BioMarin Pharmaceutical Inc. (BMRN) Earnings Call Transcript & Summary

Good afternoon, everyone. Thank you for joining us at the Jefferies Gene Therapy Summit. I have the pleasure of hosting Hank from BioMarin, the Head of BioMarin's R&D engine, and I really appreciate him taking the time to join with us, especially right after earnings which is always kind of a crazy time. So thanks so much, Hank.

Pleasure.

Now -- and by the way, I'm Akash Tewari, a senior analyst covering biotech at Jefferies. So because it's a Gene Therapy Summit, I think it's only fair that let's start off by talking a little bit about vosoritide. But like I think obviously, there's -- you have 2 very critical FDA approval decisions, one on ROCTAVIAN, one on VOXZOGO. I think on the call yesterday, a question that I feel like we were honing in on and so is The Street. Talk to me about how the agency looks at average growth velocity as an endpoint for achondroplasia and then also final adult height. And where has -- reading the tea leaves, where has their kind of thinking on both of those endpoints evolved over time? And then when you think about the regulatory package, you're submitting, which is you're taking 5 years worth of follow-up data from your Phase II study, how do you feel like you've looked -- you've adequately addressed both of those questions, average growth velocity for vosoritide and then also final vosoritide.

Even in the Gene Therapy Summit conference, the question is quite relevant actually anyway because the 2 applications, vosoritide, VOXZOGO and Valrox have in common in the U.S., a greater consideration of durability and FDA desire for conviction around durability. In the case of VOXZOGO, just sort of spurring out an extra. 1.5 centimeters, I think there's no real coral from the agency's perspective, but that's statistically significant. The question is that for that to be meaningful in the life of an individual what they kind of please that's got to add up to something that's more than 1.5 centimeters of height gain, and we showed them what 2 years does in 110 patients, and we showed them what 5 years does in like 10 or 15 or 20 patients from the open label ongoing study. And I think their judgment is just around -- the judgment they have to make is, is that enough to give us conviction that AGV is itself clinically relevant by itself. Or do they want to hold beat to the fire and say, as a post-marketing commitment that we can legally hold you to, we want to see what it adds up to. And I think that's the ongoing debate at the agency. One version looks like an accelerated approval calling AGV, an intermediate endpoint, another looks like another version of that is kind of like blood fee is now accepted as a clinical endpoint. And PKU and maybe that's where they get in regard to AGV. It's -- I don't -- I don't have a crystal ball about what their thinking is. And just to be sort of fully transparent about this, we don't know that they're going to approve the drug. I mean, this has to make its way through the final, final, final stages. But that's my impression of why they were asking for longer-term studies in the first place. There's a whole literature about and regulatory precedents around growth hormones, where they approve things for AGV, but they're really interested in final adult height. I think they think about growth hormone differently than the vosoritide. Because growth hormone gives you a big bang initially, but then tapers off to nothing, which is why they got so interested in final adult height for the growth hormone indications. On the vosoritide side, you don't have that big bang, so it's got to add up to something. And so that's what I think they're wrestling with I think, because we don't have infinite visibility into what they're thinking and discussing. But the end result of all this, I think, is that we have really strong solid clinical data on improvements in annualized growth. There's no debate as to whether we improve annualized growth velocity. And there really isn't much debate about its durability past the first year or the second year. I think it's really more -- I feel like it's more about a regulatory policy kind of a decision around accelerated approvals. And I think the same kinds of considerations are in play different center at the FDA. It's VOXZOGO's regulated by the Center of Drugs. The gene therapies are regulated by the Center of Biologics. But the same fabric of question, which is we want more confidence about how it ultimately is going to play itself out. In the case of VOXZOGO, it's unknown what we add up to in large numbers of patients. And in the case of Valrox, it's -- you've got these very theoretical considerations out there, and we'd like to make sure that we're clearing the bar more substantially than just 6 months' worth of data in 17 people, which is why they moved the goalpost to be 134 patients for 2 years. Again, so that they feel more confident that we're way above the bar of benefit risk calculus. And what's hard about Valrox is what is the safety issues that were the discussion topic of the FDA's advisory committee. So let me pause there.

