Fulcrum Therapeutics, Inc. (FULC) Earnings Call Transcript & Summary

Welcome to Fulcrum Therapeutics conference call. [Operator Instructions] I would now like to turn the call over to Christi Waarich, Director of Investor Relations and Corporate Communications at Fulcrum. Please proceed.

Thank you, operator. Good morning, and thank you for joining us today to review the results from the ReDUX4 Trial with Losmapimod in FSHD. Earlier this morning, we issued a press release on these results, which you can access in the Investor Relations section of our website at fulcrumtx.com. Please be reminded that remarks made during this call may contain forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. These may include statements about our future expectations and plans, clinical development timelines and financial projections. While these forward-looking statements represent our views as of today, they should not be relied upon as representing our views in the future. We may update these statements in the future, but we are not taking on an obligation to do so. Please refer to our most recent filings with the Securities and Exchange Commission for a discussion on certain risks and uncertainties associated with our business. With me on today's call are Bryan Stuart, President and CEO; Judith Dunn, President, Research and Development; and Chris Morabito, Chief Medical Officer. Chris Moxham, Chief Scientific Officer; and Michelle Mellion, the [indiscernible] responsible for the study will also participate during the Q&A portion of the call. Let me quickly run through this morning's agenda. Bryan will begin with opening remarks. Chris Morabito will review background on FSHD and our approach to the ReDUX4 trial, and Judy will provide a review of the data. We'll then open the call for Q&A. With that, it's my pleasure to turn the call over to Bryan. Bryan?

Thank you, Christi. Good morning, everyone, and thank you for joining us today. Today's ReDUX4 results are extremely exciting. Losmapimod demonstrated slow disease progression and improved function in patients with FSHD. These are unprecedented results in the severe and progressive form of muscular dystrophy, which lacks any current therapeutic alternatives. As a reminder, our hypothesis was that losmapimod could reduce FSHD-related muscle degeneration through reduction of DUX4, leading to benefits in muscle health, function and patient-reported outcomes. Our goal was to develop a disease-modifying therapy that stops or slows FSHD progression. These results strongly support our belief that losmapimod has the potential to be a transformative therapy for FSHD patients. While the study did not meet the primary endpoint of a reduction in the experimental biomarker DUX4-driven gene expression due to several technical and biological variables that diminished our ability to detect DUX4 changes. ReDUX4 did demonstrate the downstream functional benefits we would expect from reduced DUX4 that are the ultimate goal of a potential therapy. We observed statistically significant and clinically meaningful changes in muscle health, muscle function and patient recognized benefit, which we believe are the most important outcomes for FSHD patients. Evidence of losmapimod's impact on muscle health was reflected via MRI, measure of reduced fat infiltration. Patients treated with losmapimod showed better functional scores versus placebo as measured by reachable workspace in dynamometry and patients reported feeling better when treated with losmapimod compared to placebo through the patient global impression of change survey. Consistent with prior studies in more than 3,600 subjects, losmapimod exhibited favorable safety and tolerability in ReDUX4. With these promising data in hand, we plan to meet with the FDA in the second half of 2021 to discuss potential next steps, including the potential for accelerated approval. We are also -- continue to be very excited by the productivity of our FulcrumSeek platform, which identified losmapimod as a treatment for FSHD as well as FTX-6058 for selecting hemoglobinopathies. As a reminder, our product engine is designed to identify targets that modulate the root cause of genetically defined diseases. Our patient-centric approach utilizes cellular models that embodies a proper disease context and that can be interrogated with our proprietary small molecule and CRISPR libraries. We are able to survey the transcriptome, cellular biology pathways and networks at scale through the use of high content technologies and machine learning. This approach allows us to identify highly derisked targets that address the root cause drivers of disease biology and create the opportunity to quickly identify and develop transformative therapies for severe diseases. Chris Morabito and Judy Dunn will take you through today's ReDUX4 results, which is the only clinical program for FSHD. I'd like to thank all of the participants, families, caregivers, physicians and investigators for continuing this trial through challenging times. We are very grateful for their time and their courage. I'll now turn the call over to Chris Morabito, our Chief Medical Officer. Chris?

