Cytokinetics, Incorporated (CYTK) Earnings Call Transcript & Summary

Hello, everyone. I'm Diane Weiser, Senior Vice President of Corporate Affairs. I'm pleased to welcome everyone to our investor event to discuss the primary results from ACACIA-HCM, the pivotal Phase III clinical trial of aficamten and nonobstructive HCM. Today's event is being hosted in a hybrid fashion. I'd like to welcome those here in the room with us in Munich as well as those joining online. I'm pleased to introduce from Cytokinetics, Dr. Fady Malik, Steve Heitner and Robert Blum, and I'm also thrilled to welcome 3 leading experts in HCM for our panel discussions today. Dr. Marty Maron, Dr. Ahmad Masri and Dr. Christina Paitazoglou. Today's agenda will begin with -- is it on the slide? Today's -- we are going to pause for a minute. Turning to today's agenda. Fady Malik will provide some brief opening remarks. Next, Dr. Masri and Maron will provide an encore presentation of the ACACIA-HCM results that were presented earlier today in the online session at the Congress. Then Steve will facilitate a panel discussion and Q&A session. And finally, Robert Blum will close us out, providing some remarks. Those online, today's slides are available for download in the webcast. You can submit questions to the panel at any point during the event using the Ask a Question tab on the upper right-hand side of the webcast. For those in person, please raise your hand during the panel discussion to ask a question. Before I continue, as you can see on this slide, today's discussion will include forward-looking statements, which are subject to risks and uncertainties. Please refer to our SEC filings for a discussion of these factors. And with that, I'll turn it over to Fady.

Thanks, Diane. I'm very pleased to be here alongside my colleagues and leading experts in HCM and the global cardiology community -- maybe we need to turn the mic off. As we share the primary results from ACACIA-HCM, the pivotal Phase III clinical trial of aficamten in patients with nonobstructive HCM. These results were just presented a short while ago in the hotline session at the Congress, and we're accompanied by 2 simultaneous publications, one in the New England Journal of Medicine and one in the Journal Circulation. These are the first -- this is the first ever positive Phase III clinical trial in nHCM. And the results show that the treatment with aficamten was associated with statistically significant and clinically impactful improvements in both exercise capacity and symptom burden as well as cardiac biomarkers. As we'll discuss today, these results create the potential for aficamten to become the first and only cardiac myosin inhibitor to address the full spectrum of the symptomatic HCM offering the first therapy that may directly treat the underlying disease process of HCM. The response from the HCP community at the Congress today was overwhelmingly positive. It's clear these results are resonating and there's genuine interest in what they could mean for the future in an HCM. Since sharing the top line results earlier this summer, we've moved swiftly to preparing a supplemental new drug application that we plan to submit to FDA later this year for aficamten and HCM. So we're extremely proud of these results and of the exemplary execution of this clinical trial. To that end, I'd like to express my gratitude on behalf of Cytokinetics to the patients, to the families investigators and the study staff who help conduct ACACIA-HCM. We thank you for your dedication. And with that, I'll hand it over to Dr. Masri and Dr. Maron to present the primary results from ACACIA-HCM. Dr. Masri.

