Pfizer Inc. (PFE) Earnings Call Transcript & Summary

All right, I think we're live. So good morning, and welcome to the Guggenheim Healthcare Talks, our signature Genomic Medicine & Rare Disease Conference Day. This is the second of our 9 panels, and this will primarily focus on Duchenne muscular dystrophy. I am Debjit. I'm one of the senior analysts at Guggenheim, and joining us today is Michael Binks, the Vice President of Clinical Research from Pfizer; Sam Wadsworth, Clinical -- Chief Scientific Officer at Ultragenyx; Arthur Levin, Chief Scientific Officer at Avidity, Carl Morris, Chief Scientific Officer at Solid; and Ian Estepan, Chief Financial Officer at Sarepta.

So given we have a very packed panel, why don't I start first at Pfizer and more on sort of the challenge that we, on the outside are facing in trying to understand all the emerging data in the field. So even in younger patients, do you guys think that there is a point of no return where any intervention, especially with micro-dystrophin gene therapy might not be clinically responsive?

So I mean, we think it's certainly the case that the earlier we intervene and the healthier the state of a muscle, the more likely it is that we'll get high levels of transduction of the gene and expression of the mini-dystrophin protein. And so we are targeting younger boys before they lose their ambulatory function in the first instance. However, from a patient point of view, the impact on quality of life, if you're in a wheelchair and losing respiratory power and so on, the impact on your quality of life of having a 50% improvement in respiratory muscle function or in the function of your index finger that allows you to maintain computer control or control of your wheelchair, those may be very important. So although it's -- we think it's likely to be more effective earlier on. We don't discount the profound impact that these therapies could have on the quality of life of more advanced patients, even though they may start with a poorer muscle state. We -- any residual fibers may be improved in function as a result of dystrophin replacement.

Got it. So maybe to you, Ian. We had sort of 2 challenges with one of -- the Study 102, the potential imbalance in the older patients and issues with dosing because you have 3 batches of dose -- of the drug product and some patients got underdosed. So if you were to take those two, what's contributed to the [ missed ] patient versus a static benefit in the younger patients?

Sure. So first off, I want to apologize. I think people were expecting Louise Rodino-Klapac to be on the panel today, and I am a pretty poor stand-in for her. She got vaccinated yesterday, and is feeling significantly under the weather. So apologize that Louise isn't on. So I'm going to do my best. I'm certainly outmatched by the rest of the panel members here. Now to answer your question, Debjit though, obviously, yes, there were 2 factors contributing to some of the results that we saw from Study 102. The -- if you look at the data, it's very clear, though, that the imbalance in the 6- to 7-year-old group really contributed to the overall miss in stat sig. And the reason why I say that is because, obviously, we hit that sig in the 4- to 5-year-old group. And the lot-to-lot variability was equally distributed amongst both patient population. So the fact that the baselines were so different in the 6- to 7-year-old group, that definitively led to the reason why we didn't see that stat sig in that group. However, obviously, the lot-to-lot variability contributed overall because if we got better expression from all the patients, that potentially could have still had a more profound impact. And so the younger patients obviously performed very well despite the fact that we had lot-to-lot variability and some of the patients got a lower dose associated with that group.

So you guys obviously have had a lot of time to digest that initial data set. Was there a clear correlation between even in the older or the younger, patients who got the right dose had a more profound clinical benefit and not just the expression of the micro-dystrophin?

So there is some variability as it relates to the biopsies themselves because -- it's interesting, Louise has a very good schematic of this. I mean even in one -- in the same muscle, if you look at different sections of the muscle, you'll actually see different expression within that same muscle, right? If you look around the edges, it's less. More -- in the center, it's actually more. And so there is some distribution as it relates to dystrophin within the entire muscle. That being said, there is a good correlation between -- there's a correlation between function and expression. Obviously, it's not perfectly one to one. You and I, both, as everybody on the panel, let's assume you all have similar dystrophin levels, but some of us will win in a race, right? And so obviously, function is not strictly correlated to dystrophin itself. There are other factors that obviously play into the overall results that you see from -- on a functional endpoint, but you do see a loose correlation between expression and function.