Thanks, Bryan. It's really a very exciting time to have joined Fulcrum and these results are remarkable. I'll start by walking through some background on FSHD and the rationale and designed for ReDUX4. FSHD is a rare, serious, progressive and disabling disease characterized by severe muscular degeneration that occurs as skeletal muscle is replaced by fat. The genetic root cause of FSHD is the apparent expression of DUX4 in skeletal muscle, which leads to progressive muscle cell death and fat infiltration resulting in muscle weakness and eventual loss of function. FSHD initially affects movement of the face and eventually progresses to the trunk and legs. We estimate there are approximately 16,000 to 38,000 patients in the U.S. and similar incidents worldwide, leading to an estimated global patient population of 300,000 to 780,000. And there are currently no approved therapies for FSHD. We were very pleased that the FDA held a Patient-Focused Drug Development day last year with the FSHD community. As you can see on slide 10, regulators were able to hear directly from patients, family members and caregivers who shared information about the physical and emotional suffering, stress and spare caused by FSHD and their urgent desire to have a therapy that can slow the progression of the disease. The results we are reporting today increase our enthusiasm and belief that losmapimod could have a meaningful impact on patients. As Bryan mentioned, Fulcrum's proprietary discovery engine, FulcrumSeek identified p38 MAP kinase as a target with the potential to prevent apparent DUX4 activity in skeletal muscles and stop or prevent functional impairment. Losmapimod as selective p38a/ß MAP kinase inhibitor and preclinical data shown here on Slide 11 demonstrate losmapimod's proof of mechanism. On the left, an FSHD myotubes, losmapimod reduced p38 MAP kinase activity, DUX4 expression, DUX4 activity and cell death. The right panel shows myotubes from 11 people with FSHD. This preclinical study clearly demonstrates losmapimod's ability to decrease DUX4-driven gene expression. As Bryan mentioned, the hypothesis that guided Fulcrum's ReDUX4 trial was that by decreasing DUX4-driven gene expression and activity, losmapimod will slow or stop muscle apoptosis and fat infiltration. The slowing of FSHD disease progression would be demonstrated by preservation of muscle health, function and patient benefit. ReDUX4 was a randomized, double-blind placebo-controlled clinical trial designed to assess the efficacy and safety of losmapimod and people living with FSHD. After a 28-day screening period, patients were randomized 1:1 to receive either dose losmapimod 15-milligram BID or matching placebo. As you may recall, the trial was initially designed to last only 24 weeks, but due to the COVID-19 pandemic and to ensure collection of key data at trial sites, the trial was extended to 48 weeks. As shown on Slide 13, DUX4 activity was measured using the muscle needle biopsy at baseline, and then at either week 16 or week 36, and while measures of PK, safety, muscle health and clinical outcome assessments were performed through week 48. Following the completion of the trial, 99% of eligible participants elected to continue to the open-label extension, whether previously on losmapimod or placebo in the blinded portion. Enrollment in the open-label extension was allowed for those participants who completed 24 weeks of blinded treatment prior to the implementation of the COVID-19-related amendment or who completed 48 weeks after the amendment. Statistically, hypothesis and testing occurred on the primary and key secondary endpoints, all analyses presented are prespecified, except where noted. Hierarchical assignments for exploratory analysis were not prespecified. As I mentioned earlier, apparent DUX4 expression in skeletal muscle is the root cause of FSHD, leading to myofiber death and fat replacements. We focused on DUX4-driven gene expression as a primary endpoint because we believe that we'd be able to translate the robust preclinical effects of losmapimod to a 24-week clinical trial. To assess DUX4-driven gene expression in affected muscles, a novel experimental assay using quantitative PCR was validated and served as a measurement for the primary end point. Baseline muscle needle biopsies were initially performed on STIR positive muscles identified via MRI under lower extremities of the participants. STIR positive regions of muscle were selected as these regions are more likely to express DUX4. Our repeat needle biopsy in close proximity to the original -- initial biopsy site was conducted at week 16 or 36, DUX4-driven gene expression in each biopsy was reported as a quantitative average of the expression of 6 known DUX4 target genes. On Slide 15, you can see fat infiltrating muscle, which leads to a loss of function. In ReDUX4, we evaluated skeletal muscle quality using a novel whole-body MRI. This approach captures the intra and inter muscular heterogeneity of the disease and overcomes the limitations of older traditional MRI measures by allowing for both a holistic and quantitative picture of muscle health. There are 3 primary measures that were derived from whole-body MRI shown here. They include a measure of muscle fat infiltration or MFI, which is an indicator of muscle quality, a measure of muscle fat fraction MFF, which is a measure of overall fattiness of the muscle, and last, lean muscle volume or LMV, which is a measurement of the amount of lean or contractile muscle tissue in the muscle. Integrating these various measures of skeletal muscle as determined by MRI. We classify muscles in a prespecified manner into 3 categories shown here on Slide 16. Those muscles that were classified as A are normal muscles and do not appear to be affected by disease. B muscles are classified as intermediate. They haven't been so severely fat replaced that they've lost all function, but they are the most likely to progress, and C muscles, which are end-stage have lost, likely most, if not all, function. Beyond the structural assessments, we also evaluated multiple relevant clinical outcome assessments, or COAs. Importantly, these measures are relevant to people living with FSHD and the progression of their disease. These included the traditional, Timed Up and Go, in which a patient starts from a stated position stands up, walks 3 meters, turns around then comes back and sits down. We also developed FSHD Timed Up and Go. In contrast to the classic Timed Up and Go, here, patients start from a supine or laying down position, sit up, stand, walk 3 meters, come back and return to that in supine position. We use classic dynamometry to measure muscle strength, we used an endpoint called reachable workspace, which is a very relevant clinical outcome assessment because it allows one to quantitatively measure upper extremity movement along 5 different domains and has been shown to correlate closely with activities of daily living. And we use a quantitative assessment of motor function abilities. Importantly, we also wanted to hear from patients and utilize patient-reported outcomes as part of this trial. This included the patient's global impression of change in which participants were asked to respond at the end of the trial to the query shown here, since the start of the study, my overall status is, they then selected a score that best described their perceived overall status. In addition, we also deployed another patient-reported outcome called the FSHD Health Index. In summary, ReDUX4 was designed to capture a wide range of FSHD disease progression in addition to DUX4-driven gene expression, measured by repeat muscle needle biopsy. Importantly, beyond this experimental biomarker, the trial interrogated the safety and tolerability as well as the pharmacokinetics and target engagement of losmapimod, structural and functional clinical and patient-reported outcome assessments. Bryan has already told you that losmapimod has favorable impacts on disease progression across structural, functional and patient-reported endpoints despite not demonstrating a difference in DUX4-driven gene expression versus placebo. These results that Judy will now go through with you show in detail the positive benefit risk of losmapimod in FSHD. I do want to remind you that the analyses and results presented now are all prespecified according to the statistical analysis plan, except where noted, because the p-value for the primary endpoint was not less than 0.05, all statistically significant differences are called nominal. Judy?

