Ovid Therapeutics Inc. (OVID) Earnings Call Transcript & Summary

Good morning, ladies and gentlemen. Thank you for standing by, and welcome to the Ovid Therapeutics Conference Call. [Operator Instructions] It is now my pleasure to turn the call over to Jeff Rona, Consulting CFO of Ovid Therapeutics. Please go ahead, sir.

Thank you, operator, and thank you, everyone, for dialing in today. Earlier this morning, Ovid and Takeda announced positive top line results from the Phase II ELEKTRA trial of soticlestat in Dravet syndrome and Lennox-Gastaut syndrome. The press release can be found on Ovid's website at www.ovidrx.com. Please note that we will have a webcast slide to accompany this call. For those dialing in, it may be helpful to go to the Investors page of the Ovid website and access the webcast player now if you haven't done so already. A PDF of the webcast slides will also be made available. Joining me on the call this morning are Dr. Jeremy Levin, Chairman and Chief Executive Officer of Ovid; and Dr. Amit Rakhit, President and Chief Medical Officer. Jeremy and Amit will walk you through the data, and then we will go ahead and open the line for questions. Jason Tardio, Chief Commercial Officer, will also join us for the Q&A. As a reminder, during today's call, we will be making certain forward-looking statements. Various remarks that we make during this call about the company's future expectations, plans and prospects constitute forward-looking statements for purposes of the safe harbor provisions under the Private Securities Litigation Reform Act of 1995. Forward-looking statements contained in this call include, but are not limited to, statements regarding the potential benefits of clinical and regulatory development and commercialization of soticlestat. The potential value and benefits of the collaboration with Takeda, the anticipated reporting schedule of clinical data, the likelihood that data will support future development and the association of data with treatment outcomes. Actual results may differ materially from those indicated by these forward-looking statements as a result of various important factors, including the preliminary nature of the data that we will be discussing on today's call. Please refer to Ovid's filings with the SEC, which are available from the SEC or on Ovid's website for information concerning the risk factors that could affect the company. In addition, all information we provide on this conference call represents our views only as of today and should not be relied upon as representing our views as of any other date. While we may elect to update these forward-looking statements at some point in the future, we undertake no obligation to update or revise any forward-looking statements we may make on this call on account of new information, future events or otherwise. I will now turn this call over to Jeremy.