Thank you, Dr. Malik, and hi everyone here online. I will present our final results on the ACACIA-HCM. So nonobstructive HCM is a common condition. It affects a lot of patients with hypertrophic cardiomyopathy. And it leads to reduction in exercise capacity and development of progressive symptoms. There are currently no available therapies that have proven to be effective in this disease. Aficamten, which many of you are aware, is available in obstructive HCM in many jurisdictions has favorable effects on diastolic function as well as a whole host of other key areas that are involved in HCM pathophysiology. And aficamten has favorable pharmacological profile, which allows for rapid up titration and dose adjustment with good tolerability. So this is the design of the trial. Probably you've seen this before. Patients with symptomatic nonobstructive HCM were randomized to aficamten versus placebo, 517 total patients, 1:2:1 randomization. Aficamten, dose range was 5 to 20 milligrams. All the patients were followed up to 72 weeks in a blinded fashion with a 4-week withdrawal period. The study continued until the last patient crossed week 36. In terms of our study visits and assessments, they are listed up there to the left of the screen. We've conducted for the dual primary end point that KCCQ assessment was done in every visit but the peak VO2 and cardiopulmonary exercise tests were only done at baseline and at week 36. Here are our endpoints, the primary dual primate endpoint change from baseline to week 36 in KCCQ clinical summary score as well as peak VO2. And in the secondary endpoints, we are tested in a hierarchal fashion, going from baseline week 36 in NYHA class, Z-score, which includes both maximum and submaximal exercise metrics, NT-proBNP change, left atrial volume index and time to first composite cardiovascular event. Safety outcomes typical of a myosin inhibitor trial, where we looked at LVF heart failure. But something really important to note, which is unique to aficamten at this stage is that patients who get EF between 40% and 50% due to us pushing the dose higher, they can get down titrated while staying on the drug without a drug holiday interruptions that are mandated. And so per protocol, only those with EF less than 40% have to interrupt the therapy, and they can be restarted on, but they have to enter therapy. These are our baseline characteristics. Again, as I mentioned, 1:2:1 randomization to 58 aficamten to 59 placebo. This is really, in my opinion, at least well-conducted study. The reason behind that, if you look at how many patients had family history of hypertrophic cardiomyopathy or pathogenic -- likely pathogenic variant for the disease, 2/3 of that. This is different from when you look at any patient with LVH. This was a study conducted to enroll such patients, 75% of them were on beta blockers, and they were highly symptomatic with reduced KCCQ scores. They had -- 1/3 of them had NYHA Class III biomarkers were elevated and peak VO2 was significantly reduced. Here are the primary results. So the first of the dual primary end point is KCCQ. The primary endpoint was aligned at week 36 for KCCQ and peak VO2. As you can see, the difference was 3 points at week 36 between aficamten and placebo in favor of aficamten with a p-value of 0.021. What you see here is that aficamten beyond week 12 consistently improved KCCQ clinical summary score. And at each time point, aside from week 36, you have a difference that, that is from 4.8 to 7 points, including the end of treatment, which every patient in the trial is supposed to go out at the end of the treatment from the trial, 5.6 points difference. During washout, these scores within 4 weeks were reduced back to similar to placebo without any difference. So the reason behind what you see here is the fact that placebo patients had fluctuations and our benefit as it reflects on KCCQ, which you can see clearly here between, for example, week 24, 36, 48 and end of the treatment. The second dual primary end point was peak VO2 affected influenced positively peak VO2 with a difference of 0.67 per kilogram per minute with a p-value of 0.003. If you look at many prespecified subgroup analysis, the treatment effect on both KCCQ and peak VO2 was consistent. And we here highlight to you some of the groups that some might feel are important, including beta blockers, intracavitary obstruction and genotype status. Again, as I mentioned during the presentation, I walked into the trial thinking we might end up having a super responder or a less responsive group. The reality is when you look at the prespecified analysis here, most of