Good morning. My name is Robert Blum. It's my pleasure to welcome you to the Cytokinetics webcast to discuss the results of the FORTITUDE-ALS trial that were announced here yesterday at the American Academy of Neurology Annual Meeting in Philadelphia. This meeting is convening here in Philadelphia but is also being webcast. We welcome those people who are dialed into the webcast. Today, we're going to elaborate on some very important data that were presented yesterday at the AAN relating to the development of reldesemtiv for the potential treatment of ALS. I'll be making, others will be making, some forward-looking statements. We will refer you to our SEC filings with regard to caveats to those statements. We do not undertake any obligations to update those statements.
As you know, to set some context, Cytokinetics is a pioneer in developing a new pharmacology relating to modulators of the contractility of muscle, different types of muscle. Today, we're going to be, in particular, talking about reldesemtiv, a next-generation activator of troponin found specifically in fast skeletal muscle. As we are advancing our pipeline, we're aiming towards the treatment of severe diseases of muscle dysfunction and weakness. Today, I'll be joined by colleagues Dr. Fady Malik, who leads our research and development and has been with Cytokinetics since we founded the company in 1998. It's his vision around these modulators of muscle biology that we've been prosecuting as we advance our pipeline. You'll hear from him relating to the underlying biology and mechanism of reldesemtiv that sets the table for the discussion of the results to follow.
I'm also joined by Andy Wolff, our Chief Medical Officer. Andy Wolff has been with us 15 years. Like Fady, is a physician and has overseen the conduct of most of the clinical trials that we've conducted at Cytokinetics. He'll be moderating a Q&A session to follow. I'm especially pleased to be joined by two leading ALS treatment physicians and clinical research investigators, people who have pioneered the advancement of new medicines in ALS for decades. Jeremy Shefner, who is lead investigator of FORTITUDE-ALS, Professor and Chair of Neurology at the Barrow Neurological Institute, Professor, Chair of Neurology, University of Arizona in Phoenix.
Jeremy, a founder of the NEALS Consortium and Principal Investigator of this study, will be in a position to provide context for these data, not only with respect to this mechanism, but also in particular as it relates to the development of investigative medicines for ALS for many years. Angela Genge, a member of the steering committee for the FORTITUDE-ALS trial, Director of the Clinical Research Unit in Montreal, overseeing the Neurological Institute there. I should point out, one of the lead enrollers of patients, not only in this trial, but also for the BENEFIT-ALS and VITALITY-ALS trials, having a lot of experience with this mechanism over quite a long period of time. Thank you both for joining us on the panel here this morning.
I'm going to make some brief introductory comments, Fady will join me in a discussion of our pipeline and how it's advancing. First, may I underscore that this is a picture of the sarcomere. It's the fundamental unit of muscle contractility around which we've been focusing our drug discovery and development activities for many years. As you may know, we are advancing now a broader portfolio of compounds that are all directed to this structure. Activators and inhibitors that we divide into two verticals, those that pertain to cardiac muscle, like omecamtiv mecarbil in phase III for the development of heart failure. Also other compounds like AMG 594, a cardiac troponin activator, and CK- 274, a cardiac myosin inhibitor. I'll have more to say about these later in the morning.
We're going to primarily focus to reldesemtiv, a small molecule activator of troponin, as you'll hear from Fady, that gives rise to a therapeutic hypothesis that we were testing in FORTITUDE-ALS. Putting that into another context, here's our pipeline. We are advancing, as you can see here, in mid-stage and late-stage clinical trials, investigative medicines, all of which have been discovered and optimized and characterized by scientists at Cytokinetics in our laboratories and now in advanced clinical trials. We've conducted over 50 clinical trials over these many years. Thousands and thousands of patients enrolled in published, peer-reviewed, placebo-controlled clinical trials. FORTITUDE-ALS ranks amongst those as yet another study, as you'll hear, that serves to represent opportunity moving forward. I'm going to set the table a little bit now in terms of a discussion of FORTITUDE-ALS, I'll come back and make some other comments.
What I'll say is, as we go through the program this morning, please keep in mind that we do phase II clinical trials for a reason. We do phase II clinical trials in order to understand the potential for consistency of effect across doses, time points, end points, and subgroups in order to be able to properly design a phase III trial that could have the robustness with a statistical significance of P less than 0.05 in order to support potential registration. I would put forward, as you'll hear now in further detail, that FORTITUDE-ALS most definitely accomplished the objectives that we established, in that way, lend support for progression of reldesemtiv into further clinical trials. I'm going to turn the podium over to Fady.
He'll walk you through more of what we understand for the mechanistic biology, he'll turn it over to Jeremy to present not only those slides that he presented yesterday to the audience at the AAN, but also elaborate on those with additional data, he'll be joined in conversation led by Andy and together with Angela.