Thanks, Jeff. Good morning. And good morning to everyone, and thank you for joining us today. We're delighted to share more detail on the positive top line results from the Phase II ELEKTRA trial. As we do so, I also want to remind you that these results come at a time when a lot of what Ovid has been working towards since its founding in 2014 is coming to fruition. Indeed, this is a transformational year for Ovid. We've already delivered positive results in Fragile X syndrome and other clinical programs. Additionally, as we look to the long term, we've deepened our scientific base and are building a leading pipeline in neurodevelopmental disorders and rare epilepsy. Today, our pipeline is advanced, focused and continues to expand. We are building for the future. We have an unwavering commitment to developing medicines that transform the lives of people with rare neurological diseases. And not only are we making substantial program -- progress, it's our intent to lead this field. But today, I want to focus on another announcement of positive top line results this time from the Phase II ELEKTRA trial of soticlestat in Dravet syndrome and Lennox-Gastaut syndrome and the associated open-label extension study called the ENDYMION study. It's an important step in our deliberate strategy to grow Ovid with world-leading medicines. Soticlestat, also known as OV935 and TAK-935, is an orally administered small molecule inhibitor of cholesterol 24-hydroxylase, an enzyme found only in the brain. Importantly and impressively, it was discovered in the laboratory of Takeda Pharmaceutical and is the subject of our global development and commercialization collaboration. We entered into this One Team alliance in early 2017. Unlike many other partnerships, this is a true 50:50 alliance, not only with people but we share developmental costs and profit share split equally between our 2 companies. I want to take note. Takeda is a tremendous partner, and we are very happy to be working closely with them. Soticlestat has an extensive preclinical, translational and Phase I data package, including in multiple models of epilepsy. We, along with Takeda, are running a broad comprehensive clinical program with this molecule in multiple types of rare epilepsy. Today, we'll discuss the top line data from the Phase II ELEKTRA trial with soticlestat in 2 different disorders, Dravet, with a clear etiology and more clear definition of clinical spectrum, and Lennox-Gastaut, or LGS, with multiple different etiologies in a broad range of severity. ELEKTRA was a comprehensive, randomized placebo-controlled trial and enrolled 141 patients. As I mentioned, we also have an open-label extension study that is ongoing with soticlestat called ENDYMION. All patients who've completed the ELEKTRA trial have elected to roll over into the open-label ENDYMION study. We will present data today from the ELEKTRA patients who rolled over into ENDYMION. Soticlestat is a very exciting molecule with great potential. And by working together with Takeda, we are maximizing the opportunity to define optimal development and commercialization potential of this molecule. Now let me summarize today's results from ELEKTRA. This is an important study. As I said, it represents the first randomized, double-blind, placebo-controlled trial with soticlestat. This trial conducted with Takeda was superbly run and so we completed enrollment well ahead of our initial expectations. We and Takeda are very pleased with the results. The trial met its primary endpoint with high statistical significance in reducing seizure frequency in the combined Dravet and Lennox-Gastaut syndrome study population. Multiple secondary efficacy endpoints were also achieved with statistical significance. There was a very clear, statistically significant reduction in seizures in Dravet syndrome patients treated with soticlestat. Based on these strong results, we indicated a plan to initiate a Phase III registrational program of soticlestat in Dravet syndrome. We also saw a numerical trend for seizure reduction in Lennox-Gastaut patients, although the trial did not meet statistical significance for these patients. Additional analysis are now being conducted to better understand the potential next steps for the development of soticlestat in this highly diverse LGS patient population. Amit will speak to this more in a moment. Soticlestat was also well tolerated and demonstrated a safety profile consistent with the findings of previous studies. No new safety signals were identified. This is important for a therapy that may be used as an adjunctive therapy for difficult-to-treat seizures. In fact, based on the data to date, we believe that soticlestat safety profile is a clear differentiator versus other therapies. Soticlestat represents a new opportunity to tackle seizure disorders. It's a highly selective, potent, first-in-class inhibitor of the enzyme cholesterol 24-hydroxylase or soticlestat is, to our knowledge, is the only molecule with this mechanism of action in clinical development. Cholesterol 24-hydroxylase is predominantly expressed in the brain, where it converts cholesterol into 24-hydroxycholesterol or 24HC, which is believed to adjust the homeostatic balance of brain cholesterol and, in addition, modulates the NMDA receptor and glial cell. Literature indicates 24-hydroxycholesterol amplifies the excitatory glutamatergic pathway through modulation of the NMDA channel, suggesting potential in diseases where excessive glutamate may play a role, such as epilepsy. Inhibition of cholesterol 24-hydroxylase by soticlestat reduces the neuronal levels of 24-hydroxycholesterol and may improve excitatory inhibitor balance of NMDA channel activity. In other words, modulation of the NMDA channel may interrupt the cycle that leads to seizures. We previously presented preclinical data showing that soticlestat over time reduces stimulus-induced extracellular glutamate levels, consistent with an additional activity on glutamate release and/or uptake possibly through glial modulation. Reducing 24-hydroxycholesterol level may mediate reduced glutamatergic signaling, modulation of glial cell, for example, astrocytes and microglia, and thereby their function and activation and thereby reduced inflammation. Because of this, we believe soticlestat has a novel mechanism of action, which, unlike any other antiepileptic drugs, or AEDs, which are direct seizure suppressants. Soticlestat may potentially provide for both antiseizure action as well as anti-epileptogenic effects. Due to this mechanism of action, the target engagement of soticlestat with cholesterol 24-hydroxylase can be measured with a biomarker. Measuring plasma 24HC levels is a means to understand soticlestat's dosing and target engagement. Accordingly, 24HC is an area of great interest with the intent -- with our intent to explore its developmental potential as a biomarker for the program. Indeed, data from the ELEKTRA trial show that soticlestat reduced mean plasma 24HC level in children from baseline by about 74% to 80% consistent with data previously reported in adults. Before I turn the call over to Amit, I want to spend a few minutes discussing Dravet syndrome and Lennox-Gastaut syndrome, or LGS. Both of these epilepsies disorders typically become apparent during infancy or early childhood and are highly refractory to many antiseizure medications. There are few approved treatments for both. We chose to combine them in the trial to ensure that we could gain maximum efficiency in seeking information on the impact of soticlestat on these disorders and, at the same time, ensuring that those running the trial would be highly trained while minimizing placebo effects so often seen in test trials running different seizure populations. While Dravet syndrome and LGS are both types of developmental and epileptic encephalopathies, these are indeed very different disorders. Dravet is most commonly caused by a genetic mutation in the SCN1A gene. The etiology of LGS, however, remains much more heterogeneous. It can be caused by any number of reasons, including severe -- several gene mutations, perinatal insults, congenital infections, brain tumor or malformations and genetic disorders such as tuberous sclerosis and West syndrome. The cause of LGS are much more unclear. And as a result, this patient population is much more variable. In fact, in some cases, for LGS, no cause at all can be identified. Dravet syndrome and LGS differ by seizure type. Dravet is characterized by convulsive seizures. LGS is characterized by atonic or drop seizures. Children with Dravet syndrome experience developmental disabilities if seizures increase. Other common symptoms of Dravet syndrome include changes in appetite, difficulty in balancing and a crouched gait while walking. Children with LGS syndrome may develop cognitive dysfunction, delays in reaching developmental milestones and behavioral problems. It's estimated that between 45,000 and 52,000 patients are affected by these conditions in the United States. Novel treatment options are greatly needed, and that's why we're so encouraged by the results of the Phase II ELEKTRA trial. And now with great pleasure and excitement, I'm going to turn the call over to Amit, who will walk you through the data.