these patients have derived the same degree of benefit all in favor of aficamten. In terms of NYHA class, which is the first of the secondary endpoints tested in a hierarchal fashion, throughout the trial, you had more in favor of aficamten improvement NYHA class by 1 class or more. At week 36, it was 14% difference with a p-value of less than 0.001. This is the Z-score. It's a slightly complicated thing to explain, but the bottom line is, it's a comprehensive exercise metric where you have one exercise metric for maximal exercise programs that the patient can achieve and one is for submaximal exercise, which is what a lot of the patients do from day to day. They operate at the submaximal level, which is called ventilatory efficiency or VE VCO2. Aficamten positively affected VE VCO2 with minus 0.91. So with VE VCO2 slope lower is better with peak VO2, higher is better. And that translates, obviously, into our positive results on the z score with statistical significance. NT-proBNP was affected favorably from the first visit -- follow-up visit that the patient have continued to decline throughout the titration period and stay stable with the washout showing you how there is a quick rebound there. Slight -- [ somewhat ] advancing. All right. And so what -- so sometimes the question comes, what is the significance of having reduction in NT-proBNP or elevation in NT-proBNP. This is the recent study that we actually have been conducting for more than a decade now. It's an NIH-funded study, HCM registry, or HCMR showing that a really important predictor of outcomes in hypertrophic cardiomets NT-proBNP. And if you look at how they essentially reported the log and terminal naturally peptide during follow-up, you can see how by reducing NT-proBNP with the magnitude that was seen in the trial, you're actually shifting those patients on a curve that is different from the natural history created for NT-proBNP. So that's the message from this is that you end up achieving almost 50% reduction in the hazard ratio if you use that natural history study as your reference study. In terms of other endpoints, aficamten did affect less actual volume index, but it didn't reach significance with a p-value of 0.058 and time to first cardiovascular event was not different between the 2 groups. In terms of our safety outcome, aficamten was well tolerated. There were no new safety signals here. Aficamten in this trial was titrated to the maximum tolerated dose. It's different from our prior trials with aficamten, where we used the dose that was required to eliminate left ventricular outflow tract obstruction. And this is evident by having 81% of the patients on the maximum doses available 15 and 20 milligrams. Despite all of that, and despite the nonobstructive HCM population being a special population when it comes to sensitivity for cardiac myocin inhibition. The least square mean difference in LVF was 4.4%. This is similar to what was seen in the obstructive HCM trials as well with eficanten. And only 3% of the patients on aficamten had to hold aficamten or had treatment interruption because they have reduced left ventricular ejection fraction. A little bit more on the safety outcomes. I'm going to just focus on heart failure here. 4.7% aficamten, 1.2% of placebo. And all of this happened during the titration -- initiation and titration phase because we had to design a trial in a way that we can get patients to the maximum dose available of the drug. And these are -- while they're heart failure, these are events where it can be managed by dialytic and outpatient management. 10.5% of the patients had an [indiscernible] 50%, but only 2 of them had a serious cardiac event of heart failure. And importantly, in my opinion, that this 27 number 21 of them remained on aficamten, either at the same dose or the lower dose during the trial. So there was no treatment interruption here. One thing that I would like to draw your attention to is new onset atrial fibrillation, which is an important -- it's becoming a really important metric in HCM trials. There was no difference between aficamten and placebo in the new onset of AFib. One important reminder as well is that the safety goes through the whole period of the trial up to 76 weeks. So in conclusion, aficamten improved exercise capacity patients-reported health status and symptoms as compared to placebo. And over 36 weeks, we've shown you that aficamten hit on the dual primary end point as well as 3 of the 5 secondary end points. and more than 80% of these patients elected to enroll in forest HCM to look at the longer-term safety and efficacy of aficamten. Thank you.

