, if you look at our dose escalation and NHP studies, we were able to establish that sort of dose curve, right, to say as we go up in dose, we can get to near 100% hepatocyte editing. And so we want to recapitulate the dose curve in humans with our dose escalation. And by which we'll complete our dose escalation, and hopefully, we'll have the readout from Novartis' trial. That'll be a big moment, I think, in biotech when that data reads out. And that'll tell us how much you need to reduce it. There's a school of thought that says if you look at the different quintiles of Lp(a) levels, there's a bit of a threshold effect. If you get below 50 mg/dL, your risk level is very low. It doesn't matter if you're 25 versus 50.
So you may not need 90% lowering of LPA, even for high LPA, even if we're picking populations that are above 125 mg/dL. Even for them, you may not need a 90% reduction. But I think we'll establish the ability to do that with our dose escalation. And then we'll have to make our decisions on what dose we want to pick for pivotal trials.
But we have seen two things that sort of help that cause, right? We have seen Intellia re-dosing patients safely. And it turned out to be just as efficacious. And the variability that you see post-genome editing is very modest compared to any siRNA therapy. So by controlling that variability and having the ability to redose, why even shoot for 80%-90% of the bad? Go low. Make sure it's safe. At least let the first two big phase 3s read out, and then shoot for the moon.
Absolutely. I mean, I think that's exactly what we're saying to people who ask the question as well. Not everybody appreciates the ability to redose. And you do see it's not like if you have 50% editing with the dose, you dose twice and you get to 100%, right? You tend to get to like 60%, 70% with the second dose. And we're establishing that ability to redose in NHPs. And we'll eventually establish that in humans as well. But I think that gives us a lot of flexibility in ultimately picking the right dose.
Are you still actively pursuing the AGT program in hypertension?
Yes.
How are you thinking or de-risking any hypertensive events?
Interestingly, if you look at some of the data from siRNA, et cetera, angiotensinogen is produced not just in the liver. It's also produced in kidneys and other parts of the body. The reason you don't see as many hypotensive events, even if you have 100% knockout in liver, is because it's being produced somewhere else. The other part of this I'll say is the target population for us is refractory hypertension, patients who are on multiple agents, multiple medications to control their hypertension, right? Some of these patients are taking four to five medications a day to control it. Again, to your earlier questions around Lp(a) and how low do you want to go, one approach here is we could start titrating angiotensinogen knockdown where we get these patients from four or five medications a day to one.
You're reducing the baseline to an extent that makes much more controllable. That patient can then be titrated based on that one medicine that remains, whether it's an ACE or an ARB. Again, there's a lot to learn here. What we're seeing is, again, tremendous enthusiasm from leading KOLs who are saying they want to get their hands on this as soon as possible in the clinic. We're making every effort that we can to get it to the clinic as soon as we can.
Fortunately, the clock's run out. I wish I had more time to talk about type 1 diabetes, but thank you so much, Sam.
Yeah, thank you for having us.
Appreciate the time.