All right. All right. Thank you so much, everybody, for tuning in. We've got Editas, Gilmore, and Amy are joining me this morning. Thanks so much for being with us.
Thank you.
Let's get started with a brief overview. Obviously, there's been a recent pivot in the programs that you're addressing. We're going to spend all today talking about LDLR and what your plans are there. Can we start with maybe a high-level overview of the target and how you're planning to address it?
Absolutely. LDLR and our targeting strategy represent the overall strategy for Editas, which we articulated for about the last three years. We really sort of culminated that sort of pivot with its announcement. Essentially, we are a CRISPR editing company focused on delivering high-potency, low-cost-of-goods therapies that will actually change the standard of care. LDLR is a wonderful representation of one of the core CRISPR strategies, which is we use CRISPR to do things that other modalities cannot do. In other words, we are functionally upregulating or increasing the levels of a protein such as LDLR that actually have a significant positive impact on health and disease. LDLR is the LDL or low-density lipoprotein receptor and essentially takes up low-density lipoprotein cholesterol or LDL from the blood and essentially clears it from the blood. LDL cholesterol is so-called bad cholesterol.
Sort of the general increasing view is that there is no level of LDL cholesterol that is good for you. You can actually get cholesterol that you need, obviously, for other vital living processes, physiologic processes from other particles other than LDL. With our program, we have demonstrated to date that across a range of doses in non-human primates, we are getting an at least 6x increase in LDL receptor, which is driving a 90% reduction in cholesterol, which is an unprecedented level of reduction. It is important to emphasize that non-human primates have historically predicted both the effect sizes of CRISPR, in vivo CRISPR editing of the liver, as well as predicted very much the effect sizes of cholesterol lowering. We feel very good about this because this 90% reduction in LDL cholesterol really has the potential to transform the space.
That's not just our view, but the view of experts who look at it. More importantly, if you look at non-human primates and human experiments with PCSK9, those reduce LDL cholesterol by about a mean of 60%. That 90% is a meaningful change.
Absolutely. Now, you've mentioned that you targeted LDLR and this upregulation because of natural human variants, existing genetic studies that have paved the way here. The edits that you're making are a little bit different than the naturally occurring variants, and they have different effects. Can you talk about the potential implications of those differences?
Yeah. We actually think that from a the implications are actually marginal. What we have actually done in making those that slight change from the variant and by the way, let me pause for a second and say that we use the natural occurring human variants to identify disease-modifying or disease-mitigating proteins that could be upregulated to de-risk them from a safety and efficacy point of view. In the case of LDLR, an Icelandic kindred has a number of members who have a 2.5 kb deletion of a regulatory element, 3' untranslated region of the gene. We have made some slight changes only because we walked across that area just to maximize the potency. It does not actually have any change or does not change the LDL. The LDL receptor that is expressed is the same as the LDL receptor that exists. It is a normal.
It's just an mRNA lifetime.
Essentially, we have just stabilized the messenger RNA. The LDL receptor itself is the same, is unchanged in the person who has edited. It's just increased in its levels.
OK. How does this approach compare to other MOAs in the CV space? Obviously, this is an area that people have been predicting is going to get more fragmented as different mechanisms of action more specific to particular subpopulations start to get advanced.
Yeah. I think the way to look at it, first of all, is our technology and our approach can actually be applied across a broad swathe of people with hypercholesterolemia. That is the first thing. I think from a mechanistic point of view, the LDL receptor has been an area of interest for over 30 years. It has indirectly increased in levels by HMG-CoA reductase inhibitors, which are the statins, which were the first medicines that really potently could change cholesterol. They have a mean reduction of about 40%. PCSK9s came along, and they also indirectly increase LDL receptor by essentially blocking the degradation of the LDL receptor. They all hit a ceiling. I think that ceiling is because they are reducing degradation.
In the end, the ceiling is going to be set by the amount of LDL receptor that's being manufactured by a given cell. We break through the ceiling by essentially increasing directly the level of LDL receptor. In many ways, there is a common theme. I think the challenge has been that about 75% of patients who are treated for hypercholesterolemia do not achieve their targets. Those targets are necessarily ambitious because if you want to actually stabilize plaque, you need to get your cholesterol to about 70 mg/dL or lower. If you want to reduce or actually get the plaque to reverse—the plaque is the narrowing of the blood vessel caused by the accumulation of cholesterol—you need to get levels well below 50 mg/dL .
The nice thing about our approach is that we can do that conceivably with one injection, one infusion, and get a 90% reduction, which really it doesn't matter where your starting level is. You're going to be in that level, or you're going to get to an achieved cholesterol level of less than 50 mg/dL .
Great. Let's talk then about some of those patient subpopulations. You've mentioned the heterozygous familial groups. Obviously, that's a little bit different, and the bar is a little bit different compared to the homozygous, compared to hypercholesterolemia patients. How do you, even though all of these folks have elevated LDL and LDL needs, so what populations are the low-hanging fruit here, and what should we be expecting from first-in-human data coming next year?