Okay. Great. Thanks, Ahmad. So I'm going to just get into some of the additional data here in a second. It's going to touch on some considerations around treatment benefit and safety. But I wanted to just start, maybe if I could, just with a couple of comments. Some of these comments, we'll get back to maybe at the end in the question-and-answer period. But I'm older than Ahmad by actually quite a bit. He's absolutely a superstar. I'm a little bit older and I've got more gray hair. And this gray here, by the way, I want to tell you, a lot of it comes from having taken care of HCM patients for almost 3 decades. No. It's a long time. I've taken -- have been in trenches, take care of these patients for that period of time. And I've seen a lot of them. And I got to tell you that we had experience, which is vast, I never actually thought we would see the day that we are seeing today, actually, where we have advanced a treatment for this population of sick nonobstructive patients. And I say that because I thought it was always a very challenging bar to reach. And I think the fact that we have shown that today is not just a milestone, but actually a kind of a transformative period. And I don't think -- I not use that term likely, I mean I'm not just saying that because I'm here talking to you guys. I actually believe that to be true. And a lot of it comes down to being able to have the opportunity here to make these patients feel to function better, of course. But a lot of it stems from the enormous amount of frustration these patients feel because there's nothing available for them. And so this has met in a way one of the greatest unmet treatment needs they've had in HCM. Okay. And it has to be framed, I think, in that context here when we're looking at the numbers and trying to understand them and relate to them, we've got to take a step back we've got to really understand that this is a pivotal moment. And so with that said, I'm going to touch on 2 additional considerations that I think are on people's minds that have to do with treatment and safety. And those come out of -- as Ahmad has alluded to this sub study that was done, it's published as well now in circulation. It's about the global efficacy of the drug in this population. And again, this is reiterating but again, huge unmet need in this population related to morbidity. This is a drug that addresses the underlying basis for why these patients are frustrated, improving diastolic function. We'll get to that more in a minute. We talked about the outcomes that were achieved ACACIA and the primary results. And so the question then is why did we do this substudy then? What was the purpose of doing this since you've just seen the primary data. And I'll just make one important point. And I think it's really important because I see this happen a lot is that you just saw very elegantly shown the mean changes in indication, right? The mean changes between the aficamten and placebo arms. But sometimes those mean changes, they're hard to contextualize in terms of clinical meaningfulness. We get that a lot, right? What does that mean when you see those mean changes in KCCQ, peak VO2. And how does it -- when we talk about clinical mean, what does it mean for the patients, okay? And so we address that point here. Okay. And the way we do that is that we contextualize the treatment benefit as it relates to patients. It's the within patient change within clinical outcomes that are relevant here. I'll show you what those are in a second. That's what we call a responder analysis because it contextualizes or translates this data that you just saw into how we want to think about it as it relates to a treatment benefit for the individual patient. So that's the purpose here. And if we look at -- sorry, this is -- okay, that was just the clinical trial that you just heard this trial design. So I'll go back over that, but we saw that we'll skip right to, again, this concept of why we did this. The responder analysis, again, takes the vantage point of what is the treating physician and the patient, for that matter, too, want to know if they were in a position to use this drug clinically. First and foremost, because it's the most important treatment goal we have in this disease, a low mortality disease, it's about morbidity. It's about making my patients feel better, right? I mean that is the primary and most important thing we can achieve. And we analyze that through 2 ways. PGIC, it's just a questionnaire that is a one question questionnaire that asks the question, how are you feeling now on treatment compared to before? The same, better, much better, very much better. worse, very much worse, et cetera. And that PGIC metric is important for 2 reasons. One, it's been used scientifically by regulatory agencies and [indiscernible] to anchor to establish whether it is a meaningful change in health status with other kinds of patient-reported outcome measures like KCCQ, okay? That's number one. And number two, it's elegant in its simplicity. This is how we work, right? This is what we want to know is how we interact with patients. We ask them that question and then we get something back. And that is our bar clinically. And so the PGIC reflects clinical practice in that way. Of course, we also want to know do my patients function better, that's measured objectively with crapola exercise testing. And in this case, we want to be able to support those important clinical changes by demonstrating that the therapy is biologically active in a way that makes sense to support the clinical benefit. And in here, that we're looking at measures of diastolic function. ProBNP, a direct reflection of LV filling pressures, left atrial pressures, cardiac structure, left data volume, again, an indirect measure, but a very strong one of left ventricular filling pressures and diastolic function with septal E prime and noninvasive measure of LV or left ventricular relaxation. And so the way this works is simple. We just take those measures, and these are the endpoints in this substudy analysis, okay? In each of these measures, by the way, each of the 5 have each been related to adverse outcome in ACM. And so they are predictors of bad things happening to nonobstructive patients. So they in and of themselves are very important predictors of what happens in the future to patients. And so therefore, they are relevant in that way, and that's how they are defined as -- and then we define them as a responder analysis by putting a clinically meaningful threshold of change that we attach or associated to each one of these, okay? So for improvement of symptoms, it's a one or more improvement in NYHA class plus a one categorical change in PGIC, a 0.5 or more improvement in exercise capacity, a greater than 10% decrease in left atrial volume, a greater than 10% improvement in septal E and a greater than 50% reduction in BMP. Okay? So those define a threshold for which -- that if we achieve, we would consider to be clinically meaningful based on other supportive data for that individual patient. So if we look at that then in terms of what we saw. So that's the responder analysis. That's what we did, and this is what we found. So if you look at each of those 5 outcome measures, and you look at what the response was, what the treatment benefit was in this symptomatic functionally limited non-obstructive population, you see that over 70% of patients achieved the threshold of a clinically meaningful symptom improvement. It's a lot. And then you've got about path the patients achieve a functional improvement in peak VO2 of 0.5 million per kilogram per minute or more. You've got about 1/3 with that improvement in left atrial volume, more than half with the improvement in LV relaxation, and over 63% with a substantial reduction in BNP, again, reflecting a very important and impactful biological effect on the HCM heart with that kind of reduction in BNP as well, okay? And so those are the outcome of what we saw with aficamten treatment. And then if you look at that a little bit more -- a little bit more detail, sorry, can you go back 1 sorry. This is a little bit hard to advance. So I got -- it stuck. Can you go back one?