Thank you, Robert. I'll just remind you of the work we've done in pioneering the discovery and development of modulators of the sarcomere muscle, sarcomere being that fundamental engine that drives muscle contractility. It's an elegant biological machine whose components can be taken apart and reconstituted in ways that facilitate drug discovery, and faithfully are replicated from biochemical to intact muscle, to the activity at the level of an intact organ or intact muscle. This is an area of biology that we've used as an engine for drug discovery at Cytokinetics for nearly 20 years now. The assay that underlies most of our drug discovery is this high-throughput screening assay of a functional sarcomere, where we can replicate the activity of muscle in a test tube, so to speak, or a plate of compounds, 384-well plate of compounds whose activity may modulate that of this reconstituted sarcomere.
Taking it apart into its fundamental components, actin the roadway, the S1 head of myosin, which is the motor domain, and then the regulatory domains of troponin and tropomyosin that regulate that actin-myosin interaction in a way that's faithfully replicated in this biochemical assay. Over the years, we've conducted nearly 20 million screens of this system in cardiac, skeletal, smooth muscle of various different types. This fundamental biological assay has led to discovery of compounds and the development of the compounds that Robert introduced at the beginning of the session here. A fast skeletal muscle troponin activator is a small molecule that binds to troponin. Troponin is fundamentally the transmission in the sarcomere. Once it binds to calcium, the myosin is allowed to engage the active filament and for muscle contraction thereby to ensue. Reldesemtiv binds to troponin in a way that increases its affinity for calcium.
It changes the gearing ratio in the sarcomere. Fundamentally, we can show that in this assay that's shown on the slide here, where the affinity for calcium is demonstrated. The consequences of increasing the affinity of calcium within the sarcomere, sensitizing the sarcomere are threefold. First, skeletal muscle troponin activators amplify the response to motor neuron input. This is important for what we'll be talking today, given that ALS is a neuromuscular disease. Second, they increase muscle power. Third, they improve muscle fatigability by reducing the requirement of both nerve and calcium in order to drive muscle contractility. We've shown this in a number of preclinical models across a number of different disease conditions, including ALS, SMA, myasthenia gravis, models of heart failure, and so forth. Much of that work has been published.
Importantly, what we're able to show and we've shown now both with the first compound in this field, which is tirasemtiv, and now with the second reldesemtiv, is that the fundamental mechanism of action translates from the preclinical setting to the human setting. In this case, what we're looking at is a force-frequency relationship of muscle. In an intact animal or in an intact human, if you stimulate the nerve to muscle, the muscle is activated. The faster you stimulate the nerve, the greater the muscle is activated, and you see more force that is developed. If you dose in the presence of reldesemtiv, a troponin activator, you see that curve shift to the left because the muscle is now more sensitive to nerve input and generates more force.
Either the muscle at the same stimulation frequencies, the muscle generates more force, or for the same amount of force, it requires less nerve input. On the right, you see the human translation there, where in an assay that where you stimulate the nerve on the outside of the knee, you see an activation of the muscle in the front of the leg and a concentration exposure-dependent increase in the force that's developed there as you stimulate the nerve. This certainly shows that the mechanism of action is active in humans and also describes its exposure-response relationship. In going from our first compound in this area, tirasemtiv, to our second reldesemtiv, what you can see in the two graphs on the right, both conducted in humans, that reldesemtiv provides a greater pharmacodynamic response.
An increase in the force that is produced in response to nerve input to a greater extent than we saw with tirasemtiv. Reldesemtiv was also better tolerated in phase I studies than tirasemtiv was, fundamentally, these results supported progress of reldesemtiv into a phase II program. Now we have an extensive experience in trials in ALS, as you probably know. We've conducted several trials with tirasemtiv that gave us experience with its mechanism of action and also provided fundamentally the belief that there was biological activity of this mechanism of action in patients with ALS. Now FORTITUDE-ALS, which I think builds upon those, confirms that this mechanism of action is relevant to ALS, as Jeremy will go over with you. Almost 2,000 patients have been enrolled from a network of 80 sites in 12 countries in the U.S., North America, and Europe.
With that, I'll turn it over to Jeremy Shefner, who will walk you through the results of our latest study, FORTITUDE-ALS. Thank you, Jeremy.
Good morning. Just to dive right into what the study looked like, this was a large phase II trial with a duration of 12 weeks of active treatment and inclusion criteria that were fairly typical of many ALS trials, but somewhat less restrictive than has been the case for a few phase III trials, and in particular, more restrictive than the edaravone study, which led to approval using a group of patients that were quite early in their disease but had diffuse disease and were rapidly progressive. We took a broader approach. We did allow both edaravone to be used as well as riluzole, but patients had to be stable on either or both. The primary efficacy endpoint was slow vital capacity, and the analysis was changed from baseline to 12 weeks.
That was true for the ALS Functional Rating Scale Revised and for Muscle Strength Mega-Score, which is a quantitative measurement of muscle strength using a handheld dynamometer and with individual muscles normalized so that you could create an average strength. We also looked at safety markers as well as two pharmacokinetic days, I'll show those data as well. The study was predominantly enrolled in North America, More than half of the patients came from the United States. Approximately 100 patients came from Canada, Small numbers of patients came from the Netherlands and Ireland, slightly more from Spain, and then Australia had five sites and contributed 20 patients. This is the study schematic. After screening, patients were randomized equally to one of three doses of reldesemtiv and placebo. They were studied on active treatment for 12 weeks, withdrawn from drug, Then studied again four weeks later.