No, no. That's really helpful. And let's think about in a situation where you get accelerated approval and you have to run, let's say, some post-approval confirmatory study or some type of requirement from the FDA. Could that -- would that be as simple as taking your Phase III population and continuing to see how they have a durability of effect going out to 3 to 4 to 5 years? Or would that require another clinical trial? Or is that just really too early to speculate at this point?

Well, that's too early to talk about us where it tightly settles and that's the work over the next like 20-something days between now and the PDUFA. I think that it would be tough to do a new study with a placebo control over a longer period of time. And I feel like it's -- that would be a big ask of an FDA of a patient community. And in fact, the FDA has asked sponsors for example, in Turner syndrome, they ask the growth hormone sponsors to do a randomized study to final adult height, and that didn't really work out very well. So I think that everybody is going to try to figure out how to use the patients that we have as the leading edge of telling us what we need to know about durability. We have a lot of natural history data that's actually really strong. I mean we have like I think it's like 4 different studies. We have the placebo run-in. So we can speak to the FDA about the appropriateness of the natural history comparator population for comparison with long-term efficacy outcomes. So I think we have a strong case to be made that if you really want to see what this adds up to, which, by the way, we do too, that the best group to count on to get a meaningful answer that was going to be 110 patients who are originally randomized. And since there won't be an ongoing placebo control, you have to synthesize that control. And I feel like a very good case has been made as to why natural history studies are in reasonably good control. And so then it gets down to, okay, how strong is the requirement going to be. Is it a post-marketing requirement consequent to a full approval versus a post-marketing condition consequent to an accelerated approval? I think that's where the zone of the discussion is going to be. I think with Valrox, that's zone of the discussion under ability is just patients as are on control. They were bleeding at this rate while they were on Factor VIII replacement therapy. We took it away, how long do they go without bleeding. So there -- I don't think there's like a new randomized controlled trial. I think that's why they just sort of kicked the can to 2 years ought to be enough for our purposes.

Makes sense. And you made an interesting comment on the call yesterday, where you kind of mentioned we only have 4 patients worth of data in the Phase II study who actually reached final adult height. And that's -- I think like a lot of investors are trying to figure this out. I'm trying to figure this out. Can you talk to us about like, well, what is final -- does that mean after 5 years, we say stop? Or does that mean when the growth plate stop actually -- when the growth plate closes, when you actually reach final adult height, like what is that endpoint? And how is that kind of tracked from a clinical perspective?

Yes. That's a very technical question or it has a very technical underpinnings underneath it. So I think the most legitimate scientific definition of final adult height is when there's documented epiphyseal closure -- growth plates are closed. And that's a radiographic thing. I think that people don't really love the idea of submitting children frequent radiographs for the documentation of epiphyseal closure. I think what the natural history data indicate is past a certain age, the population incremental height gain is not very high. So we could go back and forth with the agency to say, do you want us to document epiphyseal closure. Or do you want to just say what is the height, let's say, age 18 in the reference population compared to the treated population. I think we've taken the sort of pragmatic view of in the first instance, let's think about that as a function of age where we see the growth curves to sort of flatten out in the untreated interest of population. And I think so far that book floats. But it's -- again, the ink's got to dry on the pen on the approval.

Understood. Look, I will say one thing that I noticed when you read the FDA advice -- the adcom on achondroplasia from a couple of years ago. There's literally quotes where someone says "jee haan," like if a company shows efficacy in a Phase III trial, it's going to be really hard to make them run another study. I think it's notable that at least talking to right now, the idea that the FDA is going to take what's the most common form of dwarfism in the United States, and you have a drug that has shown efficacy and safety in two well-controlled trials, to not allow them to actually launch for 4 to 5 years as you have to run another Phase III. It seems very unlikely to me talking to you, that's kind of the impression I get. But I'd love to get your take on like that worst-case scenario.