[indiscernible]

[indiscernible] Okay. Sorry. Got it. I understood. Okay. And so if we look at that a little bit more as it relates to the effect of those thresholds over the treatment period and also as it relates to placebo, if we take the 2 outcome measures, again, that clinically are the most impactful, making patients feel better on the left with that one or more improvement in NYHA as well as that one categorical improvement in PGIC. And then you've got the -- on the right, the peak VO2 functional there. And you can see the red bars are aficamten and the blue bars are placebo. And you could see with improvement in symptomatic benefit First of all, again, that's 71% at the end of the treatment period or the end of the assessment period of 36 weeks. But you also see that, that improvement in symptom status starts early. It starts at 2 weeks with a greater proportion of patients on aficamten with an improvement compared to placebo, okay? And it's maintained throughout the treatment period, and you can see that benefit in functional status as well that we talked about at almost half the patients receiving a benefit of an improvement in exercise capacity. And then here, if you combine those 2 measures, feel and function, you see a greater proportion of patients on aficamten achieving that feel and function combination than placebo. In terms of diastolic function, same story here. Patients greater proportionate, aficamten patients achieving the threshold of improvement in left atrial volume improving diastolic filling with relaxation and then the BMP, which is probably the most visually impressive here performance here, undoubtedly telling us that there is a clear physiologic effect here in terms of lowering LV pressures. And again, if you look at the benefit here across the platform, so to speak, across both clinical, structural and functional measures. If you look at this, that you've got more than half of the patients on aficamten achieving 3 or more of those outcome measures versus 13% on placebo. And then you can translate this. I'm going to finish up here in 2 seconds, you can translate this data then into a metric that we all can relate to as well, which is the number needed to treat. And if you look at the number needed to treat across the 5 outcome measures, it ranges between 1.7 for BNP and 14 for exercise capacity and symptom status is 5.5%. To give you some context here, the SGLT2 inhibitor drugs, for example, which have mean improved its in KCCQ of 2 to 3 in HFpEF and HFrEF, for example, have a number needed to treat to improve KCCQ incentive status of 14 to 18. So this 5 million far superior in terms of number needed to treat than a lot of the other traditional heart failure medicines that we have for non-HCM heart failure. Finally, to finish up here -- okay. Again, in summary, then over a relatively short period of time, and that's the other point. We're talking about 36 weeks I showed you. You see here that clinically relevant measures were achieved 3 or more achieved much greater in aficamten than placebo with the number needed to treat that we talked about. And again, this all underscores a broad effect, a broad treatment effect here in this symptomatic functional limited population of nonobstructive patients that supports a potential benefit of aficamten in this significantly underserved population. And I'll finish with just 2 or 3 slides about safety for a minute because this is really obviously very important as well. And as was probably noted, there were -- is a slight imbalance between -- in serious heart failure events in the aficamten-versed placebo groups. The point there is that those low-water events were independent of ejection fraction. Okay? Independent of ejection fraction, and I'll show you why that is in a second. And therefore, titration of aficamten was, as we all know, is algorithmic-driven, right, in the trial, you went up on the dose based on ejection fraction. That was a blinded decision that we've made, and that was the up-titration strategy. For us, the open-label extension in which patients could roll over into different story, unblinded look at the echoes you could include physician judgment in addition to the EF to make decisions about up-titration, and I'll show you that forest data in just a second. Again, you've seen this that slight imbalance in serious heart failure events in the afi versus placebo groups of 12 versus 3. And then if we look at that group, the 12 versus 3, here it is in forest. So these are patients that were in ACACIA, they roll...