Thanks, Chris. I'm also pleased to have joined Fulcrum at such an important time. I'm excited to share today's data. Before jumping into the data, you can see there was excellent randomization and participation across ReDUX4 with 40 participants in both the losmapimod and placebo arms. There were no differences in baseline characteristics between the groups, and the enrolled population is a representative sample of people living with FSHD. There were 3 discontinuations. One participant was receiving losmapimod, 2 were receiving placebo. None of these discontinuations were related to study drug. Here, you can see blood concentrations of losmapimod, they were consistent with previous studies as expected, and muscle exposures were also within the expected range. Additionally, sustained target engagement was observed with losmapimod in ReDUX4. We have spent our time so far this morning, carefully outlining the trial design and the methodology. Now I would like to share the data. This forest plot outlines the outcomes on all endpoints in the trial. I'll walk through many of these individually, but at a high level, it's evident from this graphic that losmapimod had a positive impact on multiple outcomes. These include muscle health, muscle strength, function, and importantly, patient-recognized benefit. Let's step through the data now in more detail. We selected the novel biomarker of DUX4-driven gene expression in muscle needle biopsies as the primary endpoint for ReDUX4. This endpoint had never been assessed in an interventional clinical trial. We selected it because we hypothesized that it's the most proximal measure of disease change prior to structural or functional changes. To determine the sample size of this study, we relied on preclinical in vitro and in vivo work as no human data had existed. In ReDUX4, change from baseline DUX4 gene expression was assessed at week 16 or week 36. Combined changes from baseline were not observed in either group. Pool changes from baseline at either group were not observed and the difference between losmapimod and placebo was not demonstrated. Thus, the primary endpoint was not met. Subgroup analysis by quartile of DUX4-driven gene expression also showed no difference and there was no detectable difference between change from baseline levels at week 16 and week 36. We've shown previously that losmapimod reduces DUX4. We've also developed and validated a quantitative PCR assay to measure with high precision, the relative abundance of selected DUX4 gene targets, yet, the ability to detect changes in DUX4-driven gene expression was confounded by significant variability across biopsies at baseline and upon repeat biopsy in both the placebo and the losmapimod groups. So why was the translation to the clinic challenging? We believe there were several contributing factors. The stochastic and dynamic nature of DUX4 expression combined with the scarcity of micronuclei shown to express DUX4 as well as the fact that a needle biopsy samples relatively small muscle segment from a heterogeneous cell environment. Thus, we learned that sampling a dynamic, scarce signal and a heterogeneous cell population with the needle biopsy was not robust enough to detect any changes over time. Inter and intrapatient heterogeneity likely introduced additional variability compounded by the relative in precision in the biopsy procedure across multiple clinical trial sites. So now I'll walk you through the clinical trial results of the patient specified, prespecified secondary and exploratory endpoints. The first set of results are from the MRI analysis assessed after 48 weeks. These analyses present holistic structural assessments of muscle. Focusing first on intermediate muscles. Those are most likely to change. Losmapimod-treated participants showed significantly less muscle fat infiltration compared to placebo. MFI on the left, sensitively detects changes within muscle, and prior studies have shown that changes in muscle fat measured by MRI are correlated with changes in function. Importantly, placebo-treated patients had approximately 1.5% increase in MFI compared to losmapimod. Data in our biomarker study had previously demonstrated a meaningful correlation between MFI and various measures of function. There was a trend towards less progression in muscle fat fraction and losmapimod-treated participants compared to placebo, and no difference in the muscle volume, which we believe would take additional time to demonstrate. In this time period, detectable MFI changes may be related to the sensitivity and proximity of this measure to the earliest detectable disease pathology. Observable changes to MFF and LMV may take more time to be observed. There are meaningful correlations among MRI changes and changes in clinical outcomes assessed in FSHD. These data will be presented during a poster session at the IRC conference. With MRI, we also looked at normal appearing or A muscles. As you can see, consistent with our observation in the intermediate muscles, we also observed trends of less progression in these normal appearing muscles and the muscle A group had no evidence of progression at all. These muscles appear to be preserved with losmapimod over 48 weeks. Among placebo-treated participants, there was a trend towards more fat infiltration muscle, which was not seen in losmapimod-treated participants. It's striking that normal appearing muscles progress over this time period in the placebo group, and that losmapimod slowed this progression. In total, these MRI data show there is not only less fatty infiltration in muscles, but also apparent preservation of normal muscles in patients treated with losmapimod compared to placebo. The following data will extrapolate the structural findings to function. Dynamometry is a quantitative assessment of upper and lower limb strength, fat infiltration to muscle results in progressive decrease in strength, ultimately resulting in decreased muscle function. Over 48 weeks in a post-hoc analysis, losmapimod slowed loss of muscle strength and in some cases, improved strength compared to baseline. Study participants in the placebo group lost about 15% of strength in the shoulder and ankle after 48 weeks. Study participants in losmapimod group showed trends of slower progression or even improvements in the strength of these muscle groups. Other muscle groups that were evaluated showed no difference between placebo and losmapimod. These data are important because they show directly that strength in these muscle groups that are involved with activities of daily living are less affected by progressive pathology in participants in the losmapimod group compared to placebo. Reachable workspace is a measure of muscle function. Reachable Workspace provides