Outcome measures were assessed at baseline, at screening, then approximately every four weeks after that, except for the two-week period as well. This is the description of the baseline demographics, it's a very busy slide. The things to note are that average age, most of the baseline characteristics of these patients were both well-balanced across the groups, but also typical of ALS trials. The average ALS Functional Rating Scale value at entry was about 37, and the average slow vital capacity was about 85% predicted. About 25% of patients were on edaravone at the time of their study enrollment. This shows the patient disposition. It's quite a busy slide, but there's some important points here. First, 606 patients were screened to randomize 458 patients. 457 of them ultimately went on drug, divided equally into the four groups.
If you look at the red bars, you can see that early terminations from study treatment were actually quite well-balanced across groups. There was not a tendency for more patients to drop out over time in the treated groups versus placebo. On the other hand, if you look at the bars below that, the blue bars, you can see that dropouts, early terminations from study treatment due to adverse events did have somewhat of a dose-dependent pattern, although the numbers were still quite small. Finally, the purple bars below that, there was an increased number of dropouts to perceived progressive disease in the placebo group as compared to the other treatment groups. Small numbers, but still interesting. This slide shows the pattern of when patients terminated early.
If you have seen the tirasemtiv data, you know that this is a markedly different pattern, where for the prior drug, there was very significant dose-related early terminations from study treatment. Here, the groups are all quite similar, the drug tolerability is quite good. The pharmacokinetic data at week 2 and week 12 is presented here, You can see that reldesemtiv has an increased serum concentrations in a dose-dependent fashion, That if you look at the differences between week 2 and week 12, there really aren't any, There's no tendency for the drug to accumulate over time. Looking at these concentrations and relating them to one of the figures that Fady showed, the pharmacodynamic demonstration of force in normal human volunteers. This shows the increase in force that tirasemtiv is associated with at different stimulation frequencies of the nerve.
You can see approximately where the different dose levels take you on the pharmacodynamic pattern. You can see that they're at the point where you should see a signal for all doses. The two higher doses are actually fairly close to each other in terms of what you might expect to see based on the concentration relationships to increases in muscle force. Here is the primary endpoint, which is changed from baseline to 12 weeks in slow vital capacity. The gray bar is the decline in slow vital capacity for the placebo group, and it's approximately 6.5 percentage points because this measure is expressed in % baseline. This is somewhat slower than what we expected and what it was seen in many previous ALS trials. Clinically and experimentally, we usually think that vital capacity drops by about 3% per month.
In this study, it dropped by a little bit more than 2.15. A slower decline in these patients with regard to this measure. There was a dose-dependent effect of this drug towards reducing the decline further. The primary analysis, which was a mixed model for repeated measures analysis with a weighting of placebo and the lowest dose to a smaller extent versus the two higher doses. That analysis reached a significance level of 0.11, approaching but not meeting a 0.05 level. If you look at the pattern of responses over time in the three groups and in the placebo group, you can see the gray bar is the placebo group drop, and you can see that at all time points and all dose levels, there is a tendency for the treated patients to do relatively better than the placebo patients.
The effect is largest at 12 weeks and actually is maintained, for the most part, for four weeks after that, even though patients were off drug. We did an analysis where we combined all patients on drug and compared them to placebo. Here's a graph depicting that. The P value for the comparison of drop at 12 weeks is 0.1, but the relative risk reduction, the relative reduction in rate of progression in slow vital capacity is 27%, which is, I think, if verified, would be thought of as a clearly clinically significant change in ALS progression. ALSFRS looked very similar at virtually all time points and at all dose levels. Patients progressed less quickly on the ALS Functional Rating Scale than placebo.
The mixed measures analysis showed a P value of 0.09. Again, there was a tendency for the difference to be maintained four weeks after withdrawing from study treatment. Comparing all three doses together in the same way we did previously for vital capacity, the difference seemed to get larger with time, and the comparison at 12 weeks, comparing all three groups to placebo, was quite highly statistically significant. It was significant at the 0.01 level. Again, you can see the persistence of the effect. If you're ALSFRS aficionados, you know that it's basically 12 questions, and you can divide them into four domains: a fine motor domain, a gross motor domain, a bulbar function or facial and tongue domain, and a respiratory domain.
The last two actually change much less over time and so contribute less in the total score to the drop than both fine motor and gross motor. You can see that in terms of the effect of tirasemtiv, the effect is primarily seen in those two domains, in gross motor more than fine motor. There are P values for each comparison. There are many of them, and it's probably not worth taking them too literally, but they do suggest quite a prominent effect. This is the strength evaluation. Again, you can see it's the same pattern that for, in this case, most time points and all doses, there is a tendency for patients to lose strength less rapidly than patients on placebo.