These were [indiscernible] ACACIA.

These were patients that were in ACACIA that rolled over into forest.

But this is the timing of the patient's in ACACIA still.

I'm sorry, on the wrong side. This is ACACIA, sorry, you are -- absolutely sorry. This is still ACACIA, sorry. So I'm showing you the 12 serious adverse heart failure events in ACACIA sorry, where you can see them in the aficamten pink circles, all of them, all 12 correspond to the titration period, all 12, every single one of them. Again, forced up titration based on the af in the placebo group, difference, okay? So all heart failure events in the Acacia trial occurred in patient aficamten in that titration period. All of those patients, by the way, responded very well to minimal heart failure therapy, usually low-dose diuretic therapy and were discharged within a very short period of time. Now if you look at forest then for a minute, taking those patients that rolled over into forest as part of the long-term extension therapy of aficamten really important. I think this is compelling here, which is that if you look at these patients, first of all, the patients that were on aficamten in ACACIA who rolled over into forest, there are no pink circles. That's because none of those patients had a heart failure event in forest. The only heart failure events were those patients, and there were 2 in placebo arm in ACACIA, who rolled over into forest on aficamten. And those 2, which is 1% of the population occurred during the titration period, right? And I think what this means -- to translate this, what this means is that when we look at the heart failure events, it tears that these could be significantly mitigated using a strategy of dose titration that is reflective of clinical practice, which is what forest titration does. It uses an unblinded assessment of the echo with clinical judgment to make decisions about titration. This is how we practice clinically in the real world. That's what forest mimics, and this is what I think we could expect to see based on that experience. All right. My last slide, and then we'll open it up, which is a conclusion slide. okay. So as anticipated, escalation of dose of aficamten in isolation clinical data resulted in greater events when compared to dose selection that incorporates clinical judgment and Echo assessments. The current algorithm with forest essentially mimics clinical practice as well as what we have for eaten in symptomatic obstructive HCM and the incidence of serious heart failure in forest, including patients never exposed to aficamten before is substantially lower as the that noted in ACACIA. Thank you.

So thanks, Marty. Thank you, Dr. Masri, and I'd like to just welcome Dr. Christina Paitazoglou, up to the stage, and we'll get a Q&A session on the go now.

So I'm actually going to start out with you, Dr. Paitazoglou. So your HCM doctor and heart failure doctor here in Germany. Can you just tell us a little bit about yourself, what your practice might involve? And then as a second question, a follow-up, you really were not part of the study and you saw the results really for the first time, just a couple of hours ago, maybe you can contextualize those results tell us what you think about them and what they mean for the patients in your clinic.

Yes. Thank you very much. I first of all, congratulations to your trial and the results, very impressive. My name is Christina Paitazoglou. I'm a heart failure specialist. I work at a university clinic, I'm Head of the Heart Failure Department at our clinic. And I am also the leader of the hats network for the State Department with the Ministry of Health. So we get a lot of referrals to our university clinic, not only HCM also other heart failure pathologies, but we see a lot of patients with HCM because in Germany, those patients treated with a myosin inhibitor now and titrated is in the university clinic, so we've treated all of our patients. Not all of our patients have obstruction. We get a lot of referrals with patients who do not have obstruction and there's a as Dr. Martin said, a clear unmet need. We don't have options for those patients. They are still very symptomatic and we can only treat them according to the guidelines of heart failure. And we did not yet have so far a targeted treatment. And that's why it's a great milestone. It's the first positive trial in this difficult population. You do not have an obstruction. You give a specific drug, and the patient gets better. So for us, it's a new opportunity. We have a new tool. We can help the patient, we can give them hope for treatment. And your results are for me, very encouraging because, as I said earlier to you, I was not expecting the same the same results as an obstructive because it's a different patient population, the patients with non-obstructive symptoms derived from -- it's more complex derived from many factors. It's a diastolic dysfunction. It's microvascular disease. And the drug is very specific and it led to an improvement of the patient and that is very important. Not only hemodynamically, the patient feels better and we know the safety profile. I've treated in Germany, a lot of patients with aficamten and obstructive. I feel very safe with this drug. And now it gives me hope for this patient population. I don't know if it answers exactly your question. It was very widely asked.