Great. So let's stay with the micro-dystrophin for the time being. And maybe to Carl, you guys had challenges initially with sort of 50% of the capsids being empty and now you've moved to an improved manufacturing process so the fill rate is much better. So from an immunogenicity perspective, the thing that you saw with AAV9 vectors, does anti-capsid contribute to a higher immunogenicity versus capsids with micro-dystrophin gene cassette?

Yes. So I think a 50-50 full-empty ratio is actually quite good within the field, not many people talk about the full-to-empty ratio of the capsids. I think what we've done with our second-generation process is to remove what we had defined as impurity is really the empty capsids impurities and the enriched for the full micro-dystrophin containing capsids. So what that has allowed us to do is dose at the same, in fact, the genome copy level per kilogram, but reduce our overall capsid load. The total viral load is how we sort of refer to it. And so by doing that, we think that we can sort of reduce the hit to the immune system a little bit by sort of reducing the overall amount of antigen or virus that we're administering to the patient. That, on top of the other, the clinical mitigation strategy is the increased steroids and the pretreatment with eculizumab and C1 esterase inhibitor should suppress the immune system sufficiently, and that's what we saw in that first patient that we dosed after coming off hold last month, as we announced. It's really not different in capsids. The full-empty, I don't think there's really any difference in the immunogenicity of both the full versus the empty. I think it's just an antigen, and it's a virus that we're administering. And so that's how the body responds.

Great. So maybe switching over to you, Arthur. You guys are taking a totally different approach here using an antibody targeted delivery of PMO. One, could you talk to sort of the differences in the TfR1 expression between healthy tissues versus potentially fibrotic issues in DMD patients? And then the number -- second point would be, well, the splicing events still happen in the nucleus, right? So even if you sort of deliver a lot of PMOs to the cytoplasm, can you talk to us how the splicing pricing improves?

Sure. Sort of interesting questions. So we've actually looked, at least in the models that we have available to us, for expression in healthy muscle versus disease muscle, and actually transparent level receptor levels may actually increase in some of the disease states in DMD, in particular, at least in the MDX mouse. On the other hand, to your point, obviously, if the tissue is now replaced and -- just back to Michael's point that he made earlier, the tissue is now replaced with fibrotic tissue and adipose, we're not going to be transecting that with our specs on skipper. So like all of us, we're going to have better -- like all these treatment modalities we're talking about today, we will have better efficacy in -- when there is more muscle present. With respect to your last question, there's no question about the fact that we are delivering our PMOs in this case and getting effective splice switching. I think what you've seen, if you've seen our corporate presentation or seen any of the data that we did presenting at meetings, is that compared to a naked PMO, we're talking about a quantum leap in the amount of skipping that we get with a simple administration. So single administration, for example in the MDX mouse, produces less than 0.1 -- less than about 0.1%. Skipping, we do 50-fold better with that with a single administration of the same dose when it's conjugated to a transferrin monoclonal antibody. So that the transferrin monoclonal antibody makes what is a relatively safe class of drugs. The PMO is much more effective and will allow us to lower doses and ultimately produce greater amounts of skipping for these underserved patients.

Got it. So maybe switching to you, Sam. You're -- Ultragenyx has partnered with Solid. As you sort of think through the next steps and bring the program in-house, your AAV8 vectors like presume is where the gene cassette is going to land up. Given the homology between AAV8 and rh74, do you think there's going to be pretty significant changes to the steroid protocols, right? Or the prophylaxis that Solid is currently using versus what you guys might need to use in the clinic?

Well, so I think it's -- our program is decidedly [indiscernible] really comment on any sort of immunosuppression protocol that we might employ. But obviously, some sort of intervention is going to be required, and we're actively researching that in nonclinical models today in preparation for that clinical. So a little bit early for a nonclinical program to answer that question.