Thanks, Jeremy, and good morning, everyone. As Jeremy mentioned, we're very excited about these data, and I look forward to walking you through it right now. So let me start with the study design as shown on this slide, #10, for those of you following along. ELEKTRA was a global, multicenter, randomized, double-blind, placebo-controlled trial designed to evaluate treatment with soticlestat in pediatric patients aged 2 to 17 years with highly refractory epileptic seizures associated with Dravet syndrome or Lennox-Gastaut syndrome. We enrolled 141 subjects into the trial, and this trial conducted across 43 centers in the U.S., Canada, Europe, Israel, Australia and China. The study consisted of a 4- to 6-week screening period to establish baseline seizure frequency, followed by a 20-week double-blind treatment period, which included an 8-week dose optimization period followed by a 12-week maintenance period. During the dose optimization period, patients were titrated from 100 milligrams to 200 milligrams to 300 milligrams twice daily doses of soticlestat and for children weighing less than 60 kilograms weight-based dosing was utilized. To be included in ELEKTRA, patients were required to have continued refractory seizure activity after a trial of at least 2 antiepileptic drugs, which are allowed to be on -- and they were allowed to be on 1 to 4 concomitant AEDs at the time of the study. Further, patients enrolled in the study were required to have a minimum number of seizures, at least 3 convulsive seizures for the Dravet population or 4 drop seizures in the LGS population during the minimum 4-week prospective baseline period. At the end of the 20-week treatment period, patients were offered the opportunity to enroll into the ENDYMION open-label extension study to continue receiving open-label soticlestat. As Jeremy mentioned, all eligible patients who completed the study have chosen to enroll into the ENDYMION trial, and I'll review these data on subsequent slides. Next slide, 11. As I mentioned, 141 subjects were enrolled into ELEKTRA. 126 patients completed the study, a modified intent-to-treat analysis, or a mITT analysis, consists of an analysis of 139 patients to evaluate the efficacy endpoints. And it includes any patient who enrolled in the study, but they also have received at least 1 dose of study drug and had 1 evaluable post baseline data point. In terms of the patient population split between the 2 syndromes, the mITT population includes 51 patients with Dravet syndrome and 88 patients with LGS, to note that these are not evenly balanced. Baseline characteristics were well balanced across the study groups. The majority of patients were concomitantly treated with at least 3 anti-epileptic drugs. The most common anti-seizure medicine is taken by the patients were valproate, clobazam, levetiracetam and topiramate. As you can see in the table here, the -- in the LGS cohort, a greater number of patients who were on 3- or 4-anti-seizure medicines which we believe is representative of a more refractory and difficult-to-treat patient population for LGS. Looking at baseline seizure frequency on the bottom right table, you see that the seizure ranges from 2.5 to 125 seizures in the Dravet population and range from anywhere from 4, all the way up to more than 5,000 for the LGS population. Baseline seizure frequency is based on the primary seizure types evaluated in the study. So convulsive seizures for Dravet and drop seizures for LGS. Note that the baseline severity for Dravet syndrome is fairly in a tight range and consistent with the known range of seizure activity in Dravet. And of course, the majority of reported seizures in Dravet syndrome are typically of the generalized convulsive type. Note that the LGS population has quite a broad range as low as 4 seizures per 28-day period to as high as 5,000 plus. This is a known feature of LGS with several subjects with LGS having these very high baseline seizure frequencies, consistent with the heterogeneous condition. And LGS, as you may, is often a diagnosis of exclusion and a catchall diagnosis for anyone with developmental delay and refractory epilepsy with often time unknown causes. So turning to what the efficacy endpoints were the study -- for the study. The primary endpoint of the study was median percent change from baseline in seizure frequency during the 12-week maintenance period of the study. So that's the latter half of the study, the 12 weeks of maintenance. This is based on the combined cohort of convulsive seizures for Dravet syndrome patients and drop seizures for LGS patients. Secondary efficacy endpoints included the percent change from baseline in seizure frequency during the full 20-week treatment period for the combined cohorts. Also, we looked at the percent change from baseline in seizure frequency for each respective cohorts of Dravet syndrome and LGS separately. We also looked at proportion of responders during the maintenance period with who had a greater than 50% reduction in seizure frequency from baseline in each of those Dravet and LGS cohorts. You looked at the investigator caregiver reported clinical global impression of change. And as Jeremy mentioned, we reviewed plasma levels of 24HC, which we have been studying in all of our OV935 programs as a potential plasma biomarker of target engagement. So let's look at the efficacy endpoint, the primary efficacy endpoint in the full mITT population. This is for the combined Dravet syndrome and Lennox-Gastaut cohorts. The study achieved its primary endpoint, demonstrating a 27.8% median reduction from baseline in convulsive seizures and drop seizure frequency compared to a 3.1% median increase in patients on placebo during the 12-week maintenance period. So this is a placebo-adjusted seizure reduction of 30.5%. And as Jeremy mentioned, this result is statistically significant with a p-value of 0.0007. The primary endpoint result is based on the efficacy analysis set of 120 patients with seizure data in the maintenance period. These 120 patients are part of the mITT patient population. In addition, Dravet syndrome and LGS patients treated with soticlestat demonstrated a 29.8% median reduction in convulsive seizure and drop seizure frequency compared to no change in patients taking placebo during the full 20-week treatment period, which, as I mentioned before, includes the 8-week dose optimization period plus the 12 weeks of maintenance. This is a median placebo-adjusted seizure reduction of 25.1% and is also statistically significant with a p-value of 0.0024. Just to note, that the placebo-adjusted changes are not purely additive. And that's mentioned at the footnote below the statistical methodology mentioned to you on the bottom of the slide, which is the Hodges-Lehmann Estimation is the appropriate statistical methodology to address the median of the differences. And this is a sound methodology that has been adopted in other studies as well. Now moving to the Dravet syndrome cohort specifically, patients treated with soticlestat demonstrated a 33 -- 36.5% median reduction in convulsive seizure frequency compared to a 10% increase in the placebo arm during the 12-week maintenance period. When you look at the full treatment period, which is important because this is the analysis that we look for regulatory filings. We see that soticlestat demonstrated a 33.8% median reduction in convulsive seizure frequency compared to a 7% increase in patients taking placebo during the full 20-week treatment period. This is the placebo-adjusted reduction of 46% and also statistically significant at a p-value of 0.0007. And I didn't mention that the p-value for the 12-week maintenance period that I just mentioned previously, it was also statistically significant with p equals 0.0002. As Jeremy mentioned, these seizure frequency results in Dravet syndrome are highly encouraging, and we plan to meet with the regulatory authorities later this year to discuss the initiation of a Phase III registrational program in Dravet syndrome. Turning now to the LGS cohort. We saw numerical trends in patients treated with soticlestat demonstrated a 20.6% median reduction in drop seizure frequency compared to a 6% median reduction in patients taking placebo during the full 20-week treatment period of the study as shown on the right side of the slide. For the 12-week maintenance period, soticlestat demonstrated an 18.9% median reduction in drop seizure frequency compared to a 2.2% median reduction in patients taking placebo for a placebo-corrected value of 17.1% reduction. As Jeremy mentioned, additional analyses are being conducted to better understand the potential next steps for the development of soticlestat in LGS and these analyses will also include our findings from the LGS patients who rolled over into the ENDYMION trial, which I'll speak to in a few minutes. Moving to responder rates. In the Dravet syndrome cohort, shown on the left side of this slide, 30.8% of patients treated with soticlestat demonstrated at least a 50% reduction in convulsive seizure frequency from baseline compared to 0 patients on placebo. Further, several soticlestat treated patients achieved a more substantial reduction in seizure frequency with at least 19.2% of the patients experiencing seizure reduction greater than or equal to 75%; and 1 soticlestat-treated patient experienced total seizure freedom during the 12-week maintenance period, which is a highly clinically meaningful fact in this difficult-to-treat and refractory patient population. On the right side, in the LGS cohort, 16.3% of patients treated with soticlestat demonstrated at least a 50% reduction in drop seizure frequency from baseline compared to 13.3% of patients on placebo. And 11.6% of patients on soticlestat achieved at least a 75% reduction in seizures compared to 0% on placebo. Turning to safety. The key goal of this area, of course, was also to evaluate the safety and tolerability of soticlestat. The safety analysis is based on the full 141 patients enrolled into the study. Overall, soticlestat was well tolerated in the study, and the safety findings are consistent with previous studies, with no new safety signals identified. The incidence of treatment-emergent adverse events were similar in both the treatment and placebo groups, with 57 soticlestat patients experiencing at least 1 treatment-emergent adverse event compared to 52 placebo patients. The most frequent treatment-emergent adverse events reported in soticlestat-treated patients with 5% or more difference from placebo were lethargy and constipation. The incidence of serious adverse events were similar in both the soticlestat and placebo groups, with 11 patients on soticlestat experienced at least 1 treatment-emergent serious adverse event compared to 13 on placebo. Overall, we're very pleased with the safety and tolerability profile of soticlestat. And as Jeremy mentioned, we believe this is a clear differentiator and is important for a therapy that may be used adjunctively with other antiseizure medications. Looking -- turning now to the Clinical Global Impression of Change or CGI-C, this data was consistent with seizure frequency reduction. And as you can see on this slide, more soticlestat-treated patients demonstrated improvement compared to placebo across the full mITT population cohort, as shown on the left graph, and also individually in the Dravet syndrome and LGS cohorts as shown on the right. And now finally, turning to the ENDYMION trial. To spend a few minutes here discussing this long-term extension study, it's a prospective, multicenter, open-label extension study of soticlestat in patients who have participated in a previous soticlestat clinical study. The data shown here is based on the ELEKTRA patients in ENDYMION. The primary objective of ENDYMION is to assess the long-term safety and tolerability of soticlestat, in patients with rare epilepsies; and secondarily, to evaluate the effect of soticlestat on seizure frequency over time. As I mentioned, all eligible patients who completed the ELEKTRA trial elected to roll over into the ENDYMION open-label extension study. Overall, data from the ELEKTRA patients who have rolled over into the ENDYMION study are supportive of the results in the core study. The data suggests maintenance of effect over 6 months in those patients originally randomized to soticlestat. So just walk you through this slide. On the left side in the gray box is the overall treatment period in ELEKTRA, as I've mentioned before. So it is the full treatment period reductions as you see on the left. Once you cross into the ENDYMION, the dark blue, which is the treated group, continued to maintain effect over time, and we have split this up into month 1, month 3 and month 6. The numbers start to drop off after month 6 because the patients have not reached that far in terms of time duration of therapy in the ENDYMION trial. And as you see, the soticlestat arm continues its effect as in the ELEKTRA trial. And the placebo group then transitions to open-label study drug, and that's going from the gray to the lighter blue. And you see that there is a decrease or a seizure frequency reduction evident in the first month, which is then maintained over the 6 months of follow-up that we have shown here in the Dravet syndrome cohort. So for Dravet syndrome, there's a positive response in the placebo crossover group and continued efficacy in the soticlestat patients. Now turning to the LGS cohort. We observed a reduction in drop seizure frequency in the placebo crossover group as well and the continued effect in soticlestat patients. So again, this study, again, shows the decrease in the results from the ELEKTRA period into full treatment arm the first 2 columns highlighted in gray, and then as patients move into the ENDYMION treatment period, the dark blue is soticlestat with continued maintenance of effect after 6 months and then the gray, which transitioned to the lighter blue when placebo patients crossed into the open-label study drug, they start having a continued response that you can see here displayed on the graph. So as with the ELEKTRA data, longer term results from both the Dravet and LGS patients who enrolled in ENDYMION continue to demonstrate a safety profile and efficacy profile consistent with previous findings. So in summary, we're very pleased with the results of the ELEKTRA trial, which achieved its primary endpoint with high statistical significance and achieved several secondary endpoints with statistical significance. We have very strong efficacy results in Dravet syndrome with a 46% median difference from placebo in convulsive seizure frequency reduction over the 20-week treatment period of the study versus placebo and the 50% median difference from placebo in convulsive seizure frequency reduction during the 12-week maintenance period. Data from the Lennox-Gastaut syndrome cohort demonstrated numeric reductions in seizure frequency compared to placebo and longer term data from the ELEKTRA patients who have rolled over into the ENDYMION study showed continued maintenance of seizure frequency reduction over 6 months, with a continued well-tolerated safety profile consistent with previous studies. As Jeremy mentioned, soticlestat also reduced the mean plasma 24HC levels from baseline by 74% to 80%, and that's consistent with data previously reported in adults and shows the continued potential for using 24HC as a plasma biomarker for target engagement. In terms of next steps, we look forward to continuing our collaboration with Takeda to initiate a Phase III registrational program for soticlestat in patients with Dravet syndrome, while continuing to analyze the data from patients with LGS in both the ELEKTRA and ENDYMION studies for the next potential steps in development. We do plan to present the full data from the ELEKTRA study at future scientific meetings. And before I turn the call back over to Jeremy, I'd like to thank our partners at Takeda for their hard work and support, our investigators globally who work through the challenges of COVID and ensured continuity of the trial and the LGS and Dravet patient community for their commitment and enthusiasm. It's been a pleasure to work together. I'll now turn this call back over to Jeremy.

