Ultragenyx Pharmaceutical Inc. (RARE) Earnings Call Transcript & Summary

Good afternoon, and welcome to the Ultragenyx DTX401 Phase III Results Conference Call. [Operator Instructions] It is now my pleasure to turn the call to Joshua Higa, Vice President of Investor Relations.

Thank you. We have issued a press release and posted a slide deck to our website describing the topline results from our Phase III study in glycogen storage disease type Ia, both of which you can find on our website at ultragenyx.com. Joining me on this call are Emil Kakkis, Chief Executive Officer and President; Eric Crombez, Chief Medical Officer; and Howard Horn, Chief Financial Officer. I'd like to remind everyone that during today's call, we will be making forward-looking statements. These statements are subject to certain risks and uncertainties, and our actual results may differ materially. Please refer to the risk factors discussed in our latest SEC filings. I'll now turn the call over to Emil.

Thanks, Josh, and good afternoon, everyone. Ultragenyx has had a strong year so far, with successful data showing that the UX111 gene therapy for MPS IIIA can reduce toxic substrate and improve cognitive outcomes at the World Congress in February. We then presented compelling GTX-102 data on Angelman syndrome in AAN in April, showing we have a product that can change the development of these patients that are heading to a Phase III trial. At the same time, both Roche and now Biogen have stopped or opted out of their own ASO programs. Today, we are presenting excellent Phase III results from our first randomized placebo-controlled study of DTX401 gene therapy in glycogen storage disease Ia. I first suggest this indication to Dimension Therapeutics years ago as a member of their SAB, and so it's particularly exciting to see this Phase III come out with persuasive statistical evidence in support of the efficacy of this gene therapy. We learned a lot about GSDIa in these last few years about the urgency, fears and the unmet medical need of patients fighting for their life every day by drinking flurries of cornstarch every few hours to stay alive. So it's particularly satisfying to bring forward a new option for these patients, knowing how much they go through. It's as good as it gets in biotech when you can get a result like this and have the potential of change for a genetic disease. Now I'll turn the call over to our Chief Medical Officer, Eric Crombez, to walk you through the successful results from the DTX401 Phase III gene therapy program.