Got it. So maybe, Michael, I'll switch back to you. So obviously, given the potentially dichotomous effect that you might see in healthy patients versus patients who are more fibrotic, will you think, for an expensive gene therapy like a micro-dystrophin gene therapy for DMD, would patient preselection based on quantitative MRI or qMRS be more reasonable to choose patients who are more likely to respond as opposed to just giving it to everybody and praying for the best?

Well, I think from what we know about the MR-based measures, whether it's imaging or spectroscopy, there is a pretty good correlation between what's measured by these modalities and baseline age and baseline function. So yes, it's possible there might be some increased resolution that can be eked out there. But right now, we simply don't have the data to say either that MRI is definitely more sensitive or a better indicator of muscle health than age and function. I mean it's -- the data is starting to emerge. It looks that way. But we certainly don't have evidence that, that kind of selection is going to profoundly impact the long-term benefit in these patients. The age of the patient is very relevant to where the patient stands on functional -- on the functional curves that you've seen that boys basically improve up to the age of 7 and then start to decline on average after that. But at the moment, we don't have the data to say MR is -- well, even that baseline function is a good predictor of the level of transduction that's achieved or the benefit in terms of function that's achieved. So I think the jury is out. It's a good question, but I certainly don't think the evidence is there at the moment to make that assumption.

Got it. So Ian, switching back to you. The MDA presentation recently. When we saw the curves in the 6- to 7-year olds, knowing that there was a total imbalance in the functional status of those patients at baseline, but we didn't really see any effective gene therapy. Those patients kind of stayed flat. Now was that primarily sort of -- is this sort of one, is that the ceiling effect that you're probably seeing in the older patients? Or are these patients -- I don't know if there was any data on the muscle fat content or anything that you can capture in these patients, but they did nothing on micro-dystrophin or the fact that they didn't decline was a great thing. Well...

Yes, your last point is the most relevant point. You have to have a good comparator to see if the drug's having an effect, right? There's going to be -- stabilization in some of the older patients is dramatic, right? If they're on a downslope, and you're able to flatten their progression out, that's amazing. That's a fantastic breakthrough. And so having a good comparator would have likely led to seeing that there's actually a profound effect. If you looked at the baseline NSAA of the patients in the 6- to 7-year-old treated group, they were below 20. Looking at natural history and obviously the appropriate caveats associated with that, what we would have expected from natural history was about a 4 to point decline. Subsequently to the data that we produced, we started looking at some other analysis, which also seems to confirm that the boys like looking at [ rise times ] and things like that, that these boys are doing way better than what we would have expected from natural history. And so the fact that you are flattening out the declining patient population is very supportive of the treatment effect and very supportive of exactly what Dr. Binks was saying, which is being able to stabilize the muscles that are still there has a huge impact on the quality of life. And so that's what the data seems to be suggesting right now. And obviously, in Study 301, we will generate more placebo-controlled data that hopefully confirms it.

Got it. Switching to you, Arthur, again. So in your antibody PMO construct, are we in the realm of oncology, where you have 3 or 4 or 4 to 5 PMOs per antibody or a single PMO per antibody? And given that the transferrin receptor also transports iron, from a side effect perspective, what would you be more concerned about overall?

