Precision BioSciences, Inc. (DTIL) Earnings Call Transcript & Summary

Hi, everyone. My name is Maury Raycroft, one of the biotech analysts at Jefferies. It's with great pleasure that I'd like to welcome Derek Jantz, the CSO of Precision BioSciences. Thanks for joining us today, Derek.

Hi, Maury. How are you?

Good. I'm good. So we're going to do a fireside chat. And maybe to start off, if you want to give a 1-minute intro to the company.

Sure. So Precision BioSciences is a gene editing company. We're based on proprietary gene-editing technology called ARCUS, which is not CRISPR-Cas. The technical cofounders of the company, myself and Jeff Smith, actually developed ARCUS with an eye toward clinical translation. So we wanted to develop a gene-editing technology that had the properties that you would want if you were going to use this to treat patients. That's what Precision BioSciences is doing today. Our R&D efforts are approximately equally split between allogeneic CAR T, and we currently have 4 CAR T programs in the clinic. And then we also -- the other half of what we're doing is in vivo gene editing, so actually with the goal of using ARCUS to correct mutations associated with rare genetic disease. We have a pretty deep pipeline of in vivo editing projects that are also marching toward the clinic.

Got it, very good. And maybe for the ARCUS technology, if you can talk a little bit more of the history there and also highlight some of the key advantages over other gene-editing technologies.

Sure. So the ARCUS platform is based on a naturally occurring gene-editing enzyme called I-CreI that comes from algae. It's a very rare example of an enzyme that actually evolved in nature for the purpose of gene editing in a eukaryotic cell with a large genome. So the CRISPR, TALEN, zinc finger, insofar as the other editing technologies exist in nature, they're all bacterial restriction enzymes. So they're enzymes that bacteria use to defend themselves against viruses. And as a consequence of that, they have a really fundamentally different enzymology from the I-CreI enzyme that we used for that actually evolved to edit a single site in a large eukaryotic genome. First and foremost, really the property that I-CreI has that attracted us to it in the first place is the enzyme has the ability to turn itself off after it has made its gene edit. So it finds its target site in the genome. It edits its site in the genome, and then it actually changes confirmation and inactivates itself. And by doing so, it prevents the enzyme from sort of wandering off and finding unintended sites in the genome to edit. And that's really what we're looking for when we think about therapeutic gene editing. So probably first and foremost, the biggest advantage is really exquisite levels of specificity due to its ability to self-inactivate. A second advantage that the technology has, it's a very practical one, is it's very, very small, so it fits very comfortably into AAV vectors. It's actually about 1/5 the size of pyogenes Cas9. So it just gives us a much greater array of delivery technologies that we can think about for in vivo editing and actually delivering ARCUS to patients. Third advantage that the technology has is, because of the particulars of how it edits and the particular DNA repair machinery that it recruits to edit the genome, it's very good at inserting DNA into the genome. So ARCUS is not just a tool for knocking genes out. We can also use it to insert DNA and knock genes in very efficiently. And again as we think about rare genetic disease, most of the highest unmet need applications actually would involve inserting a gene that is defective into the genome. And then the fourth advantage the technology has is it is proprietary to Precision. We invented it. We own and control all of the IP. And as these gene-editing therapies do get closer and closer to market, IP is going to become increasingly important.

Got it. That's a really helpful summary of the technology. And you guys have published a decent amount of work on some of the in vivo programs. Maybe if you want to talk about how some of these technology advantages have been proven out in some of the preclinical data. I think the PCSK9 example is an interesting one. I don't know if you want to elaborate on what's been shown there.

Yes, yes. That's a great example to start with. This is actually a project that we started with Jim Wilson's group at Penn in the beginning of 2017, so a long time ago. And the objective back then was really to be the first to demonstrate gene editing in a large animal. No one had actually done it at that point. There were several groups that had edited primate embryos and then implanted those embryos and made an edited monkey as a consequence of that, but no one had actually delivered a gene-editing technology to an animal and edited that animal. So we're really the first to demonstrate that working with Jim's group, and we published that work in 2018. The objective there was to knock out a gene called PCSK9 that is involved in lipid trafficking. It's a good target for cholesterol control because, if you can suppress PCSK9 in the liver, or knock it out in our case, you significantly reduce serum cholesterol levels. So we basically showed back in 2018 that we could put an ARCUS enzyme that recognizes the PCSK9 gene into an AAV and deliver that to a group of nonhuman primates. And pretty much immediately following a single administration of the vector that took about 10 minutes, we saw PCSK9 levels start to drop. And serum LDL cholesterol levels dropped correspondingly. So about a 50% reduction in LDL cholesterol in the animals at the time that we published that first paper. Then more recently, a month or 2 ago, we published a long-term follow-up study from that same group of monkeys that were treated in 2017. We now published 3 years worth of follow-up data, and really what we were looking for is was the gene editing durable. The whole promise of gene editing is that, if you can go into a cell and edit the DNA, it should stay edited. It should be permanent and potentially curative because of that. And in fact, what we reported was, after 3 years, the nonhuman primates looked exactly the same as they did 1 year following treatment, meaning the PCSK9 gene that we knocked out years ago had stayed knocked out despite the fact that hepatocytes have a pretty high turnover rate. So over that 3-year window, all of the hepatocytes that we initially transfused with the vector when we treated the animals in 2017, those hepatocytes were gone. They had turned over and been replaced by daughter cells produced by hepatocyte stem cells. Those daughter cells had a PCSK9 gene edit, so really for the first time ever, we were able to demonstrate that, yes, in fact, gene editing is durable in a large animal and potentially permanent. So very exciting outcome, I think, showing the differentiation between gene editing versus, for example, RNAi or conventional AAV gene therapies which no one really thinks of that as being a permanent solution. It's going to require, in the case of RNAi, repeated administration for life. And in the case of AAV, depending on the particulars of the therapy, it is going to have a relatively limited lifespan.