quantitative data that correlate well with changes in activities of daily living and the measure is not prone to bias. It's an important measure of disease progression in FSHD as the disease tends to progress from the upper body down and loss of shoulder movement leads to loss of mobility and impact on activities of daily living. Reachable workspace uses a centrally read evaluation of an individual's global upper extremity range of motion and strength, including the shoulder girdle and arm. ReDUX4 used a quantitative video surveillance of muscle function to measure deterioration of reachable workspace. Losmapimod resulted in the slowing of progression and even improvement on multiple reachable workspace metrics compared to placebo. Reachable workspace assessments were performed with and without weight to further stress the systems and provide more robust assessments of function. As shown here, there are clinically relevant differences in reachable workspace, which can be impactful on activities such as independent self-care. Study participants in the placebo group lost about 2% to 4% of total surface area. While participants in the losmapimod group showed trends of slower disease progression and even improvements up to 1.5% in the accessible surface area when using a 500-gram weight on their wrists compared to placebo. More details of that reachable workspace are presented on the next slide. If we look at changes of relative surface area by domain, the losmapimod arm showed either no progression from baseline or improvements across multiple domains, particularly in 1 and 3, which assess upper regions of the reachable workspace compared to placebo and saw a general progressive loss of relative reachable service area. Prior studies have shown that Q1 and 3 are the regions that are most likely to change. Combined with the shoulder of doctor dynometry data, these results suggest that losmapimod results in positive impacts on upper extremity strength and function. These are especially important regions for people living with FSHD because they translate to activities of daily living, such as personal hygiene, including brushing hair or teeth and reaching for things on shelves. From the dynamometry data, which show differences in ankle dorsiflexors strength favoring losmapimod, we would expect to see changes in the Timed Up and Go and indeed using the classic Timed Up and Go, placebo-treated patients showed a trend in increased TUG time, representing progression compared to losmapimod. The data showing changes in Timed Up and Go are noted here. There's an increase of about 3/4 of the second in the placebo group and an increase of less than 1/4 of a second in losmapimod group. This difference of 1/2 of a second has been shown to be clinically meaningful in other muscular dystrophy studies. Taking this result back to the MRI data, this difference in magnitude approximates what we predict based on differences in muscle fat infiltration assessed by the MRI at 48 weeks. Now we'll move to a more global assessment of function. Patients' Global Impression of Change, or PGIC, evaluates how participants are feeling by asking them to score change from baseline assessments of status using 7-point scale noted here. This is a common tool in clinical trials to assess treatment effects. Here, we note that after 48 weeks, study participants who received losmapimod reported a significant and meaningful improvement of 0.6 points in impression of change scores relative to placebo. Now we'll look more closely at this relevant result, the change in PGIC. As noted here, losmapimod-treated participants showed improvements not seen with placebo. On the left in green, more participants improved versus placebo at 48 weeks. Further, fewer losmapimod participants reported worsening compared to placebo at week 48 shown here in purple. In total, as stated earlier, there was a clinically meaningful difference of 0.6 points between the groups, which favored losmapimod. These data reflect a drug effect making losmapimod a first treatment to demonstrate meaningful clinical benefit in FSHD. Finally, to wrap up the results. We did not detect relevant changes in the FSHD TUG, the motor function measure or the FSHD high assessment, as noted by the placebo results, which showed no progression. These measures appear not to be sensitive to change over the 48-week time period. Again, these are summary data from ReDUX4 -- this forest plot shows that across the board, we have demonstrated evidence of impact on multiple dimensions of disease progression. We see favorable impacts on muscle structure using innovative tools that sensitively detect change through the whole-body imaging. We see favorable impacts on function across multiple clinical outcome assessments, including reachable workspace, Timed Up and Go and dynamometry and importantly, favorable impact on quality of life as measured by the PGIC. The wealth of evidence supports DUX4 hypothesis in FSHD, and we know that losmapimod has potent effects on DUX4 expression in clinical experiments. This study demonstrated the challenges of measuring DUX4-driven gene expression by muscle biopsies in patients. And here, its ability to correlate with treatment-related changes in functional or clinical assessments demonstrated by losmapimod. We've also been able to advance the understanding of tools that may be used clinically to monitor disease progression and perhaps lead to further innovations. Finally, losmapimod continues to demonstrate favorable safety and tolerability. Previous data from over 3,500 people exposed to losmapimod across almost a dozen disorders, generally showed no clinically significant differences in safety and tolerability compared with placebo. And these findings now extend to ReDUX4. The majority of events in ReDUX4 were rated as mild or moderate, and there were no drug-related treatment emergent adverse events that caused study discontinuation or withdrawal. The data reported today exceeded our expectations. We're encouraged by these results, benefiting patients on several different outcome measures. We believe these positive results are unprecedented in any muscular dystrophy. We're committed to advancing losmapimod for the treatment of FSHD and plan to meet with regulators in the second half of the year to determine the best path forward for this program, including the potential for accelerated approval. To close, I'd like to thank my colleagues for their diligence and persistence toward our mission of helping patients. We want to thank our clinical trial investigators and sites for their guidance and support. And most importantly, we want to thank the participants and their families for helping us advance this important program. We can now open the call to Q&A. Operator?