Yes. So first of all, the concept that our antibodies have been well-designed and well engineered needs to be really discussed here. We've engineered each of the components of our oligonucleotide antibody conjugate for the specific task. So for DMD, we're actually using a DAR, which is greater than 1. It's a DAR that's closer to 4 in this case. We've engineered our other programs that are using sRNAs to have a DAR of 1, and we can specifically control that and tightly control that to all the kinds of specifications that one would need. So for DMD, we know that we need to deliver the maximal amount of drug to the muscle as possible. The skipping mechanism isn't as potent as is an sRNA mechanism. But more to the point, we have been able to demonstrate not only do we get good binding to the transparent receptor, but that binding to the transparent receptor and its transport doesn't compete for the natural ligands of the transferrin receptor, which is transferrin. In fact, one actually needs to do that design because there's so much transferrin in the blood, you wouldn't want to actually be competing with transferrin. So yes, by design, our oligonucleotide antibody conjugate is not going to bind to the binding site and compete for iron. So many of the toxicities that you'd expect with a drug that was competing for iron have been designed out. In addition, again, back to the concept that we engineer each and every bit of our construct is that we've also engineered out antibody directed cytotoxicity from that molecule. So these are effector function antibodies. So that we're not going to get ADCC on cell types to which the transferrin monoclonal antibody binds. So again, through careful engineering of both the DAR and where we're linking the particular oligonucleotides and the antibody itself, it's giving us a platform which has a remarkable safety profile. And we've already discussed publicly our non-GLP data and the remainder of the program is not far behind. Sorry. The results -- sorry, and the results of those programs are like minor detail. The results of that is that there are no really -- there are no clinically meaningful changes in histopath, in hematology or clinical chemistry. Sorry, I forgot the punchline.

No. That's great. So just one more follow-up on that then. Unlike oncology, where you want to be as close to homogeneous distribution as possible so you can have the maximum impact of the DAR on the proliferating cells, do you have to tightly control? Does your -- does it -- is your product very homogenous for DAR 4? Or you have a distribution because it shouldn't really matter in DMD as much, right?

So what we're doing and what we need are maybe 2 different things. We're going to do what produces the best quality manufacturer, which will be to control it as tightly as necessary. In the end, it comes down to how much PMO you deliver and that's -- so yes, we will tightly control it because that's good pharmaceutical science, and that produces elegant formulations. Whether you need to do that or not is a different story, but we're going to control that carefully.

Got it. So back to the micro-dystrophin and mini-dystrophin story. So you basically are looking at 3 different constructs in the clinic right now. For each of Pfizer, Solid and Sarepta, the constructs that have been put in the clinic, do you see the -- internally, at least, are you seeing the evidence of the differentiation between the different constructs? Is that manifesting in the clinical data so far? Maybe we start with Carl.

Well, I think all 3 constructs are designed to bind internally and at the membrane and stabilize the membrane. So I think all 3 should have that base functionality, as both Michael and Ian have said, there's a -- the mechanism of the micro-dystrophin should be to stabilize or maintain the muscle membrane integrity. And so therefore, minimize some of the contraction-induced injury that's occurring. So in the older patients with -- they will actually stabilize the functional -- function. We shouldn't really anticipate sort of major gains in function, except for when they're actually maturing in the younger patients. I think that's sort of evident in what Ian's talked about in the 4- and 5-year olds where beyond that, really stabilization or maintaining function is a wonderful outcome when you look at the natural history. We have a unique -- we put in the nNOS bonding domain because we think that, that may provide some additional benefit. By having the nNOS binding domain there, we localize the nNOS to the membrane and enable nitric oxide to be produced and enhance our blood flow when necessary and to reduce some of the bouts -- potential bouts of functional ischemia. So we charge that as 2 of our domain. The other constructs, I'll leave it to them, but they selected different portions of the full-length dystrophin to sort of maintain that overall shock absorber functionality that incorporated some other aspects to show it as a functional surrogate of the full-length dystrophin.

So Michael, do you want to take the question in terms of the advantages to the mini-dystrophin?

Sure. So I mean, there clearly are differences. Another view of your question was has evidence emerged yet in the clinic that these are meaningfully different? And I would have to say these are all early in development. And until we see more full data sets from clinical trials, it's going to be difficult to make those comparisons. And even when we have them, because of the differences in populations in clinical trials, it may be difficult to make those comparisons. But certainly, well, I think all 3 of the constructs involve a highly selective muscle specific promoter, which means that the protein is not expressed outside of muscle and the heart. And that in itself has some impact on our overall safety profile. Our particular mini-dystrophin, as we call it, because we don't think these are a millionth the size of dystrophin, they're about 1/5 the size of dystrophin. Anyway, our mini-dystrophin construct contains the H3 hinge region, which is very highly conserved in ontogeny. And we believe that, that may indicate some important functional contribution of the H3 hinge region. But this is fairly speculative in terms of the differentiation that, that may make in terms of functional benefit. Right now, all we have to go on our rodent data and the clear translation between mini-dystrophin expression and benefits in function and the early clinical data in small numbers of patients that there does seem to be a relationship there. And there are -- there is initial evidence of clinical benefit. But we're still very early on in the development cycle here, and we don't have very much data to differentiate these things.