There are two P values here, one for a comparison of slope, which had a P value of 0.3, the other is a change from baseline to week 12, similar to the other analyses, and that had a P value of 0.13. That's there. If you average all groups together and look at the difference, it looks very similar. The magnitude effect is slightly smaller. It's 0.21, and the P value for that difference between all treated patients versus placebo at 12 weeks is 0.2. We did a variety of subgroup analyses, These are disease-related subgroups.
Patients who entered the study with greater or less than 80% vital capacity, patients who entered the study at greater or less than a given value of ALSFRS, greater or less than a two-year symptom onset. Site of onset and a pre-study evaluation of how fast people were progressing with respect to ALSFRS. This is a forest plot where every point to the right of the line, which is a zero effect line, suggests a benefit of the study drug compared to placebo. You can see that the subgroups really don't isolate a group that was responsive to the drug versus others. All of the points are to the right of the no effect line. The effect was really quite consistent about across all of these subgroups. That's true for vital capacity shown here.
It also is true for the ALS Functional Rating Scale shown here, It was also true for Muscle Strength Mega-Score. Again, quite, I think, amazing consistency of the data. With regard to safety, mortality was very rare in this study. There was one death in a patient on active treatment, and that patient was on placebo. There were two deaths in the four-week follow-up period. One was on placebo and one on the highest dose. Very rare occurrences. Serious adverse events were quite evenly distributed across the spectrum of placebo to all dose groups. Rare and not significant, not higher in treated patients. With respect to adverse events, both fatigue and nausea, although rare, had a dose-dependent increase in being reported as compared to placebo. The other clinical adverse events listed here don't have a clear trend like that.
In particular, it's probably worth pointing out dizziness, which was a big problem with tirasemtiv, seems to be not a problem at all with this drug. There were some laboratory events to be aware of. Cystatin C increased in a dose-dependent fashion, since we calculated GFR based on cystatin C, GFR decreased in a dose-dependent fashion as well. There was some tendency for transaminases to be increased in a dose-dependent fashion as well. Looking at those data in a little bit more detail, here's the average changes in GFR based on cystatin C over time for all three groups versus placebo.
You can see there's no change in the gray bars, there is a dose-dependent decline in GFR that reaches its maximum level quickly by two weeks, stays very constant during the course of the study, then when you withdraw the drug, comes back towards normal. The maximum in the group, the sort of biggest change is on the order of a 15% drop. This was noted. It doesn't seem to be severe. It wasn't associated with signs of renal toxicity, as Andy will discuss. The GFR dropped, but other renal markers didn't change. Although it's something to be aware of and was the biggest reason for withdrawal from study seems to be a modest effect. Transaminase elevations were more than five times the upper limit of normal for a few patients. I think there are six patients displayed here.
You could see that they could happen at various times during the study, then when study drug was withdrawn, the transaminases returned towards normal. Our summary is that the pre-specified analysis for the primary and secondary endpoints were not statistically significant at the criterion we set, the p-values approached significance for the first two, but not for strength. That patients on reldesemtiv declined in all measures less than placebo, that there was a trend towards the differences getting larger with time for both SVC and ALSFRS revised. Reldesemtiv, in terms of the estimate of the size of effect, showed clinically meaningful effects in reducing the decline of vital capacity, 27% at three months, ALSFRS 25% in three months, and in strength as well, which is 21%. When you combine the active groups and compare them to placebo.
The incidence of early treatment discontinuations, serious AEs, and clinical AEs were overall similar between the four groups. That there were elevations in transaminases and declines in eGFR that were dose-related, the GFR change was the leading cause of termination from study treatment. It's our view that these results support the continued investigation of reldesemtiv into a pivotal Phase III trial. Here are the investigators and members of the steering committee and the data safety monitoring committee. Thanks.
I think we're going to open it up now for Q&A, and it's going to be moderated by Andy. Some questions have already come in that I think, Andy, you're going to maybe start with, maybe we'll come to the questions here in the audience as well. Is that okay?
I'll start with a question maybe each to our two panelists open it up to the audience, I'll continue to ask questions to our panelists, it's intended to be interactive. Let me start with you, Jeremy. How can you put this data into context of phase II studies in patients with ALS that we've seen in the past 10, 20, even 30 years?
My involvement in ALS trials started with the riluzole study, I think this is the most positive and consistent demonstration of potential benefit that has been shown in any phase II study for any drug in ALS. That may be superseded by results that are going to be presented tomorrow for antisense treatment of a very small number of ALS treatments with a specific mutation. For sporadic ALS, the consistency and magnitude of these effects is greater in phase II than any other study that I'm aware of.
I'd like to bring people's attention to the fact that this was a broad population, some of whom had had ALS for up to four years, that would speak to the fact that they are slow progressers. We've actually, in this study, expanded dramatically. We've done the other end of the spectrum from the development of several other drugs, yet we have such strong results.
Angela, we've already heard from a few people who see the p value of 0.11 on the primary endpoint, and they don't seem to look any further. What do you have to say to them?