Got it. And I think part of that paper too is that it shows some examples of the iterative steps in making an optimal ARCUS. And maybe if you can talk a little bit about that and what you've learned from that.

Yes. I told you all the good stuff about ARCUS. I didn't tell you the bad stuff. So the reason that not everybody out there is working with ARCUS, the reason everybody loves CRISPR-Cas is because it is very difficult to make an ARCUS enzyme for a new target sequence. So for example, if we want to edit the PCSK9 gene, we have to make an ARCUS enzyme that recognizes and edits the DNA sequence somewhere in the PCSK9 gene; and producing that editing enzyme is actually very challenging. It's an iterative process in which we make a first-generation version of the enzyme. We test it. We look for off-target gene editing. And usually a first-generation version of an ARCUS will have a relatively high frequency of off-target gene editing. We then make a second generation, a third generation, a fourth generation, sometimes even a fifth to get to the point where we have what we would consider to be a clinical-grade editing enzyme. And that is an enzyme that edits its intended target site with high efficiency but does not cut unintended, off-target sites to a level that we can detect. That entire process from start to finish takes at least 6 months, usually even longer than 6 months to get to the clinical candidate, so there is a pretty -- upfront investment in just getting the enzyme made, but once it is made, it is very, very good. The PCSK9 papers had data from both a first-generation ARCUS enzyme and a second-generation ARCUS enzyme. And we showed that, between the first- and second-generation iterations of that particular editing enzyme, we did see a significant reduction in the frequency of off-target gene edits. And importantly we were actually able to follow the off-targeting profile over that 3-year span that we've been following the animals. It's actually now out to almost 4 years. And what we've seen is that the spectrum of off-target gene edits, even with that very early-generation nuclease, haven't changed over time. So there were some off-target edits that we detected very early on in experiment. And what we see today looks exactly the same, which again speaks to the fact that ARCUS is able to inactivate itself. It's able to turn itself off. Otherwise, we would be seeing the accumulation of off-target gene edits over time, and we just haven't seen that.

Got it. It's interesting. And you guys are supposed to nominate a candidate for PH1 later this year. So when you do that, investors can appreciate that a lot of the iterative steps and a lot of the background work has gone into this final product that you're going to be nominating. And so maybe if you can talk a little bit about the PH1 candidate and just kind of what next steps are with that program.

Yes. We are going to be hosting an R&D Day at some point midyear, meaning later in the summer. And the R&D Day will be focused on in vivo gene editing specifically, so no CAR T, just in vivo adding. And at that R&D Day, we will give not only updated preclinical data for multiple programs, including PH1. We'll also be giving time lines and sort of what's left to do to get those programs into the clinic. I will say that we have shared already quite a bit of nonhuman primate proof-of-concept data showing -- that we have for the PH1 program, in addition to PCSK9 and our TTR programs. We have already demonstrated that we can get therapeutic levels of gene editing in a nonhuman primate. So as far as what steps are left to actually get those programs to the clinic, it is a, I would say, relatively routine -- insofar as gene therapy is routine, a relatively routine development pathway from here.

Got it. And you mentioned ATTR, and you recently reported some data on that. Can you say how far along that one is in the process?

Yes. We reported at the ASGCT conference a few weeks ago some, again, long-term nonhuman primate data in which -- and this was also working with Jim Wilson's group, using AAV to deliver a nuclease and knock out TTR. And we were able to achieve very high levels of gene knockout. So about 95% reduction in serum TTR levels sustained over time. So that's another example of a program that has achieved long-term durable efficacy in a nonhuman primate model and is therefore in a good position to move into preclinical studies in development.

Got it. You're not putting a time line on it, though. Potentially...

Not yet.

Not yet. Okay...

R&D Day.

Okay, okay. And let's see. And you guys recently announced a collaboration with Eli Lilly as well, and it's a pretty interesting one. I'm just wondering if you can talk more about the progress so far on DMD and any other details on the other 2 undisclosed targets.

Yes. So what we have shared publicly is, as you say, there are 3 initial targets in that partnership, the first of which has been named publicly, which is Duchenne muscular dystrophy. We have shared data showing proof of concept in human patient cells from a DMD patient essentially establishing that we can use a pair of ARCUS enzymes simultaneously in the cell to delete a really large piece of the dystrophin gene. It's about 0.5 million base pair segments of the dystrophin gene that we can delete and effectively remove a mutation hotspot that is responsible for about 60% of the cases of Duchenne muscular dystrophy. And in doing so, we can now restore expression of dystrophin in patient-derived cells that previously didn't make any. So we've shared [ at ZAT ] cell-based data for the DMD program. We have not yet disclosed what the other 2 programs are. And I would expect that, at the R&D event that we will be hosting later, we will give an update on the DMD program at least.

Got it. And will you reveal more on time lines for that one at the R&D Day as well?

I cannot commit to that because we are partnered with Lilly. And so we have to respect their wishes and what they would like us to present, so I would say possibly, but I can't commit to it.