[Operator Instructions] Our first question comes from Ted Tenthoff with Piper Sandler.

Congratulations on unprecedented clinical benefit in these patients who have great medical need. I wanted to dig in a little bit more with respect to the muscle fat infiltration and this differentiation between normal tissue and diseased or at-risk tissue. How well validated is that differentiation? And is this something that the FDA and/or community understand? And then I have a quick follow-up.

Yes. Ted, thanks for the question. Let me turn it over to Chris Morabito to speak to that.

Yes, this is a validated categorization. We're the first people to use the terms A, B and C, but it has been documented in previous research that the amount of muscle fat infiltration and the amount of muscle fat fraction do contribute to the understanding of those muscles that are likely to progress versus those that are not. So we sat down and talked with our colleagues in the investigator space and asked the question, how would we categorize this in a clinical trial and came up with these assessments of categorizations of normal, A, intermediate B, and end stage, C. The data before we started the study suggests that those muscles that are categorized as intermediate and specifically with parameters involving MFI and MFF are the ones that are more likely to progress. They have the higher probability of progressing in FSHD, and those muscles that have low muscle fat infiltration less than 10% are those that are normal and healthy appearing are less likely to progress overall but can progress in this disease. And then finally, the end-stage piece, these are muscles that have a very significant amount of muscle fat fraction and they are essentially end-stage. They have no function left. If they have some, it's incredibly small and essentially are significantly remodeled to a point where they will not contribute to the working of the muscle.

Right. And the diagram for the pictures that you included were really helpful to see that differentiation. Quick question follow-up on reachable space. Is it similarly -- is it traditional to add weight in that exercise. And does that sort of exacerbate any -- or make it more of a real-world setting?

Yes, absolutely. I'll turn it back over to Chris and we can speak to that, Ted, as well as the work that has been done by Jay Han and others and has previously been published.