Before I go to Ian, Dr. Binks, Pfizer just recently changed the Phase III protocol for -- from an enrollment perspective. I believe it was 4 to 7 versus, right now, it's 4 and above. So what -- any thoughts on what's driving that change?

Well, we certainly have -- I'm not actually aware of that change. So I'll have to get -- to challenge your fact or check back.

No worries, we can come back to you later. So Ian, back to you on sort of the micro-dystrophin sort of question and, in the end, what matters?

We'll look function matters, obviously, and I agree with Carl and Michael that the data is very early, and it may take time to actually see that emergence. Obviously, from an expression standpoint, we think that's important, especially from a durability perspective, too. But as it relates to the construct specifically, one of the things that -- Louise did a lot of empirical research to try to find and optimize construct. And one of the things she did when she was looking at all sorts of elements to include in the ultimate construct was she found that R2 and R3 being included where you're actually attaching to the sarcolemma membrane was really important for muscle force. And so this is the reason why we've included that in some preliminary data, even Jeff Chamberlain actually wrote a paper where including those 2 regions are incredibly important and showing a difference in muscle force. So we're going to have to see ultimately, how that translates exactly, to Michael's point, how that translates into the clinical and functional benefit that we're going to see over time.

Got it. Question on the translation of limb-girdle data to DMD, especially in terms of the drop in vector genomes per nucleus like between the first biopsy and the second biopsy, you guys saw about a 73% loss in vector genomes per nucleus. Is that a direct read-through into DMD? And what does that tell you about long-term durability? Is it more like hemophilia A? Or can you even ambition a situation it could be more like hemophilia B, where you've gone 10-plus years and there has been no decline in factor IX?

Yes. We were actually really, really encouraged to see the limb-girdle out to 2 years. From my knowledge, it's the longest expression data we've seen before. [ Then days ] had data out to 1 year, where you saw no [indiscernible] of expression out to 1 year. Now we have limb-girdle low-dose cohort out to 2 years where you're seeing exactly the same amount of expression. Looking at both positive fibers, protein expression, you're not seeing any change at all. To your point, the vector genome copy numbers did go down. I don't think that, that's actually -- that's an artifact of very small numbers. Again, there's some variability probably driving that, which is the reason why we're seeing that change. But we don't think that -- because you're seeing all the other expression measures being essentially equal that, that's predictive of a decrease in expression. So I think it's an artifact of just one patient having a low number kind of driving the B down.

So maybe to Sam, given your sort of experience with OTC and the glycogen storage disorders, how do you think the DMD gene therapy is likely to evolve under Ultragenyx in terms of dose? Do you have the flexibility to take the dose higher compared to the AAV9? Or you don't think that's going to be required, and in the end, higher the dose, more potential issues both from immunogenicity and long-term safety perspective?