Well, so again, I mean, some policymaker at the agency who is high up may do that. I mean they say screw you patient community get to work. You got to do your job here. At the FDA Advisory Committee, it was a little bit unnoticed, but they actually had an emphasis from the FDA present and talk about ethical considerations in the conduct of pediatric clinical trials. And one of the things that she pointed out was that unlike clinical trials in adults, which start with equipoise, meaning uncertainty as to the man who tested it whether there is a treatment benefit. In clinical trials in children, there must be a prospect of benefit, else, it's deemed unethical to enroll children in clinical trials. And what you're really -- the burden of expectation there is proof of safety and that, that and it's about the magnitude of benefit, not whether there is benefit. And I think that those people that you're talking about responding to at the committee is their judgment about how long is it reasonable to ask people to participate in placebo-controlled trials. And clearly, the community has -- there's normally a tension between the investigative community and the agency. The agency is always like go longer and the investigative community is you don't have to actually talk to the patients. And so that discussion lives there, and it kind of got propagated when they said, we'd like to see the 2-year data. And so then when we provided 2-year data, their mood kind of shifted to, okay, well, even the 2-year data doesn't really answer the ultimate question. So I think that they recognize from an ethical perspective that it's going to be hard to push the ultimate question. You make the company running much longer placebo-controlled trials. The patient community, it's been trapped. It's failed in both one of those internals, for example. So I think the -- I think it's gotten a little bit less -- it's gotten less notice, but there really are important ethical considerations about how long you can run the perfect -- FDA likes things. They always say, you're going to make my job as a reviewer much easier on the back end if you do a 5-year placebo-controlled trial. Never mind the impossibility of talking people into that. They want to make their job easier for them. And I have sat with an empathy for the challenge that, that represents. They have to apply judgment. At the end of the day, all the regulatory decisions or judgment decisions around does the benefit outweigh the risk? And I think we gave them a lot of good information to enable them to make that decision. The treatment effect is absolutely comprehensively real, and the biology indicates that the durability is going to be just fine and the clinical evidence collected to date is -- and that was the framework that the advisory committee put the question of durability into at the committee. The FDA just more better. But again, we're getting closer to when a real-world decision has to take effect. What you also do to the community is very willing to extrapolate. And again, this is another place where the community tends to be in one place and the agency tends to be in a slightly different place. And the agency -- so the community said to the agency, look, if you have positive data in patients over 5, the children under 5 provided it's safe are going to be really in seeking this treatment. And the agency's typical view is to say, after we've gotten pivotal data from that population. Now the EMA, I think, exemplifies a little bit more of the pragmatic like live-in-the-real-world kind of things like we recognize this is a fact that, it's genetic. It starts from birth. The earlier you intervene, you're better off. You've shown us some safety and some pilot data, that's good enough for us. That doesn't tend to be how the FDA works. But let's see, still some water to flow under the bridge.

It's -- recently you mentioned that, Hank is, my associate, Leo and I, we were looking at your EMA review. And in that, you guys did a pretty interesting extrapolation analysis on the potential height going out to age 16 with your Phase II patients. And you -- I mean, you had something between 9 centimeters all the way to 20 centimeters, there's obviously a lot of variables there. But it did seem like the agency, at least in Europe, was supportive to that type of analysis. And they also agreed that, that additional Phase II -- that 5-year Phase II data was supportive here as well. That -- I'm sure you presented that analysis to the FDA. What has been kind of the feedback you've gotten on these type of extrapolation studies and their usefulness in terms of predicting clinical effect.