Thank you, Emil, and good afternoon, everyone. This is an important milestone for our GSDIa program, and I am happy to be able to share in more detail the positive top line results from our Phase III GlucoGene study evaluating DTX401, an investigational gene therapy for the treatment of patients with glycogen storage disease type Ia. Next slide. In this presentation, we will be making some forward-looking statements. And just as a reminder, DTX401 is currently an investigational drug and not yet approved by any regulatory authority. On Slide 4, I will provide some brief background on the disease and our gene therapy. GSDIa is a severe metabolic disease caused by deficiency in glucose-6-phosphatase, which leads to the liver's inability to release glucose into circulation between meals or during times of fasting and leads to the accumulation of glucose stored as glycogen in the liver. The inability to release glucose to the body puts patients at risk for repeated episodes of hypoglycemia, which can be severe and potentially life-threatening. Current treatment involves the use of cornstarch, which is slowly broken down in the GI tract as a way to provide a source of glucose. It is essentially like providing a continuous oral glucose replacement therapy to do what the liver cannot. The treatment has been life-saving, with GSDIa being considered universally fatal before cornstarch was introduced. Though life saving, it is very burdensome, with the need to take cornstarch every 2 to 4 hours including overnight, and the regimen does not adapt well to situations like exercise or illness. The picture shown here is a child with GSDIa and the amount of cornstarch he takes on a daily basis, which includes waking at least one time during the night. Amounts consumed can range between 300 and 400 grams per day, which is essentially a pound of raw cornstarch every day. Children are particularly vulnerable, and a single misdose can be life-threatening, especially in the middle of the night. Because glucose-6-phosphatase expression is highly regulated, DTX401 and AAV8 gene therapy includes the native promoter and enhancer for the gene, which provides the ability to respond to the normal physiological signals that coordinate and control glucose levels. Insulin suppresses the gene after eating and glucose is [ upped ]. Glucagon induces a gene when fasting or exercising and glucose is low. Cortisol induces a gene when stressed and there is a need for more energy. This responsive control of the transgene replicates the same metabolic regulation as endogenous G6Pase. Establishing the normal responsiveness to physiologic signals for glucose control like a normal liver cell is a critically important feature of treatment with DTX401. This enables the patient's liver to start responding to the signals their body makes based on their glucose needs. Next slide. This is an illustration of the Phase III study design, 44 patients were included in the modified intent-to-treat population and were randomized 1:1 to DTX401 or placebo and followed in a blinded manner for 48 weeks. The Phase III included a longer screening period compared to the Phase I/II to allow for optimization of cornstarch before treatment randomization. While the double volume design was important, it meant that site investigators could not have real-time direct access to the glucose information from CGM, which presented challenges to safely managing cornstarch dosing in both the treated and control arms of the study. In order to protect the blind and to not put placebo patients at unnecessary risk, we included an unblinded physician that received the CGM information and then guided the PIs who remain blinded on cornstarch reduction options. This is a far more constrained process than what was used in our open-label Phase II study, where physicians and patients have direct access to real-time feedback on glucose levels directly from CGM. The approach did allow us to maintain the blind and conduct the experiment required to demonstrate the impact of DTX401. On Slide 6, I will go into study results. The primary endpoint, the percent change from baseline to week 48 in daily cornstarch intake for the DTX401 group compared with the placebo group showed a 41% reduction in DTX401 treated patients versus 10% in the placebo group. The relatively high level of reduction in the placebo group was particularly driven by one outlier patient as the median change was only 2%. On the graph on the left, you see the steady reduction in cornstarch use in the DTX401 group during the 24- to 48-week interval, showing that patients are continuing to progress on treatment. This contrasts with the placebo group that showed an initial small decrease, probably due to further cornstarch optimization and then no further cornstarch reduction from 24 to 48 weeks. We attribute the more gradual titration of cornstarch in the treated group compared to earlier and more rapid reduction in the Phase I/II to the blinding in the Phase III and the lack of real time and complete feedback to the investigators. With the continued reduction through week 48, we expect further titration in the open label setting beyond week 48. On the right, we present a responder analysis of patients with at least a 30% reduction in total cornstarch use. We found that 13 out of 19 DTX401 treated patients or 68% reached that level of reduction compared to 3 out of 23 placebo patients. A 50% or greater reduction in cornstarch was achieved in 7 DTX401 treated patients or 38% compared to a single placebo patient. I'll get more into this a bit later on in the presentation, but the data generated in this study showed that a 30% or greater reduction in cornstarch was clinically meaningful to patients based on their scores from the Global Impression Scale. On the next slide, we get into the reduction in total amount and number of doses of cornstarch, both of which are important to patients. Patients treated with DTX401 showed significant reduction in frequency and quantity of day and night time cornstarch compared to placebo. On the left, you see reduction in total doses of cornstarch taken over the course of 24 hours, with DTX401 treated patients dropping a full dose compared to a small adjustment in the placebo group. On the right, you see reduction in the need for overnight cornstarch, which requires a setting of alarms and backup alarms to ensure doses are not missed. Patients treated with DTX401 were able to reduce doses by 0.4, with an equal increase in doses in the placebo group. Total grams of overnight cornstarch use was reduced by 44% at week 48 from baseline in the treated group compared to an increase of 7% in the placebo group. For both of these parameters, we expect continued improvement as the patients adapt to lower cornstarch doses and make less insulin over time, as we saw in the Phase I/II study over the 1 to 3 years following treatment with this gene therapy. With all patients now changing to management in an unblinded manner, we do see the potential for continued improvement with treating physicians and patients having access to glucose values in real time. Next slide. Controlled fasting challenges were conducted, and DTX401 treated patients showed an almost fivefold greater improvement in time to hypoglycemia, with an increase of 12.8 minutes per gram of cornstarch administered compared to 2.9 minutes per gram of cornstarch in the placebo group. What this means is that, with the liver now producing glucose, each gram of cornstarch was able to maintain glucose levels for fivefold longer with DTX401 treatment than cornstarch alone. Next slide. In addition to controlled fasting challenges, glucose control was evaluated by continuous glucose monitoring by CGM devices and by self-monitoring of blood glucose or finger sticks to assess glucose control in the real world setting. By all 3 evaluation methods, treated patients showed statistically significant noninferiority for glucose control after clinically important and statistically significant reductions in cornstarch. In the DTX401 group, we do see some variability in glucose control as we saw before, as these patients can have some reactive hypoglycemia, and this is more challenging to manage in a double-blind study. When you evaluate the more significant hypoglycemia level of less than 54 milligrams per deciliter, we see that the values are now essentially the same and within the range of variation. This is good to see that patients were able to maintain glucose control while significantly reducing cornstarch. On Slide 10, we show the results from the PGIC or Patient Global Impression of Change score. The DTX401 treated patients reported moderate to much improved PGIC scores with a median score of 2 compared with only 1 for the placebo group. These data also established that a 30% or greater reduction in cornstarch was correlated with moderately or much improved score, establishing 30% as a clinically meaningful threshold. This is an even lower threshold than we anticipated, demonstrating the significant burden daily doses of cornstarch have on patients. Standardized exit interviews were also conducted and highlighted the benefits of establishing the normal metabolic pathway to break down glycogen to provide an endogenous source of glucose that is independent from dietary intake. The safety of DTX401 was acceptable and consistent with what we saw in the Phase I/II study. Elevated [ LFTs ] were seen as expected with a liver-directed gene therapy and well-managed with steroids. Two infusion-related reactions were seen in the actively treated group, but 2 were also see in the control group and attributed to anxiety associated with the monitoring involved with dosing. No gene therapy class effects of DGR toxicity or thrombotic microangiopathy were seen. To step back and look across the Phase III results, the study showed the important clinical benefit of DTX401. The study achieved persuasive statistical significance across the primary endpoint in 2 of 3 key secondary endpoints. As time goes forward, we expect to see continued reduction in cornstarch use for these patients. We also expect declines in insulin levels and improved metabolic control, with much less dependence on cornstarch and far less fear of their disease. GSDIa does not have an approved treatment option. The Phase III data show that our gene therapy could become a powerful tool to help physicians and patients manage their disease by directly addressing the underlying cause of disease. Clearly, there is a large unmet need here, and I look forward to partnering with my commercial colleagues who are well versed in successfully launching products in the rare disease space. I'll close with a few next steps. These results put us in good position for a successful pre-BLA meeting with the FDA planned for the second half of this year. The BLA submission is planned for 2025, which will allow us to complete our manufacturing transfer to our facility in Bedford, where manufacturing runs have begun. We look forward to providing you with updates as we progress through these steps and sharing additional data at an upcoming scientific conference. With that, let's move on to your questions. Operator, please provide the Q&A instructions.