Look further. It's a very basic concept. This is a phase II study in which we didn't control for slow progressers in the placebo arm. We didn't control for a lot of other details. We wanted maximum exposure for ALS patients to this drug across the disease state, and yet we still saw this benefit. I would draw your eyes to both having moved the bar on the SVC and also to have moved the bar on the gross motor subset of the ALSFRS. For patients, this is a critical change or response.
Thanks. Are there questions from the floor? I think yeah. There's a microphone coming your way.
Jeremy, can you comment on the maintenance or the potential maintenance of the SVC after drug withdrawal? How should we be thinking about that mechanistically?
It's very interesting. It wasn't just for SVC, it was for strength and ALSFRS as well. It's not a new observation. This was seen in the signal that we saw in BENEFIT-ALS, where there was a strong SVC effect of the drug. It was also maintained after 28 days. It's not a pharmacologic effect in the sense that it's not due to maintenance of the drug or its metabolite in the system. I think there's 2 choices. It has to be some sort of long-term effect on muscle or of nerve. With respect to muscle, I think it's possible that if you make your muscles more useful, you use them more. There may very well be a conditioning effect that outlasts the actual drug that's in the body.
Alternatively, if you're not having to fire motor neurons as fast to get the level of activation that you need to make the movements that you require, you may be stressing them less. That may have an effect on long-term motor neuron survival as well. We don't have assays for either of those possibilities right now, but if you think of ways that you might look at this, there are some potential ways to look at it in the future.
I'll just add that not only have we seen this persistence of effect on vital capacity with tirasemtiv, and now here, vital capacity and ALSFRS-R with reldesemtiv in patients with ALS. Recently, we presented data from a study of reldesemtiv in older children and adults with spinal muscular atrophy. There, the improvement in six-minute walk distance that was evident after 4 and 8 weeks was also persistent 4 weeks after the last dose of study drug. We've seen it now in 2 different disease populations on at least 2 or possibly 3 different endpoints.
Thank you. Joe Pantginis from H.C. Wainwright Thank you very much for all the additional information. My question is really looking at 2 layers of perspective from the panel. Number 1, Dr. Shefner, you certainly touched on 1 of them. Number 1 is, for example, when you look at the 27% rate or slowing of the rate of decline, can you put this into real-world perspective about why this is meaningful to a patient? How does it help them? Secondly, as part of this perspective, you talked about comparing these data to other phase IIs in the past, but how does it compare to the edaravone data that led to its approval based on how broad this is?
Would you still go?
First, with respect to what do percentage points mean in terms of patients, you can ask that for any of these outcomes. I guess it's probably easiest to discuss it with vital capacity because respiratory failure is the main cause of death. If, in fact, you can slow the rate of progression in vital capacity by a quarter or more, because it does seem to increase with time. Even 25% means that you require non-invasive ventilation 25% less frequently, which for a disease with a survival on the order of 4 years is up to a year later. You potentially prolong survival by about that same amount. I think that level of prolongation of disease, if true, if we were speaking to a group of hematologist- oncologists about a cancer chemotherapy, that would be a home run.
With respect to predictors of death, vital capacity is the strongest. With respect to ALSFRS, it's an imperfect scale, but if you look at the individual values, changes of one point, many of them are incredibly significant. Going from intelligible speech to unintelligible speech is a one-point change. The scale drops by, on average, about one point per month overall. If you can take away a loss of one point every year, for example, that's a 10% chance. That's important. If you can change the level of disability by three points at the end of the year, that, depending on which particular points they are, can be incredibly meaningful. I think these are things that make a difference to ALS patients. With respect to what's the perspective compared to edaravone?
The edaravone.
Remember, edaravone was 6 months, and it was a 33% change in rate of progression in ALSFRS. We had a 25% change at 3 months. If you compare apples to apples, looking at the edaravone data at 3 months, there is about an 8% change in ALSFRS at that point. In terms of magnitude at that time, it's clearly greater. Is it likely that the effect will continue to diverge from placebo as the longer study progresses? That's something we'd have to see. Even at 25%, it's the same order of magnitude. To me, if you start adding 25% effects onto 25% effects, at some point you get a drug combination of real meaning for patients.
The other important issues to highlight when you attempt to compare apples to oranges is that the edaravone study took only the most ideal patients. They didn't take all comers in the phase III. They took a subset of patients who were two things. They were very early in the disease, and they had proven progression rates. They couldn't stay in that study unless they were actively progressing. With this particular program, we took everyone. Not everyone, but 4 years is a lot of everyone. Because of the way patients are entered into studies, we therefore took slower progressors. Even in the placebo arm, we took slower progressors, and they progressed more slowly than in other studies.
If you actually look at what patients will be taking a drug like this, we actually mirrored what patients will be most interested in taking a drug like this. The other thing to remember is edaravone was allowed in this study. This is in addition to real-world use of edaravone. It is not an edaravone naive or restricted population. We're not actually comparing to only placebo or placebo riluzole. Patients were allowed edaravone as well. I think that makes it a more complete perspective.
