We're going to go ahead and get started. Welcome, everyone. Thanks for joining in. I'm Gavin Clark-Gartner, one of the senior biotech analysts here at Evercore ISI, and I'm really happy to be here with Stephen Uden, who is the CEO of Rallybio and co-founder. Thanks for joining us, Stephen.
Thanks, Gavin, and thanks for inviting us.
Absolutely. Why don't you just start off with a quick overview of the company where things stand today, and then we'll go through the programs in more detail.
Yeah, absolutely. So Rallybio, we're a rare disease portfolio company based in New Haven. We have four programs under development. Our lead program in this rare hematological condition that impacts the fetus while it's still inside the mother's womb. We have a complement program, a complement factor five inhibitor that would be given subcutaneously, very, we think, would be best in class. A Matriptase-2 antibody for treatment of iron overload and also a small molecule therapy for hypophosphatasia or rare bone disease. And one last thing I would say to you, Gavin, is you very kindly invited us to this conference last year. There was a big question mark over the whole portfolio other than FNAIT, and in the last 12 months, everything has turned around.
Complement's back on its feet, Matriptase-2, we've sort of made it best in class, and lo and behold, the small molecule approach in hypophosphatasia seems to be working gangbusters. So back to you.
Awesome. All right, let's start off on the complement C5 side just because you guys had an update this week. So maybe just give us the higher level summary. You gave an awesome amount of detail on the assay side. Maybe just kind of give us a higher level summary of that update where the program sits today.
Absolutely. So again, back to wind the clock back to this time last year, we'd just taken our complement C5 inhibitor based on the Affibody platform. We'd taken it into a SAD-MAD study, and there were two bits of data that came out of it that were sort of not what we expected. The first was mild tolerability issues reminiscent of what had caused the program to come to a crashing halt when it was in Sobi's hands at 10 mg. We'd got up to over 100 mg, and we started to see adverse events, which we suspected were reminiscent of bacterial host cell proteins due to the manufacture. Very not severe, but similar to those that were seen at doses of two and 10 mg back in the day.
The second thing that we just couldn't understand is, despite all of the preclinical work that we'd done with the assay, the ability of the 116, as we call it, to reduce free complement factor five was less than we'd predicted. I just think the idea is, remember, this drug binds the complement factor five. Inevitably, there's going to be a small amount there. It's a protein-protein interaction, and it's that free C5 that causes the damage. But with the assay that we were using, it would appear that there was more complement free C5 than be expected. The update, to summarize, to answer your question directly, Gavin, is number one, we put the drug into an enhanced manufacturing process, which used an additional resin to remove host cell proteins, which we licensed from Affibody themselves. They'd seen similar issues with other parts of their portfolio.
We presented data from mass spec analysis that showed we'd virtually eliminated all of these except for one and reduced that two log fold lower, and we went back and looked at these free C5 data, which didn't make any sense. Just as we expected, when we used hemolysis, which is the very labor-intensive way of looking for free C5, there was no hemolysis at the concentrations we would expect that we studied. We've sort of interrogated it, and we realized it's due to limitations of the assay. When you reset the free C5 level specific to the assay, we're well within 99% knockdown of free C5.
Suddenly, the C5 program is back where we wanted it to be, a subcutaneous one to two ml injection that we could be given every once a week, stored in the bathroom cabinet, and do everything that Ultomiris or Soliris does. I'll stop there.
Yeah, that makes perfect sense. And then thinking about next steps for this program, what you laid out will give you confidence on the efficacy/pharmacodynamic side, also the safety or tolerability side. You're running some additional cohorts for the phase one portion now. Maybe just kind of quickly frame what you're looking to see there. And then as you think about next steps for the program, what's that?
Great. No, a couple of good questions. So the first thing, yeah, we are going to go back into volunteers. We're not going to mess about 100 mg. That was well tolerated. So we're going to look at 150 mg, and it'll be an adaptive design. So probably 225, but we'll have the ability to adjust up or down. While we're confident that 100 mg would work for the vast majority of patients, what we learned working with Soliris and Ultomiris in our previous lives is there are patients where you do need to be able to increase the dose. 85% do great on, for example, Soliris, 85%, a significant number do really well on 900 mg, but there are those who need that 1200 mg. We want that flexibility. So we'll be looking for good toleration.