If you just look at the United States, through the lens of the United States, you end up at 70 million patients or people in the United States have elevated cholesterol that would meet the need for treatment. The HeFH, the heterozygous familial hypercholesterolemia patients who are eminently treatable with an LDLR upregulation strategy such as ours represent about a million patients. Refractory certainly represent more than half of those. They are obviously a very clear niche population to start with, as regulators necessarily will want to understand the benefit-risk. There is a very clear benefit-risk for these patients who really have a very relatively rapidly progressive accumulation of cholesterol and a high risk of developing significant cardiovascular and arteriosclerotic cardiovascular events early in life.
The other populations that we're interested in and that we would expand into would be, for example, patients who have had significant coronary disease already and have a high risk of a further event. These are the next area of expansion. Essentially, what we're going to do is walk along from a risk-benefit point of view as we characterize the safety as regulators and physicians prescribers become more and more comfortable with the mechanism, we can expand. The beauty of this expansion lifecycle, we can do it with one therapeutic strategy, with one medicine, which is our EDIT-401. You touched on homozygous. What I would say is that in the context of homozygous, homozygous really means that they have loss of function or reduction in function mutations in both copies of the LDLR gene.
We would anticipate that it would be more challenging to treat them. Now, if you upregulate some people with a hypofunctional, you could certainly reduce cholesterol. And we've seen that partially validated by the use of PCSK9s, where they have a much lower impact, but some impact on patients with homozygous familial hypercholesterolemia. The heterozygous would be where we start, and then we will actually move, and I think we could move relatively rapidly into other subpopulations such as that secondary high risk and then ultimately primary high risk.
OK. What should we be expecting from data next year? Obviously, it's first-in-human. It's going to be a relatively modestly sized cohort. How do we interpret the results there as we look across different potential patient subpopulations? Should we be expecting to see consistent effects across everybody that ends up enrolled in your trial, or are there particular subpopulations where we'll see different?
I think the way to think about the data is that our primary focus for the human proof of concept, which we plan to have by the end of 2026, so we're tracking for that, will be in heterozygous familial hypercholesterolemia patients. Obviously, we have to finalize the details of a study with regulators. We can actually look into additional populations or subpopulations such as the ones I've described. I would say that we anticipate that the effect size should be the same. The reason I say that is because not only have we seen a 90% reduction in non-human primates, but we saw that 90% reduction in mice with a very high level of LDL cholesterol based on a high-fat diet. Also, very importantly, we saw it in mice who are heterozygous for loss of function of LDL.
We've looked across a sort of a swathe, if you will, of animal models that physiologically represent that swathe of niche or subpopulations of the hypercholesterolemia patient population and saw similar effect sizes.
OK. Let's talk about that preclinical data. Let's talk about the editing efficiency there. The data you showed recently suggests that even at very low doses, you can get excellent reductions in LDL, greater than 90%, as you say, even with very modest allele editing. Your dosing curve that you showed recently was really quite flat in non-human primates. What should we be expecting from low doses in human escalation? How should we be thinking about the efficacy curve across dosing?
I think this is a very important point in that to date, we have sort of seen a relatively flat dose response with the doses we've explored to date. The data we have shared, obviously, limited between 1.5 mg/ kg and 4 mg/ kg in the non-human primates. We are actually continuing a dose ranging finding study, which is going to higher levels just to push the safety because the doses we've tested to date were very well tolerated chemically, physiologically, behaviorally by the animals. We are actually also going lower, looking for a minimally efficacious dose. Obviously, when we have those answers, we'd have a much clearer view of what that dose response could look like in the human clinic.
It's worth saying, though, that because you're upregulating a protein instead of knocking it out, you don't necessarily have to touch every cell in the liver.
That's a very good point. We believe that the potency we're seeing—sorry, let me go back. We actually think we've got quite a potent approach with this augmentation or upregulation strategy because we've noticed that we're getting that 6x upregulation and that significant reduction with only 10%-40% allelic editing.
Which is quite different across the liver.
That's 6x across the liver.
Not just in an edited cell.
Yes. We've looked at the total liver. Obviously, when we're looking at allelic editing, we don't have cell-by-cell upregulation data. Obviously, it's generalized. We have seen it with 10%-40% allelic editing, which represents about 15%-50% of hepatocytes being edited. That suggests there's a very potent effect size driven by our selection of an augmentation strategy, which I think is very valuable and potentially actually means that we may be able to expect or look to a potentially better safety profile.
What about durability? Obviously, we're very early days here. Is there anything we can conclude about durability from NHP data so far or theoretically?
To date, we've only presented one-month data, up to one-month data for non-human primates. We do have three-month data in our murine, and we've actually seen a maintenance of that effect. Our durability data has been very good. We would anticipate a good durability, frankly, up to lifelong durability based on the experience of others. Others who have done in vivo liver editing, granted they're doing knockdown, but in the end, a CRISPR edit is a CRISPR edit. How you use it is really what's differentiating us. We're actually seeing durable effects both in their long-term animal follow-up, up to six months, I think people have presented. Indeed, other companies in the CRISPR space have demonstrated up to a year's durability already in sort of early readouts from their phase I studies.