[Operator Instructions] Our first question comes from Dae Gon Ha with Stifel.

Congrats on the data. I'm guessing -- well, I guess I'm getting a number of inbounds of investors somewhat perplexed by the data presentation. So Emil or Eric, I guess if you can take a step back and walk us through sort of the rationale behind the changing of the cornstarch reduction methodology used here. I guess what was particularly concerning about the Phase I/II type of approach because ultimately, the magnitude of reduction shown here is meaningfully different from what we saw in the Phase I/II. And then maybe on the baseline characteristics, I didn't see a table outlining the features here. So wondering if you can talk a little bit about any particular characteristics to note that might have contributed to either more or less on the magnitude of cornstarch reduction benefit. And can you also comment on the proportion of R83C patients on both arms?

I'll start with the first question, and then, Eric, maybe you can do the second or third. So first of all, I think cornstarch reduction is excellent that we're seeing. And when we had to run this in a blinded study, and I think Eric explained in the call that we couldn't let the doctors see the CGM data because they would know immediately what was going on. So we had an independent doctor looking at the CGM daily and then guiding them. But that process is very stealthy. So it doesn't allow for more rapid changes. And we know from before that when we instituted CGM, it was definitely more powerful to rapidly change, particularly deal with things like reactive hypoglycemia. So we expect it to be slower, but there really wasn't anything different about how we're running the cornstarch, just had to do more of the fact that we had to have a constrained process. If you don't do that, you're not -- you might break the blind and that would be problematic. So this is an experiment that just shows that the delta is real. But we'd expect to be seeing the same reductions ultimately that we saw before. I think if you look at the data, you can see at week 24 to 48, it's still declining. And the patients afterward continue to decline. So we feel comfortable about the same efficacy is being seen. There's really not a significant change. It's more about the operational part. So we had to do it that way in order to maintain the blind, so that's that main point. And then Eric, maybe you can answer the other questions. I don't know if we know the R83 story.