And then, in addition to the assay with the reset cutoff, we will use hemolytic assays as well just to double confirm that we've got what we want. So we'll be getting data from that in the second half of next year. The rate limiting step now is simply turning the drug product into drug, no, drug substance into drug product and getting the study started. The next steps, what we're going to be looking to do is we will need to capitalize off that, but go in and study a small number of PNH patients. Two reasons for that. There's no better syndrome to confirm that this is a real complement inhibitor than PNH. PNH is fundamentally a C5 disease. And there is actually, you'll see in our presentation, huge unmet need for this.
Patients don't want to have to go to the hospital every two to six weeks or whatever. And then we're looking at two other syndromes. One is something called antiphospholipid syndrome, recurrent thromboses. It is a rare disease. I personally treated patients with APS as a junior hematologist. The treatment back a few years ago was warfarin. It's still warfarin. And there are probably about 20,000 patients a year who don't respond, and we think there would be a, and there's lots of evidence that complement is involved. And as an aside, APS is associated with recurrent miscarriages, so hence the sort of interest on the FNA pregnancy side. Probably wouldn't go straight into pregnant women, women trying to get pregnant. And then obviously, again, look at our presentation, GMG, generalized myasthenia gravis, still a desire for that. So that would be another option.
So depending on where we are and where the capital markets are, definitely PNH, as soon as possible, APS, and then GMG would be a nice add-on. I'll stop there.
All right, that makes sense. Let's stick with the rest of the earlier pipeline. We'll come back to FNA, not because it's less important, but because there's been a lot of conversation over that last couple of years, including us specifically. So maybe let's go over to Matriptase-2. For those who haven't looked at this program as closely, what exactly is the mechanism? And you're showing some non-clinical data at ASH. Maybe just remind us what that is.
Yes. So Matriptase-2, it's the sort of membrane-bound protein that is responsible for removing hepcidin from circulation. Now, hepcidin is the, I guess, the hormone or the, yeah, the hormone that controls iron levels. I think everybody in this room will know that the body is extremely good at absorbing iron, not very good at excreting it. Hepcidin is the hormone that says, bring in more. Or bring in more iron. By inhibiting Matriptase-2, sorry, stops you from bringing in too much iron. By inhibiting Matriptase-2, the hepcidin level goes up, so you put a block on the absorption of iron and the release. So you're not going to keep cranking this stuff in, and then also it prevents the release from iron stores. Now, that's really important for patients who require recurrent or frequent blood transfusions because the body can't excrete iron.
So if you're one of these patients, for example, like thalassemia or polycythemia or these other sickle cells that require frequent transfusions, you feel great, but eventually the iron overload catches up with you. The hypothesis is by increasing hepcidin levels by blocking Matriptase-2, you prevent iron absorption, you prevent iron release, and enable the patients to cope with their frequent blood transfusions.
Yeah. Excuse me. Maybe remind us to, while you're having a coughing fit, I'll keep going. Remind us to what other Matriptase-2 programs are in the clinic, what clinical data has been generated, is still being generated, and how does that inform your own development?
PNH. Quick bottle of water. It's how you get out of.
Quick water break for everyone on the line.
Yeah, so when we were very proud of ourselves when we found this, dear me, this monoclonal. However, no sooner had we brought it in than Regeneron came out of stealth mode and said, haha, you never guess what. We're just going into the clinic. Now, what they've been able to show is that the mechanism works, hepcidin goes up, you can control iron levels. So sometimes it's good to be a follower because the path is being mapped out for you. The one thing that we did notice was that the monoclonal was a pretty ho-hum monoclonal in terms of duration of action and volume of drug that needed to be given. It just so happens that one of our team was the inventor of Ultomiris, the long-acting Soliris.