Well, thank you so much, Amit. What a great pleasure to hear your story and terrific. And in addition, I want to thank you, Takeda. Jointly, we're extremely pleased by these results. It's really something very special for patients and invest in great opportunity. As Amit said, it's our intent to publish all the data in the scientific forum as soon as we're able to do so. So in closing, 2020 has been a pivotal year for Ovid, and we have more to come. With soticlestat, we expect to report data from our open-label Phase II ARCADE study in patients with CDKL5 deficiency disorder and Dup15q Syndrome later this quarter. Along with further updates from the ENDYMION trial, which we anticipate will include updating data on the patients who enrolled into ENDYMION from our Phase Ib/IIa adult study. And importantly, in the fourth quarter, we expect to report top line data from our pivotal Phase III NEPTUNE trial with OV101 for the treatment of Angelman syndrome. We're also planning, and I hope you'll join us to host an educational webinar on OV101 in October. It's certainly been an incredibly busy year for us. And I want to thank all of the Ovid employees that have been tireless, and particularly notable in the face of the constraints imposed by COVID-19. I also want to thank our Scientific Advisory Board and our Board of Directors for their advice and support. And as we go further into this, we're very enthusiastic about working with Takeda into the next phase of our collaboration. And we are going to drive forward as one team. Lastly, I want to thank all of the patients who participated in the ELEKTRA trial and the other trials that we have ongoing. We look forward to advancing soticlestat and hope it will become a meaningful treatment option for those with rare epilepsy. I'd now like to turn the call back over to the operator and look forward to your questions. Thank you very much. Operator?