So reachable workspace is what we view as a clinical outcome measure. It correlates quite well with activities in daily living. But I'll answer your first question first and then get to the ADL aspect second. So the weights are added to reachable workspace assessments in order to improve the robustness of the assay, they increase the sensitivity of the test by stressing those muscle systems that are involved. And as you know, we look at both sides here, left and right, dominant nondominant to give us a better comprehensive assessment about what the total surface area is in 3 dimensions that one can reach to. So Han and others have spent a lot of time thinking deeply about what are the correlations between changes in reachable workspace and changes in activities of daily living. And there are very good data that show direct correlations between those changes and those changes in ADLs, specifically in patients with degenerative diseases.

Our next question comes from Joseph Schwartz with SVB Leerink.

Congratulations on an impressive array of secondary endpoint data. It all seems to consistently point in the same direction. And it seems like it could open the door to more analyses to look for patterns and correlations within and across patients and endpoints and perhaps tease out more insight about how the drug could be helping and substantiate its effects. So I was just wondering what you might have planned in this regard going forward. Can you talk about whether you have any more analyses planned and what those might be?

Yes. So thanks for the question, Joe. I'll turn it over to Chris to address that. And one thing, Joe, as you were alluding to, and I think one of the elements that we're particularly enthusiastic about is obviously, the consistency that we saw here, as you referenced in the secondary and exploratory endpoints. So within the course of a 48-week trial, to be able to see these changes in muscle health in function and patient-reported outcomes. And that consistency between the 3 was obviously something we were very excited about and really did exceed expectations. But I'll let Chris speak to additional analyses.

Yes. So carrying on from that consistency theme, there's already a visible consistency among the clinical outcome assessments. Judy referred to the reachable workspace and the dynamometry results, we saw changes in shoulder muscle, shoulder abductor strength by dynamometry, and we saw changes in the reachable workspace. And you put those 2 together and it gives you a sense already about how these clinical outcome assessments work together to give us a sense about what is actually happening to disease progression with people not on losmapimod, and then loss of disease progression or prevention of disease progression and in fact, improvements when you look at dynamometry for people who are on losmapimod. So we're going beyond just those observations. We're digging very deeply into the data, and we're looking specifically at relationships among all of the endpoints to give us a sense about what our responder variables and ultimately, how do these responder variables link to one another. This is called predictive modeling, it's work that's been kicked off and ongoing. And hopefully, soon, we'll have a more comprehensive analyses to share.

Great. And then you've obviously put a ton of thought into designing the study and executing it. So now that you have this data, I was wondering with the benefit of this information, what would a confirmatory trial look like in your view, if the FDA would have asked for one, either in advance of approval or to substantiate an accelerated approval as a confirmatory study. Now that you've generated all this data? Is there any -- are there any endpoints in particular that you think would serve as a better endpoint than DUX4 gene expression, and would you lean more towards a functional endpoint as primary now? Can you give us some insight into why DUX4 for gene expression was selected previously and how you're thinking about what might be the best way to design a study in this setting now?

Yes. Sure, Joe. Why don't we -- maybe we'll split that up into 2 questions, and I can look to Chris to address here. But firstly, why did we select DUX4-driven gene expression as the primary endpoint in this trial? And would we utilize it again? And then maybe secondly, as we think about our interactions with the regulators, and as you mentioned, Joe, we do anticipate going to the FDA, talking to them about what next steps are, whether that's the potential for accelerated approval or an additional trial. At a high level, obviously, we're extremely enthusiastic that we were able to demonstrate efficacy and tolerability on multiple efficacy measures within this time period, which we think positions us very well for those conversations. But let me turn it over to Chris.

Thanks, Bryan. Thanks, Joe, for the questions. So FSHD is a disease driven by a ReDUX4 expression, and people who've this disease have high levels of DUX4. We know from not only FulcrumSeek, but also our preclinical data and preclinical data generated by labs external to ours, that losmapimod very selective p38a/ß MAP kinase inhibitor has pretty potent effects on reducing DUX4 and DUX4-driven gene expression. And also, as I pointed out in my presentation piece, things like cell death and other in vitro assessments of actual pharmacodynamic activity. So we're pretty confident as we move this program along that this is a drug that would have effects on DUX4. We designed the study in -- for Phase II to see whether those effects that we saw in the lab would translate into a human clinical trial, believing, as the data show, that by changing DUX4 expression in humans who have this disease we'd be able to have long downstream consequences on disease progression. So we designed a 24-week trial to assess changes in DUX4 at 16 weeks. We powered the study based on in vitro data, Chris Moxham and his team on the research side, spent a lot of time developing a very sensitive quantitative PCR analysis to look at DUX4-driven gene transcripts to help us quantify those differences in the course of this trial. When we designed the study, we wanted to look to see what else this can do, whether we could start to see any changes downstream in patients that would be related to changes in DUX4-driven gene expression in their disease. So we included these clinical outcome assessments, so on and so forth. And the results are, as you see, we were unable to detect changes in DUX4 mostly because of the heterogeneity and the variability associated with going into a human study, a patient study, but importantly, we saw these pretty profound effects on disease progression. So that translates into the question about what the Phase III study would do -- would look at. Phase III studies of confirmatory trials are meant to look at clinical outcomes as primary endpoints, they're not designed here to look at biomarkers. So we would absolutely not use DUX4-driven gene expression as primary endpoint here. We will look at -- and I'll give you an example here, we'll look at reachable workspace, which is a clinical outcome assessment known to correlate well with activities of daily living and quality of life. And frankly, we know that this drug has effects on reachable workspace that are pretty profound and meaningful for patients living with this disease. We could look at things like the patient-reported outcome, PGIC or other patient-reported outcomes that would be sensitive to change in the study to provide meaningful representations of how a patient feels and functions, which ultimately is the approvable guidelines set by the U.S. regulators. So that's how we were thinking about the Phase III study. Obviously, as Bryan said, we're -- and Judy said, we're looking to meet with regulators coming up, and we'll have conversations with them about how to define this.