Yes. So we haven't yet analyzed these by different genotypes. It is difficult to do even with this relatively larger number of patients compared to the Phase I/II. We did look at that with our Phase I/II patients and there's just too much variability to really be able to draw meaningful conclusions. What I do think is important for this patient population, and again, similar to Phase I/II is with GSDIa, you don't see the type of variability or differences in severity that you're seeing a lot of inborn errors of metabolism. Patients really are fairly consistent in their phenotype. I don't think the difference in the rate of titration of cornstarch has anything to do with patient characteristics. I do absolutely think it's due to the blinding. I mean you have to remember that at the time of diagnosis, along with every follow-up visit, you stress to patients and their families the importance of sticking to the strict regimen of cornstarch because of the real fear of a serious hypoglycemic event. So it makes sense that in a blinded fashion, the principal investigators are going to be more cautious. And I think you see that reflected in the slope of that line. But importantly, they were continuing to titrate through week 48. And that's why in the open-label setting, we do expect further titration as we follow up with them over time.

Our next question comes from Tazeen Ahmad with Bank of America.

Emil, just in terms of the tech transfer that you're expecting to have to make before applying, you've given guidance that you would apply in 2025, but this would be the first time that Ultragenyx would be undertaking such an endeavor. So can you just walk us through what needs to be done and give us a sense of how long you think this process could take? And then I have a follow-up.

Yes. So first of all, the plan has been up and operating under GMP control now for a good part of the year. So it's operating and run multiple runs already. The PPQ runs of call lots for GSDIa have already started and are running well in the plant. So we're running all the things, but we need to run them, test release, provide some stability data, et cetera, fill, finish, which will all be done in the plant. So it takes some time to do that in order to have the information in hand to be able to file. And our feeling was, we made the decision is that we'd rather have this in control. We run -- it will be more cost efficient and more reliable than the contract manufacturer we've been using. And -- but we feel very confident. The team's running it well. And we don't think -- there's no real tech transfer done. It's already running. The GP runs are running for GSDIa. So I don't see any big issue with getting through. It just takes time to run runs and test release, et cetera.

Okay. So when you say 2025, do you think that it's more likely to happen earlier in the year? Or could it take until the second half of the year to get all those sorted?

So it's probably in the middle of the year. It's not at the beginning of the year, and it's not second half probably, but in the middle of the year. We have to go talk to the FDA about pre-BLA meeting too, and get some pieces done. But you have to run the runs, test release, put some stability, et cetera. It will be some time in that range. We haven't set it. We haven't defined it precisely yet. And we'll provide more guidance as we move along, but especially after our pre-BLA meeting when we find out exactly what FDA wants to see because sometimes they have requests. But we just wanted to be clear, we weren't going to do it this year because of the CMC changes.

Our next question comes from Anupam Rama with JPMorgan.

Congrats on the data. So following up on Tazeen's question, you guys outlined some of the gating factors in the U.S., the pre-BLA meeting, the CMC work. I'm assuming these are similar gating factors to a potential EU or rest of world strategy, or any other different regulatory considerations for OUS?

It's the same issue to file in Europe. We would generally want to make sure the [ referrers ] have seen what we're doing before we go and file. So there will be some follow-up with Europe. But I would say the U.S. is the primary driver of the timeline and the execution because this is the most important market for the program. So really, U.S. drives all the decision-making, but we have to manage the EU and the PIP requirements, et cetera. But -- so yes, it's pretty similar. But I would say whatever is happening, the timeline is driven by U.S.

Our next question comes from Maury Raycroft with Jefferies.

Congrats on the update. I was wondering, for the 7 patients getting to greater than 50% cornstarch reduction, do these patients seem to be trending similarly in profile and response to patients in your Phase I/II where you expect they could come off of cornstarch completely and liberalize their diets? And then for filing for approval in 2025, do you plan to include a cut from the crossover portion of the study or could that even include the 96-week data in that -- in the pipeline?