[Operator Instructions] First question comes from Brian Abrahams with RBC Capital Markets.

A couple for me. So I guess starting out, I was wondering if you saw any patterns with respect to potential predictors of response and whether there were any correlations between the 24HC biomarker and response?

Brian, just to recall and to refresh everybody on the call, in our Ib/IIa very small sample of patients, we saw a very good correlation between 24-hydroxycholesterol and seizure reduction. And I'll hand over to -- we don't have that information yet for this particular trial that's coming later on, I hope. But in the meantime, I'll just hand it over to Amit to elaborate a little bit more on 24-hydroxycholesterol.

Brian, sure. So the 24HC continues to be something that we are looking at in all our clinical trials. As Jeremy mentioned in the adult study, we looked at 24HC as a marker of target engagement. So with increasing dosing, you see the level of 24HC in the plasma drop. In the adult, multi-ascending dose studies, we saw that you can reach anywhere from 90% to 95% target engagement without having adverse events and those adverse events are ketamine like events, like psychogenic events. So we haven't seen that in our program because we titrate the dose. In this study, we see target engagement of 74% to 80%, and that's by the reductions in 24HC levels from baseline. So we're still continuing to analyze the data in terms of its correlation with seizure frequency or other types of activity. And that's something we're looking at specifically in both populations, Dravet and Lennox-Gastaut. But it does look -- continue to be a promising potential biomarker for target engagement, which hopefully down the road will be able to fine-tune kind of dosing in terms of what level of target engagement is appropriate for individual patients or individual kind of disease populations. So that data is still being analyzed.

Got it. And then a question on your future plans in LGS and what you may be looking for any additional data analyses to guide next steps. It did look to us certainly placebo group that crossed over in ENDYMION seem to respond better to soticlestat than the initial randomized group to drug. And when you look at CGI-C, that did look more substantial in terms of improvements in the randomized portion versus the seizure reductions. And I know you did point out some of the heterogeneity in baseline seizure counts. So I guess I'm just curious if your impressions right now in terms of any initial signals in any particular subpopulations with respect to etiology or severity guided by the totality of the data that might shape your next steps and whether or not you think you may have enough to potentially move into a Phase III based on some of those additional signals.

I'll leave Amit, if you wouldn't mind me discuss this, and we're actually pretty interested and excited by the opportunity in LGS. Lots of work to be done there. But let me ask Amit, please, to walk you through that.

Sure. Thanks, Brian. So as you know, the placebo-adjusted reduction for LGS across multiple trials, I mean, this is a difficult-to-treat patient population. And when you adjust for placebo, which we had a very low placebo rate compared to other studies isn't very comparable when you're looking around at 14% to 18% reduction versus anywhere from 18% to 25% let's say, in other studies. So a couple of things driving that. So there is a heterogeneity of illness, right? So the baseline seizure frequency, as you saw the range in LGS is huge. It's anywhere from 4 to 5,000 plus. That's one thing that we're looking at is that there is a -- who respond the most? Is it higher number of seizures, potentially lower number of seizures, is there a seizure threshold? Also different in our studies, we allow up to 1 to 4 antiepileptic drugs concomitantly and most studies allow up to 3. So are we looking at a more severe population? If you look at 24HC levels and target engagement, like I mentioned, we reached 75% to 80% in total. Can you potentially push the dose? And is the dose different in different diseases like LGS? LGS is pretty heterogeneous. We saw in our adult study that if you had structural lesions like hematomas or tumors of the brain or corpus callosum defects, you potentially didn't respond as well as if you had intrinsic disease. So LGS is a combination of factors that may have comorbid condition. So things like neonatal ischemia or structural defects. And that's something that we're looking into. But we do think that there are possibilities to move forward with LGS. We just want to make sure that we look -- have a good handle on all the data, and then we'll be able to describe a little bit more our plans for advancing development in LGS. These are some of the things we're looking at.

Amit, one last quick one for me. In Dravet, what might a Phase III look like? What might timelines look like there? And would you indicate a plan to discuss any potential for expedited approval in that indication, given today's data?

Amit, would you kindly take that?

Yes. So we're in discussions right now with the Takeda team has been really -- the teams are really aligned and moving this forward as quickly as possible, we anticipate a regulatory interaction before the end of the year, and we'll have more details after that. I think there's a couple of possibilities. But we're very excited about the data, and it's strong. It's highly statistic significant and could make a huge difference for patients to have a potential new treatment option. So more to come as we get those kind of finalized details done.

Our next question comes from Charles Duncan with Cantor Fitzgerald.

And congratulations on these pretty interesting results. I had a couple of questions. Just kind of a follow-up to the last question regarding next steps in Dravet syndrome. I know you haven't really or can't really talk about details here, but when we look at the kind of variability in the results, both in this one and as well ENDYMION for longer term results, what do you think about, call it, the sizing and timing of the next steps in Dravet syndrome? Can you give us any kind of guidepost as to what the next study could look like?

Sorry, Charles, you said variability. I'm sorry, I don't understand. The Dravet syndrome looks very clean to us and very straightforward, very positive results. I'm not sure I understand quite the word variability.

Yes. Jeremy, if you take a look at the air bars in Slide 15 in the responder rate, obviously, there's going to be a range of responders. But I guess if you don't agree that there's a range of response, I guess, I'm wondering what you think are some of the variables. But more importantly, what we should anticipate in terms of the size of the next Phase III in Dravet?

Amit, do you want to tackle that, please?