Our next question comes from Matthew Harrison with Morgan Stanley.

This is Kostas on for Matthew. Two questions from us. The first one is around your thoughts around what is considered clinically meaningful on the muscle changes?

I'm sorry, Kostas. Can you repeat the question, clinically meaningful on the muscle changes?

Correct.

Sure. So why don't I turn it over to Michelle Mellion, who can speak to that.

Thank you for the question. So the changes that we're seeing on dynamometry have really not been seen before in other clinical trials with FSHD. We see clinically, clinical improvement, especially in shoulder abduction as well as in ankle dorsiflexion, which certainly translates to meaningful impact on patient functioning such as walking associated with the TUG as well as reaching up over the shoulder and translating into activities of daily living. The -- we are evaluating the exact clinically meaningful differences based on this clinical trial. But the fact that we are seeing any differences at all in this patient population in such a short period of time is meaningful for this neuromuscular disease.

And one more question, please, around the sampling method. How are you thinking that you could support your thinking there that one of the reasons that the primary endpoint was missed was the sampling method, the variability in sampling method?

Yes. Thanks, Kostas. Let me turn it over to Chris Moxham, our Chief Scientific Officer, and we can just speak in a little more detail about some of the challenges that we saw in terms of translating the reductions in DUX4-driven gene expression that we and others have observed preclinically to utilization with the primary endpoint.

Sure. Thanks, Bryan. So again, as Chris Morabito indicated, we built a very precise and highly robust qPCR assay to quantify these 6 genes that are known to be DUX4 target genes. The issue is not the asset. The issue is, in fact, as Judy pointed out, more technological or biological variability that's introduced by the fact that, we know that DUX4 expression is, in fact, stochastic. It's on and off. It's high and low. It's also fairly scarce that 1,000 myonuclei, or perhaps even 1 in 3,000 myonuclei are shown to express DUX4. And so in effect, we were using a needle biopsy to try to sample in a heterogeneous cell population and trying to track down a signal that is both stochastic and scarce. And I think those issues of biologic and technical variability then contribute to the variability that you see across the various biopsies that really can found in our ability to detect change either in placebo, but in particular to detect change in response to losmapimod treatment. The other factors that contribute, as Judy pointed out, is that, as we know, FSHD is heterogeneous, both within the patient and across patients. And lastly, performing the biopsy procedure across 17 different sites just inherently introduces additional variability that all of those submit to contribute to the variability that we have observed in this study.

Our next question comes from Dae Gon Ha with Stifel.

The update call this morning. I was wondering if we can circle back to the DUX4 aspect of the data. I think you showed the Hsp27, the phosphorylated versus the total as part of your target engagement data. But can you just speak to any other downstream effectors of losmapimod or p38 that could point to perhaps the justification that you are indeed targeting DUX4. And what did you actually see there if there were other quantifications that were done? And then second question is with regards to the muscle MRI. I know you kind of honed in on the muscle fat infiltration recognizing that the fat fraction of the lean muscle volume may come later. I guess what would you consider to be a clinically meaningful difference there in MFF and the LMV? And now that OLE is the study portion that's an open-label extension, is ongoing. Is there any way for us to know what those long-term benefits may be coming out of that MFF and the LMV measurement?

Sure. Yes. Thank you for the question, Dae Gon. And maybe we can answer that in 2 questions, and I'll turn it over to Michelle Mellion. So the first question about target engagement. And it would be great, and we can also speak to the initial Phase I study that we did to determine both muscle penetration and target engagement of losmapimod. And then we can speak in a little more detail to the whole-body MRI that we utilize what we saw with MFI and what we would then expect to see with MFF but also the correlations, the correlations that we've seen between fat infiltration and function, which is something that we shared earlier this year at the NPA Conference. Maybe I'll turn it over to Michelle, and we could start with target engagement.