Yes. I mean that was a beautiful description. Maybe just to dig a little bit deeper. You mentioned to me earlier on that you were actually an expert in cardiopulmonary exercise testing. You've done some advanced training in New York City. So you -- and you also treat other pools of heart failure. So you have a special understanding of the magnitude of the treatment effect in terms of the peak VO2, for example. Maybe you can just go down on that a little bit to put some perspective over there.

The CPET training, yes, we do a lot of CPET in our heart failure patients. And that is not a very common examination for cardiologists. Not many have the expertise in CPET. I did it in New York at the Columbia University also in Heidelberg and we performed CPET in all of our patients, even advanced heart failure. And I know these parameters, you presented the peak VO2, even the stop they are parameters, prognostic parameters for heart failure. And usually, in my patients, they go down. not the slope. The slope goes up, which is bad. And in your trial, we saw an improvement, and it's very difficult to change those parameters in the CPET, an improvement in peak VO2 it's very difficult to achieve. Even for those patients and exactly for those patients without obstruction. So maybe there a little bit more broader patient collective like HFpEF patients. It's I don't say impossible, but it's very difficult to improve them. And many trials have not chosen this parameter as a primary end point, I think, because they are afraid of the results. and they may have failed. So an improvement, consistent with an improvement in symptoms and the NT-proBNP going so much on after 36 months is for me as a clinician, a very great sign, and it makes me happy to see that. And you have to leave the patient longer on this medication, maybe this parameter even get better because the medication targets exactly the disease.

All right, beautiful. Thank you so much. Maybe moving on to you Dr. Masri. How would you characterize the overall safety and tolerability data that you presented today? And what did you and your investigators actually see and manage over the course of the trial? And what do you see in some of the patients that you might actually have now in the forest openly extension?

That's a great question. So as we alluded to is that the numbers that you're looking at, you have to really understand what it means when we say that there was a reduction in LVF. So I think of them as kind of almost 3 buckets. One is, what was the average reduction in LVF brought the trial or at x86, for example, and that was 4.4%. The second bucket is we are dosing to titrate 2 LVF. So like you do with blood pressure, if you want to dose to blood pressure, you will have instances of reduction in blood pressure. Same with EF, you are dosing to it. There will be insensitive reduction. The question is, can you keep the patients on the drug or lower dose can you continue without having cycles of stop and start? And are you causing these patients with the reduced ejection fraction, either irreversible issues or something really significant that is having them exit or just say, "I don't want to do this. And this is not our experience. We've already shown you the numbers. I'm not going to repeat those. But even our local experience where we have now more than 40 patients, or so with nonobstructive HCM on aficamten as part of these clinical trials in the open forest HCM study. This is not our experience. Our experience is, generally speaking, in the blinded period. I don't know anything about what happens to these patients in the blinded trial. But if these patients suffered a significant event or something meaningful per protocol, I would have known about it, for example, and this was in the case. And in forest, as Dr. Martin alluded to, we actually are not dealing -- we dosed all of these patients in forest, and we are not dealing with any of these major issues. Even if you end up seeing a reduction in EF less than 50%, if the patient is asymptomatic, the reason why we go down on the dose is because we don't know what the long-term meaning of that. So we just want to keep that threshold. And to me, essentially, the safety is on par with what we have expected it to be. It's just you have to put into perspective, the design of the clinical trial, where we're taking 517 patients, randomizing them 1:1 and for those 289 patients we're telling them all of you have to go through the same process. This is not clinical practice and not forest HCM design either.

Got it. Thank you. Dr. Maron, so which of the patients in your practice, do you think that the results of Acacia are most applicable to? In other words, do you think that there are patients who would drive more benefit than others. And thinking a little bit more ahead, how do you think the conduct of ACACIA-HCM may translate into future diagnoses of new nonobstructive HCM patients.