Sure. Sure. So I think the -- so as I had mentioned, the details of the Phase III program are still being worked out. But a couple of considerations just to have in the back of your mind. So we do see a very strong effect size here. So that always drives your sample size estimation, depending on the effect size you expect to see. So both in the 12-week maintenance period as well as the full 20-week treatment period, which is important from a regulatory perspective, you see a very strong effect and highly significant results. And so that -- we had 51 patients in Dravet syndrome in our study driving that result. So when you look at other Phase III studies that are out there in Dravet, they're anywhere -- they're a little bit, right, so they're about 90 to 120 or so. And so that would be typical for a Phase III, but that also depends in the effect size. And that also depends on what you see for our placebo rate. We had a very low placebo rate and certain patients actually got worse. In historical studies, you see placebo rates in the Dravet population go anywhere from nothing like we have, like, 1% all the way up to 20%. So we want to factor that in as well. So I think if you look at historical example that's similar. But it kind of depends on the overall design of the study, which we're still working out in terms of the full design, but that should give us kind of a whole guidepost in terms of what might be possible.

Okay. That's helpful, Amit. And would you anticipate some of these additional analyses to be available for presentation, say, in December at the American Epilepsy Society Meeting?

So we just have the results. We are -- we definitely want to present it at -- as soon as possible for one of the major scientific meetings and epilepsy is -- AED is one of the ones on our radar. I think we'll have to work and wait from the organizers. But we definitely are planning to present full data at a scientific meeting or multiple scientific meetings coming forth. So I would just say stay tuned on that one, but that is a hope of ours that we would present fairly soon at epilepsy meetings.

Okay. That's helpful. I look forward to those additional analyses. My last question is regarding the announced equity raise. I guess I'm kind of wondering about that. It seems to be odd timing given the data, I would have thought that you'd present the data and then perhaps, if interested going out to do an equity raise. And can you help us understand the timing? Investors were brought over the wall on this data before it was presented publicly.

Yes. A good question. Chaz, we had a lot of inbound interest earlier this year, and it's the best way to accommodate some of the significant investors, who are going to be long-term supporters of the company, and others who already are. The bottom line, it provides for less distraction for management and less risk mitigation -- and risk mitigation of an incredibly volatile market environment. At the same time, it removes any questions of an overhead and sets us up to take the maximum advantage of that and prepare for what we hopefully beat the successes that we see in NEPTUNE. At the same time, these fluctuations that we see in the market are troubling. And we want to make sure that we have protected our shareholders in the best possible way by ensuring a very strong balance sheet.

Makes sense. It has been a wild market. Do you believe this $50 million will fully fund the next Phase III with soticlestat?

Well, we are very confident and comfortable with our current strong balance sheet, Chaz, which is a -- this is a step go forward in a strategic financing approach that we've adopted. And our goal now, when we couple this with what we have now, not only what we financed earlier on in the year, but then the additional partnerships that we -- the money that we got from the partnership in with Angelini, we feel that we have a very clear financial strategy in place, and we're going to accommodate all of the trials we need. So this is terrific. Looking now and into the future, we're in very good shape.

Congrats on the results.

Chaz, we want to appreciate. Thank you for the question. Thank you.

Our next question comes from Ritu Baral with Cowen.

Can you just address the, I guess, the placebo rate that you guys saw in the study? It was much slower than we were anticipating based on historical precedent. Can you talk about what may be driving that? And then I've got a follow-up on how the titration...

Ritu, I'm very sorry. Could you repeat -- could you -- you asked the question, but I didn't hear the key, what would you like us to address?

The low placebo rate? Our expectations and historicals.

Yes. Great question. Amit, would you be kind enough to deal with that?

Sure. Ritu. Yes, I think that is -- no, I think we were expecting a historical placebo rates in epilepsy studies across both Dravet and Lennox-Gastaut, I probably like 10% to 20% is what we are expecting. And so yes, we were also surprised by the low placebo rates. But a couple of things happened in this study, which may be different from other studies. First of all, we work very closely with the epilepsy consortium. So the epilepsy consortium, as you know, is a group of experts in epilepsy across multiple academic sites who actually reviewed our entry baseline patient seizure information to ensure that we were capturing correctly, Dravet patients and Lennox-Gastaut, both convulsive and drop seizures appropriately. And then they also were instrumental in making sure that the training across individual types and folks happened. And we're a good kind of oversight for the program. I think that's probably the power of the epilepsy consortium that gives you this consistency and making sure that there is ability to run the trial tightly. And it's a placebo-controlled study, so it was blinded as well, but we had baked in a 10% to 20% placebo rate. But I think it's mostly because of the epilepsy consortium's results plus I think it was a very well-run study and had good oversight across the centers that we use, which were high academic centers that follow patients with Dravet and Lennox-Gastaut. So I think those 2 contributed to the lower placebo rate that we see in our study. And of course, it's not completely unknown if there are other studies out there in the Dravet space, specifically that have low placebo rate, 1% or so. So that is something that's been seen in other studies. But yes, we're pleasantly surprised and very happy with the results. And thanks to the collaboration that we've had with our external experts.

Got it. And then what percentage of your patients were able to titrate to that high 300-milligram BID dose? And did you see any correlation with titration final dose and adverse events?

So we don't see -- so I don't have the exact data on the number that reached that 300-milligram twice a day. Apparently most -- everybody did reach that dose. We had very low tolerability issues. I think if you look across our overall discontinuation rate due to adverse events, which is where most of the tolerability, it's an adverse event. So you would discontinue. It's very low, less than 5% of folks would discontinue. We had only a few people that discontinued actually from the placebo group and then completed the study. So we don't see a correlation specifically with the way we titrate the dose and adverse events. And that's one of the reasons we titrate because if you remember from the adult studies, the adverse event profile was only when you kind of gave the dose immediately at the 300 twice a day, so 600-milligram total dose is when you started seeing a potential adverse events. So when you titrated the dose, we were able to mitigate that. So we were able to mitigate those adverse events fairly well. The safety profile that we've seen is fairly consistent with the other studies where we've titrated as well. But we're looking into that a little bit further, and that's something that we'll address I think in the future scientific meetings when we look at that effect with titration.