Hi, I'm Dave Liebowitz from Morgan Stanley, here for Jeff Hung. Specifically, given the results of this trial and the results of the tirasemtiv phase III from the past, when you look forward and envision a phase III trial for this program, what is this trial going to look like? What will the patients be? How might they be different? Which of the endpoints would you choose to select for that phase III program? Is 12 weeks the ideal duration, or might you consider looking at 24 weeks, such as other phase III programs?
Just I'm going to make one quick comment and let them talk about the planned study. The most striking thing about the tirasemtiv trial as compared to the reldesemtiv was the side effect profile and therefore dropouts due to side effects. That difference is not inconsequential. In fact, it was very consequential with tirasemtiv. The fact that reldesemtiv has an incredibly strong retention rate for an ALS trial, they stayed in the duration, really speaks to the tolerability of this drug and that we were able to actually see the benefit in all comers. We didn't have that big dropout that adversely affected the analysis of the tirasemtiv study. There are still people who take tirasemtiv, which speaks to if you could tolerate it. If you didn't have side effects, people are still very much invested in that drug as well.
Although reldesemtiv has a much more interesting and more tolerable profile.
In terms of design, I think we met in late April to look at these data for the first time. I think we've been more focused on looking at these data than designing the next study. I think clearly it's going to be longer. 6 months was the time for the primary outcome measure in Vitality. I think that's probably an appropriate time for that, although duration of effect beyond that, it's going to be important to look at as well. We chose vital capacity for the Vitality study based on what we saw with tirasemtiv, not necessarily because of our expectation that something is special about the diaphragm and couldn't be seen more generally with extremity function. I think, the real difference of this drug versus tirasemtiv is that we can really get a better picture of the range of effects, not colored by tolerability.
Since we saw a strong signal in ALSFRS, and it's already been an approvable endpoint, my tendency would be to go there first, but it's not something that we've decided. Beyond that, Angela's mentioned the sort of inclusivity of this patient population, and my own view is that you learn more if you look at a broader group of patients. I think we might want to be more careful to make sure that the patients that we select are less likely to be very slow progressors. There's a variety of ways to do that.
Angela, you've already spoken to the much better tolerability of reldesemtiv compared to tirasemtiv. Now we've touched on the fact that in this trial now, for the first time, we've seen a pretty clear effect on ALSFRS-R and vital capacity, whereas with tirasemtiv, we really saw only effects on vital capacity, and the ALSFRS-R was not so positively impacted. How could you explain that?
Well, certainly the ability of the patients to take the drug helps in your assessment long term. That's obvious. If you lose a significant number of patients to a secondary side effect, you struggle. What we see with this data, and is reported by the patients, is that they have a very clear feeling of their ability to continue to do their activities. Activities of daily living and the ALSFRS is a reflection of the ability to not lose ability. They maintain their activity. In many cases, they maintain their work schedule, they maintain their life schedule. The one thing that has evolved between the tirasemtiv development and this reldesemtiv trial is an evolution in our trial design, really drilling down on the subscores of the ALSFRS and even some other subscores that are relevant to the regulators and the patients.
An improvement or a lack of decline of your gross motor function has a huge impact. I think that we've got some very positive results to build on here across the board.
If I were to summarize, you're saying that better tolerability probably underlies an ability to see an effect on the ALSFRS-R that we didn't see before?
Absolutely.
Jason Butler, JMP Securities. First question, obviously, you've only had the data a short period of time, but have you had a chance to look at any responder analyses yet? Are there any subpopulations other than that you showed there that you thought about, for example, patients on edaravone, whether they're showing the same response, different responses? Second question, just on the renal safety, both from a mechanism of action and clearance perspective for the drug, any explanations for what might be happening there? Have you seen this previously in other with tirasemtiv or in the SMA study? Just from a practicing perspective, what would this kind of magnitude of change matter to patients? Are ALS patients at more risk of having renal insufficiency? Thanks.
I'll start with what we know about what is and isn't happening, and then I'll ask our panelists to comment on what they think the implications of that are for future studies. As we said several times, we've only had the data for a little while, but we have had the opportunity to consult a couple of expert nephrologists, and we will consult more. To answer one of your questions, we have seen this across different populations, in healthy volunteers, in older kids and adults with SMA and in ALS, and it seems to be a very consistent pattern. The input that we've gotten from our experts tells us more about what they think it's not than what they think it is. What they think it's not is pretty reassuring.
They point to the fairly rapid evolution, the fact that this effect on GFR appears to be fully evolved after two weeks is faster than one might expect for a drug-induced nephritis. Another thing that is very impressive to them, and we didn't show you the data, but the urinalyses are very clean. There were no casts in the urine. In 457 patients, the only casts in any of the urinalyses were at baseline and none after treatment. That's a clear marker of renal injury, cellular death, sediment in the urine, also red cells, white cells, very infrequent. That's very impressive to them, and I always reflect upon the fact that I think we look at urinalyses in clinical trials often as sort of a, "Oh, do we really need to do that?" It was very important to us here in this trial.