I think I would agree that the edit, we should expect the edit to be stable. The follow-up question would be, what about the cells? Is there something fundamental about upregulating a protein like this that could reduce the survivability?
That's a very good point. These are things that we actually have thought very carefully about, and we actually feel very good. Obviously, one would be, do you see, is there a potential risk to the cell for upregulating LDLR, which, by the way, is a low-abundance protein? In contrast to things like an alpha-1 antitrypsin or albumin, which are high-abundance proteins, you can, and sorry, alpha-1 is a good example of where a high-abundance protein where there is a misfolding mutant could actually result in liver disease. That is not something we've anticipated here. We've done a lot of work in thinking about that because, again, we are dealing with a low-abundance protein. I think the second issue is, is there a hypothetical risk that a cell that's taking up a lot of cholesterol will be stressed by that?
Indeed, our histopath to date, looking at our animals, have shown no increase in droplets or lipid droplets. Indeed, where steatosis has been an issue, it really is driven by triglycerides rather than cholesterol esters. We have, what do I say, our understanding of steatosis. We have our understanding of our own data. To date, we actually see a low risk. From that point of view, I think it's very good. We don't believe that editing a cell would actually give it a survival disadvantage. Of course, as you know, an edit as a cell divides, the daughter cells will inherit the edit. Both daughter cells will inherit the edit, unlike, for example, an AAV where the episome will only go to one of the two daughter cells. Dilute out your effect.
That does bring us to a safety. Obviously, safety is a very sensitive topic in gene therapy these days. Upregulation is a novel strategy in gene therapy and gene editing. How have your initial interactions with the FDA been or any particular concerns related to this approach?
I think the good thing is we have not sort of disclosed yet any interactions with the FDA around this program. I will say that we have had multiple interactions with the FDA and other agencies around upregulation strategies because, of course, upregulation was the strategy we used for upregulating fetal hemoglobin or gamma globin specifically to treat sickle cell disease. Obviously, that was an ex vivo approach. We have flipped it to an in vivo, but that is still in discovery. Our interactions at that time and our own experience with upregulation actually was very positive. In fact, we believe it validated our strategy of differentiating our company by focusing on upregulation and using naturally occurring human variants to inform the selection of a target.
I say that our upregulation strategy for fetal hemoglobin, others have done it too, was actually driven by an observation that naturally occurring human variants that upregulate fetal hemoglobin would actually be well tolerated and indeed potentially beneficial. Overall, the one nice thing about this upregulation strategy, or one of many things, is that we've actually had that experience both in the clinic. That experience continues. We've had a lot of interactions about that approach with agencies.
OK. You've also said that you can lean on that Icelandic cohort for comfort about safety given that they all have elevated levels of LDLR relative to the general population. That makes sense. You've also said that you are upregulating LDLR substantially more than they do. Does that change the bar in phase I in terms of just number of patients that you might need to feel confident?
No, I don't think it does. Obviously, that will be determined when we have discussions with regulators. Overall, I think the approach in many ways, we're just increasing the levels of a naturally occurring low-abundance protein. The Icelandic cohorts have been up to 2.5x. Indirectly, the suggestion is that the PCSK9 gets to those levels as well. We're going a little bit higher. We believe that with the low-abundance protein, we have a very potent rationale for doing that safely. By the way, it's not just about the regulators. It's about us, what do we think we can do. We feel very comfortable about what we're doing and what the risk, or rather what the risk profile would be to patients.
Fair enough, although I typically think of the regulators being less comfortable than the sponsors.
Actually, we drive a company where we want to be less comfortable than the regulators. That's our philosophy. We're the last line of defense, I say.
Fair enough. One more about safety then before we close out in depth. Your NHP histology data does show possible, although obviously minimal, off-target edits in liver and in ovary. Is there a difference in concern with this approach versus some other gene editing approach from an off-target perspective?
I wouldn't expect so. We wouldn't expect so. Indeed, those are what I've called the first pass. Those, again, were very low edits. We are still obviously determining which specific cell types were actually edited, which is important because obviously, if you see a very low degree of editing or delivery in a tissue, for example, a gonad, you have to make sure really it's the gametes that you actually have to worry about. We anticipate that just like others have seen, when they have actually expanded their data sets and looked at more detail, this will not be a new special concern. I wouldn't expect a special concern. We would not expect a special concern with an upregulation versus a knockdown strategy.
All right, Amy, we might jump. I guess we've got to ask you a question.
Sure.
How about cash runway and bandwidth and what we can get through with these initial human cohorts?
Sure. Gilmore just spoke about our data and how we're progressing. We're seeking to file an IND or CTA in the middle of 2026 next year, and then obviously human PoC at the end of 2026. Our cash runway right now extends to Q3 2027. We have more than enough cushion to get us through those preliminary readouts and continue to progress the program.
Excellent. With that, I think we're out of time. I appreciate all the time.
Thank you.
Thank you for joining us.
Thank you very much.