So to be clear for those. Go ahead, Jeremy.

Just to add one other thing. It was important, although probably nuanced comment by Amit earlier on, when he said that the 24-hydroxycholesterol, which is a good measure of target engagement, we're actually lower than one actually could get to before you see the adverse -- any kind of adverse event for this drug. So there's some flexibility also in dosage, which is interesting.

Got it. And so to clarify, when you present the data, we'll know for those patients that didn't titrate all the way up to 300, why they stopped, what was driving it?

Yes, we'll look into that for sure, but we did present an adverse event related withdrawal from the study, and there was 7. It was evenly balanced across. It was 4 in this treatment group, 3 in the placebo group.

Got it. What drove -- what symptomatology drove the withdrawals?

So AE withdrawal, so it's either an infection. So this was unrelated, potentially infections or potential -- that was I think, the biggest one I've seen. I have to look into that. Ritu, I can get back to you about that one.

Sure. Yes that would be great. And then last question. Can you just talk about the time course of benefit over the titration period or in general, just sort of the shape of the curve? The reason I'm asking is just because the -- in the Dravet cohort on Slide 13, understandably, the full treatment period seizure reduction is less than the maintenance period, but it's actually sort of the inverse -- sort of the inverse for Lennox-Gastaut on Slide 15. So I'm just wondering what you guys sort of saw on those curves.

Yes. So I think this -- so Jeremy, I can go ahead and answer this. It's basically you see both, right? You see both an acute effect as well as the long-term effect. And I think it's easiest probably to see it from the median slide that we've shown, which kind of breaks it up into different months and different time course. So we see -- and also from the maintenance versus full treatment. So the maintenance is the first 12 weeks -- first 8 weeks, sorry, and then the 12 weeks after. So you still see an effect in the maintenance period, which is strong in the Dravet, you see a trend in the LGS and then in the ENDYMION, when you see the placebo rate -- placebo groups move into the open-label, you see an effect within the first month. So you see acute effects, but then you also see maintenance of effect and durability over time. So it's a mixture of both.

Got it. Got it. Okay. Great.

Sure. And I'll follow-up with you about the related trials.

Our next question comes from Tim Lugo with William Blair.

This is Lachlan on for Tim. A couple of questions here. First, just in terms of -- as you're thinking about the future trial in Dravet syndrome, I noticed none of the patients in this trial had tried or failed EPIDIOLEX or Fintepla. Obviously, that's going to be changing moving forward. So for sort of Phase III or next trial, would you plan to include or exclude patients that have previously failed either of those new therapies? And then the second question, on the topic of the 24HC levels. You mentioned that you saw a pretty good correlation there. Did you notice any differences between the indicators? So anything different there between the Dravet and Lennox-Gastaut cohorts?

So just as a prelude to Amit answering this in detail. Any approved drug, one is allowed into our trials with the exception of perampanel where we -- as we pointed out in our previous study. This is a rather infrequently used medicine, but we did see increases in seizure rates when you use it in combination. And that is the only drug that we've seen any interaction with. And we've now, if you look at the data presented in front of you, you will see that the cohorts that we had between 2 and up to as many as 6 different additional AEDs onboard. So with regards specifically to the cannabidiol and the Fintepla et cetera, Fintepla, as you know, is not approved prior to this trial initiating. So therefore people couldn't be on it. What kind of the dials were, and Amit can give you the exact details of how that played out?

Okay. So yes, so 2 parts of your question. So in terms of the cannabidiol question, so in the protocol allows up to 1 to 4 AEDs and they have to be approved. So in this study, we did have people on cannabidiol, so that is included. 10% of the total -- 8% of the total population is on cannabidiol kind of split with 4% in the Dravet group, 10% in the LGS group were on branded cannabidiol. So that would be Epidiolex. No, Fintepla was not approved yet. So when during this recruitment period for this program. On the 24HC question, we have the complete total data, total cohort population, which I mentioned, is 74% to 80% reduction in plasma 24HC levels versus placebo as a measure of target engagement. We are still analyzing that data for the differential cohort. So the individual cohorts is something that we need to look at still, and it's being analyzed. But as Jeremy mentioned before, we have seen the ability to go up higher in terms of target engagement without adverse events. So that's something we're looking at, for sure, in terms of individualizing kind of each cohorts kind of target in future studies?

Okay. And on that point, is that something that you could build into a Phase III study if you wanted to look at titrating patients up to higher levels of engagement? Or would that require a separate study?

That's a good question. I think the more things you do in a study, the more complexity it adds. So there is simplicity of a trial and speed versus complexity and adding more time. So I think we're looking at that. We always collect 24HC levels in all our studies thus far. And I think that's something that we'll continue to have a high interest in. But that depends, I think, on what the ultimate study design is just to ensure that it's not so complex and difficult for folks. So I think stay tuned and so we have a better understanding of what our Phase III program looks like.

Our next question comes from Yigal Nochomovitz with Citi.

Jeremy and Amit, this is Samantha on for Yigal. Just sticking with the plasma 24HC levels for a minute. You said the range 74% to 80%. Can you just give a bit more context in that range and how that's calculated for the overall cohort?

Amit, why don't you handle it? But Samantha good to hear your voice. Thanks for joining us. Bottom line is, look, there are no other biomarkers like this out there. So we have consistently attempted to understand exactly the role of this be it on simply looking at target engagement. And as we do that, we'll have lots to publish on this in the future. But we're excited by what we saw in this reduction and I'll ask Amit to see if he can give you some more color. But at this moment, this is the data that we've presented is confirmatory of what we've seen in the past and leaves us to be very encouraged that not only do we have a good overall safety profile and a good efficacy profile, but we may have a good biomarkers here that could be helpful. So Amit?