It does not appear to be a true toxicity. There is no evidence of inflammation. It doesn't appear to be renal damage. The fact that it recovers, if we follow patients or volunteers long enough, that it essentially recovers completely, also suggests that this isn't a toxicity to the kidney, but some pharmacodynamic effect that, to be honest, we don't understand. There are other drugs, notably ACE inhibitors, that disrupt intraglomerular hemodynamics, such that they really do affect a decline in glomerular filtration, but it recovers when you withdraw the drug. That might be what's happening here. That's one thing that's been hypothesized is that we have some effect on the intraglomerular pressures so that the filtration fraction falls, eGFR falls. It recovers when you stop the drug. What do you think this means for further development?
The kidneys are really far away from what I feel like I'm expert about. I think that it's a signal. I think that we're going to have to be careful, and we're going to have to have a lower limit, below which if somebody drops, they'll probably have to withdraw from study treatment in the future study. I think that if you look at that sort of a threshold that has been proposed to us, that would've been a very rare occurrence in this study.
Just comment, we had about six early terminations due to the drop in eGFRs.
With regard to your first question, which was any further subgroups, the one you mentioned, we did start to look at them. The effect is seen both in patients who were taking edaravone and those who weren't, and the general magnitude of the effect were similar in both cases. I think responder analyses are very interesting to look at, and we're going to do that. Beyond that, we haven't yet.
I'll just add again from our expert consultants, their advice was that a 25% decline in GFR was probably too stringent when you're starting with a population whose average glomerular filtration rate is over 100. They would be willing to treat patients until they fell below 60, especially because they're reassured that if you stop the drug, it will recover.
Yeah. The change in GFR would have no effect on the clinician's decision whether or not to start this drug unless you were dealing with a renally impaired patient. As you know, ALS patients are young, healthy, active, and typically have no significant other medical problems. They would be the last group that you would worry about what we're seeing so far.
Thank you for all the updates. The baseline characteristics all look pretty balanced, and I assume you controlled for some of them, the rest were handled by randomization. The one mismatch seemed to be bulbar onset for the higher group, the 450 milligram dose group. Your own subgroup analysis showed that that was the one group that didn't really respond at all. That makes sense, of course, clinically and other studies, the bulbar onset progresses faster and so forth. I wondered if there was a comment about the ability to measure the bulbar onset and the seal around the flow and things like that, and if there's any qualitative, quantitative measure of if that was. The sites look all very experienced and so forth, if there was some comment around that.
The second question is, do you have a baseline trajectory such that in post-hoc analysis you can look at the slow, fast progressors? Do you just have the inclusion criteria and cannot subgroup it out going forward, at least in the post-hoc?
So-
Let me comment just on the technique and the problem with bulbar patients in clinical trials where you have an SVC. Patients can have trouble with the test and still be in good functional condition. It's one of the tricks of actually dealing with bulbar onset patients in these ALS studies. We can't exclude them, they really do struggle with the seal, as you mentioned. It is the reason why you really should never only focus on an SVC in an ALS trial because of these effects. It would be very difficult, a rapidly progressive bulbar, it would be very difficult for any drug to change their ability to seal around the tool that we use. Go ahead, Jim.
With respect to that question, you're right. On SVC, there was essentially no signal. In terms of the other measurements, the bulbar site of onset didn't seem to have any effect on the subgroups. I would look at this in terms of consistency of subgroup changes. I'm not impressed by that. What my group does at Barrow is we manage the training for outcome measures for this study and lots of other studies. All of the points you made about difficulty that bulbar patients have in performing that test are certainly true and adds to the variability. The variability around the decline in SVC in this study was quite small. I think we did a good job of training sites. You had asked the question about can we somehow look at rates of progression.
We have no interim look, the way people have estimated rates of progression in prior studies is with this formula. You ask the patient what's the month of your first symptom, and you look at their ALS Functional Rating Scale at the time that you're examining them, and you estimate the slope from 48, which is the highest value of the ALSFRS-R at the date of their first symptom to what they are now, and you can draw a slope between that. You can use that kind of calculation to estimate rate of progression. If you look at that last set of comparisons of subgroups, that's exactly what we did.
Just to emphasize, these are not based on the rate of progression during the study, the rate of progression prior to enrollment of the study, and then they're
Stratified into 3 tertiles. You see it makes sense if your ALSFRS, as Angela says, hasn't declined over the last 24 or 48 months by very much, over the next three months in which they're enrolled in this trial, it's not likely to decline very much, and you wouldn't see much of a signal as a consequence.
All right. Gil Blum for Chad Messer, Needham & Company. Just a quick question about the dose for phase III, given the ongoing phase I in healthy volunteers and the effects you guys seen on eGFR.
Well, we've got to figure that out. I think for the most part, if you look at the efficacy and relative instance of side effects in this study, you might be tempted to go with the middle dose. I'm actually also very interested in the fact that if you look at all of the doses that we use, we're nowhere near the top of the pharmacodynamic range. If we could find a way to safely manage the adverse events, I think looking higher at some point in the development process of this drug would be pretty important.
It just begets to ask the question, is looking at healthy volunteers relevant at this point?
I'm sorry.
Is looking at healthy volunteers still relevant at this point?