Sure. So in our Phase I study, we were able to show that losmapimod does impact -- penetrate the muscle at clinically relevant concentrations, and also that we were able to show that target engagement occurred in muscle at levels at which we expect about a 35% to 65% decrease in DUX4. And looking at this data, what we had also seen was that there was a nearly 1:1 correlation between what we have seen in the muscle and what we have seen in plasma. So we can extrapolate from there knowing that what we see in plasma that we are, in fact, decreasing DUX4 by looking at the target engagement parameters and by what we know from our preclinical data set as well as this Phase I data set, for the reduction of DUX4. So Hsp27 is very robust in showing the activity of losmapimod and the target engagement in that losmapimod in fact, is working in the target organ of interest, which in this case is the muscle.

May be, you want to also speak to the MRI findings.

Sure. So in terms of the MRI findings as we've discussed previously, the muscle fat infiltration is the most proximal measure of pathology in terms of MRI. And the scale in which we measure MFI is 0 to 50%. And what we see and the difference is when we're looking at the correlation of MFI with TUG is that we see that there's a very strong correlation. So decreases in MFI essentially are a 1:1 almost decrease in TUG. So there's a very a strong correlation there with how performance is between MFI and TUG. In terms of muscle fat fraction, there the differences or changes in muscle fat fraction and lean muscle volume, as we have said previously, we do see a trend in muscle fat fraction, but don't necessarily see any differences in lean muscle volume, and we anticipate that, it's likely that it might just take longer to see those differences. You have to remember that this study was not necessarily powered to see differences on MRI. So again, seeing any differences on these measures is meaningful in this population.

Okay. Great. If I can just clarify, and apologies, but on the Hsp27 data, the 35% to 65% reduction number that you just cited, is that based on Phase I? Or is that also seen in Phase IIb, the ReDUX4 trial that we're presented today?

Yes, it was the same. It was the same way. It was consistent across both Phase I and Phase II.

Our next question comes from Judah Frommer with Crédit Suisse.

This is Mark on for Judah. I guess building on the last question, I was just curious, in your study, was there a correlation between those that had reduced muscle fat and filtration and those that saw improvements in reachable workspace, is that an analysis you've done already? And if so, did you see kind of those that had the greatest improvement and one had the greatest improvement in the other. And then just a housekeeping question. I was wondering, I think I saw for the reachable workspace there is data on, I think, give or take, 55 or 60 patients. So I was just wondering if you could reconcile that with the 80 patients in the study?

Sure. Why don't I turn it over to Michelle to speak to both the correlations as well as the reach for workspace.

So to reiterate your question, is your question, reachable workspace is that, patients that recognize improvements that these are also the patients in which we saw improvements in reachable workspace? And...

Yes, yes. And in muscle fat infiltration, like those that saw the great improvement in muscle fat infiltration saw the greatest improvement in RWS.

Yes. So in our studies for our cross-sectional analyses, we do see a strong correlation -- moderate to strong correlation between the muscle fat infiltration as well as our clinical outcome assessments, not only with reachable workspace, but also with TUG and FSHD TUG across the board. Additionally, we have looked at those patients that did report improvement and those patients that did improve in reachable workspace were also those patients that reported improvements on the PGIC as well. So there are correlations there.

So, is there...

Sorry there's one more question.

The second question was about resolving the ends, the samples for the various things. So 2 factors came into play. One is that some patients actually rolled over into the long-term extension before the 48-week endpoint. So those who completed the 16-week biopsy before the amendment, before COVID hit, were allowed to roll over into the long-term extension. There were 16 of those. And we have 3 discontinuations, all of which were unrelated to study drug. So that leaves a total of 61 and they're all accounted for in the course of the trial. We do have individual points that are missing, and that's related to an ability of the individual participants to get to a center for the visit during COVID, during the pandemic.

[Operator Instructions] Our next question comes from Tazeen Ahmad with Bank of America.

This is Stephen on for Tazeen. You mentioned earlier that you'll be speaking with regulators about the potential for accelerated approval. I just wanted to know your thoughts on if you think European regulators might have the same view as the U.S. on approvability?

Sure. Thanks for the question. Let me turn it over to Judy Dunn.

Sure. The U.S. regulators have been very clear recently that what's important to them is the benefit to patients in terms of how patients think and function. We have no reason to believe that the European regulators would be far astray of that opinion. That what is important to measure in a clinical trial is the benefit to patients. And so we will continue to meet with the FDA, but also be going into the EMA shortly.

Thank you. And I'm currently showing no further questions at this time. I'd like to turn the call back over to Bryan for any closing remarks.

Thank you. Thank you, everybody, for joining us today. We appreciate the continued support of Fulcrum. Have a great day.

This concludes today's conference call. Thank you for participating. You may now disconnect.