Well, this is a really critical question across the field. And our learnings, from our earlier clinical trials, OTC, GSD, also Wilson disease [indiscernible] CDD and the CNS coming on, we have a great deal in the manufacturing technology side, and we believe we're in an ideal situation for a high patient dose, high patient number disease like Duchenne, and that's because we've developed a clonal producer cell line platform. Now why do I say this is ideal? Well, there's a number of things that I think the [indiscernible] recognize the elements of this. We've eliminated the plasma supply chain due to the clonal nature of the PCL. So that's a tremendous cost and complication savings. We have high product yield of 1E14 vector genomes per liter are many times greater proportion of full vector particle certainly fit-out of the cell. That's before [ purification ] and this can be 70% to 80% full. And this simplifies both the cost and complexity of downstream purification. This platform is extremely robust. We can transfer a 200-liter process from our one GMP process [ development ]. So that's a tenfold jump in GMP manufacturer. I mean we can do this without validation [ weapons ], which both saves tremendous amount of time and money. And in addition, our latest iteration of this PCL platform, which we have called HeLa 3.0, can deliver 3 to 5x more product than the version that we're currently. So all in all, if you add this together, we believe this is ideal for Duchenne, high product yield and a robust platform. It leads to lower cost of goods, high proportion of full particles right out of the box, which is -- leads to a lower total vector load or viral load as already been referred to and then coming back to the nNOS domain containing micro-dystrophin. We obviously believe in the functionality of this domain. And when you put all of this together, we believe we're in a very good position to be helpful to these patients, perhaps even older patients, larger patients as we go forward.

Got it.

Can I just get back to -- you asked me about the age of inclusion changing in our pivotal trial protocol that started in December. There has been no change to the age inclusion, which is 4 to less than 8 years. So just as a point of clarification, I'm not sure where you got that information from.

Got it. My bad, I thought there was. But anyway, [indiscernible]. My mistake, apologies. And so to the micro-dystrophin constructs that are in the clinic right now, we talked about micro-dystrophin expression levels, we talked about vector genomes for nucleus, we talked about localization to the fibers -- micro-dystrophin positive fibers. Which one of these 3 things is the best predictor for a functional benefit or each of them are important for different aspects of it? Because in the end, I guess, everybody is interested in that functional benefit. And if you have to look at these 3 things, what could be the most important factor? Ian, you want to start?

Sorry, I couldn't -- sorry, you might have faded out a little bit. Sorry, could you just repeat the question? Apologies.

Sure. So with micro-dystrophin gene therapy or any of these exon skipping approaches, we are talking about de novo expression of the protein, we are talking about micro-dystrophin positive fibers, and then we're talking about vector genomes for nucleus, right? So of the 3, of these 3, which is the most important predictor of functional benefit?

Look, again, we're going to have to see what data emerges and, obviously, it's going to be long-term data that informs us. You see a very good correlation between both expression from being properly localized in the membrane and the intensity of that -- of those fibers and then also as it relates to the actual protein that you're seeing via -- whether it be by mass -- that you're measuring by mass spec or measuring by western blot. And so those 2 seem to be the best correlate of function, but we're going to have to continue to see over time since, obviously, there is some variability, even between blocks, there's some variability between expressions. So you're not going to see, like I said, perfect correlation. But I think over time, the more dystrophin that you have is likely going to lead to a more durable outcome. If you look at the preclinical models that we've seen, and I'm sure both Carl and Michael have very similar data, there is a threshold that you get to where you're seeing improvement. So from our perspective, we've seen -- if you get expression between 0% and 10%, you're seeing some decrease in progression. If you see between 10% and 20% expression, you're seeing stabilization in the MDX mouse model. If you're seeing greater than 20%, then you're seeing improvement. Now obviously, if you're at 70%, and then over time, you're seeing a decrease in expression, which we haven't seen yet, but assuming that, that happens over some period of time, you would suspect that the higher expression levels on the mean would lead to a more durable functional outcome. So -- but again, we're going to have to see how that plays out over time as we continue to follow these patients for many, many years to come.

So Carl, your thoughts, especially given that you saw some pretty interesting FVC data emerge from -- in your recent update?