I'm sorry. I think it's relevant in terms of bioavailability. We've seen no differences between what you can determine there and what you determine for the most part in patients with dialysis. It's a place to begin looking. It's a lot faster and easier to study bioavailability of different drug formulations in healthy volunteers. Yeah, the definitive experiment then is going to be in the population of interest.
I think the current study was intended to examine a range of doses, 150, 300, and 450. It was intended to inform selection of dose. I think the selection of dose will be balanced by what we think of as the benefit versus potential safety of the doses that we examined. Probably in this particular study, the balance was in the middle there around 300. As Jeremy said, there's rationale to think about, at some point, exploring a higher dose if we can.
Any other questions from the floor? Any other thoughts from our two expert panelists that they want to say that they haven't had a chance to say yet?
Just that it should be clear, I think these are really exciting results. I am anxious to see a phase III study go forward that really tests the hypothesis.
This is an oral medication with impressive results that is well-tolerated, and those are the three distinguishing features. We've tried it in all comers. I think we have something to run with now.
I want to thank our panelists. I want to thank my colleagues for those presentations and that discussion. I'd like to now end the program by putting some more context around these data and our corporate outlook. I'll first start by coming back to one of the comments I made in my introductory comments. This is a webcast where primarily the audience is investors. Admittedly, there's already being demonstrated some skepticism around these results in communications that have been put forward regarding this trial. You will not hear Cytokinetics underscore this data as having met the pre-specified efficacy analysis. We laid out a pre-specified efficacy analysis around which we achieved not conventional statistical significance, but certainly approaching that and borderline statistical significance with a P equals 0.11.
However, I hope that what you've heard today would underscore that these are impressive data in the context of the development of medicines for ALS, and especially as was intended by the conduct of a phase II study to inform the potential progression to phase III. As I said, why do you do a phase II study? You do a phase II study to understand whether you have a high probability of achieving success in a phase III study, where there the bar should be 0.05 or lower. What do you learn from a phase II study that enables you to enrich and amplify a response that would be observed in a phase III trial as you apply learnings from the phase II study? We do have learnings from this study.
We learned that with reldesemtiv, we've advanced the field beyond that which we achieved with tirasemtiv, most especially with regard to tolerability. Underlying that is now evidence to support the mechanistic hypothesis that reldesemtiv is affecting not only slow vital capacity in a large clinically meaningful way that seems to be increasing over time and persistent, but also absent intolerable effects. We're seeing evidence of effect there too on ALSFRS, which is a registratable endpoint as demonstrated by a recent approval. We see these data as I hope you've also heard from our panelists, and as I can tell you, was unanimous when these data were shared with investigators. We see these data as positive and lending support for progression of reldesemtiv to a next-level clinical trial. Certainly, there's more analyses we need to do.
Certainly, we need to understand if there's ways maybe by eliminating slow progressers or otherwise, taking a look at those subgroups, we may be able to augment the effect, and certainly demonstrate durability of effect, and that's a task upon us that we relish. I look forward to being able to provide updates to how we may now move forward with reldesemtiv in next-level clinical trials, underscoring that that's a conversation we also need to have with our corporate partner at Astellas. We and Astellas have only had a matter of a few days to review these data, and there's a lot more we need to do together and with regulatory authorities in terms of understanding around which these data may lend support for progression into a specific phase III trial. Underscoring that we need to then know what does that trial look like, endpoints, timelines, and budget.
What I can say is that it's highly unlikely that any such phase III trial would occur in 2019. Just knowing what the task in front of us are, it's more likely that that would be underway, we would hope, in 2020. Given that, and given the pipeline that I shared with you before, I think as we point forward to milestones in 2019. In February, we laid out for investors what would be those key milestones around which our company and progress and plans would pivot for the second half of the year and through 2020. We're now three for three on those milestones. Earlier this year, we announced that the GALACTIC study of omecamtiv mecarbil had successfully passed through an interim futility analysis, enabling us to proceed to completing enrollment of that study.
We'll have our earnings call later this week, and we'll provide updates with regard to GALACTIC, but suffice it to say that that study is proceeding to conclusion of enrollment, as is the METEORIC study now enrolling patients, and both of those trials would be expected to deliver data in 2021 or possibly sooner. Put these data into further context, I also had shared with you that we would be looking to CK 274, our cardiac myosin inhibitor in phase I to inform potential progression to phase II. That phase I study is concluding. We'll have more to say about it also later this week on our earnings call, but we are already proceeding to plan for the design of a phase II program in patients with hypertrophic cardiomyopathy for CK 274.
That's also occurring as we and Amgen are conducting a phase I study with regard to AMG 594, our cardiac troponin activator for the potential treatment of heart failure. I want to thank you all for your interest and attention to these data. I hope we set a proper context for reldesemtiv and the development of a program directed to patients with ALS, and also how that fits into the overall corporate picture and corporate development for Cytokinetics. With that, we'll conclude this program. Thank you for joining us either here in person or on the webcast. We invite you to ask us questions and also engage us in comments as we can further elaborate on this for your benefit. Thank you very much.