So realizing that we were having to conserve capital, Doug said, "Right, well, I know what I'll do in the spare time. I'll re-engineer our molecule so that it'll be sort of best in class in terms of duration of action." So the intent would be once we capitalized, we've got the cell banks. So as soon as we're capitalized enough, we can get on with GMP manufacturing talks and what have you, and then sort of follow the path that Regeneron and Disc Medicine are very kindly laying out for us, but we think we will have best in class. Now, ASH will be presenting those preclinical data that demonstrate the work that Doug did on the engineering and why we think we've got something special here.
Awesome. Look to hear more. On the oral ENPP1 side for hypophosphatasia, remind us, how does this mechanism differ from Strensiq? What room does Strensiq leave on the table to improve upon? What are you trying to do with this mechanism?
Great. So quick reminder, hypophosphatasia, it's due to either total absence of alkaline phosphatase or significantly, i.e., more than 95% absence of alkaline phosphatase. Strensiq is enzyme replacement. Incredibly expensive molecule. It is, in all fairness to Alexion, now AstraZeneca. It's a very complicated molecule. Also fairly burdensome. You need up to six injections a week, sometimes more than one injection each time. Weight-based dosing, so incredibly expensive. So at the moment, Strensiq is only used, and probably for the foreseeable future, is only used in patients with total absence of alkaline phosphatase. These children that are born with no calcification of their bones, and they're going to be on this for life. We now know from genetic studies that HPP in its entire spectrum is actually quite frequent. Probably about 80% of so-called HPP patients. Symptoms don't manifest themselves till later in life, teenage, adulthood.
They're often misdiagnosed as osteomalacia, osteogenesis imperfecta, dental problems. For these patients, there is no treatment. So this would be a therapy that would treat the other 80%, and Strensiq would be doing what's needed for those patients with total catastrophic loss. The approach, so whereas Strensiq replaces the enzyme, our approach is to reduce the substrate. So alkaline phosphatase turns pyrophosphate into phosphate. And there's been sort of quite intense academic debate as to is hypophosphatasia primarily a problem of lack of ability to make phosphate to bind to calcium, or is it actually the problem of a buildup of pyrophosphate, which is in fact inhibitory for calcification of bone? And this is, academics love to argue over this. I think we're going to help them answer that question, as our industry often does. They argue for years.
We bring along the drug, and then that's the end of the discussion. It would appear that certainly substrate reduction is going to be a big player. And one of the reasons that was that question is there are other ways that pyrophosphate can be converted into phosphate. So there's always some residual activity. So hopefully that answers the question. This will be a drug for anybody who's got impaired alkaline phosphatase activity, and it will be a once-a-day pill rather than six times a week injection sort of thing. So I'll stop there.
Yeah, great. I'll just ask one more on the ENPP1 side, and we'll go over to FNA. You've been working with Exscientia, I think on lead optimization for a fair chunk of time now. What exactly has been the rate limiting step there?
People are probably aware ENPP1 inhibitors are being looked at in oncology. But of course, the profile. All drugs have to be as good as they possibly can, but oncology is one thing. This is a drug that would be given to somebody from probably infancy all the way through to adulthood. And fundamentally, the issue has been all around. We'd have a molecule that was highly bio available, but there was a small problem with liver enzyme induction. So you'd fix that, and then so it's really sort of trying to get to an across the entire, and we agreed before we started a list of all of the criteria for a drug that could be taken for the patient's life. And we finally got something that literally the Goldilocks drug that does everything just right. I'll stop there.
Yeah, that's great. All right, so let's go over to FNAIT last, but certainly not least. We probably start with a 30-second overview of what FNAIT is and then go into lay out your trial design and kind of the cadence of dosing since you're just getting that off the ground.
So FNAIT, rare syndrome, it's a syndrome where the mother of a baby makes antibodies against her own fetal platelets. So although the baby's all wrapped up and nice and safe in the mother's womb, the antibodies that the mother makes cross the placenta and destroy the fetal platelets to the extent that even the smallest bleed, bleeds and bleeds and bleeds and bleeds, and that causes intracranial hemorrhage, miscarriage, major problems during delivery. And it can be from everything that we know from a similar, more prevalent red blood cell problem and from everything that we've studied in various models of this syndrome, it can be prevented by giving the mother a very small dose of the very antibody that you don't want it to make. So this antibody cleans any fetal platelets that get into the mother's circulation are quickly removed so she doesn't have an immune response.