There probably aren't two dirtier words to use with an FDA regulator then extrapolation and the other one is modeling. They're just really grounded in like the reality of present evidence. There's -- I think one of the challenges that the agencies have with all the vacancies they have is that scientific layer that -- so like reviewers are like a little bit like traffic cops. It's like, okay, you're 6 inches over the curve, ticket. They don't think about like, okay, there's no obstruction, right? The leadership level above that is the group that's supposed to assemble those reviews and then make recent judgments about things. But that layer at the FDA, they've had a difficult time hiring and filling people into those layers. And I think that might explain some of those sort of weird decisions that get made. Where reviewers come out in one place and then FDA management comes down in a very different place. They've got some wood to chop in terms of bringing those different perspectives together on their side, which is why I keep harping on the it's not over until it's over. But the flip side of it is the evidence here is really strong. The biology here is really strong. The impact on patients is really clear. Where the patient community stands on this matter is not particularly difficult to interpret. I mean it should work, right? We should be able to get this across the finish lines.

Makes sense. Well, I guess with that, let's actually ask a gene therapy question. But I think the way you framed it and you were using that and that comparison between ROCTAVIAN and VOXZOGO is relevant. To me, if you look back at the CRL, like -- yes, there was -- okay, it wasn't a one-shot cure. The factor levels wane over time, like sure. I guess that's one thing that we found out with the updated Phase III data. But to me, when I looked -- I put myself in the FDA seat, maybe they weren't sure whether your Phase II data set with those 7 patients and those 2 hyper responders and your Phase III data set were the same commercial product, right? Like that seems like the Nexus of the -- what the FDA was trying to figure out. Now that you've started to -- you've gotten what is "the more real result" and real clinical benefit for ROCTAVIAN as you've treated it in more and more patients. Do you feel like -- do you agree with that assessment that it was really -- they didn't know if these were the same 2 products that -- that led to the CRL on ROCTAVIAN in the first place. And now that you've presented more data and you're going to have data at the start of the year with the Phase III, how do you feel like you adequately answered that question to the agency?

What they wrote in the CRL is crystal clear. What motivated them to write that is not as clear. Because there hadn't been a lot of discussion that went something like, okay, we're willing to accept the durability of the Phase II study as evidentiary supportive durability of the commercial product ROCTAVIAN, under certain conditions of similarity of the Phase II and the Phase III results. That discussion never happened. So it -- that's why I was so out of the blue and it's like, okay. Yes, we acknowledge that there's a difference here between the Phase II and the Phase III, but we don't ascribe meaning to that difference because even at very low factor levels, there's excellent hemostatic efficacy. And we also showed them that the magnitude of the change in Factor VIII expression from the end of year 1 to the end of the year 2 is informed by the peak of the Factor VIII expression. So I think their worry was, okay, it's not going to go -- it's not going to launch as high. And if it degrades as fast as the original 60 cohort degrades, the whole thing could be over by the end of year 2. And so they have -- one point there's a common is like anything could happen, this thing could fall off in a bit. And I think they just -- again, it goes down to the reviewers. If you haven't proven something, anything else is possible until you've proven that, that didn't happen. And I think that together with what the community refers to as the unknown unknowns. We don't know how it's going to long it's going to last for sure combined with there are these unknown unknowns, let's hit the can. And by the way, this community is served by pretty good alternative products. So we can -- we feel like we can kick the can and so if we surprise you sponsor about this, our bad. And that's kind of my interpreter. I don't know exactly what motivated them to say lean left instead of lean right on this one. But that's -- it's subject to interpretation. I will say that since the CRL, they've been very consistent up to year day-to-day. It's not like Wall Street pushed us hard on can you get them to get started faster. We push too, but their consistency was remarkable. And as discussions have gone further, it really seems that's where their itch is. The 2-year data. It doesn't feel like that ask questions about beyond 2 years, beyond 3 years in the efforts. But they've surprised us before. I hope they don't surprise us again. We're working very hard to make sure that they have an opportunity to express themselves fully so we can more fully understand what it is that they're looking for and set appropriate expectations. The expectation they've given us is that they want to see the 2-year data.