Yes. Great. Thanks, Steve. So I'm going to get to that in one second. I just want to address one other point that was raised in is important is that -- and this is something I'm kind of seeing and hearing, there was -- I think that there was this expectation, I think, in some people's minds a little bit that we would see a treatment benefit here that was similar to the obstructive HCM across, right. There was that expectation, I think, psychologically or otherwise. And of course, that was never going to be the case because in obstruction, these drugs so effectively lower acutely that high LV pressure that it translates into such a magnitude of benefit that our patients feel that you weren't going to ever be able to achieve that kind of degree of improvement in this different population. That doesn't mean that the clinical benefit we're seeing in non-obstructive is less significant, necessarily, but that to compare the 2 experiences isn't fair and shouldn't really be done. And I think I see a little bit of a psychological thing going on here with people were expecting those same treatment benefit others and not seeing them psychologically has created a little bit of consideration. And I think that, look, we made the case here, I think, with this responder analysis that the benefit is clearly there in this population. It's different than nonobstructive. Number two, I'll make the point that and I'll support that point with this following anecdote, which I think to me, tells the whole story almost. If you look simply at the patients that were in ACACIA, who rolled over into forest, of which there several hundred, okay? If you look at those patients, you will see that 99% of them remain in forest on aficamten, months, many months later, okay? Almost none have dropped out. I can tell you -- there's one thing I can tell you, it is that these patients, if they were not deriving a clinical benefit, would not slip back in for all of the visits that are necessary for Forest and everything else, if they did not feel that there was a true treatment benefit that they were getting with the therapy, okay? That kind of level of commitment would not happen. And it speaks, I think, essentially, to the treatment benefit that we're talking about. And so if we take that for a minute then, Steve, to answer your question, is that Ahmad showed that there really was no heterogeneity in treatment benefit across all of these different subgroups. And so the answer of who would potentially benefit from this drug, it would be nearly all the obstructive patients who fit into the criteria of ACACIA, symptomatic functionally limited, patients would be potential candidates for the therapy. And by the way, I mean it's tough to sit across from these patients that are really frustrated but how they're doing. And you know that you've got a therapy that can improve how they feel and function across the nonobstructive spectrum and not potentially think about that option for them. okay? I mean that is an important principle. And so I think we're going -- what we will do in practice is that we will potentially consider all the patients that fit into the ACACIA trial profile as potential candidates for therapy.

All right. Perfect. Thank you. beautifully, ad. We're going to switch over to some questions from the audience, Joanna. There you go.

Cory Kasimov with Evercore. Dr. Maron, I just want to follow up on a point you were making and just trying to ask it a little bit more bluntly. So the discussion in trial, I think balanced overall again, brought up this point of a modest effect when describing the study. Given the dearth of options for these patients with nonobstructive AGM, in other words, having none. Are there patients who you think would not be candidates for aficamten assuming approval or perhaps subgroups of patients where you would not want to at least try this product?

So I think -- well, 2 things. One is that I think I would push back a little bit on the characterization that came up to you about the treatment benefit was modest. I think what we're -- what I'm trying to say here, based on the data and our experience, is that I don't think that, that is necessarily characterizing what's going on here in terms of treatment. I think the treatment benefit is much more than that for the vast majority of patients that are on therapy, okay? Again, it may not reflect as intuitively in mean changes, we saw the benefit that I think supports that with the responder analysis where you've got 70% of patients feeling better by a clinic meaningful change, okay? So that's really important. So I don't look at it, by the way, by the data were in my own practice as a modest change. I think it's much more than that. Two is that I think that -- look, I think we don't really have any idea about whether there's any kind of subgroups, we don't really have a good idea about whether there are specific subgroups that would be better responders or not so much good responders in this population yet. I think what we do know is what Ahmad showed from the primary paper, which is that there was no treatment difference among a lot of different prespecified relevant subgroups here, all of them were getting benefit here. And I think we have to go on that persumption right now that, that is the case. And that's how we'll practice clinically.

Tess Romero from JPMorgan. So just to piggyback off here, maybe just thinking about how you compare and contrast the treatment effects that you've seen across both of these patient populations. What were the true surprises in the data relative to your expectation? And then how should we be thinking about sort of longer-term treatment effects and sort of where this could go over time?