So our therapeutic, as I say, is two-one-two. It's an antibody that will remove fetal platelets from the mother's circulation. We've tested this incredibly extensively. I'd sort of challenge anybody. I know I'm going to be in a meeting like this and I'm put in my place, but we struggle to think of any therapeutic where the first use in patients are pregnant mothers. Okay, so I've worked on programs where five years after approval, you get label extension, but we're going from phase one straight into pregnant mothers. So we've had to, we've done extensive maternal fetal toxicology. We had to work to create a model for that. We've tested this in models of the disease by injecting platelets into volunteers, PKPD, and we're now ready. We've now defined the dose, sorry, the exposure level that we need.
So what we need to do is tread the path by making sure that there's adequate drug to mop up platelets, these bleeds, but not so much that the antibody that could be toxic gets into the mother. And we think we've got a pretty safe margin there. As we go into pregnant mothers, as I say, there are three risks, I think, to the program. Risk number one is, will the regulators allow you to treat pregnant women? I'm happy to say we've checked that box. We've got the green light to start phase two from the Europeans, where the phase two study and in fact have started screening already in Norway. The second question is, of course, operational. Will a mother actually sign up to take part in the study? Certainly two things.
All of the reassurance we've had from doctors who look after these moms say, yes, they will. Secondly, as I said earlier, screening has started. So women have had this all laid out to them and are still prepared to be tested for it with the idea that they would go into a clinical trial. They haven't actually been told, right, you're at risk, let's take the drug, but they're signing up all over the place. And then the third risk is purely technical: how do the changes that occur in pregnancy impact pharmacokinetics? We've modeled it. We've had it in animal models of it, blah, blah, blah. We've just recently published on how we're going to select the dose. But at the end of the day, we need to put it into a pregnant woman.
As the blood volume increases during the pregnancy, as the metabolism changes, the presence of the fetus itself with the antigen, how does that impact the drug levels? And we'll be answering that as we go. One other thing, the good news is that this is a totally objective endpoint drug level. So unlike many phase two studies, as patient by patient, we will be able to share data because there's no sort of chance of bias here. So yeah, so that's the fundamental design. Oh, sorry, the design is we'll treat one mother, total safety. We'll then treat three mothers through pregnancy and then four. So it'll probably take sort of two, two and a half years to complete the study. I'll stop there.
Yeah, that makes sense. And we've spoken about the therapeutic window in the past since it's a very important concept. And the paper you put out was wonderfully detailed, which we're not going to go through today. I just wanted to advertise that quickly. And then final question just on the cadence of enrollment and data updates. When are you dosing pregnant women in relation to the pregnancy? So is this going to be the first sentinel mother, wait nine months, give or take, and then go for three more? Is that kind of the cadence of updates?
Yeah, so just a quick bit. We will only treat women who've reached at least week 16 of the pregnancy beyond because that's after that the antigen starts to appear, and we don't want to get confused. Did the mother actually end up immunized before we blah, blah, blah, but that will mean that we will be following that. Now, we all know pregnancy lasts nine months, but I think we all know that women don't rush to the doctor and say, I think I've got pregnant last night, so it'll probably be from starting dosing to the baby being born, we've predicted about six months, so it'll be a, we'll follow that mother. As soon as we see that the baby is healthy and hasn't got thrombocytopenia, God forbid, we will start to recruit the next cohort of three women.
We will follow that mother to make sure that she didn't alloimmunize, but as soon as we've got a fix, and the design of the study enables us between each cohort to adjust the dose. It isn't a sort of old school dose response x, two x, four x. Each group of data will go into our PKPD model. We will make any adjustments, and then the next cohort will have that updated. It's data-driven dose selection rather than best guess before we even start.
Yeah, that sounds great. So progress on all fronts, main takeaway. So I'm looking forward to exciting 2020.