Yes. So one of the things that's happening, in some of the patients that do better, it's just that their doctor is being more aggressive is titrating the starch so the patient is more comfortable with it, frankly. What patients don't know, what they -- they can't look at their own glucose. So this creates a challenge for them being comfortable doing what the doctor is telling them to do. So when the patient is more comfortable, the doctor is comfortable, then you see them moving along. We had a number of patients that got -- that did really well, as you would normally see, but people that are more cautious needed a lot of encouragement to take -- make moves when they weren't 100% sure. So yes, we think those patients are behaving more like the others, but I actually think all of them will get there. And I think we're pretty comfortable with that they'll all really get down there. I don't think that group is going to be -- I think all the patients could potentially get off. But if you remember from the Phase I/II, it took to almost in the third year before people came off completely. And the point to be made here is that they have a lot of insulin. They make -- they have excessive insulin production. Until their beta cells start settling down and stop making insulin chronically, you can't really wean them off completely. So you have to push them down step-by-step and let their bodies adapt to this new regime. They've been getting -- most of these patients are like Type 2 diabetics. They've been getting so much starch and have so much glucose driven by -- that their insulin levels are like a Type 2 diabetic. So we have to change 20 years of that regimen here when we adjust. So I do think we're going to get there with these patients the same. I think it's just a little more constraining in the beginning, a little bit like our cohort 2. In terms of the crossover, we'd expect to have the crossover patients. There are patients are already crossing over. They're doing well. And our expectation is that open-label crossover will be because now we'll have real-time data, so it will be a lot easier for people to make the adjustments and we'll do better. So the crossover data might add to the pool of data we have in addition to the open-label extension of this group. So a little bit of time on the CMC probably gives a chance to have a little more extension data, which I think will be important in looking at the overall outcome for these patients.

Next question comes from Joon Lee with Truist Securities.

Congrats on the data. This is Mahdi on for Joon. So basically, to me, it seems that the daily cornstarch intake on the graph shows a bit of flattening. So how do you expect these cornstarch reduction and also the global -- Patient Global Impression change over time in longer term and basically in the real-world setting? And I have a follow-up question.

Yes. Well, in the real-world setting where they can see, what will happen is during the first 12 weeks or so, there will be a more rapid decline and they'll end up at a lower level. The higher their starch is, the higher their glucose and their insulin. So a lot of these patients that you're looking at, we haven't shown you, we'll show later, we put the data out a meeting, [indiscernible] patients have 34% of the time are hyperglycemic right now. They're all on too much starch actually. So they're all hyperglycemic because they're so afraid of hypoglycemia, but they're actually taking too much. So I have very good confidence we can get them to wean off their starch now if they know what they've got and they're not going to have that fear. So we think that the cornstarch reduction will be there. When you look at the scores of those that have 30% or 50% of cornstarch reduction, the PGIC -- the Global Impression Scale scores are better. They're clearly much better if you get down because once you get down and the effect of being like a type 2 diabetic and hyperglycemic and sick all the time starts to fade and they start feeling better and more energetic and so forth. So this will all get better. I think in the real-world setting, look, Phase III studies are contrived experiments. And this is one of the examples where you can't have the right loop, but we're very comfortable with what we're looking at. We know they are still hyperglycemic, in fact, and so there's a lot of room for improvement in moving their starches down. I think we'll get to the same place we were before. Now they can -- the patient and the doctor can know what their starch is like. What's your follow-up?

That's very helpful. The question is, what is the mechanism by which you think this single placebo patient achieved like 50% reduction in cornstarch intake?

Yes. Well, it's a little bit mysterious. I think in placebo-controlled trials, there's always one oddball. You must have been -- some of the patients at baseline were clearly on too much starch to begin with. They tend to be overstarched. And there was a period of 6 to 8 weeks in the beginning where they're supposed to be titrated. But that guy clearly wasn't titrated enough. He was -- just probably had too much starch. But I would say that variation is normal. It's a little bit hard to explain, but we're not concerned about it. Just one odd patient we think was getting too much starch and he didn't get it optimized well enough.