Yes. I think you asked about the 3 measures. And really, what you want to see is widespread distribution, make sure that your virus does the right thing and goes where it needs to get to. So showing a large widespread distribution of micro-dystrophin positive fibers, I think is important because it can then potentially stabilize the overall muscle. The expression level, 10%, 20%, that's -- there's a lot of factors that play into how you actually measure the normal. And so that's probably less important as long as you establish a quantified level of protein expression. There are some other things as well. You also want to look and make sure that, there's indicators of biological function or biochemical function of the micro-dystrophin itself. So finding the dystroglycan complex members, things like beta cyclic glycan or beta dystroglycan, really sort of can also highlight the functionality of the protein itself. And even localization towards the membrane sort of indicates that it's moving -- it's being translated and moved to the right place. So I think there's a whole sort of broad picture that will help sort of establish that micro-dystrophin is acting as a functional surrogate of the full length. In terms of -- you asked about FVC, we are very sort of happy with those data, and that we're actually seeing, even in the older patients, pretty much either a stabilization or even a little bit of an improvement in respiratory function. So we're happy. But again, tiny, tiny numbers, and we have to really sort of wait, as Ian said, wait and sort of dose a lot more patients and look for a little bit longer to see the translation of this effect into the functional benefit.

And Michael, in terms of the sort of using LCMS versus western blot, I know Pfizer is sort of being the pioneer in the LCMS aspects of it. Your thoughts on the choice of that characterization technique and where the agency might land up or sort of feedback on that technique itself, given that the western blot that Sarepta and everybody else is working with was initially kind of developed with the FDA back in 2017, if I'm not mistaken? Or actually earlier than that.

No. And it's been used for the accelerated approvals of the exon skippers. So it certainly has some validity. From a biochemical perspective, though, the LCMS method does seem to be less variable. There are fewer assumptions that you have to make in equating it to a normal dystrophin, and it's a direct molar measurement of concentration of the peptide derived from dystrophin in the tissue. So from a pure quantitative perspective, it's a step forward. How much difference it will make in the overall program? I don't know. It's certainly the direction of travel that regulators are in and academics are interested in moving, just to iteratively improve the methodology. But I don't -- I have no -- and we've obviously discussed it with multiple regulators, and they are happy that that's our only measure of dystrophin in our pivotal trials. But that's not to say that they will no longer accept data based on the older western blot-based method.

Got it. So, maybe Arthur, you guys are, I think, the first program for your antibody targeted PMO construct is going to go after exon 44, if I'm not mistaken. Any reason why you decided to go with exon 44, given that it's sort of slightly slower declining than the 51, 53 or 45s? Doesn't that sort of elongate the progression? Or the precedence is set to be approved based on expression levels?

So I think, first of all, 44 is an underserved population. And for me to go to tell a 44 parent that we're going to go to 51 because it's more rapidly progressing is totally unsatisfactory. And so we're really looking towards the patient and patients that have a significant need here. And we understand that this is still an unrelenting progressive disorder. And so getting therapies to these boys who may actually even benefit greater from it because they are less advanced in the disease when we can begin treating them is actually an advantage. I might remind you that we also have programs in 45 and 51, but 44 is underserved at this point. And so I think going after 44 was a -- it was a rational decision on our part both because it's underserved and because it potentially has potential benefit for the program in general. So this is a matter of getting to these patients as quickly as possible while we can save as much function as possible. And I think that's really what you heard from all the panelists all along. With respect to some of the other questions you're asking about gene therapy, I probably want to remind you that using the skipping approaches, you get not a mini- or micro-dystrophin, you're getting a much longer construct with much more of the functionality built in. So I think in that regard, there is a -- there are potentially significant advantages to a therapeutic modality like exon skipping, if we can take these 2 very safe therapeutic modalities. So that is monoclonal antibodies and PMOs, link them together and create something, which is even -- which has that same kind of safety profile. So that's really I think the goal here, and the goal is -- in 44 is to treat this underserved population.

So as we sort of wrap up, I've got like 5 minutes left. I just wanted to get your thoughts of the micro-dystrophin approaches right now or the mini-dystrophin approach. There seems to be a difference, obviously, in treating the younger boys versus the older boys in terms of a steroid regimen, right? Because the younger boys probably are just getting on the steroid regimen at 4, whereas the 6-, 7-year olds have been on steroids for a while. Do you think the early -- or the duration of steroid treatment impacts the tolerization of the gene cassette, which might be detrimental? Or you don't -- do you think that's not going to be a -- it's not an issue at all? Yes. Maybe, Michael, do you want to -- yes. Maybe, Michael, do you want to start?