Understood. Understood. And when -- let's define what they mean by the 2-year data and what success. Because you powered your trial to show superior standard of care. That was something that you felt as important originally wasn't powered that way. You powered it how to do so, a, what is your degree of confidence? You said like the agency hates extrapolation and modeling. And I feel like that's all we do. But when we extrapolated out, we personally have a pretty high degree of confidence that will show superiority over standard of tariff. But I'd love to get your take in terms of what your expectations are with that Phase II data or that 2-year data? And then what do you think the agency specifically is looking for to feel like this is not "falling off of a cliff, " and it kind of meets their internal thresholds.

Yes. We're staying away from modeling and extrapolation because of the hate -- the love that comes back to us when we talk to them about that. So we stay pretty grounded in here are the facts. So the facts are that in the end of 17 group that prospered all the way through the end of the second year, their Factor VIII expression levels were higher than the 4E group was at the same time point. It was lower than the 2E -- than the 6E group predominantly because it launched lower. So it declined more intermediate between the 1-year pretty significant fall in the 6E group and the 4E pretty undramatic fall. So we're starting year 2 at a higher Factor VIII level than the 4E group did, and the 4E group didn't bleed in year 3 or in year 4. So that gives us reasonable optimism that the data beyond 2 years will fit with what we've seen in the 4E Group. And we have confidence that the end of 17 is going to be reasonably predictive of the end of 134 because their Factor VIII levels were so similar to each other at the 1-year mark. And we think that the 1-year Factor VIII level is going to be prognostic of what the factor VIII level at 2 years is. So I think it's a fact pattern, which gives us confidence that ABR is going to be fine. Factor VIII levels are going to be fine at the end of year 2. It's not like there's going to be a fall on Factor III expression. But if you're talking about 4, 5 years of bleed-free, protein free, that's a pretty significant transformation for these people.

Yes. I think -- I mean -- and to your point, honestly, if you just remove the 2 hyper responders in Phase II, your Phase III and Phase II data sets look...

That's another thing that they don't like sort of removing basis.

Right, sure. Well, we talked about the hyper responders. But I'm curious, does the agency think about nonresponders? How do they think about them from an approval perspective. And I often get asked by investors like, hey, what's a good nonresponder rate versus a bad one. And I feel like there's this arbitrary number of around 15%, 20% because that's kind of what you had in your data so far, so you just want that trend to hold. But let's say that as you get more Phase III data, that your "nonresponder rate" gets into the 20s, right? How would the agency theoretically look at that? And have they talked to you about nonresponders in relation to getting approval on ROCTAVIAN.

They have a little bit. The issue of what's a meaningful Factor VIII level is the kernel of this question. We do not -- and the way -- and the reviewers who like I say, they're very specific and concrete. And they're very literal and they're very fact-based. What they said was you're going to prove statistical significance in Factor VIII expression in a very small number of patients because of the size, check. You're going to demonstrate a benefit on ABR as a result of that Factor VIII expression level, check. What you won't be able to do before registration is determine the correlation between a given factor level and protection from bleeding because that's going to require a much larger number of patients. We can corroborate that because we only have 2 patients who reverted to prophylaxis so far. And because we see excellent hemostatic efficacy down to Factor VIII levels that are actually undetectable by the chromogenic assay, they're still detectable by the one-stage assay, which is the original historical standard. So these patients are still mildly affected in their -- so they wouldn't be candidates in the real world for prophylactic therapy. So I think we're going to have to build -- we're going to have to follow more patients for a longer period of time as factor levels drift down. Their interest in this, I think, is a patient management interest, which is if it's going to decline, then some people are going to decline faster than other people. How does a doctor know when to restart prophylaxis? That's their question. Now the good news about that is just nothing about ROCTAVIAN that would preclude you from restarting Factor III prophylaxis or emicizumab. So that's really greatly reassuring. You're not taking options away from patients. And so in that context, I think their interest in nonresponders is more of a patient management topic than it is a benefit risk topic. I don't think they think -- I don't get the impression that they think that the failure to launch fraction is so high that causing question of the benefit of this thing.