Yes. I think the key difference there is those patients who were followed long term by the effects for PIs versus patients who were referred in for enrollment in the trial, and we did see that some of these patients who were treated out in the community didn't have as optimal care as the patients treated by the true experts.

Yes. That patient was one of the referral patients, right? So he wasn't being handled by one of the expert sites? That's right.

Our next question comes from Whitney Ijem with Canaccord Genuity.

Congrats on the data. In terms of filing timelines, I guess I just wanted to clarify, is CMC and kind of the stability the rate limiting step to filing? Or is there also a specific amount of data you're hoping to capture from the open-label follow-up or the crossover patients in terms of that real-time glucose monitoring?

Yes. So the filing time is really driven off of CMC. That's the primary driver of it. While that's happening, we'll collect whatever data we can collect. The better -- the more data, the better. But we're driving the timeline all on CMC.

Got it. Okay. That's helpful. And then just curious if you've had any early kind of payer discussion, not obviously with this data, but with the Phase I/II data or even kind of [ TPP ] discussions around their views of kind of clinical meaningfulness and kind of value of kind of what you're seeing here?

We haven't done deep dives on this topic yet. I think it's important to do. I think they're going to have to look at the picture, not just cornstarch, but the overall picture of how patients are doing and what it means to them physiologically to get the full breadth of the information about it. But we haven't done extensive discussions. Given that the pricing of a lot of gene therapies have gotten higher and higher, I feel like the price point we have been thinking about, I think, is well within the range. And given the lethality of this disease and the patient demand for something better, I think that I feel comfortable this is a type of therapy that will be urgent, that will get well adopted, unlike other situations where people are really comfortable maybe in their current care and they're having -- questioning about where they want to take the chance. We had no problem enrolling. Instead, we had people complaining they couldn't get in the study, in fact. And because people are anxious to get off this treadmill. They're running around with a gun to their head, hoping it doesn't go off because they forgot to take starch. So you have to imagine what that feels like. If you can take the heat off, in fact, stop taking starch at night, are not highly dependent that you're going to crash suddenly, it's a big difference. So we have to make sure we're making that statement. It's about relief of the fear and pressure of the cornstarch that providing an independent source of glucose to give these patients.

And our next question comes from Yigal Nochomovitz with Citi.

This is [ Ashik Obaigon ] on for Yigal. Just one from us. What will be the key focus of the pre-BLA meeting? Is there anything in particular you're going to get clarity on? Maybe around manufacturing or the aggregation of this clinical data? Or is this more of a box to exercise out of the filing?

Normally, it's just going to be procedural. We agreed on the primary endpoints and other endpoints with the agency. We've hit them with persuasive statistical significance. So the clinical data question should be answered from this randomized controlled study. So it's going to be more about determining how much data they want from the CMC side, how much stability data is enough data and any other particular details they want or analyses done from the clinical data that would be important to include in support of the product. And it might include how much extension data they want. But our expectation is that we will have some extension data and also crossover data from patients, which will -- usually, the crossover data, I would say, is pretty important. In most [ rate ] of these programs I've been in, we've included the crossover, placebo crossover data because there's not that many patients treated. So if you have another chunk, double the number of patients treated, showing the same effect or a better effect even since they're going to crossover in an open-label format, I think it allows us the opportunity to help bolster the case for the program, but it will be mostly about getting the bits and pieces. I don't think there's going to be any other criteria set at the pre-BLA meeting for filing.

Our next question comes from Yaron Werber with TD Cowen.

This is Brendan on for Yaron. Congrats on the data, guys. Maybe just zooming out a little bit here and kind of touching on some of the earlier questions. I guess when you think about the broader GSDIa population, whether you're looking at the Phase I/II patients and the Phase III, how should we think about who gets this drug right off the bat? I mean these are obviously meaningful reductions in overnight cornstarch. I mean not quite down to 0 or curative, but I mean we're just trying to understand like if you've identified certain patients across the board that you think would be better candidates or maybe faster adopters that you could target from the jump, just kind of trying to give us a sense there.