Sure. Well, so first of all, I think it's important to say we -- and I'm not sure anybody else has seen any clinically relevant evidence of anti-transgene product immunogenicity. So we haven't seen a lot of responses to the novel dystrophin that's been created. And so your question on the age and duration of steroid treatment, the effect on that isn't quite there. In terms of the AAV immune response and safety related, well, there could be some adaptation to chronic steroid dosing impacting the relative anti-inflammatory immunomodulatory effects of steroids in a previously steroid-naive relative to the chronically treated patients. But I don't think we really understand that in enough detail to know that it's important. But the transgene tolerance we're not seeing as an issue in any of our clinical trials.

Yes. We've seen absolutely no effect to steroids in either population on any of the outcomes or immunogenicity either.

Got it. Great. The other question that keeps coming up in terms of getting these micro-dystrophin and mini-dystrophin constructs over the clinical hurdle is the rate of disease progression. Do you think the rate of disease progression of skeletal muscles is very different than the cardiac and the diaphragm? Or could those be sort of used as alternative endpoints if NSA is variable and hard to interpret? Maybe, Michael, I'll start with you again.

Well, there almost certainly are differences in rate even within skeletal muscle in different parts of the body. So I'm not sure we've adequately characterized all of those differences. In terms of the cardiac and respiratory endpoints, of course, we're all measuring force vital capacity, the FVC. But in younger boys, that usually is barely declining, and it's a more useful measure in older boys. But it's not necessarily less variable than the North Star. And in terms of cardiac function, where we're tending to focus on ejection fraction, really, I think all of these studies are requiring that, that's normal at baseline. So -- and it's too early, given the rate of decline in ejection fraction in Duchenne to really expect to see a difference. So maybe when we've treated hundreds of patients, we'll be able to look back and say, oh, well, actually there's a bigger impact on the rate of cardiac decline or on respiratory decline. But we just don't have that data at the moment, and it's hard to speculate as well. I'll ask Carl, the muscle biologist, just to carry on there.

Yes. I mean nothing much to add. In the younger boys, looking at the 4 to 7, there's really no noticeable cardiac decline as yet. And it's very age-dependent as you transition out, as Michael was saying, with respiratory and cardiac function, there is a different progression. The scalable decline happens early and starts at about 6.5, 7 years of age is where they sort of plateau out in overall motor function. And so anytime after that, you're looking at potentially trying to maintain the functions there and minimize any loss or a disease progression. So we don't know yet. There's no major, major difference in the variability and the endpoints that are measured. There's heterogeneity within Duchenne, and we're all dealing with that. It's a bit of a graveyard of drug development actually in Duchenne, which is unfortunate for the patient. We haven't sort of figured out how to avoid that heterogeneity. And so I think we all have to be better at identifying ways to find homogeneous population prespecified, so specific stratification as we go forward.

Yes. One thing just quickly to add, what we've seen is you're getting -- one of the reasons why we picked the MHCK7 promoter was because of the good expression that you see in the cardiac muscle. And then to the other panelists good point, the 4- to 7-old population is going to be challenging, but one of the goals is for all of us to expand the patient population. And we get great expression in the diaphragm also. So when we start going to an older patient population, some of these endpoints may become more relevant. So look at, obviously, the patients are not ambling, you could be looking at upper limb performance, you could be looking at ejection fraction, you could be looking at FVC, like Michael said. So these -- as we expand to a more broad patient population, these endpoints likely will become more important.

Anybody has anything else to add here? We are right at 9:50.

Thank you very much for the opportunity to join this distinguished panel.

I appreciate everybody's time. Thank you so much, Ian, for filling in for Louise and Dr. Binks and Carl, thank you so much for your time today. Really appreciate it.

Many thanks, Debjit.

Thank you.

Have a wonderful day.

Bye.