Understood. That's pretty helpful. Okay. And I guess for the last question because I know we're nearing the end of your time. This is more of like, look, the perception of gene therapy honestly has never been more negative, I would say, than ever. What I mean like there's -- it's true. Now what's interesting is you -- your team legitimately made some -- like the way that you go after gene therapy targets is different than other companies that I've seen because you're going after -- you're trying to produce a lot of protein, right? You go after HAE. I mean you have to produce an amount of protein that I wouldn't think a gene therapy would per se be suited to do. And say we'll see what happens with 307 your PKU, but you did legitimately optimize that vector in a way that it seems like -- like the issue was the vectors for genome copy needed for PKU. That was the kind of physiological hurdle. Now the question I had, like, obviously, there's -- you have these like double mutated mice, that may be a reason why you had those -- the integration of the [ tumors ] that were in those mouse livers. But are you concerned that there's some type of like therapeutic window for gene therapy, where the more you optimize the vector and the better that they're actually getting into the cells, you might also be increasing the risk that you could get integration? Like it's such a theoretical question at this point, but I'm curious what your take would be.

Well, I mean, I do think that, that was a bit of a conclusion of the ASGCT and the FDA adviser giving you roundtable, which is that the more DNA you put in there, the more chance there's going to be for these random integration events that occur. Now nobody knows whether integration in the mouse genome has any relevance to integration in the human genome. And I think it's pretty clear that you can document integration of AV and in nonhuman primates and even in humans. I think uniQure is talking about that as well from that liver biopsy from the 1 patient who had an HCC, integrations happen. The thing that people are losing a little bit of track of is that, this was all expected. And all of this was known at the time of approval, for example, for Zolgensma. And people accepted that, that the therapeutic benefit was enough and that a theoretical risk didn't cause nonapproval of Zolgensma. And not much has changed really in the mouse literature since that decision was made. So I think as long as we stay in dose levels that have been -- that the people are comfortable with, I think we should be fine. That was part of the purpose of the experiment, right? Everybody concludes some as good it more. That was why we did a pharmacology experiment. We wanted to test the hypothesis that more isn't necessarily better. We were completely not surprised that we saw oncogenicity. That wasn't the purpose of the experiment. The purpose of the experiment was, is 2e14 better durability wise than 6e13. And the answer is that 6e13 in a mouse appears to be just as durable as 2e14 and you're not burning out the cell by going up to 2e14. That's important information for us as we feed the clinical trial. Like, for example, if after 4, 6 months, we start to see waning of the 6e13 effect we might say, well, that's not unexpected, and we should go to a higher level of exposure. And so we're ready for that. We've enabled that. As to how that then means on oncogenicity risk, it's clearly going to be a notch higher for the reasons we talked about. But again, it's going to be notch higher for the mice. Nobody has any idea what it means for humans and -- I think the experts that we talked to -- hired experts tell us not what we want to hear, but tell us how agency you're going to think about this. Their comments are, drugs get approved with oncogenicity risks all the time. It's really more of a labeling manner than you do in AMS or you do a carcinogenicity study, what appears in the label. I mean nobody really knows whether that's going to cause cancer new. So I think that's kind of the flavor of where that field is. It's like what are the policy considerations, what are the labeling considerations. I don't think people are showing up like wanting to stop the development of gene therapy for serious conditions.

Understood. Well, look, this is a fantastic talking to you. I think we're at the end of our time. But this is a blast. If you guys have any questions for Hank, you can reach out to Hank and Tracy directly or I'm happy to facilitate a discussion, but I just want to say I appreciate you taking the time, and have a good one.

Thanks, Akash.