Yes. Sure. So for the 24HC levels, which we are still continuing to analyze, as I mentioned, but we do have the full cohort and so the magnitude of mean sustained effect on 24HC as a potential biomarker is about 74% to 80% reduction in these children. It's very comparable to what we've seen in the adults, even though we pushed it further higher even in adults. And the way we do it as we, on day 1, before your dose, we'll measure the 24HC levels. And what that baseline is. We'll have a good sense of what the baseline 24HC plasma concentrations are in nanograms per ml. And then we measure again at week -- for example, at day 15, at 1 month, at 3 months, and then at end of study at day 141, both in the placebo arms and the treated arms. And so we have a good sense that within the first 2 weeks or so, you start seeing a good reduction in those plasma 24HC levels if you're on the treatment group. And that's where you get that 74% to 80%. And it seems to be maintained over that same period of time as long as you're on study drug. In the adult studies, we've seen that once you come off, you do will come back up. So that's -- it's reversible, too. So that's another good sign for this type of potential biomarker. So we're analyzing that further. But current indications are, it's very comparable to what we've seen in our adult programs and other programs, and it has an effect within the first 2 weeks, at least, and then it's maintained over time as long as you're continuing on the medication or so the study drug -- the actual drug.

That's very helpful. And then I guess, just a follow-up, what -- are you able to complete these analyses before the start of the Phase III? And would you consider using 24HC as a biomarker in the Phase III Dravet study?

Amit, why don't you take it? Thank you.

Sure. I'll take that. So I think it remains high potential to be used as a biomarker. I think we don't want to delay our potential for effect to bring a potential new therapy forward for Dravet. I think we'll incorporate as much as possible creating a biomarker for use or an approved biomarker for use has its own pathway. I think we will definitely integrate as much as possible, but the focus really is on starting our Phase III program in Dravet and starting a study that we can then have the efficacy Phase II results and the Phase III to support an ultimate registration. But I think we still have high hopes for this as a potential plasma biomarker that's kind of working in parallel along the side. So we'll try to integrate as much as possible, but it may also be somewhat separate.

Our next question comes from Michael Higgins with Ladenburg Thalmann.

Congratulations. Great to see the results this morning, especially impressive in DS. We have a strong response over placebo, which is really flat, as we discussed. So a question then goes to LGS, where you're going to continue to look at the results, talk with the FDA. So just trying to get a better sense for where are you so far at this point. You believe it performed in which LGS patients, just trying to get a sense if you're aiming towards a subpopulation, maybe a patient characteristics we can see in the baseline, there's obvious differences there even in what such thing as weight. Any sense at this point as to what you're teething out as to where you may find "ideal" LGS patient?

Mike, thank you very much for your observation really well, and you put a thing on a very important question. We're very encouraged by the trends we saw in LGS. Having said that, this is such a diverse population with so many different baselines, as you saw there. But nevertheless, we are, as we saw in the other results from it, we were getting very interesting trends across the signals. So with that, Amit, perhaps you will outline. I can -- I won't give -- I won't go into the details here because we're still working through some of the questions, but perhaps you can give you some color on how we're thinking. But bottom line here, and we're encouraged by the LGS.

Yes. Thanks, Michael. So yes, so I think definitely, we're encouraged by the LGS. As I mentioned, if you look at the placebo-adjusted rate, it's still in this very difficult-to-treat population that's pretty heterogeneous. We're still encouraged by the results, and we're looking at this. I think you've hit on some of the key areas that we are focused on in terms of how do you look at the subpopulation, but the appropriate inclusion/exclusion criteria for people with LGS. So areas that we're looking at, for example, are the number of concomitant antiseizure meds. We are out up to 4, other studies up to 3, maybe it's a less severe population potentially. You're looking at the 24HC target engagement in this population. I only have the total data right now. But is there something with the dose, for example, that may be different in LGS versus Dravet? We're looking at co-morbid conditions. If you have structural defects, as I mentioned before, versus more intrinsic disease, does that cause a differentiation in the probability of effect? And the third, I think you mentioned is that the baseline frequency of seizures is such a highly heterogeneous range from as low as 4 to all the way up to 5,000-plus, is there a cutoff threshold that might be useful to look at in terms of the appropriate population to include? So I think those are all pieces that we're looking at and all still very encouraging because we have much of all that data. So once we kind of fine-tune our approach, I think we'll be able to state a little bit more about how we got to the area. But we're still very happy with what we see with LGS in the trends. And I think there's -- we look at something to come forward, hopefully soon once we do the analyses, like, as you mentioned, to all those areas.

Just one follow-up then. Is it possible to consider an even further dose with some of these patients? Maybe those a little heavier, those are a little "sicker", we tend to see a greater response in clinical trials from the sicker patient population. But here, it may be a different challenge. Just your thoughts on pushing the dose a bit higher, but obviously, a little bit longer duration of that titration period.

Yes. I think that if you look at the 600-milligram daily dose, which is the adult dose, it's about -- for a 60 -- 60-kilo or 70-kilo, it's anywhere from 8 to 10 milligrams a kilogram in adjusted kind of pediatric population. But we see that from at least evidence from our target engagement, which is the closest indicator we have of in addition of the enzyme that we can measure with plasma 24HC, I think that's part of the beauty of having a potential biomarker here. So we reached 74% to 80% versus we reached higher with that dosing in the adult population. So is there differences among different diseases and the susceptibility of the enzyme to be inhibited if you have different intrinsic conditions. And so that's something we'll definitely be looking at because not one, potentially, there might be a different dose for Dravet versus LGS. I don't know that. But I think having this potential biomarker to kind of titrate that for individualized target, at least target engagement, is something we'll definitely explore. I know you're right, maybe there's a potentially different dose and you could go higher, for example, in LGS. And we know that when we titrate up, we can -- we mitigate any potential adverse events you have from tolerability. So I think that's something we'll definitely look at as well, along with the other factors that we just talked about.

And then a last follow-up, if I could. How quickly do you get the 24HC levels might that be used to adjust your dosing?

Yes. So I think that's a good question. I think right now, we batch produce those. But I think that's something, as we move down a path that if this still pans out as a potential biomarker, we'll be looking at more kind of ways to assay this in parallel. But right now, we batch kind of collect them and run them. They don't take that longer run once you have them run. So -- but is it -- can you get as immediate? I think that's something to be seen as we kind of invest behind this as a potential biomarker if it holds up.

Thank you. I'm showing no further questions at this time. Thank you for joining the call today. You may now disconnect.