Yes. Well, first of all, I wouldn't look at the Phase III data as the definitive determination of what the efficacy is. We may -- we think of that for a lot of programs, but this is a disease that's going to take a couple of years of evolution. We've shown that already in Phase I/II. It takes a couple of years of time. We didn't want to run a randomized trial for [ Phase II ], it's not necessary. We can track them in open-label extension. So we think the overall benefit will be seen over the couple of year period, not necessarily just the Phase III data. We would pitch that with our extension information, et cetera. So that's first off on how you determine. And when you look at the population of GSDIa, 80% of the patients have severe genotypes, null genotypes. So there's a lot of people that are highly dependent on oral cornstarch and do not have any -- or are essentially brittle. I think any of those patients are the ones that are at most at risk of having a crash event. And those are the ones that are going to be high drivers. So I think the majority of the patients with GSDIa are the type that will want to get treated. The numbers that are very mild are relatively small and infrequent. And so I really would say the vast majority of patients would be addressable type patients for this. And all of them want to get off the treadmill here that they've been running for their whole lives. So I really think you're going to find a lot of patients want to do anything to get off the treadmill. So think about the data broader than just the Phase III and understand the severity of disease and what people like, I think we'll put this and the majority of the market will be addressable, and I think it will be adopted and utilized.

Okay. Maybe really quickly, if I could. Just looking at maybe some of the liver responses. Do you think there is -- has it come up in your conversations with FDA if they would recommend or if the label would potentially require steroid co-treatment with the gene therapy? Or that's -- you haven't talked about it yet?

Well, steroids are part of the protocol. We gave blinded steroids in this trial. So patients got steroids if they have LFT elevations. As a standard part of it, I would expect it to be part of the labeling. So that is -- the way we have done it in this trial, we did not give steroids right at the beginning. We gave it a couple of weeks after because actually giving steroids to GSDIa patients can be dangerous when they don't have any transgene. Without the transgene, the steroids can actually cause them to drop their glucose, which would be dangerous. Once it gets to transgene, though, it is responsive to steroids. So we actually know that the transgene is being turned on because the steroid treatment actually induces hyperglycemia usually in these patients after the gene therapy. It gives us actually a nice way of demonstrating that the drug is working. So yes, steroid will be needed as part of the plan. We have -- did not have a real significant problem. I don't know if there's any comment on safety, Eric? But I thought safety was pretty good. I didn't see any issues there.

Yes. No, absolutely. And I think the use of steroids or some sort of immunosuppression is important to control for that immune response to the vector.

Our next question comes from Gena Wang with Barclays.

So I wanted to ask you, I know you hear the data on both the frequency and quantity, the day and the lifetime point. So which one do you think will be more meaningful to patients? And I know over [indiscernible] your expected [indiscernible].

Gena, we broke up, but I'll answer the question you asked, and then we'll get you back. So you asked, what's more important, the amount or the frequency. I think both are important, frankly, because the amount of cornstarch is how much slurry you have to drink. It also is the amount of glucose you're dumping into the system, the amount of insulin you have to create to manage it. And the frequency is just a convenient thing. The way the titration was being done primarily was through reduction in amount of cornstarch without changing frequency, but there obviously was a change in frequency because we had a statistical reduction in total number of doses per day. If you looked at the nighttime dose, it went down by 40-plus, 44%, I think. So the nighttime dose went down quite a lot, which I think is a sign that patients are -- the therapy is helping them get through the night. I'd expect as we saw later than most of the patients might stop taking nighttime cornstarch, especially those that don't like getting up. And then for a lot of others, they may skip starch during the day and they'd have it a little in the morning if they need it. But might -- we've had some that not take it during the day and start reducing the total number. But you start by reducing the amount and then you start dropping out the numbers as you get going. But each patient wants to -- is going to craft their own story. And we don't think it really matters either way. The total amount of starch given is really the thing and the frequency is a patient-specific preference. I don't know if we lost Gena, but let's go on to the next question then.

And our next question comes from Kristen Kluska with Cantor Fitzgerald.

Let me also add my congratulations on these data. I wanted to ask how you're thinking about the market opportunity. I think in the past, you've cited 6,000 patients worldwide. So can you give us a sense of if that's still how you're viewing the market? And also, because you're in a unique position that you have a global footprint, how you're starting to think about maximizing that opportunity?